The Continuum

Companion to Hume

9780826443595_Prelims_Final_txt_print.indd i 2/9/2012 9:38:07 PM

9780826443595_Prelims_Final_txt_print.indd ii 2/9/2012 9:38:07 PM
 THE CONTINUUM
COMPANION TO HUME

GENERAL EDITORS

Alan Bailey
Dan O’Brien

9780826443595_Prelims_Final_txt_print.indd iii 2/9/2012 9:38:07 PM

 First published in 2012 by

Continuum International Publishing Group

The Tower Building 80 Maiden Lane

11 York Road Suite 704
London SE1 7NX New York, NY 10038

Companion to Hume
EISBN 978 1 44111 461 7

© Continuum, 2012

British Library Cataloguing-in-Publication Data

A CIP record of this title is available from the British Library

All rights reserved. No part of this publication may be

reproduced, stored in a retrieval system, or
transmitted in any way or by any means, electronic,
mechanical, photocopying, recording or otherwise,
without the written permission of the copyright holder.

Typeset in Sabon by
Newgen Publishing and Data Services
Printed and bound in Great Britain

9780826443595_Prelims_Final_txt_print.indd iv 2/9/2012 9:38:07 PM

 CONTENTS

List of Contributors vii

Abbreviations for Works Written by Hume ix
Acknowledgements xi
David Hume – A Timeline xiii

INTRODUCTION 1
1. HUME’S LIFE, INTELLECTUAL CONTEXT AND
RECEPTION 20
Emilio Mazza
2. HUME’S EMPIRICIST INNER EPISTEMOLOGY:
A REASSESSMENT OF THE COPY PRINCIPLE 38
Tom Seppalainen and Angela Coventry
3. HUME’S ‘SCEPTICISM’ ABOUT INDUCTION 57
Peter Millican
4. THE PSYCHOLOGY AND EPISTEMOLOGY OF
HUME’S ACCOUNT OF PROBABLE REASONING 104
Lorne Falkenstein
5. CAUSATION AND NECESSARY CONNECTION 131
Helen Beebee
6. HUME ON SCEPTICISM AND THE MORAL SCIENCES 146
Alan Bailey
7. THE SELF AND PERSONAL IDENTITY 167
Harold Noonan
8. ‘ALL MY HOPES VANISH’: HUME ON THE MIND 181
Galen Strawson
9. ACTION, REASON AND THE PASSIONS 199
Constantine Sandis

9780826443595_Prelims_Final_txt_print.indd v 2/9/2012 9:38:07 PM

 CONTENTS

10. FREE WILL 214

James A. Harris
11. HUME ON MIRACLES 227
Duncan Pritchard and Alasdair Richmond
12. DAVID HUME AND THE ARGUMENT TO DESIGN 245
Andrew Pyle
13. PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF 265
David O’Connor
14. HUME’S SENTIMENTALIST ACCOUNT OF
MORAL JUDGEMENT 279
Julia Driver
15. HUME AND THE VIRTUES 288
Dan O’Brien
16. HUME’S HUMAN NATURE 303
Russell Hardin
17. HUME AND FEMINISM 319
Lívia Guimarães
18. HUME ON ECONOMIC WELL-BEING 332
Margaret Schabas
19. ‘OF THE STANDARD OF TASTE’: DECISIONS,
RULES AND CRITICAL ARGUMENT 349
M. W. Rowe
20. HUME ON HISTORY 364
Timothy M. Costelloe
21. HUME’S LEGACY AND THE IDEA OF BRITISH EMPIRICISM 377
Paul Russell

Bibliography 396
Index of Names 437
Index of Topics 441

vi

9780826443595_Prelims_Final_txt_print.indd vi 2/9/2012 9:38:07 PM

 LIST OF CONTRIBUTORS

Alan Bailey Lorne Falkenstein

Visiting Lecturer in Philosophy Professor of Philosophy
School of Law, Social Sciences and Department of Philosophy
Communications University of Western Ontario
University of Wolverhampton Canada
UK
Lívia Guimarães
Helen Beebee Lecturer in Philosophy
Professor of Philosophy Department of Philosophy
Department of Philosophy Universidade Federal de Minas Gerais
University of Birmingham Brazil
UK
Russell Hardin
Timothy Costelloe Professor of Politics
Associate Professor of Philosophy Department of Politics
Department of Philosophy New York University
College of William and Mary USA
USA
James Harris
Angela Coventry Senior Lecturer in Philosophy
Associate Professor of Philosophy Department of Philosophy
Department of Philosophy St Andrews University
Portland State University UK
USA
Emilio Mazza
Julia Driver Associate Professor
Professor of Philosophy Institute of Human Sciences, Language and
Department of Philosophy Environment
Washington University in St Louis Università IULM, Milan
USA Italy

vii

9780826443595_Prelims_Final_txt_print.indd vii 2/9/2012 9:38:08 PM

 LIST OF CONTRIBUTORS

Peter Millican Department of Philosophy

Gilbert Ryle Fellow and Professor of University of Edinburgh
Philosophy UK
Hertford College
University of Oxford M.W. Rowe
UK Senior Lecturer in Philosophy
Department of Philosophy
Harold Noonan University of East Anglia
Professor of Philosophy UK
Department of Philosophy
University of Nottingham Paul Russell
UK Professor of Philosophy
Department of Philosophy
Dan O’Brien University of British Columbia
Senior Lecturer in Philosophy Canada
Department of History, Philosophy and
Religion Constantine Sandis
Oxford Brookes University Reader in Philosophy
UK Department of History, Philosophy and
Religion
David O’Connor Oxford Brookes University
Professor of Philosophy UK
Department of Philosophy
Seton Hall University Margaret Schabas
USA Professor of Philosophy
Department of Philosophy
Duncan Pritchard University of British Columbia
Professor of Philosophy Canada
Department of Philosophy
University of Edinburgh Tom Seppalainen
UK Associate Professor of Philosophy
Department of Philosophy
Andrew Pyle Portland State University
Reader in Philosophy USA
Department of Philosophy
University of Bristol Galen Strawson
UK Professor of Philosophy
Department of Philosophy
Alasdair Richmond University of Reading
Senior Lecturer in Philosophy UK

viii

9780826443595_Prelims_Final_txt_print.indd viii 2/9/2012 9:38:08 PM

 ABBREVIATIONS FOR WORKS
WRITTEN BY HUME

Abs An Abstract of a Book lately Published; Entituled, A Treatise of Human Nature

Cited by paragraph and page number from the texts included in the editions of THN listed
below.

DIS A Dissertation on the Passions

Cited by section and paragraph number from A Dissertation on the Passions; The Natural
History of Religion, The Clarendon Edition of the Works of David Hume, ed. T.L. Beauchamp
(Oxford: Clarendon Press, 2007). Hence DIS 1.2 = Section 1, paragraph 2.

DNR Dialogues concerning Natural Religion, ed. N. Kemp Smith, 2nd edn with supple-
ment (London and Edinburgh: Thomas Nelson, 1947).
Cited by part and page number (e.g. DNR 1.135 = Part 1, page 135).

E Essays, Moral, Political, and Literary, ed. E.F. Miller (Indianapolis: Liberty Classics,
1987).
Cited by page number.

EHU An Enquiry concerning Human Understanding, ed. T.L. Beauchamp (Oxford:

Oxford University Press, 1999).
Cited by section and paragraph number, and supplemented by the corresponding page num-
ber in Enquiries concerning Human Understanding and concerning the Principles of Morals,
ed. L.A. Selby-Bigge, 3rd edn with text revised and notes by P.H. Nidditch (Oxford: Clarendon
Press, 1975). Hence EHU 5.1 / 40 = Section 5, paragraph 1 in the Beauchamp edition and
page 40 in the Selby-Bigge edition.

EPM An Enquiry concerning the Principles of Morals, ed. T.L. Beauchamp (Oxford:
Oxford University Press, 1998).
Cited by section and paragraph number, and supplemented by the corresponding page num-
ber in Enquiries concerning Human Understanding and concerning the Principles of Morals,
ed. L.A. Selby-Bigge, 3rd edn with text revised and notes by P.H. Nidditch (Oxford: Clarendon

ix

9780826443595_Prelims_Final_txt_print.indd ix 2/9/2012 9:38:08 PM

 ABBREVIATIONS FOR WORKS WRITTEN BY HUME

Press, 1975). Hence EPM 2.2 / 176–7 = Section 2, paragraph 2 in the Beauchamp edition and
pages 176–7 in the Selby-Bigge edition.

H The History of England, from the Invasion of Julius Caesar to the Revolution in
1688, ed. W.B. Todd, 6 vols (Indianapolis: Liberty Classics, 1983).
Cited by volume and page number (e.g. H1.155 = Volume 1, page 155).

LDH Letters of David Hume, ed. J.Y.T. Greig, 2 vols (Oxford: Clarendon Press, 1932).
Cited by volume, page number, and letter number (e.g. LDH 1.11–12, 5 = Volume 1, pages
11–12, letter 5).

LFG A Letter from a Gentleman to his Friend in Edinburgh, ed. E.C. Mossner and
J.V. Price (Edinburgh: Edinburgh University Press, 1967).
Cited by page number.

MOL My Own Life.

Usually cited by page number from Volume 1 of LDH above.

NHR The Natural History of Religion.

Cited by page or, where applicable, part and page number from Dialogues and Natural
History of Religion, Oxford World’s Classics, ed. J.C.A. Gaskin (Oxford: Oxford University
Press, 1993). Hence NHR 134 = page 134; NHR 1.135 = Part 1, page 135.

NLH New Letters of David Hume, eds. R. Klibansky and E.C. Mossner (Oxford:
Clarendon Press, 1954).
Cited by page and letter number (e.g. NLH 5–6, 3 = pages 5–6, letter 3).

THN A Treatise of Human Nature, eds. D.F. Norton and M.J. Norton (Oxford: Oxford
University Press, 2000).
Cited by book, part, section and paragraph number, and supplemented by the corresponding
page number in A Treatise of Human Nature, ed. L.A. Selby-Bigge, 2nd edn with text revised
and notes by P.H. Nidditch (Oxford: Clarendon Press, 1978). Hence THN 1.4.7.1 / 263 =
Book 1, part 4, section 7, paragraph 1 in the Norton edition and page 263 in the Selby-Bigge
edition.

References to the Appendix of the Treatise make use of the abbreviation App. and are then
given by paragraph and page number (eg. THN App. 1 / 623 = paragraph 1 of the Appendix
in the Norton edition and page 623 in the Selby-Bigge edition).
References to the editorial material of the Oxford Philosophical Texts or Clarendon Critical
Edition versions of the Norton edition are by page number and the abbreviations THN-P and
THN-C respectively.

9780826443595_Prelims_Final_txt_print.indd x 2/9/2012 9:38:08 PM

 ACKNOWLEDGEMENTS

Special thanks to our copy editor, Merilyn Holme, for coaxing and prodding the book to
completion; a far from easy task particularly given, in the final months, the looming festive
season. We are immensely grateful to her and to Sarah Douglas at Continuum for commis-
sioning the project. Thanks are due as usual to the Edwardian Tea Room in the Birmingham
Museum and Art Gallery for refreshments and scholarly ambience, and Birmingham Central
Library provided from its stacks a range of books and journals that would have done credit to
a major university library. We would also like to take this opportunity to express our gratitude
to the following colleagues at Oxford Brookes University, the University of Wolverhampton
and Keele University: Stephen Boulter, Mark Cain, Beverley Clack, Geraldine Coggins, Meena
Dhanda, Giuseppina D’Oro, Cécile Hatier, William Pawlett and Constantine Sandis.
A. B. and D. O’B.
Much of my work on this book has been timetabled around the ongoing DIY house project
that Lucy and I are undertaking. This is something that I am sure Hume would appreciate.
If not hands-on, he was certainly no slouch when it came to the upkeep of the home. When
the ‘[p]laister broke down in the kitchen’ in his house in James’s Court, Edinburgh, he tells us
that:

[his repairman] having thus got into the house, went about teizing Lady Wallace [Hume’s
tenant], and telling her, that this and the other thing was wrong, and ought to be mended.
She told him, as she informs me, that everything was perfectly right, and she wou’d trou-
ble the Landlord for nothing. Yet the Fellow had the Impudence to come to me, and tell
me that he was sent by Lady Wallace to desire that some Stone Pavement under the Coal
Bunker shoud be repair’d. I, having a perfect Confidence in Lady Wallace’s Discretion,
directed him to repair it, as she desir’d. Having got this Authority, which cannot be good
as it was obtain’d by a Lye, he not only pav’d the Bunker anew, but rais’d a great deal of
the other Pavement of the Kitchen and laid it anew, nay, from his own head, took on him
to white-wash the Kitchen: For all which, he brought me in an account of 30 Shillings.
(NLH 115, 206)

The contributors to this book may have explored Hume’s contributions to metaphysics,
morality, religion and epistemology, but I have sympathetic appreciation of his knowledge of

xi

9780826443595_Prelims_Final_txt_print.indd xi 2/9/2012 9:38:08 PM

 ACKNOWLEDGEMENTS

that great human pursuit of house-building. And my deepest gratitude goes to Lucy my fellow
plasterer, drywaller, spark and plumber and to Dylan who is still ‘patiently’ waiting for the
kitchen to be finished. The skirting boards will be attached next week . . .
D. O’B.
Particular thanks on my behalf go to Ross Singleton for his longstanding friendship and our
many lengthy conversations about religion, philosophy and international politics. Linda Dai
has patiently coped with my tendency to introduce comments about Hume into a quite exces-
sive number of contexts, and her backing and encouragement have played a crucial role in
allowing me to complete my contribution to this volume. I also wish to thank my mother
Dorothy Bailey for her support during the writing and editing process.
A.B.

xii

9780826443595_Prelims_Final_txt_print.indd xii 2/9/2012 9:38:08 PM

 DAVID HUME – A TIMELINE

1702 Death of William III and the accession of

Queen Anne
1704 Battle of Blenheim; Isaac Newton, Opticks
1707 Union of England and Scotland
1711 Hume born in Edinburgh
1712 Thomas Newcomen’s steam engine
1713 George Berkeley, Three Dialogues between Hume’s father dies
Hylas and Philonous; Anthony Collins,
Discourse of Free-Thinking
1714 Death of Queen Anne and the accession of
George I
Bernard Mandeville, Fable of the Bees
1715 Jacobite rebellion in Scotland; death of
Louis XIV; Louis XV becomes King of
France at the age of five
1718 Inoculation for smallpox introduced in
England
1720 Collapse of the South Sea Bubble; Edmond
Halley becomes Astronomer Royal at
Greenwich
1722 Robert Walpole becomes the equivalent of a Hume and his brother John enrol
modern British Prime Minister as students at the University of
Edinburgh
1725 Work starts on Grosvenor Square, London;
Francis Hutcheson, The Original of Our
Ideas of Beauty and Virtue
1726 Jonathan Swift, Gulliver’s Travels Hume begins a legal education

xiii

9780826443595_Prelims_Final_txt_print.indd xiii 2/9/2012 9:38:08 PM

 DAVID HUME – A TIMELINE
1727 Death of George I and the accession of
George II; Isaac Newton dies
Robert Greene, Principles of the Philosophy
of the Expansive and Contractive Forces
1728 John Gay, Beggar’s Opera
1729
1733 Alexander Pope, Essay on Man
1734 Voltaire, Lettres philosophiques (Lettres
anglaises)
1735 John Harrison’s chronometer; William
Hogarth, Rake’s Progress
1736 Joseph Butler, Analogy of Religion
1737
1739
1740 Start of the War of Austrian Succession;
Frederick II (Frederick the Great) becomes
King of Prussia
George Turnbull, Principles of Moral
Philosophy
1741 Samuel Richardson, Pamela
1742 Walpole falls from power; Handel’s
Messiah premieres in Dublin
1745 Second Jacobite rebellion; von Kleist
discovers the ability of the Leyden jar to
store electrical charge
1746 Jacobite army is decisively defeated at
Culloden
xiv
9780826443595_Prelims_Final_txt_print.indd xiv
Hume abandons the idea of
becoming a lawyer; his ‘new Scene
of Thought’
Hume travels to London; takes
up employment in Bristol as a
merchant’s clerk; resigns or is
dismissed; relocates to France and
works on the Treatise
Hume moves from Rheims to
La Flèche
Returns from France to London
Publication of Books One and Two
of A Treatise of Human Nature
Book Three of the Treatise
Volume I of Essays, Moral and
Political published in Edinburgh
Hume fails to secure the Chair of
Ethics and Pneumatical Philosophy at
Edinburgh; becomes tutor in England
to the Marquess of Annandale
Hume’s mother dies
Hume’s appointment as a tutor
comes to an acrimonious end
Hume becomes Secretary to
Lieutenant-General James St Clair;
accompanies St Clair on a military
expedition attacking the coast of
Brittany
2/9/2012 9:38:08 PM

1748 War of Austrian Succession concludes;
Treaty of Aix-la-Chapelle
Excavations begin at Pompeii;
Montesquieu, L’Esprit des lois; La Mettrie,
L’Homme machine
1749 Buffon, first volumes of Histoire naturelle;
David Hartley, Observations on Man;
Henry Fielding, Tom Jones
1751 Diderot and d’Alembert, Volume I of
L’Encyclopédie
Adam Smith becomes Professor of Logic at
the University of Glasgow
1752 Adoption in Britain of the Gregorian
calendar
1753 British Museum founded
1754
1755 Lisbon earthquake; Samuel Johnson,
Dictionary of the English Language
1756 Seven Years’ War
1757 Robert Clive and the East India Company
are victorious at the Battle of Plassey
Edmund Burke, A Philosophical Enquiry
1758
1759 Laurence Sterne, Tristram Shandy;
William Wilberforce is born
1761 The Bridgewater Canal opens from Worsley Madame de Boufflers’s initial letter
to Manchester to Hume
1762 Catherine II becomes Empress of Russia;
Sarah Scott’s novel of a female utopian
community, A Description of Millenium Hall
1763 Seven Years’ War concludes; Peace of Paris
Catherine Macaulay, first volume of her
History of England
xv
9780826443595_Prelims_Final_txt_print.indd xv
DAVID HUME – A TIMELINE
Travels with St Clair on a
diplomatic mission to Vienna and
Turin
Publication of Philosophical Essays
concerning Human Understanding
Returns to Scotland and resides
with his brother and sister at the
family home in Ninewells
Hume moves to Edinburgh and is
later joined by his sister Katherine
Publication of An Enquiry
concerning the Principles of Morals
Hume is unsuccessful in his
candidacy for a chair at the
University of Glasgow
Elected Librarian to the Faculty of
Advocates in Edinburgh
Publication of Volume I of The
History of Great Britain
Hume resigns from his post as
Librarian
Philosophical Essays published
under the new name of An Enquiry
concerning Human Understanding
Hume accompanies Lord Hertford
to Paris and takes on the duties of
Secretary to the British Embassy
Hume and Madame de Boufflers
meet for the first time
2/9/2012 9:38:09 PM
 DAVID HUME – A TIMELINE

1764 Thomas Reid, Inquiry into the Human

Mind; Horace Walpole, The Castle of
Otranto
Mozart’s Symphony No. 1 composed in
London during the family’s European tour
1765 Matthew Boulton finishes building the Soho Hume is officially confirmed as
Manufactory in Birmingham Secretary shortly before his post
comes to an end
1766 Immanuel Kant, Dreams of a Spirit-Seer Hume returns to London,
Elucidated by Dreams of Metaphysics accompanied by Jean-Jacques
Rousseau; Rousseau accuses Hume
of being part of a conspiracy
against him
A Concise and Genuine Account of
the Dispute between Mr. Hume and
Mr. Rousseau
Hume spends the final months
of the year in Ninewells and
Edinburgh
1767 James Craig’s plan for New Town, Travels to London to take up the
Edinburgh is adopted; Royal Crescent, office of Under-Secretary of State in
Bath, started the Northern Department
1768 James Cook’s first voyage to the Pacific Hume retires from public office
1769 Josiah Wedgwood and Thomas Bentley open Hume returns to Edinburgh
their Etruria factory near Stoke-on-Trent
James Watt’s steam engine, James
Hargreaves’ spinning jenny, and Richard
Arkwright’s water frame are patented
Arthur Wellesley, later Duke of Wellington,
is born in Dublin
1770 The future Louis XVI marries Marie Hume has a house built for himself
Antoinette; Baron d’Holbach, Système de la in Edinburgh’s New Town; rumours
nature reach Paris that Hume might be
about to marry Nancy Orde
1771 Tobias Smollet, Humphry Clinker Hume and his sister move into the
new house – St Andrew Square, off
St David Street
1773 Boston Tea Party
Hester Chapone, Letters on the
Improvement of the Mind; Oliver
Goldsmith, She Stoops to Conquer

xvi

9780826443595_Prelims_Final_txt_print.indd xvi 2/9/2012 9:38:09 PM

 DAVID HUME – A TIMELINE

1774 Death of Louis XV; Louis XVI becomes

King of France
Joseph Priestley discovers ‘dephlogisticated
air’ (oxygen)
1775 John Wilkinson’s cannon-boring machine; The ‘Advertisement’ repudiating the
Jane Austen born Treatise
1776 Declaration of American Independence Hume dies in Edinburgh
Edward Gibbon, Decline and Fall of the
Roman Empire; Adam Smith, The Wealth
of Nations; Thomas Paine, Common Sense;
Jeremy Bentham, Fragment on Government
1777 Publication of Life of David Hume,
written by Himself
1779 World’s first iron bridge is completed across Publication of the Dialogues
the River Severn at Coalbrookdale concerning Natural Religion
1782 Atheism openly avowed in print in Britain
for the first time – Matthew Turner, Answer
to Dr Priestley’s Letters to a Philosophical
Unbeliever
1783 Peace of Versailles establishes independence
of American colonies
1785 Edmund Cartwright patents his power loom
1787 Founding in Britain of the Committee for
the Abolition of the African Slave Trade
1789 French Revolution

xvii

9780826443595_Prelims_Final_txt_print.indd xvii 2/9/2012 9:38:09 PM

9780826443595_Prelims_Final_txt_print.indd xviii 2/9/2012 9:38:09 PM
 INTRODUCTION
In 2005 the British Broadcasting Corporation
ran a poll asking Radio Four listeners to say
whom they regarded as the greatest philoso-
pher of all time. Such familiar philosophical
luminaries as Plato, Aristotle, Descartes and
Kant all featured prominently in the subse-
quent voting, and Marx’s immense influence
within the political arena saw him, rather
predictably, taking first place. However, the
pre-eminent British philosopher, and the
philosopher with the second highest overall
number of votes, was David Hume.
In his own lifetime Hume certainly pos-
sessed a substantial reputation as a public
intellectual. In some respects, though, it would
be more appropriate to talk in terms of his
notoriety rather than his reputation. His sup-
posedly sceptical epistemological views and
the manner in which his writings seemed to
develop a series of pointed criticisms of reli-
gious belief attracted vituperative criticism
from many of his contemporaries. It is also
a striking fact that much of his fame sprang
from his ostensibly non-philosophical writ-
ings. Until his death in 1776 Hume enjoyed a
great deal of influence as a writer on matters
of economics. Moreover, sales of his History
of England made him independently wealthy
and brought him to the attention of far more
readers than were interested in works of
9780826443595_Intro_Final_txt_print.indd 1
metaphysics and epistemology. There is, in
fact, an overwhelming case for saying that
no other eighteenth-century writer’s account
of English history came close to matching the
intellectual quality and non-partisan nature
of Hume’s own narrative, and in this particu-
lar case those genuine merits were, for once,
rewarded by the approbation of substantial
sections of the public.
Hume’s current reputation is, therefore,
something that stands in need of explana-
tion. How has an eighteenth-century Scottish
intellectual and writer who enjoyed his great-
est success amongst his contemporaries as a
historian, economist and writer of polite
essays arrived at the status and, in the eyes
of the editors of this volume, the wholly
deserved status of being viewed as the great-
est British philosopher?
In many respects the answer lies in the
very features of his philosophical writings
that saw them subjected to so much criticism
when Hume was alive. Epistemological scep-
ticism, even of a radical variety, is no longer
seen as constituting any kind of threat to
morality and social order; so it is now pos-
sible to respond to the sceptical arguments
deployed within Hume’s writings as provid-
ing us with a series of fascinating puzzles that
may succeed in pointing us towards important
2/9/2012 9:37:57 PM
 INTRODUCTION
truths about the nature of philosophy or the
incoherence of certain aspects of our self-
conception as inquirers and agents without
those arguments constituting instruments of
intellectual self-annihilation. Moreover, once
this fear of sceptical conclusions has been
dissipated, it becomes psychologically easier
to acknowledge the inadequacy of so many
of the standard supposed refutations of scep-
tical arguments. Hume’s own recognition of
the power of these arguments accordingly
comes to be seen as compelling evidence of
his own intellectual integrity and powers of
analysis.
This issue of intellectual integrity also has
a bearing on present-day reactions to Hume’s
criticisms of religion. Britain in the eighteenth
century was an overwhelmingly Christian
country, where overt expressions of disbelief
could still attract substantial prison sentences
and books regarded as attacking Christianity
were frequently subjected to determined
campaigns of suppression. Today, in contrast,
there is substantial evidence that between 30
and 40 per cent of the British population do
not believe in God or any Higher Power anal-
ogous to a person. And although the United
States signally lags behind almost all Western
European states in this regard, agnosticism
and atheism are making some inroads even
in that hitherto hostile environment. There
is, accordingly, a far more receptive audi-
ence in the current climate for arguments
challenging the metaphysical underpinnings
of a religious world-view and the compla-
cent supposition that religious convictions
constitute crucial support for moral behav-
iour and an appreciation of the value of life.
Hume’s writings provide such arguments in
abundance, and his critiques of the argument
to design and the credibility of testimony to
alleged miracles still constitute some of the
most trenchant attacks ever launched on
9780826443595_Intro_Final_txt_print.indd 2
these staples of theistic apologetics. Just as
significantly, however, Hume’s evident will-
ingness to question religious dogma at a
time when the social and cultural pressure
towards internalizing such beliefs was so
strong marks him out as a person who was
prepared to be guided by argument and the
available evidence instead of suppressing
his critical faculties at the behest of super-
stition and the power-structures of religious
authority. In this respect, Hume amply meets
the essential requirement that a true philoso-
pher, a philosopher of genuine integrity, must
answer only to the autonomous demands of
the reflective intellect.
It also seems to be the case that once
Hume’s epistemological and irreligious views
are no longer predominantly seen as views
that need to be repudiated as aggressively
as possible, other valuable aspects of his
philosophical outlook become increasingly
easy to recognize. Given the disappointing
results of attempts at a priori metaphysics,
Hume’s denunciations of the application
of a priori reasoning outside the sphere of
issues of ‘quantity and number’ (EHU 12.27
/ 163) seem amply vindicated by the histori-
cal record. Thus philosophical inquiry needs
an alternative methodology if it is not simply
to repeat past errors in ever more complex
forms. And Hume’s ‘experimental’ method,
with its commitment to being guided by
experience, seems to meet this need.
There might perhaps be some worries that
this approach actually amounts to a simple
repudiation of philosophy in favour of the
investigative methods of the sciences. In
Hume’s hands, however, it constitutes not an
abandonment of philosophy but a confirma-
tion that at least some philosophical conun-
drums can be satisfactorily dissolved by
paying due attention to the empirical facts.
Confronted, for example, by the question
2/9/2012 9:37:57 PM

‘What obligation do we have to obey this or
indeed any government?’, we might initially
be at a loss to know how to proceed. Science,
we have been repeatedly told, cannot answer
normative questions. Conceptual analysis,
it is tempting to suppose, can at best clarify
the sense of the question, and a priori rea-
soning of a non-mathematical kind cannot
be relied upon to yield anything more sub-
stantial than vacuous tautologies. Hume’s
account of human nature, in contrast, allows
us to see this question as an idle one. There
may indeed be scope for choosing which
government to follow. But our psychologi-
cal properties mean that some institutions
of government will inevitably arise in all cir-
cumstances that are ever likely to persist for
a significant length of time. Moreover, once
these institutions have arisen, their success in
securing high levels of obedience is equally
inevitable irrespective of our normative
speculations.
It would be remiss, however, of any
account of Hume’s well-merited appeal to
present-day philosophers and anyone inter-
ested in understanding the place of human
beings in the world to ignore the question
of Hume’s personal character. Although this
has frequently been traduced by defend-
ers of religion and people who mistakenly
suppose that seriousness of purpose must
be evidenced by tortuous writing, pompous
pretentiousness and a complete absence of
humour, it is clear from the record of Hume’s
life that he was a benevolent man of ami-
able temperament, a good and loyal friend,
and a master of comic self-deprecation and
subtle word-play. If one were planning a fan-
tasy dinner party, it is difficult to imagine
any philosopher in history who would make
a more winning and entertaining guest or a
more congenial host, though it would prob-
ably be advisable, given his reported skill at
9780826443595_Intro_Final_txt_print.indd 3
INTRODUCTION
whist, to avoid challenging him to any card
games involving large sums of money. In a
sense, of course, Hume’s personal virtues do
not add to the importance of his intellec-
tual achievements. But they do confirm one
important thing, namely that the philosophi-
cal outlook embraced by Hume is one that is
entirely compatible with a flourishing human
life that combines generous concern for the
well-being of other people with ample enjoy-
ment of a full range of social and intellectual
satisfactions.
This combination of the power of Hume’s
thought and the engaging quality of his per-
sonality has undoubtedly helped in bringing
together the contributors to this volume. As
editors, we were repeatedly pleasantly sur-
prised by the enthusiasm expressed for this
project by potential contributors, and we
hope that the finished anthology succeeds
both in illuminating Hume’s own achieve-
ments and in suitably showcasing the com-
mitment to Humean scholarship manifested
by all the authors represented in the follow-
ing pages.
Emilio Mazza opens the volume by draw-
ing us into Hume’s world, one far from the
ivory tower – a world of business, military
expeditions, international diplomacy and
Parisian ladies. But Le Bon David always
sought refuge from this heady world in
work, in friendships and in his pursuit of
literary fame, his ‘ruling passion’. His aca-
demic legacy and fame, however, are perhaps
rather fortunate given that he would have
been happy to stay at home in the borders
of Scotland, if his brother had not married,
or to join the army if he had discovered its
pleasures and camaraderie at a younger age.
Mazza’s evocative biography illuminates
a life of travel and friendships with a por-
trait of a cheery, avuncular man toddling
around Edinburgh, being dragged out of a
2/9/2012 9:37:57 PM
 INTRODUCTION
bog, saving a man from execution after that
man’s unsuccessful attempt to commit sui-
cide in Paris, and having deep and sometimes
stormy relationships with the literati of his
day including, amongst many others, Adam
Smith, Rousseau, d’Alembert and Lawrence
Sterne. We come away with the impression
that Hume had a good life – one with much
friendship, fame and fortune – and if one can
ever say this, Hume also had a good death.
To the end he was in good spirits, reading his
beloved classics, and revising his Dialogues
concerning Natural Religion.
Tom Seppalainen and Angela Coventry
take a ‘fresh look’ at Hume’s theory of ideas
and impressions. The notion of liveliness or
vivacity that distinguishes mere ideas from
beliefs – beliefs being vivid ideas – is usu-
ally taken to be a phenomenological one
and various interpretations of the nature of
vivacity are considered. It has been charac-
terized in terms of qualitative feel although
Seppalainen and Coventry argue that think-
ing of perceptions in this way ignores their
intentional content and the way perceptions
seem to be of the world. An improvement,
then, is to read the phenomenology of per-
ception not in terms akin to those describ-
ing the intensity of colour in a picture, but
in terms of ‘presentedness’ (or, according
to another intentional reading of Hume, in
terms of verisimilitude and the feeling of
reality). Experience presents the world to
one. Seppalainen and Coventry applaud
such intentional readings, but they argue
that Hume does not use vivacity to refer to
the phenomenological qualities of individual
perceptions, but rather to sequences of ideas
and impressions. Only patterns of change
can be vivid in the requisite sense. Our very
notion of the existence of the external world
depends on the constant and coherent flux
of our perceptions and, they argue, such flux
9780826443595_Intro_Final_txt_print.indd 4
is the source of the vivacity of our experi-
ence, of its intentional content, and of the
believability of our ideas. This account is
contrasted with Descartes’s theory of ideas
and with interpreters of Hume who see him
as a proto-logical positivist.
Peter Millican turns to Hume’s account
of inductive reasoning and his ‘famous argu-
ment’ to the conclusion that we have no
warrant for our beliefs about unobserved
matters of fact. Millican spells out the steps
of Hume’s argument as articulated in the
Treatise, the Abstract and the first Enquiry.
All beliefs concerning matters of fact are
grounded in causal reasoning but, Hume
argues, knowledge of causal relations can-
not be acquired a priori, nor can it be gained
via inductive reasoning. In place of such sup-
port Hume provides an account of belief
grounded in custom or habit. However con-
vincing Hume’s arguments may be, there is
undoubted tension between his seemingly
sceptical conclusions and his embrace of
inductive science, his ‘experimental’ approach
to the study of human nature, and his empiri-
cal approach to history. Some interpreters of
Hume take him just to be concerned with a
psychological description of thinkers and not
with issues concerning justification and war-
rant. Millican, however, argues that Hume
is interested in normative questions; it is
important to be clear, though, on the target
of Hume’s scepticism – and that is Locke’s
conception of reason, what Millican calls his
perceptual model. Such scepticism, however,
does not engender what has come to be called
‘The problem of induction’. The purpose of
Hume’s form of ‘mitigated’ scepticism is not,
as with Descartes, to prompt us to discover
a sure path to certain knowledge, but rather
to instil in us a suitable level of modesty and
caution concerning our epistemic practices.
Furthermore, Hume’s naturalistic account
2/9/2012 9:37:57 PM

provides an explanation of how human
beings can – and actually do – reason induc-
tively and this, given the impossibility of any
rational foundations for such reasoning, pro-
vides us with all the support we require for
our causal inferences.
Lorne Falkenstein further explores Hume’s
account of causal reasoning and the ‘system’
of the Treatise. It is constant conjunctions
in experience that impel the imagination to
form beliefs concerning causes and effects.
Habits of thought guide our reasoning and
not rational argument or judgement. As dis-
cussed by Coventry and Seppalainen, beliefs
are seen as vivacious ideas, and Falkenstein
stresses that Hume does not think of vivacity
in terms of the intensity of an image. Beliefs,
rather, amount to dispositions of the mind
(for example, to incline one to act in certain
ways and to focus one’s attention). Beliefs,
then, are the product of the principle of asso-
ciation of causation, although Falkenstein
also suggests how Hume might have included
the relations of contiguity and resemblance
in his account of reasoning.
If, however, belief formation is just a mat-
ter of habit, how can it be so that some ways
of forming beliefs are seen as better than oth-
ers? Hume suggests that we follow certain
general rules, those that are learnt from past
experience, such as, like objects in like cir-
cumstances will have like effects. Particular
inferences we make can be assessed against
such general rules. This is Hume’s logic of
causal inference.
Such rules can also be extended to cover
probable reasoning. It is not the case that we
have uniform experience, but this does not
lead to us rejecting such general rules. Instead,
in cases where like causes do not seem to lead
to like effects we come to believe – again via
habit grounded in past experience – that
there is some hidden cause that is disrupting
9780826443595_Intro_Final_txt_print.indd 5
INTRODUCTION
the usually apparent regularity of nature.
And Hume’s account of the vivacity of belief
can show how the strength of our beliefs
depends on the uniformity of our experience.
This account does not depend on a mathem-
atical calculation of probabilities, but rather
on associationist psychological processes of
vivacity transfer. Falkenstein concludes by
showing how Hume takes his scepticism as
supporting his ‘system’ and his account of
empirical reasoning.
Helen Beebee heads into stormy water –
into Hume’s account of causation, a hotly
debated topic: the so-called New Hume
Debate focusing on the question of whether
Hume takes there to be real causal powers
in nature – oomphs pushing billiard balls
around tables – or whether he thinks that
there are really only constant conjunctions
and regular brute patterns in the world.
Beebee investigates this question, concentrat-
ing on three claims to which Hume seems to
be committed. First, Hume suggests that we
project causal connections onto the world,
connections that are not really there. Second,
the concept of causation includes the notion
of necessity. We think of causal connections as
those that are necessarily connected together:
the red ball must move off in that direction
given that it was struck in that way by the
white ball. And third, causal talk is objective
in that we can talk correctly of causes and
we can sometimes make false causal claims
about the world. These commitments seem
to be inconsistent since it is not obvious how
causal talk can be objective when Hume does
not think that there are really causal connec-
tions in nature and that they are projections
of our cognitive processes. Beebee discusses
various ways to resolve this (perhaps only
apparent) inconsistency. The traditional
interpretation of Hume claims that causation
is just constant conjunction. The sceptical
2/9/2012 9:37:57 PM
 INTRODUCTION
realist, in contrast, sees Hume as accepting
that there are causal powers in nature; it
is just that we cannot come to have know-
ledge of them. The projectivist interpretation
adopts a non-cognitivist stance: our claims
concerning causal relations are subject to
norms, but these norms are constituted not
by features of the world independent of our
judgements concerning its causal structure –
by real causal powers in nature – but by
certain ‘rules’ which we have come to appre-
ciate concerning how we judge of causes and
effects, rules that enable us to override errant
judgements in particular cases.
A clue to the correct interpretation can be
found in Beebee’s claim that Hume is driven
by his opposition to the Cartesian Image of
God Hypothesis. There are two aspects to this
hypothesis: we are, as the Bible says, made
in God’s image, and we have epistemic abil-
ities in line with such an origin. The nature
of reality is accessible to human reason – we
can come, through a priori reasoning, to
have knowledge of the world and specifically
of its causal structure. Beebee notes certain
aspects of this picture in the sceptical realist
approach and thus argues that this cannot be
the correct interpretation of Hume. Of the
remaining options, Beebee favours projectiv-
ism over the traditional interpretation.
Alan Bailey then undertakes an examina-
tion of the equally vexed issue of the pros-
pects for providing a unified account of
Hume’s philosophical outlook that satisfac-
torily accommodates both his ambitions to
construct a science of human nature and the
sceptical elements of his thought. If Hume’s
putative scepticism actually amounted to
nothing more than a modest epistemologi-
cal fallibilism, as some recent commentators
have supposed, there would be no real ten-
sion to overcome here. Bailey argues, how-
ever, that Hume’s scepticism is a much more
9780826443595_Intro_Final_txt_print.indd 6
radical position than mere fallibilism. In
Bailey’s judgement, this scepticism is better
interpreted as a stance that does not endorse
any beliefs as possessing a positive degree
of epistemic justification except for beliefs
about very simple necessary truths that can
be grasped without going through any proc-
ess of inference and beliefs about the content
of our present ideas and impressions.
It is clear, however, that if such scepticism
is an integral component of Hume’s thought,
then it co-exists with Hume’s assent to a
detailed and carefully constructed account
of human nature that is supposed to be both
true and useful. Even a moment’s reflection
on the Treatise’s subtitle, which is Being
an Attempt to Introduce the Experimental
Method of Reasoning into Moral Subjects,
indicates that it would be a disastrous mis-
reading of Hume’s views to construe him as
simply a destructive sceptic.
Bailey’s solution to this interpretative con-
undrum is to claim that Hume views radic-
ally epistemological scepticism and properly
conducted empirical inquiries as mutually
supportive. Hume thinks that sceptical
arguments are indeed successful in placing
us in a position where only our acceptance
of the view that scarcely any of our beliefs
are rationally justified can allow us to deny,
without being guilty of bad faith, that such
sceptical arguments provide rationally com-
pelling grounds for that assessment of our
beliefs. However, Bailey argues that Hume
does not see this as posing any threat to the
ability of our belief-forming mechanisms
to generate and sustain in existence all the
beliefs we need to guide our actions. Nor
indeed does Hume view it as undermining
our capacity to assent to relatively sophisti-
cated scientific theories. Where such theor-
ies are constructed using systematized and
reflective versions of common sense methods
2/9/2012 9:37:58 PM

of inquiry and are accordingly supported by
experience and experiment, sceptical discov-
eries are incapable of preventing us from
giving our assent. And where those theories
are not supported by experience and experi-
ment, Hume can, as an empiricist, rejoice
in their destruction. Thus Bailey holds that
Hume sees scepticism and the proper experi-
mental method of inquiry working in tan-
dem. Sceptical arguments curb the power of
the imagination to generate beliefs that are
not the products of the observed correla-
tions that give rise to causal inferences. And
the experience-based beliefs towards which
we accordingly gravitate generate a plaus-
ible picture of the workings of the human
mind that makes it even more difficult for
us to represent ourselves as capable of arriv-
ing at many beliefs that genuinely qualify as
rationally justified.
Harold Noonan and Galen Strawson both
explore what Hume calls the labyrinth of
personal identity. Noonan considers various
arguments in Hume against the Cartesian
conception of personal identity, against, that
is, the existence of an enduring self, identi-
cal from moment to moment and from day
to day. Hume’s empiricism demands that we
have an impression of such an entity, but this
we do not have – all we find, on introspec-
tion, is a bundle of variously related percep-
tions. Hume’s ‘master-argument’ establishes
that all perceptions are logically capable of
an independent existence. There is thus no
need for the ‘unintelligible chimera of sub-
stance’ (THN 1.4.3.7 / 222) in which prop-
erties must inhere. This is so for physical
things, such as wax – contra Descartes, wax
for Hume is just a collection of properties –
and for human beings: we do not require an
enduring soul underlying our ever-changing
properties. Hume thus criticizes a suppos-
ition of both materialists and immaterialists,
9780826443595_Intro_Final_txt_print.indd 7
INTRODUCTION
that is, that the soul or mind is a substance,
be it physical or non-physical.
All we can do is provide an account of
what causes us to have the mistaken belief
that there are enduring selves. Such an
account includes certain identity-ascribing
mechanisms of the imagination – those
grounded in the principles of association of
resemblance and causation – mechanisms
that generate belief in the self as well as in
the continued existence of the external world
and of bodies.
Hume, however, is dissatisfied with his
conclusion. He thinks that he is committed
to two inconsistent principles: that distinct
perceptions are distinct existences, and that
the mind never perceives any connection
between them. Perplexingly, however, these
principles are not inconsistent, and uncover-
ing why Hume claims them to be so is a key
difficulty for interpreters of Hume’s views
on the soul and the self. Noonan suggests
that Hume realizes that his account does
not explain our continuing belief in personal
identity. One can accept Hume’s empiricist
conclusions with respect to the external
world and give up the notion that there is a
substance or substrata underlying the prop-
erties of bodies, but we cannot accept this
with respect to the self. Why not? Hume did
not know.
Galen Strawson has a distinct account of
why Hume’s ‘hopes vanish’. Hume discov-
ers – late in the day, in the Appendix to the
Treatise – that his whole empiricist philoso-
phy depends on a conception of the mind
that his empiricism does not allow him to
have. His genetic account of our belief in
an enduring self relies on the principles of
association – it relies on the assumption that
we have a ‘Principle-Governed Mind’. This
explains our belief in the self as well as our
belief in the external world and in causation.
2/9/2012 9:37:58 PM
 INTRODUCTION
But such a conception of the mind goes
beyond a loose association of distinct percep-
tions. In order to legitimately ground one’s
philosophy in such an account of the mind it
is required either that there is an observable
real connection between the perceptions that
make up the mind or grounds for claiming
that such perceptions inhere in some kind of
soul-substance. But Hume has argued against
both possibilities.
A possible response here is to take Hume
as sheltering in his scepticism: the essence
of the mind is unknowable to us and thus it
cannot be this – the lack of knowledge of the
Principle-Governed Mind – that leads him to
despair. But, Strawson argues, such agnosti-
cism cannot do the trick. Hume does need
to, and does, assume a certain notion of the
mind – a rule-governed one. He can perhaps
remain agnostic about just how it works, but
he cannot be agnostic about its very exist-
ence – and its very existence is what is incom-
patible with Hume’s empiricism. Strawson
claims that Hume’s despair is a result of his
acknowledgment of this deep inconsistency
in his philosophy.
Constantine Sandis explores Hume’s
account of action and in so doing considers
how reason, the will and the passions are
related. Hume’s account of action is an empir-
ical one: we acquire knowledge of a person’s
reasons for acting from careful observation
of human behaviour. This ‘science of man’
grounds Hume’s History of England and
the study of this work highlights how Hume
sees character as playing a key motivating
role in our behaviour. Further, the historian
is best placed to uncover the truth about
human nature since he does not aspire to the
detached perspective of the philosopher, nor
is he too close to his subjects and thus prone
to bias or the distorting influence of his par-
ticular interests and circumstances.
9780826443595_Intro_Final_txt_print.indd 8
Sandis also turns to interpretations of
Hume’s famous claim that ‘[r]eason is, and
ought only to be the slave of the passions’
(THN 2.2.3.4 / 414) and argues that the
received Humean theory of motivation is
unfounded. This is the view that an agent
cannot be motivated by belief alone, but
only by a belief along with an appropri-
ately related desire. Sandis claims, though,
that such an account is not to be found in
Hume. It is also suggested that Hume does
not equate belief and opinions with judge-
ment. Ideas and beliefs are distinguished by
their vivacity, and the vivacity of judgements
should be seen as lying somewhere between
that of ideas and beliefs. This is relevant to
Hume’s account of morality: Hume does not
talk of moral judgements but, on Sandis’s
account, this still allows Hume to have an
account of moral beliefs and of their motiv-
ational force.
James Harris turns to liberty and neces-
sity, and to what Hume calls ‘the most con-
tentious question of metaphysics, the most
contentious science’ (EHU 8.23 / 95). Hume
is often thought to be an advocate of an early
version of what is now called compatibilism,
and it has been claimed that there is noth-
ing distinctive about his position. Locke and
Hobbes had also discussed this question and
suggested compatibilist answers. Harris,
however, argues that Hume is not just rehash-
ing their arguments.
Importantly, it is claimed, Hume is not a
determinist in the modern sense, unlike, for
example, Hobbes. Determinism is a meta-
physical stance and Hume eschews metaphys-
ical questions. His claim is not that we have
reason to think that the laws of nature cannot
change – that they are determined; everyday
experience, rather, leads us to expect that
people behave in regular ways and we inter-
act with them in light of these regularities.
2/9/2012 9:37:58 PM

Harris also notes various changes in
emphasis between the Treatise discussion of
this topic and that of the Enquiry. In the latter,
Hume more squarely targets metaphysicians.
Once we clarify the nature of liberty and
necessity, long-running metaphysical and, in
particular, religious disputes concerning God’s
prescience and responsibility for evil will be
undermined. Philosophy should ‘return, with
suitable modesty, to her true and proper prov-
ince, the examination of common life’ (EHU
8.36 / 103). Hume’s discussion of liberty and
necessity is not a case of Hume engaging in
metaphysical debate, the question then arising
of whether his contribution is original or not –
he is, rather, agnostic about all such issues, his
discussion reflecting his empiricist attitude to
questions concerning the regularity of human
behaviour and morality.
At this point the contributors turn to the
subject of Hume’s views on the truth and
utility of religious beliefs. In his own era he
was interpreted as attacking Christianity and
all forms of theistic belief. However, his argu-
ments were frequently dismissed as incon-
sequential sophistries motivated not by a
concern for the truth but by a desire to secure
personal notoriety and increase the sales of
his books. Such an assessment of the force of
his arguments and his motivation for advan-
cing them is now wholly discredited. Yet the
recognition that he wrote on this particular
topic in good faith and with a commitment
to seeking the truth and promoting human
well-being has led some present-day com-
mentators to suggest that he was actually a
defender of some philosophically purified
form of theism that might even be compat-
ible with the truth of Christianity. Such an
interpretation seems to be based on nothing
more substantial than his invariable cour-
tesy when debating matters of religion and
an almost inexcusable failure to recognize
9780826443595_Intro_Final_txt_print.indd 9
INTRODUCTION
the extent to which he and other irreligious
thinkers of his time were forced to engage
in misdirection and linguistic contortions in
order to avoid social ostracism and the offi-
cial suppression of their writings. The authors
of the three chapters in this anthology that
focus primarily on Hume and religion are
therefore unanimous in presenting him as
a rigorous and intellectually honest thinker
who deploys a formidable set of arguments
against any form of religious outlook based
on the truth of theism or even a robust form
of deism.
Duncan Pritchard and Alasdair Richmond
investigate Hume’s notorious arguments on
the topic of the credibility of testimony con-
cerning miracles. They are careful to locate
these arguments within the broader frame-
work of Hume’s reservations about our abil-
ity to justify expectations about the future
in a rational, non-circular manner and his
pragmatic response to those sceptical wor-
ries. Although causal reasoning cannot be
supplied with a non-circular argumenta-
tive defence, it remains the case that human
beings find such reasoning persuasive and
continue to use it, even after exposure to
sceptical arguments, as a touchstone for
assessing whether particular beliefs are ones
they are content to endorse or ones that are
no more than mere foolishness. Consequently
Pritchard and Richmond construe Hume as
attempting to show that no testimony about
the occurrence of miraculous events capable
of serving as the foundations of a system of
religion has ever met the standards of dox-
astic acceptability that normally prevail in
less contentious cases when we are weigh-
ing human testimony concerning an alleged
event against the implications of our obser-
vations of past natural regularities.
Pritchard and Richmond reject the suppo-
sition that any form of a priori conceptual
2/9/2012 9:37:58 PM
 INTRODUCTION
argument forms part of Hume’s case against
belief in miracles, and they also maintain
that it is a mistake to construe him as arguing
that the kind of regularity in experience that
would need to be interrupted in order for an
event to qualify as a plausible candidate for
being a miracle would be so well entrenched
and confirmed that no possible amount of
human testimony could render it appropriate
to believe that an interruption had occurred.
They emphasize that for Hume it is always
a contingent matter whether the testimony
offered is weighty enough to overcome the
initial presumption that a hitherto well-con-
firmed natural regularity with no previously
known exceptions has not abruptly come to
an end. Nevertheless the standards of dox-
astic acceptability we embrace in practical
contexts when we are making judgements
in a careful and reflective manner are such
that this testimony needs to overcome an
exceptionally high hurdle. Unless the plau-
sibility that this testimony is mistaken or
deliberately deceitful is even lower than the
extremely low plausibility that attaches to
the supposition that a pervasive and well-
tested regularity that has previously mani-
fested itself throughout all human history
has been breached at a particular time and
place, it is not appropriate for us to accept
that this testimony is correct. And although
testimony of this quality is at least conceiv-
able, Pritchard and Richmond hold that
even when we assess Hume’s arguments
from the perspective of Bayesian reasoning
or the non-reductionist view that testimony
can possess some independent credibility
that does not ultimately derive from non-
testimonial sources, it is apparent that Hume
manages to present a strong case for the
conclusion that such exemplary testimony
has never yet been forthcoming in the case
of any allegedly miraculous event presented
10
9780826443595_Intro_Final_txt_print.indd 10
as confirmation of the truth of religious doc-
trines or teachings.
If Hume is right to maintain that reports
of alleged miracles fail to offer any genuine
support for the bold claims advanced by
religions about the ultimate nature of real-
ity, where might such support be found?
Hume’s religious contemporaries placed
great confidence in the probative value of
the design argument, and Hume undertook
a detailed examination of this argument in
his Dialogues concerning Natural Religion,
which were published posthumously in
1779.
Andrew Pyle accordingly presents in his
chapter an overview of the complex discus-
sion that occurs between Hume’s principal
characters in the Dialogues, and he arrives at
the conclusion that this work was intended
to show that the design argument cannot
legitimately support theistic conclusions and
that a naturalistic explanation of the orderly
nature of the universe is, when judged by
everyday standards of good causal reason-
ing, more acceptable than a theistic one. An
interpretation of the Dialogues along these
lines might initially be thought to overlook
Hume’s repeated suggestions that our intel-
lectual faculties are wholly inadequate when
confronted by the task of arriving at satisfac-
tory conclusions about such rarefied matters
of inquiry. Those pronouncements appear to
give very strong support to the conclusion
that Hume holds that the only legitimate
response to questions about the ultimate ori-
gins of the universe is a stance of complete
neutrality and suspension of judgement.
However, Pyle draws an important distinc-
tion between a commitment to a particular
hypothesis as more probable than all com-
peting hypotheses with equivalently detailed
content and a comparative judgement that
a particular hypothesis is more likely to be
2/9/2012 9:37:58 PM

true than some specified rival hypothesis.
He accepts that Hume is indeed inclined to
maintain that no detailed hypothesis that
we can formulate about the ultimate origins
of the universe and the order it displays is
worthy of endorsement as an explanation
that is more likely to be true than false. But
he maintains that Hume takes the view that
the empirical evidence, inadequate as it is in
terms of favouring a particular determinate
theory as the most likely theory, does at least
marginally favour a naturalistic account of
the universe as more plausible than the
world-view represented by theism and con-
ventional deism.
As Hume was only too well aware, the
apparent paucity of good empirical evidence
for theism or conventional deism has not
prevented the emergence of popular theis-
tic religions with vast numbers of nominal
adherents. How can the widespread preva-
lence of this form of belief be explained if
it is not a response to evidence; and even if
theistic religions potentially lack the virtue
of offering a true description of the universe
and our place within it, could it be the case
that the existence of such religions is a vital
bulwark of morality and an expression of the
highest and most sublime aspects of human
nature?
David O’Connor investigates the account
Hume provides of the psychological origins
of religious belief, and he contends that this
account is a strongly deflationary one. Not
only does Hume describe in the Natural
History of Religion a set of psychological
mechanisms that explain how religions can
arise and sustain themselves even if their core
metaphysical and historical claims are both
unwarranted and untrue, but he also presents
some of those mechanisms as dependent on
such unedifying aspects of human nature as
ignorance, fear and servile self-abasement.
11
9780826443595_Intro_Final_txt_print.indd 11
INTRODUCTION
O’Connor emphasizes the role Hume allo-
cates in his causal story to the human dis-
position to anthropomorphize phenomena.
People living in relatively sophisticated soci-
eties, where information about the genuine
hidden causes of otherwise puzzling phe-
nomena is fairly widespread, often engage
in what we might term playful anthropo-
morphism as an amusement or a deliberately
chosen form of metaphor. O’Connor argues
that Hume contrasts such playful anthro-
pomorphism with a literal-minded anthro-
pomorphism that characteristically emerges
when people have little or no grasp of the
true aetiology of striking or potentially dan-
gerous phenomena. In those circumstances
the human proclivity to think in terms of
otherwise unexplained phenomena as arising
from agency and intention exerts itself with
full force; and as no observable agents can
be detected with the appropriate intentions
and purposes, the idea develops of multiple
invisible and intelligent powers that have a
concern with human affairs. Such speculative
notions are, of course, theoretically distin-
guishable from actual beliefs. But O’Connor
locates the mechanism that takes people from
a spontaneous conception of invisible and
intelligent powers to belief in the existence
of such powers in human fear and an acutely
distressing sense of vulnerability. Once those
lively and pervasive passions are engaged, a
mere picture of the world is transformed into
a set of beliefs that guide people’s actions.
In particular, people attempt to relieve their
helplessness and sense of vulnerability by
treating these hidden agents as susceptible to
manipulation by flattery and supplication.
Ironically many theists would probably be
happy to endorse this or some similar account
of the origins of polytheism. O’Connor, how-
ever, argues that Hume’s account of the psy-
chological genesis of theistic religion is every
2/9/2012 9:37:58 PM
 INTRODUCTION
bit as subversive as his explanation of poly-
theism. Hume, in O’Connor’s judgement,
rejects as wholly inconclusive the supposed
evidence for theism from miracle reports
and such arguments as the cosmological
argument and the design argument. And the
emergence of theism from a background of
antecedent polytheism is explained by Hume
in terms of an attempt by people to ingratiate
themselves with a particular invisible power
by assigning to that agent ever more impres-
sive attributes and abilities in much the same
way that one might seek to curry favour with
a murderous human despot by eulogizing his
or her non-existent qualities of wisdom, jus-
tice and benevolence. This practice of base
flattery inevitably corrupts over time even
the judgement of the original flatterers, and
Hume sees its impact on the beliefs of other
people in society, especially when aided by
education and religious instruction, as even
more profound and pernicious. In this man-
ner, some previously negligible deity of com-
parable status to a petty human princeling is
potentially exalted over many generations,
without any assistance from cogent truth-
oriented reasoning, into the supposedly
omnipotent, infinitely perfect and wholly
just creator of the entire universe.
Of a piece with his attitude towards, and
arguments against, religion Hume provides a
secular moral theory, one in which there is
no place for God. Julia Driver turns to this
account and its grounding in the natural
emotional responses we have to the happiness
and suffering of those around us. As O’Brien
and Hardin also go on to discuss in subse-
quent chapters, such emotional responses
have their source in our sympathy with oth-
ers – sympathy, for Hume, seen as the capac-
ity we have to share the emotions of our
fellows. And, when the actions of a person
lead us to feel pleasure, via sympathy with
12
9780826443595_Intro_Final_txt_print.indd 12
those affected by that action, we are drawn
to feel a certain moral sentiment of approba-
tion. The feeling of such a moral sentiment
constitutes, for Hume, a moral judgement.
Given that moral judgements depend
on our emotional responses to others, they
must be, at least in some sense, subjective
since they are not independent of our natu-
ral human responses to each other. Driver,
however, explores how moral judgements
can nevertheless possess a kind of objectiv-
ity in that moral truths are independent of
an individual’s particular responses to a
certain action. Such objectivity is supplied
by our ability to adopt the general point of
view. We can ‘correct’ our sometimes mis-
guided moral judgements because we are
able to adopt a perspective divorced from
our own, a perspective encompassing the
‘narrow circle’ of those affected by a certain
action and not biased by our own concerns
or interests. Can, then, a moral judgement be
true or false? Driver argues that it can, the
ultimate grounding for the truth of a moral
judgement lying in the utility of the actions
that we judge to be virtuous from the general
point of view.
Dan O’Brien continues to explore Hume’s
account of morality, focusing on his concep-
tion of virtue and vice. Virtues for Hume
are those character traits that are useful and
agreeable to ourselves and to others, and
thus people manifest many different kinds of
virtue and many different vices. Hume denies
that all virtues are innate and God-given and
highlights the importance of artificial virtues,
traits that people in societies have developed
in order to aid the social interactions within
those communities. We come to be able to
see certain traits as virtuous through sympa-
thizing with the effects that a person’s behav-
iour has on those around them. Benevolence
is virtuous because I resonate to the pleasure
2/9/2012 9:37:58 PM

that the benevolent person’s actions bring
to his friends and acquaintances. Reason,
however, also plays a role here, but only in
helping us to appreciate what sentiments we
should feel if we are to be impartial in the
requisite way.
O’Brien goes on to explore how Hume
subverts the religious conception of virtue or
what Hume calls the ‘monkish virtues’; ele-
vating pride to his first natural virtue, a due
sense of pride being agreeable to ourselves
and ultimately useful in our social engage-
ments. That many traits are useful and agree-
able is uncontroversial, but a distinction is
also usually drawn between traits that have a
moral dimension and those that do not: ben-
evolence and compassion are of the former
kind; dexterity and wit, the latter. Hume,
however, thinks that distinctions hereabouts
are not at all sharp and that moral virtues
are not different in kind from other benefi-
cial ways of behaving.
Russell Hardin moves the discussion
away from questions concerning the nature
of moral judgement and what constitutes a
good character to a consideration of Hume’s
analysis of how self-interest can give rise
to social conventions and organizations
that promote public benefits. Hardin views
Hume as deliberately eschewing the attempt
to show that moral principles are true or
rationally justified in favour of a scientific
investigation into how beings who are pre-
dominantly motivated by self-interest never-
theless create institutions and social practices
that serve the collective good. Numerous
philosophers have implausibly purported to
show that altruistic behaviour is a funda-
mental dictate of reason or a requirement
imposed upon us by some divine lawgiver,
but Hume is seen by Hardin as adopting the
radically different and substantially more
illuminating approach of explicating how
13
9780826443595_Intro_Final_txt_print.indd 13
INTRODUCTION
behaviour that serves the interests of other
people is primarily a product of psychologi-
cal mirroring and the conventions that arise
when self-interested agents of limited power
are attempting to maximize their own ben-
efits from repeated interactions with other
similarly self-interested agents.
Hardin describes psychological mirroring
as an automatic response that people show
to the actions and emotional states of other
people. It is readily observable that human
beings have a strong tendency to mimic
the behaviour of the people around them.
However, it also seems to be the case that
most of us find other people’s observed emo-
tional states similarly infectious. Observing
a person showing clear signs of distress or
fear tends to give rise to analogous emo-
tional states in the spectator. And behaviour
manifesting joy and gladness has at least
some tendency to raise the spirits of a person
who observes such behaviour. Hardin credits
Hume with being one of the first thinkers to
explore in any detail the implications of this
phenomenon for human actions and choices.
Given the existence of psychological mirror-
ing, the psychological states of other people
cannot be a matter of complete indifference
to us. No matter how self-interested we hap-
pen to be, our own lives are more satisfac-
tory, all other things being equal, when the
people around us are also faring well. And
this responsiveness to the psychological states
of other people is what Hardin identifies as
lying at the core of the Humean principle of
sympathy.
Sympathy alone, however, is an inadequate
explanation for the range of circumstances in
which people seem to accept some check on
their self-interest so that the interests of other
people can be safeguarded or promoted.
Hardin accordingly attaches great importance
to Hume’s exploration of the way in which
2/9/2012 9:37:58 PM
 INTRODUCTION
repeated interaction between people comes to
shape social behaviour in ways that see collec-
tive benefits emerging from self-interest. Even
a purely self-interested agent needs to return
favours if he or she is to have much chance
of securing the co-operation in the future of
people who are aware of past performances.
And if we advance to a more sophisticated
level, justice, in the sense of social stability
and good order, is something that we all have
some interest in promoting even if the institu-
tions and habits that maintain stability and
order sometimes prove inconvenient to us
on particular occasions. Hardin accordingly
presents Hume as someone who succeeds in
setting before us a detailed account of the
self-interested strategies that lead to the evo-
lution of some of the most salient social prac-
tices and forms of organization that serve to
enhance our collective well-being.
One important aspect of any human soci-
ety is the relationship between the sexes,
and Hume’s views on this relationship and
on gendered differences have aroused con-
siderable controversy. Some commentators
have accused him of acquiescing in and even
actively seeking to defend sexist forms of
discrimination and oppression. Other read-
ers of his writings have, in contrast, seen him
as someone who seeks to enhance the status
of women and also wishes to see some of
the values and psychological characteristics
traditionally associated with women dissemi-
nated more widely throughout society.
Lívia Guimarães mounts a strong defence
of the view as elsewhere there should be an
accent on one of the i’s of Livia that it is the
latter reading of Hume’s position that offers
the more illuminating perspective on his
attitude towards women and conventional
distinctions between masculine and femi-
nine characteristics. She helpfully reminds
us of Hume’s many close relationships with
14
9780826443595_Intro_Final_txt_print.indd 14
women throughout his life, relationships
often based on mutual respect and shared
intellectual interests rather than transient
sexual or romantic passion – although it is
also clear from her account that such passion
was certainly not wholly alien to Hume’s
character. Even more importantly Guimarães
identifies Hume’s writings as showing a great
willingness to deconstruct gender dichoto-
mies. In his History of England, female
characters are frequently portrayed as active
agents endowed with energies, drives and
reasoning abilities that are at least equivalent
to anything possessed by the men surround-
ing them. And Guimarães argues that when
Hume is explicitly engaged in the study of
human nature at a more theoretical level,
his emphasis on human beings as embodied
mammalian animals responding to the influ-
ence of concrete conditions including social
circumstances and personal relationships
means that he avoids the trap of construct-
ing an idealized account of our nature that
uncritically sees its essence as lying in such
allegedly masculine virtues as pure ration-
ality and the suppression of the passions.
Moreover, Hume’s account of human reason
and inference further subverts traditional
gender categories by presenting such reason-
ing as founded in associations of ideas, our
passions and the faculty of sympathy.
Guimarães also notes that when Hume is
working within traditional gendered catego-
ries, he usually speaks in favour of the wider
diffusion of supposedly feminine character-
istics. Such virtues as tenderness, benevo-
lence and mildness are not simply seen by
Hume as appropriate for women. Instead
he argues that society, as a whole, would
benefit from these virtues being more wide-
spread amongst men as well. Hume read-
ily acknowledges that the martial virtues of
aggression, fierceness and intransigence have
2/9/2012 9:37:59 PM

utility in primitive societies, where violence is
needed to maintain order and to repel inva-
sion and despotic oppression. But in more
civilized societies Hume sees other virtues
as more effective at promoting the general
well-being, and Guimarães draws our atten-
tion to the fact that Hume frequently indi-
cates that these less abrasive virtues are best
spread throughout society by increasing the
opportunity for women to exert their influ-
ence on men. Guimarães therefore concludes
her chapter with the striking suggestion that
Hume can be seen both as sketching out an
ideal society that would constitute a femi-
nist utopia and as recommending a greater
emphasis on supposedly feminine virtues and
attributes as an effective means of improving
existing societies.
Margaret Schabas investigates Hume’s
economic thought. Schabas points out that
Hume differs from most present-day econo-
mists by emphasizing the greater value to the
individual of mental well-being rather than
material wealth. Nevertheless Hume’s inter-
est in all aspects of human nature and the
social forces that shape people’s lives meant
that he corresponded and published quite
extensively on matters of economic policy
and theory. Indeed Schabas argues that the
circulation and influence in the eighteenth
century of Hume’s economic essays entitles
him to be seen as one of the foremost econo-
mists of his era. His pre-eminence amongst
British theorists was eventually to be usurped
by his friend Adam Smith. Schabas reminds
us, however, that this was not to happen until
the 1790s, despite the publication of Smith’s
The Wealth of Nations in 1776 and Hume’s
death in that same year.
Schabas concentrates in her chapter on
Hume’s response to the conspicuous rise in
the wealth of Western Europe in the seven-
teenth and eighteenth centuries. She draws
15
9780826443595_Intro_Final_txt_print.indd 15
INTRODUCTION
our attention to the care that Hume took to
investigate this phenomenon by seeking out
the best available data and using his exten-
sive acquaintance with classical authors to
compare Europe with the civilizations of the
ancient world. Moreover, Schabas maintains
that Hume was right to explain this rise in
wealth by invoking the combined impact of
division of labour and the increased supply
of silver coinage made possible by the mines
of the New World.
Hume also emerges as equally astute in
his reflections on the consequences of such
additional wealth for human happiness and
welfare. Unlike conservative critics of wealth
and luxury who saw and often still affect to
see such things as harbingers of moral decay,
Hume held that the modern commercial
world and the opportunities that it gener-
ated had an improving effect on civil soci-
ety and people’s characters. Schabas presents
him as arguing that trade and manufactur-
ing promoted civility, gave a new impetus to
learning and human ingenuity, and enhanced
liberty and equality. Hume saw these benign
influences as most readily impinging on those
located in the middle ranks of society, but a
flourishing middle class helped to bind all of
society together in ways that ultimately bene-
fited everyone. So although Hume was fully
prepared to disparage the rapacious acqui-
sition of expensive trinkets, Schabas locates
in his writings an ingenious account of how
wealth indirectly promotes human happi-
ness. Hume’s predisposition towards Stoic
values meant that he viewed material posses-
sions in themselves as being of little conse-
quence once the necessities of life had been
supplied, but the process of acquiring wealth
through trade and participation in manufac-
turing served the crucial role of giving people
the opportunity to gain personal satisfaction
and a sense of purpose from the exertion
2/9/2012 9:37:59 PM
 INTRODUCTION
of their mental and physical powers in an
undertaking that was immediately appealing
to them. As might be expected with Hume,
this confidence in the ameliorative powers of
the commercial world is hedged around with
substantial reservations. Schabas indicates
that these reservations presciently included
worries about the destructive power of pub-
lic debt in the hands of politicians and the
prediction that the American colonies and
China would eventually eclipse Britain and
other European nations in terms of trade
and manufacturing output. But Schabas
amply succeeds in showing that Hume’s case
for supposing that commerce and the pursuit
of wealth can often improve people’s disposi-
tions and moral character remains a useful
antidote to the unthinking prejudice that
morality and personal development are best
promoted by austerity and the eschewal of
luxury.
Mark Rowe, in contrast, is less sanguine
about the merits of Hume’s account of a
standard of taste in matters of aesthetic
judgement. Rowe argues that Hume is con-
cerned to reconcile a form of subjectivism
about aesthetic taste with the supposition
that some aesthetic judgements can be mis-
taken. The subjectivism is a product of
Hume’s commitment to the view that aes-
thetic qualities are projected onto the world
rather than discovered in the world. But the
need to find some room for the concept of
an error in aesthetic judgement arises from
Hume’s conviction that some judgements of
aesthetic merit would be as perverse and as
obviously illegitimate as some patently false
pronouncements about physical objects and
such qualities as size and shape.
A part of the answer to Hume’s dilemma
lies in the concept of a qualified judge, and
Rowe enumerates the attributes that Hume
regards as essential if someone is to be viewed
16
9780826443595_Intro_Final_txt_print.indd 16
as such a judge. The thought here is that aes-
thetic pronouncements issuing from people
who lack those attributes can be dismissed in
much the same way as the colour judgements
of someone known to be suffering from a
fever or viewing an object in non-standard
lighting conditions carry no weight with us
in respect of our assessments of the object’s
real colour. This, however, raises the question
of whose judgement is to be accepted if and
when people who count as qualified judges
disagree. In some specific instances Hume
is perfectly content to say that the disagree-
ment is irresoluble. But Rowe points out
that Hume is not always so accommodat-
ing: sometimes he seeks a standard of taste
that can override or correct the judgement of
qualified judges.
Rowe maintains that all Hume’s attempts
at explaining how such corrections can be
given legitimacy are unsuccessful. Hume
sometimes appeals to rules of composition,
but Rowe powerfully argues that these rules,
on a Humean account, can be nothing more
than inductive generalizations that summa-
rize the characteristics displayed by works
that people usually find pleasing and beau-
tiful. Thus they lack the normative force
that Hume requires. Similarly, Rowe rejects
Hume’s alternative appeal to a consensus
amongst qualified judges. Even if majority
opinion were against your personal verdict,
would it make sense for you, as a person
with the attributes requisite for being a
qualified judge, to treat your judgements as
wrong simply because you are in a minority?
Finding yourself in a minority might well give
you grounds for reviewing your reactions to
a particular work or artistic performance
again. But being in a minority of qualified
judges is not constitutive of being wrong in
your aesthetic judgement. Rowe concludes
that although Hume rightly sees the need to
2/9/2012 9:37:59 PM

accommodate critical discussion when con-
sidering questions of aesthetic merit, his view
of what constitutes argument in that arena
is too impoverished to explain how genuine
debate can take place and real discoveries
can be made.
Timothy Costelloe introduces us to Hume’s
conception of ‘philosophical’ or ‘true’ history,
and his analysis of what Hume means by
such history helps to bind together Hume’s
more explicitly philosophical works and his
History of England. Some of Hume’s more
malicious critics have accused him of effect-
ively abandoning philosophy after the poor
reception afforded the Treatise in order to
pursue money and popular fame through the
alternative means of writing a best-selling
history. However, Costelloe brings into focus
numerous important continuities between the
philosophical project pursued in the Treatise
and Hume’s aspirations for his History.
Costelloe argues that Hume sees ‘philo-
sophical’ history as an attempt to combine
responsiveness to evidence, rather than the
promptings of partiality or the imagination,
with a reconstruction of the past that sus-
tains the reader’s interest and gives him or
her a lively sense of the truth of the events
portrayed. A mere propagandist concentrates
only on the second of these two tasks; but if
the author makes no attempt to select events
and shape the narrative in a way that will
appeal to the reader’s imagination and pow-
ers of empathy, then the resulting work will
be entirely unreadable. Thus a writer of the
kind of history that Hume views as worthy
of a philosophical author needs to have the
skills to weigh testimony carefully, a passion
for the pursuit of the truth that promotes
impartiality and overrides any temptation
to flatter influential patrons, and the capac-
ity to keep the imagination in check so that
fanciful associative links do not crowd out
17
9780826443595_Intro_Final_txt_print.indd 17
INTRODUCTION
conclusions supported by real evidence. At
the same time the writer requires many of the
skills of the poet or dramatist: the truth must
be shaped and ordered so that a reader read-
ily enters into the narrative and engages with
the character and situation of the personages
presented to him.
Costelloe emphasizes that Hume sees his-
tory written in this manner as serving the
crucially important function of laying out
the past before us so that it can serve as a
guide to future conduct. Firstly, it allows dis-
tant past events to be used by the scientist of
human nature as a means of confirming or
refuting hypotheses about our mental mech-
anisms and dispositions. Thus it provides the
philosopher, in his role as a psychological
anatomist, with the data he needs to guide
and refine his conclusions. However, it also
serves the second function of improving our
moral judgements. People and events close to
us in time and space are frequently assessed
through a prism of partiality that prevents
us from seeing how they strike other people,
and important potential consequences have
often not yet had a chance to manifest them-
selves. In contrast, if we are reading a histori-
cal narration of events that took place many
years ago that involved people not intimately
connected to us, we have an opportunity
to arrive at less biased and better-informed
moral judgements, a habit that can then be
transferred to situations in which our own
interests are at stake. Thus Costelloe main-
tains that Hume regarded his philosophical
and historical investigations as seamlessly
intertwined. Just as abstruse philosophy is
depicted in the first section of the Enquiry
concerning Human Understanding as guid-
ing us to a better understanding of human
nature and easy philosophy is portrayed as
inspiring us to act virtuously, so too ‘philo-
sophical’ history continues the task of laying
2/9/2012 9:37:59 PM
 INTRODUCTION
bare the hidden mechanisms of human action
while simultaneously depicting virtuous and
vicious characters in such a light that we
sympathize with the former and are repelled
by the latter.
Paul Russell brings the volume to a con-
clusion with a discussion of changing trends
in the interpretation of Hume’s philosophi-
cal position. Russell distinguishes between
the interpretation of a philosopher’s position
and the legacy of that position. Interpretation
is a matter of arriving at an understanding
of a philosopher’s original aims and inten-
tions, whereas the legacy is constituted by
the reception of his or her views and their
fruitfulness over time. The various compet-
ing interpretations that might arise form
part of that reception, but there need be
no correlation between the dynamism of
the interpretative framework and its philo-
sophical fecundity. Similarly, a sensitive and
well-balanced interpretation might reveal
itself over time to be nothing more than
the accurate signposting of a barren philo-
sophical cul-de-sac, whereas an interpreta-
tion that is little more than a caricature of
a philosopher’s actual project might for-
tuitously inspire subsequent philosophical
developments of great value and independ-
ent interest. Russell, however, warns against
the error of supposing that an important and
interesting legacy confirms the accuracy of
the interpretation that generated it. He also
points out that if we complacently allow an
interpretation of a philosopher’s views to go
unchallenged because it is linked to a valu-
able legacy, we are in danger of forgoing
important philosophical developments that
might arise from reflection on some plausi-
ble alternative interpretation.
In the specific case of Hume, Russell
argues that much of the fruitfulness of his
legacy up to this point has arisen from a
18
9780826443595_Intro_Final_txt_print.indd 18
pattern of interpretation that stresses his rad-
ical empiricism and his affinities with Locke
and Berkeley. Critics of empiricism have seen
Hume as providing the valuable service of
exposing how radical empiricism ultimately
collapses into incoherence, whereas philoso-
phers of empiricist sympathies have often
purported to find in Hume the inspiration for
developing what they hoped would be a suc-
cessful version of empiricism that eschewed
all a priori metaphysical speculations. Russell
is not inclined to deny that this interpretative
tradition has inspired much important philo-
sophical work, but he does deny that it offers
the resources required to construct an accu-
rate interpretation of the underlying nature
of Hume’s philosophical stance.
The difficulty that Russell identifies is
that it has become increasingly clear that
the empiricist elements of Hume’s thought
co-exist with a naturalistic programme
that involves the construction of an intri-
cate science of human nature that purports
to be based on experience and experiment.
However, the interpretation of Hume as a
radical empiricist seems to have sceptical
implications that are inconsistent with the
development of such a science. Yet if we
water down the empiricism in order to make
it more compatible with the positive side
of Hume’s philosophy, it remains the case
that Hume’s philosophical writings appear
to contain an array of explicitly sceptical
arguments that do not need to be embedded
within a framework of radical empiricism in
order to pose a serious challenge to his natu-
ralistic project.
Russell strikingly sums up the situation as
generating the worry that Hume’s philosoph-
ical outlook is ultimately broken-backed.
The sceptical aspect of his philosophy, which
seems to be clearly present even if it is not to
be construed as generated by a radical form
2/9/2012 9:37:59 PM

of empiricism, does not initially appear to
cohere well with Hume’s science of human
nature. Russell accordingly maintains that the
way forward is to place Hume’s philosophical
writings in a new interpretative framework,
one that sees Hume not as part of a trium-
virate of British Empiricists or a follower
of Newton or Hutcheson but as someone
who is actively attacking the metaphysical
and moral foundations of Christianity as a
member of a partially concealed tradition of
‘speculative’ atheism. In Russell’s judgement,
recognition of Hume’s irreligious inten-
tions as manifested even within the Treatise
provides the key to an account of Hume’s
writings that can reconcile his sceptical
19
9780826443595_Intro_Final_txt_print.indd 19
INTRODUCTION
arguments with the more positive aspects of
his philosophical position. Whether Russell
is right to imply that this is both necessary
and sufficient to permit such a reconcilia-
tion is not yet clear. However, other work by
Russell has certainly undermined the suppo-
sition that the contents of the Treatise lack a
substantial connection to issues of religion.
And it can safely be asserted that solving the
puzzle of how to harmonize the sceptical
and positive sides of Hume’s philosophy is
now widely acknowledged to be one of the
principal tasks that needs to be accomplished
if we are ever to possess a truly satisfactory
interpretation of all the essential elements of
Hume’s philosophical perspective.
2/9/2012 9:37:59 PM
 1
HUME’S LIFE, INTELLECTUAL
CONTEXT AND RECEPTION
Emilio Mazza
These are a few particulars, which may
perhaps appear trifling, but to me no
particulars seem trifling that relate to so
great a man (W. Cullen to J. Hunter, 17
September 1776)1
1. ‘WAKE-MINDED’
‘Our Davie’s a fine good-natured crater, but
uncommon wake-minded’, Hume’s mother
is supposed to have said in a piece of familial
assessment.2 And with regard to Hume’s reli-
gious principles, his brother John ventured the
opinion: ‘My brother Davie is a good enough
sort of man, but rather narrow minded’.3
This latter judgement echoes in its choice of
words Hume’s own recollection that in Paris
they ‘used to laugh at me for my narrow way
of thinking in these particulars’ (LDH 2.273,
484), and his description of Rousseau – ‘a
very agreeable, amiable Man; but a great
Humourist’ (LDH 2.13, 303; see LDH 2.130,
381) – indicates that Hume shared with his
brother a partiality for verbal sallies that com-
bined initial restrained praise with a less com-
mendatory ending. We seem too to find in the
Enquiry concerning the Principles of Morals, a
riposte on Hume’s part to his mother’s assess-
ment of his character: ‘any remarkable defect
20
9780826443595_Ch01_Final_txt_print.indd 20
in . . . understanding, dignity of mind, would
bereave even a very good-natured, honest man
of this honourable appellation. Who did ever
say, except by way of irony, that such a one
was a man of great virtue, but an egregious
blockhead?’ (EPM App. 4.2 / 314).4
Young Hume was troubled by a ‘weak-
ness’ of spirits; later on he would see a sig-
nificant relationship between ‘delicacy’ and
‘weakness’ of the mind (LDH 1.17, 3; 1.397,
214). The first as well as the last edition of
his Essays open with ‘Of delicacy of taste
and passion’, and only in 1772 does Hume
stop claiming a ‘very considerable connex-
ion’ between these delicacies in the original
frame of the mind (E 603). His mother’s
supposed saying has been discussed for 150
years by those who seek to defend the repu-
tation of ‘one of the greatest philosophers of
any age, and the best friend to mankind’, as
d’Holbach calls Hume, without contradict-
ing a woman of ‘singular merit’, as Hume
calls his mother.5
At the end of his Life Hume celebrates
himself as ‘a man of mild dispositions, of
command of temper, of an open, social,
and cheerful humour’, and his temper as
‘naturally cheerful and sanguine’ and not
‘very irascible’ (MOL, LDH 1.1–3). In
2/9/2012 9:39:42 PM

1757, somewhat between jest and earnest,
Hume says he is a ‘good-natured man of a
bad character’ (LDH 1.264, 139) and also
(if the text be Hume’s) a ‘very good man,
the constant purpose of whose life is to
do mischief’.6 The history of his writings
shows him as a man of ‘superior genius’, a
quality that he does not even recognize in
d’Alembert, who was simply a man of ‘supe-
rior parts’, even though after Paris Hume
considered him ‘with some few exceptions
(for there must always be some exceptions)
. . . a better model of a virtuous and philo-
sophical character’ (LDH 2.110, 363).
2. MY OWN (UNSUCCESSFUL)
WRITINGS
In 1734, when he begins ‘to despair of ever
recovering’ from his ‘Disease of the Learned’,
Hume wrote ‘a kind of History of my Life’
(LDH 1.13, 3; 1.17, 3); in 1776, when his life
is really ‘despaired of’, he writes ‘the History
of my own Life’ or My own Life (LDH 2.318,
522A; LDH 2.323, 525). This short ‘funeral
oration of myself’, Hume says, contains ‘little
more than the History of my Writings’, since
‘almost all my life has been spent in liter-
ary pursuits’ (MOL, LDH 1.7). The rhythm
of the Life is the alternation of learning and
business, expectations and disappointments,
which recalls that of action and repose in
the ‘Refinement in the Arts’ (E 270). Every
disappointment is overcome by character,
‘command of temper’ and ‘cheerful humour’
(MOL, LDH 1.7). Hume’s Life is also a his-
tory of the reception of his writings, where he
commonly distinguishes between immediate
and gradual success. The want of it is chiefly
measured by the standard of silence. The
Treatise and the first volume of the History
21
9780826443595_Ch01_Final_txt_print.indd 21
HUME’S LIFE
are the most ‘unfortunate’ books, and the
Political Discourses a work ‘successful on the
first publication’ and ‘well received abroad
and at home’. The more sustained success
begins in the 1750s, when Hume discovers
symptoms of a ‘rising reputation’, including,
for example, a ‘railing’ reaction of the clergy
(MOL, LDH 1.2–4).
3. ‘NEVER TO REPLY TO ANY BODY’
In about 20 years (1739–61) Hume publishes
almost every kind of writing: a Treatise,
its Appendix and Abstract; a Letter to a
friend and a True Account of the conduct of
another friend, the Essays, the Philosophical
Essays and the Enquiries; the Discourses and
the Dissertations; the History, the Natural
History and a Dialogue. He also receives
almost every kind of answer. In 1766 he
observes: ‘I could cover the Floor of a large
Room with Books and Pamphlets wrote
against me, to none of which I ever made the
least Reply, . . . from my Desire of Ease and
Tranquillity’ (LDH 2.92, 351).
With regard to the years 1749–51, in 1776
Hume declares he has ‘fixed’ and ‘inflexibly
maintained’ a resolution ‘never to reply to
any body’ (MOL, LDH 1.3). He starts assert-
ing this resolution in the second half of the
1750s, as a reaction to the ‘Warburtonian
School’, but in 1760 he declares that he
formed it ‘in the beginning of my Life, that is,
of my literary Life’ (LDH 1.320, 172), which
seems therefore to begin with a commitment
to refuse any literary controversy. Like many
official claims, however, this is not completely
reliable even though it does contain a sub-
stantial admixture of truth. He often replies
indirectly to his critics in his writings, and
sometimes he is even driven to the expedient
2/9/2012 9:39:43 PM
 HUME’S LIFE
of explaining that he is giving an answer and
that it should be extended to different adver-
saries, as in the 1775 ‘Advertisement’ to the
Enquiries (LDH 2.301, 509).
In 1757 someone suggests that he has
deliberately ‘so larded his Work with
Irreligion’ that the first volume of the
History ‘might sell’,7 and Hume observes
that the few ‘Strokes of Irreligion’ are of
‘small Importance’, even though they are
likely ‘to encrease the sale’ (LDH 1.250, 132;
1.256, 136). A few months later he allows
that he would accept the challenge to defend
The Natural History of Religion against
Warburton’s criticisms were he attacking his
‘principal Topics’. As he tells the bookseller:
‘The Hopes of getting an Answer, might
probably engage [Warburton] to give us
something farther of the same kind; which
at least saves you the Expence of advertising’
(LDH 1.265–67, 140).
Concerning his no-reply resolution, in
1758 Hume still maintains that he ‘shall
probably uphold it to the End of [his] life’.
The Concise and Genuine Account of his dis-
pute with Rousseau recalls that Hume ‘hath
seen his writings frequently censured with
bitterness . . . without ever giving an answer
to his adversaries’, yet, in the case of the
Rousseau imbroglio, the ‘circumstances’ were
such as to draw Hume into a scandal, ‘in spite
of his inclinations’. Consistent with them, he
authorizes the editors to declare ‘that he will
never take the pen again on the subject’.8
4. NOT UNFIT FOR BUSINESS: ‘THE
ARMY IS TOO LATE’
Like Lucian in De mercede conductis and
Apologia pro mercede conductis, which Hume
first quotes in 1751–2, he takes satisfaction
22
9780826443595_Ch01_Final_txt_print.indd 22
both in the fact that he has never ‘preferred a
request to one great Man, or even . . . [made]
advances of friendship to any of them’ and in
the fact that he has nevertheless found him-
self on good terms with such people in his
personal affairs, public business, and while
composing his History (MOL, LDH 1.5–6;
LDH 1.113, 63; 1.295, 156; 1.355, 191;
1.427–28, 232; 2.188, 422). As a man now
beyond middle age working for the Northern
Department, he finds that ‘to a Man of a lit-
erary turn, who has no great undertaking in
view, . . . public Business is the best Ressource
of his declining Years. Learning requires the
Ardor of Youth’ (LDH 2.385, 137). Thirty-
three years before, in spring 1734, trying to
leave his distemper behind and working on
the Treatise, he found ‘two things very good,
Business & Diversion’, and resolved ‘to seek
out a more active life’, laying ‘aside for some
time’ his pretensions in learning (LDH 1.17,
3; MOL, LDH 1.1).
In 1767 the Earl of Rochford remembers
that Hume is ‘unfit for business’, and Hume
himself has already admitted that the office
of secretary requires ‘a Talent for speaking
in public to which I was never accustomd’
(LDH 1.519, 289).9 However, Hume’s essay
‘Of Eloquence’ (1742) attacks, following
Swift and La Bruyère the ‘antient Prejudice
industriously propagated by the Dunces in
all Countries, That a Man of Genius is unfit
for Business’ (E 621), and the first Enquiry
(1748) claims that the accuracy of abstruse
philosophy is ‘subservient’ to every art or
profession: the politician, the lawyer and the
army general may take advantage from it
(EHU 1.9 / 10). In part, at least, this sounds
like a defence of those aspects of his life and
career that were not directly connected to
his literary and philosophical pursuits, for
at various times he found himself taking on
the roles of clerk for a merchant in Bristol
2/9/2012 9:39:43 PM

(1734), companion and tutor of a marquess
near St Albans (1745–6), secretary to a gen-
eral and judge-advocate in Lorient (1746–7),
secretary and aide-de-camp to the same
general in Vienna and Turin (1748), library-
keeper in Edinburgh (1752–7), secretary to
the Embassy and Chargé des affaires in Paris
(1763–6), and Under-Secretary of State for the
Northern Department in London (1767–8).
He also considers (but ultimately rejects) the
‘not agreeable’ life of the ‘Travelling Tutor’
(LDH 1.18, 3; 1.17–8, 24; 1.35–6, 14; 1.
57–8, 24; NLH 26, 10), even though he is
often ‘mortally sick at sea’ (LDH 2.206, 432;
1.214, 105; also see LDH 1.114, 64; 2.95,
352) and claims that ‘Shortness . . . is almost
the only agreeable Circumstance that can be
in a Voyage’ (LDH 1.105, 56).
Every time he is enjoying his solitude
Hume receives an ‘invitation’ he cannot refuse
(MOL, LDH 1.2, 4, 5–6). According to the
correspondence, his life is a permanent yearn-
ing for (philosophical) retreat and leisure, con-
tinuously thwarted by external circumstances,
leading him into some practical business: ‘I
lived several years happy with my brother at
Ninewells, and had not his marriage changed
a little the state of the family, I believe I should
have lived and died there,’ he says in 1759
about his own ‘reluctance to change places’,
even though in 1763 he has ‘so often changed’
his places of abode that he comes to think that
‘as far as regards happiness, there is no great
difference among them’ (LDH 1.295, 156;
1.415, 224; see also LDH 1.243, 128; 1.246,
130; 1.531, 295; 2.189, 423). With regard to
the years spent with General St Clair, the Life
claims that ‘these were almost the only inter-
ruptions which my studies have received dur-
ing the course of my life’ (MOL, LDH 1.2–3),
even though 15 years before he has allowed:
‘I have frequently, in the course of my life,
met with interruptions, from business and
23
9780826443595_Ch01_Final_txt_print.indd 23
HUME’S LIFE
dissipation; yet always returned to my closet
with pleasure’ (LDH 1.451, 244).
In 1746 Hume receives an ‘unexpected’
invitation from St Clair to go with him as sec-
retary in his military expedition, which was
planned to be an attack on French Canada
but came to its conclusion on the coast of
Brittany (MOL, LDH 1.2; LDH 1.382, 206;
1.92, 51; NLH 24, 10). He arranges his
‘Departure for America’ (‘Such a Romantic
Adventure, & such a Hurry’) with one box
of books and one of paper in his trunk (LDH
1.90, 50).10 Being asked whether he ‘would
incline to enter the Service’, he answers that
at his years he could not ‘accept of a lower
commission than a company’ (LDH 1.94,
52). One year after he says that for the ‘Army
[it] is too late’ (NLH 26, 10). The expedi-
tion is a ‘failure’, but it gives rise on Hume’s
part to a beautiful letter to his brother, a
brief journal or hypomnema, a piece on ‘The
descent on the coast of Brittany in 1746’, and
possibly an article (LDH 1.99, 54; 1.94–8,
53). The expedition also shapes Hume’s
opinions about soldiers. Major Alexander
Forbes, for example, is described as ‘a Man
of the greatest Sense, Honour, Modesty,
Mildness & Equality of Temper in the
World’: ‘His Learning was very great for a
man of any Profession, but a Prodigy for a
Soldier. His Bravery had been try’d & was
unquestion’d.’ When Forbes kills himself as
a result of anxiety and fear that he may have
been guilty of a dereliction of duty, Hume
maintains that in the course of dying from
his self-administered injuries, he expressed
a ‘steady Contempt of Life’ and ‘determind
philosophical Principles’. And after Hume
has seen his friend die in front of him, it is
probably Hume who also undertakes the
official duty of recording that one Dougal
Steuart was made Captain ‘in room of Alexr
Forbes deceast’ (LDH 1.97, 53).11
2/9/2012 9:39:43 PM
 HUME’S LIFE
In 1734 Hume had compared the soldier’s
courage to the devotee’s devotion (LDH
1.21, 4). In 1748 he publishes ‘Of National
Characters’, where a few pages could be enti-
tled ‘The Soldier and the Priest’. It is a double
reaction to his academic and military adven-
tures in Edinburgh and Lorient. In 1743
Hume reads Leechman’s sermon on prayer,
and sends him some remarks on argument
and style, together with 22 small faults that
the author does not even take into consid-
eration. The sermon, Hume concludes, una-
voidably makes his religious author ‘a rank
Atheist’ (NLH 10–14, 6). Despite his youth-
ful claim that ‘there is nothing to be learnt
from a Professor, which is not to be met with
in Books’,12 in 1744 Hume attempts in vain
to become professor at the University of
Edinburgh, and declares himself extremely
surprised that the ‘accusation of Heresy,
Deism, Scepticism, Atheism &c’ is supported
by the ‘Authority of Mr Hutcheson & even
Mr Leechman’ (LDH 1.58, 24).
In 1741 Leechman published another
sermon on the character of the priest. ‘Of
National Character’ is also an answer to
him. Leechman claims we can never clearly
‘unvail’ to mankind their ‘hidden hypocrisy’,
nor justly contempt the devout worship-
pers by calling the outward displays of their
inward devotion ‘solemn grimaces, and hypo-
critical airs’;13 Hume replies that the clergy-
men ‘promote the spirit of superstition, by a
continued grimace and hypocrisy’ and this
‘dissimulation often destroys’ their ‘candor’
(E 200n). Hume denounces their conceited
ambition, professional faction and perse-
cuting spirit. In contrast, soldiers are ‘lav-
ish and generous, as well as brave’, ‘candid,
honest, and undesigning’. Since ‘company’ is
their sphere they can acquire ‘good breeding
and an openness of behaviour’ and a ‘con-
siderable share’ of politeness (E 199). Even
24
9780826443595_Ch01_Final_txt_print.indd 24
though soldiers have their exceptions, Hume
passed his St Clair years ‘agreeably, and in
good company’ (MOL, LDH 1.2-3): reliable
officers, learned physicians, whist-players
and humorous people who dedicate them-
selves to the ‘service of the Ladys’. Among
priests, on the contrary, ‘gaiety, much less
the excesses of pleasure, is not permitted’
(E 201n). He is ‘in the Army’ and he calls
it ‘our Family’ (LDH 1.97, 53; 1.132, 64).14
With these ‘friends or confidents’ – he says
with Quintilian and Svetonius, or more sim-
ply with Voiture – he can be free ‘in seriis et
in jocis, – amici omnium horarum’ (in grave
and jocular manners, – friends of all hours)
(LDH 1.102, 56).
In 1747, when St Clair invites Hume to
go over to Flanders with him (LDH 1.108–9,
61), he has ‘a great curiosity to see a real
campaign’, notwithstanding his fears of the
‘expense’ and looking ridiculous as a result
of ‘living in a Camp, without any Character
& without any thing to do’ (NLH 23, 9).
Nothing could be ‘more useful’ to his ‘histor-
ical projects’. Hume looks forward to pick-
ing up a great ‘military knowledge’, by ‘living
in the General’s family, and being introduced
frequently to the Duke’s’ (ibid.). In 1748 he
attends St Clair in his mission to Vienna and
Turin, notwithstanding an ‘infinite regret’ for
leaving ‘stores of study & plans of thinking’
(LDH 1.109, 61; 1.111, 62). In accordance
with the opinions of Lucian, Bayle, Addison
and the Guardian, and following the advice
contained in a volume by Polybius, which
he keeps in his hand (LDH 1.100, 54), he is
looking for ‘an opportunity of seeing Courts
& Camps’:
this knowledge may even turn to account
to me, as a man of letters, which I con-
fess has always been the sole object of my
ambition. I have long had an intention,
2/9/2012 9:39:43 PM

in my riper years, of composing some
History; . . . some greater experience
of the Operations of the Field, & the
Intrigues of the Cabinet, will be requis-
ite, in order to enable me to speak with
judgement upon these subjects (LDH
1.109, 61).
St Clair arrives in Turin on the 8 May, Hume
and St Clair’s nephew, Sir Harry Erskine,
about eight days later; on 29 November
1748 they all set out.15 The result of the mis-
sion is a ‘long epistle’, which he calls a ‘sort
of Journal of our Travels’ (LDH 1.114, 64;
1.132, 64). Before the departure Hume opti-
mistically contrasts his situation with that of
the ‘severe’ Lord Marchmont, who, ‘entirely
employed in the severer studies’, suddenly
opens his eyes on a ‘fair nymph’ aged just 16
and marries her in a few days: ‘they say many
small fevers prevent a great one. Heaven be
praised, that I have always liked the per-
sons & company of the fair sex: For by that
means, I hope to escape such ridiculous pas-
sions’ (LDH 1.110, 61). Ten days after his
arrival he is already ‘troubled’ by an ‘indis-
position’ connected with the ‘pretty women’
of Turin. After two months he declares an
‘attachment’ for a Countess of 24.16 The
Turin-based Madame Duvernan anticipates
the Parisian Madame de Boufflers and their
extrovert public reputations stand some-
what in tension with Hume’s claim that he,
like Mandeville’s perfect sociable benevolent
man,17 took a ‘particular pleasure in the com-
pany of modest women’ and had therefore
no reason to be ‘displeased with the recep-
tion I met with from them’ (MOL, LDH
1.7). In summer 1764 he reminds his rever-
end friend Jardine that ‘A Man in Vogue will
always have something to pretend to with the
fair Sex’, and Jardine banters: ‘An inordinate
Love of the fair Sex . . . is one of those Sins,
25
9780826443595_Ch01_Final_txt_print.indd 25
HUME’S LIFE
that always, even from your earliest Years,
did most easily beset you’ (LDH 1.438, 237;
2.353, App. C. V).
In Turin Hume becomes bored and sick.
Admiral John Forbes called him ‘the sleeping
philosopher’, someone says he was ‘affected
by a most violent Fever’, some other that he
‘received Extreme Unction in a dangerous ill-
ness’. He hangs around with Lord Charlemont
and reads Montesquieu’s Esprit (LDH 1.133,
65).18 Consistent with his Treatise, and in the
name of ‘sympathy’, he enjoys the pleasure
and beauty of extended, fertile, cultivated
plains. He wishes to make ‘a short Tour thro’
some of the chief Cities of Italy’, but appar-
ently the Duke of Newcastle rejects the
request.19 He does the accounts (as he did in
Bristol) and examines the Sardinian docu-
ments in the Commissary’s office.20 He writes
St Clair’s official letters and copies them into
a letter book. He probably suggests passages
for St Clair’s letters (like the observation of
the historians that ‘Britain has commonly lost
by Treaties what she gain’d by Arms’) and cer-
tainly receives suggestions for his future writ-
ings: ‘Of the Balance of Power’ discusses the
peace of Aix la Chapelle, and the dying Hume
is still remembering those inconceivably ‘good
terms’ that France had granted to Britain.21
Hume’s experience in Turin resumes that
begun in Lorient and prepares the way for
his 1760s appointments in Paris and London.
General St Clair, Lord Hertford and his
brother General Conway all wanted Hume
with them. St Clair ‘positively refused to
accept of a Secretary from the Ministry’, and
Hume goes ‘along with him’; some 15 years
later in 1763 Hertford is ‘resolved never to
see, or do business with his Secretary, and
therefore desired [Hume] should attend him’
(LDH 1.111, 62; 1.421, 228). In March 1767
Hume is ‘deeply immersed in study’, when
Hertford surprisingly urged him to ‘accept of
2/9/2012 9:39:44 PM
 HUME’S LIFE
the office of Depute-Secretary of State under
his brother’. He cannot refuse and sees him-
self ‘embarked for some time in state affairs’.
Yet, he says, ‘I foresaw also that a place was
offered me of credit and confidence; that it
connected me with General Conway’ (LDH
2.123, 374; also see 1.511, 282). Hume says
he feels like a ‘banished man in a strange
country’. He is not ‘hurry’d with Business’
and commonly attends on the Secretary
‘from ten to three’. He has ‘no more Business
than would be requisite for [his] Amusement’
in London (LDH 2.123, 374; 2.127, 377).
Hume is not only employed in ‘cypher-
ing and decyphering’: during his public
activities he does not forget his opinions.
When the burden of diplomatic work at the
embassy in Paris is falling entirely upon him
(LDH 511–12, 282–3), and friends start call-
ing him ‘a man of Business’ (LDH 1.421n,
228),22 he saves from prison and death an
Englishman who attempted to kill himself
in the Seine. Marischal Keith congratulates
Hume: ‘you have done many good works in
your Ministerial functions, I am sure it was
one to save a pour fellow from the gallows,
who chose rather to drown than starve’.23
And Diderot has the complete story:
They fished him out alive. They brought
him to the Grand Châtelet, and the
Ambassador had to interpose his author-
ity to prevent them from putting him to
death. Some days ago Mr. Hume told us
that no political negotiation had been
more intriguing than this affaire and that
he had been obliged to go twenty times to
see the first president before he could make
him understand that there was no article,
in any of the treaties between France and
England, that forbade an Englishman
from drowning himself in the Seine under
penalty of being hanged. And he added
that, if his compatriot had unfortunately
26
9780826443595_Ch01_Final_txt_print.indd 26
been jailed, he would have risked igno-
miniously losing his life for having or not
having drowned himself.24
It was the time of extravagant requests, like
that of the ‘Apulien Philosopher’ Vincenzo
Maria Gaudio (1722–74). In January 1764
he wrote to Hume asking him two questions
‘for the good of human kind’: ‘How many
and which physical and moral causes pro-
duce the variety and contrariety of opinions
among men?’; ‘How to reduce the sum of
evils and increase that of goods?’.25
When he is Under-Secretary in London,
‘degenerated into a petty Statesman’ and
‘entirely occupyed in Politics’ (LDH 2.128,
379), he meets another extravagant case: ‘one
Giraldi, an Italian Physician’. Giraldi, who
is in London and needs protection in Italy,
addresses himself to Hume; Hume reporting
to Lord Shelburne:
[He] seems to me a man of sense and
learning, and whose orthodoxy has
of consequence been brought under
great and I suppose just suspicions. . . .
It seems a Cardinal, in his absence, fell
in love with his wife, and has taken
her into keeping; and on the physician
expressing some displeasure at this treat-
ment, his Eminence, who has great credit
in the Holy See, has threatend to have
him put into the Inquisition . . . . He has
addressed himself to me, on the suppos-
ition, no doubt, that I woud sympathize
with his cause. I conjure therefore your
Lordship, if there be any virtue, if there
be any praise, if there be anything comely
or of good report, to save the poor her-
etic from the flames . . . his case wou’d
puzzle Rhadamanthus himself: as a cuck-
old, he ought to go to heaven; as a her-
etic to hell. But, without joking, his case
is worthy of compassion; and I recom-
mend it to your Lordship’s humanity.26
2/9/2012 9:39:44 PM

Some ‘fresh intelligence’ discovers to Hume
that Giraldi ‘lives in intimacy with Gemino,
no great sign of his orthodoxy’. His project
was to retire to the Island of Capri, which
Giraldi ‘represents as an earthly paradise’,
and – Hume concludes – ‘indeed the only
paradise he ever expects to go to’.27
5. MY OWN FORTUNE
‘Money – says the Concise Account – is not
universally the chief object with mankind;
vanity weighs farther with some men’.28 Not
entirely exempt from vanity, Hume never
abandons the money that belongs to him
‘of right’, like the quarter salary from the
Annandale Estate and the half-pay military
pension from the Treasury: after more than
15 years he is still fighting for it. But he is
also ready to retract his application at the
Advocates’ Library, retain the office and give
a friend a bond of annuity for the salary. In
1747 he calls himself ‘a good Oeconomist’
(NLH 26, 10). Riches are valuable ‘at all
times, and to all men’ (E 276), and in his
short Life he spends some words celebrating
his income.
He says he was ‘of a good family’, but ‘not
rich’. As a younger brother, his ‘fortune’ was
‘very slender’ and therefore unsuitable to his
literary plan of life. So he laid down a rule:
‘to make a very rigid frugality supply my
deficiency of fortune’. In 1745 his Annandale
appointments made a ‘considerable accession
to [his] small fortune’; in 1746–8 the St Clair
appointments earned him a ‘fortune’ that he
calls ‘independent’. He wanted ‘to maintain
unimpaired [his] independency’ and he is
now ‘master of near a thousand pounds’. In
the 1750s the Faculty of Advocates gave him
‘little or no emolument’ but the ‘command
27
9780826443595_Ch01_Final_txt_print.indd 27
HUME’S LIFE
of a large library’. At the beginning of the
1760s ‘the copy-money given [him] by the
booksellers, much exceeded any thing for-
merly known in England’ and Hume is ‘not
only independent, but opulent’. In 1766 the
Parisian Secretary returned to Edinburgh,
‘not richer, but with much more money, and
a much larger income’ than he left it. He
was now ‘desirous of trying what superfluity
could produce’. In 1769 the London Under-
Secretary returned to Edinburgh ‘very opu-
lent’ (he ‘possessed a revenue of 1000 l. a
year’) and with the double prospect of long
enjoying his ‘ease’ and of seeing the increase
of his ‘reputation’ (MOL, LDH 1.1–6).
Thanks to Hertford’s family he really was, as
he once wrote from Paris, ‘in the high Road
to Riches’ and ‘in the high Road to Dignities’
(NLH 78, 38; LDH 1.421, 228).
6. STRIKE OUT STERNE:
FASHION IN PARIS
Hume was in Paris, Reims and La Flèche
in the 1730s, Paris in 1748 and Paris again
in the 1760s. He constantly saw himself
through the French looking-glass: the first
philosophical readings and the successful
French translations of his writings (in 1761
the Essais Philosophiques earn themselves a
place in the Index Librorum Prohibitorum),
the Embassy, the Court, the Great Ladies
(Madame de Boufflers) and the Philosophes
(Rousseau). In 1745 Hume first expresses
the slightly melancholy intention of retiring
to the South of France (NLH 17, 7). In the
Life he remembers living in Paris as a ‘real
satisfaction’: ‘I thought once of settling there
for life’ (MOL, LDH 1.6). Everyone affects
to consider him ‘one of the greatest geniuses
in the world’ (LDH 1.410, 223), since in
2/9/2012 9:39:44 PM
 HUME’S LIFE
Paris, unlike London, a man of letters ‘meets
immediatly with Regard & Attention’ (LDH
1.497, 272).
‘Anglomania was the manner of the place,’
Charlemont observes, and ‘Hume’s Fashion’
was ‘truely rediculous’: ‘no Lady’s Toilet
was compleat without Hume’s attendante.’
Walpole is more concise: ‘Mr. Hume is fash-
ion itself.’29 Indeed, he was more celebrated
for his name rather than his writings, for
his economical, historical and anti-religious
writings rather than his philosophical opin-
ions, and for his general opinions instead
of his precise arguments. The French mode
entailed ‘excessive civilities’ (MOL, LDH
1.6), but what was ‘at first oppressive’ in
two months ultimately sat ‘more easy’, espe-
cially as he gradually recovered the ‘facility’
of speaking the language (LDH 1.417, 225;
1.414, 224; 1.498–9, 272).30 The Life sums
up: ‘Those who have not seen the strange
effects of modes, will never imagine the
reception I met with at Paris’. And Hume
reports, in a remark that he was later to
strike out, that ‘Dr Sterne told me, that he
saw I was [celebrated in town] in the same
manner that he himself had been in London:
But he added, that his Vogue lasted only one
Winter’ (MOL, LDH 1.6).31
In 1762 Sterne does not worship the
French goddesses, but, he says, he has ‘con-
verted many unto Shandeism’. In 1764 he
preached a sermon deemed ‘offensive’ (he
calls it ‘innocent’) at the Embassy Chapel.
Hertford has just furnished the new and
‘magnificent’ Hôtel de Brancas, which gave
‘the subject of conversation to the polite
circles of Paris, for a fortnight at least’.32
Sterne preaches on the Book of Kings and
Hezekiah, who foolishly showed all the pre-
cious things that were in his house; even his
wives and concubines, adds the preacher.
Behind ‘urbanity or the etiquette of courts’,
28
9780826443595_Ch01_Final_txt_print.indd 28
Sterne unveils Hezekiah-Hertford’s ‘vanity’
and ‘ostentation’. Later on, at Hertford’s
table, Sterne had a dispute with Hume (a ‘lit-
tle pleasant sparring’, he says). In his sermon
Sterne had celebrated integrity and miracles,
and blamed pride and hypocrisy. At dinner
‘David was disposed to make a little merry
with the Parson; and, in return, the Parson
was equally disposed to make a little mirth
with the Infidel’. Sterne concludes: ‘it is this
amiable turn of character, that has given more
consequence and force to his scepticism, than
all the arguments of his sophistry.’33
At the end of 1765 Sterne publishes his
Sermons with a probably less ‘unlucky’ and
offensive version of ‘Hezekiah’. He is ready to
‘quarrel’ with Hume by calling him a ‘deist’,
if he will not add his name to the ‘most splen-
did list’ of subscribers. The Sermons came
out, but Hume’s name was not in the list. In
1767 Hume recalls the ‘usual extravagance’
of Sterne’s productions (NLH 160, 80), and
in the Sentimental Journey Sterne plays with
Hume the historian, his ‘excellent heart’
and bad knowledge of French. When Sterne
dies, Hume subscribes five guineas for his
widow.34 In 1773 Hume detects in Brydone’s
Tour through Sicily and Malta ‘some Levities,
too much in the Shandean Style’, which he
advises the author to ‘obliterate’. He also says
that Tristram Shandy is ‘the best Book, that
has been writ by any Englishman these thirty
Years . . . , bad as it is’ (LDH 2.269, 482).
Three years later he first writes and then
strikes Sterne’s name out of his Life.
7. LIFELONG LUCIAN AND
THE IRISH SKYTHS
‘Lucien est votre auteur favori, et . . . je
l’aime bien autant que vous’ (‘Lucian is
2/9/2012 9:39:44 PM

your favourite author, and . . . I love him as
much as you do’), Morellet reminds Hume
in 1766.35 Lucian follows Hume through-
out his literary career. In 1742 he allows that
‘some Dialogues’ of Lucian are among the
few excellent pieces of pleasantry in ancient
literature (E 134). The explosion of Lucian
occurs in the second half of the 1740s. ‘The
Sceptic’ (1753 version) suggests that we can
improve our mental disposition by reading
the ‘entertaining moralists’ and engaging
with the ‘imagination of Lucian’, if nature
has endowed us with a ‘favourable’ tem-
per (E 179n). Moreover, the moral Enquiry
assesses Lucian as ‘licentious with regard to
pleasure’ but a ‘very moral writer’ in other
respects, and accordingly regards it as highly
significant that he ‘cannot, sometimes, talk
of virtue, so much boasted, without betray-
ing symptoms of spleen and irony’. In Great
Britain, adds Lucianic Hume, such a ‘contin-
ued ostentation’ of public spirit and benevo-
lence inclines men of the world ‘to discover a
sullen incredulity on the head of those moral
endowments’ (EPM 6.21 / 242).
In the first Enquiry, where he laments the
‘harsh winds of calumny and persecution’
directed against philosophy, Hume bitterly
observes: ‘it does not always happen, that
every Alexander meets with a Lucian, ready to
expose and detect his impostures’ (EHU 11.2 /
132–3; EHU 10.23 / 121). In all antiquity, says
the ‘Populousness of ancient nations’, there
is not a philosopher ‘less superstitious’ than
Lucian (and Cicero). The ‘agreeable’ Lucian,
says the Natural History, had ‘employed the
whole force of his wit and satire against the
national religion’ (E 463n; NHR 12.174).
Morellet is not the only translator of
Lucian with whom Hume was acquainted.
In Turin he met Edward Murphy (1707–77).
Murphy’s repeated ‘grand query’ to Hume
concerns a cypher he invented, and the use of
29
9780826443595_Ch01_Final_txt_print.indd 29
HUME’S LIFE
cyphers was part of Hume’s official duties in
both Turin and Paris. Murphy was also the
editor of The Select Dialogues of Lucian, first
printed in 1744. In 1767 Hume compares
him to the ‘Royal philosopher Anacharsis’.
Murphy usually calls himself ‘Ô Murraghoo
Rex’, Anacharsis is one of Lucian’s dialogues
and the name of a character in Scytha sive
hospes.
In 1765 Hume had refused to go to
Ireland with Hertford: the Dubliners and
the Londoners did not want the Scottish phi-
losopher to make such a visit. Hertford had
prepared him an apartment in the Castle
of Dublin, but Hume thought it ‘not worth
while’: ‘It is like Stepping out of Light into
Darkness to exchange Paris for Dublin’
(LDH, 1.514, 285). In Ireland the philoso-
pher and historian was ‘excessively disliked’.
It will be ‘an Age or two at least’ before the
Irish can perceive his doctrines, and ‘perhaps
an age or two more’ before they can relish
them, writes Chaplain Trail: ‘I could almost
as soon promise Antichrist himself a welcome
Reception.’36
Possibly alluding to Hume’s account of
the ‘most barbarous’ cruelty allegedly perpe-
trated by the Irish during the ‘universal mas-
sacre’ of the English in 1641, where ‘[n]o age,
no sex, no condition was spared’ (H 5.55,
341), Murphy says in June 1767 that Hume
considers the Irish ‘Savages’ because they ‘eat
Human Flesh when [they] can get it good’.
The native Irish, adds Murphy, are ‘provd by
History to be Scythians by Descent, or rather
. . . Skyths, which word has been corrupted
into Scots’.37 In a swift Lucianic style Murphy
invites Hume to Ireland, ensuring him he will
be treated ‘as safe, as kindly . . . as ever [he]
was in Paris, or Edenburgh’:
We do not devour Strangers who visit us
as Friends; not even such as, we know,
2/9/2012 9:39:44 PM
 HUME’S LIFE
come to rob us. If you will not send me
a good Answer to my Query, I will go at
the Head of my Mighty Men and extir-
pate your Nation. Yea, we would eat you
all up, Man, Woman, and Child; but that
we are a little nice about the Quality of
our Fleshmeat. I never saw a Piece of a
Scotchman (though we have a whole
Province stock’d with them) at any gen-
teel Table here.38
The pretended ‘Reverend Murphy’, who in
Rome bought a papal plenary indulgence
for three crowns and made some remarks on
‘the pope and his fellow jugglers’, denounces
Lucian’s ‘entire want of Candour, while he
talks against the Christian Religion’, yet,
he adds, ‘it is impossible not to admire his
matchless Abilities’.39 Murphy’s translation
of Lucian seems to have been a (Greek) text-
book at Trinity.40 Three months before dying
the ageing Hume invites his nephew David
to read Lucian ‘sometimes’ and not to forget
his Greek, to mix the volumes of ‘taste and
imagination’ with ‘more serious reading’, as
the young Hume used to read at his pleasure
‘sometimes a Philosopher, sometimes a Poet’.41
8. HUME AND SMITH: A LIVING
SUMMER DIALOGUE
Lucian’s writings have something to say
about Hume’s death and legacy, the publica-
tion of Hume’s Life and Dialogues, and the
role taken by Adam Smith: an alternation
of intentions and second thoughts, trust and
worries, will and codicils. In 1773 Smith is
not well and leaves Hume ‘all [his] literary
papers’ and the publication of the ‘history of
the Astronomical Systems’;42 on 4 January
1776 Hume is seriously sick and leaves Smith
30
9780826443595_Ch01_Final_txt_print.indd 30
‘all [his] manuscripts . . . desiring him to
publish [the] Dialogues’. Hume reserves him-
self the power to alter his will at any time,
‘even in death-bed’.43 And so he does. He first
leaves ‘entirely’ to Smith’s discretion ‘at what
time . . . , or whether’ to publish the Dialogues
(LDH 2.316–18, 522–522A). Smith accord-
ingly thinks he has persuaded Hume to
allow him to leave the Dialogues concerning
Natural Religion unpublished if he views that
as advisable. Hume feels that his own death
is imminent, and begins to think in terms of
printing a ‘small edition’ and giving his edi-
tor Strahan the ‘literary property’ (LDH
2.323, 525).44 In a first codicil he makes
Strahan ‘entirely Master’ of his manuscripts
(LDH 2.325, 527): the Dialogues must be
‘printed and published any time within two
Years after [his] Death’, and the Life ‘pre-
fixed to the first Edition of [his] Works’.45 In
a second codicil he ordains: ‘if my Dialogues
. . . be not publisht within two Years and a
half after my Death, as also the Account
of my Life, the Property shall return to my
Nephew’.46 Smith criticizes this ‘unnecessary
clause’,47 and Hume leaves Smith a ‘security’
copy (LDH 2.334, 538). Finally, two days
before dying, he informs Smith and Strahan
that he is leaving his nephew the ‘property of
the Manuscript in case by any accident [to
Strahan’s Life] it should not be published
within three years after [his] decease’ (LDH
2.336, 540).
Smith is trying to move away from the
Dialogues: ‘If you give me leave I will add a
few lines to your account of your own life.’
It would make ‘no disagreeable part of the
history’ to relate Hume’s ‘want of an excuse
to make to Charon, the excuse you at last
thought of, and the very bad reception which
Charon was likely to give it’. Inspired by a
preceding letter to Wedderburn, Smith wants
to celebrate Hume’s ‘steady cheerfulness’
2/9/2012 9:39:44 PM

towards death.48 In the January will Hume
had asked Smith to take the pains of ‘correct-
ing and publishing’ the Dialogues; now Smith
offers to publish an Addition to Hume’s Life
and to ‘correct the Sheets of the new edition
of [his] works’, the Dialogues excepted.49
Hume answers Smith that he is ‘too good in
thinking any trifles that concern [him] are so
much worth of [his] attention’ and gives him
‘entirely liberty to make what Additions [he]
please[s] to the account of [his] Life’ (LDH
2.336, 540). Strahan tells Smith that the
Addition will be ‘highly proper’, like ‘every
particular respecting that great and good
man’.50 At the beginning of October Smith’s
Addition is ready: ‘I think there is a propri-
ety in addressing it as a letter to Mr Strahan
to whom [Hume] has left the care of his
works.’51 Acknowledging that Smith’s narra-
tive is the consequence of his ‘request’ and
Hume’s ‘approbation’ of it, Hume’s brother
‘much’ approves it and suggests a few altera-
tions, consistent with the ‘short and simple
a manner’ of the Life.52 Smith adopts his
remarks and sends Strahan his ‘small addi-
tion’ to Hume’s ‘small piece’ (LDH 2.318,
522A; 2.323, 525), and Strahan likes it
‘exceedingly’.53 In 1777, contrary to Hume’s
dispositions, Strahan publishes the Life and
the Addition ‘separately’ from Hume’s work.
The Dialogues are published in 1779 (three
years after his death) by Hume’s nephew,
possibly with some help from Strahan.
According to friends and doctors, in his last
months Hume ‘amuses himself’ with reading
(classic) ‘amuseing books’.54 And in his last
days he is reading ‘the dialogues of Lucian’,
the ‘Dialogues of the Dead’ and ‘the dialogue
entitled Kataplu’. On 8 August 1776 Hume
has a Lucianic conversation with himself. By
mixing up Megaphentes’ excuses (‘Cataplus’)
with Socrates’ final resistance (Dialogues of
the Dead), he invents ‘several jocular excuses’
31
9780826443595_Ch01_Final_txt_print.indd 31
HUME’S LIFE
to obtain a delay from Charon.55 As a writer
very busy in ‘correcting [his] works for a new
edition’, Hume asks for ‘a little time’ to see
‘how the public receives the alterations’; as
an anti-religious writer ‘endeavouring to
open the eyes of people’ he asks for ‘a few
years longer’ to have ‘the satisfaction of
seeing the downfall of some of the prevail-
ing systems of superstition’. Charon is not
convinced, Hume wants ‘a lease for so long
a term’: ‘there will be no end’ of correcting
and opening people’s eyes will take ‘many
hundred years’.56 By the way, Lucian – says
the Enquiry – ‘entirely opened the eyes of
mankind’ (EHU 10.23 / 121).
Hume, says Smith, ‘diverts himself with
correcting his own works’, he makes ‘many
very proper corrections, chiefly in what con-
cerns the language’, and Smith is ready to
‘revise the sheets and Authenticate its being
according to his last corrections’, as he has
‘promised’ to Hume.57 In the Addition Smith
prepares the excuse of correcting by telling
the reader that Hume ‘continue[s] to divert
himself, as usual, with correcting his own
works for a new edition’.58 This agrees with
Hume’s self-representation as an author
extremely ‘anxious of Correctness’ (LDH
1.175, 82; 2.304, 511; see ibid. 1.38, 16;
1.175, 82; 2.239, 455). ‘There is no End of
correcting’, proclaims Hume in 1763, and in
1771, ‘an Author may correct his works, as
long as he lives’ (LDH 1.379, 203; 2.246–7,
457). As he tells the printer, ‘I am perhaps the
only Author . . . who gratutiously employ’d
great Industry in correcting a Work, of which
he has fully alienated the Property’ (LDH
2.239, 455).
Hume, who finds ‘curious’ that an author
could have no patience to read over his pub-
lished works (NLH 62, 31; LDH 2.31, 314),
recalls in the 1770s a ‘saying’ of Jean-Baptiste
Rousseau the poet and offers his own version:
2/9/2012 9:39:44 PM
 HUME’S LIFE
‘a man might spend his whole Life in cor-
recting one small Volume, and yet have inac-
curacies in it’ (LDH 2.243, 456; 2.304, 511;
2.250, 461).59 In 1742 he makes a mistake in
quoting the aforementioned Rousseau (‘C’est
la politesse d’un Suisse / Dans la Hollande
civilisé’), and six years later he readily cor-
rects it (‘En Hollande civilisé’).60 Most of
his corrections ‘fall upon the Style’, he easily
allows, and locating and eliminating inaccur-
acies gives him a ‘sensible pleasure’ (LDH
2.151, 394; 2.188, 422; 2.243, 456; 2.250,
461). At the end of July 1776 he employs
himself in correcting the sixth volume of
the History and the moral Enquiry. It is the
trifling occupation of a dying author:
You will wonder, that, in my present
Situation I employ myself about such
Trifles, and you may compare me
to the modern Greeks, who, while
Constantinople was besieged by the
Turks and they themselves were threat-
ened with total Destruction, occupyed
themselves entirely in Disputes concern-
ing the Procession of the holy Ghost.
Such is the Effect of long Habit! (LDH
2.329, 531)
Strahan takes these lines as a ‘living Evidence’
that we are ‘much interested in what is to
pass after our Deaths’.61 In his last letter
to Strahan, concerning the moral Enquiry,
Hume announces: ‘This is the last Correction
I shall probably trouble you with: . . . all shall
be over with me in a very little time.’ (LDH
2.331–32, 537) Hume’s excuses in Smith’s
Addition sound like a little pleasantry about
his own final remarks:
I have made [my new edition] extremely
correct; . . . if I were to live twenty Years
longer, I shoud never be able to give it
32
9780826443595_Ch01_Final_txt_print.indd 32
any further Improvements. This is some
small Satisfaction to me in my present
Situation; . . . it is almost the only one
that my Writings ever afforded me: For as
to any suitable Returns of Approbation
from the Public . . . they are yet to come
(LDH 2.322, 525).
In his Lucianic dialogue Hume is concerned
with the effect of his writings on religion. In
his Addition Smith tries to soften the effect
of this imaginary conversation with the
mythical ferryman: he moderates the lan-
guage and introduces the excuse that Hume
desires yet a further opportunity to correct
his manuscripts. According to Smith’s letter
to Wedderburn, Hume is dying with ‘great
chearfulness . . . and more real resignation
to the necessary course of things, than any
Whining Christian ever dyed with pretended
resignation to the will of God’; according
to the Addition, he is simply dying with ‘the
utmost cheerfulness, and the most perfect
complacency and resignation’.62 Hume’s
‘pleasure of seeing the churches shut up, and
the Clergy sent about their business’ is turned
into that of ‘seeing the downfall of some of
the prevailing systems of superstition’.63
Hume’s brother remarks that Hume did
not say ‘I am dying as fast as my worst
enemies could wish’, as Smith writes in
the first draft of the Addition, but ‘as my
enemies, if I have any, could wish’.64 Smith
accepts the correction and adds that Hume
has ‘no enemies upon whom he wished to
revenge himself’.65 This fits with My Own
Life, where Hume (ironically and by a
marginal addition) declares that he is ‘little
susceptible of enmity’ and has never been
attacked by the ‘baleful tooth’ of ‘calumny’,
even though he ‘wantonly’ exposed himself
to ‘the rage of both civil and religious fac-
tions’ (MOL, LDH 1.7). The Life sounds
2/9/2012 9:39:45 PM

like the Manifesto of the Humean Party:
‘English, Scotch, and Irish, Whig and Tory,
churchman and sectary, freethinker and reli-
gionist, patriot and courtier, united in their
rage against [me]’ (MOL, LDH 1.4; also see
NLH 194, 107). Before publishing the third
Book of the Treatise, Hume had already
realized that the clergy are ‘always Enemys
to Innovations in Philosophy’, and only after
the publication of the History does he rather
belatedly resolve ‘to be more cautious than
formerly in creating myself Enemies’ (LDH
1.37, 15; 1.352, 189).66
As Strahan realizes and Smith endeav-
ours to forget, Hume shows an ‘extreme
solicitude’ to publish the Dialogues.67 He has
been accused of being ‘as great an atheist as
Bolingbroke’, even though Bolingbroke’s vol-
umes contain ‘so little Variety & Instruction’
and the clergy have ‘no Reason’ to be
‘enrag’d against him’ (LDH 1.168, 78; 1.209,
101; 1.214, 105). Yet, for the sake of his
Dialogues, Hume does not scruple to com-
pare himself to Bolingbroke. And he tells both
Smith and Strahan that Bolingbroke’s editor
was not ‘any wise hurt by his Publication’,
as ‘he always justify’d himself by his sacred
Regard to the Will of a dead Friend’. If he
leaves Strahan the Dialogues ‘by Will’, his
‘executing the Desire of a dead Friend, will
render the publication still more excusable’
(LDH 2.316, 522; 2.324, 525). According to
the last codicil, the ‘duty’ of his nephew in
publishing the Dialogues ‘as the last Request
of his Uncle, must be approved of by all the
World’.68
In the 1750s Hume is extracting from the
classics ‘what serv’d’ concerning the ancients
(LDH 1.152–3, 71): from Lucian, he says,
we may ‘gather’ that ‘leaving great sums of
money to friends’ was a common practice
in Greece and Rome (E 400n), even though
the Dialogues of the Dead, like Jonson’s
33
9780826443595_Ch01_Final_txt_print.indd 33
HUME’S LIFE
Volpone, talk of childless rich old men and
parasitical clients rather than real friends:
perfect friendship does not contemplate
money, as Lucian remembers in ‘Toxaris’
and Montaigne in ‘L’Amitié’. In Hume’s ‘A
Dialogue’, ‘by his Will’ Usbek makes his ‘inti-
mate Friend’ Alcheic ‘his Heir to a consider-
able Part of his Fortune’ (EPM Dial. 8 / 326).
By his own will Hume leaves 200 pounds
sterling to his ‘friends’ Ferguson, d’Alembert
and Smith. D’Alembert accepts the legacy.
He had helped Hume with the publication of
the original French version of A Concise and
Genuine Account of the Dispute between
Mr. Hume and Mr. Rousseau, and Hume
had even asked Smith to tell d’Alembert he
had made him ‘absolute Master to retrench
or alter what he thinks proper’ (LDH 2.83,
348).69 Smith did not help Hume with the
correction and publication of the Dialogues:
he discharges the legacy because he cannot
‘with honour accept it’.70 Hume had ‘hith-
erto forborne’ to publish the Dialogues,
desiring to ‘keep remote from all Clamour’
(LDH 2.323, 525); Smith is ‘uneasy about
the clamour’ they could excite. Worried that
they could ruin his tranquillity, he tries to
make Strahan apprehensive that they could
ruin his interest.71 Hume assures Smith that
the Life is a ‘very inoffensive’ piece (when
compared to the Dialogues), which ‘will
be thought curious and entertaining’ (LDH
2.318, 522A; 2.323, 525); Smith agrees, and
is later offended by the reception accorded
his own Addition: ‘a single, and as I thought
a very harmless Sheet of paper, . . . concern-
ing the death of our late friend Mr Hume,
brought upon me ten times more abuse
than the very violent attack I had made
upon the whole commercial system of Great
Britain’.72
In 1747 Hume says he could not see ‘what
bad consequences follow, in the present
2/9/2012 9:39:45 PM
 HUME’S LIFE
age, from the character of an infidel’ (LDH
1.106, 58); 30 years afterwards Horne vitu-
peratively attacks the ‘foolish and insens-
ible’ Hume for reading Lucian and inventing
droll conversations with Charon. Hume’s
excuses are an attack on ‘superstition’, and
‘we all know . . . against what Religion his
shafts are levelled, under that name’. Smith
wished that the Addition could be agree-
able; Horne thinks that he who can reflect
‘with complacency’ on Hume, amusing him-
self at his death ‘may smile over Babylon in
ruins’.73
Hume was reading Lucian. ‘It is an idle
thing in us,’ he writes in June, 1776, ‘to be
concerned about any thing that shall hap-
pen after our Death’; yet, he adds, ‘I often
regretted that a Piece, for which I had a par-
ticular Partiality, should run any hazard of
being suppressed after my Decease’ (LDH
2.325–6, 527). Again Strahan takes it as a
sign that ‘our Existence will be protracted
beyond this life’.74 With regard to the
Enquiry, in 1754 Hume had allowed: ‘I am
willing to be instructed by the Public; tho’
human Life is so short that I despair of ever
seeing the Decision’; in 1776 he laments: ‘I
shall not live to see any Justice done to me’
(LDH 1.187, 91; 2.322, 525). ‘If I live a few
Years longer,’ Hume tells Smith, ‘I shall pub-
lish [the Dialogues] myself’ (LDH 2.316,
522); and Smith tells Strahan: ‘I could have
wished [they] had remained in Manuscript
to be communicated only to a few people
. . . [they] never should have been published
in my lifetime.’ In 1763 Elliot and Blair had
expressed the same sentiments.75
In his last days Hume was ‘revising’ the
Dialogues (LDH 2. 334, 538), but despite these
exertions on the part of a dying man his good
friend Smith remained determined to have
nothing to do with the publication of what he
saw as so incendiary a piece of writing. Smith,
34
9780826443595_Ch01_Final_txt_print.indd 34
says Strahan, ‘knew [Hume] well, and loved
him much’, and it ultimately falls to Smith
to tell Strahan and the world of the death of
‘our most excellent, and never to be forgot-
ten friend’. After 11 years he also unforgiv-
ably celebrates ‘the abilities and virtues of the
never to be forgotten Dr Hutcheson’.76
9. STUCK IN A BOG (FISHWOMEN
FOR THEOLOGIANS)
There is an anecdote about one of Hume’s
misadventures in Edinburgh, which much
amused the historian Sir Leslie Stephen and
his daughter Virginia Woolf, and reminds
us of Thales’ tumble and the laughter of the
Thracian maid.77 It agrees with the pagan
frolic tone of the Natural History and says
something about Hume’s style, his attitude
towards religion, and his not-so-serious pro-
nouncements on women, monks and super-
stition (NHR 12.144). It also says something
about the distinction between popular and
pretended philosophical religion, and some
Humean resolutions to ‘keep [him]self in
a proper disposition for saying the Lord’s
Prayer, whenever [he] shall find space enough
for it’ and to ‘proceed directly to attack the
Lord’s Prayer & the ten Commandments &
the single Cat’ (LDH 1.148, 70; NLH 43,
25). Finally, it says something about his real
expectations as regards the effect of his work:
‘this did not convert the generality of man-
kind from so absurd a faith; for when will
the people be reasonable?’ (NHR 12.173).
Hume is engaged in building a ‘small
House, I mean [he says] a large House for an
Author’ (he did finish it), which is the ‘second
great Operation of human Life’ (LDH 2.232,
451; 2.235, 453). As always, he is out of
the common road:
2/9/2012 9:39:45 PM

On his daily visits to inspect the work,
he was in the habit of taking a short
cut across what was then a swamp . . .
he made a slip, fell over, and stuck fast
in the bog. Observing some Newhaven
fishwomen passing with their ‘creels’, he
called aloud to them for help; but, when
they came up, and recognised the wicked
unbeliever David Hume, they refused
any assistance, unless he first repeated, in
a solemn tone, the Lord’s Prayer. This he
did, without pause or blunder, and was
extricated accordingly. He used to tell
this story himself with great glee, declar-
ing that the Edinburgh fishwives were
the most acute theologians he had ever
encountered.78
Even when getting out of his bog, the sceptic
will always be ‘the first to join in the laugh
against himself’ (EHU 12.23 / 160).
NOTES
1
Early Responses to Hume’s Life and Reputation,
2 vols, 2nd edn, ed. J. Fieser (Bristol: Thoemmes
Continuum, 2005), vol. 1, p. 294.
2
These alleged words are often treated as an
indication that Hume’s mother viewed him
as weak-minded; other interpreters main-
tain that the sense of the words in the local
‘Vernacular’ of the time was such that his
mother was drawing a contrast between
his good nature and his ‘uncommon acute-
ness’ (see H. Calderwood, David Hume
(Edinburgh: Oliphant Anderson & Ferrier,
1898), p. 14; D.F. Norton, ‘An Introduction to
Hume’s Thought’, in D.F. Norton (ed.), The
Cambridge Companion to Hume (Cambridge:
Cambridge University Press, 1993), pp.
1–32, p. 2; E. Mazza, ‘La mamma di Hume.
Interpretazioni di un detto apocrifo’, in Il
mestiere di studiare e insegnare filosofia
(Milano: Wise, 2000), pp. 93–152).
3
J.H. Burton, Life and Correspondence of
David Hume, 2 vols (Edinburgh: W. Tait,
35
9780826443595_Ch01_Final_txt_print.indd 35
HUME’S LIFE
1846), vol. 1, p. 294n1; S.J. Pratt, Supplement
to the Life of David Hume (London: J. Bew,
1777), pp. 33–4.
4
Hume seems to reflect on La Rochefoucauld’s
maxim: ‘everybody speaks well of his own
heart, nobody dares to speak well of his own
understanding’ (La Rochefoucauld, Maximes
(Paris: Garnier, 1991), p. 309.
5
Letters of Eminent Persons addressed to
David Hume, ed. J.H. Burton (Edinburgh:
W. Blackwood, 1849), p. 252.
6
Burton, Life, vol. 1, p. 226.
7
J. Brown, An Estimate of the Manners and
Principles of the Times, 2nd edn (London: L.
David and C. Reymers, 1757), p. 57.
8
E.C. Mossner, ‘New Hume Letters to Lord
Elibank, 1748–1776’, Texas Studies in
Literature and Language 4 (1962), p. 445;
A Concise and Genuine Account of the
Dispute between Mr. Hume and Mr. Rousseau
(London: T. Becket and P.A. De Hondt, 1766),
pp. iii, iv, viii.
9
Horace Walpole’s Correspondence with
Madame Du Deffand and Wiart, ed. W.S.
Lewis and W.H. Smith (New Haven: Yale
University Press, 1970), pp. 253, 266.
10
BL, Add. MS 36638, f. 22r.
11
NLS, MS 25692, f. 33v.
12
E.C. Mossner, ‘Hume at La Flèche, 1735:
An Unpublished Letter’, University of Texas
Studies in English 37 (1958), pp. 30–3,
p. 32.
13
W. Leechman, The Temper, Character, and
Duty of a Minister of the Gospel, 5th edn
(Glasgow: R. Foulis, 1749), pp. 34, 41.
14
J.C. Hilson, ‘More Unpublished Letters of
David Hume’, Forum for Modern Language
Studies 6(4) (1970), pp. 315–26, p. 321.
15
LDH 1.131, 64; NLS, MS 25708, f. 36v; MS
25703, f. 187v.
16
NLS, MS 25703, ff. 188r, 212v.
17
B. Mandeville, The Fable of the Bees, ed.
P. Harth (London: Penguin, 1989), p. 341.
18
NLS, MS 25703, 188r; ff. 210r, 212v; Early
Responses to Hume’s Life and Reputation, ed.
J. Fieser, vol. 2, p. 203; Charlemont, ‘Anecdotes
of Hume’, RIA, Charlemont MSS 12.R.7,
f. 518v.
19
NLS, MS 25708, ff. 40v; LDH 1.132, 64.
20
NLS, MS 25708, ff. 46rv, 50v.
21
NLS, MS 25708, ff. 33r, 41v; E 2.7.15, 339;
2/9/2012 9:39:45 PM
 HUME’S LIFE

Home, J., ‘Diary of a journey with Hume to Archiwum Historii Filozofii I Mysli Spolecznej
Bath’, in Early Responses, vol. 1, p. 284; NLS, 9:5 (1963), p. 138; LDH 1.10, 1.
42
MS 25708, f. 33r. The Correspondence of Adam Smith, ed. E.C.
22
Letters of Eminent Persons, p. 165. Mossner and I.S. Ross (Indianapolis: Liberty,
23
Ibid., p. 67. 1987), p. 168.
24 43
D. Diderot, Correspondance V, ed. G. Roth, D. Hume, ‘Disposition and Settlement’, 4
trans. E. Mazza (Paris: Les Editions de Minuit, January 1776, in ‘Record of Testaments
1959), p. 132. for the year 1781’ (8 March 1781), NAS,
25
NLS, MS 23163, f. 100. CC8/8/125, ff. 863, 865.
26 44
NLS, Acc 11139, f. 24r. The Correspondence of Adam Smith, p. 211.
27 45
NLS, Acc 11139, f. 25r. D. Hume, ‘Codicil to my Will’ (7 August
28
A Concise and Genuine Account, p. 90. 1776) (1st codicil: ‘In my latter Will and
29
Charlemont, ‘Anecdotes of Hume’, f. 521; Disposition’; 2nd codicil: ‘I also ordain’), NLS,
Horace Walpole’s Correspondence with MS 23159, item 24, f. 17; see LDH 2.318,
Hannah More et al., ed. W.S. Lewis et al. 522A; 2.323, 525.
(New Haven: Yale University Press, 1961), 46
NLS, MS 23159, item 24, f. 17; LDH 2.323, 525.
pp. 47, 49. 47
The Correspondence of Adam Smith, pp. 199,
30
Mossner, ‘New Hume Letters to Lord Elibank’, 206, 211.
p. 452. 48
Ibid., pp. 203–4, 206.
31
NLS, MS 23159, item 23, f. 10. 49
Hume, ‘Disposition and Settlement’, f. 863;
32
L. Sterne, Letters, ed. L.P. Curtis (Oxford: The Correspondence of Adam Smith, p. 206;
Clarendon, 1967), pp. 157, 219. see ibid., pp. 203, 219.
33
Sterne, Letters, p. 218; L. Sterne, ‘The Case of 50
The Correspondence of Adam Smith, p. 212.
Hezekiah and the Messengers’, The Sermons of 51
Ibid., pp. 214, 216.
Mr. Yorick, vol. 3 (London: T. Becket and P.A. 52
Ibid., pp. 206, 208, 214–15.
De Hondt, 1766), pp. 30, 42. 53
Ibid., pp. 216, 221, 222.
34
Sterne, Letters, pp. 219, 239, 445 (also see 54
Ibid., pp. 207, 218; J. Home, ‘Diary of a
pp. ivi, 235, 254); L. Sterne, A Sentimental Journey with Hume to Bath’, p. 281.
Journey through France and Italy by Mr. 55
The Correspondence of Adam Smith, pp. 203,
Yorick (London: T. Becket and P.A. De Hondt, 219; Correspondence of Dr William Hunter,
1768), pp. 92–3. vol. 2, p. 226.
35
Lettres d’André Morellet. Tome I: 1759–1785, 56
The Correspondence of Adam Smith, p. 219.
ed. D. Medlin et al. (Oxford: The Voltaire 57
Ibid., pp. 204, 206, 211.
Foundation, 1991), p. 72. 58
Ibid., p. 218.
36
The Correspondence of Dr William Hunter, 59
J.-B. Rousseau, ‘Preface’ (1712) to Œuvres
1740–1783, ed. C.H. Brock, 2 vols (London: diverses, 2 vols (London: J. Tonson and J.
Pickering & Chatto, 2008), vol. 1, pp. 229, Watts, 1723), vol. 1, p. iv.
230–1. 60
E 127. The initial error and subsequent correc-
37
Letters of Eminent Persons, pp. 171–72; see tion are displayed in D. Hume, Essays, Moral
J. Curry, An Historical Review of the Civil and Political (Edinburgh: A. Kincaid, 1742),
Wars in Ireland, 2 vols (Dublin: L. White, p. 80, and Essays, Moral and Political, 3rd
1786), vol. 1, p. 215n. edn, corrected (London: A. Millar; Edinburgh:
38
NLS, MS 23156, f. 77. A. Kincaid, 1748), p. 177.
39
Dublin, Trinity College, 543/2/11 /E.2.19; 61
NLS, MS 23157, item 69, f. 294.
E. Murphy, The Select Dialogues of Lucian 62
The Correspondence of Adam Smith, pp. 203,
(London and Dublin: Edward Exshaw, 1746), 218 (see also p. 206).
pp. xi–xii. 63
Ibid., pp. 204, 219.
40
Notes and Queries 3 (8 April 1899), 64
Ibid., p. 215 (see Correspondence of
p. 262a. Dr William Hunter, vol. 2, p. 226).
41
T. Kozanecki, ‘Dawida Hume’A Nieznane Listy 65
The Correspondence of Adam Smith, pp. 215,
W Zbiorach Mezeum Czartoryskich (Polska)’, 219.

36

9780826443595_Ch01_Final_txt_print.indd 36 2/9/2012 9:39:45 PM

 HUME’S LIFE

66 75
See Burton, Life, vol. 1, p. 226. The Correspondence of Adam Smith, p. 211;
67
The Correspondence of Adam Smith, p. 212. LDH 1.380, 203; NLH 71, 35; 72, 36; NLS
68
NLS, MS 2319, item 24, f. 17. MS 23153, 51, f. 157.
69 76
Letters of Eminent Persons, p. 252; The Correspondence of Adam Smith, pp. 220, 309.
77
J.C. Hilson and J.V. Price, ‘Hume and L. Stephen, ‘Hume’, in S. Lee (ed.),
Friends, 1756 and 1766: Two New Letters’, Dictionary of National Biography (London:
The Yearbook of English Studies 7 (1977), Smith, Elder, 1891), vol. 28, p. 223b; V.
pp. 121–7, p. 126. Woolf, To the Lighthouse (1927) (London:
70
The Correspondence of Adam Smith, pp. 209, Grafton, 1988), pp. 62, 70; G.L. Strachey,
214, 210, 215. ‘Hume’ (1928), in Portraits in Miniature and
71
Ibid., pp. 211, 216–17. Other Essays (London: Chatto & Windus,
72
Ibid., p. 251. 1931), pp. 151–52.
73 78
G. Horne, A Letter to Adam Smith LL.D. On Selections from the Family Papers Preserved
the Life, Death and Philosophy of his Friend at Caldwell, pt II, vol. 2 (Glasgow: 1854),
David Hume (Oxford: Clarendon Press, 1777), pp. 177–8n1; see The Scotch Haggis
pp. 9–11, 12–13, 29–30. (Edinburgh: D. Webster and Son, 1822), p. 77;
74
NLS, MS 23157, item 69, f. 297. Burton, Life (1846), vol. 2, p. 458.

37

9780826443595_Ch01_Final_txt_print.indd 37 2/9/2012 9:39:46 PM

 2
HUME’S EMPIRICIST INNER
EPISTEMOLOGY: A REASSESSMENT
OF THE COPY PRINCIPLE
Tom Seppalainen and Angela Coventry
1. INTRODUCTION
Vivacity, the ‘liveliness’ of perceptions, is
central to Hume’s epistemology. Hume
equated belief with vivid ideas. Vivacity is a
conscious quality and so believable ideas are
felt to be lively. Hume’s empiricism revolves
around a phenomenological, inner epistem-
ology.1 Through copying, Hume bases viv-
acity in impressions. Sensory vivacity also
concerns liveliness or patterns of change.
Through learnt skilful use, vivacity tracks
change specific to intentional sense-percep-
tual experience consisting in Hume’s ‘coher-
ent and constant’ complex impressions.
Copying, in turn, communicates vivacity to
ideas where it becomes an indicator of the
believability of ideas. Hume’s copying con-
cerns then the causation of conscious skills
required for the identification of empiric-
ally warranted structures. Copying allows
Hume to combine a radically externalist
empiricism with a phenomenological inner
epistemology.
38
9780826443595_Ch02_Final_txt_print.indd 38
2. COPYING AND ITS TARGETS
The copy principle connects ideas to impres-
sions. The actual process is left without much
clarification by Hume. We know it is a causal
one and that impressions are causally prior
to ideas. We also know that causation holds
only for concrete particulars. These ideas,
together with Hume’s general scepticism
about the intelligibility of causal relations,
suggest that a lack of insight into the process
is to be expected. From Hume’s standpoint,
there are not many empirically founded con-
ceptual insights to share. Although such a
state of affairs is common for fundamental
posits of scientific theories including those
of psychology, it has caused much dismay in
philosophy and for a good reason. At stra-
tegic philosophical occasions, when Hume
challenges the intelligibility of commonly
accepted philosophical beliefs, he introduces
copying as a test or justificatory principle
(see, for example, EHU 1.13 / 13). Copying
is Hume’s distinct operationalization of
2/9/2012 9:39:33 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
empiricism; the paucity of information on
the process leaves Hume’s empiricism with
a ‘conceptual’ problem. This is the context
for our reassessment of the copy principle.
Despite the conceptual problem, copying is
the centerpiece of Hume’s empiricism. Why
this is so can be best explained in program-
matic terms.
The distinguishing features of the few
broad epistemological orientations are
clear.2 Empiricism embodies an externalist
explanatory strategy for whatever it seeks to
explain or clarify. In Hume’s case, the copy
relationship establishes an externalist intel-
lectual strategy for the cognitive realm of
ideas through the sensory realm of impres-
sions. Externalism differs from internalist
paradigms of epistemology such as rational-
ism and constructivism. These emphasize the
role of inherent, innate, or just a priori pos-
its for cognitive and behavioural outcomes.
Empiricism minimizes the significance of
such posits even if the mind is not construed
as a Lockean tabula rasa. Epistemological
internalism is not the only relevant contrast.
Externalism comes also in a form that cou-
ples the mind (or parts of it) to the world
or environment (or parts of it). This type
should be called ‘objectivism’. Hume’s ‘exter-
nalist’ empiricist epistemology concerns the
role of the sensory environment for acts of
cognition. Copying is a relation between two
mental domains and does not entail object-
ivism. In Hume’s case, objectivism follows
only if the sensory system is explained by
factors of the environment, ones external
to it. We leave open the issue of objectiv-
ism. An externalist hypothesis about Hume’s
empiricism leads to a straightforward heur-
istic for clarifying copying. Since the pro-
cess is unclear, we must focus on its targets.
39
9780826443595_Ch02_Final_txt_print.indd 39
Once we articulate the targets of copying, we
can form an exact judgement about Hume’s
externalist empiricism.
Prima facie, copying indicates externalism.
Copies share significant properties with their
originals. Think of scale models and photo-
copies. Both are results of copying and both
appear like their originals. The terminology
Hume uses indicates a commitment to this
strong externalist form of copying. Ideas for
him are ‘faint’ versions of impressions (THN
1.1.1.1 / 1; EHU 2.5 / 19). Copies can also
function like their originals and not just seem
or appear like them. Often they do both. This
point echoes through the history of debates
between internalists and externalists. One
philosopher’s empiricism is another’s nativ-
ism when it is not specified whether proper-
ties or processes or both are shared across
systems. The historical reminder is important
in the case of Hume. He uses potentially
misleading terminology according to which
‘perceptions of the mind’ are ‘objects.’ But
copying need not be limited to the passing of
objects and their properties to cognitions for
processing. All kinds of ‘entities’ can be cop-
ied, including functions or activity types. We
pose the question about the targets of copy-
ing in an open-minded way.
According to collective wisdom, copy-
ing indicates Hume’s ‘meaning empiricism’.
According to this ‘semantic hermeneutics’,
linguistic meaning is the target of copying.
The interpretation is codified in slogans:
copying gives the ‘empirical cash value’ of
terms.3 Hume’s concern with the meaning
of linguistic expressions has textual support.
He claims to apply the copy principle to
philosophical terms (EHU 2.9 / 22). Despite
this, the semantic hermeneutics distorts the
causal nature of copying. Copying becomes
2/9/2012 9:39:33 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
interpreted in analytic or definitional terms
such as operational definitions. This is not
Hume but empiricism of a much later ori-
gin. And semantic hermeneutics is not just
bad historiography. It has led to scathing cri-
tiques of Hume. Although we believe in his-
tory serving current concerns, these critiques
should by now be seen as vitiating any such
edifying function for the semantic hermen-
eutics. To us, they raise only one question.
Why study Hume if his central causal prin-
ciple is without merit, theoretical function,
or promise? For us, copying targets informa-
tion of the senses and information comes in
many forms, including semantic ones. Let us
start with Hume’s ideas about the targets of
copying that fit clearly under the heading of
information and then turn to the less obvious
informational targets.
Hume’s senses include atomic elements or
‘simple impressions’. Any particular impres-
sions of sight, smell, taste, sound, touch,
pleasure or pain, are copied to the realm of
ideas as simple ideas. But this is not all. Most
information is in a form more complex or
structured than atomic perceptions and this
is also the case for Hume. Both senses and
cognitions contain complex impressions and
ideas, respectively. Some of this structured
information of cognition owes its origin to
copying but not all of it. Hume does not say
whether all structured information in the
senses is copied to cognitions; it is clear that
‘cognitive’ processes of imagination generate
complex information that is novel relative to
the senses. It is novel since no such specific
complex structures exist in the senses – even
if the elemental information does. The idea
of a unicorn is the classic exemplar. So far
everything is clear and uncontroversial.
The copying of even some complex impres-
sions shows that copying also targets struc-
ture. Given that some complex impressions
40
9780826443595_Ch02_Final_txt_print.indd 40
are copied, senses and cognitions resemble in
terms of their structural features. This holds
even if no ‘abstract’ structure ever occurs in
the systems independently, without ‘objects’
or contentful complex information. Structure
need not be independent for it to be a psy-
chologically and epistemologically critical
factor. It only needs to occur robustly across
circumstances, have a cognitive function and
be knowable. The robust presence of particu-
lar complex impressions is enough for the
first condition. The latter two conditions will
be taken up next.
Copied structure opens up the possibility
for further targets of copying. Structure itself
has an origin. This may be similar for impres-
sions and ideas, at least under some circum-
stances. Generally, if complex structure is the
end result – effect – of some sort of activity –
its cause – such activity may itself be copied.
More specifically, if the structured informa-
tion in the senses is a function of activity,
Hume’s system allows for its copying. Since
Hume’s psychology is a ‘genetic’ or develop-
mental theory this is a genuine possibility.
We believe that it is more than a theoretical
possibility.
The fact that Hume never discussed the
copying of activities should not prejudice this
theoretical option. A direct treatment is pro-
hibited by Hume’s scepticism about the intel-
ligibility of all causal relations qua relations.
It applies to activity. For this reason, Hume
included in his psychological taxonomy an
inherently action-related quality, ‘vivacity’,
the ‘liveliness’ of perceptions. Hume’s syno-
nyms, ‘strength’, ‘violence’ and ‘vigour’, are
all action-related terms. To explore the copy-
ing of activity we must address the copying
of vivacity.
Hume never clearly mentions the copying
of vivacity. It is obvious in the light of his
methodology and theory. Targets of copying
2/9/2012 9:39:33 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
can be explored through a resemblance-
based interpretative methodology. This
coheres fully with Hume’s theory and use.
For Hume, ideas are not only caused by but
also resemble impressions (THN 1.1.1.6 / 4).
Copying results in a resemblance between
impressions and ideas as ‘objects’. Moreover,
if ideas are particular images, then ideas are
literal ontological copies. Copying can also
concern activities and still be operational-
ized in terms of resemblance. And finally,
copying is an associative causal process yet
unique in being based in the ‘natural’ rela-
tion of resemblance instead of contiguity.
Importantly, a resemblance methodology can
be used in a two-way fashion. Facts concern-
ing the realm of impressions can be used to
interpret the realm of ideas. Nothing prohib-
its using resemblance in reverse, in theorizing
and interpretation. Hume arguably did both
if only because ordinary consciousness is
ambiguous between the presence and contri-
bution of ideas and impressions. Let us apply
this methodology to vivacity.
Impressions and ideas resemble in terms of
vivacity in the way that ideas and impressions
generally resemble. Idea copies are ‘faded’
versions of impressions also with respect
to the quality of vivacity. Clearly, vivacity
is a target of copying. But is it not already
included in the general idea of a faded object-
copy? The question points at the information
carried by vivacity. That is of a kind no indi-
vidual ‘perceptual object’ can contain. Let us
see why.
For Hume, vivacity is a ‘quality’ of per-
ceptions. This does not mean that vivacity is
an ordinary ‘inherent’ property. It cannot be
transferred to the realm of ideas with a copied
object. As liveliness, ‘vivacity’ is a quality of
a ‘series’ of impressions and of the changes
between them. Slow change is not lively. As a
quality of change, liveliness has an indicator
41
9780826443595_Ch02_Final_txt_print.indd 41
role. Hume uses relative vivacity in psycho-
logical taxonomy to classify perceptions into
ideas and impressions, distinguish memory
from other ideas, and distinguish beliefs
from mere fictions of the imagination. Thus,
through vivacity, information about mental
activity acquires a fundamental taxonomic
role in Hume’s psychology. Even if we cannot
know activity itself, we can know its quality,
vivacity. Vivacity is the knowable indicator
of activity. But this is not all.
Hume equated the central epistemologi-
cal notion, belief, an idea type, with vivacity
(THN 1.3.9.15 / 106). This has caused much
puzzlement in the secondary literature.4
Much of it can be resolved by articulating
the informational nature of vivacity. ‘Belief’
is an ambiguous term and combines informa-
tion with attitude. When we believe some X
we ‘have’ the informational content X ‘in our
mind’. But we also ‘have’ a specific doxastic
attitude towards the content (THN 1.3.7.1
/ 94). ‘Belief’ is normally used for both. In
our analysis, ‘belief’ refers to the former and
‘believability’ to the latter, doxastic attitude.
For Hume, vivacity is a quality of perceptual
objects instead of the object as such. Thus,
vivacity should not be identified with belief
as such. Instead, vivacity concerns believ-
ability. Believability is based in the ‘lively’
or active dimension of ideas. Since vivacity
is copied, the doxastic attitude of believ-
ability is copied. The copying of both vivac-
ity, a quality of perceptual objects, and the
informational content of perceptual objects
couples belief-ideas tightly with the senses.
This is understandable for an empiricist.
For rationalists, ‘judgement’ stands behind
‘believability’ as an independent internal
epistemological process. In Hume’s external-
ism, both the doxastic attitude and (much of)
the contents of belief are copied to cognitions
from the realm of the senses.
2/9/2012 9:39:33 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
Above we dubbed vivacity an ‘indicator’
for psychological taxonomy. The idea applies
also to belief. The quality of vivacity is an
indicator of believability. Vivid ideas cause the
doxastic attitude. This shows that structured
information has a central cognitive role for
Hume. Through its origins, that role extends
to the activities responsible for structure. The
second condition for taking complex struc-
ture as a target of copying is satisfied. To
complete the sketch, we must address knowl-
edge of structure and its origins.
Vivacity is a feeling for Hume. This is not
surprising. All central posits of his mental
mechanism have a conscious dimension. Yet
vivacity is unique among these. It is a quali-
tative indicator of activity. Hume’s scepti-
cism about ‘relations themselves’ covers all
activities. It also spans all methods includ-
ing ‘inner’ ones. The evidentiary limits of
phenomenology for activities are severe and
comprehensive. No wonder Hume aspired
to be the Newton of the Mind (THN Intro.
7–8 / xvi–xvii). He professed no comprehen-
sive phenomenological epistemology for the
‘metaphysical’ nature of mental associative
relations that make the mind move; with
his ‘non fingo’, Newton professed the same
for all observational methodologies con-
cerning force or gravitational ‘associations’
that make objects with mass move. Vivacity
remains ‘only’ an indicator of activity. There
is a further limit. All perceptual objects are
vivid, yet, for Hume, not all of them are
believable. Vivacity is only ‘effects-based’
information about activities, yet sufficient
to discriminate among critical activity types.
Vivacity is a conscious indicator whose use
or ‘indicator function’ must be learned. It
is a skill of conscious activity or conscious
skill, for short. The idea of a learnt conscious
epistemic skill is best introduced through its
alternative, Cartesianism.
42
9780826443595_Ch02_Final_txt_print.indd 42
Descartes’s methodology of clear and dis-
tinct ideas is the famous ‘inner’ or phenom-
enological epistemology. Despite the ‘Rules
for the Direction of the Mind’,5 the conscious
skills of thinking needed to attain clear and
distinct ideas remain unsatisfactory. The
flavour of a pseudo-methodology, a mere
‘inner gaze’ without a clear learning history,
remains. The reason is in the innate nature of
Cartesian epistemologically appropriate ideas.
Descartes’s metaphysics of consciousness com-
plements the epistemology. Consciousness is
independently in all mental systems, govern-
ing and unifying them. Hume’s metaphysics
is externalist – ‘genetic’ or developmental and
his ‘inner’ epistemology complements it. His
emphasis on ‘experience’ and external causal
factors opens the possibility that epistemo-
logically relevant conscious skills are learnt
through a specific causal history.
Vivacity has a central role in the specific
causal history of skills of conscious activ-
ity. Tracking change in perceptions through
vivacity is of epistemological use only if
vivacity discriminates among different types
of change. More specifically, vivacity allows
for distinguishing beliefs from mere concep-
tions only if we are able to ‘read’ change.
How should this skill be conceptualized?
Copying makes it possible that we learn from
vivacity within the sensory realm and then
use it within the realm of cognition. So what
does sensory vivacity teach us?
Sensory vivacity also tracks change. One
‘type’ of sensory change has a clear cogni-
tive function for Hume. Complex impres-
sions that have the ‘qualities’ of coherence
and constancy are involved in the causation
of the belief in an external world (THN
1.4.2.15–20 / 194–5). And this belief-idea, in
turn, is conscious information about the exist-
ence of a world independent of us. Thus, only
some complex impressions through copying
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
partake in the causation of the intentionality
or ‘about-ness’ of ideas. Through their phe-
nomenological nature and this causal role,
coherence and constancy themselves are con-
scious sensory information about the world.
The skilful use of vivacity is learnt in this con-
text. The correlation between different sense-
perceptual outcomes, coherent and constant
complex impressions and their opposites, and
different profiles of sensory vivacity grounds
vivacity’s indicator function in the senses. The
conscious skill is learned in the senses. More
specifically, in the copied form that indicates
believability, vivacity informs about those
changes in ideas that are relevant for empir-
ically warranted believability. In the senses,
such changes cause sense-perceptual inten-
tionality. They create complex impressions
with constancy and coherence. In general,
Hume’s felt indicator of vivacity has then the
function to track mental actions relevant for
the creation of an empirical world, one with
structured stability. When copied to ideas, it
grounds beliefs and demarcates beliefs from
mere imaginings.
Let us take stock. Hume’s ideas and
impressions are systems that are causally
coupled by the copy-relation. Copying tar-
gets information in the sensory system. This
information is based on structure, elements,
properties and relations. Much of the struc-
ture is based in activities since much of it
is not hardwired or innate. Moreover, both
systems are mental. Mental activities are not
knowable but their effects are. Vivacity is a
conscious quality which tracks such effects
and profiles of change and it informs about
skilful mental activities when specific percep-
tual outcomes correlate with specific profiles
of change. This indicator of skilful mental
activity is copied to cognitions along with
the rest of the information in the senses. The
broad informational matrix ranging from
43
9780826443595_Ch02_Final_txt_print.indd 43
elements through structure to activities and
conscious qualities that track them consti-
tutes the targets of Humean copying. As a
result, copying grounds Hume’s epistemol-
ogy causally and creates an externally gov-
erned inner phenomenological epistemology.
To bolster the idea that copying communi-
cates to ideas a conscious indicator of believ-
ability, for the remainder of our chapter we
focus on the secondary literature. We analyse
four interpretations of vivacity directly rel-
evant to our sense-based account. Each has a
kernel of truth yet each fails in an instructive
way. The critical analysis is intended both to
clarify and support our position. If it fails to
do so, a critical review of the secondary lit-
erature stands on its own and fits well with
the spirit of a reassessment.
3. VIVACITY AS ‘QUALIA’
Much of the secondary literature on vivac-
ity fits under the theoretical concept from
current philosophy of mind, ‘qualia’. The
‘qualia-interpretation’ of vivacity encom-
passes three dimensions: ‘folk psychological’,
‘epistemological’ and ‘metaphysical’.6 Let us
briefly analyse these and then turn to the sec-
ondary literature on vivacity.
In the folk psychological sense, qualia
refer to the qualitative, ‘felt’ content of
conscious states. The milky blueness of the
northern sky on a clear spring day would
be an example. In its epistemological sense,
qualia constitute the ‘sensuous’ foundation
of knowledge. The qualia epistemology of
empiricists coincides with the Cartesian epis-
temological ‘marks of the mind’.7 The third
dimension concerns the intrinsic nature of
qualia.8 Metaphysically, qualia are non-
relational properties. Together, the three
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
dimensions form the ‘internalist’ Cartesian
picture of the mind: intrinsic qualia are dir-
ectly known, without inference and represen-
tations, and the object of such knowledge is
an intrinsic quality, fully internal to the mind
and independent of anything external to it.
For Hume, vivacity varies in strength
across mental state types. ‘Vivacity’ may be
taken to refer to the strength in folk psy-
chologically understood qualia. Bennett’s
and Stroud’s interpretation of vivacity as
‘phenomenological intensity’ does just that.9
For them, vivacity is Hume’s description of
a directly phenomenologically known varia-
tion in intensity. For example, if the quale is
the colour red, ‘vivacity’ describes its vary-
ing intensity so that colour impressions are
highly vivid and memory colours less so. The
interpretation gains support from Hume’s
phenomenological methodology and specific
remarks about the uniform nature of vivacity
across perceptions: ‘the idea of red we form
in the dark and that impression which strikes
our eyes in sun-shine differ in degree, not in
nature’ (THN 1.1.1.5 / 3).
The qualia interpretation does not just
concern the folk psychology. Qualia are a
close cousin of the epistemological primitives
of neo-classical British empiricism. Their
‘sense data’ and ‘the given’ characterize the
foundational epistemological role of ‘sensu-
ous’ experience. Price, for example, defines
sense data as that which cannot be doubted
when having a sensory-experience and iden-
tifies them with Humean impressions.10 And
since all twentieth-century British neo-classi-
cal empiricists are typically seen as Humeans,
the qualia-interpretation results in historical
continuity. More specifically, as an epistemo-
logical qualia-concept, vivacity indicates the
foundational role of the senses. Yet common
epistemological inclinations do not exhaust
the qualia-interpretation.
44
9780826443595_Ch02_Final_txt_print.indd 44
Empiricist epistemology bottoms out in
perceptual theory. The transparent, indubi-
table epistemology of qualia rests on their
metaphysical and phenomenological intrin-
sicality. This is evident in the jargon used for
knowledge of qualia. They are ‘felt’, ‘had’, or
‘acquainted with’ instead of explored and
represented. The epistemological puzzles
indicate the same. Qualia foundationalism
inherits Descartes’s ‘problem(s) of knowl-
edge’ about anything external to inner qualia
including the world. Empiricist perceptual
theory has a symmetrical problem: sense-
perceptual consciousness is intentional,
about the world, yet qualia are intrinsic, not
about anything. The ‘problem of perception’
concerns how sense-perception ‘reaches’ to
an external world from intrinsic qualia. An
example clarifies this model of the senses and
how it influences Hume interpretation.
Price describes sense data in the experi-
ence of a tomato as ‘a red patch of a round
and somewhat bulgy shape’.11 Qualia,
sense data, hues and shapes are the given
foundation for his empiricist epistemology.
But they are also the given for his percep-
tual theory – ‘something from which all
our theories of perception ought to start,
however much they may diverge later’.12
For ‘Humean’ empiricists such as Price,
intrinsic qualia form the elements of
sense-perception; they are the ‘sensations’
of sensory-psychologists. The intentional
sense-perception of a tomato is created out
of non-intentional hue and shape qualia.
How this occurs, however, is a problem in
the philosophy of perception. When qualia
are identified with sensations of perceptual
theory, they are typically also phenomeno-
logically non-representational/intentional
states. Vivacity, in the qualia-interpretation,
inherits the metaphysics and phenomenol-
ogy of sensory intrinsicality.13
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
Sensory vivacity is central for the qualia
interpretation so let us focus on it for criti-
cal purposes. The qualia model of senses is
programmatic to neo-classical empiricism.
But is it Hume’s sensory metaphysics? In psy-
chology, as elsewhere, Hume emphasizes the
‘non-relational’ and thus intrinsic aspects of
perceptions: ‘every distinct perception, which
enters into the composition of the mind, is a
distinct existence, and is different, and distin-
guishable from every other perception’ and
that there are no ‘two distinct impressions,
which are inseparably conjoin’d’ (THN
1.4.6.16 / 259, 1.2.6.3 / 66). In his theory of
senses, Hume emphasizes that simple impres-
sions, colour, taste, smell and touch, ‘admit of
no distinction or separation’ (THN 1.1.1.2 /
7–8). These are typically also interpreted as
indivisible perceptual ‘atoms or corpuscles’
that make up the idea of space (THN 1.2.3.15
/ 30).14 Simple impressions are clearly non-
intentional. They lack relations and thus
structure, including spatial structure. Because
of the intrinsic nature of simple impressions,
Hume’s theory of the senses is often seen as
the paradigmatic ‘atomistic’ one.15
But are Hume’s complex sensory impres-
sions similarly intrinsic? The common empiri-
cist interpretation of Hume’s account of the
origin of clearly intentional perceptions,
belief-ideas, shows that complex impres-
sions are not seen to add much to atomic
qualia. Price’s classic account of Hume’s
theory of the external world, the belief-idea
of a separate and independent world, is an
example.16 For Price, Hume’s ‘fleeting, per-
ishing impressions’ form a ‘gappy’ sequence
of sense data (THN 1.4.2.20 / 195). Even
complex impressions give only ‘gappy’
information. Price thinks that higher-order
cognitive processes of the imagination sup-
ply the needed ‘gap-fillers’. He concludes
that Hume’s mind ‘postulates’ to fill in the
45
9780826443595_Ch02_Final_txt_print.indd 45
gaps in sensory information and to create a
belief in an external world.17 And given that
beliefs for Hume are phenomenologically
vivid ideas, Price’s gap filler ‘hypothesis’ also
concerns perceptual phenomenology. Our
awareness of a ‘smooth and continuous’
world requires intra-mental higher-order
‘gap fillers’.
Contrary to the qualia account, we believe
that Hume’s complex impressions have
intentional information and that Hume gives
a causal explanation of it. His central theor-
etical sense-perceptual concepts are two con-
scious qualities, constancy and coherence’
(THN 1.4.2.15–20 / 194–5). Information
on the constancy and coherence of sensory
objects concerns complex impressions.18
These qualities causally influence imagin-
ation to form similar qualities in ideas,
‘continued and distinct existing objects’.
Imagination does not ‘supply’ the latter qual-
ities central to the belief-idea in the exter-
nal world. Instead, imagination processes
complex sensory information that already
is coherent and constant. Price’s dynamical
metaphor of cognitive processes ‘working
upon data’ is then misleading. First, it dimin-
ishes the information in complex impres-
sions. Second, it overlooks the similarity in
information between ideas and impressions,
the similarity of the quality pairs, ‘continued
and distinct’ and ‘coherent and constant’.
Hume relies on this similarity in explan-
ation. According to him, the belief in exter-
nal objects ‘must arise from a concurrence’
of some ‘qualities’ of ‘certain impressions’
combined with ‘qualities of the imagination’
(THN 1.4.2.15 / 194; emphasis added). In
Hume’s full account, idea-intentionality and
belief in external bodies is caused by constant
and coherent complex impressions. Copying
‘communicates’ causally similar qualities to
ideas, viz. distinct and continuous objects.
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
The similarity encompasses the felt inten-
tional object-involving qualities of both per-
ception-types. Principles of the imagination
are also involved in the causal process but
these cause identity in the intentional objects
of ideas. In Hume’s own words, these innate
processes ‘render the uniformity as compleat
as possible’ once the mind is ‘in the train of
observing uniformity among objects’, and the
postulation of continued existence ‘suffices’
for this purpose (THN 1.4.2.22 / 197–8).
Hume’s account differs significantly from
the Cartesian-Kantian nativist approach to
cognition. For Hume, internal imaginative
processes exaggerate the intentionality of
ideas. Ideas are neither a cognitive achieve-
ment nor a priori. Moreover, the causes of
exaggerated ‘postulation’ extend beyond the
mind to language – as they did for Hume’s
nominalist predecessors. But whatever their
causation, the felt ‘distinct and continued’
objects of ideas – Hume’s pun on Descartes’s
distinct and clear idea-objects – are only simi-
lar to sensory objects. Since intentional objects
and, at times, their causes differ for impres-
sions and ideas, vivacity is also included in
Hume’s causal model. Let us see how.
Both of the qualities of constancy and
coherence pertain to sensory change. For
Hume, coherence is directly about the change
of complex impressions. When their change
profile is lawlike, complex sensory informa-
tion is ‘coherent’.19 The quality of constancy
also concerns change. Hume characterizes
it in terms of the ‘order’ of appearance of
objects with respect to changes in the per-
ceiver such as his or her head movements.20
This is a mere illustration of the perpet-
ual change in circumstances of perception.
Despite these changes, the objects remain
constant yet ‘only’ so instead of identical.21
For Hume, lawlike change is the key
causal factor in sensory intentionality and
46
9780826443595_Ch02_Final_txt_print.indd 46
‘stable experience’. It is created from change
in impressions. The causally central facts of
change include the variation in stability of
complex impressions and the fact that stable
ones do not just pop inexplicably into exist-
ence. The additional causal posit – vivac-
ity – tracks change and informs about stable,
coherent and constant change. Since this
type of change is necessary and sufficient
for sensory intentionality, vivacity is con-
scious information on intentionality. And,
once copied, vivacity causes the association
of ideas with an ‘external world’ instead of
internal (causal) fancy. Vivacity is a quality
indicative of warranted intentional objects,
sense-based ones.22
Let us conclude with a summary. The
qualia-interpretation leaves vivacity without
an epistemological function. At best, vivacity
informs about inner qualitative change. But
such change does not inform even about the
existence of a world external to the qualia
let alone distinguish beliefs from imaginative
creations. But for Hume, vivacity informs
about the lawlike or constant and coherent
change of intentional sensory objects. This
function is copied to the idea-realm where
vivacity informs about believable ideas
through their (intentional) objects’ vivacity.
The epistemological emptiness of the qualia-
interpretation follows from a mistaken view
of Hume’s phenomenology and metaphysics
of the senses. Constant and coherent com-
plex impressions are both metaphysically
and phenomenologically intentional.
4. VIVACITY AS PRESENTEDNESS
We are not alone in underscoring the inten-
tional nature of Hume’s senses. In this section,
we analyse Dauer’s intentional interpretation
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
of vivacity as the ‘what it’s like of conscious-
ness’. For Dauer, vivacity is Hume’s way of
describing what it is like to undergo sense-per-
ception, remembering and believing. Although
vivacity applies broadly in phenomenology,
the core meaning emerges from sensory
experience. Dauer defines sensory vivacity in
terms of a ‘sense of presentedness’.23 Sensory
consciousness presents whereas thinking rep-
resents and the two, for Dauer, can ‘never be
confused’.24 Thus vivacity demarcates cat-
egorically between impressions and ideas and
gives the unique phenomenology of sensing.
Let us look at the details.
Dauer develops the sensory phenomenol-
ogy of presentedness through the ‘what it’s
like’ concept popularized by Nagel.25 Dauer
interprets Nagel’s scenario about the dif-
ferent subjective points of view of bats and
humans in terms of shape perception. For
bats that echolocate, sounds present shapes.
That is their vivid experience or ‘what it’s
like’ of sensory consciousness. Humans dif-
fer subjectively since colours present shapes.
For Dauer, it is ‘an essential feature of visual
experience’ that ‘colours give us shape’.26 The
presentation of ‘shape in colours’ constitutes
Humean (visual) vivacity for humans.
In Dauer’s account, Humean sensory phe-
nomenology is intentional. It is ‘about’ some-
thing. Colour impressions do not just ‘give’
colour as qualia would but ‘give’ shape.
Colours are about shapes. Bat phenomen-
ology is also intentional, about shape. Dauer’s
presentations of shape occur across different
types of sensory experience. This state of
affairs, in turn, is ‘definitional’ for intention-
ality. Intentionality is ‘aspectual’: intentional
states reach out to their objects ‘under this
aspect rather than another’.27 Bats’ sensory
states reach out to the ‘object’ of shape under
the aspect of heard sound; humans’ sensory
states reach out to that object under the
47
9780826443595_Ch02_Final_txt_print.indd 47
aspect of seen colour. We will at times use
the term ‘presentational-representedness’ to
underscore the intentional nature of Hume’s
senses for Dauer.
Presentational-representedness causes prob-
lems both for the intended phenomenological
sense of vivacity and the categorical distinc-
tion between impressions and ideas. Bats and
humans differ in sensory phenomenology
yet their vivid sensory states are about the
same intentional object, shape. We will label
it ‘shape’ to underscore its phenomenologi-
cal opaqueness. ‘Shape’ cannot reduce to the
‘transparent’ qualitative shape or colour of
human visual experience because bats access
‘shape’ too without experiencing colours. The
opaque phenomenology undermines Dauer’s
distinction between representational ideas
and presentational impressions. For Dauer,
representations are many-one ‘mappings’.
But both ‘presentations’ and ‘representations’
are in a many-one relationship to a common
object. ‘Shapes’ are presented by sounds and
colours; ‘shapes’, for Dauer, are represented
by equations and sonar. What is the differ-
ence between the two many-one mappings?
Since presentationally represented objects
are opaque in phenomenology, Dauer cannot
ground the difference in a phenomenology
of the given. The problem applies directly to
vivacity.
For Dauer, Humean senses do not reduce
to qualia. Vivid colours, for example, do not
reduce to ‘phenomenologically intense’ col-
our qualia. Vivacity is an intentional ‘quality’.
Dauer also distinguishes presentational-rep-
resentations from representations in terms
of the former’s vivacity. Yet both are inten-
tional. Dauer cannot base this distinction in
either qualia or intentionality. As a result,
we do not know what role qualia have in an
intentional sensory phenomenology, or vice
versa. Dauer’s predicament is common for
2/9/2012 9:39:34 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
intentional accounts of sense-perception. The
relationship, if any, between the phenomeno-
logically given and representational object/
content is the core puzzle for such accounts.
Without an answer to it, colours could be
seen to ‘give’ shape non-phenomenologically.
Indicator-theories of sensory content do just
that. Colours serve the function of shape per-
ception without being ‘felt’.28 But a merely
‘indicated’ intentional object is not a given in
phenomenology just as a represented object
need not be felt in its representation. Both
are opaque representations.
Is there a way to resolve Dauer’s prob-
lem in order to advance a representational-
intentional account of Humean senses that
includes vivacity? Dauer’s predicament rests
on the specificity of the representational
object, ‘shape’, and representer, colour,
together with the opaqueness of phenom-
enology. This potentially initiates a problem-
atic slide towards indicator theories. ‘Shape’
could though refer to the felt world-involv-
ing feature of sensory experience as such. If
so, ‘shape’ is neither a specific type of quale
nor an opaquely indicated domain. Instead,
‘shape’ refers to the intentional ‘feeling’ of
vision. If this is Dauer’s view of Humean
vivacity, it is promising, although it is not
an accurate account of Hume’s view. Let us
see how colours might play a role in vivid or
intentional sensory experience according to
Hume’s theory of senses.
Colours qua simple impressions do not
‘give us’ anything in phenomenology. Simple
impressions are non-intentional qualia.
Not even all Humean ‘coloured’ complex
impressions ‘give’ us anything. For example,
a still and unchanging or ‘uniform’ colour
mosaic will lack intentional appearance.
This is because, for Hume, only complex
impressions with felt coherent and constant
change-profiles appear intentional. Colours
48
9780826443595_Ch02_Final_txt_print.indd 48
play a role in intentional sense experience if
changing colour impressions have a role in
the creation of constant and coherent com-
plex impressions. It is very likely that col-
ours have a systematic role in this. Sensed
objects have systematic colour-profiles of
change. Such ‘coherent and constant’ change
concerns change in the colour of objects
relative to our action and changes in the
objects and their perceptual environment.29
For Hume, colours have a role in ‘presented-
ness’ of sense-perception because they have
a role in the causation of intentional sensory
objects.
What is the role and nature of vivacity
for intentionality? Vivacity is an indicator of
change and lawlike profiles of change make
intentional sensory objects in the first place.
Vivacity has a central causal role. But what is
the phenomenological nature of vivacity rel-
ative to philosophical terms of art such as the
‘given’? Dauer equates vivacity with a given
sense-perceptual presentedness. But liveli-
ness cannot characterize momentary experi-
ences, states or ‘perceptual objects’. Vivacity
pertains to change and characterizes events.
If ‘the given’ excludes definitionally the phe-
nomenology of events, vivacity cannot be
given. But to understand Hume we must find
room for an event phenomenology. The con-
scious qualities of constancy and coherence
of complex impressions themselves charac-
terize change. Intentional sensory objects are
created from such change and are always in
a process of making. Nothing in an experien-
tial state corresponds to the quality of vivac-
ity; nothing in experience corresponds to a
fixed sense-perceptual object. As Noë, the
enactivist perceptual theorist, puts it, there
are no ‘snapshot pictures’ in experience.30
Our Hume would agree.
Still, ‘presentedness’ does capture a fun-
damental feature of sensory experience.
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
To articulate this ‘given’, we must bring
into view the change-oriented nature of the
presented world. Noë refers to it as a ‘felt
presence’.31 Sensory awareness of a world
concerns the availability of further detail
about its objects. We feel this and this is
because of change in the sensed objects and
their circumstances. Our Hume would agree.
He would underscore that the ‘felt presence’
is a ‘given’ only once we have learnt the
indicator function of vivacity with respect
to profiles of change in sensory objects. We
are accustomed to ‘just feel’ how the sensed
world will change according to learnt pro-
files of change. We learn associations and feel
them. A better word for this phenomenology
is a felt ‘giving’. As a ‘given’, vivacity must be
seen in intentional terms, as being about the
further information available to us. That is
what vivacity tracks and what it has taught
our sensory phenomenology.32
Our account of ‘the giving’ in vivac-
ity applies to ideas in Hume’s comparative
sense. The intentional objects of ‘idea-events’
change less and for several reasons. Cognition
is ‘off-line’ and not open to external change
in the manner ‘on-line’ sense-perception is.
Sense-perception allows us to access informa-
tion from the world instead of just (copied)
memories. Moreover, ideas traffic in ‘ficti-
tious’ categories based in the philosophical
relation of identity. Extra-sensory, intramen-
tal processes ‘reify’ their intentional objects
in terms of identity, whose effects are not felt
in the senses – as each of us can verify by
attending carefully to the perpetual change
of sensory objects. For these reasons, ideas
lose in vivacity to the ‘constant and coherent’
intentional sensory objects.
Hume’s epistemological message with viv-
acity is that a fully empirical, impression-
related indicator applies to ideas. Hume’s
empiricism is externalist. The causal processes
49
9780826443595_Ch02_Final_txt_print.indd 49
behind intentional objects overlap only for
‘sensing and thinking’. Contentful ideas have
many causes. In epistemology, Hume is con-
cerned with errant causes, imaginings and lin-
guistic habits, and their fictitious idea-effects.
Equating believability with an indicator cop-
ied from the senses addresses this concern.
Whether vivacity should be seen as a ‘given’
according to the conceptual habits of some
later empiricist movements is far less import-
ant than knowing what vivacity teaches us.
For Hume, it informs us of the sensory origin
of some representational objects and teaches
us which ideas are worthy of belief.
5. VIVACITY AS VERISIMILITUDE
In this section, we analyse another intentional
interpretation of vivacity, Waxman’s ‘verisi-
militude’. The term was originally used by
Popperians in philosophy of science to refer
to the ‘truth-likeness’ of scientific theories.
But Waxman’s ‘verisimilitude’ is not a concept
of metaphysics or even an epistemological
judgement of representations. Waxman uses
it to describe the ‘truth-likeness’ of Humean
perceptions. For Waxman, ‘before there can
be belief in the existence of any perception,
consciousness must respond to it with the
feeling that it is real’.33 Verisimilitude is this
phenomenological ‘reality-likeness’ of per-
ceptions, their felt ‘presence and reality’.34
Let us first explore Waxman’s idea by com-
paring it with the two earlier interpretations
and then turn to a critical analysis.
Waxman’s verisimilitude reading differs
from the qualia-interpretation. Verisimilar
perceptions are a response to qualia, whereby
the intrinsic states become about reality.
Waxman’s interpretation of vivacity is simi-
lar to Dauer’s ‘presentedness’. Yet Waxman
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
overcomes the ambiguity about the object
that is presented in phenomenology and the
un-Humean narrow applicability of pre-
sentedness only to impressions. Waxman’s
account underscores the general world-
involving nature of vivacity for the phenom-
enology of both ideas and impressions. He
distinguishes vivacity from Humean ‘percep-
tual objects’. The latter are mere qualia or
‘appearances’.35 Verisimilitude is not their
quality (of intentionality) but a ‘quality of
consciousness’. As a result, Waxman would
disagree with Dauer’s contention that, in vis-
ual phenomenology, colours ‘give’ or ‘present’
us with shape. Colours as perceptual objects
are qualia that present or ‘give’ nothing.
Once colours are verisimilar, they are inten-
tional but not about any specific object (such
as shape) nor opaquely so (as they would be
if they were about ‘shape’). Verisimilar visual
experience presents a full-blown, intentional,
coloured reality.
To achieve interpretative overlap and
conceptual coherence, it is best to analyse
Waxman’s distinction between qualities of
consciousness and perceptual objects as two
different ‘styles’ of consciousness. In one,
perceptions are merely ‘present’ as qualia to
consciousness.36 For Waxman, this style of
consciousness is not vivid. The other style
is. Consciousness ‘responds’ to idea and
impression qualia with a feeling of reality
or verisimilitude. These perceptual objects
are intentional. For Waxman, both styles
of consciousness are a matter of ‘human
nature’ or ‘reflect its constitution’.37 In other
words, both are innate. The terms Waxman
uses for the role of consciousness in imbuing
Humean qualia with verisimilitude – ‘act of
the mind’, a ‘regarding’, a ‘stance’ – are then
misleading in their causal, process perspec-
tive.38 All Humean perceptual objects acquire
vivacity just ‘in’ consciousness, so vivacity is
50
9780826443595_Ch02_Final_txt_print.indd 50
something ‘attending perceptions’.39 Waxman
concludes that vivacity is an ‘original div-
ision’ of consciousness for Hume.40
Waxman’s analysis of vivacity incorpo-
rates the causal history of the relevant phe-
nomenology. Waxman’s Hume is a nativist:
‘one could in principle distinguish impres-
sions and ideas from the first moment of
one’s conscious life.’41 This nativism is based
in an innate vivacity. Nativism entails that
intentionality is not caused by sensory expe-
rience or intramental cognitive processes
such as judgement. Or more generally, since
vivacity is not a quality of perceptual objects,
it cannot be affected by the causes that
affect Humean perceptual objects. Nativism
also influences Waxman’s phenomenology.
Vivacity is an intentional ‘frame’ of phenom-
enology and cannot vary across specific types
of intentional objects.
Let us turn to critical analysis. Waxman
pairs nativism with naturalism. By leav-
ing verisimilitude with ‘human nature’ and
by conceptualizing it as having a ‘primitive
character’ that ‘applies to already animals
and infants’, Waxman intends to do justice to
Hume’s naturalism.42 But, historically, nativ-
ism is associated with the anti-naturalism of
the rationalists. Not surprisingly, Waxman’s
nativist reading contradicts Hume’s causal
theory of intentionality. Let us briefly revisit
Hume’s account to see why.
Waxman ignores the distinction between
simple and complex impressions and their
qualities. Simple impressions have no intra-
systemic causal history for Hume. If we access
these in phenomenology, they are qualia.
Simple impressions are ‘original’ with respect
to Hume’s theoretical psychology and maybe
to his account of consciousness. Neither is true
of all complex impressions. First, all complex
impressions have a causal history in process-
ing. Much of this falls below the threshold
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
of consciousness with respect to ‘felt’ indi-
cators. To the extent that it does, the causal
origins of complexes have no corresponding
original divisions in ‘qualities’ of conscious-
ness. But this is not true of the causation of
constant and coherent complex impressions.
Such structures are caused by constant and
coherent change. It is reflected in conscious-
ness through effects, constant and coherent
sensory objects and vivacity. Furthermore,
these qualities, together with copying, cause
the belief in a separate and distinct external
world. And this idea, in turn, has precisely
the phenomenological identity Waxman
refers to with ‘verisimilitude’, the felt pres-
ence of an external reality. As a consequence,
neither verisimilar ideas nor verisimilar com-
plex impressions are phenomenologically
‘original’ or causally innate. Both have a
causal history that extends to our conscious-
ness of them. And when we add that Hume’s
account also concerns specific intentional
objects and not just the intentional ‘frame’ of
phenomenology, we see that Waxman’s inter-
pretation misses its mark also in the role of
learning. The indicator function of sensory
vivacity with respect to specific coherent and
constant objects is learnt. Only through a
successful causal learning-history can copied
vivacity ground the epistemology of believ-
ability within ideas. Epistemologically war-
ranted, verisimilar consciousness is a learnt
style of consciousness without an influence
from unnaturalistic ‘judgemental’ processes.
6. VIVACITY AS FUNCTIONAL ROLE
Vivacity is a copied epistemological indicator.
In accordance with this idea, we have ignored
interpretations of vivacity that emphasize
the internal processing of perceptions.43
51
9780826443595_Ch02_Final_txt_print.indd 51
The heuristic leaves the functionalist inter-
pretation straddling the fence between
externalism and internalism. Since our own
view underscores mental functioning, a few
remarks on it are in order to bring our ana-
lysis of the secondary literature to a close.
For functionalists, vivacity concerns the
system-wide significance of perceptions, their
‘force’ in influencing other perceptions.44
Vivacity is the intramental ‘functional role’
of perceptions. And such roles vary since
perceptions differ in their degree of inter-
connectedness and their associated felt
‘force’. Functionalists have an explanation of
Hume’s identification of belief with vivacity.
Believable ideas have interconnections and a
felt influence on other ideas, whereas mere
ideas do not.
Sensory vivacity is problematic for func-
tionalists and especially when impressions are
seen as non-relational qualia. In her seminal
functionalist interpretation, Govier solves the
problem by dividing Hume’s vivacity-terms
into two categories.45 ‘Forceful’ and ‘vigor-
ous’ are functional terms. ‘Vivid’ and ‘lively’
are not but describe the clarity and amount
of detail of a perception. Govier argues that
the functional concept applies to ideas and
that sensory vivacity concerns only clarity
and detail.
Govier’s Hume turns out to be a coher-
entist instead of an empiricist. Vivacity of
beliefs has nothing to do with vivid impres-
sions. Copied sensory vivacity concerns only
‘informational quality’ – clarity and detail –
which is without epistemological significance
for Govier. Hume’s epistemology rests fully
on the interconnections among ideas. This is
a distorted picture. Yet Govier’s interpreta-
tion can be diagnosed in a manner that sup-
ports Hume’s central epistemological idea,
the relationship between sensory informa-
tion and cognitive function.
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
Only intentional complex impressions
have Govier’s informational qualities of
detail and clarity. Only a specific type of
change in impressions, a specific ‘functional
role’, causes impressions with both detail
and clarity. The phenomenology of ‘felt pres-
ence’ or ‘giving’ shows that sensory objects
take clear form through detail. Sensory
information integrates functional profiles
with informational quality. This allows for
an inner epistemology for ideas: through
copying, detail and clarity become criteria
for the intentional objects of ideas which
themselves are an effect not just of copying
but also of ‘independent’ intramental cogni-
tive functioning. The skill of thinking and
internal functioning is itself governed by
skilful sensory functioning through shared
criteria for the intentional objects of both.
Vivacity is a ‘functional role’ indicator but
an empirical one, without a trace of system-
wide coherentist epistemological functions.
Unintentionally, Govier gives evidence in
favour of our integrated interpretation.46 In
seventeenth-century English, her term for
clarity and detail, ‘vivid’, meant ‘lively’, her
term for functional role. Perhaps Hume’s
use of ‘vivid’ caused a natural association
between ideas of change and stability in the
minds of his audience?
7. CONCLUDING REMARKS:
HUME’S DEVELOPMENT OF
THE THEORY OF IDEAS
Our interpretation has followed a traditional
internalist historiography. Its goal is coher-
ence. Hume’s theory becomes coherent when
vivacity is understood as a phenomenologi-
cal indicator of an ‘inner methodology’ cop-
ied from the senses; Hume’s epistemology
52
9780826443595_Ch02_Final_txt_print.indd 52
becomes a logical extension of his ‘genetic’
psychology of sensory intentionality. But
coherence can make myths. A historical
hypothesis must also be fruitful. Our con-
cluding remarks are intended to achieve just
that, offer ideas for a reassessment of Hume’s
place in the history of philosophy.
‘Idea’ is the epoch-making notion of pre-
Kantian, ‘early-modern’ philosophy. Hume’s
philosophy is one theory of ideas. History
has not dealt kindly with the ‘idea-idea’.
Linguistic and/or logically structured propos-
itions replaced ideas without a trace from the
analytic tools of epistemologists. Ordinary
languages, forms of life, and other holistic
epistemologies buried ideas under mangled
practices. ‘Ideas’ are so thoroughly discred-
ited that addressing Hume’s version appears
of antiquarian interest, at best. But the cur-
rent wisdom is itself importantly historical:
critics of early empiricism learnt their early
empiricism from much later empiricists. If
only for this reason, Hume’s ‘ideas’ deserve
a fresh look.
Hume’s empiricist predecessors never
recovered ideas from the Cartesian men-
tal depths. But Hume did and by meeting
Descartes on his turf. Hume too articulated
an ‘inner methodology’ for ideas. Cartesian
clear and distinct ideas were replaced with
vivid ideas. This indicator of believable
ideas was copied from the senses. Hume’s
externalist epistemological internalism was
usable. Fictitious ideas including dreams are
felt to be unbelievable, phenomenologically,
through idiosyncracies in vivacity. Fictions
can be identified ‘internally’, with vivacity
of copied memories and idea-acts – without
further experience. Hume thought that the
inner methodology could even fight ‘custom’
and its imaginative necessary connections.47
The ‘internal arm’ of Hume’s empiricism
was decisive on the ‘problem of the external
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
world’ and with facts closest to home, con-
sciousness of change in perceptions.
Despite their overlap, Hume did not
extend Descartes’s project. To Hume’s detri-
ment, it was Descartes’s project that through
neo-Kantians won a decisive (interim) vic-
tory in the competition over the sociological
aspects of epistemology. Cartesian-Kantian
epistemology focuses on ‘form’, representa-
tional frameworks. For understanding Hume,
the differences between the form of ideal-
ized languages and clear and distinct ideas
is irrelevant. Hume had no epistemology of
form because he believed no representational
framework existed nor was required for the
mind to know. Abstract ideas are fictions
because no abstract mental representational
vehicle exists. Without one, information can-
not be structured by identity and difference,
whether codified in clear and distinct ideas
or analytic definitions.
His disregard of form entails that the text-
book image of Hume as the forefather of log-
ical positivism is misleading. His structured
knowledge is inexpressible through defini-
tions, whether sensation, observation or
operation language-based. The reason is that
his knowledge-mechanism does not follow
a syntax or grammar. Thus his empiricism
does not rest on empirical atomic proposi-
tions – Tractarian or Russellian logical atoms
that mirror empirical facts – that turn into
complex propositions by truth-functional
connectives. Hume’s mechanism is not rule-
governed at all because it is causal; it is
governed by all of the causally robust dimen-
sions of acts of experience. No logic governs
these causal principles; complex information
is a contingent matter. Nor can it be stud-
ied in the abstract or normatively as gram-
mar and logic are. In short and unlike that
of the logical positivist, Hume’s empiricism is
naturalistic. Still, he did drag ‘ideas’ from the
53
9780826443595_Ch02_Final_txt_print.indd 53
Cartesian internal depths even if he stopped
short of turning them into logical, social arte-
facts. But did he drag ideas to the world? His
perceptions constitute an ‘empirical world’.
But this is not based in transcendental rea-
sons of a priori cognitive achievements. The
innate a priori categories of Kant’s nativist-
empiricism ‘schematized’ in the senses are
not a limit for or origin of Hume’s empirical
world. It is a labile causal interface of per-
ception and action. Complex structure is cre-
ated from change and its only resting place is
coherent and constant change accessed from
the world, as such, unknowable.
Hume was then an empiricist about the
nature of empiricism, unlike Kant and the
logical positivists. In this ‘naturalism’, Hume
is reminiscent of today’s epistemologists
who deny ‘first philosophy’. He never gave
a ‘first version’ of his empiricism either. Yet
the modern naturalism is based in Quine’s
view of holistic knowledge, one that ignores
psychology and the knower’s awareness.
Modern naturalism is eliminativist in its
externalism: a knower need not know that
he or she knows to know. Justification exists
below and beyond the radar of the knower,
in phylogenetic processes (evolutionary epis-
temology), in the social epistemology of
scientific practices (Quine and Putnam’s div-
ision of linguistic labour), and unconscious
neural processes (Churchlands’s eliminativ-
ism). Each demands that the agent defers
justification to external forces and loses
touch with the felt dimension of knowledge
to know.
In Hume’s naturalism, only the phenom-
enological afterglow of associations can cre-
ate justified beliefs and an epistemic agent.
He never let go of the role of the subject’s
learnt skilful awareness for knowledge. In its
developmental, individualistic and phenom-
enological orientations, Hume’s account is
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
too limited to solve many of today’s burn-
ing epistemological problems. The contrast
in naturalisms appears to us then as much
historical as it is substantive. Historically,
Hume’s epistemology is of and for the
Enlightenment. From the Enlightenment
perspective, current naturalism is not only
psychologically implausible but socially irre-
sponsible. Substantially, if the Enlightenment
philosophers tackled problems still rele-
vant today, Hume offers a systematic nat-
uralistic empiricism for addressing them. In
both cases, our reassessment of his copying
should help to see the history of philosophy
and especially empiricism in a novel, fruitful
manner.
NOTES
1
A comprehensive analysis would also cover
Hume’s other quality, ‘facility’ (THN 1.3.13.6 /
146; 2.3.5.5 / 422; EHU 4.2 / 25). We will not
explore facility on this occasion, but merely
remark that facility is clearly dependent on
vivacity since no ease of transition within
perceptions can occur without identification of
transitions or patterns of change.
2
See P. Godfrey-Smith, Complexity and the
Function of Mind in Nature (Cambridge:
Cambridge University Press, 1996), chap. 2.
3
See H.H. Price, ‘The Permanent Significance
of Hume’s Philosophy’, in A. Sesonske and
N. Fleming (eds), Human Understanding:
Studies in the Philosophy of David Hume
(Wadsworth Publishing, 1968), p. 7. See
also A. Rosenberg, ‘Hume’s Philosophy of
Science’, in D.F. Norton (ed.), The Cambridge
Companion to Hume, 2nd edn (Cambridge:
Cambridge University Press, 1993), p. 66;
G. Dicker, Hume’s Epistemology and
Metaphysics: An Introduction (London:
Routledge, 1998), p. 11ff.; J. Bennett, Locke,
Berkeley and Hume: Central Themes (Oxford:
Clarendon Press, 1971), chap. 9; Learning
from Six Philosophers, vol. 2 (Oxford:
Oxford University Press, 2001), chap. 32; and
54
9780826443595_Ch02_Final_txt_print.indd 54
‘Empiricism about Meanings’, in P. Millican
(ed.), Reading Hume on Human Understanding
(Oxford: Oxford University Press, 2002),
chap. 3; A. Flew, David Hume: Philosopher
of Moral Science (Oxford: Blackwell, 1986),
pp. 22–3 and D. Pears, Hume’s System (Oxford:
Oxford University Press, 1990), pt 1. More
recent advocates of a semantic interpretation
of copying focus on ‘concepts’, stored mental
representations that are used in thinking or
cognition more broadly (see, for example,
J. Fodor, Hume Variations (Oxford: Clarendon
Press, 2003). Although this idea has important
differences from the other one, we will not
differentiate the two on this occasion. The main
reason is that the overall outcome for the ‘fate’
of copying has been largely the same here as it
has in the more directly linguistically-oriented
semantic traditions. For an exception, however,
see J. Prinz, Furnishing the Mind: Concepts and
Their Perceptual Basis (Cambridge, MA: MIT
Press, 2002), chap. 5.
4
For an overview of the main puzzlements that
surround Hume on belief see M. Gorman,
‘Hume’s Theory of Belief’, Hume Studies 19(1)
(1993), pp. 89–102. No doubt the fuel to this
scholarly fire is provided by Hume’s own lin-
guistic dissatisfaction with the account of belief
in the Appendix to the Treatise, admitting that
he is at a ‘loss for terms to express [his] mean-
ing’ (THN 1.3.7.7 / 628).
5
R. Descartes, Rules for the Direction of the
Mind [1628] in J. Cottingham et al. (eds), The
Philosophical Writings of Descartes, vol. 1
(Cambridge: Cambridge University Press,
1984–91).
6
These three dimensions map onto Dennett’s
influential analysis although he does not label
them; see ‘Quining Qualia’ in W. Lycan (ed.)
Mind and Cognition: A Reader (Cambridge,
MA: MIT Press, 1990), pp. 519–47.
Originally in A. Marcel and E. Bisiach (eds),
Consciousness in Modern Science (Oxford:
Oxford University Press, 1988).
7
Dennett includes Cartesian marks such as pri-
vacy and direct knowledge in ‘Quining Qualia’,
p. 519 (see also, e.g., J. Kim, Philosophy of
Mind, 2nd edn (Boulder, CO: Westview, 2006),
pp. 18–22).
8
Dennett lists other key ideas used to charac-
terize the ‘intrinsicality’ of qualia: intrinsic
2/9/2012 9:39:35 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
qualia are non-relational properties – dynami-
cally – they do not change depending on the
experience’s relation to other things, and, syn-
onymously or interchangeably, intrinsic proper-
ties are ‘simple’, ‘homogenous’, ‘unanalyzable’
and/or ‘atomic’ (‘Quining Qualia’, p. 519).
9
See Bennett’s Locke, Berkeley and Hume,
19
p. 225; Stroud’s Hume (London: Routledge
& Kegan Paul, 1977), p. 29; J. Broughton,
‘Impressions and Ideas’ in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise (Oxford:
Blackwell, 2006), p. 45; Dicker’s Hume’s
Metaphysics, p. 5 and D. Owen, Hume’s
Reason (Oxford: Oxford University Press,
20
1999), p. 73.
10
See H.H. Price, Perception (London: Methuen
& Co., 1932), pp. 3, 19. See also Bennett’s
Locke, Berkeley and Hume, p. 222 and Flew’s
David Hume, p. 26.
11
Price, Perception, p. 3.
12
Ibid., p. 19.
13
Since vivacity attaches to all Humean percep-
tions, according to the qualia-interpretation,
all mental states have a non-representational
dimension since all of them have a qualia
21
dimension of vivacity.
14
D. Raynor, ‘Minima Sensibilia in Berkeley and
22
Hume’, Dialogue 19 (1980), pp. 196–200;
D. Garrett, Cognition and Commitment
in Hume’s Philosophy (Oxford: Oxford
University Press, 1997), pp. 60–2; and
W. Waxman, Hume’s Theory of Consciousness
(Cambridge: Cambridge University Press,
1994), pp. 44–6.
15
T. Holden, ‘Infinite Divisibility and Actual
Parts’, Hume Studies 28(1) (2002), pp. 3–26,
p. 4; M. Frasca-Spada, ‘Simple Perceptions in
Hume’s Treatise’, in E. Mazza and E. Ronchetti
(eds), New Essays on David Hume (Milan:
FrancoAngeli, 2007), p. 42.
16
H.H. Price, Hume’s Theory of the External
World (Oxford: Clarendon Press, 1940).
17
Ibid., chap. 2. For other qualia accounts on
this score see Stroud’s Hume, chap. 5 and
Dicker’s Hume’s Metaphysics, p. 169.
18
Constancy and coherence cannot be qualities
of simple impressions for they are ‘fleeting’,
‘perishing existences and appear as such’
(THN 1.4.2.15 / 195). That these qualities are
of complex impressions is clear from Hume’s
examples such as mountains, houses and trees,
55
9780826443595_Ch02_Final_txt_print.indd 55
or a bed, table, ‘books and papers’, ‘sun or
ocean’. Hume also gives the scenario where he
leaves the room for an hour and finds when he
comes back that the fire is not in the same way
as he left it (THN 1.4.2.18–19 / 194–5). These
illustrations all involve complex impressions
with clear intentional objects even if processes.
Re-entering the room he finds the fire not in
the same way as he left it, but he is accustomed
in other instances to see a ‘like alteration
produced in a like time, whether [he] is absent,
present, near or remote’ so this coherence of
‘their changes is one of the characteristics of
external objects’ (THN 1.4.2.19 / 195).
In the case of constancy he writes: ‘Those
mountains, and houses, and trees, which lie at
present under my eye, have always appear’d to
me in the same order; and when I lose sight of
them by shutting my eyes or turning my head,
I soon after find them return upon me without
the least alteration. My bed and table, my
books and papers, present themselves in the
same uniform manner, and change not upon
account of any interruption in my seeing or
perceiving them’ (THN 1.4.2.18 / 195).
See, for example, A. Noë, Action in Perception
(Cambridge, MA: MIT Press, 2004).
The idea that vivacity informs about stable
complex impressions creates conceptual uni-
formity among the many alternative terms
Hume uses for ‘vivacity’ such as strength,
violence, liveliness, vigour, firmness, steadiness
and solidity. From an ordinary language point
of view, the list contains unrelated and even
contradictory terms. For example, ‘solidity’
seems unrelated if not antagonistic to ‘live-
liness’. Solid things such as statues are inert,
unmoving and generally lacking in liveliness.
But from our interpretative perspective, the
two terms are naturally connected. ‘Solid’ com-
plex impressions are ones whose intentional
objects remain constant despite change and
because they display a robust ‘lively’ profile
of change. The object remains stable amidst
change by emerging from lawlike change.
The solid objects are also found out to be so
through our own action which, from a sensory
end point, consists of ‘liveliness’ or patterned
change in impressions. In short, solidity and
‘firmness’ emerge from liveliness or vivacity in
the same way that constancy amidst change
2/9/2012 9:39:36 PM
 HUME’S EMPIRICIST INNER EPISTEMOLOGY
and coherence of change complement each
other. We leave it to the reader to work out
how other apparently unrelated if not antagon-
istic terms Hume uses to characterize vivacity
indicate the dynamic between stability and
change. That Hume’s apparently unrelated lin-
guistic descriptors form a coherent whole lends
support to our interpretative perspective.
23
F.W. Dauer, ‘Force and Vivacity in the Treatise
and the Enquiry’, Hume Studies 25(1&2)
(1999), pp. 83–100, p. 86.
24
Ibid.
25
T. Nagel, ‘What Is It Like to Be a Bat?’,
Philosophical Review (1974), pp. 435–50.
26
Dauer, ‘Force and Vivacity’, p. 86.
27
G. Graham, Philosophy of Mind (Cambridge:
Cambridge University Press, 1993), p. 136.
28
In indicator theories, sensory content is identi-
fied with representational content. It has no
phenomenological ‘intrinsic’ features. In some
versions, we just ‘see through’ experience to its
representational content instead of accessing
qualia (e.g. G. Harman, ‘The Intrinsic Quality
of Experience’, in Philosophical Perspectives 4
(1990), pp. 31–52). In Dretske’s version, col-
ours can represent surface spectral reflectance
without it being transparent in phenomenology
(except in colour constancy); see F. Dretske,
Naturalizing the Mind (Cambridge, MA:
MIT Press, 1995). We can ‘see’ nothing about
the object represented but still our content
co-varies with and provides information on it.
All versions are problematic both for Dauer’s
idea of a phenomenologically given vivacity
and the distinction between impressions and
ideas in terms of intentionality or vivacity.
29
For details, see, for example, Noë’s Action in
Perception and J. Broackes, ‘The Autonomy of
Colour’, in K. Lennon and D. Charles (eds),
Reduction, Explanation, and Realism (Oxford:
Oxford University Press, 1992), pp. 421–65.
30
See Noë’s Action in Perception, p. 35.
56
9780826443595_Ch02_Final_txt_print.indd 56
31
Ibid.
32
Hume’s distinction between present in fact and
present in power can be easily used to articu-
late this phenomenology and the dynamic
between change and stability. Hume does so
himself for ideas; see THN 1.1.7.7 / 20.
33
Waxman supports the verisimilitude reading
from passages where Hume talks about ‘acts
of the mind’ that have a reality-revealing phe-
nomenological indicator quality; see Hume’s
Theory of Consciousness, p. 33ff.
34
Ibid., p. 34.
35
Ibid., pp. 38–9.
36
Ibid., p. 34.
37
Ibid., p. 35 and also Waxman’s ‘Impressions
and Ideas: Vivacity as Verisimilitude’, Hume
Studies 19(1) (1993), pp. 75–88, p. 77.
38
Waxman, Hume’s Theory of Consciousness,
pp. 33–5, 38, and ‘Impressions and Ideas’,
pp. 77, 81.
39
Waxman, Hume’s Theory of Consciousness, p. 42.
40
Ibid., p. 27.
41
Waxman, ‘Impressions and Ideas’, p. 75.
42
Waxman, Hume’s Theory of Consciousness,
pp. 38, 42.
43
For example, Stevenson’s account of conscious-
ness as an impression of reflection in ‘Humean
Self-consciousness Explained’, Hume Studies
24 (1998), pp. 95–130.
44
See, for example, S. Everson, ‘The Difference
between Feeling and Thinking’, Mind 97,
(1988), pp. 401–13.
45
T. Govier, ‘Variations on Force and Vivacity
in Hume’, Philosophical Quarterly 22 (1972),
pp. 44–52, p. 45ff.
46
Govier, ‘Variations’, p. 46.
47
For learning contexts geared towards novelty
and general applicability, such as those of
science, the inner arm had to be refined with
a codified and more rigorous external arm,
in essence, Baconian or ‘Millian’ methods of
causal inference (THN 1.3.15 / 173ff.).
2/9/2012 9:39:36 PM
 3
HUME’S ‘SCEPTICISM’
ABOUT INDUCTION
Peter Millican
‘Is Hume a sceptic about induction?’ This
might seem to be a fairly straightforward
question, but its appearance is misleading, and
the appropriate response is not to give a direct
answer, but instead to move to a more funda-
mental question which is suggested by Hume
himself at the beginning of his definitive dis-
cussion of scepticism in Enquiry Section 12:
‘What is meant by a sceptic?’ (EHU 12.2 /
159). His point here is that ‘sceptic’ can mean
many things, and what counts as ‘sceptical’
will often depend on the relevant contrast.
Someone who is sceptical about morality or
the existence of God, for example, need not
be sceptical about the external world. And
someone who is sceptical about the rational
basis of inductive inference need not be scep-
tical at all – in the sense of dismissive or crit-
ical – about the practice itself.
This crucial point about the varieties of
scepticism is often overlooked in discussions
of Hume on induction, generating a great
deal of misunderstanding. Commonly, the
debate will be framed in terms of a simple
contest between ‘sceptical’ and ‘non-scep-
tical’ interpretations. Then on the one side,
a case is made drawing on Hume’s famous
negative argument which apparently denies
induction any basis in ‘reason’.1 Meanwhile,
57
9780826443595_Ch03_Final_txt_print.indd 57
on the other side, appeal is made to Hume’s
writings as a whole – including the Treatise,
Essays, Enquiries, Dissertations, History
and Dialogues – which display a clear com-
mitment to induction, and even reveal their
author to be a fervent advocate of inductive
science. The evidence on each side is then
judiciously weighed, and an appropriate con-
clusion drawn depending on which way the
balance falls. But this whole procedure is mis-
directed, because once we recognize the varie-
ties of scepticism, it becomes clear that these
two bodies of evidence are not in conflict.
1. A SCEPTICAL ARGUMENT, WITH
A NON-SCEPTICAL OUTCOME
In this chapter, I shall maintain that Hume’s
argument concerning induction is indeed a
sceptical argument, in the sense of showing
that inductive extrapolation from observed
to unobserved lacks any independent rational
warrant. To avoid any misunderstanding
on the way, however, it will help to be clear
from the start that this is entirely compatible
with his wholehearted endorsement of such
extrapolation as the only legitimate method
2/9/2012 9:38:48 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
for reaching conclusions about ‘any matter
of fact, which lies beyond the testimony of
sense or memory’ (EHU 12.22 / 159). The
two may initially seem incompatible, but if
so, this is because we are taking for granted
that a method of inference is to be relied upon
only if it can be given an independent rational
warrant. And one of the central messages of
Hume’s philosophy is that this assumption is
itself a rationalist prejudice that we should
discard, even though it is shared by both
the Cartesian dogmatist and the extreme
‘Pyrrhonian’ sceptic. In the contest between
those two extremes, the Pyrrhonist ‘seems
to have ample matter of triumph’ while he
‘justly’ urges Hume’s own ‘sceptical doubts’ of
Enquiry 4 (the famous argument which is then
summarized at EHU 12.22 / 159). However,
the appropriate response, as Hume himself
explains, is not to follow the dogmatist in
vainly attempting to challenge the argument
that yields these doubts, but rather to ask the
Pyrrhonist: ‘What his meaning is? And what
he proposes by all these curious researches?’
(EHU 12.23 / 159). What, after all, does he
really expect us to do in response to this scep-
tical argument, even if we fully accept it? Is
he seriously proposing that we should stop
drawing inferences about the unobserved?
That would obviously be absurd:
a Pyrrhonian . . . must acknowledge, if
he will acknowledge any thing, that all
human life must perish, were his princi-
ples universally and steadily to prevail.
All discourse, all action would immedi-
ately cease; and men remain in a total
lethargy, till the necessities of nature,
unsatisfied, put an end to their miserable
existence. (EHU 12.23 / 160)
Theoretically, the Pyrrhonist might try to
deny any such disastrous consequences, on
the ground that if induction is unwarranted,
58
9780826443595_Ch03_Final_txt_print.indd 58
then we have no good reason for supposing
that human life will indeed perish in these cir-
cumstances. But Hume suggests that even the
Pyrrhonist – whatever his theoretical commit-
ments – will be quite unable to insulate him-
self from such common-sense beliefs: ‘Nature
is always too strong for principle. . . . the first
and most trivial event in life will put to flight
all his doubts and scruples, and leave him the
same, in every point of action and specula-
tion’ with the rest of us (EHU 12.23 / 160).
Hume cannot, of course, prove that
putting total scepticism into practice will lead
inevitably to disaster, at least not to the sat-
isfaction of the Pyrrhonist who consistently
refrains from induction. Nor can he prove
that common life will always trump scepti-
cal principle. But if in fact Hume’s inductive
conclusions about human psychology are
correct, then he does not need to prove these
points to any such opponent:
Nature, by an absolute and uncontroula-
ble necessity has determin’d us to judge
as well as to breathe and feel; nor can
we any more forbear [making inductive
inferences], than we can hinder ourselves
from thinking as long as we are awake,
or seeing the surrounding bodies, when
we turn our eyes towards them in broad
sunshine. Whoever has taken the pains to
refute the cavils of this total scepticism,
has really disputed without an antagonist,
and endeavour’d by arguments to estab-
lish a faculty, which nature has anteced-
ently implanted in the mind, and render’d
unavoidable. (THN 1.4.1.7 / 183)
So if in fact the sceptic’s doubts will be spon-
taneously ‘put to flight’ as soon as common
life intrudes, then Hume’s point is practic-
ally successful even if theoretically unproved.
And recall again that Hume himself need
not be committed to accepting only what is
2/9/2012 9:38:49 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
theoretically provable – that is the very preju-
dice which he is aiming to undermine.
Hume’s subtle approach to scepticism
is made harder to appreciate by the vigour
and rhetoric of some of his negative argu-
ments and conclusions (especially in the
Treatise, where his ultimate position on
scepticism remains relatively obscure), but
also, I suspect, by the widespread tradition
of approaching scepticism initially through
Descartes’s Meditations. Descartes sees the
sceptic as an opponent to be refuted out-
right, through rational argument of such
overwhelming force as to be immune to any
possible doubt. He thus takes on the onus of
providing an ultimate justification of human
reason, with any ineradicable doubt telling
in favour of his sceptical opponent. Hume
succinctly points out the fundamental flaw
in this approach immediately after having
raised the question ‘What is meant by a scep-
tic?’ at the beginning of Enquiry Section 12:
There is a species of scepticism, anteced-
ent to all study and philosophy, which
is much inculcated by Des Cartes . . .
It recommends an universal doubt . . .
of our very faculties; of whose verac-
ity . . . we must assure ourselves, by a
chain of reasoning, deduced from some
original principle, which cannot possibly
be fallacious or deceitful. But neither is
there any such original principle, which
has a prerogative above others, that are
self-evident and convincing: Or if there
were, could we advance a step beyond
it, but by the use of those very faculties,
of which we are supposed to be already
diffident. The Cartesian doubt, there-
fore, were it ever possible to be attained
by any human creature (as it plainly is
not) would be entirely incurable; and no
reasoning could ever bring us to a state
of assurance and conviction upon any
subject. (EHU 12.3 / 149–50)
59
9780826443595_Ch03_Final_txt_print.indd 59
Such antecedent scepticism is utterly unwork-
able, because in refusing to trust our facul-
ties from the start, we are denying ourselves
the only tools that could possibly provide
any solution. The proper alternative, Hume
seems to be saying, is to accord our faculties
some initial default authority, and to resort
to practical scepticism about them only ‘con-
sequent to science and enquiry’, in the event
that those investigations reveal their ‘falla-
ciousness’ or ‘unfitness’ (EHU 12.5 / 150).
Thus the onus is shifted onto the sceptic to
give reasons for mistrusting our faculties,
and in the case of induction, that onus is at
most only partially fulfilled. Admittedly,
The sceptic . . . seems to have ample mat-
ter of triumph; while he justly insists,
that all our evidence for any matter of
fact, which lies beyond the testimony of
sense or memory, is derived entirely from
the relation of cause and effect; that we
have no other idea of this relation than
that of two objects, which have been
frequently conjoined together;2 that we
have no argument to convince us, that
objects, which have, in our experience,
been frequently conjoined, will likewise,
in other instances, be conjoined in the
same manner; and that nothing leads us
to this inference but custom or a certain
instinct of our nature; which it is indeed
difficult to resist, but which, like other
instincts, may be fallacious and deceitful.
(EHU 12.22 / 159)
But this result – as we have seen – gives no
practical basis for scepticism. Certainly it
raises a ground for theoretical concern, and
highlights ‘the whimsical condition of man-
kind, who must act and reason and believe;
though they are not able, by their most
diligent enquiry, to satisfy themselves con-
cerning the foundation of these operations’
2/9/2012 9:38:49 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
(EHU 12.23 / 160). But unless we are in the
grip of the rationalist prejudice that Hume
rejects, we should not see this lack of theoret-
ical satisfaction as sufficient reason to aban-
don our only respectable method of inference
about the unobserved. That would be to take
the sceptical considerations to a ridiculous
(and anyway unachievable) extreme. Instead,
the appropriate response is less dramatic but
far more valuable: to recognize our ‘whimsi-
cal condition’ as a ground for modesty about
the depth and extent of our powers, and to
adopt a ‘mitigated scepticism’ which is cor-
respondingly diffident and cautious (EHU
12.24 / 161–2), and which confines our atten-
tion to the subjects of common life, ‘avoiding
distant and high enquiries’:
While we cannot give a satisfactory rea-
son, why we believe, after a thousand
experiments, that a stone will fall, or
fire burn; can we ever satisfy ourselves
concerning any determination, which we
may form, with regard to the origin of
worlds, and the situation of nature, from,
and to eternity? (EHU 12.25 / 162)
This sentence is Hume’s last word on the
question of inductive scepticism, and rep-
resents the conclusion of a coherent line of
thought which can be traced from the begin-
ning of Enquiry Section 12, his most clear and
explicit – and repeatedly refined – treatment
of scepticism as a whole. So far, then, we have
a clear outline of his mature position.
2. HUME’S SCEPTICAL ARGUMENT
The main aim of this chapter is to understand
the logic and significance of Hume’s famous
argument, and in particular its implications
60
9780826443595_Ch03_Final_txt_print.indd 60
for his notion of ‘reason’ and for the rational
status of inductive inference. These issues
are far from straightforward, partly because
the argument appears three times in Hume’s
works, with many differences between the
three presentations – some of them highly
significant – and clear evidence of a system-
atic development in his views. But for our
purposes, it will be enough here just to high-
light the most salient points.
2.1 THE ARGUMENT OF THE TREATISE
In the Treatise, the famous argument occurs
within the context of Hume’s rather ram-
bling search for the origin of the idea of
necessary connexion, which he has previ-
ously (THN 1.3.2.11 / 77) identified as the
key component of our idea of causation. Not
having ‘any certain view or design’ on how
to trace the impression(s) that could account
for this crucial idea, he sets off to ‘beat about
all the neighbouring fields’ in the hope that
something will turn up (THN 1.3.2.12–13
/ 77–8). His first such ‘field’ concerns the
basis of the Causal Maxim ‘that whatever
begins to exist, must have a cause of exist-
ence’ (THN 1.3.3.1 / 78), but after conclud-
ing that this Maxim cannot be ‘intuitively or
demonstratively certain’,3 he quickly moves
on to a related question, ‘Why we conclude,
that such particular causes must necessar-
ily have such particular effects, and why we
form an inference from one to another?’
(THN 1.3.3.8–9 / 82). He soon narrows his
focus onto what he considers the paradigm
case of a causal inference, from a sensory
impression of one ‘object’ (for example, we
see a flame), to forming a belief – a lively
idea – of its effect or cause (for example, we
expect heat). He then analyses such an infer-
ence into its component parts: ‘First, The ori-
ginal impression. Secondly, The transition
2/9/2012 9:38:49 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
to the idea of the connected cause or effect.
Thirdly, The nature and qualities of that
idea.’ (THN 1.3.5.1 / 84). The remainder of
Section 1.3.5 discusses the first component,
then 1.3.6, entitled ‘Of the Inference from the
Impression to the Idea’, comes to the second
component, the causal inference itself.4
Hume’s first move, in discussing this
paradigm causal inference, is to insist that it
cannot be made a priori, simply from obser-
vation of the cause:
There is no object, which implies the
existence of any other if we consider
these objects in themselves, and never
look beyond the ideas which we form of
them. Such an inference . . . wou’d imply
the absolute contradiction and impossi-
bility of conceiving any thing different.
But as all distinct ideas are separable,
’tis evident there can be no impossibil-
ity of that kind. When we pass from a
present impression to the idea of any
object, we might possibly have separated
the idea from the impression, and have
substituted any other idea in its room.
(THN 1.3.6.1 / 86–7)
Here Hume is appealing to the principle that
if an inference is to be a priori, there must
be an absolute contradiction and impossi-
bility of conceiving things as turning out
differently: an a priori inference has to yield
total certainty. He also seems to be taking for
granted that such a contradiction in concep-
tion implies a contradiction in fact, which is
closely related to his Conceivability Principle
that whatever we conceive is possible (this
makes a more explicit entrance shortly, at
THN 1.3.6.5 / 89). Note also his appeal to
what is commonly called his Separability
Principle, that ‘all distinct ideas are separ-
able’ (cf. THN 1.1.3.4 / 10, 1.1.7.3 / 18–19,
1.3.3.3 / 79–80), which plays a major role
61
9780826443595_Ch03_Final_txt_print.indd 61
in the Treatise but disappears from his later
writings.5
Hume has now established one of the
most important results of his philosophy:
‘’Tis . . . by experience only, that we can
infer the existence of one object from that of
another.’ (THN 1.3.6.2 / 87). And he imme-
diately goes on to explain that the kind of
experience which prompts such a causal
inference is repeated conjunctions of pairs of
‘objects . . . in a regular order of contiguity
and succession’. Where we have repeatedly
seen A closely followed by B, ‘we call the
one cause and the other effect, and infer the
existence of the one from that of the other’.
Hume enthusiastically trumpets this relation
of constant conjunction as the sought-for key
to the crucial notion of necessary connexion,
with a clear allusion back from THN 1.3.6.3
/ 87 to 1.3.2.11 / 77,6 and he celebrates the
progress of his rambling journey of discov-
ery. Admittedly there is still some way to go,
because mere repetition of conjunctions does
not seem to generate ‘any new original idea,
such as that of a necessary connexion’. But
the line of investigation seems clear:
having found, that after the discovery of
the constant conjunction of any objects,
we always draw an inference from one
object to another, we shall now examine
the nature of that inference . . . Perhaps
’twill appear in the end, that the neces-
sary connexion depends on the inference,
instead of the inference’s depending on the
necessary connexion. (THN 1.3.6.3 / 88)
This last sentence provides an elegant
epitome of the link between Hume’s theories
of induction and causation, anticipating the
eventual outcome of his quest for the elusive
impression of necessary connexion (which
will come much later, at THN 1.3.14.20 /
164–5). For present purposes, however, we
2/9/2012 9:38:49 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
can forget about that quest, and focus on the
nature of inductive inference.
Having established that causal inference
‘from the impression to the idea’ (e.g. from
seeing A to expecting B) depends on experi-
ence, Hume goes on to pose the central ques-
tion that his argument aims to answer, namely
which mental faculty is responsible for the
inference: ‘the next question is, Whether
experience produces the idea by means of the
understanding or imagination; whether we
are determin’d by reason to make the tran-
sition, or by a certain association and rela-
tion of perceptions?’ (THN 1.3.6.4 / 88–9).
If the faculty of reason were responsible,
Hume says, this would have to be on the basis
of an assumption of similarity between past
and future, commonly called his Uniformity
Principle: ‘If reason determin’d us, it wou’d
proceed upon that principle, that instances, of
which we have had no experience, must resem-
ble those, of which we have had experience,
and that the course of nature continues always
uniformly the same.’ (THN 1.3.6.4 / 89). So
the next stage is to see whether there is any
argument by which reason could establish
this principle, and if there is not, then Hume
will conclude that reason cannot be the basis
for our inductive inferences.
Following the standard categoriza-
tion deriving from John Locke,7 just two
types of argument are potentially available,
demonstrative and probable, and Hume
now eliminates each in turn. First, demon-
strative arguments proceed with absolute
certainty based on self-evident (‘intuitive’)
relationships between the ideas concerned;
these sorts of argument are capable of yield-
ing ‘knowledge’ in the strict sense, and are
mostly confined to mathematics.8 But no such
argument can possibly prove the Uniformity
Principle, because that would mean the prin-
ciple is absolutely guaranteed, which the
62
9780826443595_Ch03_Final_txt_print.indd 62
Conceivability Principle shows it cannot be:
‘We can at least conceive a change in the
course of nature; which sufficiently proves,
that such a change is not absolutely impos-
sible. To form a clear idea of any thing, is an
undeniable argument for its possibility, and
is alone a refutation of any pretended dem-
onstration against it.’ (THN 1.3.6.5 / 89). As
for probable arguments (that is, arguments in
which we draw conclusions – typically about
things in the world of our everyday experi-
ence – with less than total certainty), these
must be based on causal relations, because
causation is ‘The only . . . relation of objects
. . . on which we can found a just inference
from one object to another’ (THN 1.3.6.7 /
89).9 But Hume has just argued that causal
inference is ‘founded on the presumption of a
resemblance betwixt those objects, of which
we have had experience, and those, of which
we have had none’ (an argument that he reca-
pitulates at THN 1.3.6.6–7 / 89–90, echoing
the discussion of THN 1.3.4.1–4 / 82–9).10
And since probable inference relies on causal
relations, ‘’tis impossible this presumption
[of the Uniformity Principle] can arise from
probability’, on pain of circularity.11 So nei-
ther demonstrative nor probable arguments
can provide any solid basis for the Uniformity
Principle, and Hume quickly concludes that
reason cannot be responsible for causal
inference:12
Thus not only our reason fails us in the
discovery of the ultimate connexion of
causes and effects, but even after experi-
ence has inform’d us of their constant con-
junction, ’tis impossible for us to satisfy
ourselves by our reason, why we shou’d
extend that experience beyond those
particular instances, which have fallen
under our observation. We suppose, but
are never able to prove, that there must
be a resemblance betwixt those objects,
2/9/2012 9:38:49 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
of which we have had experience, and
those which lie beyond the reach of our
discovery. (THN 1.3.6.11 / 91–2)
Instead, such inference must derive from asso-
ciative principles in the imagination (THN
1.3.6.12 / 92), and in particular, from a mech-
anism which Hume calls custom (e.g. THN
1.3.7.6 / 97, 1.3.8.10 / 102) or habit (e.g.
THN 1.3.10.1 / 118). Experience of constant
conjunction between A and B establishes an
associative connexion between them, mak-
ing our mind habitually move easily from
the idea of one to the idea of the other. When
we then see an A, the ‘force and vivacity’ of
that sense impression is transferred through
the associative link to our idea of B, enliven-
ing it into a belief. Hume accordingly goes on
to define a belief as ‘a lively idea related
to or associated with a present impres-
sion’ (THN 1.3.7.5 / 96), and to expand on
this theory of belief formation over the sub-
sequent sections.
2.2 FROM THE TREATISE TO THE ABSTRACT
Given the fame that it has subsequently
enjoyed, Hume’s argument in Treatise 1.3.6
is surprisingly inconspicuous. It occurs within
a detour (at THN 1.3.3.9 / 82) from a ram-
ble through fields (THN 1.3.2.13 / 77–8); the
core of it occupies only six fairly short para-
graphs (1–2 / 86–7 and 4–7 / 88–90); and its
primary role seems to be to identify custom
as the ground of causal belief – as a compo-
nent in Hume’s larger theory of belief – rather
than to emphasize its own apparently scep-
tical conclusion. He does later remark on the
striking nature of this conclusion:13
Let men be once fully perswaded of these
two principles, that there is nothing in
any object, consider’d in itself, which can
63
9780826443595_Ch03_Final_txt_print.indd 63
afford us a reason for drawing a conclu-
sion beyond it; and, that even after the
observation of the frequent or constant
conjunction of objects, we have no rea-
son to draw any inference concerning any
object beyond those of which we have had
experience; . . . and this will throw them
so loose from all common systems, that
they will make no difficulty of receiving
any, which may appear the most extraor-
dinary. (THN 1.3.12.20 / 139)
But this again is within a context where his
aim is to develop his theory of belief, now
focusing on inferences involving probability
where the relevant past conjunctions are not
constant.
Books 1 and 2 of the Treatise were pub-
lished at the end of January 1739, but well
before the end of that year, Hume seems to
have radically reassessed the significance of
his philosophy. By then he had written his
Abstract of the Treatise, which appeared
in print in March 1740, and which devotes
8 paragraphs out of 35 (paragraphs 8 and
10–16) to the famous argument. From being
a very small part of a much larger system,
suddenly it becomes the prime focus of
his philosophy, as it remained in the first
Enquiry of 1748, which can indeed be seen
as mainly constructed around the argument
and its implications.
The declared purpose of the argument in
the Abstract is to understand ‘all reasonings
concerning matter of fact’ (Abs. 8 / 649),
rather than limiting discussion to the para-
digm case of a causal inference – ‘the infer-
ence from the impression to the idea’ – which
had been the topic of Treatise 1.3.6. But
Hume then immediately states that all such
factual reasonings (to coin a shorthand term)
‘are founded on the relation of cause and
effect’, thus making clear that causal infer-
ence is still the focus. However, this initial
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
move is helpful in both emphasizing the gen-
erality of the argument and also streamlining
it, avoiding the need for the recapitulation of
his treatment of causal reasoning which had
occupied THN 1.3.6.6–7 / 89–90. Now, in
proving that all causal reasoning presupposes
the Uniformity Principle, he will have proved
at the same time that ‘all reasoning concern-
ing matter of fact’ – and hence all probable
reasoning – has such a dependence.14
To facilitate discussion, Hume introduces
the simple example of one billiard ball strik-
ing another and causing it to move (Abs. 9–10
/ 649–50). He then presents a vivid thought-
experiment, imagining the first man Adam,
newly created by God, and confronted with
such an imminent collision:
without experience, he would never be
able to infer motion in the second ball
from the motion and impulse of the
first. It is not any thing that reason sees
in the cause, which makes us infer the
effect. Such an inference, were it possi-
ble, would amount to a demonstration,
as being founded merely on the compari-
son of ideas. But no inference from cause
to effect amounts to a demonstration.
Of which there is this evident proof. The
mind can always conceive any effect to
follow from any cause, and indeed any
event to follow upon another: what-
ever we conceive is possible, at least in a
metaphysical sense: but wherever a dem-
onstration takes place, the contrary is
impossible, and implies a contradiction.
There is no demonstration, therefore,
for any conjunction of cause and effect.
(Abs. 11 / 650–1)
Compared to the equivalent passage in the
Treatise (THN 1.3.6.1 / 86–7), this is clearer
and more straightforward, proving by dir-
ect appeal to the Conceivability Principle
a general lesson which he states even more
64
9780826443595_Ch03_Final_txt_print.indd 64
forthrightly elsewhere: ‘that to consider the
matter a priori, any thing may produce any
thing’ (THN 1.4.5.30 / 247, cf. 1.3.15.1 /
173, EHU 12.29 / 164).
So experience is necessary to ground any
causal inference (and hence any inference
‘concerning matter of fact’). And Hume goes
on to explain that the type of experience rele-
vant to his thought-experiment would be of
‘several instances’ (Abs. 12 / 651) in which
Adam saw the collision of one ball into
another followed by motion in the second
ball. Such experience would condition him
‘to form a conclusion suitable to his past
experience’, and thus to expect more of the
same. ‘It follows, then, that all reasonings
concerning cause and effect, are founded
on experience, and that all reasonings from
experience are founded on the supposition,
that the course of nature will continue uni-
formly the same.’ (Abs. 13 / 651). So as in
the Treatise, we reach Hume’s Uniformity
Principle, and he now proceeds accordingly
to consider what rational basis this principle
could be given:
’Tis evident, that Adam . . . would never
have been able to demonstrate, that the
course of nature must continue uni-
formly the same, and that the future
must be conformable to the past. What
is possible can never be demonstrated to
be false; and ’tis possible the course of
nature may change, since we can con-
ceive such a change. (Abs. 14 / 651)
As in the Treatise, we have an appeal to
the Conceivability Principle to show that a
change in the course of nature is possible,
which in turn implies that uniformity cannot
be demonstrated.
Nay, . . . [Adam] could not so much as
prove by any probable arguments, that the
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
future must be conformable to the past.
All probable arguments are built on the
supposition, that there is this conformity
betwixt the future and the past, and there-
fore can never prove it. This conformity is
a matter of fact, and if it must be proved,
will admit of no proof but from experi-
ence. But our experience in the past can
be a proof of nothing for the future, but
upon a supposition, that there is a resem-
blance betwixt them. This therefore is a
point, which can admit of no proof at all,
and which we take for granted without
any proof. (Abs. 14 / 651–2)
Here the logical circularity of attempting to
give a probable argument for the Uniformity
Principle is more explicitly spelled out than
in the Treatise. With both demonstrative and
probable argument eliminated, Hume briskly
concludes that ‘We are determined by custom
alone to suppose the future conformable to the
past. . . . ’Tis not, therefore, reason, which is the
guide of life, but custom.’ (Abs. 15–16 / 652).15
2.3 THE ARGUMENT OF THE ENQUIRY
In the Enquiry concerning Human
Understanding of 1748, the famous negative
argument occupies virtually all of Section 4,
with the positive account in terms of custom
appearing in Section 5. Compared with the
versions in the Treatise and Abstract, the
argument is clarified and greatly expanded,
leaving little doubt that Hume considers this
his definitive presentation.
Section 4 starts with an important dis-
tinction now commonly known as ‘Hume’s
Fork’, between relations of ideas – that is,
propositions (notably from mathematics)
that can be known to be true a priori, just by
examining and reasoning with the ideas con-
cerned – and matters of fact – that is, prop-
ositions whose truth or falsehood depends on
65
9780826443595_Ch03_Final_txt_print.indd 65
how the world is, and so can be known (if at
all) only through experience. Some matters of
fact we learn directly by perception, and can
later recall.16 But what of the rest? Hume sets
himself to address this key question: ‘what
is the nature of that evidence, which assures
us of any real existence and matter of fact,
beyond the present testimony of our senses,
or the records of our memory’ (EHU 4.3 /
26)? On what basis do we infer from what we
perceive and remember, to conclusions about
further, unobserved, matters of fact?
Hume calls such inferences ‘reasonings
concerning matter of fact’ (EHU 4.4 / 26),
a term we saw introduced just once in the
Abstract but which now becomes his stand-
ard way of referring to what he had pre-
viously called ‘probable arguments’. The
reason for this terminological adjustment
seems to be to avoid the infelicity of calling
such inferences merely ‘probable’ even when
they are based on vast and totally uniform
past experience that yields complete ‘moral
certainty’ (that is, practical assurance). In a
footnote to the heading of Section 6, Hume
will accordingly draw a distinction – within
the class of ‘reasonings concerning matter of
fact’ – between probabilities and proofs, the
latter being ‘such arguments from experience
as leave no room for doubt or opposition’, as
when we conclude that ‘all men must die, or
that the sun will rise to-morrow’.17
In Enquiry 4, the famous argument now
proceeds much as it had in the Abstract, albeit
greatly filled out. The appendix to this chap-
ter lays out a structure diagram involving 20
stages,18 with the stages numbered accord-
ing to the logic of the argument. The same
numbers will be followed here, within square
brackets, to enable easy cross-referencing.
First, we learn that [2] ‘All reasonings concern-
ing matter of fact seem to be founded on the
relation of Cause and Effect’ (EHU 4.4 / 26),
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
since [1] ‘By means of that relation alone we
can go beyond the evidence of our memory
and senses.’ As in the Abstract, starting in this
way has the virtue of streamlining the argu-
ment that follows, so that conclusions to be
drawn about causal reasoning will automatic-
ally apply to the entire class of factual reason-
ing. The first of these conclusions, as before, is
that [5] all knowledge of causal relations must
be founded on experience: ‘the knowledge
of this relation [i.e. causation] is not, in any
instance, attained by reasonings a priori; but
arises entirely from experience, when we find,
that any particular objects are constantly con-
joined with each other.’ (EHU 4.6 / 27). Again
we get a thought-experiment involving Adam,
but this time with water and fire, illustrat-
ing the general truth that [3] ‘No object ever
discovers [i.e. reveals], by the qualities which
appear to the senses, either the causes which
produced it, or the effects which will arise
from it’. This is relatively easy to see when the
phenomena are untypical or unfamiliar, such
as the unexpected adhesion between smooth
slabs of marble, the explosion of gunpow-
der, or the powers of a (magnetic) lodestone,
where we have no temptation to imagine
that we could have predicted these effects
in advance (EHU 4.7 / 28). But with com-
monplace occurrences, such as the impact of
billiard balls (EHU 4.8 / 28–9), we might sup-
pose that the effect was foreseeable a priori.
To prove that this is an illusion, Hume asks us
to imagine how we could possibly proceed to
make such an a priori inference, arguing that
we could not, on the grounds that the effect is
a quite distinct event from the cause (EHU 4.9
/ 29), while many different possible effects
are equally conceivable (EHU 4.10 / 29–30).
Summing up [4]:
every effect is a distinct event from
its cause. It could not, therefore, be
66
9780826443595_Ch03_Final_txt_print.indd 66
discovered in the cause, and the first
invention or conception of it, a priori,
must be entirely arbitrary. And even after
it is suggested, the conjunction of it with
the cause must appear equally arbitrary;
since there are always many other effects,
which, to reason, must seem fully as con-
sistent and natural. (EHU 4.11 / 30)
Note the strong emphasis on arbitrariness,
making clear that it is not just the conceiv-
ability – or mere theoretical possibility – of
alternative outcomes which makes any a pri-
ori inference from cause to effect impossible;
it is the fact that from an a priori point of
view, there is nothing to suggest one outcome
over another.19
If causal relations cannot be known a pri-
ori, then factual inference cannot be a pri-
ori either (given [2] that factual inference is
founded on causation). [6] ‘In vain, therefore,
should we pretend to determine any single
event . . . without the assistance of observation
and experience.’ Hume now brings Part 1 of
Section 4 to a close, with two very important
corollaries for his philosophy of science. The
first is that since we cannot aspire to a priori
insight into why things work as they do, the
appropriate ambition for science is instead
to aim more modestly for systematization
of those cause and effect relationships that
experience reveals: ‘to reduce the princi-
ples, productive of natural phænomena, to
a greater simplicity, and to resolve the many
particular effects into a few general causes,
by means of reasonings from analogy, experi-
ence, and observation.’ (EHU 4.12 / 30). Then
follows Hume’s most explicit account of
applied mathematics (which he calls ‘mixed
mathematics’), emphasizing that although
mathematical relationships are a priori, the
laws through which they are applied to the
world – his example is the Newtonian law of
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
conservation of momentum – remain unam-
biguously a posteriori: ‘the discovery of the
law itself is owing merely to experience, and
all the abstract reasonings in the world could
never lead us one step towards the know-
ledge of it’ (EHU 4.13 / 31).20
Part 2 starts by summarizing Hume’s
results so far, and anticipating his eventual
conclusion [20]:
When it is asked, What is the nature
of all our reasonings concerning mat-
ter of fact? the proper answer seems to
be, that they are founded on the rela-
tion of cause and effect. When again it
is asked, What is the foundation of all
our reasonings and conclusions concern-
ing that relation? it may be replied in one
word, Experience. But if we still carry
on our sifting humour, and ask, What is
the foundation of all conclusions from
experience? this implies a new question
. . . I shall content myself, in this section,
with an easy task, and shall pretend [i.e.
aspire] only to give a negative answer to
the question here proposed. I say then,
that, even after we have experience of
the operations of cause and effect, our
conclusions from that experience are not
founded on reasoning, or any process of
the understanding. (EHU 4.14–15 / 32)
Having established that experience is required
for any factual inference, Hume goes on to
explain how experience plays that role:
we always presume, when we see like sen-
sible qualities, that they have like secret
powers,21 and expect, that effects, simi-
lar to those which we have experienced,
will follow from them. . . . But why [past]
experience should be extended to future
times, and to other objects, which for
aught we know, may be only in appear-
ance similar; this is the main question . . .
(EHU 4.16 / 33–4; emphasis added)
67
9780826443595_Ch03_Final_txt_print.indd 67
This passage seems to be saying that [7] when
we draw conclusions from past experience,
we presuppose a resemblance between the
observed and the unobserved, extrapolating
from one to the other.22 Later, when appar-
ently referring back to this passage, Hume
confirms such a reading: ‘We have said, . . .
that all our experimental conclusions pro-
ceed upon the supposition, that the future
will be conformable to the past.’ (EHU 4.19
/ 35). So his ‘main question’ at EHU 4.16 /
34 concerns, in effect, the foundation of the
Uniformity Principle.23 He repeats (cf. EHU
4.6 / 27) that [3] ‘there is no known con-
nexion between the sensible qualities and
the secret powers’ of any object, and infers
from this that [9] ‘the mind is not led to form
such a conclusion concerning their con-
stant and regular conjunction, by any thing
which it knows of their nature’ (EHU 4.16
/ 33). So the Uniformity Principle cannot be
established on the basis of anything that we
learn directly through sense perception, in
which case [10] any foundation for it will
have to draw on past experience, which for
the sake of the argument can here be taken
as infallible: ‘As to past Experience, it can be
allowed to give direct and certain informa-
tion of those precise objects only, and that
precise period of time, which fell under its
cognizance . . . . (EHU 4.16 / 33). The ‘main
question’ is then urged: how to justify the
step from past experience to the assumption
of future resemblance?
These two propositions are far from
being the same, I have found that such
an object has always been attended with
such an effect, and I foresee, that other
objects, which are, in appearance, simi-
lar, will be attended with similar effects.
I shall allow, if you please, that the one
proposition may justly be inferred from
the other: I know in fact, that it always is
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
inferred. But if you insist, that the infer-
ence is made by a chain of reasoning, I
desire you to produce that reasoning.
The connexion between these propos-
itions is not intuitive. There is required
a medium, which may enable the mind
to draw such an inference, if indeed it
be drawn by reasoning and argument.
(EHU 4.16 / 34)
So because [11] the inference from past
experience to future resemblance is not intui-
tive (i.e. not immediately self-evident), [12]
there must be some medium, some ‘connect-
ing proposition or intermediate step’ (EHU
4.17 / 34) if indeed the inference is ‘drawn by
reasoning and argument’.24
The long paragraph that we have just
been discussing (EHU 4.16 / 32–4) includes
steps that have no parallel in the Treatise
and Abstract, where, as we saw, Hume sim-
ply takes for granted that if the Uniformity
Principle is to be rationally well founded,
then this must be on the basis of some chain
of reasoning, either demonstrative or prob-
able. Here in the Enquiry, he explicitly rules
out both sense experience and intuition as
sources of foundation for the Uniformity
Principle, and only then comes to consider
demonstration and probability, which are in
turn dismissed in the familiar way, but again
with the structure of the argument made
somewhat more explicit:
[13] All reasonings may be divided into
two kinds, namely demonstrative reason-
ing, or that concerning relations of ideas,
and moral reasoning, or that concerning
matter of fact and existence.25 [15] That
there are no demonstrative arguments
in the case, seems evident; since [14] it
implies no contradiction, that the course
of nature may change, . . . Now what-
ever is intelligible, and can be distinctly
conceived, implies no contradiction, and
68
9780826443595_Ch03_Final_txt_print.indd 68
can never be proved false by any demon-
strative argument or abstract reasoning a
priori. (EHU 4.18 / 35)
As in the Treatise and Abstract, Hume appeals
to the Conceivability Principle, though
slightly differently: here he expresses it as the
principle that what is conceivable implies no
contradiction, rather than saying that what is
conceivable is possible.26 Moving on now to
probability:
[16] If we be, therefore, engaged by
arguments to put trust in past experi-
ence, and make it the standard of our
future judgment, these arguments must
be probable only, or such as regard mat-
ter of fact and real existence, . . . But
. . . there is no argument of this kind,
. . . We have said, that [2] all arguments
concerning existence are founded on the
relation of cause and effect; that [5] our
knowledge of that relation is derived
entirely from experience; and that [7] all
our experimental conclusions proceed
upon the supposition, that the future
will be conformable to the past. [17] To
endeavour, therefore, the proof of this
last supposition by probable arguments,
or arguments regarding existence, must
be evidently going in a circle, and taking
that for granted, which is the very point
in question. (EHU 4.19 / 35–6)
Note in passing how Hume just assumes
here some obvious inferences, linking [2]
with [5] to deduce that [6] all factual infer-
ences (‘probable arguments’, ‘arguments
concerning existence’) are founded on
experience, and then combining this with
[7] to deduce in turn that [8] all factual
inferences ‘proceed upon the supposition’
of the Uniformity Principle.27 He also now
leaves the reader to piece together the final
2/9/2012 9:38:50 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
stages of his argument.28 First, that since the
Uniformity Principle cannot be established
by either demonstrative or factual inference,
it follows that [18] there is no good argu-
ment for the Uniformity Principle. Secondly,
that therefore (given [12]),29 it follows that
[19] the Uniformity Principle cannot be
founded on reason, and finally, that since
[8] all factual inferences are founded on the
Uniformity Principle, it follows that [20] no
factual inference (i.e. no ‘reasoning concern-
ing matter of fact and existence’) is founded
on reason. Hume had anticipated this con-
clusion at EHU 4.15, quoted earlier:30 ‘I say
then, that, even after we have experience
of the operations of cause and effect, our
conclusions from that experience are not
founded on reasoning, or any process of the
understanding.’ (EHU 4.15 / 32). Also in the
following section – most of which is devoted
to sketching his theory of belief as based on
‘Custom or Habit’ (EHU 5.5 / 43) – Hume
refers back to this argument and states its
conclusion explicitly, once purely negatively
and once alluding to his positive theory:
‘we . . . conclude . . . in the foregoing sec-
tion, that, in all reasonings from experience,
there is a step taken by the mind, which is
not supported by any argument or process
of the understanding; . . .’ (EHU 5.2 / 41);
‘All belief of matter of fact or real existence
. . . [is due merely to] . . . a species of nat-
ural instincts, which no reasoning or pro-
cess of the thought and understanding is
able, either to produce, or to prevent.’ (EHU
5.8 / 46–7).
2.4 THE ESSENTIAL CORE OF HUME’S
SCEPTICAL ARGUMENT
We can now distil the essence of Hume’s
argument from these three different presen-
tations, into eight main stages:
69
9780826443595_Ch03_Final_txt_print.indd 69
(A) The argument concerns all inferences
to matters of fact that we have not
observed: what the Enquiry calls ‘rea-
sonings concerning matter of fact’ (here
factual inferences for short). Although
the Treatise version starts with a nar-
rower focus on causal inference ‘from
the impression to the idea’, it later
requires the lemma that all factual
inferences are based on causal relations
(stated at THN 1.3.6.7 / 89). So the
argument is improved both structur-
ally and philosophically by starting with
all factual inferences, as in the Abstract
and the Enquiry, and then deriving this
lemma as its first main stage (Abs. 8 /
649; EHU 4.4 / 26–7).
(B) Hume next argues that causal relations
cannot be known a priori, and hence
are discoverable only through experi-
ence (THN 1.3.6.1 / 86–7, Abs. 9–11
/ 649–51; EHU 4.6–11 / 27–30). This
is a major principle of his philosophy,
wielded significantly elsewhere (e.g.
THN 1.3.15.1 / 173, 1.4.5.30 / 247–8;
EHU 12.29 / 164).
(C) From this principle, together with the
lemma from (A), Hume concludes that
all factual inferences are founded on
experience, the relevant experience
being of those constant conjunctions
through which we discover causal rela-
tionships (THN 1.3.6.2 / 87, Abs. 12 /
651; EHU 4.16 / 33).
(D) Factual inferences thus involve extrapo-
lation from observed to unobserved,
based on an assumption of resem-
blance between the two. Initially in the
Treatise, Hume seems to suggest that
such an assumption of resemblance –
commonly called his Uniformity
Principle (UP) – would be necessarily
implicated only if reason were respon-
sible for the inference (THN 1.3.6.4 /
88–9). But his settled view, expressed in
all three works (see note 10 above), is
that UP is presupposed by all factual
2/9/2012 9:38:51 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
inferences,31 simply in virtue of their
taking for granted a resemblance
between observed and unobserved.
(E) Hume now proceeds to investigate criti-
cally the basis of UP itself. In the Treatise
(THN 1.3.6.4 / 88–9) and Abstract
(Abs. 14 / 651–2), he appears to assume
immediately that any foundation in reason
would have to derive from some demon-
strative (i.e. deductive) or probable (i.e.
factual) inference. In the Enquiry, how-
ever – which hugely expands this part of
the argument from the cursory treatment
in the earlier works – he considers demon-
strative and factual inference only after
first (EHU 4.16 / 32–4) explicitly ruling
out any foundation in sensory awareness
of objects’ powers, or in immediate intu-
ition (i.e. self-evidence).32
(F) Any demonstrative argument for UP is
ruled out because a change in the course
of nature is clearly conceivable and there-
fore possible (THN 1.3.6.5 / 89; Abs.
14 / 651–2, EHU 4.18 / 35). Any factual
argument for UP is ruled out because,
as already established at (D), such argu-
ments inevitably presuppose UP, and
hence any purported factual inference
to UP would be viciously circular (THN
1.3.6.7 / 89–90; Abs. 14 / 651–2; EHU
4.19 / 35–6).
(G) The upshot of this critical investigation
is that UP has no satisfactory founda-
tion in reason, though Hume expresses
this in various ways:
’tis impossible to satisfy ourselves by
our reason, why we shou’d extend
[our] experience beyond those par-
ticular instances, which have fallen
under our observation. We suppose,
but are never able to prove, that there
must be a resemblance betwixt those
objects, of which we have had experi-
ence, and those which lie beyond
the reach of our discovery. (THN
1.3.6.11 / 91–2)
70
9780826443595_Ch03_Final_txt_print.indd 70
This [resemblance between past and
future] is a point, which can admit of
no proof at all, and which we take for
granted without proof. (Abs. 14 / 652)
it is not reasoning which engages us
to suppose the past resembling the
future, and to expect similar effects
from causes, which are, to appear-
ance, similar. (EHU 4.23 / 39)
(H) Since UP is presupposed by all factual
inferences (D), and UP has no foundation
in reason (G), Hume finally concludes
that factual inference itself has no foun-
dation in reason. Again he expresses this
conclusion in various ways (and note
here the narrower focus of the Treatise on
causal inference ‘from the impression to
the idea’, as pointed out at (A) above):
When the mind . . . passes from the idea
or impression of one object to the idea
or belief of another, it is not determin’d
by reason (THN 1.3.6.12 / 92)
’Tis not, therefore, reason, which is the
guide of life, but custom. That alone
determines the mind . . . to suppose
the future conformable to the past.
However easy this step may seem,
reason would never, to all eternity, be
able to make it. (Abs. 16 / 652)
I say then, that, . . . our conclusions
from . . . experience are not founded
on reasoning, or any process of the
understanding. (EHU 4.15 / 32)
in all reasonings from experience, there
is a step taken by the mind, which is not
supported by any argument or process
of the understanding (EHU 5.2 / 41)
Note also that two of these quotations – from
Abs. 16 / 652 and EHU 5.2 / 41 – could just
as appropriately have been cited as illustra-
tions of (G), because both refer to that ‘step’
2/9/2012 9:38:51 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
which is precisely the presupposition of the
Uniformity Principle. Since factual inference
operates by extrapolation from past to future,
Hume takes it to be obvious that the founda-
tion of such inference must be the same as the
foundation of the principle of extrapolation.
Hence he does not consistently distinguish
between (G) and (H), making the last stages
of his argument less explicit than one might
wish (cf. the end of section 2.3 above).
3. THE NATURE OF HUME’S
SCEPTICAL CONCLUSION
Hume usually expresses the conclusion of his
famous argument in a way that seems to imply
some incapacity on the part of human reason.
The Uniformity Principle is something that we
‘are never able to prove’ (THN 1.3.6.11 / 92),
and which indeed ‘can admit of no proof at
all’ (Abs. 14 / 652). Because of this, ‘’tis impos-
sible to satisfy ourselves by our reason’ (THN
1.3.6.11 / 91) concerning the inferential step
from past to future, a step which ‘reason
would never, to all eternity, be able to make’
(Abs. 16 / 652) and ‘which is not supported by
any argument or process of the understand-
ing’ (EHU 5.2 / 41). Hume also frequently
uses similar terms within the argument itself,
when saying that various would-be proofs
of UP are impossible, refutable, circular or
lack any ‘just foundation’ (THN 1.3.6.5 / 89,
1.3.6.7 / 89–90, 1.3.6.10 / 91; EHU 4.18 / 35,
19 / 35–6, 21 / 37–8), denying that human
knowledge ‘can afford . . . an argument’ that
‘supports the understanding’ (EHU 4.17 / 34)
in reasoning from past to future, and conse-
quently denying that our factual inferences
‘are built on solid reasoning’ (THN 1.3.6.8
/ 90). In both the Treatise (see section 2.2
above) and Enquiry (see section 1), he later
71
9780826443595_Ch03_Final_txt_print.indd 71
glosses the conclusion of the argument in
apparently very negative terms, as showing
that ‘we have no reason’ to draw any factual
inference (THN 1.3.12.20 / 139), and that
‘we cannot give a satisfactory reason, why we
believe, after a thousand experiments, that a
stone will fall, or fire burn’ (EHU 12.25 / 162).
In this light, it seems entirely appropriate that
he should entitle Enquiry Section 4 ‘Sceptical
Doubts concerning the Operations of the
Understanding’, and describe it as appearing
to give the sceptic ‘ample matter of triumph’
(EHU 12.22 / 159).
As discussed earlier, however, the issue
of Hume’s inductive ‘scepticism’ is not so
straightforward, and it is far from clear that
he sees the acknowledged incapacity of rea-
son to ‘prove’ or ‘support’ the Uniformity
Principle as any sort of genuine problem.
Certainly he does not infer from it (either in
the Treatise, the Abstract or the Enquiry) that
induction is unreasonable in any pragmatic
sense. And indeed the line of thought sketched
in section 1 above, drawing on Section 12 of
the Enquiry, somewhat suggests that he con-
siders it inevitable that our most basic prin-
ciples of inference – precisely because they
are so basic – will lack any ultimate justifica-
tion beyond their fundamental place in our
mental economy. That being so, the central
upshot of Hume’s argument might be sim-
ply to identify the Uniformity Principle as a
basic principle of this kind, and the sceptical
flavour of his reasoning – in demonstrating
reason’s incapacity to prove UP – need not
carry over at all into the theory of human
inference that he draws from it. Nevertheless,
the sceptical flavour of the famous argu-
ment itself would remain, in denying UP a
source of rational support that more opti-
mistic philosophers might have expected it
to enjoy. And although the argument also
delivers the important positive principle that
2/9/2012 9:38:51 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
the Uniformity Principle is presupposed by
all factual inference (D), even in the Enquiry
we have to wait until the final section to see
this wielded as part of an effective theoretical
defence against the ‘Pyrrhonian’ sceptic.33
3.1 DEBATES ABOUT HUMEAN ‘REASON’
What we get much sooner, of course, and in
all three works, is Hume’s positive account of
how our inductive inferences operate through
custom or habit – what he calls in the title of
Enquiry Section 5 his ‘sceptical solution’ to the
earlier ‘sceptical doubts’. But as David Owen
observes, it seems odd to suppose that a psy-
chological account of how belief functions
could in any way ‘solve’ genuine epistemo-
logical doubts; Owen accordingly suggests
that the famous argument is itself best under-
stood as exclusively concerned with psycho-
logical mechanisms, and as having nothing
to do with ‘the warrant of probable reason-
ing or the justification of belief’.34 The argu-
ment’s conclusion, that factual inference is not
founded on reason, may initially seem unam-
biguously normative, but Owen interprets
‘reason’ here as signifying what he takes to
be a Lockean conception of the mechanism
by which human reasoning operates, namely,
through stepwise inference via intermediate
ideas. Thus he is able to read Hume’s conclu-
sion that factual inference ‘is not determin’d
by reason’ as purely descriptive: as denying
that we actually draw factual inferences in the
stepwise, mediated manner that Locke suppos-
es.35 The famous argument accordingly serves
to reject this Lockean conception of probable
reasoning in favour of the more immediate
and instinctive Humean model, thus providing
a contribution to empirical psychology, rather
than an exercise in sceptical epistemology.
Owen is by no means the first to read
Hume’s famous argument as designed to
72
9780826443595_Ch03_Final_txt_print.indd 72
reject some non-Humean notion of reason;
indeed this style of interpretation became
extremely popular in the 1980s. Before then,
the general image of Hume was of a highly
destructive sceptic, intending through his
famous argument to maintain that induct-
ive arguments lack all rational justifica-
tion. Barry Stroud wittily expressed what
he took to be Hume’s conclusion: ‘As far as
the competition for degrees of reasonable-
ness is concerned, all possible beliefs about
the unobserved are tied for last place.’36
Some of these extreme sceptical interpret-
ations – most influentially those of Antony
Flew37 and David Stove38 – took Hume to
be starting from the assumption of deduc-
tivism, that an inference is rationally justi-
fied only if it is logically guaranteed.39 But
deductivism sits very uneasily with the (fal-
lible but reasonable) empirical judgements
that abound within Hume’s contributions to
‘the science of human nature’, for example
his discussions of the passions, his Essays
on politics and economics, and his vari-
ous pieces on religion. The ‘wise man’ of
Enquiry 10.4 / 110, who ‘proportions his
belief to the [empirical] evidence’, clearly
cannot be a deductivist; hence Flew, in dis-
cussing ‘Of Miracles’, was forced to accuse
Hume of ‘flagrant and embarrassing’ incon-
sistency.40 Even more flagrantly inconsist-
ent, from this perspective, were the passages
in which Hume, after his famous argument
had apparently denied causal, factual infer-
ence a place within the realm of ‘reason’,
then quite explicitly treated it as one of rea-
son’s central operations, for example:41
with regard to reason . . . The only con-
clusion we can draw from the existence
of one thing to that of another, is by
means of the relation of cause and effect
(THN 1.4.2.47 / 212)
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
reason, in a strict and philosophical
sense, can have an influence on our con-
duct . . . by informing us of the existence
of something which is a proper object
of [a passion]; or when it discovers the
connexion of causes and effects, so as to
afford us means of exerting any passion.
(THN 3.1.1.12 / 459)
But such gross inconsistency was hard to
credit, and given Hume’s evident commit-
ment to inductive moral science (and to its
reasonableness in comparison with ‘super-
stition’), it seemed to most later scholars far
more plausible to interpret his famous argu-
ment not as genuinely sceptical, but instead
as a way of rejecting or undermining the
deductivist notion of reason on which it was
thought to be based (by revealing its com-
plete incapacity to underwrite any factual
inference). From this it would follow that
Hume must use the term ‘reason’ in at least
two distinct senses: one narrowly deductiv-
ist or ‘rationalistic’ notion within the famous
argument, and a broader, more ‘naturalistic’
notion elsewhere, such as in his discussions
of the passions and morality.
This anti-deductivist style of interpretation
was pioneered in 1975 by Tom Beauchamp
and Thomas Mappes,42 whose work was
quickly followed by numerous variations on
the theme.43 However it lost favour after I
and Don Garrett (independently, in 1995 and
1997)44 pointed out the implausibility, under
careful analysis, of reading Hume’s argument
as employing a deductivist notion of rea-
son. Deductivism proves very hard to square
with the argument’s logic,45 and it also seems
strange – if Hume’s purpose is to wield the
argument in order to reject the notion of reason
that it employs – that he should then continue
to assert its sceptical conclusion in the appar-
ently sincere terms we saw earlier.46 My own
73
9780826443595_Ch03_Final_txt_print.indd 73
proposed alternative was to see the argument
as presupposing a perceptual model of reason,
according to which we draw inferences through
the perception of evidential connexions.
This had the virtue of identifying a plausible
(and substantial) target of Hume’s argument,
namely, Locke’s view – complacently assumed
and stated in his Essay concerning Human
Understanding47 but never worked out in any
detail – that probable reasoning is founded
on the perception of probable connexions.
Against this, Hume’s rival model of probable
inference based on custom – and introduced
immediately after his famous argument had
refuted the alternative – stood out as a radi-
cal (and highly sceptical) departure. Moreover
the ubiquitous hold of the traditional percep-
tual view of reason, which goes back to the
ancients and was shared by Hume’s contem-
poraries, could help to make sense of his own
apparent assessment of the famous argument
as having a significant sceptical impact. A mere
denial that induction has deductive force, or
yields total certainty, would hardly be worthy
of notice in the wake of Locke.48 But denying
that induction is founded on perception of any
good reason whatever would have vastly more
sceptical significance.
Garrett’s approach was quite different,
and in context more radical. He insisted –
against the prevailing orthodoxy – that
Hume employs but a single sense of ‘reason’,
taking this to be Hume’s name for ‘the gen-
eral faculty of making inferences or produ-
cing arguments – just as it was for Locke’.49
Leaving aside the (debatable) attribution to
Locke,50 the obvious advantage of this inter-
pretation was precisely its lack of any need to
posit an ambiguity in ‘reason’:
Hume . . . [is] making a specific claim,
within cognitive psychology, about the
relation between our tendency to make
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
inductive inferences and our inferential/
argumentative faculty: he is arguing that
we do not adopt induction on the basis
of recognizing an argument for its reli-
ability, for the utterly sufficient reason
that there is no argument (‘reasoning’
or ‘process of the understanding’) that
could have this effect. . . . this does not
mean that inductive inferences are not
themselves instances of argumentation
or reasoning . . . His point is rather that
they are reasonings that are not them-
selves produced by any piece of higher
level reasoning: there is no argument
that could lead us to accept the conclu-
sion that inductive reasonings will be
reliable if we did not already accept that
conclusion in practice. Hence, in just
this sense, they are a class of ‘reasonings’
(inferences or arguments) that ‘reason’
(the faculty of making inferences or giv-
ing arguments) does not itself ‘determine’
(cause) us to make.51
Like Owen, Garrett saw Hume’s argu-
ment as an exercise in descriptive psych-
ology rather than normative epistemology,
delivering a result about the causation of
inductive inference rather than its ‘eviden-
tiary value’. However, this sat uneasily with
Garrett’s emphasis (as in the quotation
above) on the recognition of higher-level
arguments for the reliability of induction,
and under challenge, he modified his ori-
ginal reading of Hume’s conclusion to make
it more general. Here are the relevant parts
of the passage above, edited to reflect the
adjusted reading:52
Hume . . . [is] making a specific claim,
. . . about the underlying causal mechan-
ism that gives rise to inductive inferences:
namely, that it is not itself dependent on
any reasoning or inference. . . . His point
74
9780826443595_Ch03_Final_txt_print.indd 74
is . . . that [inductive inferences] are rea-
sonings which are not themselves caused
by any piece of reasoning (including, of
course, themselves). Inductive inferences
require that we bridge a gap between
observation and prediction, and for
someone not already disposed by nature
to bridge that gap, no argument for doing
so would be persuasive. Hence, . . .53
One surprising effect of this change was to
bring Garrett’s interpretation rather close to
Owen’s, because his detailed analysis of how
mediated inferences operate made them ipso
facto inferences that are ‘determin’d by rea-
son’. On this account, a demonstrative infer-
ence from A to D, mediated by the intuitive
connexions of A to B, B to C, and C to D, will
include as part of its processing the intermedi-
ate inference connecting B to D.54 This makes
the latter inference, according to Garrett, a
cause of the overall inference from A to D;
hence that overall inference is ‘determin’d by
reason’ in the sense of being caused by another
inference. Thus Garrett agrees with Owen
that Hume’s conclusion involves a denial
that induction proceeds by stepwise ratiocin-
ation. All this may seem somewhat artificial,
and increasingly distant from anything to be
found in Hume’s text, but it has the nice fea-
ture of accommodating a genuinely Humean
point – that probable inference is characteris-
tically immediate and instinctive rather than
mediated and reflective – within a framework
which, unlike Owen’s, avoids any need to treat
the notion of ‘reason’ that is operative in the
famous argument as one that Hume rejects.
Garrett has consistently urged this last point
against rival interpretations: that Hume’s
famous argument gives little internal clue that
he is employing some special notion of reason
which he aims to reject. And although both
Owen and I sought to mitigate the impact of
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
this criticism on our interpretations (by stress-
ing that we saw Hume’s response to the argu-
ment as changing not the scope of his notion
of reason, but rather its presumed method
of operation),55 the lack of any obvious and
deliberate ambiguity or equivocation on
Hume’s part has remained by far the strongest
weapon in Garrett’s armoury.
Garrett’s interpretation has also seemed
attractive for a more specious reason,
namely, the extent to which Hume expresses
his conclusion in terms of the impossibility
of founding induction on ‘reasoning’ (e.g.
THN 1.3.6.8 / 90; EHU 4.15 / 32, 4.16 / 34,
4.23 / 39), ‘proof’ (e.g. THN 1.3.6.11 / 92;
Abs. 14 / 651–2), or ‘argument’ (e.g. Abs. 15
/ 652; EHU 4.16–17 / 34–5, 4.21–3 / 38–9,
5.2 / 41). Today we read these terms as sig-
nifying complex inference involving inter-
mediate steps, but in Hume’s day they were
understood rather differently. Johnson’s dic-
tionary of 1756 tells us that ‘reasoning’ is
derived from ‘reason’, and defines it simply
as ‘argument’.56 The first sense of ‘argument’
is given as ‘A reason alleged for or against any
thing’, and Johnson implicitly confirms that
he takes this as its primary sense in specify-
ing – as one of the non-discursive senses of
‘reason’ – ‘Argument; ground of persuasion;
motive’. Likewise the first sense of ‘proof’
is given as ‘Evidence; testimony; convincing
token’, supplemented in later editions by
the clauses ‘convincing argument; means of
conviction’. The words that Johnson favours
for stepwise inference are ‘deduction’ and
‘ratiocination’,57 as in the specification of
the two discursive senses of ‘reason’: ‘The
power by which man deduces one propos-
ition from another, or proceeds from prem-
ises to consequences’, and ‘Ratiocination;
discursive power’. These both contrast with
‘intuition’, which is ‘Knowledge not obtained
by deduction of reason’, and ‘intuitive’,
75
9780826443595_Ch03_Final_txt_print.indd 75
which means ‘Having the power of discov-
ering truth immediately without ratiocin-
ation’. All this seems to fit with Hume’s own
usage: he refers to ‘deductions’ and ‘ratiocin-
ation’ in contexts where stepwise argument
is clearly intended (e.g. THN 1.3.14.2 / 156,
EHU 5.22 / 55, EPM 1.4 / 170; EHU 4.23
/ 39, 12.17 / 155), and he is happy to refer
to ‘arguments’, ‘inference’ and ‘proof’ that
are ‘intuitive’, and hence do not proceed in
a stepwise fashion (THN 1.3.14.35 / 173,
2.3.2.2 / 408; EHU 4.21 / 37, 8.22n18 / 94n,
LDH 1.187, 91).58 Overall, therefore, the lan-
guage in which Hume expresses his famous
conclusion is no argument (sic.) in favour of
Garrett’s interpretation.
Perhaps the most serious problem for
Garrett’s interpretation, as for Owen’s, has
been in making sense of the logic of Hume’s
famous argument. For as we have seen, that
argument does not in fact put much emphasis
on a general absence of stepwise processing
or ratiocination within inductive inference.59
Instead, it focuses on the very specific step of
extrapolation from observed to unobserved –
that is, the supposition of the Uniformity
Principle (UP) – and then it attacks in turn
the props on which that principle ‘may be
suppos’d to be founded’, showing that none
of them can ‘afford any just conclusion of this
nature’ (THN 1.3.6.4 / 90). This move makes
perfect sense on an epistemological interpret-
ation of the argument, because if any essen-
tial evidential step in an inductive inference
lacks a ‘just foundation’, then the inference
as a whole will, apparently, be undermined.60
On the Garrett/Owen style of psychological
interpretation, however, the move looks
almost irrelevant – even if it sufficed to show
that UP is not itself reached through mediated
ratiocination, that would not exclude such
ratiocination from playing some other role
within inductive inference. This objection can
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
be sharpened by posing a dilemma over what
role UP itself is supposed to play here.61 If
Hume is saying that UP functions as an inter-
mediate step in inductive inference, then it
looks as though he thinks inductive inference
does involve stepwise ratiocination (via UP
itself), in which case he is contradicting the
very conclusion that Owen takes him to be
drawing from the argument. If, on the other
hand, Hume is denying that UP can play any
psychological role within inductive inference
(on the basis that it has no ‘just foundation’),
then it is unclear why he should take this to
imply anything further about the actual psy-
chological mechanism of inductive inference.
The only apparently plausible answer here is to
see Hume as placing a conditional constraint
on how stepwise ‘reason’ could work: ‘If rea-
son determin’d us, it wou’d proceed upon’
UP (THN 1.3.6.4 / 89; emphasis added). But
this conditional statement appears only in the
Treatise presentation, and even there, it is fol-
lowed three paragraphs later by the uncondi-
tional statement that ‘probability is founded
on the presumption of’ UP (THN 1.3.6.7 /
90; emphasis added).62 Owen largely builds
his interpretation around the conditional,63
but it is straining credibility to rely so heav-
ily on one statement in the Treatise, when we
have seen so much evidence in sections 2.1–3
above that the versions in the Abstract and
(especially) the Enquiry are more carefully
crafted.
On a psychological interpretation, more-
over, Hume should not be so confident that
mediated ratiocination for a factual conclu-
sion can proceed only via UP.64 At best, he
can claim that any rationally sufficient rea-
soning for such a conclusion must involve
UP. And likewise, his argument seems com-
pletely inadequate to show that UP itself
could not be believed on the basis of some
mediated ratiocination. At best, again, he
76
9780826443595_Ch03_Final_txt_print.indd 76
has shown that UP has no ‘just’ foundation
in demonstrative or factual reasoning.65 But
why should this be thought to exhaust the
possibilities of relevant reasoning? Hume
quite often refers to ‘arguments’ or ‘rea-
soning’ that he considers ‘absurd’, ‘falla-
cious’, or ‘sophistical’ (e.g. THN 1.2.4.11 /
43, 1.4.5.30 / 247; DNR 9.189–92); on a
psychological interpretation of the famous
argument, these should be as relevant to his
theory as the ‘just’ reasonings that he is able
to rule out.66 Suppose, for example, it were
suggested that induction can be founded on
the principle that every change must have a
cause, and hence the ultimate causal laws
must be consistent over time. This would
bring into play the attempted demonstra-
tions of the Causal Maxim that Hume
refutes at THN 1.3.3.4–8 / 80–2: even if
fallacious, they could still be contenders as
psychological explanations of our induct-
ive assumptions. In short, Hume has done
nothing to refute the hypothesis that UP
may be believed on the basis of an invalid
demonstrative argument; hence on the inter-
pretations of Owen and Garrett, his famous
argument is hopelessly incomplete.67
But things are even worse than this, for yet
another strategy that remains open on their
readings was actually used by Hume’s friend
and correspondent Richard Price in the
first chapter of his Review of the Principal
Questions in Morals. Price claims that the
Causal Maxim is known intuitively, ‘nothing
being more clearly absurd and contradict-
ory, than the notion of a change without a
changer’.68 Then, in a footnote a few pages
later, he explains how this can provide a
basis for inferring future events from past
regularity:
The conviction produced by experience
is built on the same principle, with that
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
which assures us, that there must be a
cause of every event . . . Because we see
intuitively, that there being some reason
or cause of this constancy of event, it must
be derived from causes regularly and con-
stantly operating . . . And the more fre-
quently and uninterruptedly we knew this
had happened, the stronger would be our
expectation of its happening again, . . . 69
Hume produced no fewer than seven editions
of his Enquiry after Price published this, and
it would be astonishing if he did not know
of it, given that the Enquiry itself – under its
original title Philosophical Essays – is men-
tioned twice within the vicinity of these quo-
tations (in footnotes on pages 12 and 41). As
interpreted by Owen and Garrett, however,
Hume’s famous argument completely fails to
engage with Price’s justification of induction.
That justification starts from what Hume
would no doubt claim to be a fallacious ‘intu-
ition’, but again there is no obvious psycho-
logical obstacle to erecting an argument on a
fallacy: humans do it all the time!
Another related issue concerns the intended
scope of the famous argument: is it supposed
to be proving something about every indi-
vidual factual inference, or about the genesis
of our general practice of factual inference?
When developing his interpretation, Owen
writes repeatedly as though it were the latter,
for example:
If the uniformity principle were some-
thing we knew or believed, prior to our
engaging in probable reasoning,70 we
could explain probable reasoning as
being based on reason. . . . [P]rior to our
engaging in probable reasoning, we . . .
neither know nor believe the uniform-
ity principle. . . . Hume’s argument . . .
has more to do with the failure of reason
to account for why we have the beliefs
77
9780826443595_Ch03_Final_txt_print.indd 77
we do than with why those beliefs are
unjustified.71
But when articulating Hume’s conclusion,
Owen apparently moves towards a claim
about the functioning of all individual
inductive inferences, and strongly contrasts
this with the alternative view as ascribed to
Garrett:
Hume is here denying that such infer-
ences can be explained as an activity of
the faculty of reason conceived as func-
tioning by the discovery of intermedi-
ate ideas . . . Garrett says that Hume ‘is
denying only that we come to engage
in this species of reasoning as a result
of any piece of reasoning about it’.72. . .
My main objection to Garrett’s inter-
pretation is that he treats Hume as ask-
ing about the cause of our engaging in
probable reasoning . . . Hume’s question
is not what Garrett takes it to be. Hume’s
question is: how is it that we manage to
make these inferences?73
As we have seen already, however, Garrett’s
interpretation in his 1997 book was
quickly modified (in his 1998 debate with
me), at which point he clarified that, like
Owen, he took Hume’s conclusion to be
one that applies to all individual inductive
inferences:
[From] Cognition and Commitment . . . ,
Millican understandably infers that on my
interpretation ‘it is only the general practice
of induction that fails to be determined by
reason, and each of our particular inductive
inferences is itself an instance of the oper-
ation of our reason.’ . . . The crucial dis-
tinction for Hume, however, is . . . between
an inference being an instance of reason-
ing and the same inference being caused by
(another instance of) reasoning.74
2/9/2012 9:38:52 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
The significance of this point is that the
bulk of Hume’s argument, and especially
his conclusion – in both the Treatise and
the Enquiry – seems to intend a result about
every particular factual inference (cf. sec-
tion 2.4 [H] above). But on the interpret-
ations of Owen and Garrett, the argument
seems to lead far more plausibly to a result
about factual inference in general. Note, for
example, that Hume takes many induct-
ive inferences to be reflective and mediated,
especially those that involve ‘inferences from
contrary phænomena’ (THN 1.3.12.7 / 133)
or the application and balancing of ‘general
rules’ (THN 1.3.13.7–12 / 146–50, 1.3.15 /
173–5). Moreover, inductive inferences from
‘only one experiment of a particular effect’
can – Hume says – be mediated by explicit
reflection on a principle which looks very
similar to UP (THN 1.3.8.14 / 104–5).75
Owen is well aware of this,76 but he does
not apparently recognize the threat to his
own interpretation, under which such medi-
ated inferences become counterexamples to
Hume’s conclusion, at least if that is read
(as Hume’s own words seem to require) as a
claim about each and every factual inference.
Garrett is also aware that ‘not all probable
inferences are immediate’, but he endeavours
to explain how nevertheless Hume’s conclu-
sion can be seen to apply even to those that
are mediated:
it may well happen, as Millican notes,
that one piece of probable reasoning is
part of another piece of probable rea-
soning . . . But as Hume states his con-
clusion . . . , it is that ‘in all reasonings
from experience, there is a step taken by
the mind which is not supported by any
argument or process of the understand-
ing’ [EHU 5.2 / 41; emphasis added].
Where a piece of probable reasoning
does occur as part of a second piece
78
9780826443595_Ch03_Final_txt_print.indd 78
of probable reasoning, it will involve
a crucial step that is not supported by
any argument . . . ; that is the point of
Hume’s discovery. Reasoning cannot
cause the crossing of an inductive gap.77
However, my objection to which this was a
response referred not to the situation where
‘one piece of probable reasoning is part of
another piece of probable reasoning’, but
rather, where one inductive inference’s con-
clusion (e.g. ‘a general principle of uniform-
ity’ as at THN 1.3.8.14 / 104–5) is then given
‘the role of a premise in further inductive
inference’.78 Garrett’s response thus preserves
what he takes to be Hume’s conclusion, as
universally applicable to factual inferences,
only by stipulatively treating the inference
which was used to establish a proposition
as itself a part of the further inference which
then takes that proposition as a premise. In
the context of discussing the epistemology
of induction, this might seem reasonable
enough: if the proposition in question has a
problematic foundation, then those problems
will be inherited by any further inference
built on it. But if Hume’s famous argument
is to be interpreted as involving the psycho-
logical mechanism of individual inductive
inferences – as Garrett intends – then the
move looks artificial and ad hoc, smudging
over the manifest difference between argu-
ing for some proposition and taking it as
assumed or established (on the basis of pre-
vious argument).
To sum up so far, we have yet to find an
interpretation that is genuinely satisfying. Any
extreme sceptical reading leaves Hume’s phil-
osophy hopelessly inconsistent. The anti-de-
ductivist reading and that of Owen both have
serious difficulty making sense of the logic of
his argument, and also have to rely on the text-
ually questionable claim that Hume’s notion of
2/9/2012 9:38:53 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
‘reason’ within the argument is a target rather
than sincere (a problem which also beset my
own previous interpretation).79 Garrett’s read-
ing has the significant merit of avoiding this
last pitfall, but again has difficulty squaring
with the argument’s text and logic, and has
been forced to adapt accordingly over time.
Initially, Garrett understood Hume’s conclu-
sion as the straightforward claim that ‘we do
not adopt induction on the basis of recogniz-
ing an argument for its reliability’.80 This soon
changed into the more complex claim that ‘the
underlying causal mechanism that gives rise to
inductive inferences . . . is not itself dependent
on any reasoning or inference’.81 Meanwhile,
since the Enquiry argument clearly ranges
beyond narrow ‘reasoning or inference’,
Garrett suggested that here Hume ‘expands
the famous conclusion to rule out any “rea-
soning or process of the understanding,”
thereby eliminating such non-inferential proc-
esses of the understanding as intuition or the
perception of a probable connection between
even a single “proof” and a conclusion’.82 But
pushing in the opposite direction, his recog-
nition that Hume acknowledges the role of
explicit (and sometimes complex) ratiocin-
ation within some inductive inferences has led
to a narrowing of the supposed conclusion, to
focus on the very specific logical step which is
‘the crossing of an inductive gap’.83 Even after
all this, as we have seen, Garrett’s defence of
the interpretation looks suspiciously ad hoc,
holding that conclusion to be true even of an
inductive inference which explicitly argues
across the inductive gap using an anteced-
ently established Uniformity Principle, sim-
ply on the basis that some previous inference
was required to establish that principle. But
by now the interpretation has been diluted
beyond recognition, and we seem to have lost
any focus on the actual psychological mech-
anism of individual inductive inferences – this
79
9780826443595_Ch03_Final_txt_print.indd 79
is looking far more like a discussion of the
epistemology of our general presumption of
uniformity.
3.2 ‘REASON’ AS THE COGNITIVE FACULTY
One promising route towards a better under-
standing of Humean ‘reason’ is to look at the
usage of Hume’s contemporaries in Scotland
and England, and especially those who –
unlike Locke – were enthusiastic about the
language of ‘faculties’ and relatively con-
sistent in their usage.84 Francis Hutcheson,
Professor of Moral Philosophy at Glasgow
and correspondent with Hume from 1739,
provides the closest spatio-temporal match
to the Treatise, having published in 1742 no
fewer than four works containing an out-
line of the faculties, at least one of which –
Philosophiae Moralis Institutio Compendiaria
(later translated as A Short Introduction to
Moral Philosophy) – he sent to Hume.85 This,
like the Synopsis Metaphysicae (Synopsis of
Metaphysics)86 was a Latin teaching text, and
contains a discussion of ‘The parts or pow-
ers of the soul’.87 The other two works were
An Essay on the Nature and Conduct of the
Affections and Illustrations on the Moral
Sense, published together as a third edition
of both. To the former, Hutheson added a
footnote on the faculties,88 and to the latter,
a new paragraph:
Writers on these Subjects should
remember the common Divisions of
the Faculties of the Soul. That there is
1. Reason presenting the natures and
relations of things, antecedently to any
Act of Will or Desire: 2. The Will, . . .
or the disposition of Soul to pursue
what is presented as good, and to shun
Evil. . . . Both these Powers are by the
Antients included under the Λόγος or
λογικòν μέρος. Below these they place
2/9/2012 9:38:53 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
two other powers dependent on the
Body, the Sensus, and the Appetitus
Sensitivus, in which they place the par-
ticular Passions: the former answers to
the Understanding, and the latter to
the Will. But the Will is forgot of late,
and some ascribe to the Intellect, not
only Contemplation or Knowledge, but
Choice, Desire, Prosecuting, Loving.89
It is clear from his alternation between
‘Reason’ and ‘the Understanding’ that
Hutcheson takes these to be one and the
same; indeed his Essay’s new footnote says as
much.90 This equivalence is also asserted (or
manifested through the same sort of elegant
variation of terminology that we see above)
by various other writers known to Hume, for
example Shaftesbury, Butler and Price,91 so it
was evidently commonplace, though writers
in the Scottish common-sense school later
preferred to use ‘reason’ more narrowly, in
much the way that Garrett favours.92 Hume
himself, however, is clearly aligned with
the former group, interchanging between
‘reason’ and ‘the understanding’ dozens of
times – purely for the sake of stylistic vari-
ation – just as he does between ‘the fancy’
and ‘the imagination’:93
the mind . . . is not determin’d by reason,
but by certain principles, which associate
together the ideas of these objects, and
unite them in the imagination. Had ideas
no more union in the fancy than objects
seem to have to the understanding, . . .
(THN 1.3.6.12 / 92; emphasis added)
There are no principles either of the
understanding or fancy, which lead
us directly to embrace this opinion . . .
The . . . hypothesis has no primary rec-
ommendation either to reason or the
imagination . . . (THN 1.4.2.46 / 211;
my emphasis)
80
9780826443595_Ch03_Final_txt_print.indd 80
This sceptical doubt, both with respect
to reason and the senses . . . ’Tis impos-
sible upon any system to defend either
our understanding or senses . . . ’
(THN 1.4.2.57 / 218; emphasis added)
Consider also the following two footnotes,
the first of which was expanded and moved
to create the second:
when it [the imagination] is oppos’d
to the understanding, I understand the
same faculty, excluding only our demon-
strative and probable reasonings. (THN
2.2.7.6n / 371n; emphasis added)
when I oppose it [the imagination] to
reason, I mean the same faculty, exclud-
ing only our demonstrative and prob-
able reasonings. (THN 1.3.9.19n22 /
117–18n; emphasis added)
Again, the switch from ‘the understanding’
to ‘reason’ looks purely stylistic, perhaps
prompted by the clumsiness of ‘. . . the under-
standing, I understand . . . ’.
There is thus overwhelming textual evi-
dence that Hume generally treats ‘reason’
and ‘the understanding’ as one and the
same. And virtually all major writers of the
period take ‘the understanding’ to refer to
our principal cognitive faculty, usually draw-
ing a general division between it and ‘the
will’.94 This division between the domains
of the understanding and the will is indeed
essentially the same as the modern distinc-
tion between cognitive and conative mental
functions, a dichotomy whose fundamental
nature is often now expressed in terms of a
‘direction of fit’ between world and mind:
the understanding aims to conform our
beliefs to the way the world is, while the will
aims to change the world to conform to our
desires. Reid characterizes this in terms of a
2/9/2012 9:38:53 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
distinction between our intellectual (or con-
templative) and active powers:
We shall . . . take that general division
which is the most common, into the pow-
ers of understanding and those of will.
Under the will we comprehend our active
powers, and all that lead to action, or
influence the mind to act; such as, appe-
tites, passions, affections. The under-
standing comprehends our contemplative
powers; by which we perceive objects; by
which we conceive or remember them;
by which we analyse or compound them;
and by which we judge and reason con-
cerning them. . . . The intellectual powers
are commonly divided into simple appre-
hension, judgment, and reasoning.95
Although Reid is somewhat critical of this
framework, which he takes to be ‘of a very
general reception’, his clear account of it is
helpful in setting the scene for Hume, whose
understanding of it – though again critical –
seems to be very similar.96
The earlier quotation from Hutcheson’s
Illustrations on the Moral Sense likewise
recognizes this ‘common Division of the
Faculties of the Soul’ between ‘Reason’ (or ‘the
Understanding’) and ‘The Will’, while suggest-
ing a hierarchical structure in which the senses
‘answer to’ the understanding, and the passions
to the will. His Short Introduction to Moral
Philosophy, however, paints a cruder picture
which is closer to that outlined by Reid above:
The parts or powers of the soul . . . are all
reducible to two classes, the Understanding
and the Will. The former contains all the
powers which aim at knowledge; the
other all our desires. . . . [Of] the several
operations of the understanding . . . The
first in order are the senses . . . Senses are
either external, or internal.97
81
9780826443595_Ch03_Final_txt_print.indd 81
This treats the senses themselves as ‘oper-
ations of the understanding’, a tendency
common enough for Price to make a point
of criticizing it.98 But Hutcheson’s Synopsis
of Metaphysics, within two sentences, first
implicitly places the senses within the under-
standing and then gives them a subordinate
reporting role, which suggests that the former
placement is just a shorthand way of indicat-
ing that the senses fall within the understand-
ing’s sphere of influence:
we might reasonably reduce [the powers
of the mind] to two, namely, the faculty
of understanding and the faculty of will-
ing, which are concerned respectively
with knowing things and with render-
ing life happy. The senses report to the
understanding, . . .99
The Synopsis goes on to give Hutcheson’s
most detailed account of his faculty frame-
work, with Chapter 1 of Part II devoted to a
categorization of the powers associated with
the understanding, including external sensa-
tion (sect. 3), internal senses or consciousness
(sect. 4), reflexive or subsequent sensations
(sect. 5), memory, the power of reasoning and
imagination (sect. 6).100 This again suggests a
hierarchical structure, with these various pow-
ers ‘reporting to’ an overseer faculty – reason
or the understanding proper – which perceives
and judges the deliverances of the subordin-
ate faculties in order to establish truth. Thus
‘Reason is understood to denote our Power of
finding out true Propositions’.101 Price talks
in a similar spirit of ‘the power within us that
understands; . . . the faculty . . . that discerns
truth, that views, compares, and judges of all
ideas and things’.102 And a similar conceptual
linkage between reason or the understanding
and the search for truth is common to many
other writers.103
2/9/2012 9:38:53 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
We are now in a position to appreciate the
significance of Hume’s repeated statements
that align him strongly with this general con-
ception of reason as cognition:104
Reason is the discovery of truth or fal-
shood. (THN 3.1.1.9 / 458);
That faculty, by which we discern
Truth and Falshood, . . . (EHU 1.4n –
1748/1750 editions);105
Thus the distinct boundaries and offices
of reason and of taste are easily ascer-
tained. The former conveys the knowl-
edge of truth and falsehood: . . . (EPM
App. 1.21 / 294);
reason, in a strict sense, as meaning the
judgment of truth and falsehood, . . .
(DIS 5.1 / 24; cf. THN 2.3.3.8 / 417).
In all these contexts, Hume is stressing that
reason, since it is purely cognitive, cannot
also be conative: that is, it cannot be contrary
to any passion or – by itself – provide any
motive to action or the will. This is the crux
of one of Hume’s three most famous argu-
ments concerning the incapacity of reason,
which concludes that ‘Since morals . . . have
an influence on the actions and affections, it
follows, that they cannot be deriv’d from rea-
son’ (THN 3.1.1.6 / 457). For the purpose of
this argument, it is enough that reason is con-
fined to the domain of truth and falsehood,
though Hume’s talk of discovery, discern-
ment and knowledge suggests a normative
bias towards truth rather than falsehood.
This normative flavour is far more explicit
in another of the three famous arguments,
this time concerning the external world (the
third, of course, concerns induction):
The vulgar confound perceptions and
objects . . . This sentiment, then, as it
82
9780826443595_Ch03_Final_txt_print.indd 82
is entirely unreasonable, must proceed
from some other faculty than the under-
standing. . . . Even after we distinguish
our perceptions from our objects, ’twill
appear presently, that we are still incap-
able of reasoning from the existence of
one to that of the other: So that upon
the whole our reason neither does, nor is
it possible it ever shou’d, . . . give us an
assurance of the continu’d and distinct
existence of body. (THN 1.4.2.14 / 193)
’tis a false opinion that any of our . . .
perceptions, are identically the same
after an interruption; and consequently
. . . can never arise from reason, . . .
(THN 1.4.2.43 / 209)
we may observe a conjunction or a rela-
tion of cause and effect betwixt differ-
ent perceptions, but can never observe
it betwixt perceptions and objects. ’Tis
impossible, therefore, that from the exist-
ence or any of the qualities of the former,
we can ever form any conclusion con-
cerning the existence of the latter,106 or
ever satisfy our reason in this particular.
(THN 1.4.2.47 / 212)
It might appear strange that in one of these
arguments, reason seems to embrace both
truth and falsehood, whereas in the other, it
is normatively connected with truth. But this
sort of linguistic variation is commonplace,
and it is worth noting that an unambiguous
identification of reason with the cognitive
faculty is consistent with a fairly wide range
of nuances of meaning. Given such an iden-
tification, ‘reason’ might most naturally be
used to refer to the human (or animal) fac-
ulty of truth-apprehension, however well
and by whatever processes it operates (as,
for example, when Hume compares the ‘rea-
son’ of people and animals at THN 3.3.4.5 /
610). But sometimes there might be debate
over these processes, in which case we could
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
find ourselves referring to processes that are
commonly taken to be involved in truth-
apprehension, even if they turn out not to
be truth-conducive (as suggested by Hume’s
‘scepticism with regard to reason’ of Treatise
1.4.1). Alternatively, we might wish to apply
a stricter criterion under which ‘reason’
would be confined to processes that operate
successfully to apprehend truth (thus giving
the normative flavour of the passages from
Treatise 1.4.2 quoted above).107 A differ-
ent strict usage is to refer to the faculty of
truth-apprehension acting entirely alone,
independently of other faculties such as the
senses or memory (this seems to be Hume’s
intention at THN 3.1.2.1 / 470).108 Finally,
there is in the early modern period a com-
mon metonymy, under which ‘reason’ is used
to refer to its product, namely true belief as
successfully achieved using our rational fac-
ulty (hence the pairing of ‘truth and reason’
at THN 2.3.3.5 / 415, 3.1.1.15 / 461, and
THN App. 1 / 623; cf. also note 84 above).
Notice that acknowledgement of all these
nuances is quite different from supposing an
ambiguity in ‘reason’, because they all arise
naturally from the core meaning, and there
need be no suggestion that the word has been
coincidentally assigned two or more distinct
meanings. With this understood, much of the
evidence that has previously been adduced
for the ambiguity of ‘reason’ is significantly
undermined, and it becomes more plausible
to suggest that the term has, for Hume, a sin-
gle core meaning, namely what we now call
cognition.
3.3 REASON AND THE IMAGINATION
If reason, for Hume, is just our overall cogni-
tive faculty, and if his general epistemological
approach is – as set out in section 1 above –
to begin by ascribing default authority to
83
9780826443595_Ch03_Final_txt_print.indd 83
our natural faculties, then one would expect
that his arguments about reason’s capabil-
ities would start from a relatively straightfor-
ward and conventional understanding of our
cognitive functions. Coming from a Lockean
background, it is no surprise to find Hume
recognizing the cognitive faculties of the
senses, memory, intuition, and demonstrative
and probable reasoning.109 The senses can
be either external (i.e. sight, touch, hearing,
smell, gustatory taste) or internal (i.e. reflec-
tion) – these provide the impressions from
which our ideas are copied, and those ideas
are represented to us either through the mem-
ory or the imagination. It follows that all of
our thinking, except in so far as it confines
itself to memory, must involve representation
of ideas in the imagination, which is appar-
ently to be thought of as something like a
multi-layered or multi-dimensional canvas
on which sense-copied ideas appear, with dif-
ferent degrees of ‘force and vivacity’ and ‘in a
perpetual flux and movement’ (THN 1.4.6.4
/ 252).110 Thus our faculties of intuition, dem-
onstration, and probable reasoning must
inevitably act on our imagination, through
such processes as bringing ideas into mind,
dismissing others, or – most importantly
given Hume’s analysis of belief (summarized
at the end of section 2.1 above) – changing
their degrees of force and vivacity. Even when
we make judgements about the deliverances
of our senses and memory, it is their force
and vivacity in the imagination which appar-
ently constitutes our assent to them.111 Hence
Hume’s comment, leading into the sceptical
anxieties of the Conclusion of Treatise Book
1, that: ‘The memory, senses, and understand-
ing are, therefore, all of them founded on the
imagination, or the vivacity of ideas.’ (THN
1.4.7.3 / 265). This comment might be read as
suggesting that the imagination is itself active,
but earlier in the same paragraph, Hume
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
makes clear that he is talking of principles
(namely experience and habit) that ‘operate
upon the imagination’. The initial framing of
his discussion of induction in the Treatise (as
quoted earlier in section 2.1) may give a differ-
ent impression: ‘the next question is, Whether
experience produces the idea by means of the
understanding or imagination; whether we
are determin’d by reason to make the tran-
sition, or by a certain association and rela-
tion of perceptions?’ (THN 1.3.6.4 / 88–9).
But Hume’s answer to his own question –
repeated numerous times – will be that our
causal reasoning is determined by custom,112
and he never says that it is determined by the
imagination itself. So at least in this context,
the imagination is apparently only the virtual
stage on which the mind’s various principles –
either of reason or custom – orchestrate their
dance of perceptions.113
In other contexts, however, the imagination
does appear as an active agent, having the lib-
erty to transpose and change its ideas (THN
1.1.3.4 / 10, 1.1.4.1 / 10–11, 1.3.5.3 / 85,
1.3.7.7 / 97; EHU 2.4 / 18, 5.10 / 47–8, 5.12
/ 49), to distinguish and separate them (THN
1.2.4.3 / 40, 1.2.5.3 / 54–5, 1.3.3.3 / 79–80,
1.4.5.5 / 233), to suggest (THN 1.1.7.15 /
23–4) or raise them up (THN 1.2.1.3 / 27),
and to generate fictions (THN 1.1.6.2 / 16,
1.4.2.29 / 200–1, 1.4.2.36 / 205, 1.4.2.43
/ 209, 1.4.2.52 / 215, 1.4.3.3–5 / 220–1,
1.4.3.11 / 224–5, 1.4.6.6–7 / 253–5).114 The
distinction between the two classes of oper-
ation seems to be explained by the footnote at
THN 1.3.9.19 / 117 which we encountered in
section 3.2 above:
In general we may observe, that as
our assent to all probable reasonings
is founded on the vivacity of ideas, it
resembles many of those whimsies and
prejudices, which are rejected under
84
9780826443595_Ch03_Final_txt_print.indd 84
the opprobrious character of being the
offspring of the imagination. By this
expression it appears that the word,
imagination, is commonly us’d in two
different senses; and tho’ nothing be
more contrary to true philosophy, than
this inaccuracy, yet in the following rea-
sonings I have often been oblig’d to fall
into it. When I oppose the imagination to
the memory, I mean the faculty, by which
we form our fainter ideas. When I oppose
it to reason,115 I mean the same faculty,
excluding only our demonstrative and
probable reasonings. (THN 1.3.9.19n22
/ 117–18n)
This note was inserted by means of a ‘cancel’
leaf, prepared by Hume while the Treatise
was going through the press, and I believe
he saw the need for this on rereading the end
of THN 1.3.9.4 / 108:116 ‘All this, and every
thing else, which I believe, are nothing but
ideas; tho’ by their force and settled order,
arising from custom and the relation of cause
and effect, they distinguish themselves from
the other ideas, which are merely the off-
spring of the imagination.’ A related pas-
sage is at THN 1.4.4.1 / 225, where Hume
addresses the complaint that he has criticized
‘the antient philosophers’ for being guided by
imaginative fancies, whilst building his own
philosophy on principles of the imagination:
In order to justify myself, I must dis-
tinguish in the imagination betwixt the
principles which are permanent, irresist-
ible, and universal; such as the custom-
ary transition from causes to effects, and
from effects to causes: And the principles,
which are changeable, weak, and irregu-
lar; such as those I have just now taken
notice of [e.g. the ‘inclination in human
nature to bestow on external objects
the same emotions, which it observes
in itself’, as attributed to the ‘antient
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
philosophers’]. The former are the foun-
dation of all our thoughts and actions, so
that upon their removal human nature
must immediately perish and go to ruin.
The latter are neither unavoidable to
mankind, nor necessary, or so much as
useful in the conduct of life . . .
All three passages point to a distinction
drawn within the class of principles that
‘operate on the imagination’ – that is, which
affect our thinking. Some of these are ‘the
foundation of all our thoughts and actions’,
the ‘permanent, irresistible, and universal’
principles that ground ‘our demonstrative
and probable reasonings’, and are therefore
appropriately dignified with the name of rea-
son or the understanding. The other princi-
ples are those that we more naturally think of
as belonging to the imagination itself: those
that ground our free play of ideas, fictions,
whimsies and prejudices. Hence in this nar-
rower sense the imagination is in opposition
to reason, though both sets of principles per-
form on the same stage – the imagination in
the broader sense – where all our non-mem-
ory ideas are represented. It is this broader
sense which enables Hume to refer, without
paradox, to ‘the understanding, that is, . . .
the general and more establish’d properties
of the imagination’ (THN 1.4.7.7 / 267).117
3.4 AN OPERATION OF REASON WHICH IS
‘NOT DETERMIN’D BY REASON’
Equipped with this understanding of the
Humean faculties, let us now try to clarify
the significance of his famous argument. In
the Treatise, he considers a paradigm causal
inference ‘from the impression to the idea’;
in the Abstract and Enquiry, he widens this
to any factual inference ‘beyond the present
testimony of our senses, or the records of
85
9780826443595_Ch03_Final_txt_print.indd 85
our memory’ (EHU 4.3 / 26). He then iden-
tifies the crucial step of such inference: the
extrapolation from observed to unobserved
which is encapsulated in his Uniformity
Principle. If this is to qualify as founded on
reason, then there must be some cognitive
operation that grounds it, and which does
so through genuine cognition (rather than
some fallacy or confusion). In the Treatise
and Abstract, Hume apparently takes it to be
obvious that the only plausible contenders
here are demonstrative reasoning and prob-
able (i.e. moral or factual) reasoning. In the
Enquiry he is more thorough, and rules out
also both intuition and sensory knowledge
as sources of foundation for the Uniformity
Principle. Since memory is taken for granted
in the experiential observations from which
the inference starts, this exhausts all the
standardly recognized sources of evidence
with which reason might operate. It is there-
fore no coincidence that the four sources
considered – and rejected – in Hume’s argu-
ment in the 1748 Enquiry match up exactly
with those itemized in his 1745 Letter from
a Gentleman: ‘It is common for Philosophers
to distinguish the Kinds of Evidence into
intuitive, demonstrative, sensible, and moral;
. . .’ (LFG 22). If reason is understood by
Hume in the standard contemporary way –
as the overall cognitive faculty – then we
should indeed expect it to embrace all four
‘Kinds of Evidence’.
Notice that this way of reading Hume’s
argument has the clear implication that
inductive (i.e. probable, moral, factual)
inference is being treated as an operation of
reason throughout, which at least strongly
suggests that it would be a mistake to inter-
pret Hume’s conclusion – that such inference
is ‘not determin’d by reason’ – as deposing
it from that status. (For if that were indeed
his intention, one might reasonably expect
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
such an apparently paradoxical move to
be far more clearly signalled.) So we need
to understand how Hume, as a result of
his famous argument, can coherently view
induction as an operation of reason which is
not ‘determin’d by’ reason.118
The obvious answer, given both our inter-
pretation of reason and the structure of
Hume’s argument, is that he views induction
as a cognitive process which depends on a
non-cognitive sub-process. So he is think-
ing at two levels, with inductive inference
being a manifest operation of our conscious
reason, causally driven by a subconscious
process that involves the customary enliven-
ment of our ideas. This underlying process
is of a type which is naturally categorized as
‘imaginative’ rather than ‘rational’, because
it works through an associative mechanism
which automatically and mindlessly extrapo-
lates beyond anything that we have perceived
or otherwise detected in the world (whether
objective events, or evidence). It is therefore
in sharp contrast with the underlying pro-
cess hypothesized by Locke, who supposed
inductive (i.e. probable) inference to be
driven by a perceptual process, namely the
rational apprehension of objective probable
connexions. Locke therefore saw induction
as a cognitive process which depends on a
cognitive sub-process – apparently ‘cognitive
all the way down’ because it is ultimately
founded on direct perception of evidential
connexions. Hume’s argument destroys this
illusion by showing that there is no plausible
source for such perception: it cannot derive
from examining relations of ideas (because it
depends on the experienced world), but nor
can it derive from experience, either current
(because the senses detect no such evidential
connexions) or remembered (because induct-
ive extrapolation is the very process whose
perceptual basis we are seeking).
86
9780826443595_Ch03_Final_txt_print.indd 86
A similar theme can be seen in the other of
Hume’s most famous arguments that assigns
a vital role to the imagination, on ‘Scepticism
with Regard to the Senses’ (THN 1.4.2 /
187–218). Here he takes on the natural and
naive assumption that external objects – dis-
tinct from us and continuous over time – are
directly and straightforwardly perceived
through the senses. To refute this, he shows
that identification of objects over time
requires a process that goes beyond anything
we perceive, latching onto patterns of ‘con-
stancy’ and ‘coherence’ in our distinct impres-
sions, and smoothing over gaps and changes
to generate an illusion of continuity. Again
the process involved is naturally categorized
as ‘imaginative’, and so Hume describes
his argument as showing that our ‘assur-
ance of the continued and distinct existence
of body . . . must be entirely owing to the
imagination’ (THN 1.4.2.14 / 193). Like
his argument concerning induction, there-
fore, this can be seen as making a significant
contribution to both cognitive science and
epistemology, by highlighting how the infor-
mational processes that are implicit in the
temporal identification of physical objects
go well beyond anything that is directly per-
ceived. Indeed modern-day cognitive science,
through the development of ‘artificial intel-
ligence’ visual systems, has provided striking
vindication of Hume, by showing how even
the identification of physical objects at a time
requires ‘imagination-like’ processes of edge-
detection, region identification, shadow and
texture interpretation, and so forth. So far
from being merely passive, visual perception
involves many active – albeit unconscious –
processes, without which the manifold of our
sensory impressions would be completely
incomprehensible.
This interpretation of Hume involves
understanding his talk of faculties as
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
descriptive of types of process rather than as
references to parts of the mind, and indeed
this seems anyway to be required in the light
of his general scepticism about any faculty
language that pretends to be more than a
functional description (THN 1.4.3.10 / 224;
DNR 4.162–3). For Hume, as for Locke, a
faculty just names a power.119 Nevertheless,
at least in the Treatise, he has an unfortunate
tendency to talk of faculties in the way that
Locke rightly deplored, as ‘so many distinct
Agents’.120 Taking such language literally, his
famous argument paints the absurd picture
of reason attempting in vain to make an
inductive inference, and needing the imagi-
nation to step in to lend a hand. But induc-
tion is such a central cognitive process that
it ought by definition to be an operation of
reason (just as remembering is by definition
an operation of memory); hence if we think
of faculties as distinct agents or areas of the
mind, custom – as the underlying process
that drives induction – should itself be part
of reason. Presumably it is this sort of con-
sideration that led Hume, in the wake of his
famous argument, to reassign ‘the general
and more establish’d properties of the imagi-
nation’ – which must surely include custom –
to reason or the understanding (THN 1.4.7.7
/ 267, cf. section 3.3 above).121 But then we
get into a muddle if we want to hold on to
his conclusion that inductive inference is ‘not
determin’d by reason’, given the frequency
with which he says that inductive inference is
indeed determined by custom. Little wonder,
perhaps, that both Hume and his interpreters
sometimes seem to exhibit confusion of the
faculties!
As so often, the Enquiry brings consid-
erable improvement, and in a number of
respects. Now the faculties are rarely spo-
ken of as agents in their own right, with the
harmless exception of those passages that
87
9780826443595_Ch03_Final_txt_print.indd 87
stress the liberty of the imagination (EHU
2.4 / 18, 5.10 / 47–8, and 5.12 / 48–50). The
contrast is especially striking in the case of
reason, because whereas the Treatise speaks
of reason itself as a determining cause (e.g.
THN 1.3.6.4 / 88–9, 1.3.6.12 / 92, 1.3.7.6
/ 97, 1.4.1.1 / 180), the Enquiry never does
so. In the later work, Hume’s preference is
to talk instead of reasoning processes (e.g.
EHU 4.23 / 39, 5.4 / 42, 9.6 / 108), which
were never mentioned as such in the Treatise.
Meanwhile, custom in the Enquiry is said
to act on the imagination (EHU 5.11 / 48,
9.5 / 106–7) and is never said (or implied)
to be itself an operation of the imagination,
thus avoiding the complications that arise
from trying to place it consistently within the
conventional faculty structure.122 Moreover,
Hume no longer refers to custom as a prin-
ciple of association of ideas (cf. THN 1.3.7.6
/ 97),123 but says instead that it is a process
analogous to the association of ideas, which
‘is of a similar nature, and arises from simi-
lar causes’ (EHU 5.20 / 53–4). He continues,
however, to draw a contrast between cus-
tom and reason (EHU 5.5 / 43, 5.20 / 53–4),
thereby retaining the core of his theory that
inductive inference is determined by a sub-
process which is not itself cognitive.
4. CONCLUSION: SCEPTICISM AND
RATIONAL FOUNDATIONS
After all this, how sceptical is Hume’s pos-
ition? His famous argument has shown
that inference from past to future crucially
involves a process of extrapolation that can-
not be independently justified by anything
within our cognitive grasp. This crucial step
is instead due to a mechanical associative
process in the mind, whereby past experience
2/9/2012 9:38:54 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
raises certain ideas about the future and enli-
vens them into beliefs. Such a process – given
its automatic, non-reflective nature, and its
lack of any rational insight or apprehension
of reality – is naturally classified as ‘imagina-
tive’ rather than ‘cognitive’, and Hume’s
faculty language is best interpreted accord-
ingly, as a way of categorizing types of pro-
cess, rather than as a theory of distinct agents
within our minds. So when he claims that the
imagination plays a crucial role in inductive
inference, he should be understood as saying
simply that our process of making inductive
inferences itself crucially involves an imagi-
nation-like sub-process.
As we have seen (in section 3.1 above),
Hume is well aware that many inductive infer-
ences also involve reason-like sub-processes,
as for example when we consciously take
into account the ‘rules by which to judge of
causes and effects’ of Treatise 1.3.15 / 173–6,
or attempt to identify underlying mathemat-
ical patterns (e.g. EHU 4.13 / 31, 7.29n17
/ 77n). But he is clearly far more interested
in the crucial imaginative step, even to the
extent of describing it as solely responsible
for the inference:
When the mind . . . passes from the idea
or impression of one object to the idea
or belief of another, it is not determin’d
by reason, . . . The inference . . . depends
solely on the union of ideas. (THN
1.3.6.12 / 92; emphasis added)
all reasonings are nothing but the effects
of custom; . . . (THN 1.3.13.11 / 149;
emphasis added; cf. EHU 5.5–6 / 43–4)
Similarly with his argument concerning our
belief in the continued and distinct existence
of body (which aims to show that it depends
crucially on various associative processes,
constructing ‘fictions’ from the passing show
88
9780826443595_Ch03_Final_txt_print.indd 88
of impressions), he expresses his conclusion
in a way that ignores the obvious and essen-
tial role of the senses:
That opinion must be entirely owing to
the imagination (THN 1.4.2.14 / 193;
emphasis added).
Hume seems to be assuming here that even
one imaginative step is sufficient to charac-
terize the entire process of which it is a part
as one that is determined by the imagination
(rather than by reason or the senses), just as
one invalid step within a sequential inference
typically renders the entire inference invalid.
Such a focus when speaking of inferential
processes is indeed quite natural, since we
are typically interested in the weakest link in
any chain of support rather than the strong-
est. The same applies to other supportive or
foundational relationships: thus a climber can
properly be described as ‘supported only by a
rope’, whether that rope itself is secured to a
mountain, a building, a heavy vehicle or any
other relatively reliable anchor.124 Likewise,
an argument or legal case which crucially
depends on some imaginative fabrication,
even if it also depends on numerous points
that are logically unassailable, can appro-
priately be said to be ‘founded on fantasy’.
But if we follow through this line of thought,
then since inductive inference depends on a
sub-process of ‘imaginative’ extrapolation
which itself has no rational grounding, we
seem forced to conclude that any proposition
that can be established only by such infer-
ence must apparently in turn be disqualified
from counting as founded on reason. Yet as
we have seen in sections 3.1 and 3.4 above,
Hume continues to treat induction as a legit-
imate operation of reason.125 There is, at the
least, a sceptical tension here: can we really
suppose that he would consider a process
2/9/2012 9:38:55 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
genuinely rational which rests on a purely
mechanical, non-reflective sub-process?
To address this worry, suppose that Hume
were to take the alternative view, that any
rational process must have a rational foun-
dation. It would then immediately follow
that for anything to be founded on reason
at all, it must be founded on reason ‘all the
way down’ (i.e. it would have to be solidly
founded on evidence or principles, which
are either immediately apprehended by
reason, or else themselves solidly founded
on evidence or principles which are either
immediately apprehended by reason or . . . ,
etc.). Hume would thus be committed to a
strongly rationalistic notion of reason, the
demands of which would be impossible to
fulfil without abandoning the heart of his
philosophy. At no point would he be able to
halt the foundational regress by acknowledg-
ing that ultimately the principles of our rea-
son can (legitimately) be grounded on basic
psychological mechanisms. So the only pos-
sible outcomes would be either rationalism
or incurable scepticism. Some interpreters
have indeed seen Hume as impelled towards
radical scepticism by precisely this kind of
regressive train of thought.126 But it would
be completely at odds with his efforts to
ground a conception of reason on the con-
tingent operations of the human mind, and
flatly incompatible – in the light of his own
investigations – with treating induction as a
genuine operation of reason.
As we saw in section 3.4 above, Locke
implicitly follows the path that Hume rejects,
by attributing probable reasoning to the per-
ception of probable connexions. And indeed
direct perception – conceived of as a process
of transparent apprehension – seems to be a
paradigm of what reason requires if it is to
be ‘cognitive all the way down’. Such percep-
tion could at once provide both a rational
89
9780826443595_Ch03_Final_txt_print.indd 89
foundation for belief and also a cause. By
contrast, ‘imagination-like’ processes such
as custom may cause belief, but they cannot
at the same time provide a cognitive foun-
dation: that is indeed precisely why they do
not qualify as processes of reason.127 Hume
seems to have embraced this distinction, if
not perhaps immediately, for his language in
the Abstract and Enquiry (though not in the
Treatise) precisely fits it. In the Treatise, he
repeatedly talks about custom (or the prin-
ciples of the imagination) as providing a
foundation for inductive inference.128 In the
Abstract and Enquiry, by contrast, he never
does, but there are no fewer than 19 passages
that describe the influence of custom in terms
that are either explicitly causal, or naturally
interpretable as such.129 This strongly sug-
gests that Hume himself came to recognize a
firm distinction between what in the Enquiry
he calls a foundation, and what he there calls
a determining cause. Thus understood, it is
clear that custom qualifies only as a cause,
whereas reasoning processes or sources of
information can potentially provide a cogni-
tive foundation.130
All this brings the possibility of posing
coherent but unanswerable questions, such
as that which introduces Hume’s discussion
in Part 2 of Enquiry Section 4: ‘if we still
carry on our sifting humour, and ask, What
is the foundation of all conclusions from
experience? this implies a new question . . .’
(EHU 4.14 / 32). If custom cannot qualify
as a foundation, then Hume’s ultimate con-
clusion that ‘All inferences from experience
. . . are effects of custom, not of reasoning’
(EHU 5.5 / 43) excludes any foundation at
all. For in competing successfully for the
causal explanatory role, custom effectively
excludes anything else from the foundational
role (which it is nevertheless unable to fulfil
itself). Perhaps, then, there is hidden depth
2/9/2012 9:38:55 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
in Hume’s declaration of intent: ‘I shall con-
tent myself, in this section, with an easy
task, and shall pretend only to give a nega-
tive answer to the question here proposed.’
(EHU 4.15 / 32). The upshot is that ‘if we still
carry on our sifting humour’ in the search
for ultimate foundations, we hit rock bot-
tom with something that has a cause but no
foundation. And that is the tendency, rooted
in our animal nature, to infer from past to
future, from experienced to not-yet-expe-
rienced. This is radically different from the
kind of perceptual foundation presupposed
by traditional conceptions of reason, differ-
ent enough to make Hume’s position seem
outrageously sceptical by comparison. But in
reality it is quite the reverse: he is providing
an account of reason which makes inductive
APPENDIX: HUME’S ARGUMENT CONCERNING INDUCTION
(FROM SECTION 4 OF THE ENQUIRY CONCERNING HUMAN UNDERSTANDING)
Hume’s Own Statement of the Propositions
Identified in the Structure Diagram
1. By means of [Cause and Effect] alone
can we go beyond the evidence of our
memory and senses. (EHU 4.4 / 26)
2. All reasonings concerning matter of fact
seem to be founded on the relation of
Cause and Effect. (EHU 4.4 / 26)
. . . all arguments concerning existence
are founded on the relation of cause and
effect . . . (EHU 4.19 / 35)
. . . all our evidence for any matter of
fact, which lies beyond the testimony
of sense or memory, is derived entirely
from the relation of cause and effect . . .
(EHU 12.22 / 159)
3. No object ever discovers, by the quali-
ties which appear to the senses, either
the causes which produced it, or the
effects which will arise from it . . . (EHU
4.6 / 27)
90
9780826443595_Ch03_Final_txt_print.indd 90
inference possible for human beings, despite
the sceptical impact of his famous argument
which shows that it cannot be founded in any
of the ways that previous tradition would
countenance.131 He reveals how we actu-
ally reason inductively, rather than falling
back on the aprioristic supposition that this
can only be through the rational perception
of evidential connexions. That traditional
notion is decisively refuted by his sceptical
argument, but his own position is very far
from sceptical. On the contrary, as we saw in
section 1 above, Hume sees very good reason
to accept our faculty of inductive inference
as it is (at least when suitably disciplined by
general rules etc.), and no good reason to
reject it. We have, indeed, no alternative, nor
any compelling reason for desiring one.132
It is allowed on all hands, that there is
no known connexion between the sen-
sible qualities and the secret powers . . .
(EHU 4.16 / 33)
4. . . . every effect is a distinct event from its
cause. It could not, therefore, be discov-
ered in the cause, and . . . the conjunc-
tion of it with the cause must appear . . .
arbitrary; since there are always many
other effects, which, to reason, must
seem fully as consistent and natural.
(EHU 4.11 / 30)
5. . . . the knowledge of [cause and effect]
is not, in any instance, attained by rea-
sonings a priori; but arises entirely from
experience . . . (EHU 4.6 / 27)
. . . causes and effects are discover-
able, not by reason, but by experience
. . . (EHU 4.7 / 28)
In vain, therefore, should we pretend
to . . . infer any cause or effect, without
2/9/2012 9:38:55 PM
9780826443595_Ch03_Final_txt_print.indd 91

(1) Only the relation of cause and

effect can take us beyond the
evidence of our memory and senses

(2) All factual inferences to the

unobserved are founded on the (7) All reasonings from experience
relation of cause and effect are founded on the Uniformity
Principle (UP)

(4) Any effect is quite distinct from (6) All factual inferences to the
(8) All factual inferences to the
(20) CONCLUSION
its cause, and many different effects unobserved are founded on
are equally conceivable experience unobserved are founded on UP No factual inference to the
unobserved is founded on reason

(14) A change in the course of (13) Two kinds of argument are

nature can be distinctly available (for proving UP):
(5) Causal relations cannot be known
conceived, and hence is possible demonstrative and factual
a priori, but can only be discovered
by experience

HUME’S ‘SCEPTICISM’ ABOUT INDUCTION

91

(15) Future uniformity cannot be (16) If there is a good argument

inferred demonstratively from for UP, it must be a factual
(3) Sensory perception of any object
past uniformity inference
does not reveal either its causes or its
effects, and there is no known
connexion between the sensible
qualities and its ‘secret powers’

(9) UP is not founded on anything (17) Any factual inference to UP (18) There is no good argument
that we learn through the senses would be circular of any kind for UP
about objects’ ‘secret powers’
(10) UP can be founded on reason
only if it is founded on experience (19) UP is not founded on reason
(of uniformity)
(12) UP can be founded on
reason only if it is founded on
argument (via some medium
enabling it to be inferred from
past experience of uniformity)

(11) The inference from past

uniformity to future uniformity is not
intuitive
2/9/2012 9:38:55 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
the assistance of observation and experi-
ence. (EHU 4.11 / 30)
. . . our knowledge of that relation [of
cause and effect] is derived entirely from
experience . . . (EHU 4.19 / 35)
6. . . . nor can our reason, unassisted by
experience, ever draw any inference
concerning real existence and matter of
fact . . . (EHU 4.6 / 27)
In vain, therefore, should we pretend
to determine any single event . . . with-
out the assistance of observation and
experience. (EHU 4.11 / 30)
7. . . . we always presume, when we see
like sensible qualities, that they have like
secret powers, and expect, that effects,
similar to those which we have experi-
enced, will follow from them . . . (EHU
4.16 / 33)
We have said, that . . . all our experi-
mental conclusions proceed upon the
supposition, that the future will be con-
formable to the past . . . (EHU 4.19 / 35)
. . . all inferences from experience sup-
pose, as their foundation, that the future
will resemble the past, and that similar
powers will be conjoined with similar
sensible qualities . . . (EHU 4.21 / 37)
8. [This proposition is implicit in the infer-
ential sequence:] We have said, that all
arguments concerning existence are
founded on the relation of cause and
effect; that our knowledge of that rela-
tion is derived entirely from experience;
and that all our experimental conclu-
sions proceed upon the supposition,
that the future will be conformable to
the past. (EHU 4.19 / 35)
9. . . . the mind is not led to form such a
conclusion concerning [sensible quali-
ties and secret powers’] constant and
regular conjunction, by any thing which
it knows of their nature . . . (EHU 4.16
/ 33)
10. [This proposition is implicit in Hume’s
transition from considering ‘a priori’
evidence for the Uniformity Principle
92
9780826443595_Ch03_Final_txt_print.indd 92
to considering experiential arguments
for it:] As to past Experience, it can be
allowed to give direct and certain infor-
mation of those precise objects only,
and that precise period of time, which
fell under its cognizance: but why this
experience should be extended to future
times, and to other objects, which for
aught we know, may be only in appear-
ance similar; this is the main question
on which I would insist. (EHU 4.16 /
33–4)
11. The connexion between these proposi-
tions [I have found that such an object
has always been attended with such an
effect and I foresee, that other objects,
which are, in appearance, similar, will
be attended with similar effects] is not
intuitive. (EHU 4.16 / 34)
12. There is required a medium, which may
enable the mind to draw such an infer-
ence, if indeed it be drawn by reasoning
and argument. (EHU 4.16 / 34)
13. All reasonings may be divided into two
kinds, namely demonstrative reasoning,
or that concerning relations of ideas,
and moral reasoning, or that concern-
ing matter of fact and existence. (EHU
4.18 / 35)
14. . . . it implies no contradiction, that the
course of nature may change . . . May I
not clearly and distinctly conceive [such
a thing]? (EHU 4.18 / 35)
15. That there are no demonstrative argu-
ments in the case, seems evident . . . (EHU
4.18 / 35)
. . . whatever is intelligible, and can be
distinctly conceived, implies no contra-
diction, and can never be proved false by
any demonstrative argument or abstract
reasoning a priori . . . (EHU 4.18 / 35)
16. If we be, therefore, engaged by arguments
to put trust in past experience, and make
it the standard of our future judgment,
these arguments must be probable only,
or such as regard matter of fact and real
existence . . . (EHU 4.19 / 35)
2/9/2012 9:38:56 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
17. To endeavour, therefore, the proof [that
the future will be conformable to the
past] by probable arguments, or argu-
ments regarding existence, must be evi-
dently going in a circle, and taking that
for granted, which is the very point in
question. (EHU 4.19 / 35–6)
18. . . . it may be requisite . . . to shew, that
none of [the branches of human knowl-
edge] can afford such an argument . . .
(EHU 4.17 / 35)
. . . we have no argument to convince
us, that objects, which have, in our experi-
ence, been frequently conjoined, will like-
wise, in other instances, be conjoined in
the same manner . . . (EHU 12.22 / 159)
19. . . . it is not reasoning which engages us
to suppose the past resembling the future,
and to expect similar effects from causes,
which are, to appearance, similar . . .
(EHU 4.23 / 39)
. . . nothing leads us to [expect con-
stant conjunctions to continue] but cus-
tom or a certain instinct of our nature
. . . (EHU 12.22 / 159)
20. I say then, that, even after we have expe-
rience of the operations of cause and
effect, our conclusions from that experi-
ence are not founded on reasoning, or
any process of the understanding. (EHU
4.15 / 32)
. . . in all reasonings from experience,
there is a step taken by the mind, which
is not supported by any argument or pro-
cess of the understanding. (EHU 5.2 / 41)
All belief of matter of fact or real
existence [is due merely to] a species of
natural instincts, which no reasoning or
process of the thought and understand-
ing is able, either to produce, or to pre-
vent. (EHU 5.8 / 46–7)
NOTES
1
The argument appears in Treatise 1.3.6 /
86–94, Abstract 8–16 / 649–52, and Enquiry 4 /
93
9780826443595_Ch03_Final_txt_print.indd 93
25–39, and is outlined in sections 2.1–4 below.
In discussions of induction it is commonly
referred to as ‘Hume’s famous argument’, a
convenient shorthand that I shall adopt. Note
also that ‘induction’ is the modern term for the
topic of his argument; he himself never uses the
word in this sense.
2
This is the summary of the Section 4 argument
alluded to earlier. Note, however, that the pre-
vious clause brings in a point from the Section
7 discussion of the idea of necessary connexion,
which does not feature in Section 4 itself.
3
Hume does not reject the Causal Maxim, but
says that it ‘must . . . arise from observation and
experience’ (THN 1.3.3.9 / 82), hinting that he
will return to it later (though he never does). For
detailed discussion, see Peter Millican, ‘Hume’s
Determinism’, Canadian Journal of Philosophy
40 (2010), pp. 611–42, sects II, IV, VI.
4
Section 1.3.7 will in due course move on to the
third component, ‘the nature and qualities’ of
the belief-idea.
5
Hume continues to mention the imagination’s
power to mix and separate its ideas (e.g.
Abs. 35 / 662, EHU 5.10 / 47–8), but the
Separability Principle as such is never again
invoked as it had been, very significantly, in the
Treatise (e.g. THN 1.1.7.3 / 18–9, 1.2.3.10 /
36–7, 1.2.5.3 / 54–5), arguably sometimes with
absurd results (e.g. THN 1.4.3.7 / 222, 1.4.5.5 /
233, 1.4.5.27 / 245–6, App. 12 / 634).
6
At THN 1.3.2.11 / 77, Hume had stressed
that (single-case) contiguity and succession
are insufficient to characterize a cause and
effect relationship, pointing out that ‘There
is a necessary connexion to be taken into
consideration’. Now at THN 1.3.6.3 / 87, he
reminds us that ‘Contiguity and succession are
not sufficient to make us pronounce any two
objects to be cause and effect’, and he expresses
satisfaction at having unexpectedly ‘discover’d
a new relation . . . This relation is their con-
stant conjunction.’ The link between the
passages is evident both from the content and
the capitalization.
7
John Locke, An Essay concerning Human
Understanding, ed. Peter H. Nidditch, (Oxford:
Clarendon Press, 1975), IV.xv.1, IV.xvii.2.
8
Humean demonstration corresponds to what
is now called deductive reasoning, in the
informal sense of an argument whose premises
2/9/2012 9:38:56 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
conceptually guarantee the truth of the conclu-
sion. For more on this, see Peter Millican,
‘Humes Old and New: Four Fashionable
Falsehoods, and One Unfashionable Truth’,
Proceedings of the Aristotelian Society, supp.
vol. 81 (2007), pp. 163–99, sect. V.
9
This result comes from Hume’s theory of
relations, at THN 1.3.2.1–3 / 73–4 (for criti-
cism, see Peter Millican, ‘Hume’s Fork, and
His Theory of Relations’, forthcoming in
Philosophy and Phenomenological Research).
In brief, THN 1.1.5 / 13–5 enumerates what
Hume takes to be the seven different kinds
of relation, which THN 1.3.1.1 / 69–70 then
divides into two classes. The four relations ‘that
depend solely on ideas’ are the sources of strict
‘knowledge’, with resemblance, contrariety
and degrees in quality amenable to intuition
(THN 1.3.1.2 / 70), and proportions in quan-
tity or number the basis for demonstration. Of
the three ‘inconstant’ relations, identity and
relations of time and place are amenable to per-
ception (THN 1.3.2.2 / 73–4), leaving causation
as ‘the only one, that can be trac’d beyond our
senses, and informs us of existences and objects,
which we do not see or feel’ (THN 1.3.2.3 /
74). Hume thus identifies probable with causal
reasoning, and the rest of Book 1, Part 3,
entitled ‘Of Knowledge and Probability’, is
accordingly devoted to ‘the idea of causation
. . . tracing it up to its origin’ (THN 1.3.2.4 /
74–5). Strangely, the word ‘probability’ does not
appear at all in this Part before THN 1.3.6.4
/ 89, except in the title of the Part itself and of
Section 1.3.2: ‘Of Probability; and of the Idea of
Cause and Effect’.
10
Notice that causal inference is categorically
stated to be founded on that presumption –
there is no suggestion here of the conditional-
ity that we had at THN 1.3.6.4 / 89: ‘If reason
determin’d us, it wou’d proceed upon that
principle . . .’ (emphasis added). Nor is there
any such conditionality at THN 1.3.6.11 /
91–2, or in either the Abstract (Abs. 13–14 /
651) or the Enquiry (EHU 4.19 / 35–6, 4.21 /
36–7, 5.2 / 41–2).
11
THN 1.3.6.7 / 90 expresses the circularity in
causal terms: ‘The same principle cannot be
both the cause and effect of another’, appar-
ently in order to make a joke: ‘and this is,
perhaps, the only proposition concerning that
94
9780826443595_Ch03_Final_txt_print.indd 94
relation, which is either intuitively or demon-
stratively certain’. The Abstract and Enquiry
make clear that the circularity is logical.
12
Before drawing this conclusion, Hume adds
(what I have called) a ‘coda’ to his argument
(THN 1.3.6.8–10 / 90–1), dismissing an
attempt to get round it by appeal to objects’
powers. This attempt is refuted by the simple
observation that induction needs to be presup-
posed to enable us to draw an inference from
the powers of past objects to the powers of
future objects. For discussion of this coda and
its implications, see Peter Millican, ‘Hume’s
Sceptical Doubts concerning Induction’, in
Peter Millican (ed.), Reading Hume on Human
Understanding (Oxford: Clarendon Press,
2002), pp. 107–73, sects 9–9.2.
13
He also refers back to it in a footnote at THN
1.3.14.17 / 163, feeding into his discussion of
the idea of necessary connexion.
14
For discussion of some of the nuances of termi-
nology for referring to this kind of reasoning,
see Millican, ‘Hume’s Sceptical Doubts con-
cerning Induction’, sect. 3.1, which distin-
guishes between probable inference, factual
inference, factual inference to the unobserved,
and inductive inference. Hume generally takes
for granted that all of these coincide.
15
The argument from THN 1.3.6.8–10 / 90–1 is
also very briefly summarized, in the last two
sentences of paragraph 15. For more on this
‘coda’, see note 12 above.
16
Notice that Hume seems entirely happy to
take perception and memory for granted here,
fitting with the strategy described in section 1
above, of allowing default authority to our
faculties. Scepticism regarding the senses is
addressed at THN 1.4.2–4 / 187–231 and
EHU 12.6–16 / 151–5, but Hume’s ultimate
attitude to it remains far less clear than his
position on induction.
17
This notion of a proof plays a significant role
in Hume’s argument concerning miracles in
Section 10 of the Enquiry.
18
This is taken from Millican, ‘Hume’s Sceptical
Doubts concerning Induction’.
19
Thus there is no evidence here, as influentially
claimed by David Stove, Probability and
Hume’s Inductive Scepticism (Oxford: Oxford
University Press, 1973), p. 50, that Hume’s
method of argument shows him to be a
2/9/2012 9:38:57 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
‘deductivist’, presupposing that only deduc-
tively valid arguments are legitimate. A similar
point, though less obvious, can be made about
the Abstract (‘The mind can always conceive
any effect to follow from any cause, and
indeed any event to follow upon another’,
Abs. 11 / 650) and the Treatise (‘When we pass
from a present impression to the idea of any
object, we might possibly . . . have substituted
any other idea in its room’, THN 1.3.6.1 / 87).
20
This case of applied mathematics (cf. also
24
THN 2.3.3.2 / 413–4) shows that Hume
is quite comfortable with demonstrative,
mathematical reasoning being applied to a pos-
teriori premises. For discussion of this point,
see Millican, ‘Humes Old and New’, sect. V.
21
Hume’s talk of ‘secret powers’ is new in the
25
Enquiry, and seems to reflect a more sophis-
ticated understanding of scientific reasoning
than is evident in the Treatise and Abstract.
In the Treatise, science is generally treated
as involving predictions of discrete types of
event based on previous patterns of conjunc-
tion or difference (as in the ‘rules by which to
judge of causes and effects’ of THN 1.3.15 /
26
173–176). The Enquiry, by contrast, evinces
an awareness (e.g. at EHU 4.13 / 31 and EHU
7.29n17 / 77n) that science more typically
27
deals with events having continuously vary-
ing characteristics – such as the velocity of a
28
billiard ball – whose prediction involves the
interplay of mathematically determined forces.
For more on this, see Peter Millican, ‘Against
the New Hume’, in Rupert Read and Kenneth
A. Richman (eds), The New Hume Debate:
Revised Edition (London: Routledge, 2007),
pp. 211–52, at pp. 232–3.
22
Hume obviously means us to infer accord-
ingly – though he does not explicitly state –
that [8] all factual reasoning, since it has to
be founded on experience, presupposes such a
resemblance (i.e. the Uniformity Principle). See
also note 27 below.
23
In EHU 4.16 / 33–4 itself, Hume oscillates
between reference to the activity of inference
from observed to unobserved, and to the
presupposition of resemblance on which such
29
inference is based. Indeed it seems that he takes
the foundation of the inference to be the same
as the foundation of the presupposition that it
manifests. This supports an interpretation of
95
9780826443595_Ch03_Final_txt_print.indd 95
the Uniformity Principle as implicit rather than
explicit, a principle we exhibit by our infer-
ential behaviour rather than one we always
consciously consider. Such an interpretation
nicely squares Hume’s repeated commitment
to the Principle’s role in all inductive inference
(see note 10) with his clear recognition at THN
1.3.8.13 / 103–4 that we characteristically
‘draw inferences from past experience, without
reflecting on . . . that principle’. See also note
31 below.
This suggests that if the inference were intui-
tive, it would count as ‘reasoning and argu-
ment’ notwithstanding the lack of a ‘medium’.
Indeed, as we shall see later (section 3.1), in
Hume’s day the terms ‘reasoning’ and ‘argu-
ment’ did not imply complex ratiocination.
Hume is fond of elegant variation, frequently
using a variety of terms for the same concept.
‘Moral reasoning’, ‘reasoning concerning
matter of fact and real existence’, ‘probable
arguments’ and ‘arguments concerning exist-
ence’ are all ways of referring to what we are
here calling factual reasoning. See note 14
above.
That he takes these to be equivalent was
made clear by EHU 4.2 / 25–6, where he first
explained the notion of a matter of fact.
For an earlier occurrence of this last implicit
inference, see note 22 above.
As in the Treatise (note 12 above) and Abstract
(note 15 above), Hume rounds off the argu-
ment in the Enquiry with a coda (EHU 4.21
/ 36–8) in which he refutes the attempt to
circumvent his argument by appeal to objects’
powers. He also adds a parting shot at EHU
4.23 / 39 which emphasises the unlikelihood
that peasants, infants or ‘brute beasts’ should
form their inductive expectations on the basis
of ‘any process of argument or ratiocination’.
Though the point is well made, however, its
philosophical significance is less clear, because
those who take induction to be rationally
founded need not be committed to supposing
that animals (etc.) function purely ration-
ally – see Millican, ‘Hume’s Sceptical Doubts
concerning Induction’, sect. 9.3.
Recall that [12] is the claim that ‘There is
required a medium, which may enable the
mind to draw such an inference, if indeed
it be drawn by reasoning and argument.’
2/9/2012 9:38:57 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
(EHU 4.16 / 34) – that is, because the
Uniformity Principle cannot be established dir-
ectly through sensory perception or intuition,
if it is to be established by reason at all, then
this must be on the basis of some stepwise
argument or ratiocination.
30
The other implicit final stages are also stated
explicitly elsewhere: [18] ‘we have no argu-
ment to convince us, that objects, which have,
in our experience, been frequently conjoined,
will likewise, in other instances, be conjoined
in the same manner’ (EHU 12.22 / 159); [19]
‘it is not reasoning which engages us to sup-
pose the past resembling the future, and to
expect similar effects from causes, which are,
to appearance, similar’ (EHU 4.23 / 39).
31
See note 23 above for the nature of this
presupposition, which need not be conscious,
but is implicitly manifested by the making of
the inference. So UP need not take any very
explicit or determinate form (contrary to the
impression given by THN 1.3.6.4 / 88–9), and
is best understood as something like a general
principle of evidential relevance between
observed and unobserved, more in line with
the expression of the Enquiry: ‘we . . . put trust
in past experience, and make it the standard
of our future judgment’ (EHU 4.19 / 35); we
take ‘the past [as a] rule for the future’ (EHU
4.21 / 38). This seems right: in taking such an
34
inference to be better informed than an a priori
inference, we are ipso facto presuming that
35
what happened in the past provides evidence
36
that is positively relevant to what will happen
in the future. For more on the Uniformity
37
Principle and its presupposition, see Millican,
‘Hume’s Sceptical Doubts concerning
38
Induction’, sect. 3.2 and especially sect. 10.2.
32
Moreover this sequence of argument seems to
39
be entirely deliberate, because it occurs very
explicitly twice, first within the main argument
at EHU 4.16 / 32–4, and then again in the
coda at EHU 4.21 / 37.
33
In the conclusion of Book 1 of the Treatise,
Hume’s attempt to meet the sceptical challenge
says very little about the issue of induction,
except as part of a general concern regarding
the role of ‘the imagination, or the vivacity
of our ideas’ (THN 1.4.7.3 / 265). There the
more pressing problems are those that threaten
inevitable error and contradiction (notably
96
9780826443595_Ch03_Final_txt_print.indd 96
the existence of body, the metaphysics of
causation, and the self-annihilation of reason),
which the simple assumption of uniformity
never does. The Enquiry’s response to the
Pyrrhonian sceptic, starting from a rejection
of extreme antecedent scepticism, would not
be nearly as effective against varieties of con-
sequent scepticism that bring to light genuine
contradictions – rather than simply lack of
ultimate grounding – in our faculties, and this
might explain why Hume very much down-
plays these more problematic topics in the
Enquiry. His attitude to them seems to be that
they are best left alone: for example, meta-
physical enquiries into the nature of matter
are likely to lead to contradiction or unin-
telligibility (EHU 12.14–15 / 153–5) unless,
perhaps, we fall back on a notion of matter so
empty as to be unexceptionable (EHU 12.16
/ 155). Hume’s final recommendation is for a
mitigated scepticism that inspires a suitable
degree of ‘doubt, and caution, and modesty’
(EHU 12.24 / 162), and which also focuses our
enquiries on ‘such subjects as are best adapted
to the narrow capacity of human understand-
ing’ (EHU 12.25 / 162), notably those where
we are able to progress either through math-
ematical demonstration (EHU 12.27 / 163) or
induction from experience (EHU 12.28–31 /
163–6).
David Owen, Hume’s Reason (Oxford: Oxford
University Press, 1999), p. 136.
Ibid., p. 132.
Barry Stroud, Hume (London: Routledge &
Kegan Paul, 1977), p. 54.
Antony Flew, Hume’s Philosophy of Belief
(London: Routledge & Kegan Paul, 1961).
Stove, Probability and Hume’s Inductive
Scepticism.
Stroud himself (Hume, pp. 56–7) reacted
against this, suggesting that what he saw as
Hume’s extreme scepticism could more plaus-
ibly be attributed to a ‘potentially regressive
aspect of the notion of reason or justification’
whereby evidence E can count as a reason
for believing P only if one has some reason R
for taking E as a reason. If we then ask about
the basis for R in turn, and continue in this
way, we get a regress which can apparently
be terminated only by ‘something we could
not fail to be reasonable in believing’ (ibid.,
2/9/2012 9:38:58 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
p. 62), such as an immediate experience or self-
evident truth. Hume’s invoking of UP within
his argument is indeed somewhat in this spirit,
but when considering UP’s own foundation, he
seems content to stop with (fallible) sensa-
tion or memory, not only with the certainty of
intuition or demonstration, while the appeal
to factual inference generates a circle rather
than an infinite regress. Nevertheless Stroud’s
account is illuminating, in stressing the seduc-
tive assumption that justification is required at
each step if scepticism is to be resisted. Hume’s
strategy outlined in section 1 above rejects this
by shifting the onus onto the sceptic.
40
Flew, Hume’s Philosophy of Belief, p. 171.
41
For some other passages in a similar spirit, see
note 125 below.
42
Tom Beauchamp and Thomas Mappes, ‘Is
Hume Really a Sceptic about Induction?’,
American Philosophical Quarterly 12 (1975),
pp. 119–29.
43
Barbara Winters, ‘Hume on Reason’, Hume
Studies 5 (1979), pp. 20–35 was perhaps most
influential in promoting the idea that Hume’s
notion of reason is ambiguous in this way.
Tom Beauchamp and Alexander Rosenberg,
Hume and the Problem of Causation
(Oxford: Oxford University Press, 1981);
N. Scott Arnold, ‘Hume’s Skepticism about
Inductive Inference’, Journal of the History
of Philosophy 21 (1983), pp. 31–55; Janet
Broughton, ‘Hume’s Skepticism about Causal
Inferences’, Pacific Philosophical Quarterly
64 (1983), pp. 3–18; and Annette C. Baier, A
Progress of Sentiments: Reflections on Hume’s
Treatise (Cambridge MA: Harvard University
Press, 1991) all gave slightly different anti-
deductivist readings, some of the nuances of
which are discussed by Don Garrett, Cognition
and Commitment in Hume’s Philosophy
(Oxford: Oxford University Press, 1997),
pp. 83–91.
44
Peter Millican, ‘Hume’s Argument concerning
Induction: Structure and Interpretation’, in
Stanley Tweyman (ed.), David Hume: Critical
Assessments (London: Routledge, 1995),
vol. 2, pp. 91–144; Garrett, Cognition and
Commitment.
45
Most obviously, the famous argument treats
probable argument as a potential founda-
tion for the Uniformity Principle, whereas a
97
9780826443595_Ch03_Final_txt_print.indd 97
deductivist must consider any merely probable
argument as evidentially worthless from the
start. If Hume were a deductivist, indeed,
then he could dismiss induction in a single
step with his Conceivability Principle. For
more detail on all this, see Millican, ‘Hume’s
Argument concerning Induction’, pp. 123–4,
136; ‘Hume’s Sceptical Doubts concerning
Induction’, pp. 155–6; Garrett, Cognition and
Commitment, pp. 86–8).
46
See in particular the passages quoted near the
end of the first paragraph of section 3 above.
These and others are cited in this connexion
by Millican, ‘Hume’s Argument concerning
Induction’, pp. 127, 136; ‘Hume’s Sceptical
Doubts concerning Induction’, pp. 161–2;
Garrett, Cognition and Commitment,
pp. 85–6.
47
Locke, Essay, IV.xvii.2.
48
e.g. Locke, Essay, IV.xv.2.
49
Garrett, Cognition and Commitment, p. 92.
50
Locke’s usage is somewhat variable, though I
consider perception to be more fundamental to
Lockean reason than inference (see my ‘Hume’s
Argument concerning Induction’, p. 137, or for
more detail, ‘Hume’s Sceptical Doubts concern-
ing Induction’, sect. 2). Note that both of these
are distinct from the intermediate idea charac-
teristic which Owen considers fundamental to
Lockean reason.
51
Garrett, Cognition and Commitment, pp.
91–2.
52
The modified interpretation first appeared in
Don Garrett, ‘Ideas, Reason and Skepticism:
Replies to My Critics’, Hume Studies 24
(1998), pp. 171–94, but his 2002 piece ‘The
Meaning of Hume’s Conclusion concerning
“Inductive” Inferences’ (in Peter Millican,
Reading Hume on Human Understanding,
pp. 332–4) was based directly on the two
relevant sections of his 1997 book (Cognition
and Commitment), reworded accordingly.
53
Garrett, ‘The Meaning of Hume’s Conclusion’,
p. 333.
54
Or, presumably, A to C. Garrett’s suggestion
(‘Ideas, Reason and Skepticism’, pp. 182–3)
is that in attempting to infer from A to D, we
first observe that A implies B intuitively (i.e.
self-evidently), leaving a gap between B and
D. We then set out to fill that gap, by noticing
that B implies C, and C implies D. We put
2/9/2012 9:38:58 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
these last two implications together, deducing
that B implies D (this is Garrett’s intermedi-
ate inference). Now from A implies B and B
implies D, we can deduce that A implies D.
55
See, for example, Peter Millican, ‘Hume on
Reason and Induction: Epistemology or
Cognitive Science?’, Hume Studies 24 (1998),
pp. 141–60, at pp. 145–7, and David Owen, ‘A
Reply to My Critics’, Hume Studies 26 (2000),
pp. 323–37, at pp. 329–30.
56
Samuel Johnson, A Dictionary of the English
Language (London, 1756).
57
‘To deduce’ is defined in three clauses: ‘1.
To draw in a regular connected series. 2.
To form a regular chain of consequential
propositions. 3. To lay down in regular order.’
‘Ratiocination’ is defined in just one clause:
‘The act of reasoning; the act of deducing
consequences from premises.’
58
Also, of course, Hume’s own theory of induct-
ive inference implies that it typically does not
proceed in a stepwise manner, but essentially
reduces to conception (see THN 1.3.7.5n20 /
96–7n); yet he never hints that terms such as
‘argument’, ‘inference’, ‘proof’ or ‘reasoning’
are thereby rendered inappropriate to these
transitions of thought. So it is hard to see how
he could consistently refuse to apply them – on
grounds of immediacy – to ‘intuitive inference’.
59
Such an emphasis comes later, with the positive
account in terms of instinctive custom (e.g. at
THN 1.3.8.13 / 103–4 and EHU 5.8 / 46–7).
60
At least, this looks like a plausible implication,
just as one invalid step within a mathemat-
ical proof is enough to render the entire proof
invalid. But as we shall see later (section 4),
things are not quite so straightforward here.
61
For another way of sharpening this sort of
objection, see Millican, ‘Humes Old and New,
sect. VI, which expands on Millican, ‘Sceptical
Doubts concerning Induction’, pp. 158–60.
There I focus on the very last step of Hume’s
argument, whereby he concludes that because
factual inference is founded on UP, and UP is
not founded on reason, it follows that factual
inference is not founded on reason. This step
looks very dubious if ‘reason’ here is supposed
to mean stepwise ratiocination (or, indeed,
higher-level argument).
62
See note 10 above for equivalent passages in
the Abstract and Enquiry versions.
98
9780826443595_Ch03_Final_txt_print.indd 98
63
See for example Owen, Hume’s Reason,
pp. 9–10, 120–2, 127–30, 141, 148. The books
of both Owen and Garrett present only the
Treatise version of the argument, and indeed
Owen’s analysis hardly mentions the Abstract
or Enquiry. Garrett takes relevant quotations
from the later works, but states without analysis
that ‘the structure and language of the other ver-
sions of the argument are parallel’ to that in the
Treatise (Cognition and Commitment, p. 82).
64
Descartes’s Meditations, for example, presents
the Ontological Argument for the existence
of a perfect God, and then appeals to God’s
non-deceptive nature to vindicate various
factual beliefs about the unobserved, all appar-
ently without any essential reference to the
Uniformity Principle.
65
Or – if we take the Enquiry version – in any
deliverance of the senses or intuition.
66
This objection goes back to my ‘Hume on
Reason and Induction’, sect. VII (1998), and is
also discussed by Garrett, ‘Ideas, Reason and
Skepticism’, p. 187; Millican ‘Hume’s Sceptical
Doubts concerning Induction’, pp. 157–8;
Louis Loeb, ‘Psychology, Epistemology,
and Skepticism in Hume’s Argument about
Induction’, Synthese 152 (2006), pp. 321–38,
at pp. 328–9; and Abraham Sesshu Roth,
‘Causation’, in Saul Traiger (ed.), The
Blackwell Guide to Hume’s Treatise (Oxford:
Blackwell, 2006), pp. 95–113, at pp. 108–11.
67
The case of faulty factual arguments (e.g. in
Hume’s coda at THN 1.3.6.10 / 91 and EHU
4.21 / 36–8) is less clear, because if they pre-
suppose UP, then the famous argument – as
interpreted by Garrett and Owen – can still
get a purchase on them. For critical discus-
sion, see Loeb, ‘Psychology, Epistemology, and
Skepticism’ (p. 329), who goes on to suggest
his own explanation of why Hume fails to
consider faulty arguments here: ‘The answer
must be that Hume imposes an epistemic
constraint on any causal explanation of induct-
ive inference: the explanation of our making
inductive inferences must be compatible with
their being justified’ (p. 330). Helen Beebee,
Hume on Causation (London: Routledge,
2006), pp. 55–6, takes a similar line, and both
are discussed in my ‘Humes Old and New’,
pp. 186–8. In brief, I find their approach text-
ually unsupported and also in tension with
2/9/2012 9:39:00 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
the sceptical tone of Hume’s famous argument
and of his later references to it. A far simpler
solution is to see ‘reason’ as referring to our
cognitive faculty – see section 3.2 below.
68
Richard Price, A Review of the Principal
Questions and Difficulties in Morals (London:
1758), p. 34.
69
Ibid., p. 40n.
70
Owen also adds a note at this point: ‘The
qualification, “prior to our engaging in
probable reasoning”, is important, because
Hume thinks that once we are engaged in
the practice of probable reasoning, we come
to believe the uniformity principle and use it
in probable reasoning. . . . This requires an
account of how we first engage in probable
reasoning, before the principle is available to
us.’ (Hume’s Reason, p. 128n30)
71
Owen, Hume’s Reason, pp. 128–30.
72
Garrett, Cognition and Commitment, p. 94.
73
Owen, Hume’s Reason, pp. 132–4.
74
Garrett, ‘Ideas, Reason and Skepticism’,
pp. 180–1.
75
Unlike Garrett and Owen, however, I do not
take the principle in question, ‘that like objects,
plac’d in like circumstances, will always
produce like effects’, to be identical to the
Uniformity Principle. The former concerns the
consistency of events within our experience,
whereas UP concerns the evidential relevance
of observed to unobserved. Without UP,
experienced consistency (or, indeed, any other
experienced pattern) could not be extrapolated
from past to future.
76
Owen, Hume’s Reason, pp. 131, 170–1.
77
Garrett, ‘Ideas, Reason and Skepticism’, p. 184
78
Millican, ‘Hume on Reason and Induction’,
p. 153.
79
This is the main respect in which my own
views have changed over time, largely in
response to Don Garrett’s criticisms. Most
other aspects of my previous interpretation
remain in place; for example it will become
clear in section 4 below that a perceptual
notion of reason makes a highly plausible
Humean target, even if we do not suppose that
he was employing such a notion himself within
the famous argument.
80
Garrett, Cognition and Commitment, p. 92.
81
Garrett, ‘The Meaning of Hume’s Conclusion’,
p. 333.
99
9780826443595_Ch03_Final_txt_print.indd 99
82
Garrett, ‘Ideas, Reason and Skepticism’, p. 184.
83
Ibid. It seems to be a logical rather than psycho-
logical point that some such step must be pre-
sent in every inductive inference, given that – as
Garrett acknowledges – the ‘supposition of UP’
that it exhibits can be entirely unconscious.
84
Locke starts his chapter ‘Of Reason’ (Essay,
IV.xvii) with the remark that ‘The Word
Reason in the English Language has different
Significations: sometimes it is taken for true,
and clear Principles: Sometimes for clear, and
fair deductions from those Principles: and
sometimes for the Cause, and particularly the
final Cause.’ He then goes on to say that his
chapter concerns yet another signification,
for that ‘Faculty in Man . . . whereby Man is
supposed to be distinguished from Beasts, and
wherein it is evident he much surpasses them’.
Earlier, at Essay II.xxi.17–20, Locke ridi-
cules the language of ‘faculties’ as a source of
philosophical error. For more on his view, see
Millican, ‘Hume’s Sceptical Doubts concerning
Induction’, sect. 2, and cf. note 50 above.
85
As acknowledged in Hume’s letter of
10 January 1743 (LDH 1.45, 19).
86
Francis Hutcheson, Synopsis of Metaphysics
(1744), trans. Michael Silverthorne, in Francis
Hutcheson, Logic, Metaphysics, and the
Natural Sociability of Mankind (Indianapolis:
Liberty Fund, 2006)
87
Francis Hutcheson, A Short Introduction to
Moral Philosophy (1747), ed. Luigi Turco
(Indianapolis: Liberty Fund, 2007), p. 25.
88
Francis Hutcheson, An Essay on the Nature
and Conduct of the Affections. With
Illustrations on the Moral Sense, 3rd edn
(London: 1742), pp. 30–1.
89
Ibid., pp. 219–20.
90
Note that the quoted paragraph also treats
‘the Intellect’ as just another elegant variation
on ‘Reason’ and ‘the Understanding’. Hume
does the same, albeit more rarely (DNR 3.153,
3.156), though he quite often refers in a simi-
lar spirit to the ‘intellectual faculties’ (THN
1.3.12.20 / 138, 2.3.8.13 / 437; EHU 5.5n8 /
43–4n, 9.6 / 108; EPM 1.9 / 173, EPM App.
1.11 / 290, 13 / 291, 18 / 293, 3.9 / 307; ‘Of
the Standard of Taste’, E 240–1). Garrett talks
of Hume as giving ‘an argument against the
intellect’ (Cognition and Commitment, p. 20),
but this is misleading unless ‘the intellect’ here
2/9/2012 9:39:00 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
is understood to mean ‘the intellect conceived
of as a faculty of non-sensory ideas’ (a con-
ception that Garrett traces through Descartes,
Spinoza and Leibniz, but is not shared by
Locke, Berkeley or Hume).
91
See, for example, Anthony Ashley Cooper,
Lord Shaftesbury, Characteristicks of Men,
Manners, Opinions, Times, 3 vols, 3rd edn [the
edition purchased by Hume in 1726] (London,
1723), vol. 2, II.ii, p. 118; Joseph Butler, The
Analogy of Religion, Natural and Revealed,
to the Constitution and Course of Nature,
2nd edn, corrected (London, 1736), I.vi,
p. 174; Price, Review, I ii, p. 23.
92
See, for example, Henry Home, Lord Kames,
Essays on the Principles of Morality and
Natural Religion, 2nd edn (London, 1758),
essay VII, p. 268n, and James Beattie, An
Essay on the Nature and Immutability
of Truth; in Opposition to Sophistry and
Scepticism (Edinburgh, 1770), I.i, pp. 37–8.
James Oswald, An Appeal to Common Sense
in Behalf of Religion (Edinburgh, 1766),
I.ii.1, p. 80 and Thomas Reid, Essays on the
Intellectual Powers of Man (Edinburgh, 1785),
VII.i, p. 671 are likewise keen to insist on a
narrow use of ‘reasoning’, distinguished from
‘judging’, though they allow both operations
to be subsumed under ‘reason’.
93
For other relevant passages from the Treatise,
see, for example, THN 1.3.6.4 / 88–9,
1.3.13.12 / 149–50, 1.4.1.1 / 180, 1.4.1.12
/ 186–7, 1.4.2.14 / 193, 1.4.7.7 / 267–8,
2.3.3.2–6 / 413–6, 3.1.1.16–18 / 462–3,
and 3.1.1.26 / 468–9. For passages from the
Abstract, see Abs. 11 / 650–1, 27 / 657, and
from the Enquiry, EHU 4.0–1 / 25, 5.5 / 43,
5.22 / 55, 7.28 / 76 and 9.0–1 / 104. Further
examples may be found in Hume’s other works.
94
Together with Hume himself (EHU 1.14 /
13–14, 8.22 / 93), see, for example, René
Descartes, Fourth Meditation, in The
Philosophical Writings of Descartes, trans.
John Cottingham, Robert Stoothoff and
Dugald Murdoch (Cambridge: Cambridge
University Press, 1984), vol. 2, pp. 39–40;
Locke, Essay, II.vi.2; George Berkeley,
A Treatise concerning the Principles of
Human Knowledge, ed. Jonathan Dancy
(Oxford: Oxford University Press, 1998),
I.27. References to Hume’s contemporaries
Hutcheson, Price and Reid will follow.
100
9780826443595_Ch03_Final_txt_print.indd 100
95
Reid, Intellectual Powers, I.vii, pp. 67–8.
96
Note, for example, the general division
within the Treatise between Book 1 ‘Of the
Understanding’ and Book 2 ‘Of the Passions’
(including Part 3 ‘Of the will and direct
passions’), and also the footnote at THN
1.3.7.5n20 / 96, where Hume criticizes the
‘universally receiv’d’ threefold ‘division of the
acts of understanding’ which Reid describes.
97
Hutcheson, A Short Introduction to Moral
Philosophy, pp. 25–6.
98
Price deprecates ‘the division which has
been made by some writers, of all the
powers of the soul into understanding and
will; the former comprehending under it,
all the powers of external and internal
sensation, as well as those of judging and
reasoning’. By contrast, he says, ‘I all along
speak of the understanding, in the most
confined and proper sense of it . . . and
distinguished from the powers of sensation’
(Review, I.ii, p. 20n). Note, however, that
Price implicitly equates the understanding
with reason (ibid., p. 23) thus using ‘reason’
in a broader sense than those such as Kames
(cf. note 92 above) who exclude intuition
from its scope.
99
Hutcheson, Synopsis of Metaphysics, p. 112.
The original Latin of the final clause is ‘Ad
Intellectum, referentur Sensus’.
100
Sect. 6 ends with a short paragraph on
‘Natural associations of ideas’, which
Hutcheson sees as playing an important role
in both imagination and memory; sect. 7
briefly discusses what is pleasing or distress-
ing to the senses, and our consequent judge-
ments (of good and evil) and passions; then
sect. 8 discusses habit, and sect. 9 relative
ideas and judgements.
101
Hutcheson, Illustrations on the Moral Sense,
I, p. 215.
102
Price, Review, pp. 19–20.
103
For example Locke: ‘the understanding
. . . is the most elevated Faculty of the Soul
. . . Its searches after Truth, are a sort of . . .
Hunting’ (Essay, Epistle to the Reader, para-
graph 1); David Hartley: ‘The Understanding
is that Faculty, by which we . . . pursue Truth,
and assent to, or dissent from, Propositions.’
(Observations on Man, His Frame, His Duty,
and His Expectations, Bath and London,
1749, vol. 1, Introduction, p. iii).
2/9/2012 9:39:01 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
104
Other passages that identify reason with the
discovery of truth, though usually less expli-
citly, are at THN 2.3.3.3 / 414, 2.3.3.5–6
/ 415–6, 2.3.3.8 / 417, 3.1.1.4 / 456–7,
3.1.1.19n69 / 464n, 3.1.1.25–7 / 467–70,
3.2.2.20 / 496, THN App. 1 / 623; EPM 1.7
110
/ 172, EPM App. 1.6 / 287. All of these occur
in a context where Hume is contrasting
reason (or the understanding) with conative
rather than cognitive notions, thus corrob-
orating its identification as the overarching
cognitive faculty.
105
This note (which can be found as a Textual
Variant on p. 177 of my edition of Hume’s
Enquiry concerning Human Understanding,
Oxford World’s Classics, 2007) is of particu-
lar interest because it credits Hutcheson with
distinguishing between ‘the Understanding’
and ‘That Faculty . . . by which we perceive
111
Vice and Virtue’, although Hutcheson himself
considered the moral sense to be one of the
‘reflexive or subsequent sensations’, thus fall-
ing within the domain of the understanding.
Price, Review, p. 12n, mentions Hume’s note
in the course of criticizing Hutcheson.
106
Notice that Hume is implying that if we
could observe a conjunction of cause and
effect, then we could form such a ‘conclu-
sion . . . concerning . . . existence’ and ‘satisfy
our reason in this particular’. So as in the
famous argument concerning morals (cf. the
quotation from THN 3.1.1.12 / 459 near
the beginning of section 3.1 above), Hume is
clearly here treating causal, factual inference
as an operation of reason.
107
Such nuances can apply with many words that
are associated with some achievement. For
example, a cure that does not work is strictly a
contradiction in terms, but it is fairly natural to
say, in appropriate circumstances, ‘that cure is
useless and ought to be banned’.
108
An analogy here would be to an accounting
error within a company, which on a broad
interpretation could refer to any error in the
accounts (including faulty data from external
sources), but on a narrower interpretation
would mean an error due to the accountants
112
themselves.
109
On the conative side, Hume hardly ever
speaks of ‘faculties’, explicitly referring to ‘the
will’ and ‘the passions’ as faculties just once
101
9780826443595_Ch03_Final_txt_print.indd 101
each (at THN 2.3.3.9 / 417–8 and 2.2.2.16 /
339 respectively). Judgemental ‘taste’ is called
a faculty in ‘Of the Standard of Taste’, E
240–1, and spoken of as having ‘a product-
ive faculty’ in a famous passage at EPM App.
1.21 / 294.
The model of a canvas is obviously most
appropriate to visual ideas, which indeed
seem to dominate Hume’s thought, although
ideas may correspond to any of the senses –
including internal ‘reflection’ – and only the
ideas of sight and touch will be spatially
arranged (not necessarily within a single
space). Note that he takes all of our ideas to
be sense-copied; hence as Garrett observes (cf.
note 90 above), Hume denies any separate
faculty that can take a ‘pure and intellectual
view’ of ‘refin’d and spiritual’ ideas, unsullied
by sensory input (THN 1.3.1.7 / 72).
At least, this seems to be what Hume is saying
in THN 1.4.7.3 / 265. At THN 1.3.9.3 /
107–8, he appears instead to take the force
and vivacity of the ‘impressions or ideas of
the memory’ – like that of ‘an immediate
impression’ – as itself constituting assent,
thus providing a basis for explaining the
assent that derives from causal inference.
The relationship between memory and the
imagination remains somewhat obscure,
though Hume’s talk of ‘impressions of
the memory’ (‘somewhat intermediate
betwixt an impression and an idea’, THN
1.1.3.1 / 8) suggests that the memory is
furnishing ideas that are sufficiently firm
and vivid – sufficiently impression-like – to
establish copy-ideas in the imagination:
‘The impressions of the memory never
change in any considerable degree; and
each impression draws along with it a
precise idea, which takes its place in the
imagination, as something solid and real,
certain and invariable.’ (THN 1.3.9.7 / 110).
If this is right, then all the ideas that are
actually involved in thinking lie within the
imagination, and the role of the senses and
memory is to supply the ‘impressions’ from
which those ideas derive.
See THN 1.3.7.6 / 97, 1.3.9.3 / 107–8,
1.3.11.11 / 128, 1.3.14.1 / 155–6, 1.3.14.31 /
169–70; and for references from the Abstract
and Enquiry see note 129 below.
2/9/2012 9:39:02 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
113
I call it a ‘virtual’ stage to reflect Hume’s
comment at THN 1.4.6.4 / 253 that ‘The
comparison of the theatre must not mislead
us. They are the successive perceptions only,
that constitute the mind’.
114
THN 1.4.2.22 / 198 gives a case intermedi-
ate between passivity and activity, in which
the imagination, having been ‘set into any
train of thinking, is apt to continue even
when its object fails it, and, like a galley
put in motion by the oars, carries on its
course without any new impulse’. Note that
the listed references involving fictions are
confined to those that involve characteristic
Humean fictions of philosophical interest,
rather than arbitrary combinations of ideas
(i.e. ‘mere fictions of the imagination’ as at
THN 1.3.5.4–5 / 85, 1.3.7.7 / 97, 1.3.9.3 /
108, 1.4.3.1 / 219; EHU 5.12–13 / 49–50,
6.3 / 57).
115
Recall from section 3.2 that the replaced
footnote at THN 2.2.7.6n / 371n said ‘the
understanding’ here instead of ‘reason’.
116
It is striking, for example, that these are the
only two occurrences in Hume’s writings
of the phrase ‘offspring of the imagina-
tion’. Presumably he was forced to place the
footnote at the end of the section to minimize
type resetting.
117
The understanding is also identified with
the imagination at THN 1.3.8.13 / 104 and
2.3.9.10 / 440.
118
Don Garrett, Cognition and Commitment,
p. 92, was, I believe, the first to note this
possibility, which is crucial if Hume’s use of
‘reason’ within his argument is to be under-
stood as sincere rather than a target. But of
course Garrett’s interpretation of ‘determin’d
by reason’ is very different from my own.
119
See, for example, EHU 1.13–14 / 13–14 for
the equation of ‘faculties’ with ‘powers’, and
also THN Intro. 4 / xv, 1.3.10.9 / 123. The
same equation is repeatedly found in Locke,
e.g. Essay II.vi.2, II xxi 15, 17, 20.
120
Locke, Essay, II.xxi.20.
121
In a recent debate (Don Garrett and Peter
Millican, Reason, Induction and Causation in
Hume’s Philosophy, IASH Occasional Paper
19, Edinburgh: Institute for Advanced Studies
in the Humanities, University of Edinburgh,
2011), Don Garrett argues that in inductive
102
9780826443595_Ch03_Final_txt_print.indd 102
inference, ‘the imagination in the narrow
sense is performing the customary transition’
(p. 22), thus denying that custom is reassigned
to reason along with ‘our demonstrative
and probable reasonings’. Hence he sees the
distinction alluded to at THN 1.3.9.19n22 /
117–18 as quite distinct from that drawn at
THN 1.4.4.1 / 225, a position I find rather
implausible, given that their stated rationale
is so similar, namely, to distinguish within
the imagination between the principles that
‘are the foundation of all our thoughts and
actions’ and those that give rise to ‘whimsies
and prejudices’.
122
Even in the Treatise, Hume never says in so
many words that custom is a process of the
imagination, though THN 1.3.6.4 / 88–9 and
1.3.7.6 / 97 strongly suggest this.
123
In the Enquiry, unlike the Treatise (e.g. THN
1.3.6.16 / 94, 1.3.8.6 / 100–1), Hume is care-
ful to distinguish between the associational
relation of causation (discussed at EHU
5.18–19 / 53) and custom (EHU 5.20 / 53–4).
124
We would not usually describe the climber as
‘supported only by’ a rock to which the rope
is attached unless the rock was considered
potentially less secure than the rope (e.g.
suppose the attachment is to a spur of rock
that is in imminent danger of cracking –
we might well then say that the climber is
‘supported only by the spur’).
125
Together with those quoted earlier, relevant
passages include: ‘. . . our reason . . . or, more
properly speaking, . . . those conclusions
we form from cause and effect . . .’ (THN
1.4.4.15 / 231); ‘. . . these emotions extend
themselves to the causes and effects of that
object, as they are pointed out to us by rea-
son and experience . . .’ (2.3.3.3 / 414); ‘. . .
the operations of human understanding . . .
[include] the inferring of matter of fact . . .’
(3.1.1.18 / 463).
126
Both Barry Stroud (Hume, pp. 60–2) and
John Kenyon (‘Doubts about the Concept
of Reason’, Proceedings of the Aristotelian
Society, supp. vol. 59, 1985, pp. 249–67,
at pp. 255–7) attribute this to Hume in the
context of his argument concerning induction,
but neither justifies the attribution, and there
is little evidence of it in Hume’s text (cf. note
39 above).
2/9/2012 9:39:02 PM
 HUME’S ‘SCEPTICISM’ ABOUT INDUCTION
127
This assumes the internalist perspective which
dominated the early modern period. A mod-
ern Humean might well take an externalist
approach, but given Hume’s explicit response
to the sceptic in Enquiry 12 (as described
in section 1 above), I am not persuaded by
130
Louis Loeb’s claim that ‘In light of the massive
evidence that Hume is not a skeptic about
induction, he must reject [the] internalist way
of thinking.’ (‘Psychology, Epistemology, and
131
Skepticism’, p. 333). Kenneth Winkler empha-
sizes how Enquiry 12 supports a more sceptical
reading of Enquiry 4: see ‘Hume’s Inductive
Skepticism’, in Margaret Atherton (ed.),
The Empiricists (Lanham, MD: Rowman &
Littlefield, 1999), pp. 183–212, at pp. 193–200.
128
At THN 1.3.9.19 / 117, 1.3.13.9 / 147,
1.3.14.21 / 165, 1.4.4.1 / 225, and 1.4.7.3 /
132
265.
129
At Abs. 15 / 652, 16 / 652, 21 / 654 (twice),
25 / 656; EHU 5.4–5 / 42–3, 5.5 / 43 (twice),
5.8 / 46, 5.11 / 48, 5.20 / 54, 5.21 / 54–5,
6.4 / 58, 7.29 / 76–7, 8.5 / 82, 8.21 / 92, 9.5
/ 106, 9.5n20 / 107, and 12.22 / 159. Note
that this contrast cannot be accounted for in
terms of Hume’s moving away from the foun-
dational metaphor more generally. On the
contrary, he says that induction is ‘founded’
on the relation of cause and effect, or experi-
ence, or the Uniformity Principle, and that it
is not ‘founded’ on reasoning, argument, or
103
9780826443595_Ch03_Final_txt_print.indd 103
any process of the understanding, signifi-
cantly more in the Enquiry – EHU 4.4 / 26,
4.14 / 32 (twice), 4.15 / 32, 4.21 / 37, 9.5 /
106 and 12.29 / 164) – than he does in the
Treatise – THN 1.3.6.4 / 88 and 1.3.6.7 /
89–90 (twice).
Other aspects of the logic of Hume’s foun-
dational relation are explored in Millican,
‘Hume’s Sceptical Doubts concerning
Induction’, sect. 10.1.
There is a thematic parallel here with Hume’s
account of causation, which is also commonly
thought of as sceptical, but in fact provides
him with a positive basis for applying causal
explanation to the human world. For more
on this, see Peter Millican, ‘Hume, Causal
Realism, and Causal Science’, Mind 118
(2009), pp. 647–712.
For numerous discussions on the topics of
reason and induction, I am extremely grateful
to Louis Loeb, David Owen and especially
Don Garrett, as well as many other members
of the Hume Society at its various conferences.
I am also grateful to Henry Merivale, Hsueh
Qu and especially Dan O’Brien for comments
on this paper, and to the Edinburgh Illumni
for providing the David Hume Fellowship at
Edinburgh, thus giving me the opportunity to
work in the delightful context of the Institute
for Advanced Studies in the Humanities, over-
lapping with Don Garrett’s tenure there.
2/9/2012 9:39:02 PM
 4
THE PSYCHOLOGY AND
EPISTEMOLOGY OF HUME’S
ACCOUNT OF PROBABLE
REASONING
Lorne Falkenstein
Over the course of Treatise 1.3, Hume pre-
sented what he called a ‘system’ of probable
reasoning. He then went on, in Treatise 1.4,
to argue that sceptical objections would
leave us entirely incapable of belief were our
natural inclinations not too strong for philo-
sophical conclusions to be able to restrain
our inferences. In the Enquiry concerning
Human Understanding he reversed this pro-
cedure, first offering ‘sceptical doubts’ about
the legitimacy of our inferences concerning
matters of fact, then a ‘sceptical solution of
these doubts’, and finally a conclusion that
we can only be legitimately sceptical of
claims in religion and school metaphysics,
not everyday experience or natural science.
Despite the more optimistic tone, the the-
ory of the Enquiry is built on the same two
principles as the ‘system’ of the Treatise: the
principle of the association of ideas, and the
principle of the genesis of belief in the unob-
served as a consequence of association with
sensory experience or memories. The common
‘system’ of the Treatise and the Enquiry is
sceptical because it takes our beliefs to be the
product of naturally occurring psychological
104
9780826443595_Ch04_Final_txt_print.indd 104
mechanisms rather than logically sound
judgement, and declares those beliefs to be
ultimately unjustifiable. Despite this sceptical
result Hume was able to provide for a logic of
probable reasoning, grounded on natural, but
unjustifiable beliefs. How he did so is still not
well understood. That he was able to do so is
one of his great achievements.
1. THE SYSTEM OF THE TREATISE
All kinds of reasoning consist in noth-
ing but a comparison, and a discovery of
those relations, either constant or incon-
stant, which two or more objects bear to
each other. (THN 1.3.2.2 / 73)
In the Treatise Hume’s presentation and
defence of his ‘system’ took the form of a tor-
turous journey down the dead-end lanes and
twisted turns of a meditative path of discov-
ery, supplemented by appeals to observations
and experiments, worries over contrary evi-
dence, and the introduction of refinements to
accommodate recalcitrant data. His project
was to inquire into the basis of reasoning,
2/9/2012 9:39:12 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
particularly probable reasoning. Reasoning
consists in inferring unknown from known
objects by means of a relation between the
two (THN 1.3.1.2 / 70, 1.3.2.2 / 73). The
relations on which all reasoning is based are
‘philosophical relations’, which are discov-
ered by comparing objects with one another
(THN 1.3.2.2 / 73). There are also ‘natural
relations’. A natural relation is not discov-
ered by comparison or appealed to in order
to discover or justify a conclusion. Whether
we are aware of it or not, it exercises an
influence on the imagination, impelling us
to form an idea of an object. This produces
a kind of instinctive, counterfeit reasoning
(THN 1.1.5 / 13–15, cf. 1.1.4 / 10–13).
Relations can be divided into two main
kinds: ‘inconstant’ relations, which can alter
even while the compared impressions or ideas
remain the same (e.g. relations of contiguity
and distance in space or time); and ‘constant’
relations, which cannot change without a
change in the compared objects (e.g. rela-
tions of resemblance). Demonstrative rea-
soning, yielding certainty, is founded on the
latter, while probable reasoning is founded
on the former (THN 1.3.1.1–2 / 69–70,
1.3.2.1 / 73).
Though we discover a number of con-
stant and inconstant relations by compar-
ing objects, Hume maintained that there is
only one relation that can serve as a basis
for demonstrative reasoning, the relation of
degrees in quantity, and only one that can
serve as a basis for probable reasoning, the
relation of cause and effect. The other con-
stant relations can always be ‘intuited’ with-
out the need for any demonstration and the
other inconstant relations can only be ‘per-
ceived’ and not used as a basis for inference
to the unobserved (THN 1.3.1.2 / 70, 1.3.2.2
/ 73–4). Thus all probable reasoning reduces
to causal reasoning.
105
9780826443595_Ch04_Final_txt_print.indd 105
2. THE PATH TO THE FIRST PRINCIPLE
One of the infelicities of Hume’s presentation
is this precipitate assertion that the causal
relation is an inconstant relation, and the only
relation on which probable reasoning can rest.
While Hume never adequately justified the lat-
ter claim,1 he went on to give an argument for
the former. Our ideas of cause and effect could
not be ideas of any of the observed qualities
of objects because all objects are causes and
effects, and there is no quality that all objects
share in common. For any quality we might
pick on, there is some object that is a cause
or an effect even though it does not have that
quality. The idea must therefore be the idea
of a relation (THN 1.3.2.5 / 75) – indeed, of
a relation that can change while the objects
remain the same.
At this point a further infelicity arises.
Hume simply assumed that causal relations
are not immediately perceived or intuited
upon comparing objects, as we immediately
perceive relations of contiguity or immedi-
ately intuit relations of resemblance, but are
instead ‘deriv’d’ from some other relation
(THN 1.3.2.6 / 75). It is only much later that
a reason is given for this assumption (THN
1.3.6.1 / 86–7, explained more fully at Abs.
11–12 / 650–1). If we could perceive or intuit
the existence of causal relations, then we
would be able to tell upon first acquaintance
with a pair of objects whether or not they
are causally related. But we cannot do so.
Granting that causal relations are not
immediately perceived but are instead derived
from some other relations, what might those
relations be? The only relations we discover
when we compare those objects we con-
sider to be causes and effects are contiguity
in space and succession in time. But while
these relations may hold between causes
and effects,2 they also hold between objects
2/9/2012 9:39:13 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
that we consider to be only accidentally con-
joined. And, Hume claimed, we think that
causes and effects are not just accidentally
conjoined but necessarily connected.
This is tricky. How can we think that there
is a necessary connection between causes and
effects if we cannot discover any other rela-
tions between them than contiguity in space
and succession in time? Occasional sug-
gestions to the contrary notwithstanding,3
Hume did not want to say that we have no
idea of necessary connection. A necessary
connection is simply a connection that has
to be present and cannot be broken. A har-
ness is a connection between a horse and
a carriage. If the harness had to run from
the horse to the carriage and nothing could
break it, it would be a necessary connection.
The idea of a necessary connection between
causes and effects is similarly the idea of
a ‘tye or connexion . . . which binds them
together’ (EHU 4.10 / 29). Hume’s claim at
this stage in his meditations was not that we
are not thinking of anything when we think
of a tie reaching across time and space to
bind cause and effect together. It was rather
that we cannot discover exactly what does
the job. We only ever see the horse and the
carriage. The apparatus harnessing the two
together is not apparent. This does pose a
problem: if we cannot discover any harness,
why do we think it is there – indeed, that it
must be there and cannot be broken?
Hume proceeded to try the patience of
his reader further by pretending to have no
answer to this question and affecting the need
to look for one by investigating two related
questions: (i) why we consider it necessary
that every event have a cause and (ii) why
we consider this particular cause necessar-
ily to have that particular effect. In response
to the first, he argued that our belief that
every event must have a cause could not be
106
9780826443595_Ch04_Final_txt_print.indd 106
founded on intuition or demonstration and
concluded that it must therefore be based
on experience (THN 1.3.3 / 78–82). Oddly,
he made no attempt to argue that the same
answer must be given to the second question.
It is only later (THN 1.3.6.1 / 86–7) that a
reason is given for concluding that our infer-
ences from cause and effect are not based
on demonstration. The reason goes back to
an observation on the nature of imagina-
tion Hume made at THN 1.1.3.4 / 10. The
only limitation on the imagination is that its
ideas come from things that have been pre-
viously encountered in sensory experience.
Once given ideas, it can separate and rear-
range them in any way whatsoever. Given an
object at a place and a time, the imagina-
tion can conceive any object whatsoever at
any of the contiguous places at the earlier
or later times. Since causes and effects exist
at distinct times, any particular cause could
be imagined to be followed by anything and
any particular effect preceded by anything.4
But were it intuitively obvious or demon-
stratively provable that this particular cause
must have that particular effect, any other
alternative would be inconceivable (THN
1.3.6.1 / 86–7).5
Postponing this argument for the moment,
Hume (at THN 1.3.4 / 82–4) simply took it
for granted that the connection between par-
ticular causes and their effects can only be
known by experience and proceeded to ask
what sort of experience does this. In the pro-
cess, he presented himself as suddenly dis-
covering a new, third relation obtaining in
cases of causality (THN 1.3.6.2–3 / 87–8).
The relation is not discovered when compar-
ing individual instances of cause and effect,
but only when comparing multiple instances.
All instances resembling the cause are spa-
tially contiguous with instances resembling
the effect, and precede them in time.
2/9/2012 9:39:13 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
Hume continued to express puzzlement
over how this new relation of ‘constant con-
junction’ could lead us to conclude that this
particular cause must necessarily have that
particular effect. We do not discover anything
in multiple instances that could not be found
in just one instance, and we do not discover
anything in one instance that would justify
the conclusion. Nor could we take the expe-
rience of a constant conjunction to establish
the likelihood of a necessary connection, or
the likelihood that the causes contain some
quality, unknown to us, that gives them the
power to bring about their effects.6 Since
all we perceive are the observable relations
between causes and effects, none of which is
a necessary connection, and the observable
qualities of the causes, none of which is a
power, the most we could infer is that, in the
past, objects like the cause have been con-
tiguous to and followed by objects like the
effect, and that, in the past, the set of quali-
ties characteristic of the causes has included
some further, unknown power. But we are
in no position to infer that similar rela-
tions must obtain in the future or that simi-
lar collections of observed qualities will be
accompanied by similar powers in the future
(THN 1.3.6.8 / 90 and 1.3.6.10 / 91). The
new relation of constant conjunction could
only lead us to draw these inferences with the
aid of a further supposition, that what has
been observed to occur regularly in the past
will continue to be observed to occur in the
future. But this principle is not demonstra-
bly true, because there is no contradiction in
conceiving a change in the course of nature
(THN 1.3.6.5 / 89). Moreover, it cannot be
proven by appeal to past experience, because
the question at issue is precisely why we
should take regularities in past experience to
establish a rule for what will happen in the
future (THN 1.3.6.7 / 89–90).7
107
9780826443595_Ch04_Final_txt_print.indd 107
At this point in the course of his medita-
tions, Hume finally felt prepared to reveal8
the first principle of the ‘system’. Though
constant conjunction provides us with no
justification for inferring that causes and
effects are necessarily connected, it is a
‘natural’ relation, which impels the imagin-
ation to call up an associated idea when
presented with its partner. It therefore
produces a kind of counterfeit, instinct-
ive ‘reasoning’.9 Observing objects of one
sort being customarily followed by objects
of another sort trains a habit of thinking
into the mind. Once developed, the habit
induces the imagination to form an idea of
an object upon encountering its customar-
ily conjoined partner, even in the absence of
perceiving any tie binding the two together,
even in the absence of having any reason to
suppose that the future will be like the past,
and even in the absence of any recollec-
tion of or reflection upon the past instances
(THN 1.3.8.13 / 103–4). Although Hume
did not draw the conclusion until very much
later (THN 1.3.14), this is why, even though
we cannot discover any tie that necessarily
binds cause and effect together, we think
that there must be such a thing.
3. THE PATH TO THE SECOND
PRINCIPLE
After this long journey of discovery, the
second principle of the ‘system’ was quickly,
though not easily, uncovered. Hume began
by noting that causal inference takes place
only when one of the two associated objects
is experienced or remembered and the other
is not. When both objects are experienced or
remembered, there is no occasion to imagine
either. And when neither is either experienced
2/9/2012 9:39:13 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
or remembered, we feel impelled to imagine
the one upon having occasion to imagine the
other, but do not form any belief in the exist-
ence of either (THN 1.3.2.2 / 73–4, 1.3.6.2 /
87, 1.3.4 / 82–4).
Seeking for an explanation for this variation,
Hume noted that the objects of experience and
memory are believed to exist (THN 1.3.5.7 /
86). This suggests that the belief we get as a
consequence of causal association might be
due to some sort of transfer from an experi-
ence or memory to an associated object. When
an object is believed to exist or have existed,
the relation of constant conjunction induces
us not merely to conceive an associated object
but to form a belief in that object’s existence at
a contiguous place and the appropriately earl-
ier or later time (THN 1.3.7.6 / 97).10
Hume was discontented with this bald
hypothesis and sought for a justification.
At Treatise 1.3.8.2 / 98–9 he attempted to
account for the origin of belief as a specific
instance of something that can be observed
to happen more generally: a natural tendency
to confuse readily associated objects. Hume
claimed that because the natural transition of
thought between objects that have been con-
stantly conjoined is ‘so easy’, it goes unnoticed.
Consequently, any mental ‘disposition’ that
might happen to attend the latter becomes
attached to its impostor. Where the partner
is experienced or remembered, the disposi-
tions accompanying experience or memory
are confused with the associated idea. Since
those dispositions are always bound up with
a belief in the existence of the experienced or
remembered object, we end up believing in
the existence of the associated object as well.
This account of the origin of belief is com-
plemented with an account of the nature
of belief, infelicitously inserted at Treatise
1.3.4–5 / 82–6, where it interrupts the thread
of argument for the first principle, and
108
9780826443595_Ch04_Final_txt_print.indd 108
only completed at 1.3.7 / 94–8, after hav-
ing been itself interrupted by 1.3.6 / 86–94.
The account begins with an examination of
the ‘impressions of the senses and memory’,
which are the apparent source of the belief
based on causal inference. The relevant point
about sense experience had already been
made much earlier in Treatise 1.1 / 1–25. Any
object that can be sensed can be imagined, so
that the difference between sense experience
and imagination cannot arise from what is
sensed or imagined. Since there is nonethe-
less a difference, it must be due to something
else. Hume referred to this other factor as
a different ‘manner’ in which the object is
conceived (THN 1.3.7.5 / 96), and tried to
describe further this manner of conception
by saying that sensing is more ‘forceful’ and
‘vivacious’, imagining ‘fainter’ and ‘lower’
(THN 1.1.1.1 / 1–2, 1.1.1.3 / 2–3).
Turning to memory, Hume observed that
anything that can be remembered can likewise
be merely imagined, and concluded that the
difference between what is received as a fan-
tasy and a memory must consist just in the way
it ‘feels’ to remember (THN 1.3.5.3–5 / 85–6).
Having isolated these differences between
sensing and remembering, on the one hand,
and imagining on the other, Hume inferred
that ‘the belief or assent, which always
attends the memory and senses is nothing but
the vivacity of those perceptions they present;
and that alone distinguishes them from the
imagination. To believe is in this case to feel
an immediate impression of the senses, or a
repetition of that impression in the memory’
(THN 1.3.5.7 / 86).
These reflections on the nature of sense
and memory led Hume to his conclusion. If
the belief attending causal inference arises
from a transfer from a sensed or remembered
object to an imagined object, and if the belief
in sensed or remembered objects is nothing
2/9/2012 9:39:13 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
but a more vivacious conception of those
objects, then the belief attending causal infer-
ence must likewise be just a more vivacious
conception of an object.
This is surprising. Rather than find belief in
an unperceived object to be the product of a
judgement, justified by appeal to a causal rela-
tion to an experienced or remembered object,
Hume found it to be no different from the belief
that attends sense experience and memory. It
consists just in a more vivacious conception
of the object. Perhaps because he sensed that
this conclusion would not be readily accepted,
Hume pretended to remain hesitant about it,
offering two reasons for his hesitancy:
1. Accounting for the origin of belief
involves identifying a cause. According to
Hume’s own account of causality, we can-
not discover causes merely by inspecting
their effects, nor can we be confident that
we have identified causes if we have exam-
ined only one instance. Either we must find
some analogy between the one instance we
have before us and other instances and dis-
cover some regularity in the succession of
events in the analogous cases, or we must
show how some combination of more basic,
previously established causes could account
for the effect. The appeal made at Treatise
1.3.8.2 / 98–9 to a general tendency to con-
fuse readily associated objects is a justifica-
tion of the latter sort. But Hume went on to
declare that, while he would be satisfied if
his reader found this reason compelling, he
himself placed his chief confidence in being
able to uncover a justification of the former
sort (THN 1.3.8.3 / 99; App. 3 / 624–5). He
wanted to find analogies between the for-
mation of belief as a consequence of experi-
encing or remembering an object that has
been constantly conjoined with some other
object in the past and other operations of
the mind – something that would allow us
109
9780826443595_Ch04_Final_txt_print.indd 109
to understand belief as a specific case of
something that happens more generally. His
search for these analogies had mixed results.
They will be discussed later.
2. A more pressing problem is the charac-
terization of belief. Four different characteri-
zations of belief have emerged from what has
been said about Hume’s account. At Treatise
1.3.8.2 / 98–9, Hume described belief as a
‘disposition’ of the mind. Over the course of
Treatise 1.3.7.7 / 628–9 (from THN App. 1)
and 1.3.8.2 / 98–9 this disposition is further
described as having to do with drawing and
focusing attention (‘rendering more present’,
‘weighing more in the thought’, ‘having more
force and influence’, ‘appearing of greater
importance’, ‘fixing the attention’); arous-
ing passion (‘elevating the spirits’); inspiring
deliberation (‘having a superior influence on
the imagination’); and inclining us to action.
But THN 1.3.7.4–6 / 95–7 and 1.3.7.7 /
628–9 also describe belief as a ‘manner’ in
which an object is conceived. A further char-
acterization of belief is found at THN App. 3
/ 624–5 and in passages from the Appendix
recommended by Hume for insertion in the
main body of the Treatise (1.3.5.5 / 628,
1.3.7.7 / 628–9), where belief is described as a
‘feeling different from the simple conception’
of an object. Finally, and most notoriously,
belief is described on numerous occasions
in both the Treatise and the Appendix as a
more forceful and vivacious idea, with the
terms ‘force’ and ‘vivacity’ often being sup-
plemented by a list of others (e.g. ‘solidity’,
‘firmness’, ‘steadiness’) that are not obvi-
ously synonymous, either with one another
or with ‘force’ or ‘vivacity’.
The bare fact that Hume described belief
in these different ways does not pose a prob-
lem as long as the different descriptions
can all be integrated.11 But Hume seems to
have become worried that the frequent and
2/9/2012 9:39:13 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
prominent description of belief as a more
forceful and vivacious idea had ‘not been so
well chosen, as to guard against all mistakes
in the readers’ (THN App. 1 / 623). Perhaps
this was because he found readers inclined
to take ‘force and vivacity’ to refer to some
qualitative feature of the object that is con-
ceived, like brightness or distinctness, rather
than, as he had all along wanted to insist, a
‘manner’ in which we conceive this object,
specifically, a conception with focused atten-
tion, aroused passion, and an impetus to
deliberation and action – these being ‘dispo-
sitions’ of the mind that are ‘felt’ even when
not acted upon and so not made evident
to others. In the Appendix and insertions
to the Treatise proposed in the Appendix
Hume stressed that by ‘force and vivacity’
he had meant conception of an object in this
‘manner’ – conception attended with these
dispositions.
But Hume also confessed that he found ‘a
considerable difficulty in the case; and that
even when I think I understand the subject
perfectly, I am at a loss for terms to express
my meaning’ (THN 1.3.7.7 / 628). What may
have bothered him was that appealing to ‘dis-
positions of the mind’ to explain belief does
not sit well with what he was to go on to say
about the nature of minds and mental acts (in
THN 1.4.6 / 251–63 and 1.4.5.26–7 / 244–6).
This pushed him in the direction of taking
belief to be a feeling (presumably, the feeling
of having one’s interest aroused, one’s passions
elevated, and one’s inclinations determined),
and in turn prompted worries about whether
the feeling might be separable from the con-
ception (denied over THN, App. 4–8 / 625–7,
but affirmed at EHU 5.11 / 48). But that
Hume took the presence of the dispositions to
be what is ultimately constitutive of belief, and
the feelings of being so disposed to be merely
introspective evidence for the presence of the
110
9780826443595_Ch04_Final_txt_print.indd 110
dispositions – and, importantly, evidence that
is defeasible – is suggested by THN 1.3.9.13 /
112–13 and 9.14 / 113–14.12
Whatever frustrations Hume may have
had with his efforts, it is clear that he meant
to reject the view that to believe is to perform
an act of assenting to a proposition, where
a proposition involves asserting a relation
between two or more ideas. In particular, to
believe that something exists is not to assent
to a proposition joining the idea of that thing
to the idea of existence. Hume rejected this
possibility by arguing that we have no idea
of existence distinct from whatever particular
thing we conceive to exist. To conceive some-
thing as existing is no different from conceiv-
ing it (THN 1.2.6.2–6 / 66–7, 1.3.7.2 / 94–5,
THN App. 2 / 623–4). He offered the ineligi-
bility of this account of belief as a further rea-
son to accept the alternative that to believe is
just to sense or remember or be instinctively
inclined to form a more vivacious idea.
In a footnote, Hume went so far as to
describe the division of the acts of under-
standing into conception, judgement and
reasoning, and the definitions given of these
operations as a ‘remarkable error’. These
three acts of the understanding ‘all resolve
themselves into the first, and are nothing but
particular ways of conceiving our objects’
(THN 1.3.7.5n / 96n). This is an overstate-
ment, since Hume did recognize that we
do things like comparing objects with one
another to discover relations between them,
or executing arithmetical demonstrations
in which one thing is inferred from another
by appeal to a relation between the two.
Indeed, as will be noted below, Hume went
so far as to recognize a class of ‘oblique’ or
‘explicit and indirect’ causal inferences that
are demonstrative in the classic sense. These
operations satisfy the definition of judge-
ment as the ‘separating or uniting of different
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
ideas’, and of reasoning as the ‘separating or
uniting of different ideas by the interposition
of others, which show the relation they bear
to each other’ (THN 1.3.7.5n / 96n). The
received definitions of judgement and reason-
ing apply to those operations that are con-
stitutive of knowledge in the demonstrative
sciences, particularly arithmetic, and know-
ledge of intuitive truths, such as that orange
is more like red than green.13 They even apply
to many of the judgements and arguments
found in the empirical sciences. But they do
not describe all of those operations that are
constitutive of belief in the empirical sciences
or in everyday life. In particular, they do not
describe the most fundamental of those oper-
ations. We form fundamental beliefs neither
by discerning relations between ideas nor by
inferring one idea from another by appeal to
an intermediate relation. Instead, we form
fundamental beliefs by having lively concep-
tions given to us in sensation and memory,
or by being instinctively compelled to form
lively conceptions as a consequence of asso-
ciation with what is sensed or remembered.
The latter is ‘not only a true species of rea-
soning, but the strongest of all others’ (THN
1.3.7.5n / 96n).
4. THE ARGUMENT OF THE
ABSTRACT AND THE ENQUIRY
Hume came to be dissatisfied with the ram-
bling, quasi-meditative path of discovery he
had dragged his reader down when present-
ing his ‘system’ in the Treatise. The Abstract
and the Enquiry offer a far more elegant
presentation of the same theory.14 They
replace the opening discussion of the foun-
dations of probable reasoning with the ques-
tion of how we reason concerning ‘matters
111
9780826443595_Ch04_Final_txt_print.indd 111
of fact’ or existence. Like the Treatise, they
leap to the conclusion that this can be done
only by means of causal inference. Unlike
the Treatise, they do not proceed to analyse
the causal relation in terms of a problematic
notion of necessary connection. The Abstract
analyses causality in terms of contiguity in
space, succession in time and a constant con-
tiguity and succession in like instances, mak-
ing no mention of necessary connection. The
Enquiry offers no analysis of the causal rela-
tion at all, though there are passing references
to a ‘supposed’ tie or connection between
cause and effect (EHU 4.4 / 26–7, 4.10 / 29).
In both works, necessary connection, which
had played such a large role in the Treatise,
comes up for discussion only after the two
parts of the ‘system’ have been presented.
Rather than investigate the notion of neces-
sary connection, Hume directly proceeded to
ask how causal reasoning enables us to infer
the existence of unperceived objects.
He first claimed that we cannot do this in
advance of experience, by reference to any-
thing we can find in those objects we consider
to be causes and effects. In contrast to the scat-
tered, sketchy, unconvincing, and ill-placed
arguments of Treatise 1.3.1.1, 1.3.2.5, and
1.3.6.1 his conclusion was now justified by
two different lines of argument. First, Hume
appealed to everyone’s introspection, assisted
by appeal to cases such as that of encoun-
tering an object for the first time or that of
the biblical Adam, newly created with fully
functioning, adult cognitive capacities, but no
experience. Just as Adam would be unable to
say what the effect of any given cause would
be prior to experience, even the effect of the
motion of one billiard ball towards another,
so we find ourselves unable to say what the
effect of a cause will be or what the cause
of an effect was in novel cases (Abs. 11 /
650–1; EHU 4.6–7 / 27–8). If we think we
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
do perceive causal powers in more familiar
cases, it is only because we have forgotten
what it was like to experience these things for
the first time (EHU 4.8 / 28–9). While we do
often anticipate how events will turn out in
novel situations (scientific experiments being
the prime example), the demonstrations that
we employ when doing so appeal to funda-
mental causal rules (cohesion, gravitation,
communication of motion by impulse, etc.)
that are not intuitively or demonstrably obvi-
ous, which begs the question of how we have
obtained the idea of these fundamental causal
relations (EHU 4.12 / 30–1).
Second, Hume appealed to variations on
the argument of THN 1.3.6.1 / 86–7:
• According to Abs. 11 / 650–1, effects ‘fol-
low’ from causes. Consequently, given
any cause existing at one time, we can
conceive any other object to exist at the
following time. But when something is
demonstrable, the opposite is contradic-
tory and inconceivable.
• According to EHU 4.8–11 / 28–30, every
effect is a different event from its cause.
Consequently, when conceiving the one,
it is not necessary that we also conceive
the other. If we do conceive them together,
we are conscious that nothing compels
us to do so, so that the conjunction is
effectively arbitrary. But this means that
there can be no demonstration of effects
from causes or causes from effects, again
because where there is a demonstration
the opposite is inconceivable.
The version of the argument in the Abstract
omits reference to the power of the imagin-
ation to separate different objects, but it obvi-
ously rests on that assumption and it also
makes more explicit appeal to the principle
that ‘whatever we conceive is possible, at least
in a metaphysical sense’ (Abs. 11 / 650). Like
112
9780826443595_Ch04_Final_txt_print.indd 112
the earlier argument at THN 1.3.6.1 / 86–7,
this appears to rule out the possibility of there
being any such thing as a necessary connec-
tion between the two. In contrast, the version
of the argument in the Enquiry omits any
reference to conceivability as a criterion of
metaphysical possibility.
Having established that reasoning from
causes and effects is based on past experience,
the Abstract proceeds to argue that any rea-
soning from that experience would have to
depend on the principle that the future will be
like the past. But (i) this principle is not dem-
onstrably true, since a change in the course
of nature is conceivable. And (ii) any attempt
to prove that it is most likely true would run
in a circle, since we could only appeal to the
fact that it has been true in the past to argue
that it will most likely continue to be true in
the future. Even were we to take a constant
conjunction in past experience to be evidence
of the existence of a power in causes to bring
about their effects, we only perceive the sens-
ible qualities of bodies, and we can have no
assurance that like sensible qualities will con-
tinue to be conjoined with like powers.15
The Enquiry mounts the same argument,
but addresses it to a different question – not
the question of why we suppose that the
future will be like the past, but the question
of why we suppose that like objects contain
like hidden powers (EHU 4.7 / 28). This is
not an innovation, because the same question
had been raised in the Treatise (at 1.3.6.8–10
/ 90–1) and the Abstract (at 15 / 652), though
only as an afterthought,16 and reference to a
supposition that the future will be like the
past does come up in the Enquiry over the
subsequent course of the argument (at 4.19
/ 35–6). In addition to giving the usual rea-
sons for a negative answer Hume also offered
a new argument: since peasants, infants and
animals are able to infer effects from causes,
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
either they do not do so by means of any argu-
ment or demonstration, or only by means of
the simplest and most obvious of reasons. Yet,
unless Hume was more obtuse than a peasant
or child, there are no such reasons.17
Having raised these ‘sceptical doubts
about the operations of the understanding’
the Abstract and the Enquiry proceed to offer
a ‘sceptical solution of these doubts’ – the
same, two-part solution that was presented
in the Treatise. First, our experience of what
has customarily been the case in the past
trains habits of thought into us, so that we
naturally expect the same sorts of things to
happen in the future. The expectation is not
rationally justified, but naturally induced.
Again, the Enquiry adds a new and compel-
ling argument for this conclusion: attributing
the inference to habit offers the only plaus-
ible explanation for how it is that we come
to draw a conclusion from many experi-
ments that we would not draw after seeing
just one. Secondly, we do not just infer causes
and effects from one another but believe the
absent partner to exist (at the contiguous
place and the appropriately prior or poster-
ior time).18 Because this belief does not arise
from reasoning, but from habit, it is declared
to be due to ‘a species of natural instincts,
which no reasoning or process of the thought
and understanding is able, either to produce,
or to prevent’ (EHU 5.8 / 46–7).
The Abstract and the Enquiry go on to
examine what belief is, reaching the same con-
clusions as the Treatise and, in the Enquiry,
even stating them by means of an extended
quotation from the Treatise (EHU 5.12 /
48–50, quoting THN 1.3.7.7 / 628–9 with
minor modifications). Interestingly, in the
Enquiry these further details about belief
are reserved for a distinct part of Enquiry 5,
prefaced by a remark dedicating the part to
‘such as love the abstract sciences, and can be
113
9780826443595_Ch04_Final_txt_print.indd 113
entertained with speculations, which, however
accurate, may still retain a degree of doubt
and uncertainty (EHU 5.9 / 47)’. ‘Readers of
a different taste’ are told that the part may be
neglected without impairing an understand-
ing of subsequent portions of the book.
Part 2 of Enquiry 5 (EHU 5.10–22 /
47–55) is nonetheless important. As Hume
stressed in the same breath in which he
advised ‘readers of a different taste’ to move
on, delving into the question of what belief is
and how it arises will uncover ‘explications
and analogies that will give satisfaction’. The
‘satisfaction’ Hume had in mind is not just
the satisfaction of idle intellectual curios-
ity, but the satisfaction of objections to the
account of belief laid out in the concluding
paragraphs of the Enquiry 5, Part 1. A con-
cern with uncovering ‘analogies’ between the
account of belief and other operations of the
mind is a constant of Hume’s thought about
his ‘system’ (cf. THN 1.3.8.3 / 99; App. 9 /
627; Abs. 23 / 655), for good reasons that
have already been alluded to. There is very
little positive argument to justify Hume’s
‘system’. The meditative path of discovery of
the Treatise offers only rhetorical support,19
and the Abstract’s and Enquiry’s effective
reconstruction of that path into a critique of
the view that our causal inferences are justi-
fied by appeal to facts and rules at best put
Hume in a position to claim that causal infer-
ence is not based on reasoning, not to claim
that it is based on a habit of association and
a transfer of belief from an impression or
memory. The same can be said of a new argu-
ment, presented only later, to the effect that
reasoning is too slow and uncertain in its
operations to be entrusted with an operation
as important for survival as causal inference
(EHU 5.22 / 55) and of the Enquiry’s appeal
to the abilities of peasants, children and
animals to draw causal inferences. The one
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
positive argument for the theory presented
so far is the Enquiry’s appeal to the problem
of how we draw a conclusion from repeated
experiments that we cannot draw from just
one, and that argument offers a justifica-
tion of the least satisfying sort: inference to
the best explanation. Hume hoped that by
uncovering ‘analogies’ between belief and
other operations of the mind he would be
able to offer a more compelling, Newtonian
argument by induction from the phenomena
to a general rule. The general rule would pro-
vide justification for the two-part system, as
a special case, but it would in turn be sup-
ported by induction from all the analogous
cases revealed by experience.
However, Hume had come to think that the
public had no taste for this sort of investigation,
particularly if drawn out to any great length
(Abs. Preface, 1–2 / 643). His solution was to
drastically abbreviate the argument, focus-
ing just on the exposition of analogous cases,
and to invite impatient readers to skip ahead.
In the Treatise he went on at much greater
length, not only identifying analogous cases,
but worrying about contrary evidence, refin-
ing the system to account for it, and appealing
to the system to account for a wide range of
other phenomena, thus adding a demonstra-
tion of explanatory power to the other reasons
for accepting the system. Because the Enquiry
merely repeats some of what the Treatise had
to say on this score,20 I focus on what Hume
had to say in the Treatise in what follows.
5. ANALOGIES, EXPERIMENTS,
RECALCITRANT DATA, AND
REFINEMENTS
In both the Treatise (1.3.8.3–5 / 99–100) and
the Enquiry (5.15–18 / 51–3) – the relevant
114
9780826443595_Ch04_Final_txt_print.indd 114
passages are identical in both) – Hume noted
the following ‘analogies’ between causal infer-
ence and other operations of the mind: the
picture of an absent friend ‘enlivens’ the idea
of that friend, as well as the passions that idea
occasions; the ‘mummeries’ of the Roman
Catholic religion ‘enliven’ devotion; ‘sensi-
ble types and images’, which have a greater
influence on the fancy than any other, ‘con-
vey that influence’ to the ideas they resemble;
objects that are placed in the vicinity of other
objects ‘transport’ the mind ‘with a superior
vivacity’ to ideas of those other objects (e.g.
passing the house next door on my way home
gives me an idea of my home that ‘imitates
an immediate impression’). Importantly, in
all of these cases the trigger (the picture, the
ceremony, the icon, the neighbouring objects)
must be both experienced and ‘naturally
related’ to the target (in the cases mentioned,
by relations of resemblance or contiguity); if
the trigger is merely imagined, the target is
not enlivened; if the trigger is unrelated, the
idea of the target does not even arise.
The case is the same with causal infer-
ence, as Hume proceeded to prove by appeal
to three experiments (THN 1.3.8.8–11 /
101–3): suppose the natural relation (in this
case of constant conjunction) is absent (as,
for example, when experiencing a cause for
the first time). Then the associated idea does
not arise. Now suppose a constant conjunc-
tion between the trigger and the target has
been experienced in the past. Then, solely for
that reason and without the assistance of any
intermediate process of argument or justifi-
cation or appeal to general rules, a vivacious
idea of the target arises upon experience of the
trigger. Now suppose the trigger is not experi-
enced but only imagined. Then the associated
idea does arise, but it has no vivacity.
By induction from all these phenomena,
Hume declared it to be ‘a general maxim
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
in the science of human nature, that when
any impression becomes present to us, it not
only transports the mind to such ideas as are
[naturally]21 related to it, but likewise com-
municates to them a share of its force and
vivacity’ (THN 1.3.8.2 / 98).
This maxim was no sooner justified than
Hume acknowledged a difficulty. He had
defined belief to be nothing but a more ‘viv-
acious’ idea. But he had also maintained that
belief only arises from causal inference, not
from the other natural relations of resemblance
and contiguity, notwithstanding that, according
to the maxim, they all enliven ideas. The three
claims are inconsistent (THN 1.3.9.2 / 107).
At this point Hume stood on the brink of
momentous discoveries, presented with an
opportunity to reassess his earlier, ill-consid-
ered position that belief in matters of prob-
ability can only arise from causal inference.
He had discovered that basic causal infer-
ences are inferences from the constancy of
temporal succession in resembling cases. But
we also draw inferences from the constancy
of spatial arrangement in resembling cases.
Quite apart from forming any background
beliefs about the causes of the immobility of
landmarks, we rely on the constancy of the
position of houses, trees, the pole star and
other geographical or astronomical objects to
navigate, and when we do so we reason from
experienced objects to their unperceived sur-
roundings, not to their unperceived causes or
effects.22 Hume himself recognized this with-
out realizing it when he wrote:
Suppose I see the legs and thighs of a
person in motion, while some interpos’d
object conceals the rest of his body. Here
’tis certain, the imagination spreads out
the whole figure. I give him a head and
shoulders, and breast and neck. These
members I conceive and believe him to
be possessed of. (THN App. 4 / 626)
115
9780826443595_Ch04_Final_txt_print.indd 115
Since the head, shoulders, breast and neck
are neither causes nor effects of the legs and
thighs, the inference here is not causal, even
though Hume recognized that it produces
belief. The same holds of Hume’s description
of ‘our approach to any object; tho’ it does not
discover itself to our senses’ as leading that
object to operate on the mind ‘with an influ-
ence that imitates an immediate impression’
(THN 1.3.8.5 / 100). An ‘influence that imi-
tates an immediate impression’ is just a belief.
The case of resemblance poses more of a
problem. Seeing the son of a long-dead friend
(EHU 5.19 / 53) does not produce a belief
in the existence of an unperceived object. At
best, it rouses old memories and enlivens the
associated passions.23 One reason for this is
that the resemblance relation calls up an idea
of a resembling object without giving any fur-
ther indication of where that object is placed
in space and time, leaving us with no inclina-
tion to ascribe it a location in the real world.
Our causal and geographical inferences, on
the other hand, involve not just association
of objects, but association with places and
times where the object is to be found. Not
surprisingly, therefore, when resemblance is
bound up with relations of time and place, it
has the same influence as constant succession
in time and constant conjunction in space.
This is most notably the case with our beliefs
about the identity of objects over time, where
we suppose a continuum of intermediate
states to exist unperceived between observed,
resembling, earlier and later states.
This is just a sketch of how Hume might
have gone on to investigate the possibility that
probabilistic reasoning involving all three of
the ‘inconstant’ relations of causality, contigu-
ity and identity could be grounded in the three
‘natural’ relations of constant succession in
time, constant conjunction in space and closest
resemblance at contiguous places over time.
2/9/2012 9:39:14 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
Unfortunately, Hume did not take this
path. (THN 1.4.2.15–23 / 194–9 is perhaps
the most lamentable consequence of that deci-
sion.) He did go so far as to declare that we
take our sense experiences and memories to
constitute a ‘system’ of ‘realities’, and that we
join a second ‘system’ to it, consisting of the
unperceived causes or effects of these ‘realities’
(THN 1.3.9.3–4 / 108). But he never paused
to consider that what makes the objects of the
senses and memory a ‘system’ is that each is
related to all the others in virtue of its unique
location within a single space and time, and
that what enables the causal relation to aug-
ment the system is that it directs us where
to localize unperceived objects in this space
and time – something that constant contigu-
ity and resemblance insofar as it is bound up
with identity relations could also do.
Instead, Hume maintained that because
any given object resembles and is spatially
contiguous to a huge variety of other objects,
the mind senses a certain ‘caprice’ or feeling
of liberty in making the association with just
one. This feeling of liberty prevents the easy
and unnoticed transition from one object to
another that Hume had earlier identified as
essential for the transfer of the mental disposi-
tions characteristic of belief. It also introduces
new feelings of ‘looseness’ and ‘weakness’ that
are contrary to the feelings of stability and
strength characteristic of belief. Moreover,
any tendency we might have to include objects
thought of under these conditions in the sys-
tem of ‘realities’ would produce repeated
experiences of having our expectations disap-
pointed. As a consequence, we would learn to
associate objects thought of under these condi-
tions with fictions (THN 1.3.9.6 / 109–10).24
Causal relations are very different. Any given
object is related to just one other object as its
cause and just one other object as its effect, so
there is no ‘looseness’ to the association.25
116
9780826443595_Ch04_Final_txt_print.indd 116
These reflections mandate a revision to
Hume’s maxim, although he never said so.
While impressions may transport the mind to
any ideas that are naturally related to them,
they only communicate a share of their force
and vivacity sufficient to induce belief to
ideas of those objects that have been custom-
arily conjoined with them in the past.
This revised maxim explains an attempt at
‘confirmation’ of the ‘hypothesis’ that Hume
made at THN 1.3.9.16–19 / 115–17. The
attempt appeals to an example that would
otherwise serve more to falsify than to con-
firm the hypothesis. If the repetition of a
conjunction (making it ‘customary’) plays
a more important role in producing belief
than the force and vivacity of the impression,
and if belief is a mental ‘disposition’ involv-
ing things that can be produced merely by
repetition, such as fixed attention, familiarity
and stability of the object, then we should
expect that belief could arise from the mere
repetition of an idea even in the absence of
association with an impression or memory.
Hume considered this in fact to be the case,
most notably with the beliefs produced by
education, which he considered to provide
outstanding confirmation for the hypothesis
because, as he claimed, education is respon-
sible for more than half our opinions and is
more influential than either abstract reason-
ing or experience (THN 1.3.9.19 / 117). He
also instanced the tendency of amputees and
the bereaved to be unable to accept their loss,
and of people to consider themselves to be
on intimate terms with personages they have
only read about. It is hard not to wonder
about the aptness of these examples or the
soundness of Hume’s implicit view of how
education produces belief, but people do
have a tendency to believe what they hear
from those around them simply because
everyone is saying it and even though no one
2/9/2012 9:39:15 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
is in a position to testify to the truth of what
they are saying (the belief in an afterlife being
an outstanding example).
Besides this appeal to a confirming experi-
ment, Hume justified the hypothesis by appeal
to its explanatory power (THN 1.3.9.9–15 /
110–15). The hypothesis is able to explain
such things as (i) why pilgrimages strengthen
belief; (ii) why it is wrongly supposed that
the communication of motion by collision
could be anticipated in advance; (iii) why we
have a much more vivid conception of the
vastness of the ocean from vision than from
hearing; (iv) why we are credulous, even in
the face of contrary experience; (v) why we
cannot take the infinite rewards and pun-
ishments of an afterlife seriously, even if we
believe in them; and (vi) why we enjoy reli-
gious discourses and dramatic performances
that excite the disagreeable passions of fear
and terror. In all of these cases the explana-
tion is the same. Though the natural relations
of contiguity and resemblance are not able to
produce belief on their own, when a belief
has once been formed, so that there is no
sense of ‘caprice’ in its conception, it will be
further enhanced by relations of contiguity
(i) and resemblance (ii–iv) holding between
the impression and the idea, but also weak-
ened by the opposite relation of dissimilar-
ity (v), with the weakening of belief in turn
accounting for (vi).
6. THE PSYCHOLOGICAL
FOUNDATIONS OF EPISTEMOLOGICAL
NORMATIVITY
Taken together, the two parts of Hume’s
‘system’ would appear to rule out any role
for logic in probabilistic inference – any role
for the conscious, deliberate application of
117
9780826443595_Ch04_Final_txt_print.indd 117
demonstrably truth-preserving or probabili-
ty-preserving rules to draw inferences from
the observed to the unobserved. We do not
discern a relation between causes and effects
by comparing them with one another and
then appeal to this relation to draw infer-
ences to the unobserved. Instead, we are
instinctively impelled to form ideas of objects
of a sort that have in the past been frequently
observed to be constantly conjoined with
currently sensed or remembered objects,
doing so in the absence of any memory of
those past occasions or conscious inference
from them (THN 1.3.8.13 / 103–4). And we
do not judge that these objects must exist
but are instinctively impelled to form a more
‘lively’ conception of them – a conception
that ‘gives them more . . . influence; makes
them appear of greater importance; infixes
them in the mind; and renders them the gov-
erning principles of all our actions’ (THN
1.3.7.7 / 629).
These are results that Hume trumpeted
in both the Treatise (1.3.8.12 / 103) and the
Enquiry (5.8 / 46–7), writing that because
objects have no discoverable connection with
one another, we can only draw an inference
from the one to the other with the aid of ‘cus-
tom operating on the imagination’, and that
‘all probable reasoning is nothing but a spe-
cies of sensation’ (THN 1.3.8.12 / 103), so
that when we prefer one argument to another
we do nothing but decide from our feelings
concerning the superiority of their influence,
meaning that belief is ‘a species of natural
instincts, which no reasoning or process of
the thought and understanding is able, either
to produce, or to prevent’ (EHU 5.8 / 46–7).
In the Enquiry, Hume pretended that even
though our beliefs are not drawn from obser-
vation in accord with truth-preserving or
probability-preserving rules, there is a ‘pre-
established harmony between the course of
2/9/2012 9:39:15 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
nature and the succession of our ideas’ ensur-
ing that beliefs will be produced in us in tan-
dem with the way causes and effects succeed
upon one another in nature, and providing
‘those who delight in the discovery and con-
templation of final causes’ with ‘ample sub-
ject to employ their wonder and admiration’
(EHU 5.21 / 54–5). This is a singular instance
of misdirection, inconsistent with the cand-
our that is otherwise characteristic of his
work. It is not just that, on his account, there
could at best be a pre-established harmony
between the past course of nature so far as it
has been observed by us and the succession
of our ideas. Hume’s account entails that
there should not even be that much.
According to Hume’s theory, belief is a
more vivacious idea resulting from asso-
ciation with an impression or memory. It
therefore depends on the original vivacity of
the impression or memory and the strength
of the associative link between the impres-
sion or memory and the idea. If either of
these is weakened, the belief will be as well.
But, as Hume observed in the Treatise, an
impression is more vivacious than a memory,
and a recent memory more vivacious than an
older one. A recent observation of a conjunc-
tion between types of events also produces
a stronger disposition to associate those
events than an earlier one. As the Hume of
the Treatise went on to admit, these factors
entail that the course of our ideas should
not be in harmony with the past course of
nature. Instead, it should be disproportion-
ately influenced by the most recent observa-
tions (THN 1.3.13.1–2 / 143–4).
This is not all. Hume further observed
that the strength of association is also
affected by the ease with which it is made,
so that a causal inference that needs to be
drawn by appeal to a number of intermedi-
ate causes should be less strongly believed
118
9780826443595_Ch04_Final_txt_print.indd 118
than one that is more immediate (THN
1.3.13.3 / 144). And just as there are factors
that lead us to overlook or discount connec-
tions found in the past course of nature, so
there are factors that lead us to suppose the
existence of connections that were not there.
According to the theory, we are disposed
to consider similar objects to have similar
causes or effects. But any given object is a
compound of many different characteris-
tics, only one of which may be constantly
conjoined with a cause or effect. Even if
we have learned to distinguish the essential
characteristic from the superfluous ones,
when we encounter an object that resembles
a cause or effect only in superfluous ways,
the resemblance should lead us to conceive
the associated effect or cause, and the ease
of the association together with the vivac-
ity of the encountered object should induce
a kind of bigoted or prejudicial belief,
which holds sway despite our recognition
of the superfluity of the resemblance (THN
1.3.13.7 / 146–7 and 1.3.13.9 / 147–8).
Nor are these the only such cases. Hume
noted that education ‘not only approaches
in its influence, but even on many occa-
sions prevails over that [belief] which arises
from the constant and inseparable union of
causes and effects’ (THN 1.3.9.17 / 116). He
further noted that we do not regulate our-
selves entirely by experience of the governing
principles of human nature when deciding
whether to believe testimony, but instead
‘have a remarkable propensity to believe
whatever is reported, even concerning appari-
tions, enchantments, and prodigies, however
contrary to daily experience and observa-
tion’ (THN 1.3.9.12 / 113). This is in part
due to the influence of the resemblance rela-
tion between ideas (supposed to exist in the
minds of others on the basis of their words)
and facts, which strengthens the associative
2/9/2012 9:39:15 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
relation beyond what is warranted by expe-
rience of their constant conjunction. But it
is also due to the fact that ideas that arouse
passions are reciprocally enlivened by those
passions (THN 1.3.10.4 / 120).
This looseness of fit between the course of
our ideas and the past course of nature is not
necessarily a bad feature of Hume’s ‘system’.
As a matter of fact, people’s beliefs are more
strongly influenced by recent experiences;
people are less inclined to lend credence to
the conclusions of complex arguments; and
people are disposed to bigotry, blind adher-
ence to received opinions, and credulity. The
fact that the course of our ideas fails to track
the past course of nature in just these ways is
further confirmation that Hume’s theory has
correctly captured the psychological mecha-
nisms responsible for human belief formation
and is a further instance of its explanatory
power in accounting for those inferences we
are in fact psychologically compelled to draw.
But this is still not an entirely happy result.
While many of us form blinkered, prejudicial,
obstinate and credulous beliefs, we do not all
do so. At least, we do not all do so all of the
time. At the very least, we do not all think
we should do so, even if, as a matter of fact,
we find ourselves irresistibly compelled to do
so anyway. As Hume himself remarked, infer-
ences skewed by recent experience are not
‘receiv’d by philosophy as solid and legiti-
mate’ because ‘philosophers’ do not think
that the same event or the same conjunction
between events should provide less evidence a
month from now than it provides today (THN
1.3.13.1 / 143). Furthermore, education is
not ‘recogniz’d by philosophers’ because it
‘is an artificial and not a natural cause’, and
because ‘its maxims are frequently contrary
to reason, and even to themselves in differ-
ent times and places’ (THN 1.3.9.19 / 117).
And credulity is, as Hume himself admits,
119
9780826443595_Ch04_Final_txt_print.indd 119
a universal and conspicuous ‘weakness of
human nature’ (THN 1.3.9.12 / 112).
This raises two problems. If all beliefs are
ultimately unjustifiable and all are founded
on the same operation of a transmission of
vivacity across associative links, how could
any of us have come to think that some of
them are better than others? And how could
some of us (e.g. a gambler who places bets
in accord with a calculation of the probabil-
ity of outcomes, ignoring the results of recent
games) not only think that certain beliefs are
better than others but manage to form their
own beliefs accordingly? Hume had solu-
tions for both problems.
7. EPISTEMOLOGICAL NORMS
In all cases we transfer our experience to
instances, of which we have no experi-
ence, either expressly or tacitly, either dir-
ectly or indirectly. (THN 1.3.8.14 / 105)
According to the account that has so far
been presented of Hume’s ‘system’, causal
inference is an unconscious (‘tacit’), instinct-
ive (‘direct’) operation resulting from, not car-
ried out in cognizance of, past experience. On
the first few occasions of observing objects of
one sort to be followed by objects of another
sort, we are unimpressed, and not disposed
to draw any inference when encountering
objects of either sort in the future. But as
we make more and more observations of the
conjunction of the two objects, we develop
a habit to think of the one when presented
with the other. As the number of observa-
tions increases, the habit strengthens, and as
the habit strengthens, more and more of the
mental dispositions characteristic of belief
and attendant upon experience and memory
2/9/2012 9:39:15 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
come to attend the associated idea. Belief,
therefore, is something that comes in degrees,
varying from conjecture to certainty in pro-
portion to the strength of the habit and hence
to the number of past observations up to the
point where a sufficient number of observa-
tions have been made to produce a habit that
mimics experience in its effects (THN 1.3.12.
2 / 130–1). Importantly, we do not recall the
past observations or appeal to them to justify
our belief. The past observations have made
us develop a habit and that habit alone pro-
duces the belief (THN 1.3.8.13 / 103–4).
But there are twists to this simple story.
One twist arises from the fact that past expe-
rience is not always uniform.26 Sometimes,
objects of one sort are not always followed
by objects of another sort. When that hap-
pens, the contrary experiences weaken the
habit. Over time, we end up with a habit that
would be strong or weak in proportion to the
number of confirming instances in the total
number of trials, but for the influence of the
‘unphilosophical’ factors mentioned above
(THN 1.3.12.6 / 132–3).
The belief we get from inconstant experi-
ence is still ‘tacit’ and ‘direct’. But in Treatise
1.3.12.2 / 130–1, Hume declared that there
are other kinds of causal inference. The kind
based on the gradual development of a habit
over the course of a uniform past experi-
ence is not, he claimed, to be found in any-
one ‘who is arriv’d at the age of maturity’
(THN 1.3.12.3 / 131), and the kind based
on inconstant experience is one that ‘we have
but few instances of in our probable reason-
ings’ (THN 1.3.12.7 / 133). Mature adults
are supposed to draw causal inferences after
just one experience, supposing it has been
obtained in circumstances where ‘all foreign
and superfluous circumstances’ have been
removed. And Hume maintained that when
the conjunction between causes and effects
120
9780826443595_Ch04_Final_txt_print.indd 120
is not entirely uniform, we seldom rely on a
gut reaction but instead deliberately recall
the past experiments, count up the number
of confirming and contrary instances, and
form beliefs that are stronger or weaker
in accord with a mathematical calculation
of probabilities.27 These are extraordinary
claims that at first sight seem incompatible
with the ‘system’. A habit cannot be formed
after just one experience, and a mathematical
result is based on the perception of a relation
between ideas that have no vivacity, and so
should not produce belief.28
Hume’s account of how we manage to do
these things lays the foundations for episte-
mological norms and an account of action in
accord with those norms.
8. GENERAL RULES
Hume claimed that while we do not as a
matter of fact recall any past experiences
when drawing inferences concerning con-
junctions of causes and effects that have
been constantly observed since infancy (e.g.
stones fall, fire burns, water suffocates),
we do ‘assist the custom and transition of
ideas’ by recalling past experiences when
encountering more rare or unusual objects
(THN 1.3.8.14 / 104). There is nothing
about the system that would suggest we are
prevented from doing so. On the contrary,
similar objects can jog the memory as well
as the imagination, particularly in unusual
cases. And we can be motivated to recall
past instances by passions such as curios-
ity, love of fame, fear and hope. Our causal
inferences could, therefore, be sometimes
‘express’ rather than ‘tacit’.
Just as nothing prevents us from expressly
recalling and reviewing past experiences, so
2/9/2012 9:39:15 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
nothing prevents us from drawing conclu-
sions from those past experiences ‘indirectly’,
by explicit appeal to causal rules, learned
from past experience. The chief such rule
is the general one that like objects, placed
in like circumstances, will have like effects
(THN 1.3.15.6 / 173). This rule is not jus-
tified by past experience – no causal rule is.
But past experience does lead us to form it
and believe it. It is ‘merely habitual’ as Hume
put it (THN 1.3.8.14 / 104–5). Once formed,
it can be expressly appealed to in order to
justify causal inference from as little as a sin-
gle past experience (THN 1.3.8.14 / 105).
This accounts for why people who have
reached the age of maturity will draw causal
inferences after just one experience rather
than needing to be trained by a number of
experiences.
‘Express and indirect’ causal inferences
can be based on other rules besides the
rule that similar objects placed in similar
circumstances will have similar effects. As
has already been noted, any striking resem-
blance between an unfamiliar object and a
familiar one will lead us to suppose that the
unfamiliar object has the same causes or
effects as the familiar one, even if there are
some dissimilarities between the objects and
the attendant circumstances are not quite the
same. This is the foundation of prejudice. As
we grow older, we encounter cases in which
our prejudices are disconfirmed and reflec-
tion on these cases leads us to appreciate
the importance of distinguishing between
superficial characteristics, which are often
but not always present in causal conjunc-
tions, and essential ones, which are always
present (THN 1.3.13.11–12 / 149–50). We
begin to think that it is not good enough to
suppose that any object that bears any strik-
ing resemblance to objects we have experi-
enced before will have the same causes or
121
9780826443595_Ch04_Final_txt_print.indd 121
effects. Instead, we need to be sure that we
have eliminated ‘all foreign and superflu-
ous circumstances’ as Hume put it (THN
1.3.8.14 / 104). We first consider it to be
at least possible that where two resembling
objects have different effects, all the simi-
larities between them must be foreign and
superfluous and the different effects must
arise from some hidden respect in which the
causes differ. As it turns out, our experience
of regularly finding these hidden features
upon a more exact scrutiny habituates us
to that belief (THN 1.3.12.5 / 132). A simi-
lar course of experience habituates us to the
belief that where strikingly different objects
have the same effect, all the differences
between them are foreign and superfluous
and the effect must arise from some circum-
stance common to the two. We are further
habituated to believe that any increase or
diminution in the effect must be due to a
compound cause, and that any cause that
persists over time before being followed
by its effect cannot be the sole cause of
that effect (THN 1.3.13.11–12 / 149–50,
1.3.15.7–10 / 174–5). These ‘rules by which
to judge of causes and effects’ justify other
causal inferences. Taken together, they con-
stitute a logic of causal inference (THN
1.3.15.11 / 175) – a system of rules that can
be expressly appealed to in drawing indirect
or ‘oblique’ causal inferences.
This is one part of the answer to the ques-
tion of how Hume could provide for a logic
of causal inference.29 Once we have come to
form and accept the rules, inferences that are
in accord with them will be approved of as
wise or justified, whereas those that violate
them will be condemned as foolish. This will
be the case even if the person making the
normative assessment is personally unable
to follow the rules due to the influence of
‘unphilosophical’ factors.
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
9. PROBABILITIES
Hume noted that while past experience may
have habituated us to the thought that same
objects have same effects, we do not always
find this to be the case. Contrary experience
does not, however, lead us to diminish our
confidence in the general rule because we
have been further habituated to believe that
where same objects have different effects
they will be found upon more exact scrutiny
to differ in some previously unnoticed way.
We may not always have the opportunity,
the curiosity, the leisure, or the resources to
search for this hidden circumstance. But this
just means that, in the absence of a percep-
tion of a previously hidden cause, our habits
will lead us to believe that it must exist. The
case here is similar to what Hume ought to
have said concerning the spatial contiguity
of causes and effects. If we fail to observe a
succession of cause and effect because clos-
ing our eyes or turning our heads leads us to
fail to observe the contiguous regions, then
we do not think we have observed a fail-
ure of the expected succession to occur. On
the contrary, this is exactly the sort of case
in which the habit kicks in and leads us to
form a belief in the unperceived existence of
the cause or effect at the contiguous location.
The only difference between the two cases
is that this time Hume did not fumble, as he
did at Treatise 1.4.2.21 / 197–8, but declared
that, ‘from the observation of several parallel
instances, philosophers form a maxim, that
the connexion betwixt all causes and effects
is equally necessary, and that its seeming
uncertainty in some instances proceeds from
the secret opposition of contrary causes’
(THN 1.3.12.5 / 132).
But, Hume went on to observe, even though
we may believe there is a hidden cause, in the
absence of opportunities, curiosity, leisure
122
9780826443595_Ch04_Final_txt_print.indd 122
or resources to search for it, we have no
recourse but to consider those circumstances
that tend to accompany it to be signs of its
likely existence, and so to take apparently
similar objects to have similar effects, albeit
with a diminished degree of certainty pro-
portioned to our experience of the degree of
regularity in the connection between occur-
rences of the superfluous circumstances and
the effect (THN 1.3.12.6 / 132–3). These
inferences are ‘probable’ in the strict sense
of being something less than ‘entirely free
of doubt and uncertainty’ (THN 1.3.11.2
/ 124). Importantly, they are not ‘tacit and
direct’. They do not result from a habit that
has been strengthened or weakened by past
experience – but also by a host of ‘unphilo-
sophical’ factors. They are instead ‘oblique’
or explicit and indirect. They result from a
survey of past instances and a calculation of
the proportion of confirming experiments in
the total number of trials.
It has already been noted why we might
be impelled to remember past experiences
and want to survey them. Hume’s remain-
ing challenge in accounting for our ‘explicit
and indirect’ reasoning concerning matters
of strict probability was not to come up
with a mathematical theory of the calcula-
tion of probability. It was to explain how we
come to proportion belief in accord with any
merely mathematical calculation.
The basic facts that need to be accounted
for are obvious: in cases where there has
never been an exception to a succession, we
should form a belief that is ‘entirely free of
doubt and uncertainty’; in cases where it is
no more likely that an event will occur than
not, we should have no belief, and in the
intermediate cases we should have a pro-
portionally strong or weak belief. So, where
there is what we might call a ‘fifty-fifty
chance’, we should have no belief; where
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
there is a ‘one hundred per cent chance’ we
should have certain belief; and where there
is a ‘seventy-five percent chance’ we should
have a belief that is half way between indif-
ference and certainty. In the last case, import-
antly, we do not form a certain belief in the
proposition that the event has a probability
of fifty (or seventy-five) percent;30 we form a
less vivacious conception of the event – one
that is half-way between a certain belief and
an indifferently entertained idea as measured
by the feeling that attends it and the strength
of the characteristic mental dispositions. We
should get a belief of this strength regardless
of what our views are on how to calculate
probabilities mathematically, or whether we
employ any mathematics at all beyond a bare
survey of instances.
In explaining how the belief arises Hume
distinguished between two cases, that of
belief in ‘chances’, and that of belief in
‘inconstant causes’. Chance arises when a
cause has an effect that can indifferently take
one or another of a finite number of alterna-
tive forms, e.g. the toss of a die causes it to
fall with one or other of its six sides facing
up. Causes are ‘inconstant’ when either the
indifference condition or the finitude condi-
tion is not met.
In the first case we have been strongly
habituated to associate the generic effect with
the cause. So, for example, upon witnessing a
die being tossed we form a strong belief that
it will fall with one side facing up. But now
suppose we ask ourselves exactly which side
will face up. Because the generic effect has a
number of equally possible, mutually incom-
patible forms the strong belief in the generic
effect becomes divided, with an equal por-
tion going to each alternative. But, because
the alternatives are all mutually incompat-
ible, the divided beliefs cancel one another
out. We are left with a very strong belief that
123
9780826443595_Ch04_Final_txt_print.indd 123
the generic effect will occur, but indifferent
over which form it will take. For example,
the belief that the die will fall with ‘side one’
facing up would be one-sixth the strength of
the belief that it will fall with some side or
other facing up – but for the fact that we can-
not have any degree of belief that ‘side one’
will fall face up without disbelieving that
any of the other sides will fall face up. Since,
however, all those possibilities are considered
to be equally likely, they cancel one another
out, leaving us with certainty that some side
or other will face up, but indifference about
which one.
Hume’s account of how we form beliefs
about the effects of inconstant causes grows
out of a refinement of the account of belief
in chances. If some of the chance alternatives
resemble one another, e.g. four of the six
sides have the same figure on them, a sur-
vey of the alternatives causes their portions
of the original belief to combine. In that case
the alternatives do not perfectly cancel one
another out, e.g. the possibility that the die
will land with the more familiar figure facing
up receives four of the six portions, only two
of which suffice to cancel the rival possibili-
ties, leaving us with a residual belief that the
more familiar figure will face up. This resid-
ual belief is two-sixths of the way between
indifference and conviction.31
Belief in the outcome of inconstant causes
is like this, except that in this case we do not
start off with a strong belief in the occurrence
of a generic effect. Hume instead supposed
that we develop the habit of expecting the
future to be like the past. The strong belief
in this uniformity principle plays the same
role as the strong belief in the occurrence of
a generic effect. If in the past one egg in every
crate has been rotten, we will transfer that
past experience to the future in the sense that
any subsequent impression or memory of a
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
crate of eggs will produce a strong belief in
the existence of eleven sound eggs and one
rotten one. On picking out any one egg from
the crate and forming an idea of what we
will smell upon cracking it open, that strong
belief is divided into twelve packets, eleven
of which resemble one another in being
attached to ideas of sound eggs, and one
of which is attached to the idea of a rotten
egg. Since the possibilities are incompatible,
we cannot believe them both to result from
cracking the egg. Since one of them comes
up so much more often in a mental survey
of alternatives, we cannot be perfectly indif-
ferent, either. The odd possibility is cancelled
out by one of the opposing packets leaving
us with a belief that this egg will prove to
be sound that is ten-twelfths as strong as the
original belief that there will be eleven sound
eggs in the crate.
There are oddities about Hume’s presen-
tation of this account in both the Treatise
and the Enquiry. The account offered by the
Treatise is repeated four times over (first at
THN 1.3.12.8–12 / 133–5, a second time at
1.3.12.13–18 / 135–7, again at 1.3.12.19
/ 137–8, and a fourth time at 1.3.12.20–2
/ 138–40). Whether this is because Hume
was unsure of himself or particularly proud
of his result is unclear. In Enquiry 6 / 56–9
he retreated from the attempt to provide
a calculus of the strength of belief, at first
attributing probabilistic belief to ‘an inex-
plicable contrivance of nature’ but then say-
ing that his account of belief as a firmer and
stronger conception of an object allows us to
explain matters a bit further by saying that
where ‘a great number of views . . . concur
in one event’ they ‘fortify’ the conception of
that event in the imagination and so pro-
duce belief. This reticence notwithstanding,
the full theory of belief as proportioned in
accord with the subtraction of less frequent
124
9780826443595_Ch04_Final_txt_print.indd 124
possibilities from more frequent ones reap-
pears at EHU 10.4 / 110–11.
Whatever ambivalence Hume might or
might not have had about his account, he
suggested in the Treatise (1.3.11.1 / 7–8) and
stated outright in the Abstract (4 / 646–7) that
he had offered the only adequate explanation
of probable belief that had ever been given.
There could be no demonstrative account
of probable belief because it is in principle
impossible to demonstrate that the event that
is most often observed will occur (in that case,
it would not be merely probable). There can
be no probable account of probable belief
because those who claim that a survey of
past results can at least make us certain about
which event is most likely to occur are actually
doing no more than uttering the trivial claim
that the event that has come up most often
in a survey of past results is the event that
has come up most often in a past survey of
results, not giving us a reason to believe, with
any degree of conviction, that this event will
occur on any future occasion.32 This is further
proof, in Hume’s eyes, that belief must be a
more vivacious manner of conception rather
than the product of a judgement.
10. ‘PHILOSOPHICAL’ BELIEF
By all that has been said the reader will
easily perceive, that the philosophy con-
tained in this book is very sceptical, and
tends to give us a notion of the imperfec-
tions and narrow limits of human under-
standing. (Abs. 27 / 657)
Hume had no sooner presented his account
of how we arrive at weaker or stronger
beliefs about matters of probability than he
remarked that recent experiments weigh more
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
than earlier ones in our assessments of the
probability of outcomes, whether we realize
it or not. The freshness of an experiment ‘has
a considerable influence on the understand-
ing, and secretly changes the authority of the
same argument, according to the different
times, in which it is propos’d to us’ (THN
1.3.13.1 / 143). He also said that this is an
effect that ‘has not had the good fortune’ to
be ‘receiv’d by philosophers, and allow’d to
be [a] reasonable foundation of belief and
opinion’ (THN 1.3.13.1 / 143). As already
noted, this is just one of many ‘unphilosoph-
ical’ influences on belief, influences that we
are as a matter of fact susceptible to, for rea-
sons that Hume’s ‘system’ succeeds admir-
ably well at explaining.
We might conjecture that ‘philosophers’
are led to condemn these influences as a con-
sequence of having been habituated to accept
certain rules, such as the rule that explicit
and indirect probable inferences are more
often correct than tacit and indirect ones.
But not everyone is habituated to accept the
same rules. Hume noted that ‘the vulgar’ are
not habituated to accept that same causes
must have same effects or that when a cause
fails of its usual effect a more exact scrutiny
will uncover some previously hidden circum-
stance that is the true cause. Instead, their
experiences have habituated them to accept
that causes are not perfectly regular in their
operations, even though nothing impedes
them (THN 1.3.12.5 / 132). Someone with
that outlook will be less inclined to recog-
nize a distinction between superfluous and
essential components of causes (hence more
inclined to prejudice), and less inclined to
accept that the course of nature cannot
change (hence more inclined to form their
beliefs on the basis of recent evidence). And
a philosopher is in no position to convince
them of the error of their ways since the
125
9780826443595_Ch04_Final_txt_print.indd 125
philosopher’s commitments have no other
foundation than (at best, and not always)
the past course of their own experience – a
foundation that the vulgar can appeal to as
well. The difference between the vulgar and
the philosophers is that they have had differ-
ent experiences. Each judges in accord with
what their own experiences have made them
and neither is in a position to appeal to their
experiences as a higher authority.
Nor is this the only impediment to ‘philo-
sophical’ belief. Just because ‘philosophers’
approve of the ‘rules by which to judge of
causes and effects’, it does not follow that they
will always form their beliefs in accord with
those rules. Very few of us have been habitu-
ated to think (1) that whenever a supposed
cause fails of its usual effect a hidden cause
will be discovered upon more exact scrutiny,
(2) that the bare passage of time does not des-
troy the authority of an experiment, or (3) that
beliefs that are based on recent, lurid anec-
dotes will more often prove to be false than
those that are based on an impartial survey of
cases. The reason most of us accept these and
other such rules is not personal experience in
the laboratory, but uncritical acceptance of
the testimony of others, education or the influ-
ence of a few, recent, notable errors. That is
perhaps not a bad thing. Ironically, if anything
enables ‘philosophers’ to draw their infer-
ences just in accord with the rules by which
to judge of causes and effects, it is not having
been habituated to accept them but having
been educated to accept them.33 Education is
among the most illegitimate but also the most
powerful of the factors influencing our belief.
Personal past experience conveys a degree of
vivacity to associated ideas that is vulnerable
to being artificially enhanced or diminished
by resemblance, contiguity, passion or the
passage of time, whereas education is largely
impervious to those influences.
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
Hume’s position is not entirely sceptical.
Although he never said so in quite so many
words (but see THN 1.4.7.13 / 271–2), he
would probably have agreed that a survey
would show that ‘philosophical’ beliefs have
more often turned out to be correct than
‘vulgar’ ones. The rules followed by ‘phi-
losophers’ in arriving at their beliefs, there-
fore, have a title to be considered the logic
of probable reasoning rather than just an
anthropological description of the epistemo-
logical aspirations of a certain social class.34
But the rules have this title only for those
already habituated to accept that the future
will be like the past, and to accept that expli-
cit and indirect probable inferences, based
on an impartial survey of past cases, are to
be preferred to tacit and indirect ones. Those
not habituated (or educated) to accept these
rules could justly reject any appeal to a sur-
vey of past instances as question-begging.
And even those who do accept the presup-
positions of the argument will not always be
able to resist the other factors inducing them
to form beliefs. ‘Unphilosophical’ belief will
persist, even among those who know better.
This explains the characteristic pessimism
displayed in Hume’s Essays, Natural History
of Religion, and History of England.35 Under
certain social conditions the arts and sciences
will flourish and philosophical learning will
triumph over vulgar superstition. But once
entrenched in a society, there is no guaran-
tee that the arts and sciences will progress.
A change in circumstances, beyond anyone’s
ability to control or even predict, altering the
course of people’s experiences, can change
their inferential practices and the intellectual
culture that had developed can be supplanted
by the crudest barbarism.
Hume did think that there is one means
by which the influence of ‘unphilosoph-
ical’ factors on belief might be mitigated:
126
9780826443595_Ch04_Final_txt_print.indd 126
an experience of the force of sceptical argu-
ments.36 He thought that someone who has
once been convinced of the weakness and
fallibility of our powers of knowledge will
be permanently changed by that experience.
Forever afterwards, they will be doubtful
about all their beliefs and hesitant about
forming them. This doubt and hesitancy will
naturally dispose them to distrust testimony
and education and refrain from peremptory
(tacit, direct) judgement (EHU 5.1 / 40–1,
12.24–6 / 161–3; DNR 1.133–4). It will also
extend to philosophical beliefs, as Hume
made clear in the last words of Treatise 1.4.7
/ 274.
. . . we are apt not only to forget our scep-
ticism, but even our modesty too; and
make use of such terms as these, ’tis evi-
dent, ’tis certain, ’tis undeniable; which a
due deference to the public ought, per-
haps, to prevent . . . . such expressions
were extorted from me . . . , and imply no
dogmatical spirit, nor conceited idea of
my own judgement, which are sentiments
that I am sensible can become no body,
and a sceptic still less than any other.
But a sceptical disposition will at least
put philosophical beliefs on an equal footing
with ‘unphilosophical’ ones, where they have
a greater chance of winning assent after due
consideration.
Seen in this light, Hume’s account of the
‘unphilosophical’ influences we are subjected
to and of the ultimate lack of any founda-
tion even for ‘philosophical’ belief constitutes
the best sceptical lesson, and thereby the
best lesson in logic, that anyone could have.
In the Treatise Hume overplayed this result.
Realizing the salutary effects that a sceptical
disposition could have, he set out to blast
his readers with the most extreme – but also
the most strained – sceptical arguments he
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
could invent. Not content to raise sceptical
doubts about causal inference in Treatise 1.3
/ 69–179, Hume went on to raise sceptical
arguments against the existence of an external
world, against the existence of persisting sub-
stances, and even against the validity of dem-
onstration and the existence of a self. It is as
if he thought that his reader needed first to be
driven into a deep sceptical crisis in order to
be adequately prepared to undertake a prop-
erly scientific study of the foundations of mor-
als in the passions, as taken up in Treatise 2
and 3.37 But he seems to have quickly realized
that the excessive sceptical arguments had the
opposite effect. By contesting received opin-
ions too forcefully, he had only led readers to
reject the Treatise as a whole. The Enquiry
takes a different tack. Though it mentions rea-
sons for scepticism both about the existence
of an external world and about the validity
of probable reasoning, it also discounts them
both, presenting them merely as a means to
inducing a properly scientific attitude. Hume
seems to have realized that there is no better
sceptical argument than the presentation of
the ‘system’ with its consequences. The ‘sys-
tem’, which diagnoses the problem, also cures
it by means of that very diagnosis.
NOTES
1
For a critique of such reasons as Hume gave
(at THN 1.3.2.2 / 73–4), see L. Falkenstein and
D. Welton, ‘Humean Contiguity’, History of
Philosophy Quarterly 18 (2001), pp. 279–96.
2
Hume expressed some doubts about whether
spatial contiguity and succession are always
necessary for causality, but considered the ques-
tion moot (THN 1.3.2.6 / 75 and 1.3.2.8 / 76).
3
THN 1.3.14.14 / 162 might be read as suggest-
ing that we have no idea of necessary connec-
tion. But Hume’s concern there was not to deny
that we have the idea of a connection, or even
of a necessary connection, but just that we have
127
9780826443595_Ch04_Final_txt_print.indd 127
the idea of what, specifically, connects causes
necessarily to their effects.
4
Combined with either an identification of
objects with perceptions of objects (as at THN
1.4.2 / 187–218 or EHU 12 / 149–165) or an
appeal to a conceivability criterion of meta-
physical possibility (as at THN 1.4.5.5 / 233),
this argument entails the independence of all
objects occupying distinct locations in space
and/or time, and so rules out the existence
of necessary connections between causes and
effects. The extent of Hume’s commitment
to the metaphysical impossibility of neces-
sary connections between causes and effects
remains controversial. For discussion, see
H. Beebee, Hume on Causation (London:
Routledge, 2006) and her contribution to this
volume, or the papers collected in R. Read and
K.A. Richman (eds), The New Hume Debate
(London: Routledge, 2000).
5
For more on Hume’s view of intuition and
demonstration, see D. Owen, Hume’s Reasons
(Oxford: Oxford University Press, 1999),
chap. 5.
6
At THN 1.3.6.9 / 90–1 Hume stressed that in
allowing for the possibility of hidden powers
he was making a concession to his opponent
for the sake of argument, not admitting that the
possibility is a real one.
7
The reader who has been keeping track of the
references will note that I have juggled Hume’s
order of presentation. This avoids the repetition
of the same arguments under different headings
that mars the exposition of the Treatise.
8
Here what I have charitably described as a
‘meditative path of discovery’ takes on a rhet-
orical dimension. Prepare your reader to accept
what you have to say by first inducing a deep
sense of puzzlement. Then offer what you have
to say as a solution to the puzzle and trust to
the reader’s sense of relief to induce acceptance
of your solution, even in the absence of sup-
porting argument.
9
Like necessary connection, the nature and role
of reasoning in causal inference is controver-
sial. For discussion, see P. Millican, D. Owen
and D. Garrett in the symposia on Garrett’s
Cognition and Commitment (Oxford: Oxford
University Press, 1997) printed in Hume Studies
24 (1998), pp. 141–59, 171–94 and Philosophy
and Phenomenological Research 62 (2001),
pp. 191–6, 205–15.
2/9/2012 9:39:16 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
10
I here amplify on Hume’s actual statement.
When discussing belief, Hume did not mention
14
the spatial and temporal contiguity condi-
tions he previously identified as involved in
the causal relation. He spoke just of belief in
the existence of an object, not of belief in its
existence at a prior or subsequent time and at
a contiguous place. This omission has momen-
15
tous consequences, generating pseudo-prob-
lems, most notably at THN 1.4.2.21 / 197–8.
11
An integrated account of the four features is
presented at THN 1.3.7.7 / 628–9.
12
THN 1.3.10.10 / 630–1 grapples with a fur-
ther problem that might have exercised Hume:
works of fiction focus attention and arouse
passions without prompting belief. For further
discussion of the problems with and prospects
for including mental dispositions within the
larger framework of Hume’s theory, see J.
Bricke, Hume’s Philosophy of Mind (Princeton:
16
Princeton University Press, 1980), chap. 3;
S. Everson, ‘The Difference between Feeling
and Thinking’, Mind 97 (1988), pp. 401–13;
17
L. Loeb, Stability and Justification in Hume’s
Treatise (Oxford: Oxford University Press,
2002), pp. 60–100; and J. Smalligan Marušić,
18
‘Does Hume Hold a Dispositional Account
of Belief?’, Canadian Journal of Philosophy
40 (2010), pp. 155–83. For further discussion
19
of Hume’s ambivalence about his account
20
of belief see M. Bell, ‘Belief and Instinct in
Hume’s First Enquiry’, in P. Millican (ed.),
Reading Hume on Human Understanding
(Oxford: Clarendon Press, 2002), pp. 175–85.
13
Hume was later to argue that intuition and
21
demonstration reduce to probability (THN
22
1.4.1 / 180–7). But even then his claim was
not that we do not intuit relations between
ideas or demonstrate truths in mathematics
by appeal to a chain of intuitions. It was
that because we sometimes have the wrong
intuitions, our assurance of the results of
a demonstration has to be informed by
considerations of how likely it is that we are
mistaken. Intuition and demonstration do not
‘reduce’ to probability in the sense of turning
out to be nothing other than more vivacious
conceptions of an idea. They ‘reduce’ to
probability in the sense of presupposing
second-order beliefs about the reliability
of our intuitive judgements. My intuitive
judgement that eight plus seven is fifteen has
128
9780826443595_Ch04_Final_txt_print.indd 128
no vivacity. My belief that I have correctly
intuited this relation does.
This view of the relation between the works
is controversial. For an opposed view see
P. Millican, ‘The Context, Aims and Structure
of Hume’s First Enquiry’, in P. Millican (ed.),
Reading Hume on Human Understanding,
pp. 27–65, esp. pp. 40–8.
Unlike the Treatise, the Abstract contains no
explicit pronouncement to the effect that in
speaking of ‘[t]he powers, by which bodies
operate’, Hume was indulging common but
false ways of thinking (cf. THN 1.3.6.8–10 /
90–1). In contrast, the parallel discussion in
the Enquiry contains a qualification, this time
occurring in a footnote, to the effect that the
talk of hidden powers is ‘loose and popular’
and that a ‘more accurate explication’ of the
notion would further buttress the conclusion
to be drawn here.
As noted earlier, the exposition of the Treatise
is made more elegant by bringing this after-
thought forward.
Compare THN 1.3.16 / 176–9, which appeals
to the abilities of animals as a further reason to
accept the account of belief.
As in the Treatise Hume continued, in both the
Abstract and the Enquiry, to omit this impor-
tant detail.
See note 8 above.
The interested reader is invited to consult
THN-C: lxv–lxvii, which sets out the extent
of Hume’s quotations from the Treatise in
Enquiry 5, Part 2 (EHU 5.10–22 / 47–55) and
his modifications to those passages.
This is obviously intended.
Objects can of course move around. But
just as we do not assume that two objects
are relatively immovable upon having once
seen them alongside one another, so we do
not assume that they are causally related
upon once having seen them in succession.
And just as the bare experience of a constant
conjunction in time suffices to impel us to
associate them independently of any further
justification by appeal to secret powers
producing a necessary connection, so the
bare experience of a constant conjunction in
space suffices to impel us to associate them
independently of any further justification
by appeal to the causes of their mobility or
immobility.
2/9/2012 9:39:17 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
23
The effect of resemblance in raising religious
devotion is of this sort. The icons and cere-
monies enliven an antecedent belief in the past
existence of people and events, the exception
being that in this case the antecedent belief is
grounded on testimony rather than memory.
Belief is only enhanced by the experience of the
icon or ceremony, not created.
24
This is the first appearance of the important
notion of correction by appeal to general rules.
25
This attempt to distinguish causality from con-
tiguity and identity is a failure. When one adds
a specification of direction and distance to
contiguity relations, and of temporal distance
to identity relations, they become as restrictive
as causal relations. Any given object is causally
related to a huge number of others as well, if
we do not consider whether the objects lie in
27
the direction of cause or effect, or distinguish
between proximate and remote causes and
effects.
26
When I speak here of a lack of uniformity in
past experience, I mean a verified lack of uni-
formity, where careful scrutiny of the contigu-
ous regions is unable to uncover any evidence
of the existence of the inferred object, not an
unverified lack of uniformity, where one fails
to observe a cause or effect simply because
one failed to look for it or (as in the case of
historical inference) was in no position to
observe it. Hume at one point grossly over-
stated the extent of the lack of uniformity in
our experience, pretending that the turning of
the head or closing of the eyes could prevent
us from considering a succession between
two types of objects to be perfectly constant
(THN 1.4.2.21 / 197–8). This is an artifact of
28
a mistake lamented in a number of previous
notes: Hume’s persistent neglect of the point
that, according to his own theory, causes and
effects are not merely conceived to exist, but
to exist at a certain place at a certain time. If
I am habituated vividly to imagine a cause or
effect in one place, and I consider myself to
have turned my head to look at another place
or to have closed my eyes and not be looking
at all, I am not going to suppose that I have
29
experienced a failure of my expectations. As
Hume himself pointed out elsewhere, when
we reason from causes to effects or effects to
causes, it is always the case that the object
we reason to is unperceived. Where both
129
9780826443595_Ch04_Final_txt_print.indd 129
cause and effect are present, the case is one of
perception rather than causal inference (THN
1.3.2.2 / 73). The cases where one turns
one’s head or closes one’s eyes are precisely
the occasions on which causal inference is
called for. Had Hume been right at THN
1.4.2.21 /197–8, there would be no such
thing as causal inference. There would only
be perceptions of regularities in the succession
of causes and effects and perceptions of the
failure of those regularities to occur, without
any attendant instinct to form a belief on
the latter occasions. The ‘system’ rules that
possibility out without any need to invoke
the elaborate mechanism proposed at THN
1.4.2.24–43 / 199–210 to provide for belief in
the continued existence of objects when not
perceived.
These pronouncements seem inconsistent with
THN 1.3.8.13–14 / 103–5, which declares
that ‘in all the most establish’d and uniform
conjunctions of causes and effects, such as
those of gravity, impulse, solidity, &c. the mind
never carries its view expressly to consider any
past experience’, illustrating the claim with
the point that a traveller who runs into a river
does not consult past experience when forming
the belief that walking out on the water will
be followed by sinking and suffocating. The
two passages can be reconciled if Hume’s point
in THN 1.3.12.3 / 131 and 1.3.12.7 / 133 is
taken to concern just those causal inferences
involving new, rare and unusual objects and
the formation of new causal laws. Our every-
day inferences concerning familiar objects can
be considered to proceed in accordance with
habits learned in infancy.
For Hume, mathematical results are matters
of knowledge, not belief, where the knowledge
arises from not being able to conceive things in
any other way (THN 1.3.11.2 / 124, cf. 1.3.7.3
/ 95). One can have knowledge without belief
because belief involves attention, elevation of
passion and inclination to action, and merely
appreciating the impossibility of conceiving
things any other way need have none of those
results.
For earlier versions of the account given here
see F. Wilson, Hume’s Defence of Causal
Inference (Toronto: University of Toronto
Press, 1997), chap. 2, esp. pp. 123–40 and
W.E. Morris, ‘Belief, Probability, Normativity’,
2/9/2012 9:39:17 PM
 HUME’S ACCOUNT OF PROBABLE REASONING
35
in Saul Traiger (ed.), The Blackwell Guide to
Hume’s Treatise (Malden: Blackwell, 2006),
36
pp. 77–94, esp. pp. 85–91.
30
Hume denied this at THN 1.3.11.8 / 127.
31
If we calculate probabilities in the common way,
this one-third belief would correspond to a two-
thirds probability. The point to keep in mind is
that Hume was not concerned about accounting
37
for the mathematical probability of an outcome,
but for the strength of belief in that outcome.
32
For further commentary on this argument
see C. Howson, Hume’s Problem
(Oxford: Clarendon Press, 2000), p. 14
and chap. 4.
33
For another ironic twist, see THN 1.3.13.12 /
149–50.
34
In this I follow Beebee, Hume on Causation,
pp. 71–4.
130
9780826443595_Ch04_Final_txt_print.indd 130
For notable instances see E 135–7; E 528–9;
H 5.67; and NHR as a whole.
For more on this, see D.F. Norton, ‘How
a Sceptic May Live Scepticism’, in J.J.
MacIntosh and H.A. Meynell (eds), Faith,
Scepticism, and Personal Identity (Calgary:
The University of Calgary Press, 1994),
pp. 119–39, esp. pp. 128–32.
This idea has been pursued in more detail by
Loeb, Stability and Justification in Hume’s
Treatise, pp. 6, 36, 215–29. However, Loeb
is inclined to attribute Hume’s overstated
scepticism to a ‘somewhat perverse’ (ibid., p. 16)
aspect of his ‘temperament’ (ibid., p. 229) rather
than an attempt to cause the reader to become
more reflective and hesitant about all his or her
beliefs.
2/9/2012 9:39:17 PM
 5
CAUSATION AND
NECESSARY CONNECTION
Helen Beebee
1. INTRODUCTION
It is difficult to say anything about Hume’s
views on causation and necessary connection
without making claims that are hotly dis-
puted amongst interpreters of Hume’s work.
Some interpreters take Hume to be a causal
realist, while others hold that he is a regular-
ity theorist. Some take him to hold that ‘caus-
ation’ is an irretrievably defective notion that
could not possibly apply to any worldly phe-
nomena, and some take him to be a non-cog-
nitivist about our causal talk and thought.
Some take him to hold that there is such a
thing as objective necessary connection, while
others take him to be a subjectivist about
necessity. And so on. In this chapter, I shall
strike a path through these and other inter-
pretative controversies as follows. I begin
in section 2 by sketching, in what I hope is
a reasonably interpretatively neutral way,
the bare bones of Hume’s account of caus-
ation, and in section 3 I discuss his famous
‘two definitions’. In section 4, I outline the
three main classes of interpretative position –
what I shall call the traditional, sceptical
realist and projectivist interpretations – and
briefly examine the main items of evidence
that are normally marshalled for and against
131
9780826443595_Ch05_Final_txt_print.indd 131
each interpretation. Roughly speaking, these
interpretations take Hume to be, respectively,
a regularity theorist, a non-cognitivist, and a
realist about causation. Finally, in section 5,
I sum up the current state of play, which, as
I see it, is something of a stand-off between
the projectivist and sceptical realist interpret-
ations, and say something about the specific
problems that each interpretation needs to
overcome if it is to prevail over its rival.
2. HUME’S BASIC ACCOUNT OF
CAUSATION
For Hume, causal thinking lies right at the
heart of our conception of the world: all ‘rea-
sonings concerning matters of fact and exist-
ence seem to be founded on the relation of
Cause and Effect. By means of that relation
alone we can go beyond the evidence of our
memory and senses’ (EHU 4.3 / 26). In other
words, our access to external reality, beyond
the ‘evidence’ of current experience and
memory, entirely depends on reasoning from
causes to effects (and vice versa). When I form
the belief that my dinner will not poison me,
or that what I am reading in the newspaper
2/9/2012 9:34:59 PM
 CAUSATION AND NECESSARY CONNECTION
is true, or that the kettle I turned on a few
minutes ago will have boiled by now, I do so
on the basis of such reasoning.
This being so, the primary task that Hume
sets himself is to uncover the nature of this
mysterious relation – or rather, to discover
what our idea of causation consists in, since it
is ‘impossible to reason justly, without under-
standing perfectly the idea concerning which
we reason; and ’tis impossible to understand
any idea, without tracing it up to its origin,
and examining that primary impression from
which it arises’ (THN 1.3.2.4 / 74–5). He
quickly discovers that ‘whatever objects are
consider’d as causes and effects are contigu-
ous’ (THN 1.3.2.6 / 75) and are also such that
the cause is temporally prior to its effect. But
those two conditions clearly do not exhaust
our idea of causation, for ‘there is necessary
connexion to be taken into consideration’
(THN 1.3.2.11 / 77).
Hume’s search for the impression-source
of the idea of necessary connection is a long
one, for it turns out that this crucial com-
ponent of ‘the idea concerning which we
reason’ has its source in that very reasoning
itself. So Hume needs to discover the nature
of our reasoning from causes to effects (what
I shall call ‘causal reasoning’) before he can
locate the impression-source of the idea of
necessary connection.
Hume’s investigation into causal reason-
ing – what is traditionally described as his
discussion of the ‘problem of induction’ –
yields two results that are significant for
present purposes. First, causal reasoning
proceeds on the basis of past observed regu-
larity: on observing a C, we infer that an E
will follow just when we have experienced
sufficiently many Cs followed by Es. Second,
that reasoning proceeds not by considera-
tion of any argument, but as a matter of
‘custom or habit’ (EHU 5.5 / 43). There
132
9780826443595_Ch05_Final_txt_print.indd 132
is simply a mental mechanism that, given
relevant past experience, conveys the mind
from the impression of a C (one billiard ball
striking another, for example) to a belief that
an E will follow (the second ball moving).
And it turns out, in ‘Of the idea of necessary
connexion’ (THN 1.3.14 / 155–72; EHU
7.26–30 / 73–9), that it is the operation of
this very mechanism that furnishes us with
the impression, and so the idea, of neces-
sary connection. Before Hume can establish
this latter claim, however, he needs to show
that we have no sensory impression of nec-
essary connection. Before briefly rehearsing
his argument, it is worth saying something
about the importance of the issue for Hume.
One of Hume’s main aims is to provide a
‘science of man’: an account of how the mind
works that is based on a clear-headed, ‘exper-
imental’ investigation. A plausible account of
his primary target is given by Edward Craig,1
who sees Hume’s major adversaries as those
who uphold what he calls the ‘Image of God’
doctrine (we are made in God’s image) and a
corresponding epistemological doctrine: the
‘Insight Ideal’. According to the Insight Ideal,
the nature of reality – or at least some of
it – is in principle accessible to reason; thus
philosophers before Hume had subscribed
to the self-evident or a priori status of the
claim that every event has a cause, or had
claimed that the essence of objects can be
known by what Descartes calls ‘purely men-
tal scrutiny’.2 Hume, by contrast, sets out to
undermine systematically the claim that any
aspect of the nature of reality is knowable a
priori, and, moreover, to show that no ‘mat-
ter of fact’ can be inferred a priori from any
other distinct matter of fact.
Hume takes himself to have established
this claim in his discussion of causal reason-
ing; but it is a claim to which he returns in
the negative phase of his discussion of the
2/9/2012 9:35:00 PM
 CAUSATION AND NECESSARY CONNECTION
idea of necessary connection (EHU 7.1–25 /
60–73; THN 1.3.14.1–18 / 155–65), where
he argues that the idea does not have a sen-
sory impression-source. Hume assumes, in
this discussion, that such an impression-
source for the idea of necessary connection
would have to be such that it delivers cer-
tainty that the effect will follow: ‘were the
power or energy of any cause discoverable
by the mind, we could foresee the effect,
even without experience [of past constant
conjunction]; and might at first, pronounce
with certainty concerning it, by the mere
dint of thought and reasoning’ (EHU 7.6 /
63). With this assumption (to which I return
below) in place, it is an easy matter to estab-
lish Hume’s negative conclusion, since, as we
already know from his discussion of causal
reasoning, observation of a particular event
never delivers such certainty: we can always
imagine the cause happening without its
effect, and so it is always epistemically pos-
sible that the effect will not occur.
The assumption just mentioned has caused
much puzzlement amongst commentators.
Hume is apparently arguing for a phenom-
enological claim – that on first observing
them, ‘[a]ll events seem entirely loose and sep-
arate. One event follows another; but we can
never observe any tie between them’ (EHU
7.26 / 74) – and yet his argument for this
claim proceeds by way of pointing out that
we cannot ‘pronounce with certainty’ that
the effect will follow, on observing the cause.
But how is the former claim supposed to fol-
low from the latter? In particular, is Hume
not confusing two distinct notions? On the
one hand, we have the claim that there is no
observable power, within the cause itself (e.g.
the striking of one billiard ball by another),
such that observing it would deliver certainty
that the effect will follow. (J.L. Mackie calls
such a power ‘necessity2’.3) But this does not
133
9780826443595_Ch05_Final_txt_print.indd 133
entail that there is no observable connection
or ‘tie’ between cause and effect (necessity1).
It might easily be that we can observe such
a connection – that is, we can observe the
causal relation – despite the fact that we can-
not observe any power, in the cause itself,
that produces certainty about the effect.
It seems that Hume does indeed run these
two distinct notions together; however, by
his own lights it is not clear how much of
a problem this is. Hume’s central concern,
remember, is with inference from one mat-
ter of fact to another – that is, from causes
to effects (and vice versa). The observability
of necessity1 – of a mere causal ‘tie’ between
causes and effects – would make no substan-
tive difference to the account that Hume has
already offered of such reasoning, since such
a tie could only be observed by, as it were,
observing c-causing-e as a package deal,
and could therefore not serve as the basis of
causal reasoning. For how would we then
reason, when confronted with a C? The best
we could do is reason that, since Cs have
always been observed to cause Es in our past
experience, the currently observed C will
likewise cause an E. But this inference cannot
be a priori, since it is still epistemically possi-
ble that the former is true and the latter false.
So we need to postulate a different mental
mechanism that supplies the inference, and
that mechanism would turn out (according
to Hume’s own argument) to be custom or
habit. Thus the only difference between the
account just canvassed and Hume’s own
account of causal reasoning is that accord-
ing to the former account the impression of
causation is supplied by sensation, whereas
according to Hume’s account it has another
source.
That source, of course, turns out to be the
very inference from causes to effects itself:
the ‘transition arising from the accustom’d
2/9/2012 9:35:00 PM
 CAUSATION AND NECESSARY CONNECTION
union’ (THN 1.3.14.19 / 165), that is, the
habit that takes the mind from an impression
of the cause, together with experience of past
constant conjunction, to the belief that the
effect will follow. As Hume puts it,
when one particular species of event has
always . . . been conjoined with another,
we make no longer any scruple of fore-
telling one upon the appearance of the
other, and of employing that reasoning,
which can alone assure us of any matter
of fact or existence. We then call the one
object, Cause; the other, Effect. (EHU
7.27 / 74–5)
Hence the ‘connexion . . . which we feel in the
mind, this customary transition of the imagi-
nation from one object to its usual attendant,
is the sentiment or impression from which
we form the idea of power or necessary con-
nexion’ (EHU 7.28 / 75).
Hume thus finally achieves what he set
out – nearly a hundred pages earlier, in the
Treatise (THN 1.3.2.4 / 74) – to achieve: the
impression-source for the idea of necessary
connection. Unfortunately, however, it is far
from clear what consequences Hume takes
his discovery to have for our causal talk and
thought, for he appears to endorse three posi-
tions that are mutually inconsistent:
1. He seems to think that we are apt to
project the impression – and the idea – of
necessary connection into the world: ‘the
mind has a great propensity to spread itself
on external objects, and to conjoin with
them any internal impressions, which they
occasion’ (THN 1.3.14.23 / 167). And he
seems to suggest that this projection is a mis-
take: ‘we are led astray by a false philosophy’
when ‘we transfer the determination of the
thought to external objects, and suppose any
real intelligible connexion betwixt them; that
being a quality, which can only belong to the
134
9780826443595_Ch05_Final_txt_print.indd 134
mind that considers them’ (THN 1.3.14.25
/ 168). This suggests that it is a mistake to
think that events in the world really are nec-
essarily connected to one another.
2. He does not appear to suggest that
the idea of causation can be stripped of the
component idea of necessary connection: he
does not appear to respond to the mistake
just identified by advocating a revisionary
account of the concept of cause.
3. He appears to think that our causal
thought and talk is truth-apt. (Certainly it
is subject to normative constraints: in the
Treatise the very next section lists ‘rules by
which to judge of causes and effects’ (THN
1.3.15 / 173–6).)
The inconsistency is easy to see: by (2), we
really do deploy the idea of necessary con-
nection when we engage in causal talk and
thought; and, by (3), that causal talk is in
general entirely legitimate (not least because
it is subject to normative constraints; it is
hard to see how this could be so if all such
talk was irredeemably defective). And yet, by
(1), that talk is irredeemably defective. There
can be nothing in the world that answers to
the idea of necessary connection, since that
idea derives solely from the ‘determination
of the mind’. The three broad interpretative
rivals discussed in section 4 below resolve
the inconsistency in different ways. Roughly
speaking, the traditional interpretation denies
(2),4 and both the projectivist and sceptical
realist interpretations finesse (1): they hold
Hume to the claim that there is a mistake
in the offing, but deny that the mistake in
question is that of thinking of events in the
world as causally or necessarily connected.
Finally – and this is what distinguishes these
two interpretations – the projectivist inter-
pretation finesses (3) as well: Hume does
endorse our causal talk and thought, but
that talk and thought is to be understood in
2/9/2012 9:35:01 PM
 CAUSATION AND NECESSARY CONNECTION
non-cognitivist terms. For the sceptical real-
ist, by contrast, our causal talk and thought
is straightforwardly referential: there is
something in the world that answers to our
idea of causation (or at least we believe that
there is, and there is nothing defective about
that belief).
One more piece of the already difficult
puzzle needs to be put on the table, namely
Hume’s famous two definitions of causa-
tion; this is the topic of the next section.
Before leaving Hume’s discussion of the
origin of the idea of necessary connection,
however, it is worth noting what is, in my
view, an often misunderstood feature of his
account. It is routinely taken for granted
that in Hume’s view, all events ‘seem entirely
loose and separate’. That is, phenomenologi-
cally speaking, our experience is merely as
of one event following another, even once
the habit of inference has been established,
and so even once the impression of necessary
connection is present. This has forced some
interpreters to cast the impression of neces-
sary connection as, for example, simply a
‘peculiar feeling’5 or a ‘feeling of helplessness
or inevitability’(where the inevitability is the
inevitability of one’s expectation, and not the
inevitability of the effect itself).6
Hume is not, in fact, committed to this
view. What he says is that ‘all events seem
entirely loose and separate’; he does so in
the context of ‘single instances of the oper-
ation of bodies’ (EHU 7.26 / 73) – that is,
when we first observe a pair of contiguous
events, and hence before the habit of associa-
tion has arisen. He does not explicitly state a
view about how things seem once the habit
of association has arisen, but it is plausible
to suppose that he takes the impression of
necessary connection to affect, precisely, how
things seem – that is, to affect the nature of
visual experience itself.
135
9780826443595_Ch05_Final_txt_print.indd 135
This interpretation removes the need to
think of the impression of necessary connec-
tion as a ‘feeling of expectation’ or similar;
how things seem, when the impression arises,
is, precisely, necessarily connected, just as it is
by virtue of the impression of red that things
seem red to us. It also helps to explain why
Hume offers a non-phenomenological argu-
ment for the claim that there is no sensory
impression of necessary connection. After all,
if all events really did seem entirely loose and
separate to us, Hume would not need such
an argument; unbiased phenomenological
reflection would do the job just by itself (and
would thereby establish the stronger claim
that we have no impression of either neces-
sity1 or necessity2).7
3. THE TWO DEFINITIONS
Here are Hume’s two definitions of cau-
sation – or, more precisely, definitions of
‘a cause’ – as they appear in the Treatise,
towards the end of his discussion of the idea
of necessary connection:
(D1) ‘An object precedent and contigu-
ous to another, and where all the objects
resembling the former are plac’d in like
relations of precedency and contiguity to
those objects, that resemble the latter’.
(THN 1.3.14.29 / 170)
(D2) ‘[A]n object precedent and con-
tiguous to another, and so united with
it, that the idea of the one determines
the mind to form the idea of the other,
and the impression of the one to form
a more lively idea of [that is, a belief in]
the other’. (ibid.)
Unfortunately, rather than clarifying the situ-
ation, the two definitions are problematic in
2/9/2012 9:35:01 PM
 CAUSATION AND NECESSARY CONNECTION
their own right. The first thing to note is that
they are not even extensionally equivalent.
The conditions in (D2) can easily be satis-
fied without (D1) holding, by someone hav-
ing observed an unrepresentative sample of
‘objects’ (or events), so that the two kinds of
event are constantly conjoined in their expe-
rience but not universally. And (D1) can be
satisfied without (D2) holding, for instance in
the case of constant conjunction between two
kinds of events that nobody has observed, so
that nobody’s mind is determined to move
from the idea of one to the idea of the other.
A standard solution to this problem8 has
been to claim that Hume only intends (D1)
as a genuine definition of causation, while
(D2)’s aim is different: to explain the con-
ditions under which we do, in fact, come
to make causal claims, for example. This
move seems somewhat ad hoc, however,
given Hume’s claim that the two definitions
present ‘a different view of the same object’
(THN 1.3.14.29 / 170).
A second solution, offered by Don Garrett,9
is to distinguish a ‘subjective’ from an ‘abso-
lute’ reading of each definition. Roughly, a
subjective reading of (D1) would take ‘all
objects’ to mean all objects observed so far
by a particular person. This would then be
coextensive with (D2) read subjectively,
with ‘the mind’ understood as referring to
the same person, since an ‘object’ that meets
(D1) now will be such that the mind of the
person in question is indeed determined by
the idea of the object to form the idea of its
effect. As Garrett puts it, read subjectively,
the two definitions tell us when an object
‘functions psychologically’ as a cause.10 An
absolute reading of (D1) takes ‘all objects’
to be unrestricted, so that (D1) appeals to
universal constant conjunction. This is coex-
tensive with (D2) read absolutely, with ‘the
mind’ now understood to refer to some sort
136
9780826443595_Ch05_Final_txt_print.indd 136
of ‘idealized spectator’, that is, an observer
who observes representative samples of all
kinds of constantly conjoined events.
A third solution – similar to Garrett’s
‘subjective’ reading and to Robinson’s pro-
posal concerning the second definition – is
to read both definitions not as ‘definitions’ in
the standard contemporary sense at all, but
as saying how it is we come to believe that
one thing is a cause of another. As Edward
Craig puts it, the definitions characterize the
‘circumstances under which belief in a causal
connection arises, one concentrating on the
outward situation, the other on the state of
the believer’s mind that those outward facts
induce’.11
My own view is that none of these solu-
tions are satisfactory, because they all ignore
Hume’s preceding remark in the Treatise that
‘two definitions of this relation may be given
of this relation, which are only different, by
their presenting a different view of the same
object, and making us consider it either as a
philosophical or as a natural relation; either as
a comparison of two ideas, or as an associa-
tion betwixt them’ (THN 1.3.14.31 / 169–70).
Hume’s distinction between philosophical
and natural relations is a distinction between
two kinds of mental operation. Roughly, the
former is the conscious ‘placing’ of two ideas
under a relation (hence ‘plac’d’ in (D1)), and
the latter is the unconscious ‘transition’ of the
mind from one idea to another. For example,
resemblance – which is the other relation
that is both natural and philosophical – can
operate in two distinct ways, as when I con-
sider whether a painting of a particular scene
resembles the image of a particular remem-
bered scene I have in my mind and come to
judge that it does (philosophical relation), or
when I see a picture of the Queen and my
mind is automatically drawn to the idea of
the Queen herself (natural relation). Similarly
2/9/2012 9:35:01 PM
 CAUSATION AND NECESSARY CONNECTION
for causation: I can ‘place’ two ideas under
the relation of causation, and will (or should)
do so precisely when the conditions speci-
fied in (D1) are met (contiguity, precedence
and observed constant conjunction), thereby
coming to form the judgement that one event
is the cause of the other. And I ‘naturally’
judge one event to be the cause of another
when I have acquired the relevant habit of
association: my mind is drawn from the idea
(or impression) of the first event to the idea of
(or belief in) the second, and, again, I thereby
come to form the judgement that the first
event is the cause of the second.12
Note that neither of the last two of the
four interpretative positions just outlined
delivers any verdict about the meaning of
‘cause’, since both deny that the definitions
are definitions in anything like the standard
contemporary sense. Instead, the definitions
tell us something about how it is we come to
make causal judgements. As we shall see, the
availability of these interpretative options
with respect to the two definitions under-
mines a key component of the motivation
for the traditional interpretation of Hume on
causation, according to which he is a naive
regularity theorist. We are not required to
read the first definition as a definition of
the meaning of ‘cause’, and so we are not
required (at least not required by the two
definitions) to hold Hume to the view that
causation consists in contiguity, precedence
and constant conjunction.
4. INTERPRETATIONS:
TRADITIONAL, SCEPTICAL REALIST
AND PROJECTIVIST
It is uncontroversial that Hume endorses
the following three theses. First, at least
137
9780826443595_Ch05_Final_txt_print.indd 137
in the most basic cases, we come to make
causal judgements – judgements of the form
‘c caused e’, or ‘c will cause e’ – on the basis
of the temporal priority and contiguity rela-
tions that hold between these two events, and
the observed constant conjunction between
events of the kinds that c and e instantiate.
In such cases we infer that e will occur on the
basis of having observed c, and we also think
of c as a cause of e, in a way that, somehow or
other, involves deploying the idea of necessary
connection. Second, the impression-source of
that idea is the inference just described: the
impression of necessary connection is not
a sensory impression but an impression of
reflection. Finally, there is – or can be – some-
thing awry in our deployment of the idea of
necessary connection. Perhaps the easiest way
of seeing the difference between the three
main interpretative positions is to consider
how they interpret the second and third of
the claims just described. I shall thus start
my brief account of each of the positions by
describing their attitudes to those claims.
THE TRADITIONAL INTERPRETATION
According to what I am calling the ‘tradi-
tional’ interpretation, Hume is a naive regu-
larity theorist about causation: c causes e if
and only if c is contiguous with and prior to e,
and events similar to c (the Cs) are constantly
conjoined with events similar to e (the Es).13
In other words, ‘c causes e’ just means ‘all
Cs are followed by Es’. What is awry in our
deployment of the idea of necessary connec-
tion, in this view, is, precisely, that we deploy
it at all: given that the source of that idea is
the transition of the mind, it cannot possi-
bly represent any mind-independent feature
of the world. Thus (and this is an issue on
which versions of the traditional interpreta-
tion differ) either (i) necessary connection is
2/9/2012 9:35:01 PM
 CAUSATION AND NECESSARY CONNECTION
not, in fact, part of the meaning of ‘cause’ as
we actually deploy the idea ‘cause’, or (ii) it
is part of the ordinary meaning of ‘cause’ but
Hume is in effect offering a revisionary con-
ceptual analysis of ‘cause’ (shorn of the trou-
blesome concept of necessary connection),
or (iii) it is part of the ordinary meaning of
‘cause’ and Hume offers no such revision:
‘cause’ is irredeemably defective.
The major problem with the traditional
interpretation is an almost total lack of evi-
dence in its favour. The major source of evi-
dence has standardly been thought to be the
‘two definitions’ – or rather, the first defin-
ition. However, as we saw in section 3, the
interpretation of the two definitions relied on
here (according to which the first definition
really is Hume’s attempt to offer a concep-
tual analysis of ‘cause’) has recently come
under attack. In addition, the accounts given
of the role of the idea of necessary connec-
tion offered by different versions of the trad-
itional interpretation fit badly with the text.
Against alternative (i) above, Hume does not
suggest (setting aside the first definition) that
the idea of necessary connection is no part of
the actual meaning of ‘cause’; nor, contra (ii),
does he suggest that the concept of causation
needs to be revised. He does, of course, sug-
gest that something is apt to go awry when
we deploy the idea of necessary connection;
but he does not suggest that the appropriate
response is to stop deploying it; indeed, it is
hard to see how Hume could even consider
this to be a genuine psychological possibil-
ity. Finally, against (iii), which I take to be
Barry Stroud’s position, Hume clearly and
persistently endorses a wide range of causal
claims, and indeed provides ‘rules by which
to judge of causes and effects’ (THN 1.3.15
/ 173–6). This is extremely difficult to square
with the claim that he takes all causal talk to
be equally and irredeemably false.
138
9780826443595_Ch05_Final_txt_print.indd 138
SCEPTICAL REALIST INTERPRETATIONS
In stark contrast to the traditional interpreta-
tion, the sceptical realist interpretation takes
Hume to hold that our causal talk refers –
successfully – to real, mind-independent
causal connections in nature, or what I shall
call ‘causal powers’.14 Exactly what role the
idea of necessary connection plays here is
something that sceptical realist interpreters
differ on. In particular, John Wright holds
that the idea of necessary connection refers
to genuine mind-independent necessary con-
nections, so that if only we could penetrate
into the true, underlying nature of causes,
we would be able to discern those (causal)
powers by which causes really do abso-
lutely necessitate their effects: he ‘retained
an ideal of knowledge of true causes which
was derived from the Cartesians’,15 so that
the ‘true manner of conceiving a particular
power in a particular body’ (of which we are
in fact incapable) would involve being ‘able
to pronounce, from a simple view of the one,
that it must be follow’d or preceded by the
other’ (THN 1.3.14.12 / 161).16 Thus the mis-
take we are apt to make when it comes to the
idea of necessary connection is the mistake of
holding that it derives from an impression of
sensation, and thus holding that we really do
perceive necessity. Hence our idea of neces-
sary connection is defective, in that – being
derived from an impression of reflection – it
cannot adequately represent real necessity in
nature; nonetheless, it succeeds in referring to
real necessity. In other words, Hume agrees
with his opponents when it comes to the
metaphysics of the Image of God doctrine, as
far as causation is concerned, but he disagrees
with them over the Insight Ideal: we lack the
God-like ability to penetrate into the essences
of things in a way that would reveal their true,
effect-guaranteeing, underlying nature to us.
2/9/2012 9:35:02 PM
 CAUSATION AND NECESSARY CONNECTION
According to Galen Strawson, by contrast,
Hume takes all necessity (whether causal or
logical) to be purely subjective: our idea of
necessity does not track any real necessity in
nature, but is ‘just a feeling we have about
certain things – about 2 + 2 = 4, and about
what this billiard ball does to that one, and
about the sum of the angles of a triangle’.17
Nonetheless, our idea of causation suc-
ceeds in referring to more in the world than
mere regularities: it refers to (to use Hume’s
own expression) that upon which the ‘regu-
lar course and succession of objects totally
depends’ (EHU 5.22 / 55); or, in Strawson’s
words, ‘whatever it is about the universe (or
matter) which is that in virtue of which it is
regular’.18 Again, a central thought here is
that our idea of causation is inadequate to
what it represents: we can have a ‘relative’
idea of it, but not what Strawson calls a ‘pos-
itively contentful’ idea of it. Thus Wright’s
Hume conceives of the referent of ‘cause’
to be objective necessity – the feature of the
cause that absolutely guarantees that the
effect will occur – whereas Strawson’s Hume
conceives of the referent of ‘cause’ to be a
regularity-guaranteeing feature of nature.
The starting point for the claim that there
is serious textual evidence to justify (some
version of) the sceptical realist interpreta-
tion is to take the first Enquiry rather than
the Treatise as expressing Hume’s consid-
ered view about causation. This is because it
is in the first Enquiry that Hume refers, on
several occasions, to the ‘powers and forces,
by which [the course of nature] is governed’
that are ‘wholly unknown to us’ (EHU 5.21 /
54), the ‘powers and principles on which the
influence of these objects entirely depends’
that nature ‘conceals from us’ (EHU 4.16
/ 33), and so on. It is also the place where
he says that our idea of causation is ‘imper-
fect’, and that it admits of no ‘just definition,
139
9780826443595_Ch05_Final_txt_print.indd 139
except what is drawn from something extra-
neous and foreign to it’ because we ‘have no
idea of this connexion, nor even any distinct
notion what it is we desire to know, when we
endeavour at a conception of it’ (EHU 7.29
/ 76–7).
It is possible to reinterpret these claims in
a way that does not commit Hume to belief
in causal powers. In particular, one might
take the claims about nature’s ‘secret’ pow-
ers to be mere suppositions for the sake of
the argument.19 Or one might point out that,
since Hume explicitly takes ‘power’ to be syn-
onymous with ‘cause’ (THN 1.3.14.4 / 157),
and since he doubtless thinks that there are
additional, not-yet-known regularities under-
lying the observed behaviour of objects, talk
of secret powers presents no problem for the
other interpretations.20 But there is nothing
in the text of the first Enquiry itself to moti-
vate such a reinterpretation (or so defend-
ers of sceptical realism maintain): the most
natural interpretation of the first Enquiry,
taken on its own terms, reads Hume’s claims
at face value, as expressions of belief in, but
ignorance concerning the nature of, causal
powers.
One advantage of Wright’s version of
the sceptical realist interpretation over the
traditional interpretation is that it makes
sense of the tension noted in section 2:
Hume endorses our (necessary-connection-
involving) causal thought and talk, and yet
he apparently thinks there is something
wrong with the idea of necessary connection.
Wright’s account resolves the tension by iden-
tifying what is ‘wrong’ with the idea of nec-
essary connection as our tendency to think
that necessary connections are perceivable,
and our corresponding tendency to think
that we have thus penetrated into the essence
of bodies: ‘[t]he vulgar mistake an associa-
tional connection for a genuinely perceived
2/9/2012 9:35:02 PM
 CAUSATION AND NECESSARY CONNECTION
rational connection’.21 This tendency does
not, according to Wright’s account, affect the
meaning of our causal talk, however. Our
habits of inference give rise to the belief that
there are necessary connections in nature,
and that belief is (for all we know) true; the
mistake we tend to make is the mistake of
thinking that our idea of necessary connec-
tion is fully adequate to what it represents.
When we ‘make the terms of power and effi-
cacy signify something, of which we have a
clear idea, and which is incompatible with
those objects, to which we apply it, obscu-
rity and error begin then to take place’ (THN
1.3.14.25 / 168). This is because we are in
effect claiming that there is some feature of
the external world that is adequately repre-
sented by an idea whose impression-source
is a mere transition in the mind, when in fact
there can be no such feature. But this mis-
take plays no role in the meaning of ‘cause’:
our idea of necessary connection really does
refer to real necessity, even if we are apt to be
mistaken about the nature of what it is we
thereby represent.
It is unclear whether Strawson’s interpre-
tation also has this advantage over the tra-
ditional interpretation, given his claim that
Hume takes all necessity to be subjective.
Strawson says that ‘the E-intelligible [that is,
positively contentful] meaning of the term
“causation” can only encompass certain
aspects of the experience Causation [that is,
causal power] gives rise to . . . [including]
the feeling of determination’.22 So according
to Strawson’s account, the idea of necessary
connection is only a part of the ‘meaning’ of
‘causation’ in the sense that it is the idea of
an experience that causation ‘gives rise to’.
But of course such a claim could equally
be made by a defender of the traditional
interpretation; so if Strawson’s claim here
makes adequate sense of the fact that Hume
140
9780826443595_Ch05_Final_txt_print.indd 140
endorses our necessary-connection-involving
causal talk and thought, the same can be said
of the traditional interpretation. My own
view is that Strawson’s suggestion does not
adequately capture the thought that the idea
of necessary connection really is part of the
meaning of ‘cause’: Hume really does seem
to think that our causal thought involves the
claim that causes and effects are necessarily
connected, and not merely that they happen
to give rise to a certain kind of ‘feeling’.
A second advantage that the sceptical realist
interpretation has been claimed by Strawson
to have over the traditional interpretation is
that the latter saddles Hume with the pre-
posterous (according to Strawson) claim that
there is nothing more to the world than mere
regularity (‘one of the most baroque meta-
physical suggestions ever put forward’).23 The
traditional interpretation need make no such
claim, however. According to the traditional
interpretation, Hume’s claim is only that
our thoughts cannot successfully reach out
to any mind-independent relations between
causes and effects, aside from priority and
contiguity. This is not to assert positively that
such relations do not, or cannot, exist – only
that we cannot succeed in referring to them
in our causal thought and talk (if we try, we
‘lapse into obscurity and error’). We can (as
Strawson says) form a ‘relative’ idea of such
relations – we can consider the possibility
that they exist (without being able to form
adequate ideas of what they might be like),
but to do so would be, at best, to indulge in
idle metaphysical speculation. From a seman-
tic point of view, then, the crux of the diffe-
rence between the traditional interpretation
and Strawson’s version of sceptical realism is
that for Strawson, the ‘relative’ idea we form
is a relative idea of real causal powers, which
are what our ordinary causal talk refers to.
According to the traditional interpretation,
2/9/2012 9:35:03 PM
 CAUSATION AND NECESSARY CONNECTION
by contrast, we can form a relative idea of
some possible not-further-specifiable relation
between causes and effects, but that idea is
not the idea of causation: the idea of caus-
ation is exhausted by contiguity, priority and
constant conjunction.
THE PROJECTIVIST INTERPRETATION24
The projectivist interpretation in some sense
represents a middle ground between the scep-
tical realist and traditional interpretations
(indeed Angela Coventry calls it the ‘interme-
diate interpretation’).25 It shares with scepti-
cal realism the thought that our causal talk
and thought does more than merely register
the existence of regularities in nature, and
with Wright’s version of sceptical realism
the thought that this involves the legitimate
deployment of the idea of necessary connec-
tion. However, the projectivist interpretation
shares with the traditional interpretation a
broadly meaning-empiricist interpretation of
Hume, according to which experience places
strict limits on what can be represented, via
our ideas, in our thought and talk. In particu-
lar, the two interpretations agree that noth-
ing in reality answers to our idea of necessary
connection: causation ‘in the objects’, as it is
sometimes put, amounts to no more than con-
tiguity, priority and constant conjunction.
The projectivist interpretation squares the
apparent circle by adopting a non-cognitiv-
ist approach – on the one hand, the idea of
necessary connection is deployed legitimately
in our causal talk and thought, but on the
other, no aspect of reality answers to this
idea. A standard interpretation of Hume’s
ethical (and indeed aesthetic) writing takes
him to be a ‘sentimentalist’ about moral (and
aesthetic) claims: when we make an evalu-
ative claim about an action or a person (that
they are good or bad, brave or cowardly,
141
9780826443595_Ch05_Final_txt_print.indd 141
and so on), we are expressing a sentiment or
moral attitude towards that action or person,
rather than attributing to them some mind-
independent moral property. Moreover (this
further move is admittedly more controver-
sial), we do not merely express the relevant
sentiment; we project it onto the object of
our experience or judgement, so that the
painting we are observing looks beautiful or
the murder will seem vicious thanks to the
projection of the relevant sentiment, and we
correspondingly judge them to be so thanks
to the projection of the relevant idea. Thus
Michael Smith notes that when Hume says
that you ‘never can find’ the viciousness of
a murder ‘till you turn your reflexion into
your own breast, and find a sentiment of dis-
approbation’ (THN 3.1.1.24 / 468–9), he ‘is
precisely trying to focus our attention away
from where it is naturally focused when we
judge a wilful murder to be wrong: that is,
away from the murder itself, and onto an
otherwise quite unnoticed “calm passion” he
supposes to arise in us’.26
Part of the point here, according to the
projectivist line on morality, is that Hume
is not merely making a straightforward phe-
nomenological claim; it is not supposed to
be just obvious to us that there is nothing in
the murder itself that constitutes its vicious-
ness. On the contrary, he only gets to this
claim after quite a lot of argument. So – the
thought is – it seems to us that our moral and
aesthetic judgements are responses to genu-
ine features of external objects, and their
so seeming is due to the projection of the
relevant sentiment. Our sentiment-derived
judgements are, however, subject to norma-
tive constraints – so, for example, there are
‘rules of art’ which deliver a true standard
of taste and sentiment’,27 even though ‘no
sentiment represents what is really in the
object’.28 Hence moral and aesthetic claims
2/9/2012 9:35:03 PM
 CAUSATION AND NECESSARY CONNECTION
can legitimately be regarded as correct or
incorrect (there is a ‘true standard’ for them
to meet) on the basis of those normative
constraints.29
In the case of causation, the analogue of
‘sentiment’ is, of course, the impression –
and hence the idea – of necessary connection.
According to the projectivist view, we are apt
to mistake the impression of necessary con-
nection for a sensory impression (this is a
point of agreement with Wright), and we are
apt to do this because we project the impres-
sion onto the external objects that trigger it.
(My own proposal here is that the ‘impres-
sion’ of necessary connection is in fact, for
Hume, not a self-standing ‘feeling’ at all, but
merely the modification of visual experience
that occurs once our habit of expectation is
formed; see section 2 above.) And when we
‘describe’ what is going on – when we ‘call
the one object, Cause; the other, Effect’ (EHU
7.27 / 74–5) – we similarly project the idea of
necessary connection onto those objects. Our
causal talk and thought, then, is not descrip-
tive: it does not attribute a mind-independent
relation to causes and effects, but projects
our idea of necessity onto them.30
As with the moral and aesthetic cases,
however, this is not to say that Hume is a
subjectivist about causation, for there are
norms that govern the appropriateness of
our causal claims to their objects, for exam-
ple in his ‘rules by which to judge of causes
and effects’ (THN 1.3.15 / 173–6).31 More
generally, Hume is certainly in a position to
regard our natural, instinctive causal judge-
ments as eminently revisable in the light of
the evidence.32 Thus, for example, As might
all have been followed by Bs in my experi-
ence, so that I judge that the As are causes of
Bs. But I might then find out that there are
two distinct kinds of A (call them A1s and
A2s), that all the As I have observed are in
142
9780826443595_Ch05_Final_txt_print.indd 142
fact A1s, and that A2s are often not followed
by Bs. This would give me good grounds to
revise my initial judgement that As cause Bs
in favour of the judgement that A1s cause Bs
but A2s do not, since the ‘difference in the
effects of two resembling objects must pro-
ceed from that particular, in which they dif-
fer’ (THN 1.3.15.8 / 174; rule 6). And this
would be so even if the habit of inference that
has been established naturally inclines me, on
next observing an A, to expect a B, and thus
to judge that they are causally connected
if a B does indeed follow; I know that that
judgement is hostage to information I do not
possess, namely whether or not the observed
A is an A1 or an A2. In other words, the rel-
evant rule acts as a normative constraint on
the causal judgement I am naturally inclined
to make.
Direct textual evidence for the projectiv-
ist interpretation is admittedly rather thin.
Indirect support comes largely from similari-
ties between Hume’s treatment of causation
on the one hand, and moral and aesthetic
judgements on the other. In each case, we
have the thought that our judgement (causal,
moral or aesthetic) ‘adds’ something to mind-
independent matters of fact. In the causal
case, Hume says that the mind does this via
its ‘propensity to spread itself on external
objects’ (THN 1.3.14.23 / 167); in the moral
and aesthetic cases, he postulates ‘a produc-
tive faculty, and gilding or staining all natu-
ral objects with the colours, borrowed from
internal sentiment, raises in a manner a new
creation’ (EPM App. 1.21 / 294).
A further piece of evidence – and at the
same time a response to the charge that
Hume shows no serious positive inclina-
tion towards a non-cognitivist account of
causation – comes from his selective use of
the terms ‘matter of fact’ and ‘belief’. Hume
never talks about causal ‘beliefs’ or considers
2/9/2012 9:35:03 PM
 CAUSATION AND NECESSARY CONNECTION
causal claims to fall within the class of ‘mat-
ters of fact’. We make causal judgements, but
these would appear not to have the status of
belief for Hume, nor would they appear to
be judgements about matters of fact – or at
least, Hume never claims that they are.33 In
other words, Hume appears to restrict ‘mat-
ter of fact’ – and correspondingly, ‘belief’ – to
the relata of causation. From the perspective
of the traditional and sceptical realist inter-
pretations, this is rather puzzling. For in both
interpretations, there are perfectly good facts
about causation to be had, and so our causal
‘judgements’ should count as beliefs, every bit
as much as the existence of a moving billiard
ball is a matter of fact and our expectation
that the billiard ball will move is a belief.
From the perspective of the projectiv-
ist interpretation, by contrast, there is no
real anomaly here. Causal reasoning, Hume
says, is reasoning from one matter of fact to
another. If causal judgements are projections
of the idea to which this reasoning gives rise,
then those judgements are not beliefs about
matters of fact, any more than moral and
aesthetic judgements are beliefs about mat-
ters of fact: they are projections of our habits
of thought onto matters of fact, and so do
not constitute beliefs about matters of fact.34
5. PROBLEMS AND PROSPECTS
For quite a large part of the twentieth cen-
tury, most analytic philosophers steadfastly
avoided appealing to the concept of cause:
a ‘horrid little word’, according to Peter van
Inwagen,35 and ‘a truly obscure’ concept,
according to John Earman.36 In particu-
lar, discussion of the problem of induction
overwhelmingly proceeded as though infer-
ence from the observed to the unobserved
143
9780826443595_Ch05_Final_txt_print.indd 143
is largely independent of beliefs about the
causal structure of reality. (For example ‘cau-
sation’ does not even appear in the index
of Colin Howson’s 2000 book, Hume’s
Problem: Induction and the Justification of
Belief [Oxford: Oxford University Press].)
Hume is, of course, the philosopher from
whom we are supposed to have learned that
causal thinking is both suspect and dispensa-
ble. The irony is that this is not Hume’s view
at all: for Hume, causal thinking is central to
our understanding of, and beliefs about, the
nature of reality. But how are we to under-
stand what ‘causal thinking’ amounts to for
Hume? As we have seen, the range of inter-
pretative possibilities is very wide indeed.
My own view is that the traditional inter-
pretation, in all its forms, is untenable: Hume
holds that causal thinking amounts to more
than belief in regularities, for it involves the
idea of necessary connection. Moreover, that
idea is entirely legitimate, when correctly
understood, and it does serious philosophi-
cal work for him.37 But this leaves both pro-
jectivism and at least one version of sceptical
realism still in the running, and deciding
between these possibilities is a difficult task.
Part of the difficulty lies in the differences
between the Treatise and the first Enquiry.
Reading the first Enquiry as a reworking of
the Treatise, with no substantial change in
the philosophical views presented, inclines
one towards projectivism, correspondingly
encouraging one to reinterpret Hume’s talk
of secret powers and the like in the Enquiry
so that they do not express a commitment to
the existence of real causal powers. Reading
the first Enquiry on its own, by contrast, with
no preconceptions carried over from reading
the Treatise, on the grounds that Hume took
the first Enquiry to best express his consid-
ered philosophical view, inclines one towards
sceptical realism. But of course which strategy
2/9/2012 9:35:03 PM
 CAUSATION AND NECESSARY CONNECTION
one should adopt is largely a question of his-
torical fact that further attention to the texts
themselves will not resolve.38
There is, however, at least one reason to
be sceptical about sceptical realism. Sceptical
realism (or at least Wright’s and Kail’s ver-
sions thereof) inevitably saddles Hume with
a deeply puzzling view. On the one hand, all
sides agree that Hume rejects what Craig
calls the ‘Insight Ideal’: we cannot pen-
etrate into the ‘essence’ of bodies in such
a way as to reveal any features that would
license a priori inferences from causes to
effects. Nonetheless, Wright’s and Kail’s ver-
sions of sceptical realism attribute to Hume
one aspect of the ‘Image of God’ doctrine,
the metaphysical position that underlies
the Insight Ideal, for they both attribute to
Hume ‘an ideal of knowledge of true causes
which was derived from the Cartesians’.39 In
other words, ‘true causes’ are such that we
would, if only we could penetrate into their
nature, be able to infer effects from causes a
priori. And the question is: why would Hume
commit himself to such a metaphysical posi-
tion? After all, the only motivation for such a
view is the thought that the nature of reality
must be such that God himself (and so we, if
we are sufficiently God-like) can infer effects
from causes a priori. And this is a motivation
that Hume himself clearly lacks: he has no
reason whatsoever to want to cling on to a
picture of the nature of reality that derives
from views about God’s epistemic access to
that reality.
NOTES
1
E. Craig, The Mind of God and the Works of
Man (Oxford: Clarendon Press, 1987), chap. 1.
2
R. Descartes, Meditations on First Philosophy
(1641), trans. and ed. J. Cottingham
144
9780826443595_Ch05_Final_txt_print.indd 144
(Cambridge: Cambridge University Press,
1996), p. 21.
3
J.L. Mackie, The Cement of the Universe
(Oxford: Oxford University Press, 1974),
pp. 12–13.
4
In fact, Barry Stroud’s interpretation (which
I class as a version of the traditional inter-
pretation) does not deny (2); in Stroud’s view,
we simply do, inevitably, ‘come to believe,
mistakenly, that there are objective necessary
connections between events’ (B. Stroud, Hume
(London: Routledge, 1977), p. 87). So Stroud
apparently embraces, rather than attempts to
resolve, the tension described above.
5
Stroud, Hume, p. 86.
6
H. Noonan, Hume on Knowledge (London:
Routledge, 1999), p. 142.
7
For more on this, see my Hume on Causation
(Abingdon: Routledge, 2006), sect. 4.3.
8
See, for example, J.A. Robinson, ‘Hume’s
Two Definitions of “Cause”’, Philosophical
Quarterly 12 (1962), pp. 162–71; A. Coventry,
Hume: A Guide for the Perplexed (London:
Continuum, 2007), p. 110; G. Dicker, Hume’s
Epistemology and Metaphysics (London:
Routledge, 1998), p. 115; Noonan, Hume on
Knowledge, p. 151.
9
D. Garrett, Cognition and Commitment in
Hume’s Philosophy (New York: Oxford
University Press, 1997), chap. 5.
10
Garrett, Cognition and Commitment, p. 108.
11
Craig, The Mind of God, p. 108; see also
Garrett, Cognition and Commitment, p. 106.
12
For more on this, see my Hume on Causation,
pp. 94–107.
13
Interpretations that come under this gen-
eral heading include T.L. Beauchamp and
A. Rosenberg, Hume and the Problem of
Causation (New York: Oxford University
Press, 1981); Stroud, Hume; Mackie, The
Cement of the Universe; Garrett, Cognition
and Commitment. There are important differ-
ences between the interpretations offered by
these authors, however, on the troublesome
issue of what to do about the idea of neces-
sary connection; see my Hume on Causation,
chap. 5, for discussion.
14
The major contributions to the sceptical
realist literature have been N. Kemp Smith,
The Philosophy of David Hume (London:
Macmillan, 1941); J.P. Wright, The Sceptical
2/9/2012 9:35:03 PM
 CAUSATION AND NECESSARY CONNECTION
Realism of David Hume (Manchester:
Manchester University Press, 1983); G.
Strawson, The Secret Connexion: Causation,
Realism, and David Hume (New York:
Oxford University Press, 1989); S. Buckle,
Hume’s Enlightenment Tract (Oxford: Oxford
University Press, 2001); and P.J.E. Kail,
Projection and Realism in Hume’s Philosophy
(Oxford: Oxford University Press, 2007).
15
Wright, The Sceptical Realism of David Hume,
p. 147. Kail takes the same line, holding that
the ‘causal necessity of which we may be igno-
rant can be understood as that which, were we
to be acquainted with it, would yield a priori
inference and render it impossible to conceive
cause without effect’ (Kail, Projection and
Realism, p. 103).
16
Quoted in Wright, The Sceptical Realism of
David Hume, p. 140.
17
Strawson, The Secret Connexion, p. 157.
18
Ibid., p. 126.
19
See A.J. Jacobson, ‘From Cognitive Science
to a Post-Cartesian Text: What Did Hume
Really Say?’, in R. Read and K. Richman (eds),
The New Hume Debate (London: Routledge,
2000), pp. 156–66.
20
See K. Winkler, ‘The New Hume’, in R. Read
and K. Richman (eds), The New Hume Debate,
pp. 52–87, and S. Blackburn, ‘Hume and Thick
Connexions’, in R. Read and K. Richman
(eds), The New Hume Debate, pp. 100–12.
See my Hume on Causation, pp. 180–93 for
further discussion of both options, and Kail,
Projection and Realism, pp. 118–21, for more
critical assessment.
21
Wright, The Sceptical Realism of David Hume,
p. 95.
22
Strawson, The Secret Connexion, pp. 132–3.
23
Ibid., p. 87.
24
The projectivist interpretation is suggested by
some remarks of Simon Blackburn’s; see, for
example, S. Blackburn, Spreading the Word
(Oxford: Oxford University Press, 1984),
pp. 210–12 and ‘Hume and Thick Connexions’,
pp. 107–11. It is given a fuller treatment
in my Hume on Causation, chap. 6, and in
A. Coventry, Hume’s Theory of Causation:
A Quasi-Realist Interpretation (London:
145
9780826443595_Ch05_Final_txt_print.indd 145
Continuum, 2006). For a general discussion of
the notion of ‘projection’ see Kail, Projection
and Realism, chap. 1, and for a discussion
of Hume’s (allegedly problematic) attitude
towards ‘projection’, see B. Stroud, ‘“Gilding
or Staining” the World with “Sentiments” and
“Phantasms”’, in R. Read and K. Richman
(eds), The New Hume Debate, pp. 16–30.
25
Coventry, Hume’s Theory of Causation.
26
M. Smith, ‘Objectivity and Moral Realism:
On the Significance of the Phenomenology
of Moral Experience’, in J. Haldane and
C. Wright (eds), Reality, Representation, and
Projection (New York: Oxford University
Press, 1993), p. 246.
27
D. Hume, ‘Of the Standard of Taste’, E 230.
28
Ibid.
29
See Coventry, Hume’s Theory of Causation,
pp. 117–33, for further discussion.
30
See, however, P.J.E. Kail, ‘Book Review: Helen
Beebee, Hume on Causation’, Mind 117
(2008), pp. 453–4, for a worry about the use
of the notion of projection here.
31
See Coventry, Hume’s Theory of Causation,
pp. 133–7.
32
See my Hume on Causation, pp. 160–7.
33
The same is generally true of the moral and
aesthetic cases, although, as Coventry notes,
in ‘Of the Standard of Taste’, Hume twice
appears to count matters for which there is a
‘standard’ (as in ‘standard of taste’) as ‘mat-
ters of fact’; see Coventry, Hume’s Theory of
Causation, p. 119, and Hume, E 230, 242.
34
See my Hume on Causation, pp. 152–4, for
more discussion in the context of Hume’s
distinction between reason and taste.
35
P. Van Inwagen, An Essay on Free Will
(Oxford: Clarendon Press, 1983), p. 65.
36
J. Earman, A Primer on Determinism
(Dordrecht: D. Reidel, 1986), p. 5.
37
See my Hume on Causation, sect. 6.5.
38
For a defence of the latter strategy, see G.
Strawson, ‘David Hume: Objects and Power’,
in R. Read and K. Richman (eds), The New
Hume Debate, pp. 31–51. For brief discussion,
see my Hume on Causation, pp. 221–5.
39
Wright, The Sceptical Realism of David Hume,
p. 147.
2/9/2012 9:35:03 PM
 6
HUME ON SCEPTICISM AND
THE MORAL SCIENCES
Alan Bailey
1. HUME’S RADICAL SCEPTICISM
The title of the concluding part of Book One of
Hume’s Treatise is ‘Of the sceptical and other
systems of philosophy’, and Hume wastes no
time in setting before the reader two scepti-
cal arguments that attack, when taken in tan-
dem, the justified status of the overwhelming
majority of our beliefs. In the opening section,
‘Of scepticism with regard to reason’, the
argument outlined by Hume is presented as
targeting all beliefs that are the product of any
kind of inference. According to this argument,
both demonstrative and probabilistic infer-
ences are ultimately unsuccessful in giving
their conclusions any positive degree of justi-
fication (see THN 1.4.1.1–6 / 180–3).
The attempt to found beliefs about mind-
independent physical objects on the senses
rather than some form of inference then
comes under attack in the very next sec-
tion of the Treatise. In ‘Of scepticism with
regard to the senses’, Hume rapidly dismisses
the suggestion that the senses are capable
by themselves of revealing that some of
the objects seemingly encountered by us in
perception continue to exist even after they
have ceased to appear to the senses: any such
supposition is ‘a contradiction in terms, and
146
9780826443595_Ch06_Final_txt_print.indd 146
supposes that the senses continue to oper-
ate, even after they have ceas’d all manner of
operation’ (THN 1.4.2.3 / 188). And he fol-
lows this with a repudiation of the supposi-
tion that the senses are capable of persuading
us that some of the things of which we are
immediately aware in perception have the
ability to exist independently of perception
or are representations of entities of another
kind that do possess such a power (see THN
1.4.2.4–14 / 189–93).
It seems, therefore, that if we take Hume
as endorsing the arguments that he has set
out in such detail in these key sections of
the Treatise, then we must interpret him as
accepting that no claim about mind-inde-
pendent objects and no claim that requires
support from any form of inference is ever
any better justified than the logical contra-
dictory of that claim. And as Hume appears
to hold that the inner and outer senses are
unable, without some form of inferential
supplementation, to tell us anything about
events that are not happening to us at the
present moment, this implies that Hume
holds that only beliefs about very sim-
ple necessary truths that do not need to be
grasped through a process of inference and
beliefs about the content of our present ideas
2/9/2012 9:35:12 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
and impressions are potentially capable of
being rationally justified beliefs.
It will be convenient from this point
onwards to refer to these few beliefs that
Hume is seemingly prepared to accept as jus-
tifiable as H-minimal beliefs. Now it seems
evident that an inquirer who has engaged
in substantial epistemological reflection but
nevertheless fails to accept that any beliefs
other than H-minimal beliefs are rationally
justified beliefs is someone who has embraced
a radically sceptical posture. It appears,
therefore, that we can legitimately reject the
conclusion that Hume is himself a radical
sceptic only if we can uncover substantial
grounds for supposing that Hume does not
genuinely endorse the arguments that he has
chosen to elaborate in ‘Of scepticism with
regard to reason’ and ‘Of scepticism with
regard to the senses’. There must be a pre-
sumption, however, that an author who puts
lengthy arguments before the reader without
explicitly distancing himself from those argu-
ments does endorse their conclusions. And it
is difficult to see Hume’s comments in the
paragraph that brings these two sections to
an end as anything other than an attempt to
make it clear that he stands firmly behind his
negative epistemological arguments:
’Tis impossible upon any system to
defend either our understanding or our
senses; and we but expose them farther
when we endeavour to justify them
in that manner. As the sceptical doubt
arises naturally from a profound and
intense reflection on those subjects, it
always encreases, the farther we carry
our reflections, whether in opposition or
conformity to it. (THN 1.4.2.57 / 218)
It does have to be conceded that there are
other passages in the Treatise that initially
seem to indicate that Hume has a much less
147
9780826443595_Ch06_Final_txt_print.indd 147
positive view of sceptical argumentation.
Immediately after Hume has set out the argu-
ment that no process of inference succeeds in
conferring justification on its conclusions, he
specifically addresses the question of whether
he sincerely gives his assent to this argument
and whether he is ‘really one of those scep-
tics, who hold that all is uncertain, and that
our judgement is not in any thing possest of
any measures of truth and falshood’ (THN
1.4.1.7 / 183). And his answer is one that
could easily be interpreted as a repudiation of
the argument that he has just constructed:
I shou’d reply, that this question is
entirely superfluous, and that neither I,
nor any other person was ever sincerely
and constantly of that opinion. Nature,
by an absolute and uncontroulable neces-
sity has determin’d us to judge as well
as to breathe and feel; nor can we any
more forbear viewing certain objects in
a stronger and fuller light, upon account
of their customary connexion with a pre-
sent impression, than we can hinder our-
selves from thinking as long as we are
awake, or seeing the surrounding bodies,
when we turn our eyes towards them in
broad sun-shine. (ibid.)
Commentators who doubt the propri-
ety of classifying Hume as a radical sceptic
also tend to seize on his explanation of his
purposes in presenting his master argument
against inferential justification. He indicates
that this argument has been included in the
Treatise as a way of adding credibility to his
preferred account of belief as something that
arises when ideas acquire additional force
and vivacity as a result of their associative
links with impressions:
My intention then in displaying so care-
fully the arguments of that fantastic sect
2/9/2012 9:35:12 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
[the sceptics], is only to make the reader
sensible of the truth of my hypothesis,
that all our reasonings concerning causes
and effects are deriv’d from nothing but
custom; and that belief is more properly
an act of the sensitive, than of the cogi-
tative part of our natures. (THN 1.4.1.8
/ 183)
In similar fashion, Hume’s discussion
of our belief in the existence of mind-inde-
pendent physical objects sees him surround-
ing his presentation of the relevant sceptical
arguments with observations that generate
significant perplexity about his personal atti-
tude towards such argumentation. Thus we
find that the paragraph that opens ‘Of scep-
ticism with regard to the senses’ includes the
following declamation:
We may well ask, What causes induce us
to believe in the existence of body? but
’tis in vain to ask, Whether there be body
or not? That is a point, which we must
take for granted in all our reasonings.
(THN 1.4.2.1 / 187)
It might be suggested, accordingly, that if we
must indeed hold as a presupposition of all
our reasoning that mind-independent bodies
exist, this precludes anyone from sincerely
endorsing sceptical arguments on this topic.
If we are genuinely committed to the conclu-
sion that mind-independent bodies exist, how
can we simultaneously attach any weight to
arguments that purport to call that conclu-
sion into question?
It is apparent, therefore, that the manner
in which Hume expounds his philosophical
position in the Treatise does place some dif-
ficulties in the way of the conclusion that
this work incorporates sceptical arguments
that have genuinely persuaded him that only
148
9780826443595_Ch06_Final_txt_print.indd 148
H-minimal beliefs ever possess any positive
degree of rational justification. But when we
look more closely at his explicit reservations
about sceptical argumentation, we discover
that he seems intent on distancing himself
only from the supposition that such argumen-
tation is capable of radically reshaping our
non-epistemic beliefs. He is keen to empha-
size that no amount of reflection on sceptical
arguments can permanently dislodge every-
day beliefs about the existence and proper-
ties of such things as trees, chairs and books.
This does not mean, however, that he holds
that sceptical arguments permit us to regard
such beliefs as rationally justified.
As we have seen, Hume is sometimes pre-
pared to insinuate that it would be appro-
priate to regard him as sincerely assenting to
the argument that no beliefs based on infer-
ence are ever rationally justified only if that
argument were to cause him to abandon all
inferential beliefs. Thus we find that Hume
presents our supposed inability to ‘forbear
viewing certain objects in a stronger and
fuller light, upon account of their custom-
ary connexion with a present impression’
as sufficient of itself to guarantee that it is
impossible for anyone to give their sincere
assent to this argument (THN 1.4.1.7 / 183).
It would normally be supposed, though, that
a distinction can be drawn between sincere
assent to the conclusion that a particular
belief is not a rationally justified belief and
the actual abandonment of that belief. Many
people suffering from acrophobia, for exam-
ple, sincerely accept that at least some of
their fears about being in high places are not
rationally justified fears. Nevertheless this
does not prevent them from continuing to be
afflicted by those fears. It seems, accordingly,
that Hume’s insistence that neither he nor
anyone else is capable of eschewing all infer-
ential beliefs in response to the argument set
2/9/2012 9:35:12 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
out in ‘Of scepticism with regard to reason’ is
entirely compatible with the supposition that
he sincerely assents to that argument in the
sense that he endorses as true its conclusions
about the limitations on our ability to arrive
at epistemically justified beliefs.
This latter interpretation of Hume’s pos-
ition gains additional credibility from other
remarks that he makes about the argument
against reason: these remarks are very dif-
ficult to explain if we are not prepared to
regard him as holding that only someone
who has already embraced the view that
scarcely any of our beliefs are rationally jus-
tified is in a position to deny, in good faith,
that this argument provides rationally com-
pelling support for just that assessment of
our epistemic situation. Hume boldly claims,
for instance, that the argument against rea-
son proves that the principles that lead us
to correct our judgements on any subject
through ‘the consideration of our genius and
capacity, and of the situation of our mind,
when we examin’d that subject’ would, if we
persisted in following them, reduce all puta-
tive evidence to nothing and ‘utterly subvert
all belief and opinion’ (THN 1.4.1.8 / 184).
He also explicitly bases his affirmation that
we can safely conclude that ‘reasoning and
belief is some sensation or peculiar manner
of conception, which ’tis impossible for mere
ideas and reflections to destroy’ (ibid.) on his
confidence that no one will be able to find
any error in the arguments contained in ‘Of
scepticism with regard to reason’. It seems
clear, therefore, that when we consider the
foregoing pieces of corroborative evidence
in combination with Hume’s comments in
the final section of Book One of the Treatise
that he has shown, by means of the argu-
ment against reason, that the understanding
entirely undermines itself when it operates
alone and that we are left with no choice ‘but
149
9780826443595_Ch06_Final_txt_print.indd 149
betwixt a false reason and none at all’ (THN
1.4.7.7 / 268), we have compelling grounds
for asserting that this argument’s conclusion
about the extremely limited availability of
rationally justified belief does indeed enjoy
Hume’s full support.
This same pattern is also discernible
within Hume’s discussion of the status of
our beliefs about the existence of mind-in-
dependent physical objects. It is indeed true
that he emphatically repudiates the suppos-
ition that any of his arguments could bring
about more than a fleeting alienation from
those beliefs (see THN 1.4.2.57 / 218).
Nevertheless it seems plain that he regards
the sceptical attack on their supposed
rational credentials as completely unanswer-
able. Instead of diagnosing this sceptical
critique as relying on some fallacious infer-
ential step or a premise that is either false or
readily deniable, Hume affirms that sceptical
worries about the justification of the suppos-
ition that mind-independent physical objects
exist are irrefutable: ‘This sceptical doubt,
both with regard to reason and the senses, is
a malady, which can never be radically cur’d,
but must return upon us every moment, how-
ever we may chace it away, and sometimes
may seem entirely free from it’ (ibid.). From
Hume’s perspective, no process of reflect-
ive thought can place that supposition on a
justified basis, and the intense examination
of sceptical arguments, even if undertaken
with the intention of refuting them, simply
makes their irrefutability more salient and
potentially strengthens their psychological
impact to such an extent that they can fleet-
ingly alienate us from our common-sense
belief in physical objects. Hume accordingly
rejects the project of constructing a philo-
sophical answer to scepticism in favour of
delineating a way of arriving at a psycho-
logical accommodation with scepticism that
2/9/2012 9:35:13 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
serves to minimize its potentially disrup-
tive effects. As philosophical inquiry cannot
locate any deficiencies in the arguments of
the sceptics that are not shared by all other
forms of reasoning, Hume suggests that the
best way of responding to these epistemo-
logical difficulties is to relax our intellectual
efforts: ‘Carelessness and in-attention alone
can afford us any remedy’ (ibid.).
It seems, therefore, that the most plaus-
ible interpretation of the explicit discussions
of sceptical argumentation in the Treatise
is that Hume combines the conviction that
these arguments make it impossible for him,
or indeed any inquirer who has fully under-
stood their implications, to maintain in
good faith that beliefs other than H-minimal
beliefs are ever rationally justified with
the judgement that arriving at this conclu-
sion has little psychological impact on our
everyday, non-epistemic beliefs. However,
the credibility of this assessment of Hume’s
stance also depends on the extent to which it
fits with his presentation of his views in the
Enquiry concerning Human Understanding.
And we find, in fact, that this work displays
numerous elements that prevent us from
regarding it as less sceptical than Book One
of the Treatise.
Several arguments deployed in the Treatise
are explicitly acknowledged by Hume as scep-
tical arguments only when they appear in the
Enquiry concerning Human Understanding.
Consider, for example, the revisions he makes
to the presentation of his analysis of causal
reasoning. Not only does he give the initial
stages of this analysis the forthright title
‘Sceptical Doubts concerning the Operations
of the Understanding’, but we also find that
the immediately following section begins
with a discussion of ‘the academical or
sceptical philosophy’ in which he says that
such a philosophy is harmless and innocent
150
9780826443595_Ch06_Final_txt_print.indd 150
in almost every instance and attempts to
reassure us that our natural instincts are suf-
ficiently powerful to ensure that it can never
‘undermine the reasonings of common life,
and carry its doubts so far as to destroy all
action, as well as speculation’ (EHU 5.1–2 /
41). Moreover, the piece of abstract reason-
ing Hume displays as an example of the kind
of argument held in check by these natural
instincts is nothing other than a summary of
the argument offered in Section 4 (See EHU
5.2 / 41). And even more striking is the man-
ner in which Hume shapes the discussion of
scepticism in the final section of the Enquiry.
In Part 2 of Section 12 he uses the terms
‘Pyrrhonism’ and ‘excessive scepticism’ to
refer to a form of scepticism that attempts
to ‘destroy reason by argument and rati-
ocination’ (EHU 12.17 / 155).1 According
to him, this form of scepticism purports ‘to
find objections, both to our abstract reason-
ings, and to those which regard matter of
fact and existence’ (EHU 12.17 / 156); and
when he comes to describe the most force-
ful of these sceptical objections to reasonings
concerning matters of fact, the argument
he ascribes to the Pyrrhonist is essentially a
reprise of Hume’s own negative claims about
the grounding of causal inferences.
In addition to the foregoing greater will-
ingness to identify his own arguments as
sceptical arguments, further evidence within
the Enquiry of Hume’s commitment to some
form of radical scepticism is furnished by
his discussion of the relationship between
Pyrrhonean and Academic scepticism. In the
final section of the Enquiry Hume seems to
imply that the only significant criticism that
can be levelled against the arguments used
by Pyrrhonean sceptics is that they lack the
power to dislodge our non-epistemic beliefs
(see EHU 12.23 / 159–60). And he indicates
that the causal resultant of natural instinct
2/9/2012 9:35:13 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
and Pyrrhonean scruples about rational jus-
tification is an enduring and advantageous
disposition to form one’s non-epistemic
beliefs in a cautious and undogmatic man-
ner. According to Hume, the most conveni-
ent way of acquiring this disposition is ‘to be
once thoroughly convinced of the force of
the pyrrhonian doubt, and of the impossi-
bility, that any thing, but the strong power of
natural instinct, could free us from it’ (EHU
12.25 / 162). Thus mitigated scepticism or
the Academical philosophy, which is strongly
recommended to the reader in both Sections
5 and 12 of the Enquiry, is presented not
as the outcome of recognizing flaws in the
arguments of radical sceptics but rather as
the outcome of our natural belief-forming
mechanisms causally interacting with the
realization that sceptical arguments can
only be repudiated in good faith by some-
one who already accepts that no beliefs other
than H-minimal beliefs are ever rationally
justified.2
Ultimately, then, the considerations in
favour of interpreting Hume as committed
to a radically sceptical stance in both the
Treatise and the Enquiry concerning Human
Understanding seem very strong indeed.
Prominent arguments in Book One of the
Treatise terminate in the conclusion that
no beliefs other than H-minimal beliefs are
rationally justified; and once we have given
due weight to the distinction between seeing
a belief as devoid of rational justification and
actually abandoning that belief, there seems
to be little explicit basis in the text for sup-
posing that Hume does not endorse that con-
clusion. When, therefore, we find a similarly
radical array of sceptical arguments deployed
in the Enquiry alongside seemingly robust
affirmations of their strength and irrefutabil-
ity (see EHU 12.14 / 153–4, 12.22 / 159, and
12.25 / 162), it surely ceases to be plausible
151
9780826443595_Ch06_Final_txt_print.indd 151
to hold that Hume does not embrace the
sceptical view that very few of our beliefs are
ever rationally justified.
2. THE TENSION BETWEEN
SCEPTICISM AND ABSTRUSE INQUIRY
As we have just seen, discounting Hume’s
commitment to a radically sceptical perspec-
tive in respect of the limited availability of
rationally justified belief does not appear to
be a viable interpretative strategy. However,
if we accept that reflection on sceptical argu-
ments has led Hume to the conclusion that
only H-minimal beliefs can ever qualify as
rationally justified beliefs, there does seem to
be a real danger that Hume’s overall intellec-
tual posture ceases to be a coherent one.
The subtitle of the Treatise is Being an
Attempt to Introduce the Experimental
Method of Reasoning into Moral Subjects.
And in the introduction to the Treatise, Hume
presents this work as primarily an attempt to
explain the best way of making progress in
the science of man.
Hume maintains that the science and the
philosophy of his era are replete with systems
and theories that display such undesirable
qualities as ‘principles taken upon trust, con-
sequences lamely deduc’d from them, want
of coherence in the parts, and of evidence in
the whole’ (THN Intro. 1 / xiii). These glar-
ing deficiencies ‘seem to have drawn disgrace
upon philosophy itself’ (ibid.), and they have
also given rise to what Hume describes as
‘that common prejudice against metaphysical
reasonings of all kinds, even amongst those,
who profess themselves scholars, and have a
just value for every other part of literature’
(THN Intro. 2 / xiv). As he explains, this
prejudice against metaphysical reasonings is
2/9/2012 9:35:13 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
not simply a prejudice against a particular
branch of inquiry but has come to constitute
a general aversion to ‘every kind of argument,
which is in any way abstruse, and requires
some attention to be comprehended’ (ibid.).
Hume’s declared ambition is to alleviate
this prejudice by helping to put science and
philosophy on a sound basis. His proposed
way of doing this involves applying the
methods of the natural sciences, as exempli-
fied within the most compelling theories that
have arisen within that realm of inquiry, to
the study of human nature. Hume holds that
once we have a better understanding of the
mental activities of human beings, this will
shed fresh light on a wide variety of intellec-
tual disciplines:
’Tis evident, that all the sciences have
a relation, greater or less, to human
nature; and that however wide any of
them may seem to run from it, they still
return back by one passage or another.
Even Mathematics, Natural Philosophy,3
and Natural Religion, are in some meas-
ure dependent on the science of man;
since they lie under the cognizance of
men, and are judg’d of by their powers
and faculties. (THN Intro. 4 / xv)
Of course natural philosophy is the arena in
which Hume has found examples of convin-
cing theories built on a credible investigative
methodology, an investigative methodology
he has deliberately extended to the moral sci-
ences. His strategy, therefore, involves making
use of a virtuous cycle of mutual improve-
ment. Such exemplary theories within the
natural sciences as ‘newton’s explication of
the wonderful phenomenon of the rainbow’
(DNR 1.136) and ‘the copernican system’
(ibid. 138) serve to validate the experimen-
tal and experience-based investigative meth-
odology that Hume regards as having given
152
9780826443595_Ch06_Final_txt_print.indd 152
rise to those theories. This methodology is,
in turn, something that can beneficially be
applied to the study of human nature and in
particular to the study of how human beings
make inferences and form judgements. And
with the assistance of the ensuing improved
account of human psychology, we will be
able to develop a still better methodology of
inquiry for use not only within the natural sci-
ences themselves but also within other intel-
lectual disciplines that lack acknowledged
achievements on a par with the theories of
Newton, Galileo and Copernicus.
Significantly, Hume is adamant that the
progress in the natural sciences that has
finally given us some theories that are both
intellectually satisfying and stable in the face
of examination is the product of the experi-
mental method of investigation. He specific-
ally places his project in the Treatise in the
context of an intellectual movement, exempli-
fied in his eyes by ‘some late philosophers in
England, who have begun to put the science
of man on a new footing, and have engag’d
the attention, and excited the curiosity of the
public’ (THN Intro. 7 / xvii), that is attempt-
ing to transfer the application of experi-
mental philosophy from natural to moral
subjects. Hume maintains that this attempt
forces us to acknowledge that the science of
man can only be built on a foundation ‘laid
on experience and observation’ (THN Intro.
7 / xvi). If we construct the science of man
in this manner, then our conclusions about
human psychology will assist us in refining
and improving our investigative techniques
across a broad range of subjects. But if we
abandon the guidance of experience, then we
will be led to conclusions that ‘ought, at first,
to be rejected as presumptuous and chimer-
ical’ (THN Intro. 8 / xvii).
These forthright scientific and philosoph-
ical aspirations, however, raise the question
2/9/2012 9:35:13 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
of how they cohere with Hume’s radical scep-
ticism. Given their prominence at the start of
the Treatise, how does it come about that
Book One seems to terminate in a reaffirm-
ation of sceptical arguments indicating that
only beliefs with H-minimal content are ever
rationally justified beliefs? No science can
be constructed that consists exclusively of
beliefs with such attenuated content. Should
we conclude, therefore, that Hume’s positive
project has catastrophically imploded in the
course of writing Book One and that he has
unintentionally found himself immersed in
a sceptical quagmire from which there is no
principled escape?
The supposition that Hume sees the out-
come of Book One of the Treatise as incon-
sistent with or even tangential to his initial
ambitions to put the study of human nature
on a sounder footing is difficult to reconcile
with the untroubled opening of Book Two.
In Book Two he presents an account of the
operations of the human passions that seems
perfectly in accord with his professed inten-
tions to make progress in the moral sciences,
and there is no indication whatsoever that
he regards this account as undermined or in
any way rendered problematic by the scepti-
cal reflections developed in the final stages of
Book One. Similarly, his discussion of moral-
ity in Book Three shows no signs of being
infected by destabilizing sceptical worries.
It is true that in the opening paragraph he
ventures some observations about the lim-
ited persuasive force of abstruse reasoning,
but the overwhelming tenor of his discussion
is one of quiet confidence in the satisfactory
nature of his arguments and conclusions:
I am not, however, without hopes, that
the present system of philosophy will
acquire new force as it advances; and
that our reasonings concerning morals
153
9780826443595_Ch06_Final_txt_print.indd 153
will corroborate whatever has been said
concerning the understanding and the
passions. (THN 3.1.1.1 / 455)
We also need to keep in mind that if Hume
had seen the sceptical crisis of Book One as
an unwelcome or unexpected interruption in
the unfolding of his overall line of argument
in the Treatise, then he had ample oppor-
tunity to omit it from the version of the text
sent for publication or to include additional
guidance to the reader about how to place
this element of his discussion within his posi-
tive aims. His letters confirm that he had no
qualms about deleting some of the more obvi-
ously irreligious material from the text before
finally placing it in the hands of his publisher
(see, in particular, LDH 1.25, 6). Thus the
prominent role ultimately allocated in the
Treatise to the presentation of sceptical lines
of thought presumably reflects Hume’s con-
sidered judgement that the sceptical aspects
of his thinking and the attempt to build the
science of human nature on a sound founda-
tion form part of one coherent and mutually
supportive intellectual enterprise.
Once again Hume’s revised presentation of
his views in the Enquiry concerning Human
Understanding provides important corrobor-
ation of what is implied by the text of the
Treatise. In Section 1 of the Enquiry Hume
develops a defence of ‘profound reasonings,
or what is commonly called metaphysics’
and abstract philosophy (EHU 1.7 / 9) that
presents it as providing us with, amongst
other benefits, an accurate delineation of ‘the
powers and faculties of human nature’ (EHU
1.13 / 13). Hume even suggests that the ana-
logy with the development of Newtonian
mechanics means that we can legitimately
hope that such profound and abstruse rea-
sonings will ultimately serve to ‘discover, at
least in some degree, the secret springs and
2/9/2012 9:35:13 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
principles, by which the human mind is actu-
ated in its operations’ (EHU 1.15 / 14). We
are forced to conclude, therefore, that Hume’s
intention to assist in the construction of a sci-
ence of human nature remained intact at the
time of writing the Enquiry despite the effort
lavished in the Treatise on the elaboration of
sceptical lines of argument. Yet the Enquiry
also sees Hume deploying another battery of
sceptical arguments that are scarcely any less
radical than those found in the Treatise. In
some ways, indeed, the Enquiry even seems
to place a greater emphasis than the Treatise
on the force and tenaciousness of sceptical
argumentation. It appears, therefore, that the
Enquiry concerning Human Understanding
replicates, without any obvious sign of
unease or tension on Hume’s part, the same
blend of philosophico-scientific ambitions
and sceptical conclusions about the limited
availability of rationally justified beliefs that
can be discerned in the Treatise.
Once we accept, however, that Hume
regards radical scepticism and the attempt
to construct a science of man as complemen-
tary activities, it is natural to wonder what
kind of science can exist in a context where
all the core beliefs of this putative science are
viewed by its own adherents as totally devoid
of rational justification. Moreover, if the
most profound and advanced forms of scien-
tific thinking are as compatible with radical
scepticism about justification as Hume seems
to think, what does this say about the value
of engaging in sceptical reflection?
In the final section of the Enquiry concern-
ing Human Understanding, Hume presents
what he plainly regards as a potentially dev-
astating criticism of radical scepticism:
For here is the chief and most confound-
ing objection to excessive scepticism, that
no durable good can ever result from it;
154
9780826443595_Ch06_Final_txt_print.indd 154
while it remains in its full force and vig-
our. We need only ask such a sceptic,
What his meaning is? And what he pro-
poses by all these curious researches? He
is immediately at a loss, and knows not
what to answer. (EHU 12.23 / 159–60)
Hume explains this objection in terms of the
inability of sceptical argumentation to prod-
uce stable changes in people’s beliefs and
behaviour analogous to those that can arise
from exposure to the Copernican system of
astronomy or the moral exhortations of the
Stoics and Epicureans. According to Hume,
the radical or Pyrrhonean sceptic ‘cannot
expect, that his philosophy will have any
constant influence on the mind: Or if it had,
that its influence would be beneficial to soci-
ety’ (EHU 12.23 / 160). It seems, therefore,
that Hume is suggesting that we are entitled
to dismiss reflection on sceptical reasoning as
pointless or, at best, a frivolous and esoteric
form of amusement unless there is some indi-
cation that it can alter a person’s set of beliefs
in stable and useful ways.
Somewhat ironically Hume’s own attempt
to emphasize the transient nature of the dox-
astic changes induced by sceptical arguments
actually terminates by drawing attention to
one stable change to which these arguments
can give rise. He insists that despite the initial
impact of the sceptic’s arguments, it is evident
that ‘the first and most trivial event in life
will put to flight all his doubts and scruples’
(ibid.). Thus he describes the sceptic as awak-
ening from a dream and laughing at himself
for being swayed by these abstruse episte-
mological objections. Nevertheless this reim-
mersion in quotidian life does not entirely
negate the influence of sceptical reasoning.
The sceptic may indeed be led to describe
his arguments as nothing more than amusing
puzzles, but Hume also represents the sceptic
2/9/2012 9:35:14 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
as retaining a distinctive perspective on the
rationality of his and our beliefs. Even when
the sceptical fugue that can be generated
by prolonged reflection on epistemological
arguments passes, the sceptic retains a robust
sense of an aspect of the human predicament
that escapes the attention of more superficial
inquirers:
the whimsical condition of mankind, who
must act and reason and believe; though
they are not able, by their most diligent
enquiry, to satisfy themselves concerning
the foundation of these operations, or
to remove the objections, which may be
raised against them. (ibid.)
Given that Hume draws our attention
to this enduring consequence of the careful
study of sceptical reasonings, why does he
also make the seemingly contradictory claim
that radical scepticism fails to exercise any
constant influence on the mind? The most
plausible answer here is that he is thinking
in terms of influence over our everyday non-
epistemic beliefs. The radical sceptic might
well acquire a settled disposition to deny that
any beliefs other than H-minimal beliefs are
ever rationally justified beliefs, but this does
not lead to suspension of belief on the exist-
ence of cattle, chairs and chimneys. Indeed
the radical sceptic’s expectations about what
will happen, unlike his beliefs about what
we are justified in thinking will happen,
tend to be remarkably similar to our own.
Thus the objection to radical scepticism that
Hume emphasizes in the Enquiry concern-
ing Human Understanding seems to amount
to the contention that if the argumentative
case assembled so laboriously by sceptical
thinkers has no or almost no impact on
our non-epistemic beliefs, then we have no
motive for expending the effort required to
155
9780826443595_Ch06_Final_txt_print.indd 155
follow such reasoning. Some rather unusual
people might find that the study of sceptical
argumentation serves them as a mildly enter-
taining pastime, but Cleanthes undoubtedly
speaks for most of us when he expresses
a preference in the Dialogues concern-
ing Natural Religion for less demanding
and cerebral forms of amusement: ‘But for
my part, whenever I find myself disposed
to mirth and amusement, I shall certainly
choose my entertainment of a less perplex-
ing and abstruse nature. A comedy, a novel,
or at most a history, seems a more natural
recreation than such metaphysical subtilties
and abstractions.’ (DNR 1.137)
It might be suggested at this point that the
sceptic can mount an effective response to
Hume’s objection by drawing a distinction
between the non-epistemic beliefs that are
an integral part of everyday life and the non-
epistemic beliefs generated by such activities
as philosophy and theoretical science. Even if
these everyday beliefs cannot be disturbed or
changed by sceptical argumentation, perhaps
the arguments deployed by radical sceptics
can reshape our philosophical and scien-
tific commitments. Suspension of judgement
at the everyday level is certainly not some-
thing that flows from reflection on sceptical
reasonings, but that does not guarantee that
the more abstract and sophisticated non-
epistemic beliefs found within philosophy
and the sciences are similarly invulnerable to
sceptical attack.
This indeed is Philo’s initial response to
Cleanthes’ charge that his professed scepti-
cism is an insincere affectation. Cleanthes
challenges Philo to exhibit his scepticism in
his behaviour:
Whether your scepticism be as absolute
and sincere as you pretend, we shall
learn bye and bye, when the company
2/9/2012 9:35:14 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
breaks up: We shall then see, whether
you go out at the door or the window;
and whether you really doubt, if your
body has gravity, or can be injured
by its fall; according to popular opin-
ion, derived from our fallacious senses
and more fallacious experience. (DNR
1.132)
In this crude form Cleanthes’ attack is easily
parried by Philo. Philo simply appeals to the
involuntary nature of these everyday beliefs.
The sceptic sincerely and genuinely holds
that even straightforward and seemingly
uncontroversial everyday beliefs are entirely
devoid of rational justification. However,
this conviction fails to dislodge these beliefs
because they are held in place by psycho-
logical forces that are causally impervious to
any influence that can be exerted by sceptical
arguments: ‘To whatever length any one may
push his speculative principles of scepticism,
he must act, I own, and live, and converse
like other men; and for this conduct he is
not obliged to give any other reason than the
absolute necessity he lies under of so doing.’
(DNR 1.134)
Philo also wishes, however, to leave room
for radical scepticism to affect at least some
of our non-epistemic beliefs in a potentially
beneficial way. Consequently he suggests that
abstruse theoretical beliefs can be subverted
by sceptical arguments:
But it is evident, whenever our argu-
ments . . . run wide of common life, that
the most refined scepticism comes to be
upon a footing with them, and is able
to oppose and counterbalance them.
The one has no more weight than the
other. The mind must remain in suspense
between them; and it is that very sus-
pense or balance, which is the triumph
of scepticism. (DNR 1.135–6)
156
9780826443595_Ch06_Final_txt_print.indd 156
This second component of Philo’s defence
of scepticism could be construed as overlook-
ing the possibility that refined arguments
that run wide of common life might some-
times be arguments that should be allowed
to influence us. However, it would be more
charitable to interpret Philo as attempting
to draw a distinction between complicated
and intricate arguments that are neverthe-
less rooted in the same underlying principles
as those that generate stable conviction in
ordinary life and even more ambitious and
speculative arguments that disregard those
common-sense principles. Thus we find that
Philo is happy to endorse a form of theoriz-
ing in both the natural and moral sciences
that keeps closely to patterns of inference
and evidence assessment that exert a major
influence on our quotidian beliefs:
He considers besides, that every one,
even in common life, is constrained to
have more or less of this philosophy; that
from our earliest infancy we make con-
tinual advances in forming more general
principles of conduct and reasoning; that
the larger experience we acquire, and the
stronger reason we are endowed with,
we always render our principles the
more general and comprehensive; and
that what we call philosophy is nothing
but a more regular and methodical oper-
ation of the same kind. (DNR 1.134)
It seems, therefore, that Philo’s response to
the question of the practical value of reflec-
tion on sceptical arguments is to claim that
such reflection is useful because it serves to
subvert and sweep away speculative beliefs
that are not rooted in the core psychological
mechanisms that generate our enduring every-
day beliefs. At the same time this reflection is
not dangerous because those everyday beliefs
and even the more sophisticated beliefs that
2/9/2012 9:35:14 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
arise from the methodical and systematic
application of common-sense principles of
inference are founded in permanent features
of human nature that cannot be overturned
by any form of abstract reasoning.
Is it really the case, however, that we can
satisfactorily combine the view that radical
scepticism is intellectually safe with the sup-
position that it is genuinely useful? In the
Dialogues Cleanthes returns to the attack by
alleging that self-professed sceptics invari-
ably seem to conduct even their theoretical
inquiries in the same manner as people who
do not regard themselves as sceptics:
But I observe . . . with regard to you,
philo, and all speculative sceptics, that
your doctrine and practice are as much
at variance in the most abstruse points
of theory as in the conduct of common
life. Wherever evidence discovers itself,
you adhere to it, notwithstanding your
pretended scepticism; and I can observe,
too, some of your sect to be as decisive
as those who make greater professions of
certainty and assurance. (DNR 1.136)
It is now possible to appreciate the full
force of Cleanthes’ attack on radical scepti-
cism. Initially it appeared that he was con-
centrating on the issue of whether a person
can sincerely embrace the thesis that no
beliefs other than H-minimal beliefs are ever
rationally justified. And Philo’s appeals to
psychological necessitation seemed to pro-
vide a satisfactory explanation of why the
sceptic shares so many non-epistemic beliefs
with people who would vehemently repudi-
ate radical scepticism. However, Philo’s ensu-
ing attempt to locate an area of inquiry in
which the radical sceptic’s distinctive views
about the limited availability of rationally
justified beliefs generate a plausibly benefi-
cial reordering of his non-epistemic beliefs
157
9780826443595_Ch06_Final_txt_print.indd 157
appears to leave him potentially vulnerable to
the charge that these views turn out to have
no practical consequences whatsoever. If the
radical sceptic responds at the level of his
non-epistemic beliefs to what he would call
‘putative’ or ‘alleged’ evidence in exactly the
same manner that the non-sceptic responds
to what he views as real evidence, then it
seems that even if we disregard the issue of
sincerity, there is nothing of any substantial
value to be gained from the mental exertion
required to sustain a sceptical stance. On the
other hand, if the radical sceptic genuinely
finds, as Philo suggests, that his mental pos-
ture is modified so that refined and abstruse
reasoning loses much of its usual persuasive-
ness, how can a radical sceptic be convinced
by the refined and abstruse reasoning that
underpins Hume’s science of human nature?
3. SCEPTICISM AND THE CURBING
OF THE RESTLESS IMAGINATION
It is clear that a satisfactory explanation of
how Hume sees scepticism and the science of
human nature as mutually reinforcing com-
ponents of a genuinely illuminating world-
view needs to do more than merely point to
his commitment to the supposition that many
of our beliefs are the inevitable causal prod-
ucts of psychological mechanisms that can-
not be disrupted by abstract reasoning. That
commitment accounts for Philo’s insouciance
in the face of Cleanthes’ misguided conten-
tion that a true sceptic would have no motive
for leaving via the door and stairs rather than
a high window. But it does not explain how
radical scepticism can give rise to real changes
amongst a person’s non-epistemic beliefs
while leaving refined and abstruse reason-
ing with enough persuasiveness to support
2/9/2012 9:35:14 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
an edifice of theory with the complexity
manifested by the science of man that Hume
is attempting to develop in the Treatise. In
order to find Hume’s explicit account of how
this is possible, we must turn our attention to
the final section of Book One.
Hume appears content in this section to
present sceptical arguments as serving to
expose our lack of rational justification for
our beliefs. Indeed, he even admits that these
arguments can fleetingly alienate us from our
most stable beliefs about contingent matters
of objective fact. He also emphasizes that
‘the intense view’ of the ‘manifold contradic-
tions and imperfections in human reason’
has effects that cannot be dispelled through
the efforts of reason itself (THN 1.4.7.8 /
268–9). Nature, rather than reason, ultim-
ately brings such doxastic alienation to an
end, but the initial outcome of this fortuitous
cure is a state of ‘spleen and indolence’ (THN
1.4.7.11 / 270) that is not conducive to any
further philosophical or scientific inquiry. As
time passes, however, and the inquirer grows
‘tir’d with amusement and company’ (THN
1.4.7.12 / 270), an inclination to return to
intellectually taxing investigations spon-
taneously arises again, and new attempts
at constructing intellectual systems begin.
Hume is adamant, however, that this process
operates in accordance with a natural cycle.
Until such time as a psychological propensity
to find some principles and claims immedi-
ately appealing and convincing is activated,
and the mind also recovers its enthusiasm for
intricate reasoning falling outside the bound-
aries of common life, it is simply not psycho-
logically possible to arrive at any fresh beliefs
in either the moral or the natural sciences in
the face of the influence exerted by remem-
bered sceptical arguments. But once these
mechanisms are in play once more, the con-
struction of a science of man in accordance
158
9780826443595_Ch06_Final_txt_print.indd 158
with the psychological foundations provided
by experience and observation can resume:
If the reader finds himself in the same
easy disposition, let him follow me in my
future speculations. If not, let him fol-
low his inclination, and wait the returns
of application and good humour. (THN
1.4.7.14 / 273)
Hume describes the conduct of a man who
engages in philosophical inquiry when the
inclination spontaneously arises within him,
or in ‘this careless manner’, as ‘more truly
sceptical than that of one, who feeling in him-
self an inclination to it, is yet so over-whelm’d
with doubts and scruples, as totally to reject
it’ (ibid.).4 The point here is that resisting
one’s natural propensities is something that
requires a rationale. Sceptical arguments,
however, subvert standard epistemic reasons:
it is no longer possible for a sceptic, even of
a mitigated or Academic variety, to think of
himself as striving against his inclinations in
order to construct a set of beliefs in accord-
ance with the demands of reason. Someone
who strenuously strives to shed his beliefs on
the grounds that this is what reason demands
of an inquirer who no longer views his beliefs
as rationally justified has failed to appreciate
that this alleged constraint of reason is itself
something that sceptical arguments expose as
devoid of rational justification.
Now a suspensive sceptic might possibly
claim that he finds himself convinced that
only widespread suspension of belief can pro-
tect him against erroneous beliefs that are not
balanced by equally valuable true beliefs, and
that this constitutes a motive for suspending
judgement. Hume’s reply to this contention
would presumably be that the most plausible
psychological theories available to us imply
that sceptical arguments, as opposed perhaps
2/9/2012 9:35:14 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
to insanity, cannot subvert our belief-forming
mechanisms so completely. Moreover, he
can also point out that there is no recorded
instance of anyone arriving at such a state,
rather than a state of verbally asserting that
he is suspending judgement on a radical scale,
as a result of exposure to any kind of philo-
sophical argument, and that the behaviour of
this self-professed suspensive sceptic makes it
clear that he too has numerous beliefs. Indeed
it is difficult to see how anyone, no matter
how carried away by intense philosophical
reflection, could deny in good faith that he
generally believes that if wholesale suspen-
sion of belief were to be instantiated, then ‘all
human life must perish’ (EHU 12.23 / 160).
More plausibly, perhaps, someone who
purports to be unwilling to engage in phil-
osophico-scientific theorizing and system-
building might argue that he simply finds the
results of abstruse and complicated reason-
ing, at least about matters of fact, to be so
full of ‘errors and delusions’ (THN Intro. 3 /
xiv) that he sees no point in continuing with
an enterprise that has such a disappointing
outcome. Even this position, however, seems
an untenable one. Cleanthes’ comments in
the Dialogues are particularly apposite here:
‘In reality, would not a man be ridiculous,
who pretended to reject newton’s explica-
tion of the wonderful phenomenon of the
rainbow, because that explication gives a
minute anatomy of the rays of light; a sub-
ject, forsooth, too refined for human com-
prehension.’ (DNR 1.136)
Very few people are prepared in practice
to repudiate all the results of refined and
abstruse reasoning, especially where this
yields verifiable predictions of novel phenom-
ena or generates new technologies that assist
in the manipulation of the world around us.
And if such reasoning is perforce embraced
in the case of the natural world, what basis
159
9780826443595_Ch06_Final_txt_print.indd 159
can there be for denying that it might also
be successful in uncovering the truth in the
moral sciences?
The other direction from which pressure
falls on a suspensive sceptic who professes
to suspend judgement on the results of all
abstruse reasoning is his practice of forming
numerous opinions within the arena of eve-
ryday life. According to Hume, ‘philosophi-
cal decisions are nothing but the reflections
of common life, methodized and corrected’
(EHU 12.25 / 162). As long as we remain
connected in this way to the reasoning we
employ and find persuasive outside of our
philosophical investigations, there is no
sharp division to be drawn between every-
day and philosophical reasoning, and hence
there is, it might be thought, no principled
motive for refusing to participate in philo-
sophical inquiry. After all, if one appeals
once more to the existence of unanswerable
Pyrrhonean arguments, or indeed Humean
genetic arguments for scepticism based on
alleged discoveries about the nature of the
psychological mechanisms responsible for
our beliefs,5 it is an evident fact, and one that
cannot be denied in good faith by any philo-
sophical sceptic, that these arguments tell
equally against the justification of the every-
day beliefs that all sceptics have in profusion
and are equally ineffectual at a psychological
level against these everyday beliefs and some
scientific theories. As Philo contends in his
own defence of scepticism: ‘To philosophise
on such subjects [natural and moral subjects]
is nothing essentially different from reason-
ing on common life; and we may only expect
greater stability, if not greater truth, from
our philosophy, on account of its exacter
and more scrupulous method of proceeding.’
(DNR 1.134)
Now it might be suggested that this argu-
ment from the close relationship between
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
philosophical reasoning and everyday rea-
soning overlooks Hume’s own explanation of
the inability of radical sceptical arguments to
dislodge our everyday non-epistemic beliefs.
Hume accepts that these arguments are plau-
sibly seen as continuations of patterns of
reasoning that we embrace and respond to
in the course of daily life (see, in particular,
THN 1.4.1.8–9 / 183–5). However, despite
this intimate relationship with common-sense
reasoning, they fail to have much impact on
our non-epistemic beliefs. Hume explains
this in terms of the obstacles the imagination
encounters when it attempts to participate
in complicated and subtle reasoning: ‘Where
the mind reaches not its objects with easiness
and facility, the same principles have not the
same effect as in a more natural conception
of the ideas; nor does the imagination feel a
sensation, which holds any proportion with
that which arises from its common judg-
ments and opinions’ (THN 1.4.1.10 / 185).
An important feature of this explanation,
however, is that complication and subtlety
are matters of degree. From Hume’s perspec-
tive, arguments gradually lose their persua-
siveness as their intricacy and complexity
increase: this loss of persuasiveness does
not stem from a change in the fundamental
nature of the argumentation presented to
us. Thus the sceptic who maintains that it is
appropriate for him to suspend judgement
on all conclusions reached by abstruse and
refined reasoning might be tempted to dis-
miss the significance of the close connections
between everyday reasoning and the reason-
ing used in philosophy and science. After all,
sceptical reasoning and everyday reason-
ing also have much in common, but Hume
concedes that they affect our non-epistemic
beliefs very differently.
The problem with this response is that it
fails to take into account the fact that some
160
9780826443595_Ch06_Final_txt_print.indd 160
relatively complicated arguments in the natu-
ral sciences do patently command the assent
of even the most determined sceptics. As rea-
soning in the natural and moral sciences is
‘nothing but a more regular and methodical
operation of the same kind’ (DNR 1.134) as
reasoning in common life, this affords a strong
presumption in favour of its being very simi-
lar in its persuasiveness to the reasoning that
shapes our beliefs in everyday contexts. Now
the suspensive sceptic rightly injects a note of
caution here. The relationship between eve-
ryday reasoning and arguments for radical
scepticism conspicuously fails to ensure that
sceptical arguments have a similar influence
over our non-epistemic beliefs. But when we
discover that complicated arguments in the
natural sciences do sometimes lead to sci-
entific theories that even the most hardened
sceptic would find it ridiculous to reject, then
the presumption under discussion here must
surely carry a great deal of weight.
The only recourse left open to the suspen-
sive sceptic would be to challenge the assimi-
lation of arguments in the human sciences to
arguments in the natural sciences rather than
radical Pyrrhonean arguments. Arguments
in the natural sciences and radical scepti-
cal arguments both manifest a high level of
abstraction and complexity. However, some
arguments of this kind within the field of the
natural sciences generate stable non-epistemic
beliefs, yet abstract and complicated argu-
ments of a radically sceptical kind appear to
lack this capacity. What, then, explains this
difference? Until we understand why the out-
comes are so dissimilar, we might be mistaken
in treating arguments in the human sciences
as potentially performing in the manner of
their counterparts in the natural sciences.
The explanation that Hume seems to
endorse is that arguments in the natural sci-
ences gain persuasiveness and psychological
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
force from their strong links with experience
and observation:
It is only experience, which teaches us
the nature and bounds of cause and
effect, and enables us to infer the exist-
ence of one object from that of another.
Such is the foundation of moral rea-
soning, which forms the greater part of
human knowledge, and is the source of
all human action and behaviour. (EHU
12.29 / 164)
Sceptical arguments, in contrast, are rela-
tively detached from experience, and they
represent instead attempts at exploring the
implications of various abstract epistemic
norms. Hume says little about the origins
of these norms, possibly because he believes
that all his readers will readily acknowledge
both the psychological pull exerted by these
norms and the key prescriptions they lay
down. For our current purposes, however, the
key point is that sceptical arguments seem to
rely entirely on such norms’ antecedent psy-
chological force. And as the argumentative
links with this original source of vivacity and
persuasiveness become increasingly compli-
cated and intricate, the conclusions become
less and less convincing. In the case of argu-
ments in the natural sciences, however, their
persuasiveness can be reinforced by adding
additional lively impressions to our stock of
supporting evidence or by securing obser-
vations that confirm the predictions of the
theories built upon this argumentation or the
effectiveness of the technology arising from
those theories.
The closest that sceptical arguments can
come to exploiting such links with lively
impressions is when we take ourselves to
be encountering impressions that falsify
beliefs we have formed on the basis of rea-
soning that sceptical arguments condemn as
161
9780826443595_Ch06_Final_txt_print.indd 161
insufficiently rigorous. Repeated experiences
of that kind might tend to reinforce our com-
mitment to the standards of self-criticism
exploited by radical scepticism. But even this
might, in turn, be subverted by a recognition
that such self-criticism would also suggest
that we are being excessively rash in taking
our apparent observations of falsifying phe-
nomena at face value. Essentially, therefore,
the Humean view is that compelling scientific
arguments and inferences gain persuasiveness
from being directly linked to the enlivening
effects of sense perception, whereas sceptical
arguments have little, if any, scope to exploit
this source of conviction. Thus the contest
between good scientific arguments and even
the most powerful sceptical arguments is, at
the non-epistemic level, a distinctly unequal
contest:
These [sceptical] principles may flour-
ish and triumph in the schools; where it
is, indeed, difficult, if not impossible, to
refute them. But as soon as they leave
the shade, and by the presence of the real
objects, which actuate our passions and
sensations, are put in opposition to the
more powerful principles of our nature,
they vanish like smoke, and leave the most
determined sceptic in the same condition
as other mortals. (EHU 12.21 / 159)
When it comes to epistemic appraisals,
in contrast, the struggle is one that scep-
tical arguments can win. It is one thing to be
steadfastly convinced that it is true that p and
quite another to be convinced that one’s belief
that p meets the abstract epistemic standards
required for it to qualify as a rationally justi-
fied belief. The first level of conviction can be
reached simply with the assistance of direct
perception or a causal inference based on a
lively impression and associative links created
by past patterns within one’s impressions.
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
However, the second level of conviction
requires the belief to be assessed against a set
of epistemic norms that are not themselves
directly rooted in perception. And as positive
assessments are themselves the products of a
relatively artificial and strained mental pos-
ture, they can be readily counterbalanced by
sceptical arguments.
The vulnerability of our epistemic apprais-
als derives from the fact that the realm of such
appraisals, as opposed to the realm of beliefs
about mere matters of physical and psycho-
logical fact, is itself a relatively rarefied arena
of thought that has only an indirect connec-
tion with what we are told by observation
and experience. Because it is not a natural
and instinctive turn of thought, conviction is
not as strong and as stable as it often is con-
cerning straightforward matters of fact:
No wonder, then, the conviction, which
arises from a subtile reasoning, dimin-
ishes in proportion to the efforts, which
the imagination makes to enter into
the reasoning, and to conceive it in all
its parts. Belief, being a lively concep-
tion, can never be entire, where it is not
founded on something lively and easy.
(THN 1.4.1.11 / 186)
It is not just our epistemic beliefs, how-
ever, that are vulnerable to sceptical argu-
ments. Non-epistemic judgements that arise
from over-lively imaginations that respond
to weak associations of ideas, rather than
the strong associations that underpin causal
inference, are also sufficiently weakly rooted
in our natural belief-forming mechanisms
that they too can be swept aside by radical
sceptical arguments. And it is this vulnerabil-
ity that allows sceptical argumentation to
perform its valuable function of eliminating
all those metaphysical positions that are built
upon excessively exuberant imaginations and
162
9780826443595_Ch06_Final_txt_print.indd 162
loose associative links. Theories built upon
solid causal reasoning are not similarly vul-
nerable; so they survive the encounter with
even the most powerful sceptical arguments
without any substantial damage to their
credibility. The only upshot in this latter case
is a modest prompting towards a seemly
degree of intellectual modesty and a willing-
ness to consult the opinions of other people
(see EHU 12.24 / 161–2). Thus the applica-
tion of radically sceptical arguments, when
their power and irrefutability is genuinely
internalized, is an extremely efficacious way
of orienting the mind towards theories sup-
ported by exemplary causal reasoning and
away from specious theories supported only
by weaker principles of mental association.
The outcome for a person who has what
Hume regards as the appropriate attitude
towards sceptical arguments is that some
theory-building continues even after attain-
ing a sceptical apotheosis. Causal arguments
that possess substantial experiential backing
are almost completely impervious to scep-
tical arguments at the level of psychological
persuasiveness, and they can therefore sup-
port conclusions in the natural and human
sciences of a recondite and abstruse kind.
Engaging in serious intellectual inquiry
does, however, leave the inquirer susceptible
to sudden shifts in his attitude towards his
apparent discoveries. According to Hume,
such inquiry will inevitably see us yield-
ing from time to time ‘to that propensity,
which inclines us to be positive and certain
in particular points, according to the light,
in which we survey them in any particular
instant’ (THN 1.4.7.15 / 273). Hume sees
this as an inevitable psychological conse-
quence of giving our full attention to specific
pieces of evidence and reasoning. While no
countervailing considerations are before the
mind, we are pulled along by the immediate
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
plausibility of the narrow and unbalanced
range of thoughts present to the mind at that
particular moment. No direct effort of will
can suppress these momentary lapses into
exuberant dogmatism. The only way of elim-
inating them would be to ‘forbear all exam-
ination and enquiry’ (THN 1.4.7.15 / 274).
However, the consequences of doing this at
the level of everyday enquiry would be, as we
noted before, the destruction of all human
life (see EHU 12.23 / 160). And although the
outcome of eschewing all enquiry of a sci-
entific and philosophical kind would be less
disastrous, it would still be absurd for those
with an inclination towards such researches
to engage in the struggle with their natural
impulses that this would involve.
The arguments of radical sceptics have the
psychological power to force some of us to
abandon our initial belief in the existence
of certain types of reasons for actions and
opinions. However, they do not generate any
new reasons to guide people’s behaviour and
judgements. They subvert the framework
of epistemological rationality, but they put
nothing new in its place. Moreover, it would
be a serious mistake to suppose that this
subversion fails to encompass the supposed
normative requirement to eschew beliefs that
one cannot sincerely view as rationally jus-
tified. Hume’s science of man critiques that
alleged norm at a variety of levels. Thus he
takes great delight, for example, in repeat-
edly exposing our psychological inability to
put this alleged norm into practice. But the
key observation to make in respect of the
implications of radical scepticism is that this
norm is a self-subverting norm.
Suppose that one has been driven, through
the contemplation of sceptical arguments,
to the conclusion that no beliefs other than
H-minimal beliefs are ever rationally justi-
fied. It is a commonplace of discussions of
163
9780826443595_Ch06_Final_txt_print.indd 163
radical scepticism to claim that this is a self-
refuting conclusion. Now it is true that it is,
by its own standards, not a rationally justi-
fied conclusion. But this does not entail that
it is a false conclusion, nor does this give us
any grounds for supposing that some con-
trary conclusion is better justified. Moreover,
if it is urged at this point that we are under
an intellectual obligation to believe that we
ought to eschew beliefs when we cannot
regard them as rationally justified, this sim-
ply raises the issue of the status of this sus-
pensive norm. And a little reflection seems
to establish that it is this norm, at least as it
is supposed to hold sway in anyone’s mind,
that is, unlike radical scepticism, genuinely
self-subverting.
Radical scepticism as a set of opinions
about the epistemic status of beliefs is fully
internally coherent unless it incorporates the
suspensive norm currently under discussion,
and there is no obvious basis for concluding
that it should embrace this norm. Indeed, it
appears that there are compelling grounds
for supposing that it is not a coherent norm
to endorse if one is a radical sceptic. Radical
scepticism can be true and we can believe it
to be true without transgressing any doxastic
norms internal to itself even if it does imply,
as a thesis, that it is not a position that pos-
sesses any positive justification. But if radical
scepticism across the range of topics envis-
aged by Hume is true and we believe it to be
true, then the suspensive norm undermines
itself. Let us begin by trying to assume that
this norm is true. Then it instructs us to eject
it from our minds, as the suspensive norm
comes to be its own target given its status as
a belief that we cannot regard as justified. On
the other hand, if we assume that it is a false
norm, then we clearly do not wish to allow
it any influence over us. It is, therefore, not a
norm we should allow to shape our doxastic
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
preferences, because it is either a false norm
or a norm that tells us, as radical sceptics, to
disregard it.
Only a superficial sceptic, accordingly,
would combine radical scepticism about all
beliefs with more than H-minimal content
with the judgement that this places him under
some obligation to suspend judgement on all
such matters. A more profound sceptic rec-
ognizes that thoroughgoing scepticism about
rational justification actually allows for a
level of doxastic promiscuity that is entirely
compatible with possessing as rich a set of
beliefs as would be possessed by any non-
sceptical inquirer who is guided by observa-
tion and experiment rather than the fanciful
products of an overheated imagination.
We have seen, therefore, that radical scep-
ticism of a near-global kind does not place
its adherents under an obligation to prac-
tise near-universal suspension of judgement.
It should be seen instead as merely expos-
ing the hollowness of the supposed epi-
stemic reasons that non-sceptics regard as
constraining the beliefs they ought to hold.
All other motives for holding beliefs remain
potentially intact, although the balance of
psychological forces that generate and sus-
tain our beliefs is subtly rearranged. Beliefs
founded on loose, unstable principles of
association prove to be causally incapable
of withstanding prolonged sceptical scru-
tiny, and the problems exposed within our
epistemic practices by sceptical argumenta-
tion have the effect of reducing intellectual
arrogance and dogmatism. It follows that if
one is so constituted that one takes pleas-
ure in complicated intellectual investiga-
tions, it is not the case that scepticism, even
when accepted as true, provides any reason
why one should puritanically refrain from
such activities. In the words Hume uses as
Book One of the Treatise approaches its
164
9780826443595_Ch06_Final_txt_print.indd 164
conclusion: ‘A true sceptic will be diffident
of his philosophical doubts, as well as of his
philosophical conviction; and will never ref-
use any innocent satisfaction, which offers
itself, upon account of either of them.’ (THN
1.4.7.14 / 273)
One’s willingness to accept that particu-
lar beliefs cannot be rationally justified is,
moreover, entirely compatible with firmly
regarding those beliefs as true and also hav-
ing the meta-belief that these beliefs are the
products of psychological mechanisms that
mostly latch on to the truth in nearby pos-
sible worlds. Viewing beliefs as incapable of
being supported by any form of reasoning or
putative evidence that does not exhibit circu-
larity or dependence on arbitrary and unde-
fended assumptions is something that needs
to be carefully distinguished from a stance of
regarding those beliefs as arising from psy-
chological mechanisms that would generate
false beliefs in many possible worlds close to
the one within which we take ourselves to
exist.
The first stance is one the Humean sceptic
adopts in respect of almost all beliefs. The
second stance, however, is potentially one
with much more radical implications at a
doxastic level. This latter stance is associated
with either genuine suspension of judgement
on non-epistemic matters or, in those cases
where suspension of judgement fails to super-
vene, a state of alienation from one’s beliefs
that mimics the attitude that a self-aware
kleptomaniac or sufferer from a phobia takes
towards the beliefs caused by his condition.
Widespread suspension of judgement or dox-
astic alienation is not a phenomenon associ-
ated with the Humean sceptic, and it is the
foregoing distinction that explains how this
lack of disruption manages to co-exist with
extremely wide-ranging scepticism about the
availability of rational justification. It also
2/9/2012 9:35:15 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
serves to explain how the Humean sceptic
can possess a high degree of confidence in
the abilities of human beings to uncover the
truth about even recondite matters of inquiry.
The Humean sceptic views himself as fortu-
nately located in an actual world that is one
of those possible worlds in which many key
mechanisms of belief formation latch on to
the truth and also do this in the vast major-
ity of neighbouring possible worlds. Thus
he does not see his beliefs as the product of
psychological mechanisms that generally fail
to track the truth, and he is consequently
neither alienated from those beliefs nor is he
inclined to abandon them. At the same time
he does accept that their truth is still, in some
respects, a matter of unearned good fortune
in the sense that it is impossible to construct
any argument to defend the supposition
that the actual world has the properties he
ascribes to it that is not patently arbitrary or
question begging. Thus he does regard all his
beliefs, other than his H-minimal beliefs, as
lacking rational justification.6
We arrive, therefore, at the combination
of thoughts that gives rise to the distinct-
ive character of Humean scepticism. The
Humean sceptic is confident that he has
investigative techniques at his disposal that
can reliably yield more true beliefs than
false beliefs. These are the techniques of ‘the
experimental method of reasoning’, which
is mentioned on the title pages of all three
books of the Treatise. He also accepts, how-
ever, that any argument intended to show
that this is the case must be probatively
defective. Ultimately all that underpins this
confidence is animal instinct and the posi-
tive psychological reinforcement we receive
when our beliefs steer us away from pain or
towards pleasure and also permit us to avoid
encountering too many situations where we
are driven to revise our previous beliefs.
165
9780826443595_Ch06_Final_txt_print.indd 165
It is one thing, however, to defend a habit of
inquiry from criticisms that we should aban-
don this practice and to point out that such
inquiry will inevitably lead us into momen-
tary spasms of complete conviction, and quite
another thing to maintain that such moments
provide us with beliefs that can be assembled
into a system that is beyond all criticism and
challenge. Hume confesses, therefore, that
although he has sometimes made use in the
Treatise of ‘such terms as these, ’tis evident,
’tis certain, ’tis undeniable; which a due
deference to the public ought, perhaps, to
prevent’ (THN 1.4.7.15 / 274), these terms
are to be understood merely as spontaneous
expressions of momentary flashes of convic-
tion that fail to retain their full vigour when
placed in a broader context of countervailing
considerations and general worries about the
adequacy of human reason.
Hume, accordingly, ends Book One of the
Treatise with an observation that once again
alerts us to Hume’s wish to fuse philosoph-
ical scepticism and the experimental method
into one integrated and mutually supportive
methodology for investigating both the nat-
ural world and human psychology. Hume
declares that these emphatic declarations
of certainty and conviction ‘were extorted
from me by the present view of the object,
and imply no dogmatical spirit, nor conceited
idea of my own judgement, which are senti-
ments that I am sensible can become no body,
and a sceptic still less than any other’ (ibid.).
As Hume is referring here to characteristics of
his own psychological constitution and deny-
ing that he possesses a ‘dogmatical spirit’ or
‘conceited idea of my own judgement’, there
is surely no good motive for Hume to say that
he is aware that these are dispositions that are
not appropriate for a sceptic unless Hume is,
at this point, thinking of himself as a scep-
tic.7 Thus Book One of the Treatise begins
2/9/2012 9:35:16 PM
 HUME ON SCEPTICISM AND THE MORAL SCIENCES
and ends with statements that are revelatory
of the two central pillars of his overall philo-
sophical outlook. As we have already stressed,
Hume’s subtitle for the Treatise identifies it
as seeking to apply the experimental method
to moral subjects. Moreover, the final sen-
tence of Book One constitutes Hume’s formal
unmasking of himself as a sceptic. And the
journey from that optimistic subtitle to the
concluding announcement of his commitment
to scepticism is very much an exploration of
how scepticism is not only compatible with
the successful application of the experimental
method but also assists us to apply it correctly.
As scepticism demolishes the pretensions of
that method’s putative competitors and pre-
vents us from being led astray by seemingly
attractive lines of thought that are not genu-
inely grounded in the strong associative links
that underpin compelling causal inferences,
the Humean synthesis of radical scepticism
and the experimental method represents a
new and improved form of empiricism rather
than the despairing renunciation of inquiry
and reasoning that is so often attributed to
scepticism by its detractors.
NOTES
1
In this sub-section of An Enquiry concern-
ing Human Understanding, Hume uses the
word ‘Pyrrhonism’ at 12.21 / 158 and the
phrase ‘excessive scepticism’ at 12.23 / 159.
‘Pyrrhonian’ is used as a noun twice at 12.23
/ 160, and the phrase ‘excessive principles of
scepticism’ can be found at 12.21 / 158–9.
2
The first detailed presentation of the case for
interpreting Hume’s version of Academic scep-
ticism as a causally rather than intellectually
166
9780826443595_Ch06_Final_txt_print.indd 166
mitigated form of scepticism seems to have
been Robert Fogelin’s paper ‘The Tendency
of Hume’s Skepticism’, in M. Burnyeat (ed.),
The Skeptical Tradition (Berkeley, California:
University of California Press, 1983), pp.
397–413. This paper dramatically improves
our understanding of Hume’s scepticism by
showing how we can avoid treating Hume’s
Academic scepticism as a repudiation of the
sceptical arguments found in the Treatise.
3
Hume and his contemporaries used the term
‘natural philosophy’ to refer to what we now
describe as the natural sciences, i.e. such disci-
plines as physics, chemistry and biology. The
terms ‘moral science’ and ‘moral philosophy’
were used to refer to psychology and the social
sciences along with various forms of intellectual
activity that would still be collected without
any disquiet under the banner of philosophy
even today.
4
Thus the attribution to Philo of ‘careless
scepticism’ should not be seen as Hume’s way
of insinuating some criticism of his views or
his manner of presenting those views (DNR,
Pamphilus, 128).
5
The difference between standard regressive
arguments for scepticism and Hume’s genetic
arguments is usefully explored in R.J. Fogelin,
‘Hume’s Scepticism’, in D.F. Norton (ed.), The
Cambridge Companion to Hume (Cambridge:
Cambridge University Press, 1993), pp. 90–116.
6
This characterization of the doxastic posture of
the Humean sceptic draws heavily on Duncan
Pritchard’s analysis of the distinction between
veritic epistemic luck and reflective epistemic
luck. See D. Pritchard, Epistemic Luck (Oxford:
Oxford University Press, 2005), chap. 6, ‘Two
Varieties of Epistemic Luck’, pp. 145–80.
7
The important role played by the concluding
sentence of Book 1 in the overall structure
of the work is emphasized in D. Garrett, ‘“A
Small Tincture of Pyrrhonism”: Skepticism
and Naturalism in Hume’s Science of Man’,
p. 90n2, in W. Sinnott-Armstrong (ed.),
Pyrrhonian Skepticism (Oxford: Oxford
University Press, 2004), pp. 68–98.
2/9/2012 9:35:16 PM
 7
THE SELF AND PERSONAL IDENTITY
Harold Noonan
1. INTRODUCTION
Hume discusses the self and its identity in
three main places: in sections V and VI of
Part IV of Book I in the body of the Treatise
of Human Nature and in an Appendix pub-
lished a year later with Book III. In the last
he declares himself wholly dissatisfied with
his treatment of the topic in the body of the
Treatise, but confesses that he now finds
the whole matter a ‘labyrinth’ and identifies
two principles he can neither render consist-
ent nor renounce. There is no discussion of
the topic in the Enquiry concerning Human
Understanding.
Unfortunately Hume fails to make clear in
his recantation what he finds objectionable
in his earlier account (since the principles he
labels inconsistent are in fact consistent), and
though commentators have produced a vari-
ety of suggestions, no consensus as to what
Hume’s worry was has emerged.
Section VI of Part IV of Book I of the
Treatise is entitled ‘Of personal identity’.
Our first task will be to identify exactly what
this title denotes. The title of section V is
‘Of the immateriality of the soul’. However,
this is somewhat misleading, since Hume’s
main aim in this section is not to take sides
but to criticize a presupposition of both
sides of the debate between materialists and
167
9780826443595_Ch07_Final_txt_print.indd 167
immaterialists, namely, that the soul is a
substance in which perceptions inhere. The
attack on this proposition is the foundation
of Hume’s discussion in section VI and the
basis of his insistence that the self is a fiction.
This is where we start.
2. THE FICTION OF PERSONAL
IDENTITY
In the tradition in which Hume was writing,
deriving from Locke, the problem of personal
identity was seen as that of giving an account
of what constitutes personal identity. Locke’s
own answer to this question has a negative
component and a positive component. The
negative component is that personal identity
is not constituted by identity of substance,
whether material or immaterial, any more
than is identity of man: ‘it being one thing
to be the same substance, another the same
man, and a third the same person’.1 The posi-
tive component that does constitute personal
identity is sameness of consciousness: ‘And
as far as this consciousness can be extended
backwards to any past action or thought, so
far reaches the identity of that person’.2
Thus, Locke asserts, combining the two
components, ‘it being the same consciousness
2/9/2012 9:35:27 PM
 THE SELF AND PERSONAL IDENTITY
that makes a man be himself to himself, per-
sonal identity depends on that only, whether
it be annexed only to one individual sub-
stance or can be continued in a succession of
several’.3 In subsequent discussions reacting
to Locke the role of substance in the constitu-
tion of personal identity became the key issue
and Butler, Reid and Leibniz all restored, in
their accounts, the link that Locke had bro-
ken between personal identity and substan-
tial identity.4
If we read Hume as contributing to this
debate on the constitution of personal iden-
tity, we must understand his main contention
to be an emphatic endorsement of the nega-
tive component of Locke’s account: personal
identity is not constituted by identity of sub-
stance. But, in fact, to read Hume in this way
is to misunderstand him. According to Hume,
it is not merely that personal identity is not
constituted by identity of substance; personal
identity is a fiction; the ascription of iden-
tity over time to persons, a mistake (one we
all necessarily make). There is no intelligible
question of the form ‘In what does personal
identity consist?’. There is only the genetic
problem of specifying the psychological
causes of the universal but mistaken belief in
the existence of enduring persons. This is the
problem Hume addresses in his discussion of
personal identity.
However, it is not, of course, in Hume’s
view, a peculiarity of persons that they do
not endure self-identically over time; nor
does anything else. Hume thinks that the
idea of identity is incompatible with the idea
of change: it is the idea of an object which
‘remains invariable and uninterrupted thro’
a suppos’d variation of time’ (THN 1.4.6.6
/ 253). This is Hume’s attempt to define an
idea which is ‘a medium betwixt unity and
number’ (THN 1.4.2.28 / 200), or ‘more
properly speaking’, as he adds, ‘is either of
168
9780826443595_Ch07_Final_txt_print.indd 168
them, according to the view in which we
take it’ (THN 1.4.2.29 / 201). Strictly speak-
ing, it applies to nothing, since time requires
change, but it certainly fails to apply to
objects of ordinary discourse – plants, ani-
mals, artefacts and so on – and so, as in the
case of persons, when we ascribe identity to
them, Hume says, it is only in an ‘improper
sense’. Thus, for Hume, the genetic problem
of accounting for our false belief in the exist-
ence of enduring persons is just a part of the
wider genetic problem of accounting for our
false belief in the identity over time of chang-
ing things in general. In fact, he thinks, the
same mechanism of the imagination which
accounts for our ascriptions of identity over
time to plants, animals and so on can equally
well account for our ascriptions of identity
over time to persons. This is because:
The identity which we ascribe to the
mind of man, is only a fictitious one and
of a like kind with that which we ascribe
to vegetable and animal bodies. It can-
not, therefore, have a different origin,
but must proceed from a like operation
of the imagination upon like objects.
(THN 1.4.6.15 / 259; emphasis added)
The mechanism that generates the belief in
the fiction of personal identity (the identity
we ascribe to ‘the mind of man’) is the opera-
tion by which the mind is led to ascribe an
identity to distinct but related perceptions,
however interrupted or variable. It is a mech-
anism of the imagination in the narrow sense
which is distinguished in the footnote to sec-
tion IX of Part III of Book I:
the word imagination is commonly used
in two different senses; and though noth-
ing be more contrary to true philosophy,
than this inaccuracy, yet in the following
2/9/2012 9:35:27 PM

reasonings, I have often been obliged to
fall into it. When I oppose the imagina-
tion to the memory, I mean the faculty
by which we form our fainter ideas.
When I oppose it to reason, I mean the
same faculty excluding only our demon-
strative and probable reasonings. (THN
1.3.9.19 / 117)
Hume has appealed to this identity-ascrib-
ing mechanism earlier in his account of the
genesis of our belief in an external world
or ‘body’ in section II and in his account of
the ancient philosophers’ belief in material
substrata in section III. As the distinction
made in the quoted footnote indicates, he is
emphatic that the beliefs generated by it are
not thereby provided with a rational foun-
dation, not even in the sense in which our
beliefs in the unobserved effects or causes of
observed causes or effects do have a rational
foundation.
Hume summarizes the effect of the action
of the identity-ascribing mechanism of the
narrow imagination as follows:
In order to justify to ourselves this
absurdity [i.e. the ascription of identity
to distinct perceptions], we often feign
some new and unintelligible principle,
that connects the objects together, and
prevents their interruption or variation.
Thus we feign the continued existence of
the perceptions of our senses to remove
the interruption; and run into the notion
of a soul, and self and substance, to
disguise the variation, we may farther
observe, that where we do not give rise
to such a fiction, our propensity to con-
found identity with relation is so great,
that we are apt to imagine something
unknown and mysterious, connecting the
parts, beside their relation; and this I take
to be the case with regard to the identity
169
9780826443595_Ch07_Final_txt_print.indd 169
THE SELF AND PERSONAL IDENTITY
we ascribe to plants and vegetables. And
even when this does not take place, we
will feel a propensity to confound these
ideas, tho’ we are not fully able to sat-
isfy ourselves in that particular, nor find
anything invariable and uninterrupted
to justify our notion of identity. (THN
1.4.6.6 / 254–5)
Hume indicates here the various applications
of the identity-ascribing mechanism noted
above: in addition to generating our belief in
personal identity (i.e. a soul or self), it also
generates our belief in substance (material
substrata) as well as our belief in an external
world or body, i.e. the continued existence of
the perceptions of our senses, and our belief
in the identity of plants and animals.
The important point to note is that it is an
essential element of this story, as Hume tells
it, that the propensity we have to identify
distinct perceptions is a propensity to regard
them as identical in just the sense he defines:
‘an object that remains invariable and unin-
terrupted thro’ a supposed variation of time’.
It is only if this is our idea of identity that
the psychological mechanism can operate as
he suggests. If we thought that identity were
consistent with interruption or change then
we would not be led to ‘feign the continued
existence of the perceptions of our senses’ to
remove the interruption or to ‘run into the
notion of a soul, and self and substance’ or to
be ‘apt to imagine something unknown and
mysterious’ to disguise the variations. Hence
it is essential to Hume’s account that our idea
of identity is, in fact, the one he describes,
and it is because this is so that he says:
the controversy concerning identity is
not merely a dispute of words. For when
we attribute identity . . . to variable or
interrupted objects, our mistake is not
2/9/2012 9:35:27 PM
 THE SELF AND PERSONAL IDENTITY
confined to the expression, but is com-
monly attended with a fiction, either of
something invariable and uninterrupted,
or of something mysterious and inexpli-
cable, or at least with a propensity to
such fictions. (THN 1.4.6.7 / 255)
3. THE REIFICATION OF PERCEPTIONS
Although Hume’s account of identity makes
it impossible for him to accept enduring
selves and provides him with the starting
point for his account of the genesis of our
belief in personal identity, his conception of
what the self would have to be if it existed is
another ground for his rejection of it.
One of the best known passages in Hume’s
discussion of personal identity is his denial
that he is introspectively aware of any self or
mental substance:
For my part, when I enter most intimately
into what I call myself, I always stumble
on some particular perception or other,
of heat or cold, light or shade, love or
hatred, pain or pleasure. I never can catch
myself at any time without a perception,
and never can observe anything but the
perception. (THN 1.4.6.2 / 252)
Hume writes as if it is just a matter of fact
that on looking into himself he fails to find
anything but perceptions, but this sits ill with
his emphatic denial that he has any idea of a
self distinct from perceptions. I can be confi-
dent that I am not observing a tea-kettle now
because I know what it would be like to be
doing so. But if Hume has no idea of a self
he presumably has no conception of what it
would be like to observe one. In that case,
however, how does he know that he is not
doing so?
170
9780826443595_Ch07_Final_txt_print.indd 170
Another difficulty is that, as Chisholm
puts it, it looks very much as though the self
that Hume professes to be unable to find is
the one that he finds to be stumbling – stum-
bling on to different perceptions.5 For Hume
reports the results of his introspection in the
first person: ‘I never catch myself without a
perception’, ‘I never observe anything but the
perception’.
Hume’s denial is not, therefore, the
straightforward empirical assertion it might
at first appear to be. But then what is his
basis for it?
We must recall that Hume reifies percep-
tions. He starts from a conception of mental
states according to which for a person to be
in a mental state is for a certain relational
statement to be true of that person: that he
is perceiving a certain sort of perception (‘to
hate, to love, to think, to feel, to see, all this is
nothing but to perceive’ (THN 1.2.6.7 / 67).
But if this is correct it is very natural that
Hume should deny the introspective observ-
ability of the self. For if to be in any mental
state is to possess a relational property, then
no mental state can be an intrinsic quality
of its subject. Given that the only states of
which one can be introspectively aware are
mental, then introspective awareness of a
self would require awareness of it without
any awareness of its intrinsic qualities. But
surely it makes no sense to speak of observ-
ing something introspectively if the thing has
no intrinsic qualities whatsoever which one
can observe by introspection.
These simple reflections suffice, I think,
to explain Hume’s confidence in his denial
of the introspective accessibility of the self.
But they can be taken further if we now
turn from what the Humean conception
of the mental implies about the subject of
mental states – that its only properties are
relational ones – to what it implies about
2/9/2012 9:35:28 PM

their objects, Hume’s perceptions. What it
implies, of course, is that these perceptions
are things, indeed substances, and they are
logically capable of existing independently
of being perceived. Hume is emphatic that
this is the case. Indeed, he thinks that every-
thing which can be conceived is a substance
(THN 1.4.5.5 / 233). This is a consequence
he explicitly draws from the conjunction of
two of his fundamental principles in section
V and makes him ‘absolutely condemn’ the
question of the materiality or immateriality
of the soul (THN 1.4.5.6 / 234).
The first aim of section V is to put a stop
to the endless cavils between ‘the curious rea-
soners concerning the material or immaterial
substances in which they suppose our per-
ceptions to inhere’ (THN 1.4.5.2 / 232).
He first challenges these curious reason-
ers to explain what they mean by a sub-
stance and inhesion. He demands that they
point out the impression from which their
idea derives. And then he goes on to give his
master-argument:
If instead of answering these questions,
anyone shou’d evade the difficulty, by
saying, that the definition of substance
is something which may exist by itself,
and that this definition ought to satisfy
us: Shou’d this be said, I shall observe
that this definition agrees to everything,
that can possibly be conceiv’d, and never
will serve to distinguish substance from
accident, or the soul from its perceptions.
For thus I reason. Whatever is clearly
conceiv’d may exist, and whatever is
clearly conceiv’d after any manner may
exist after the same manner. This is one
principle . . . . Again, everything which
is different is distinguishable, and every-
thing which is distinguishable, is separa-
ble by the imagination. This is another
principle. My conclusion from both is,
171
9780826443595_Ch07_Final_txt_print.indd 171
THE SELF AND PERSONAL IDENTITY
that since all our perceptions are differ-
ent from each other, and from every thing
else in the universe, they are also distinct
and separable, and may be consider’d as
separately existent, and may exist sepa-
rately, and have no need of any thing
else to support their existence. They
are, therefore, substances, as far as this
definition explains a substance. (THN
1.4.5.5 / 233)
This argument is perfectly general, of course,
and if it is a good one it establishes that not
only perceptions but qualities generally are
logically capable of an independent exist-
ence. Hume is aware of this and indeed gives
the argument himself in his criticism of the
ancient philosophers’ conception of sub-
stance in section III: ‘every quality being a
distinct thing from another, may be conceiv’d
to exist apart, and may exist apart, not only
from every other quality, but from that unin-
telligible chimera of substance’ (THN 1.4.3.7
/ 222). But if so, Descartes’s famous analogy
in the Second Meditation, in which he com-
pares the relation between a piece of wax and
its qualities to the relation between a man and
his clothes, would be an appropriate one. But
one consequence of this analogy is that the
wax is represented as hidden beneath its gar-
ments and so as in itself unobservable. This
is because the analogy implies that the asser-
tion that the wax has any quality is in real-
ity an assertion of a relation between it and
something else. And a second consequence of
the analogy is that the qualities of the wax
are represented as being themselves substan-
tial, as though they can ‘stand by themselves’,
as a suit of armour can when no man is wear-
ing it. But these consequences of the analogy,
which is an appropriate one if the Humean
argument is a good one, make it obvious that
if the wax is so conceived, its existence, as
2/9/2012 9:35:28 PM
 THE SELF AND PERSONAL IDENTITY
anything other than that of a collection of
qualities, must be regarded as highly prob-
lematic. Exactly the same is true of the self if
Hume’s argument is correct.
4. THE REJECTION OF THE
SUBSTANTIAL SELF
With the argument of ‘Of the immateriality
of the soul’ against substance behind us we
can now turn to the details of Hume’s sec-
tion on personal identity. But before doing so
we should pause to look at another remark-
able argument in section V which brings out
how far Hume is prepared to proceed on
the basis of his conception of perceptions as
logically independent entities. Hume states
that although he has already condemned as
utterly unintelligible the question of the sub-
stance of the soul, he ‘cannot forbear propos-
ing some further reflections’. Namely, that
the doctrine of the immateriality, simplicity
and indivisibility of a thinking substance is
‘a true atheism and will serve to justify all
those sentiments for which Spinoza is so uni-
versally infamous’ (THN 1.4.5.17 / 240) and
that this ‘hideous hypothesis’ [Spinozistic
monism] is almost the same as that of the
immateriality of the soul. Hume’s argument
goes as follows:
there are two different systems of beings
presented, to which I suppose myself
under a necessity of assigning some sub-
stance, or ground of inhesion. I observe
first the universe of objects or of body:
the sun, moon, stars, the earth . . . Here
Spinoza . . . tells me that these are only
modifications; and that the subject, in
which they inhere, is simple, incom-
pounded, and indivisible. After this I con-
sider the other system of beings, viz. the
172
9780826443595_Ch07_Final_txt_print.indd 172
universe of thought, or my impressions
and ideas. There I observe another sun,
moon and stars. . . . Upon my enquiring
concerning these, theologians . . . tell me,
that these also are modifications . . . of
one simple substance. Immediately . . .
I am deafen’d with the noise of a hun-
dred voices, that treat the first hypoth-
esis with detestation and scorn . . . and
the second with applause and veneration
. . . I turn my attention to these hypoth-
eses . . . and find that they have the same
fault of being unintelligible . . . and [are]
so much alike, that . . . any absurdity in
one . . . is . . . common to both. (THN
1.4.5.21 / 242–3; emphasis added)
Nor are matters improved for the theolo-
gians, according to Hume:
if instead of calling thought a modifica-
tion of the soul, we should give it the
more antient, and yet more modish name
of an action. By an action we mean . . .
something which, properly speaking, is
neither distinguishable, nor separable
from its substance. . . . But nothing is
gained by this change of the term modi-
fication, for that of action. . . . First . . .
the word action, according to this expli-
cation of it, can never be justly apply’d
to any perception. . . . Our perceptions
are all really different, and separable,
and distinguishable from each other, and
from every thing else. . . . [In] the second
place . . . may not the Atheists likewise
take possession of [the word “action”],
and affirm that plants, animals, men,
etc., are nothing but particular actions of
one simple . . . substance? This . . . I own
’tis unintelligible but . . . assert . . . that
’tis impossible to discover any absurd-
ity in the supposition . . . which will not
be applicable to a like supposition con-
cerning impressions and ideas. (THN
1.4.5.27–8 / 245–6)
2/9/2012 9:35:28 PM

There could not, I think, be a clearer illustra-
tion than this of the seriousness with which
Hume takes his reification of perceptions – if
a tree cannot be a modification of Spinoza’s
God, my idea of a tree cannot be a modifica-
tion of me!
Turning now to the section ‘Of personal
identity’ Hume proceeds very rapidly, and
confidently, for reasons that I hope will now
be perfectly understandable, to his conclu-
sion that the self is nothing more than a bun-
dle of perceptions. The whole business takes
less than two pages.
Some philosophers have thought that ‘we
are every moment intimately conscious of
what we call our self’. But: ‘Unluckily all
these positive assertions are contrary to that
very experience which is pleaded for them,
nor have we any idea of self, after the manner
it is here explained, for from what impres-
sion could this idea be derived?’ Since the
self is supposed to be an unchanging object,
any impression of self must be constantly the
same throughout the whole course of our
lives. But, Hume finds, looking within him-
self, ‘There is no impression constant and
invariable. Pain and pleasure, grief and joy . . .
succeed each other . . . . It cannot therefore,
be from any of these impressions, or from
any other that the idea of self is deriv’d; and
consequently there is no such idea’ (THN
1.4.6.2 / 251–2).
Hume goes on to raise explicitly the dif-
ficulty that his conception of perceptions
as ontologically independent creates for the
notion of a substantial self:
But farther, what must become of all
our particular perceptions upon this
hypothesis? All these are different, and
distinguishable, and separable from each
other, and may be separately consider’d,
and may exist separately, and have no
173
9780826443595_Ch07_Final_txt_print.indd 173
THE SELF AND PERSONAL IDENTITY
need of any thing to support their exist-
ence. After what manner therefore do
they belong to self; and how are they
connected with it? (THN 1.4.6.3 / 252)
It is immediately after this that he issues his
denial of the observability of a self distinct
from perceptions, and concludes that the self
can be nothing but a bundle of perceptions.
The same structure is exhibited in the
Appendix, in which Hume summarizes his
argument for the bundle theory before mak-
ing his famous confession of bafflement. After
arguing that we have no impression of self or
substance as something simple or individual
from which these ideas might be derived, he
goes on to spend no less than three paragraphs
insisting on the ontological independence of
perceptions, finally concluding that since ‘’tis
intelligible and consistent to say that objects
exist distinct and independent, without any
common single substance or subject of inhe-
sion’ (that is, it is intelligible and consistent
to reject Spinoza’s monism), ‘this proposition
therefore can never be absurd with regard
to perceptions’ (THN App. 14 / 634). In the
immediately following paragraph he denies
the observability of the self and derives the
bundle theory.
So much, then, for Hume’s arguments for
the bundle theory of the self. Taken together
with his analysis of identity, they entitle him,
he believes, to the conclusion that personal
identity is a fiction, that ‘the mind is a kind
of theatre, where several perceptions succes-
sively make their appearance . . . . There is
properly no simplicity in it at one time, nor
identity in different’ (THN 1.4.6.4 / 253).
For the idea of identity is that of an object,
that ‘remains invariable and uninterrupted
thro’ a suppos’d variation of time’. But if the
bundle theory is correct, a person is nothing
but a sequence of different (ontologically
2/9/2012 9:35:28 PM
 THE SELF AND PERSONAL IDENTITY
independent) objects existing in succession,
and connected by a close relation – something
like a thunderstorm. But ‘as such a succession
answers perfectly to our notion of diversity, it
can only be by [a] mistake that we ascribe to
it an identity’ (THN 1.4.6.7 / 255).
The only question that remains then,
Hume thinks, is to explain the psychological
mechanism that accounts for this mistake.
5. HUME’S ACCOUNT OF THE
SOURCE OF THE MISTAKE
Hume summarizes his account of this as fol-
lows. In contemplating an identical, that is,
an invariable and unchanging object, we are
doing something very different from con-
templating a succession of objects related by
links of resemblance, causation and contigu-
ity but:
That action of the imagination, by which
we consider the uninterrupted and invar-
iable object, and that by which we reflect
on the succession of related objects, are
almost the same to the feeling . . . . The
relation facilitates the transmission of
the mind from one object to another,
and renders its passage as smooth as if it
contemplated one continu’d object. This
resemblance is the cause of the confusion
and mistake, and makes us substitute
the notion of identity, instead of that of
related objects. However at one instant
we may consider the related succession
as variable or interrupted, we are sure
the next to ascribe to it a perfect identity,
and regard it as invariable and uninter-
rupted. (THN 1.4.6.6 / 254)
This identity-ascribing mechanism of the
imagination is also operative, as noted
174
9780826443595_Ch07_Final_txt_print.indd 174
previously, in generating our belief in an
external world. Hume describes its opera-
tion here in ‘Of scepticism with regard to the
senses’. According to Hume’s story, the con-
stancy of our perceptions leads us to ascribe
them a numerical identity, despite their
interruptedness, since when we consider a
constant and uninterrupted object, on the
one hand, and reflect on an interrupted suc-
cession of constant objects, on the other, in
the second situation there is the ‘same unin-
terrupted passage of the imagination’ (THN
1.4.2.34 / 203) as in the first. The second situ-
ation places the mind in the same ‘disposition
and is considered with the same smooth and
uninterrupted progress of the imagination, as
attends the view of’ the first (THN 1.4.2.34
/ 204). But ‘whatever ideas place the mind in
the same disposition, or in similar ones, are
apt to be confounded’ (THN 1.4.2.32 / 203).
Thus I confound the two situations. But since
I take the former to be a view of an identical
object, I do the same with the latter situation
and ‘confound the succession with the iden-
tity’ (THN 1.4.2.34 / 204).
But Hume insists that it is an essential part
of the notion of identity that an identical
object must be uninterrupted as well as invar-
iable in its existence. Thus I unite the ‘broken
appearances’ by means of ‘the fiction of a
continu’d existence’ (THN 1.4.2.36 / 205).
That is, I come to believe that the identical
perception which I earlier perceived has con-
tinued in existence whilst unperceived. Hume
denies that this is a contradictory belief: ‘the
continu’d existence of sensible objects or
perceptions involves no contradiction’ (THN
1.4.2.40 / 208). Thus ‘we may easily indulge
our inclination to that supposition’ (THN
1.4.2.40 / 208). In doing so, I reconcile my
initially contradictory beliefs (in identity and
interruptedness) by abandoning the second.
I thus ‘remove entirely [the interruption] by
2/9/2012 9:35:28 PM

supposing that these interrupted perceptions
are connected by a real existence of which
we are insensible’ (THN 1.4.2.24 /199), and
arrive at a consistent set of beliefs.
This is Hume’s view of the form that the
belief in body takes in the minds of the vul-
gar, the non-philosophers: they believe that
their very perceptions have a continued and
distinct existence. This vulgar view involves
no internal incoherence or concealed absurd-
ity. It is a stable stopping place for the vulgar
mind; as it were, a finished product of the
imagination.
But it is false. Merely as a matter of empiri-
cally discoverable fact, Hume thinks, percep-
tions are dependent and perishing existences.
Experiments familiar to philosophers (but
apparently not known to the vulgar) estab-
lish this (THN 1.4.2.45 / 211).
But even philosophers cannot resist the
force of the identity-ascribing mechanism of
the imagination. They too must regard the
second situation as the view of an identical
object. However, they know that, as a matter
of fact, perceptions do not continue unper-
ceived. To resolve their conflict all they can
do is to distinguish between objects and per-
ceptions, ascribing continuity and distinct-
ness to the former and interruptedness to the
latter. But such a system of ‘double existence’,
Hume thinks, is only a ‘palliative remedy’
and contains all the difficulties of the vulgar
system, with others that are peculiar to itself
(THN 1.4.2.46 / 211).
The second operation of the identity-ascrib-
ing mechanism which Hume discusses before
the section on personal identity is its opera-
tion in bringing about the ancient philoso-
phers’ beliefs in material substrata. However,
it operates differently in this case, in a way
that causes Hume to say that whereas ‘we
[i.e. the vulgar] feign the continu’d existence
of the perceptions of our senses to remove the
175
9780826443595_Ch07_Final_txt_print.indd 175
THE SELF AND PERSONAL IDENTITY
interruption’ the ancient philosophers merely
run into the notion of substance to ‘disguise
the variation’ (THN 1.4.6.6 / 254).
Hume explains the genesis of the belief
in ‘substance, or original and first matter’
(THN 1.4.4.4 / 220) as follows:
as the ideas of the several successive
qualities of objects are united together
by a very close relation, the mind, in
looking along the succession . . . will not
more perceive the change, than if it con-
templated the same unchangeable object.
But when we alter our method of consid-
ering the succession, and . . . survey at
once any two distinct periods of its dura-
tion . . . the variations, which were insen-
sible when they arose gradually, do now
appear of consequence, and seem entirely
to destroy the identity . . . . When we
gradually follow an object in its succes-
sive changes, the smooth progress of the
thought makes us ascribe an identity to
the succession . . . . When we compare the
succession after a considerable change,
the progress of the thought is broke; and
consequently we are presented with the
idea of diversity: In order to reconcile
which contradictions the imagination
is apt to feign something unknown and
invisible, which it supposes to continue
the same under all these variations; and
this unintelligible something it calls sub-
stance or original and first matter. (THN
1.4.4.3–4 / 220)
There are three differences between this
account of the genesis of the ancient phi-
losophers’ belief in substance and Hume’s
account in the previous section of our belief
in body. First, no appeal is made to con-
stancy. The belief is not the result of the effect
on the mind of a constant but interrupted set
of perceptions, but of a gradually and imper-
ceptibly changing uninterrupted succession
2/9/2012 9:35:29 PM
 THE SELF AND PERSONAL IDENTITY
of qualities. Secondly, there is, therefore, no
stable consistent stopping place akin to the
vulgar belief in body, which is arrived at sim-
ply by endorsing the identity and rejecting the
appearance of interruption in the sequence of
perceptions as a mere appearance – the vari-
ation in the sequence of qualities cannot be
rejected as a mere appearance and is incon-
sistent with the idea of identity in Hume’s
account. Hence Hume does not distinguish
in this case between the vulgar and philo-
sophical views. There is only the philosophi-
cal view. The ancient philosophers’ belief is
not a secondary product of the imagination,
although it is a product of a multi-stage proc-
ess. Thirdly, what presents us with the idea of
diversity in this case, when the succession is
viewed from a single point in time and two
distinct periods of its duration surveyed, is
not interruption, but variation. As noted, this
cannot be regarded as a mere appearance, as
interruption is in the vulgar view of body. So
reconciliation of the two points of view is not
possible and, as we have noted, Hume care-
fully speaks only of ‘disguising’ the variation
to reconcile the contradiction. The ancient
philosophers’ belief in substance is not a con-
sistent one since it involves the idea of some-
thing which is both identical and changing.
The identity-ascribing mechanism of the
narrow imagination works to produce con-
flation and error in exactly the same way,
Hume thinks, in the case of personal identity.
The succession of my perceptions is merely
a succession of distinct related objects. But
because the objects in the succession are
closely related, the action of the imagina-
tion in surveying the succession is ‘almost
the same to the feeling’ as the action of the
imagination in considering an uninterrupted
and invariable object. As in the other cases,
the similarity between the two acts of mind
leads me to confound the two situations and
176
9780826443595_Ch07_Final_txt_print.indd 176
thus to regard the succession of related per-
ceptions as really united by identity. And so
I am led to believe in the unity of the self,
which is as much a fiction as in the other
cases of the operation of the identity-ascrib-
ing mechanism, and, ‘proceed[s] entirely from
the smooth and uninterrupted progress of
the thought along a train of connected ideas
according to the principles above explain’d’
(THN 1.4.6.16 / 260). All that remains to be
said, Hume thinks, is what the relations are
in this case that link my successive percep-
tions so as to bring about this uninterrupted
progress of the thought.
His answer is: resemblance and causation.
Our perceptions at successive times resem-
ble each other for a variety of reasons, of
course, but the one Hume emphasizes is that
people can remember their past experience:
For what is the memory, but a faculty
by which we raise up the images of past
perceptions? And as an image neces-
sarily resembles its object must not the
frequent placing of these resembling per-
ceptions in the chain of thought, convey
the imagination more easily from one
link to another, and make the whole like
the continuance of one object? (THN
1.4.6.18 / 260–1)
Given this copy theory of memory Hume is
able to regard memory not merely as provid-
ing us with access to our past selves, but also
as contributing to the bundles of perceptions
which we can survey, elements which repre-
sent, and thus resemble, earlier elements; and
so – since resemblance is a relation which
enables the mind to slide smoothly along a
succession of perceptions – as strengthening
our propensity to believe in the fiction of a
continuing self. In this particular case, then,
Hume is able to say, with a nod of agreement
2/9/2012 9:35:29 PM

to Locke, ‘memory not only discovers the
identity but contributes to its production’
(THN 1.4.6.18 / 261).
But we do not remember all, or even most
of, our past actions or experiences. Yet we
do not affirm, because we have entirely for-
gotten the incidents of certain past days,
that the present self is not the same person
as the self of that time. Consequently there
must be something else which enables us to
think of our identity as extending beyond
our memory.
Here Hume appeals to causality, which has
been previously introduced in his account of:
the true idea of the human mind . . . a
system of different perceptions or differ-
ent existences, which are linked together
by the relation of cause and effect . . . .
In this respect I cannot compare the
soul more properly to anything than to
a republic or commonwealth, in which
the several members are united by the
reciprocal ties of government and subor-
dination, and give rise to other persons,
who propagate the same republic in the
incessant changes of its parts. (THN
1.4.6.17 / 261)
When we think of ourselves as existing at
times we cannot remember we do so, Hume
says, by imagining the chain of causes and
effects that we remember extending beyond
our memory of them. So the causal links
between our perceptions, as well as their
resemblances, are crucial to our belief in a
continuing self which exists at times it no
longer recalls. Consequently, Hume is able
to say, this time in agreement with Locke’s
opponents: ‘In this view . . . memory does
not so much produce as discover personal
identity, by shewing us the relation of cause
and effect among our different perceptions’
(THN 1.4.6.20 / 262).
177
9780826443595_Ch07_Final_txt_print.indd 177
THE SELF AND PERSONAL IDENTITY
According to this account of how belief
in personal identity arises via the identity-
ascribing mechanism of the imagination, it
operates as it does in producing the ancient
philosophers’ belief in material substrata and
not as it does in producing belief in body,
as Hume indicates: ‘we feign the continu’d
existence of the perceptions of our senses
to remove the interruption; and run into
the notion of a soul, and self, and substance
to disguise the variation’ (THN 1.4.6.6 /
254). Our belief in personal identity is not
explained by the constancy of our percep-
tions. (On the contrary, these:
succeed each other with an inconceiv-
able rapidity, and are in perpetual flux
and movement. Our eyes cannot turn in
their sockets without varying our per-
ceptions. Our thought is still more vari-
able . . . , and all our other senses and
faculties contribute to this change; nor is
there any single power of the soul, which
remains unalterably the same perhaps
for one moment (THN 1.4.6.4 / 253).)
And, like the ancient philosophers’ belief,
it is not in fact a successful attempt at rec-
onciliation of our belief in identity with
the variation we cannot deny, but merely
an attempt to disguise the variation. This
is why there is no distinction between vul-
gar and philosophical forms of the belief in
personal identity. In fact, what disposes us to
regard temporally distinct and sensibly dis-
tinguishable (non-resembling) perceptions
as identical is primarily their causal related-
ness. This is present when perceptions are
neither invariable nor uninterrupted, that
is, when neither of the conditions specified
in Hume’s account of our idea of identity is
satisfied. Nevertheless, Hume thinks, just as
the ancient philosophers were caused by the
imperceptibility of the differences between
2/9/2012 9:35:29 PM
 THE SELF AND PERSONAL IDENTITY
successive qualities to regard them as iden-
tical and thence were led to the notion of
substance in order to reconcile this with the
perceptible differences existing when the
succession was viewed from a single point of
time, so we are caused by the causal related-
ness of our perceptions to regard them as
identical and fall into the notion of a soul
or self in an attempt to reconcile this with
their variation, which makes their non-iden-
tity evident. (Of course, no one ever thinks
that distinguishable perceptions are iden-
tical, but it is our inclination to think this,
together with our recognition that it cannot
be so which, according to Hume, explains
our belief in a self.)
Thus, unlike our vulgar or everyday belief
in body, our everyday belief in a self is not
a consistent, stable, and merely contingently
false one, which can only be revealed to be
false by reflection on experiments to which
the non-philosophical (in particular, the ‘hon-
est gentlemen’ of England (THN 1.4.7.13 /
272)) may not be inclined. Like the ancient
philosophers’ belief in material substrata, it
is an inevitably doomed attempt to reconcile
the irreconcilable.
It needs no argument that Hume’s sup-
plementary account of our belief in the sim-
plicity of the self (THN 1.4.6.23 / 263) is to
be compared to his account of the ancient
philosophers’ belief in the simplicity of mate-
rial substrata since it is just a summary of
the account of that given in section III (THN
1.4.3.5 / 221).
6. THE LABYRINTH
In Book I Hume is confident that despite
the contradictions and difficulties that exist
178
9780826443595_Ch07_Final_txt_print.indd 178
in every system concerning external objects,
and the idea of matter, ‘the intellectual world
. . . is not perplexed with any such contra-
dictions as those we have discovered in the
natural’ (THN 1.4.5.1 / 232). By the time
of the Appendix this confidence has disap-
peared. The topic of personal identity is now
declared a ‘labyrinth’, and two principles are
identified which he can neither render con-
sistent nor renounce, viz. that all our distinct
perceptions are distinct existences, and that
the mind never perceives any real connection
among distinct existences.
Hume’s identification of his difficulty is
brief and inaccurate (since the principles he
labels inconsistent are in fact consistent). The
most a commentator can do is to identify a
flaw in his discussion of personal identity of
which he plausibly could have become aware
after completing Book I and which it is con-
sistent with the discussion in the Appendix
to suppose he might have had in mind as the
cause of the trouble.
It appears that what worries Hume is
some defectiveness in his explanation of our
belief in personal identity, and specifically
in his explanation of our initial tendency to
attribute identity to what are in fact distinct
perceptions:
Having thus loosen’d all our particular
perceptions, when I proceed to explain
the principle of connexion, which binds
them together, and makes us attribute to
them a real simplicity and identity, I am
sensible, that my account is very defec-
tive, and that nothing but the seeming
evidence of the precedent reasonings
cou’d have induced me to receive it.
(THN App. 20 / 635)
Again: ‘But all my hopes vanish, when I
come to explain the principles that unite our
2/9/2012 9:35:29 PM

successive perceptions in our thought and
consciousness. I cannot discover any theory,
which gives me satisfaction on this head.’
(THN App. 20 / 635)
As to Hume’s talk of inconsistency: the
obviously consistent principles he declared
to be inconsistent, but ones he cannot
renounce, obviously entail that the mind
never perceives any real connection among
its perceptions. Presumably Hume thinks
this inconsistent with the possibility of
explaining our attribution of identity to dis-
tinct perceptions.
Taking this focus on explanation into
account, and by the standards indicated
above for the satisfactoriness of an inter-
pretation of Hume’s discussion, a sugges-
tion as reasonable as any other is that Hume
has here come to realize that in this case his
appeal to the identity-ascribing mechanism
cannot explain the irresistible and undeni-
able belief in personal identity he thinks
we all have. As we have seen, our belief in
an enduring self, in Hume’s account, is on
a par with the ancient philosophers’ belief
in substrata. Like the latter, then, it should
easily be subverted by a due contrast and
opposition (THN 1.4.4.1 / 225). In fact,
in Hume’s account our belief in personal
identity should be even less secure than the
belief in material substrata. For, as Hume
emphasizes, in this case the crucial relation
between perceptions which causes us to
pass easily from one to another, and con-
sequently to attribute to them an identity,
is cause and effect. But nowhere else are
we disposed to identify cause and effect –
even if we are disposed to add a new rela-
tion to objects perceived as related by cause
and effect as a result of our disposition to
complete the union (THN 1.4.5.12 / 237),
we are not disposed to add the relation of
179
9780826443595_Ch07_Final_txt_print.indd 179
THE SELF AND PERSONAL IDENTITY
identity to causes and effects we perceive as
distinct in space and time and non-resem-
bling, that is, as possessing neither of the
crucial qualities in his idea of identity. In his
attempt to prove that ‘all objects, to which
we ascribe identity, without observing their
invariableness, and uninterruptedness, are
such as to consist of a succession of related
objects’ (THN 1.4.6.7 / 255), Hume identi-
fies various special features of such succes-
sions which dispose the mind to a smooth
and easy transition – a small and incon-
siderable change in a part in proportion to
the whole, gradual and insensible change,
a combination of parts to a common end,
and a sympathy or reciprocal relation of
parts, a specific identity or resemblance and
a change that is natural and essential and
hence expected. But he makes no attempt at
all to show that such features are present
in the succession of different perceptions
which ‘succeed each other with an incon-
ceivable rapidity and are in a perpetual flux
and movement’ (THN 1.4.6.4 / 252) that
we think of as ourselves. He simply insists
that, since the true idea of the human mind
is to consider it as a system of different
perceptions or different existences linked
together by the relation of cause and effect,
our ascription of identity to our distinct per-
ceptions must be caused by the ease of tran-
sition between causally related perceptions.
But this is merely an attempt to shoehorn
an account of our belief in personal iden-
tity into the explanatory framework he has
appealed to in Sections II and III of Part IV
of Book I. I suggest that in the Appendix
Hume has come to recognize the implaus-
ibility of this attempt and specifically, of its
necessary first stage, the explanation of our
disposition to regard distinct perceptions as
themselves identical.
2/9/2012 9:35:30 PM
 THE SELF AND PERSONAL IDENTITY

NOTES in A. Flew (ed.), Body, Mind and Death (New

York: Macmillan, 1964); T. Reid, Essays on the
1
J. Locke, An Essay concerning Human Intellectual Powers of Man, ed. A.D. Woozley
Understanding, ed. J. Yolton (London: Dent, (London: Macmillan, 1941); G.W. Leibniz,
1961), II.xxvii.7. New Essays on Human Understanding, trans.
2
Locke, Essay, II.xxvii.9. and ed. P. Remnant and J. Bennett (Cambridge:
3
Ibid. Cambridge University Press, 1981).
4
See J.Butler, ‘Of Personal Identity’, First Disser- 5
R.M. Chisholm, Person and Object (London:
tation to the Analogy of Religion [1736], repr. Allen and Unwin, 1976), p. 39.

180

9780826443595_Ch07_Final_txt_print.indd 180 2/9/2012 9:35:30 PM

 8
‘ALL MY HOPES VANISH’:
HUME ON THE MIND
Galen Strawson
1. ‘THE ESSENCE OF THE MIND [IS]
UNKNOWN’
Hume holds that ‘the essence . . . of external
bodies’ is unknown, and that ‘the essence of
the mind [is] equally unknown to us with that
of external bodies’ (THN Intro. 8 / xvii). His
aim in section 1.4.6 / 251–63 of his Treatise,
which addresses the question of the nature of
the mind (after having dismissed the traditional
debate between the materialists and immateri-
alists in the preceding section), is accordingly
modest. It is, first, to provide an account of
(1) the content of the empirically warranted
idea of the mind (or self or person), given that
we cannot know the essence of the mind,
and, second, to provide a causal psycho-
logical account of the origin of
(2) our belief in a single diachronically
persisting mind (or self or person), given that
the account of (1) turns out to show that no
idea of the mind as a single diachronically
persisting entity is empirically warranted.1
More particularly, it is to provide a psychological
account of how each of us individually comes to
181
9780826443595_Ch08_Final_txt_print.indd 181
believe that he or she is a single persisting person
or self or has a single persisting mind.
When Hume returns to this topic in the
Appendix to the Treatise (THN App. 10–21
/ 633–6), he can find no fault in his accounts
of (1) and (2). The consequence, in his own
words, is that his ‘hopes vanish’.
Why is this? He remains entirely happy
with his account of (2), and reaffirms it in the
Appendix (THN App. 20 / 635).2 The trouble
lies in his account of (1). But the trouble is not
that the account of (1) is wrong. On the con-
trary: Hume cannot see how it can be wrong,
on his empiricist terms, and he reaffirms it, too,
in the Appendix (THN App. 15–19 / 634–5).
The trouble is that his philosophy as a whole
makes essential use of a conception of mind
that is not empirically warranted, according
to (1). Since he is committed to an empiri-
cist approach, he needs to get more into his
account of the empirically warranted idea of
the mind. But he cannot. So his hopes vanish.
2. THE EMPIRICALLY WARRANTED
IDEA OF THE MIND
What is the empirically warranted idea of the
mind, the account of the mind at which any
2/9/2012 9:35:42 PM
 HUME ON THE MIND
philosophy that aims at clarity and distinct-
ness must aim, given that the essence of the
mind is unknown?3 Hume’s answer is plain:
the mind ‘as far as we can conceive it, is noth-
ing but a system or train of different percep-
tions’ (Abs. 28 / 657). We have no ‘notion of
. . . self . . . , when conceiv’d distinct from par-
ticular perceptions . . . we have no notion of
. . . the mind . . . , distinct from the particular
perceptions’ (THN App. 18–19 / 635).4
These are explicitly epistemologically
qualified statements of what has come to be
known as ‘the bundle theory of mind’. They
are claims to the effect that this is all we can
know of the mind. But we also find many
(mostly earlier) epistemologically unquali-
fied ontological formulations of the bundle
theory of mind. Minds or selves or persons
or thinking beings, Hume says, are ‘nothing
but a bundle or collection of different percep-
tions’ (THN 1.4.6.4 / 252). ‘They are the suc-
cessive perceptions only, that constitute the
mind’ (THN 1.4.6.4 / 253). A ‘succession of
perceptions . . . constitutes [a] mind or think-
ing principle’ (THN 1.4.6.18 / 260). It is a
‘chain of causes and effects, which constitute
our self or person’ (THN 1.4.6.20 / 262).5 A
‘composition of . . . perceptions . . . forms the
self’ (THN App. 15 / 634). A ‘train . . . of . . .
perceptions . . . compose a mind’ (THN App.
20 / 635). Hume could not be more plain:
‘what we call a mind, is nothing but a heap
or collection of different perceptions, united
together by certain relations’ (THN 1.4.2.39
/ 207). It is a ‘connected mass of perceptions,
which constitute a thinking being’ (THN
1.4.2.39 / 207), ‘a connected heap of percep-
tions’ (THN 1.4.2.40 / 207). It is a ‘succes-
sion of perceptions, which constitutes our
self or person’ (THN 1.4.7.3 / 265). ‘It must
be our several particular perceptions, that
compose the mind. I say, compose the mind,
not belong to it’ (Abs. 28 / 658).6
182
9780826443595_Ch08_Final_txt_print.indd 182
How should we take these remarks? Well,
Hume holds that ‘the perceptions of the
mind are perfectly known’ (THN 2.2.6.2 /
366), and he also holds that the essence of
the mind is unknown. So he does not intend
these outright ontological claims to be with-
out restriction, as stating the essence of the
mind. They are claims about the mind so
far as we have any empirically – and hence
philosophically – respectable knowledge
of it.7 They are claims about the maximum
legitimate content of any claims about the
nature of the mind that can claim to express
knowledge of the nature of the mind. Hume
is a sceptic, and a sceptic, even a moderate
sceptic like him, does not go around claiming
to have certain a posteriori knowledge of the
ultimate metaphysical nature of the concrete
constituents of the universe (other than per-
ceptions). He does not claim to have certain
a posteriori knowledge either of the essence
of the mind, or of the essence of objects.
– The outright ontological claims about
the mind are literally true when they are
made strictly within the philosophical
framework of ideas constituted by empir-
ically warranted (and hence clear and
distinct) ideas, since they just repeat the
definition of the empirically warranted
clear and distinct idea of the mind.
True. The point is then that this framework of
empirically warranted (and hence clear and
distinct) ideas is in Hume’s philosophy rightly
and crucially embedded within a larger scepti-
cal framework of ideas. In the larger sceptical
framework of ideas it is acknowledged that
there may be and indeed is more to reality
than what we can comprehend in empirically
warranted ideas, and words like ‘mind’ and
‘bodies’ are accordingly used in a larger sense:
‘the essence of the mind’ is ‘equally unknown
2/9/2012 9:35:42 PM

to us with that of external bodies’ (THN
Intro. 8 / xvii), and ‘the essence and compo-
sition of external bodies are so obscure, that
we must necessarily, in our reasonings, or
rather conjectures concerning them, involve
ourselves in contradictions and absurdities’
(THN 2.2.6.2 / 366).
3. WHY IS THE EMPIRICALLY
WARRANTED IDEA OF THE MIND
NOT ENOUGH FOR HUME?
Why is the empirically warranted idea of the
mind not enough for Hume? Before trying
to answer this question, we should consider
Hume’s statement of what he would need in
order to be able to put things right, now that
his hopes have vanished.
He is very straightforward about this: ‘Did
our perceptions either inhere in something
simple and individual, or did the mind per-
ceive some real connexion among them, there
wou’d be no difficulty in the case’ (THN App.
21 / 636). If we could appeal to the idea that
our perceptions inhered in something simple
and individual, all would be well. But in that
case we would of course have to have empiri-
cal warrant for the view that our perceptions
inhere in something simple and individual,
and that, Hume thinks (rightly, on his terms),
is something we will never have.
We would also be able to put things right
if we could perceive ‘some real connexion’
among our perceptions. For in that case the
idea of real connection among perceptions
would be empirically warranted (because
perceived), and could accordingly feature
as part of the empirically warranted idea of
the mind. But that, Hume says (again rightly,
on his terms), is something that will never
happen.
183
9780826443595_Ch08_Final_txt_print.indd 183
HUME ON THE MIND
4. WHAT DOES HUME NEED TO DO
THAT HE CANNOT DO?
If either of the two options specified in the
previous quotation were available, Hume
would be all right, given what he needs to do.
What does he need to do? He is very straight-
forward about this too. He needs to explain
something. What does he need to explain? He
states what he needs to explain in two differ-
ent ways in a single paragraph (THN App.
20 / 635–6). It is slightly confusing, because
he speaks of principles first in the singular
and then in the plural, although he has the
same thing in mind in both cases. He needs
‘to explain the principle of connexion, which
binds [our perceptions] together, and makes
us attribute to them a real simplicity and iden-
tity’ (THN App. 20 / 635). He needs in other
words ‘to explain the principles, that unite
our successive perceptions in our thought or
consciousness’ (THN App. 20 / 636).
One finds exactly the same shift between
the singular and the plural in the passage
about the mind (in particular the ‘imagina-
tion’) in which his problem originates, and to
which one must turn first when trying to say
what his problem was:8
nothing wou’d be more unaccountable
than the operations of [the imagina-
tion], were it not guided by some univer-
sal principles, which render it, in some
measure, uniform with itself in all times
and places. Were ideas entirely loose
and unconnected, chance alone wou’d
join them; and ’tis impossible the same
simple ideas should fall regularly into
complex ones (as they commonly do)
without some bond of union among
them, some associating quality, by which
one idea naturally introduces another.
This uniting principle among ideas . . .
(THN 1.1.4.1 / 10–11)
2/9/2012 9:35:43 PM
 HUME ON THE MIND
Why does he need to explain this principle,
or these principles, and what exactly does
he mean by ‘explain’? At the very least, he
means that he needs to make the existence of
these principles readily intelligible. Why so?
Well, for one thing, they are his fundamental
explanatory posits. The fact that the mind is
governed by these principles – the fact that
the mind is a Principle-Governed Mind – is
the fundamental explanatory posit of Hume’s
whole philosophy. He needs to explain the
Principle-Governed Mind – to make its exist-
ence readily intelligible.
As a sceptic, Hume is clear on the point
that he does not have to explain everything.
He is clear that many things about the nature
of the universe lie beyond the reach of human
understanding. In fact he is clear on the point
that he does not have to explain everything
about the very principles that are in question
when his hopes vanish. When we consider
these principles, he says, ‘the principles of
union or cohesion among our simple ideas’
(THN 1.1.4.6 / 12), we encounter
a kind of attraction, which in the
mental world will be found to have as
extraordinary effects as in the natural
. . . .[9] Its effects are every where con-
spicuous; but as to its causes, they are
mostly unknown, and must be resolv’d
into original qualities of human nature,
which I pretend not to explain.
He continues:
Nothing is more requisite for a true phi-
losopher, than to restrain the intemper-
ate desire of searching into causes, and
having establish’d any doctrine upon a
sufficient number of experiments, rest
contented with that, when he sees a far-
ther examination would lead him into
obscure and uncertain speculations. In
184
9780826443595_Ch08_Final_txt_print.indd 184
that case his enquiry wou’d be much
better employ’d in examining the effects
than the causes of his principle. (THN
1.1.4.6 / 12–13)10
Quite so. This is clear. But Hume is no less
clear on the fact that there is something
about these principles, about the Principle-
Governed Mind, that he needs to explain and
cannot explain; something that he needs to
make readily intelligible, and cannot.
5. WHAT HAS CAUSED HIS DIFFICULTY
TO ARISE IN THE FIRST PLACE?
What has caused his difficulty? Hume is
clear on this point too. It is the fact that he
has ‘loosen’d all our particular perceptions’:
‘But having thus loosen’d all our particular
perceptions, when11 I proceed to explain the
principle of connexion, which binds them
together, . . . I am sensible, that my account is
very defective . . .’ (THN App. 20 / 635). Why
is this loosening a problem? Because, he says,
in a passage already quoted:
were ideas entirely loose and uncon-
nected, chance alone wou’d join them;
and ’tis impossible the same simple ideas
should fall regularly into complex ones
(as they commonly do) without some
bond of union among them, some associ-
ating quality, by which one idea naturally
introduces another. (THN 1.1.4.1 / 10)12
It seems, then, that he should not have loos-
ened all our perceptions. If he had not, he
would not have this problem. But when he
asks what an empirically warranted idea of
the mind looks like, he finds he has to loosen
them. For all that is observable of the mind is a
2/9/2012 9:35:43 PM

‘train of perceptions’ with no observable con-
nection between them. This is his problem.
In another well-known passage, Hume
writes that ‘the true idea of the human mind,
is to consider it as a system of different per-
ceptions or different existences, which are
link’d together by the relation of cause and
effect and mutually produce, destroy, influ-
ence, and modify each other’ (THN 1.4.6.19
/ 261). Does he not here claim that he has the
‘real connexion’ he needs? No. This reference
to the relation of cause and effect brings in
nothing more than the empirically warranted
idea of cause and effect, and this, of course, is
not an idea of real connection.13 Hume is clear
on the point. The ‘true’, i.e. empirically war-
ranted, idea of the human mind, which is char-
acterized in terms of causal links, as above, is
an account of the mind in which ‘all our par-
ticular perceptions’ are ‘loosen’d’ (THN App.
20 / 635). He cannot explain the existence
of the principle/principles whose existence
needs to be explained if all he has are loose
perceptions – which are all his empirically
warranted account of the mind gives him. He
needs observable ‘real connexion’ between the
perceptions, or their inherence in ‘something
simple and individual’, which is in effect just
a particularly strong form of real connexion
(THN App. 21 / 636). That is what he says.
He could hardly be more clear on the point.
6. THE HEART OF THE PROBLEM
We can put Hume’s problem in a suitably
painful way as follows. He uses a certain
theoretical conception of the mind – the
Principle-Governed Mind described in THN
1.1.4 / 10–13 – with tremendous success in
THN 1.3.14 / 155–72 (‘Of the idea of neces-
sary connexion’), in his psychological account
185
9780826443595_Ch08_Final_txt_print.indd 185
HUME ON THE MIND
of how it is that we come to believe we experi-
ence causal necessity or power in the world,
and of how it is that we do this in spite of the
fact that the actual legitimate empirical con-
tent of our experience of causation in the
world contains no experience of causal neces-
sity or power in the world. He uses this con-
ception of the Principle-Governed Mind again
in THN 1.4.2 / 187–218 (‘Of scepticism with
regard to the senses’), and with equal suc-
cess, in his psychological account of how it
is that we come to believe in external objects
that continue to exist independent and unper-
ceived, and of how it is that we come to do
this in spite of the fact that this property of
continuous independent unperceived existence
cannot be part of our experience, or, therefore,
part of the actual legitimate empirical content
of our idea of an external object. His concep-
tion of the imagination – I will capitalize the
term ‘Imagination’ to mark it as Hume’s the-
oretical term – effectively contains his whole
conception of the Principle-Governed Mind.
It drives his account of how, given only
3. the actual empirically warranted content
of the idea of causation,
we none the less acquire
4. our belief that we experience causal power
in external objects,
as just remarked, and, equally, his account
of how, given only
5. the actual empirically warranted content
of the idea of an external object,
we none the less acquire
6. our belief in external objects.
And so far all is well and good.14 But when
he comes (in THN 1.4.6 / 251–63) to give his
account of how we acquire (2) our belief in
a single diachronically persisting mind, given
2/9/2012 9:35:43 PM
 HUME ON THE MIND
only (1) the actual empirically warranted
idea of the mind, he finds that (1), the actual
empirically warranted idea of the mind, does
not contain the materials – the machinery, one
might say – he needs to drive his account of
how we acquire (2). Nor does it contain the
machinery he needs to drive his accounts of
how we acquire (4) and (6). His account of
(1), the actual empirically warranted idea of
the mind, the only one he can use in his phil-
osophy, has deprived him of the Principle-
Governed Mind, the Imagination, which
drives all the most original parts of his philoso-
phy: the accounts of how we acquire (4), (6)
and (2). The Imagination cannot have any real
existence if all there is to the mind is a bundle
of perceptions. There is nowhere for it to be.
More moderately, and more precisely to the
point: we cannot make use of the notion of the
Imagination or the Principle-Governed Mind
in our philosophy, if our philosophy rules that
the only idea of the mind that is suitably clear
and distinct, and can therefore be legitimately
used in philosophy to make knowledge claims,
is the bundle theory of mind. No doubt what
we need exists. It exists in reality – it is the
‘essence of the mind’. But ‘the essence of the
mind [is] unknown’, and we cannot make use
of it in our empiricist philosophy.
In sum: when Hume comes to give his
empiricist account of the mind, his philoso-
phy shoots itself in the foot. That is why his
hopes vanish. His particular brand of empiri-
cism is unsustainable.15
In the first Enquiry he endorses the claim
that prompts his confession of failure in the
Appendix to the Treatise, when, in his only
direct reference to the abandoned problem,
he observes, briefly but decisively, that ‘it is
evident that there is a principle of connex-
ion between the different thoughts or ideas
of the mind’ (EHU 23 / 3.1). Here he refers
to a real connection of precisely the sort that
186
9780826443595_Ch08_Final_txt_print.indd 186
the empiricistically ‘true’ idea of the human
mind cannot countenance.16 This is the very
same ‘principle of connexion’ that he refers
to in the Appendix, the principle of connex-
ion that, he says, makes ‘my hopes vanish
. . . when I proceed to explain’ it (THN App.
20 / 635). He cannot explain it in the terms
of his empiricist theory of ideas. In turn-
ing his attention to our idea of the mind in
THN 1.4.6 / 251–63 – after having devoted
detailed attention to our idea of causation
in 1.3.14 / 155–72, and our idea of external
bodies in 1.4.2 / 187–218 – Hume finds him-
self obliged to deprive himself of ‘the unit-
ing principle’ (THN 1.1.4.1 / 10) that he had
relied on in 1.3.14 and 1.4.2: ‘the principle
of connexion, which binds . . . all our par-
ticular perceptions together’ (THN App. 20
/ 636). It cannot be part of the empirically
warranted idea of the mind. It cannot be part
of the empirically warranted idea of the mind
whether it is thought of as a ‘real connexion’
of some sort or as something ‘simple and
individual’, because neither of these things is
given in experience.17
7. THE UNANSWERABLE OBJECTION?
Some may object that Hume does not really
have the problem he thinks he has. I think he
is not so foolish as to mistake his own situa-
tion, but the objection is worth addressing,
and I will do so in due course. First, I will
restate the issue.
We can start with the two fundamental
theoretical principles that Hume says he can-
not renounce (THN App. 21 / 636). I will
number them ‘(P1)’ and ‘(P2)’:
(P1) ‘all our distinct perceptions are dis-
tinct existences’ (so far as we know)
2/9/2012 9:35:43 PM

and
(P2) ‘the mind never perceives any real
connexion among distinct existences’.
These principles are very clear and familiar
to any reader of Hume.
In the next sentence he states two options,
already noted, either of which would in his
opinion entirely solve his problem. The first
is that
(O1) ‘our perceptions . . . inhere in some-
thing simple and individual’.
The second, to spell it out a little, is that
(O2) our perceptions are distinct exist-
ences, and ‘the mind perceive[s] some
real connexion among them’.
Why would either of these two options solve
his problem? Because both can sufficiently
ground a fundamental theoretical commit-
ment of his philosophy – his commitment to
the real existence and operation of something
he has just mentioned,
(P3) the principle of connexion, which
binds . . . our particular perceptions . . .
together (THN App. 20 / 635),
or (in its plural version)
(P4) the principles, that unite our succes-
sive perceptions in our thought or con-
sciousness (THN App. 20 / 636),
most centrally, the principles of the associ-
ation of ideas, the Resemblance principle,
the Contiguity principle, and the Cause and
Effect principle. The trouble is that both
options are ruled out for him. They have
no empirical warrant. They are philosoph-
ically inadmissible from his empiricist point
187
9780826443595_Ch08_Final_txt_print.indd 187
HUME ON THE MIND
of view. The ideas of real or objective unity
or connection that the options appeal to are
conceptually clear. They are ‘perfectly dis-
tinct’ as far as they go.18 But they are empir-
ically ‘unintelligible’ in their application to
concrete reality. So Hume cannot appeal
to them in his account of the nature of the
mind. His hopes vanish.
This may not be the best way to put the
point. It may be better to say that what hap-
pens in the Appendix is that Hume realizes
that he has no answer, given his overall theory,
to an objection that begins by citing the two
options he mentions and then challenges him
to deny that at least one of them (or perhaps
their disjunction) is in effect built into what
he means by ‘mind’, and is therefore built into
what he is really taking to be the ‘true idea’ of
the mind, although they are excluded from his
official account of the ‘true idea’ of the mind.
According to this account, Hume real-
izes that he faces what one might call the
Unanswerable Objection:
– Your philosophy taken as a whole
commits you, Hume, to a view forbid-
den by your philosophy. More precisely,
it commits you to a choice between one
of two views. The first is that the self or
mind is something like an ontologically
(substantially) simple and individual
persisting something, in which succes-
sive perceptions (i.e. experiences) inhere
in such a way that it is not problematic
that they are connected or united in the
way you take them to be. If you reject
this – as you must on your empiricist
principles, given that we have (among
other things) no warrant for believing
in or appealing to the existence of any-
thing that lasts longer than the duration
of a single fleeting perception – and hold
instead that the mind is something onto-
logically (substantially) multiple, you
are no better off. For then you must hold
2/9/2012 9:35:43 PM
 HUME ON THE MIND
that the mind is something whose exist-
ence involves ‘real connexion’ (and this is
something that you have in effect already
done), and also, crucially, given your
own fundamental empiricist principles,
that this real connection is empirically
knowable, experienceable or perceivable
by us. You must hold this because the
idea of real connection is built into the
conception of mind you make use of in
your philosophy, so it must be empiric-
ally justified in order to be licensed for
use. But you must also reject the view
that it is empirically knowable, given
those same fundamental empiricist prin-
ciples, and you have indeed done so.
It seems that Hume agrees. He puts forward
exactly the same two metaphysical options,
(O1) and (O2), one straightforwardly onto-
logical, the other ontological/epistemo-
logical, and then says that either would solve
his problem, but that he cannot have either.
It may be protested that Hume cannot
really be saying this, because he takes the
idea of a persisting, simple and individual
something and the idea of real connection
to be ‘unintelligible’ tout court. But he is
saying this. Even those who want to reject
the quotation in note 18 from THN 1.4.6.6
/ 253, in support of the claim that Hume
thinks that both these two ideas are ‘perfectly
distinct’, must concede that he is taking these
two ideas to be sufficiently intelligible to be
available for use in an informative descrip-
tion of a situation in which he would not
face the philosophical difficulty he feels he
does face.19 For Hume, ‘unintelligible’ means
‘not understandable’, ‘incomprehensible’. It
does not mean ‘incoherent’, and so necessar-
ily non-existent, as it standardly does today
in philosophy. It means ‘not such that it has
any empirically warrantable applicability to
concrete reality’, hence ‘not clear in such a
188
9780826443595_Ch08_Final_txt_print.indd 188
way that it can be appealed to in empiricist
philosophy’.20
Some students of Hume have difficulty with
the idea that he makes any use at all of the idea
of objective or real connection. This is under-
standable, at least at first, but Hume’s position
is clear: if (once again) our ideas were
entirely loose and unconnected, chance
alone wou’d join them; and ’tis impos-
sible the same simple ideas should fall
regularly into complex ones (as they
commonly do) without some bond of
union among them, some associating
quality, by which one idea naturally
introduces another.21
In fact, ‘the same simple ideas [do] fall regularly
into complex ones’, and one idea ‘naturally
introduce[s]’ another (THN 1.1.4.2 / 11). This
is what actually happens, and it cannot pos-
sibly happen, Hume says, unless there exists, as
a matter of objective fact, a ‘bond of union’ – a
‘uniting principle’, ‘principles of union or cohe-
sion’ – among our ideas (THN 1.1.4.1 / 10,
1.1.4.6 / 13). The ‘causes’ of this phenomenon
are, he says, ‘mostly unknown, and must be
resolv’d into original qualities of human nature,
which I pretend not to explain’.22 But this is not
to say that these principles of cohesion are not
real. On the contrary, they are indeed real. All
we can know of them are the observable regu-
larities to which they give rise. Garrett states
the general point robustly:
Hume is not forbidden by his empiricist
principles from postulating the existence
of unperceived deterministic mechanisms
that would underlie the propensities of
perceptions to appear in particular ways.
He is forbidden by his principles only
from trying to specify the nature of those
mechanisms [in a way that goes] beyond
what experience can warrant.23
2/9/2012 9:35:44 PM

So it is that when he is discussing causal
necessity, Hume says that he is
indeed, ready to allow, that there may
be several qualities both in material and
immaterial objects, with which we are
utterly unacquainted; and if we please to
call these power or efficacy, ’twill be of
little consequence to the world. But when,
instead of meaning these unknown qual-
ities, we make the terms of power and
efficacy signify something, of which we
have a clear idea, and which is incom-
patible with those objects, to which we
apply it [it is incompatible because it is
just a feeling or impression], obscurity
and error begin then to take place, and
we are led astray by a false philosophy.
(THN 1.3.14.27 / 168)24
8. GARRETT’S OBJECTION
Many Hume commentators would reject
Garrett’s claim about Hume in this passage.
I accept it, for reasons given at the end of sec-
tion 2 above, in spite of the fact that it sug-
gests the following objection to the present
account of Hume’s problem. Look, as you
say, Hume holds that ‘the essence of the mind
[is] unknown’ (THN Intro. 8 / xvii). He takes
it for granted, in his philosophy considered
as a whole, that there is in fact something
more to the mind than just a series or bun-
dle of perceptions. He never endorses the
bundle theory of mind as the truth about the
ultimate nature of the mind. This is all true.
But this ‘something more’ is not a problem
for him, contrary to what you suggest. It is
not a problem for him because he can treat
it in the same way in which he treats many
other things – as something not further expli-
cable by us, something ‘mostly unknown’
that ‘must be resolv’d into original qualities
189
9780826443595_Ch08_Final_txt_print.indd 189
HUME ON THE MIND
of human nature, which I pretend not to
explain’ (THN 1.1.4.6 / 13), something that
is part of the unknown essence of the mind,
something ‘wonderful and unintelligible’ (i.e.
not understandable) by us, as he says of rea-
son (THN 1.3.16.9 / 179), something ‘magi-
cal’, as he says of the Imagination (THN
1.1.7.15 / 24).
This is a good objection, and it is raised
explicitly, in one form, by Garrett, so I will
call it ‘Garrett’s objection’.25 The principal
difficulty for it can be put by saying that it
seems to be an objection that must be put
to Hume himself, because it is Hume himself
who so plainly says that he has the problem
that he does not have if Garrett’s objection
is correct.
When exactly do Hume’s hopes vanish?
They vanish when he comes ‘to explain (P3)
the principle of connexion, which binds . . .
our particular perceptions . . . together’, ‘to
explain (P4) the principles, that unite our
successive perceptions in our thought or
consciousness’ (THN App. 20–1 / 635–6).
What does he mean by ‘explain the prin-
ciple of connexion which binds’ or ‘explain
the principles that unite’? What failure of
explanation does he have in mind? He tells
us. It is his failure to explain the principles of
connection that make us ‘attribute . . . a real
simplicity and identity’ to our perceptions. It
is his failure, in other words, to explain the
existence and operation of the principles of
the Imagination – I will call the whole set of
them the I-Principles, for short – that lead
us to come to believe in a single continuing
mind or self or subject. The problem, as he
sees it, is that he cannot make use of the fact
of the existence of the I-Principles in his phil-
osophy without thereby appealing to – or
rather, without being open to the charge that
he thereby appeals to – something he can-
not appeal to (O1), the idea of the mind as
2/9/2012 9:35:44 PM
 HUME ON THE MIND
‘something simple and individual’, or (O2),
some perceivable or experienceable ‘real con-
nexion’ between perceptions.26
The lamented failure, then, is a failure to
explain the existence of the I-Principles given
the resources of the empirically warranted
account of the mind. It is not a failure to
explain how the I-Principles – in particular
the Resemblance principle and the Cause and
Effect principle – lead us to come to believe
in a persisting mind or self.
9. COULD HUME’S PROBLEM BE
THE ‘PROBLEM OF DETAIL’?
Many have supposed that this is the lamented
failure. Subtle philosophers have done so. So
let us grant for the moment that the ‘when
I come to explain . . .’ passage can be read
in this way, at least when it is taken in isola-
tion from the rest of the text. Hume, then, is
despairing of his account of how the idea of
a persisting self arises in us, on the grounds
that the I-Principles (in particular the
Resemblance and Cause and Effect principles
of association) cannot really do the job.
I will say that on this view Hume’s prob-
lem is the Problem of Detail. Is this a defens-
ible interpretation (loss of all hope seems
a strangely extravagant reaction to such a
problem)? The way to find out is to look at
what he thinks might solve his problem.
But now we are back on the track already
laid out in section 3. One thing that will do
the trick is (the right to appeal to) the exist-
ence of (O1), ‘something simple and individ-
ual’ in which our perceptions inhere; another
is (O2), some perceivable or observable ‘real
connexion’ between perceptions,27 perceiv-
able in such a way that the deployment of the
idea of it in one’s philosophy when treating
190
9780826443595_Ch08_Final_txt_print.indd 190
of concrete reality is empirically warranted.
Both (O1) and (O2) will provide Hume with
the resources to explain the thing he has just
said he cannot explain. This, however, rules
out the view that his problem is the Problem
of Detail. For neither (O1) nor (O2) can help
the Resemblance and Cause and Effect princi-
ples of the association of ideas do their job in
explaining an unwarranted belief in a persist-
ing mind or self (for a doubt, see Strawson,
The Evident Connection, p. 137).
Again this is a somewhat backwards way
to put the point. A better way to put it, per-
haps, is to say that what destroys Hume’s
hopes is his realization that he cannot meet
the objection that he has in effect appealed
to one of (O1) and (O2) in placing the
I-Principles at the very centre of his theory
of human nature, in making them the great
engine of his philosophy. He has in effect
appealed to one of (O1) and (O2) although
he cannot appeal to either on his own terms.
One could put the point by saying that the
passage parses like this: ‘. . . when I proceed
to explain [the principle of connexion, which
binds them together, and makes us attrib-
ute to them a real simplicity and identity]; I
am sensible, that my account is very defect-
ive . . .’. The noun-clause inside the square
brackets denotes his problem, the phenom-
enon that is to be explained (=[P3/P4]). His
problem is to explain the phenomenon that
consists in the mind’s operating according
to the I-principles, not to explain how the
I-principles (in particular Resemblance and
Causation), once in place, can generate the
belief in ‘real simplicity and identity’. For,
once again, the two ontological solutions to
his problem that he offers in the next para-
graph cannot be construed as a solution to
that problem (the Problem of Detail).
One could say that the word ‘explain’ is
misread. It does not mean ‘expound’ or ‘spell
2/9/2012 9:35:44 PM

out’ – expound or spell out the details of
how the principle of connection that makes
us attribute simplicity and identity to our
perceptions does its job. It means ‘account
for the existence of’: account for the exist-
ence and operation of the I-principles given
the resources of a strictly empiricist account
of the mind. Hume’s problem is to give an
account of the existence of the principle of
connection that does what it does, given his
commitment to the view that the bundle view
is for philosophical purposes ‘the true idea of
the human mind’. This cannot be done.28
10. ‘EXPLAIN’?
– No. Hume can and does treat the phe-
nomenon of the existence and opera-
tion of the I-Principles in the same way
that he treats other things, as something
not further explicable by us, something
‘mostly unknown’ that ‘must be resolv’d
into original qualities of human nature,
which I pretend not to explain’ (THN
1.1.4.6 / 13). The very fact that Hume
uses the word ‘explain’ in the two crucial
passages – ‘when I proceed to explain the
principle of connexion, which binds them
together’ and ‘when I come to explain
the principles, that unite our successive
perceptions in our thought or conscious-
ness’ – proves that your interpretation
cannot be right. For your interpreta-
tion requires us to suppose that Hume is
lamenting his inability to explain some-
thing that he has repeatedly said he can-
not and does not need to explain.
This is Garrett’s objection, somewhat
extended. The best thing to do by way of
reply, I think, is to start by considering points
of agreement.
191
9780826443595_Ch08_Final_txt_print.indd 191
HUME ON THE MIND
In operating in a way that is correctly
described by the I-principles, the mind deliv-
ers all sorts of unity-and-connection expe-
riences, which we may call UC experiences
for short. It delivers persisting-physical-
object unity-and-connection experiences,
it delivers necessary-causal-relation unity-
and-connection experiences, and it delivers
persisting-individual-self unity-and-connec-
tion experiences. Is the existence of such
experiences problematic for Hume? Not at
all. He can fully explain the fact that we nat-
urally believe in these sorts of unity and con-
nection, even if our basic experience consists
of nothing more than a series of distinct and
fleeting perceptions, by appeal to the idea
that the mind operates according to certain
principles – the I-principles – that generate
such (‘fiction’-involving) UC experiences. We
cannot, however, explain the undoubtedly
real phenomena to which we refer when we
speak of the operation of the mind in accord-
ance with the I-principles by reference to the
operation of the mind in accordance with the
I-principles, any more than we can use logic
to prove the validity of logic. So the fact that
the mind operates in accordance with the
I-principles must be taken as a given (exactly
as the conformity of physical phenomena to
Newton’s law of gravity is taken as a given).
Right. This is what Hume does. The causes
of the mind’s operation in accordance with
the I-principles are, he says, ‘mostly unknown,
and must be resolv’d into original qualities
of human nature, which I pretend not to
explain’.29 He could hardly be more clear:
‘to explain the ultimate causes of our mental
actions is impossible’ (THN 1.1.7.11 / 22).
Much is unknown, then, and must remain
so. So far Hume, Garrett and I fully agree.
And Garrett and I also agree – contrary to
a cloud of commentators – that Hume does
in fact appeal to real connections throughout
2/9/2012 9:35:44 PM
 HUME ON THE MIND
Book I of the Treatise in appealing as he
does to the I-Principles – to the ‘uniting prin-
ciple’ or ‘bond of union’ that exists – the
‘uniting principles’ that exist – between our
different perceptions.30 It is also plain that,
pre-Appendix, Hume thinks that he can do
this with impunity within his empiricist phil-
osophy, because he can comfortably con-
sign that in virtue of which the I-Principles
exist to the ‘unknown essence of the mind’
in a thoroughly and indeed quintessentially
Newtonian spirit.
11. REPLY TO GARRETT’S OBJECTION
It is at this point that the agreement ends.
For I think, as Garrett does not, that Hume’s
hopes vanish when he sees that there is an
objection from which this large confession of
ignorance – this affirmation of ignorance –
cannot protect him.31 The affirmation of
ignorance sweeps up almost everything, but
it leaves a hole. Hume’s position is vulnerable
to the charge that if one relies on (P3/P4) – if
one relies on the idea of the mind’s operation
in accordance with the I-principles – then one
is obliged to accept that one of the two maxi-
mally general positive metaphysical charac-
terizations of the mind’s nature ([O1] or [O2])
must apply. But to accept this is to accept that
one must allow the applicability of terms
that are ‘unintelligible’ by Hume’s empiricist
principles (P1) and (P2). This is how (P3/P4),
(P1) and (P2), and (O1) and (O2) relate.32
Acknowledgement that one of maximally
general (O1) or (O2) must be the case is com-
patible with vast ignorance of the nature of
things, but Hume needs one of them, for he
will not otherwise be able to ‘explain’ the
I-Principles ([P3/P4]) in the following highly
general sense: he will not be able to account
192
9780826443595_Ch08_Final_txt_print.indd 192
for how they exist at all. So he will not be
able to make use of the idea that they exist.
But they are central to his philosophy.
Note that to give an explanation of some-
thing X in this sense, to give an account of
things that makes room for the bare fact of
X’s possibility, is not to attempt any further
detailed explanation of X of the sort Hume
thinks is impossible and is happy to leave as
unknown.
There is a moment when it dawns sharply
on Hume that he has a problem. He realizes
that the maximally general objection that
one of (O1) and (O2), at least, is needed, and
must in effect be allowed, given his account
of the mind, can be most powerfully pressed
against him. I suspect that it was the idea of
others coming up with this objection that was
most vivid for him as he wrote the Appendix.
One thing he then wanted to do, most under-
standably, was to be the first to make the
criticism (compare Wittgenstein’s assault on
his earlier position). His best defence was to
show complete candour and to be the first
to describe the fork – the either-a-single-
thing-or-perceivable-real-connection fork –
on which others would seek to spike him.
Imagine how you yourself would feel, and
what you might wish to do, if you discovered
a serious difficulty in your just published and
cherished theory. You would sit down and
do something comparable to what Hume did
when (probably hastily) he added the pas-
sage on personal identity to the Appendix.
One could put the point by saying that the
existence and operation of the I-principles
mean that some metaphysical description of
the mind that Hume cannot avail himself of is
knowably applicable to the mind. He cannot
invoke the mind’s ‘unknown essence’, treat-
ing this as a kind of explanation-sink that
can absorb the whole difficulty, for – this is
the direct reply to Garrett’s objection – his
2/9/2012 9:35:44 PM

opponents can happily grant that of course
much must remain unknown, while continu-
ing to insist that Hume has, in appealing to
the I-Principles, invoked something – some
sort of genuine metaphysical connection
and continuity among the perceptions of the
mind – whose existence he can make sense of
only on one of two specific conditions, nei-
ther of which is available to him.
It is hardly impressive (it is hopeless) for
him, faced with such an objection, to answer
again that much is ‘unknown’, ‘magical’,
‘unintelligible’, ‘wonderful’ and ‘inexplica-
ble’. ‘Yes, yes,’ his objectors reply in turn, ‘we
agree. The point we wish to make is much
more general (it is, in twentieth-century par-
lance, a “logical” point). In relying on the
I-Principles as you do you take a metaphysical
step you cannot take, given that you want to
give an empiricist account of the mind as well
as everything else. You incur a certain gen-
eral metaphysical debt you cannot repay on
your own empiricist principles. You cannot
rely on the I-Principles as you do and simply
refer everything else to the unknown essence
of the mind, for you cannot stop someone
replying that your reliance on the I-Principles
entails that there is at least one thing that can
be known about the essence of the mind and
that you cannot allow to be known. The thing
in question is in fact an either-or thing ([O1]
or [O2]), but that does not help. You can-
not allow this either-or thing to be known,
because it is not possible to specify what it is
without employing terms whose employment
you cannot allow, given your brand of empir-
icism, when it comes to making knowledge
claims about the nature of concrete reality.’
‘Specifically, and once again, your reli-
ance on the I-Principles entails that the fol-
lowing high-level, either-or description of the
essence of the mind – “persisting individual
single thing or really connected plurality of
193
9780826443595_Ch08_Final_txt_print.indd 193
HUME ON THE MIND
things” – can be known to apply. You cannot
make room for this because you cannot allow
any empirical meaning or (therefore) concrete
applicability to any idea of anything whose
description entails that it lasts longer than a
single fleeting perception. A fortiori you can-
not admit that any such idea has an indispen-
sable employment in your philosophy, or that
your philosophy presupposes that such an
idea has valid application. But it does. Your
philosophy entails – we are hammering the
point – that we can know at least one thing
more about the essence of the mind than you
say we do or can: we can know something
that we cannot and must not claim to know
on your empiricist principles. How else can
it possibly be the case that perceptions come
clumped in interacting groups as they do?’
To this Hume thinks, quite rightly, I believe,
that he has no effective reply. He cannot say
what he actually believes, given the dialect-
ical context of his discussion of personal
identity. He cannot say that the brain sup-
plies all the needed real continuity. And even
if he did, this would not diminish his need to
acknowledge real connection, for the brain
is certainly not a simple substance (which is,
after all, a property reserved to individual
atoms and immaterial souls).
12. A FINAL RESPONSE
– You are seriously underestimating
Hume’s resources. He is ‘not forbidden
by his empiricist principles from postu-
lating the existence of unperceived deter-
ministic mechanisms that would underlie
the propensities of perceptions to appear
in particular ways. He is forbidden by
his principles only from trying to spe-
cify the nature of those mechanisms [in a
way that goes] beyond what experience
2/9/2012 9:35:44 PM
 HUME ON THE MIND
can warrant’. But he does not try to do
this, in the case of the mind, nor does he
think he needs to. He is, again, happy to
say that what you call the ‘I-principles’
are unintelligible, inexplicable and won-
derful. He has, therefore, no problem of
the sort you describe.
This is another version of Garrett’s objec-
tion, mostly in his own words.33 I think I have
answered it. Hume does not think he can
plausibly reject the objection that he is com-
mitted to something like (O1) or (O2), caught
in a fork according to which one at least of
(O1) or (O2) is correct (he is caught because
it is a maximally general and exhaustive fork).
(O1) and (O2) are very general, but when we
consider the mechanisms to which Hume can
legitimately appeal, while holding them to be
unknown, we see that (O1) and (O2) already
‘specify the nature [or ground] of those
mechanisms’ in a metaphysical way that goes
‘beyond what experience can warrant’.
In conclusion, let me repeat the earlier sug-
gestion that Garrett’s objection has to be put
to Hume himself, because it is Hume himself
who thinks he has a problem that could be
entirely solved if he were allowed to make use
of the idea of a simple individual substance,
or the idea of (empirically observable) real
connections. This is the fundamental fixed
point, when it comes to the interpretation of
the Appendix. It is Hume himself who judges
(sees) that he is in effect committed, in his
philosophy, to the allowability of at least one
of two very high-level metaphysical descrip-
tions of the nature of the mind that can have
no empirical warrant and are therefore offi-
cially excluded from any role in his philoso-
phy. It is Hume himself who thinks that his
empiricism allows him to ignore (delegate to
the unknown, be agnostic about) all ques-
tions about the ultimate causes or sources
194
9780826443595_Ch08_Final_txt_print.indd 194
of the patterns in our experiences that lead
us to come to believe in physical objects and
causal necessity, but does not allow him to
do this when it comes to the mind itself. It
is Hume himself who thinks he has a prob-
lem he cannot solve even after he has stressed
the unintelligibility and inexplicability (and
‘wonderfulness’ and ‘magicality’) (THN
1.3.16.9 / 179, 1.1.7.15 / 24) of the work-
ings of the mind – the mind whose principles
of working are the great and indispensable
engine of his whole empiricist programme –
and who (again) thinks that he could solve
the problem immediately if the principles of
his philosophy allowed him to deploy the
notion of a simple and individual substance,
or to make empirically warranted use of the
notion of real (non-‘fictional’, non-Imagina-
tion-generated) connections. The burden on
those who favour what I am calling Garrett’s
objection is to explain why Hume feels he
has a problem he could solve if he could
appeal to a persisting individual substance or
make use of an empirically warranted notion
of real connection. It is Hume himself who
believes himself to be in a Zugzwang – a
position where he would like to be able to
make no move but feels he is obliged to make
one (or admit that he has in effect already
made one).
Old interpretative impulses may resurge:
‘For Hume, the phenomenon of conformity
to the I-principles is brute regularity; there is
therefore no need or possibility of any further
explanation of any sort, however general, in
his scheme.’ But it is far too late in the day for
such a view of Hume, and there are two more
particular replies. First, his problem stems
from the fact that he has ‘loosen’d all our
particular perceptions’ (THN App. 20 / 635);
but this loosening would not cause a prob-
lem if he took it that the phenomenon of con-
formity to the I-principles were just a matter
2/9/2012 9:35:44 PM

of brute regularity. Secondly, a reply already
made. You have to contrapose: it is Hume
himself who insists that the phenomenon of
conformity to the I-principles does need some
further explanation or grounding, however
general, and who tells us that two things that
are completely unavailable to him would do
the trick: inherence in a single substance or
real – non-regularity-theory – connections. It
is not as if he wants to say any such thing,
appealing to notions whose use in philosophy
he has ruled out as ‘unintelligible’.34 It is just
that he believes (sees) that the objection that
he must admit some such thing is correct and
unanswerable. When he moved on from (3)
his empiricist account of the content of the
idea of causation in 1.3.14 / 155–72 of the
Treatise, and (5) his empiricist account of the
content of the idea of physical objects in 1.4.2
/ 187–218, and took (1) the empirically war-
ranted idea of the mind itself as his subject in
1.4.6 / 251–63, his general ‘reductive’ empiri-
cist account of the origin of our belief in the
objective continuities, persistences and con-
nections that we take ourselves to encounter
in experience was running beautifully. It was
watertight on its own terms, and it must have
seemed that it could not fail to deal also with
the apparent or experienced continuity of the
mind. And in a sense it did, and smoothly too:
it gave at least as good an account of the ori-
gin of (2) our idea of ourselves as enduring
selves or subjects as it did of the origin of (4)
our ideas of causal power and (6) our ideas
of physical objects (which is not to say that
it was in fact empirically psychologically cor-
rect). But it relied on something more than (1)
could supply. The whole system broke down
when it came to (1), the empirically warranted
idea of the mind. It is Hume himself – one
more time – who believes that his account
of the mind is ‘very defective’, indeed hope-
less, and it is Hume himself who believes that
195
9780826443595_Ch08_Final_txt_print.indd 195
HUME ON THE MIND
his problem would be immediately solved by
one of two metaphysical provisions that his
empiricist philosophy rules out.35
NOTES
1
In Book I of the Treatise Hume takes ‘person’
to have a merely mental reference, and uses
it interchangeably with ‘mind’ and ‘self’ (and
sometimes ‘soul’). See, for example, N. Pike,
‘Hume’s Bundle Theory of the Self: A Limited
Defense’, American Philosophical Quarterly
4 (1967), pp. 159–65: ‘when Hume uses the
term . . . “person”, he generally means to be
referring only to the mind’ (p. 161), at least in
Book I of the Treatise. See also THN 1.4.6.2
/ 251, 1.4.6.5 / 253, 1.4.6.20 / 262, ‘self or
person’; 1.4.6.17 / 260, ‘mind or thinking
person’.
2
Most commentators have thought his prob-
lem lies in his account of [2]. See J. Ellis, ‘The
Contents of Hume’s Appendix and the Source of
His Despair’, Hume Studies 32 (2006), pp. 195–
231, for an interesting recent defence of this
view. See also H. Noonan, this vol., pp. 178–9.
For the opposing view (other than this paper),
see, for example, B. Stroud, Hume (London:
Routledge & Kegan Paul, 1977) and, more
recently, D. Garrett, ‘Rethinking Hume’s Second
Thoughts about Personal Identity’, in J. Bridges,
N. Kolodny and W. Wong (eds), The Possibility
of Philosophical Understanding: Essays for Barry
Stroud (New York: Oxford University Press,
2011); G. Strawson, The Evident Connexion:
Hume on Personal Identity (Oxford: Oxford
University Press, 2011), sects. 3.4, 3.10.
Hume later comes to think that Kames
gives a better account of the origin of [2]
than he does. Reading a draft of Kames’s
Essays in 1746, Hume writes to Kames that
‘I likt exceedingly your Method of explain-
ing personal Identity as more satisfactory
than any thing that had ever occurr’d to me’
(NLH 20, 8). I suspect that Hume here means
Kames’s account of the origin of our idea of
or belief in a persisting self – ‘man . . . has an
original feeling, or consciousness of himself,
and of his existence, which for the most part
accompanies every one of his impressions
2/9/2012 9:35:45 PM
 HUME ON THE MIND
and ideas, and every action of his mind and
body’ (Lord Kames, ‘Of the Idea of Self and
Personal Identity’ in Essays on the Principles
of Morality and Natural Religion (Edinburgh:
Fleming, 1751), pp. 231–2) – if only because
Kames’s further remarks (e.g. ‘this conscious-
ness or perception of self is, at the same time,
9
of the liveliest kind. Self-preservation is eve-
ryone’s peculiar duty; and the vivacity of this
10
perception, is necessary to make us attentive
to our own interest’ (ibid., p. 232)) are very
close to Hume’s own published views in Books
11
2 and 3 of the Treatise (see, for example, THN
2.1.11.4 / 317; 2.1.11.8 / 320; 2.2.2.15–16 /
339–40; 2.2.4.7 / 354; 2.3.7.1 / 427).
3
Hume holds, of course, that only an empiricist
philosophy deals in clear and distinct ideas. He
uses ‘clear and distinct’ a couple of times in the
Treatise when discussing geometrical concepts
(THN 1.1.7.6 / 19, 1.2.4.11 / 43). He uses
12
‘clear and distinct . . . idea’ and ‘clear, distinct
idea’ in the Enquiry (see EHU 12.20 / 157,
12.28 / 164), also ‘clearly and distinctly’ (EHU
4.18 / 35); otherwise he uses ‘clear’ and ‘dis-
tinct’ separately. He also uses ‘clear and precise’
(Abs. 7 / 648, THN 1.3.1.7 / 52).
4
All mental occurrences are perceptions,
in Hume’s terminology – thoughts, sensa-
tions, emotions, ideas, and so on – and
they are all (by definition) conscious. The
word that now corresponds most closely to
Hume’s word ‘perceptions’, in this use, is
‘experiences’.
5
Here the ‘causes and effects’ are particular
13
perceptions.
6
THN 1.4.6.4 / 252, 1.4.6.4 / 253, 1.4.6.18 /
260, 1.4.6.20 / 262, App. 15 / 634, App. 20
/ 635, 1.4.2.39 / 207, 1.4.2.40 / 207, 1.4.7.3
/ 265, Abs. 28 / 658. Six of these eleven
quotations are from passages where Hume is
discussing something else, or summarizing the
view and stating it in a particularly compressed
form. See also THN 2.1.2.2 / 277.
7
On this see in particular E. Craig, The Mind
of God and the Works of Man (Cambridge:
Cambridge University Press, 1987),
14
pp. 111–20.
8
As Garrett remarks, the question of what
Hume’s problem was – the question of
what Hume thought his problem was – ‘has
received what is surely a far greater number
196
9780826443595_Ch08_Final_txt_print.indd 196
of distinct answers – well over two dozen,
even by a conservative count – than has any
other interpretive question about Hume’s
philosophical writings’ (‘Rethinking Hume’s
Second Thoughts’, p. 16). I do not think this
would have happened if the discussion had
started out from THN 1.1.4 / 10.
In speaking of attraction in the natural world
he is referring to gravity.
The reference to ‘simple’ in ‘the principles of
union or cohesion among our simple ideas’
could be dropped.
There is a footnote reference letter (‘a’)
attached to the word ‘when’ in Hume’s text.
The note refers the reader to ‘Vol. I. Page 452.
This falls wholly on p. 260 in the Selby-Bigge
edition, beginning with ‘. . . if disjoin’d by the
greatest difference’ (THN 1.4.6.16) and end-
ing with ‘. . . amidst all its variations’ (THN
1.4.6.18).
Note that the problem is not just that he can-
not give an account of the fact that ‘the same
simple ideas should fall regularly into complex
ones’. That is simply the first point he considers
in THN 1.1.4 / 10–13, ‘Of the connexion or
association of ideas’. He also needs to give an
account of the operation of the other asso-
ciations of ideas that the ‘uniting principle’
accounts for, the associations of ideas based on
resemblance, contiguity and causation, which
he sets out in the rest of the section. (I think
Hume may have turned to this section to guide
his words when sketching his difficulty in the
Appendix.)
As Garrett says, ‘by “real connexion” used
as a technical term, Hume means (at least) a
connection between two objects that is more
than simply an associative relation in the
imagination’ (Cognition and Commitment
in Hume’s Philosophy (Oxford: Oxford
University Press, 1997), p. 181). Hume’s
principal example of a ‘real connexion’ is
causal necessity realistically and naively
figured as something that exists quite
independently of any activity of the fiction-
generating ‘imagination’.
Putting aside the fact that Hume’s psychologi-
cal account is wholly inadequate, empirically
speaking. See, for example, Hermer, L. and
Spelke, E., ‘Modularity and Development:
The Case of Spatial Reorientation’, Cognition
2/9/2012 9:35:45 PM

61 (1996), pp. 195–232; Carey, S., The Origin
of Concepts (Oxford: Oxford University Press,
18
2009), chap. 3.
15
For independent proof that Hume takes
there to be more to the mind than percep-
tions, see, for example, G. Strawson, The
Secret Connexion: Causation, Realism, and
David Hume (New York: Oxford University
Press, 1989), p. 130; Strawson, The Evident
Connexion, sect. 2.6. (Briefly: call a simple
impression of A an A-impression, and a simple
idea of A an A-idea. According to Hume, an
19
A-idea can arise in my mind only if I have
already had an A-impression. What happens is
that ‘there is a copy taken by the mind, which
20
remains after the impression ceases; and this
we call an idea’ (THN 1.1.2.1 / 8). But if the
mind is just a bundle of distinct experiences
with no hidden content (their contents are
‘perfectly known’), then there is no possible
way in which this can happen. For where does
the A-idea ‘remain . . . after the impression
ceases’ – given that I then go on to experience
or think about, B, C, and many other things,
and have no conscious thought of A?)
16
See also EHU 5.14 / 50: ‘nature has estab-
21
lished connexions among particular ideas.
. . . These principles of connexion or asso-
ciation we have reduced to three, namely,
Resemblance, Contiguity and Causation;
which are the only bonds that unite our
22
thoughts together . . .’.
17
This proposed solution comfortably satisfies
four of the five criteria for a successful solution
that Garrett lists in ‘Once More Into the Laby-
rinth: Kail’s Realist Explanation of Hume’s
Second Thoughts about Personal Identity’,
23
Hume Studies 36(1) (2010), pp. 77–87: the
‘Crisis Criterion’, the ‘Origin Criterion’, the
24
‘Solution Criterion’, and the ‘Scope Criterion’.
It questions whether the fifth criterion – the
‘Difficulty Criterion’ – is correct, by propos-
25
ing that Hume did not have any difficulty
in stating his problem – although he could
certainly have been clearer. The explanation of
why Hume has been found obscure lies in the
preconceived ideas about Hume that readers
have brought to the Appendix. (The clause that
has been most damagingly misread is ‘when
I proceed to explain the principle of connex-
ion, which binds them together, and makes us
197
9780826443595_Ch08_Final_txt_print.indd 197
HUME ON THE MIND
attribute to them a real simplicity and identity’.
See sects. 8 and 9 below.)
‘We have a distinct idea of an object, that
remains invariable and uninterrupted thro’
a suppos’d variation of time; and this idea
we call that of identity or sameness. We have
also a distinct idea of several different objects
existing in succession, and connected together
by a close relation; . . . these two ideas of
identity, and a succession of related objects
[are] in themselves perfectly distinct’ (THN
1.4.6.6 / 253).
Compare the move he makes in THN 1.4.5 /
232–51, discussed in Strawson, The Evident
Connexion, p.50.
Craig has an incomparable discussion of this
issue; see in particular Craig, The Mind of
God and the Works of Man, pp. 123–30, a
decisive antidote to Millican’s unfortunate
recent attempts to re-equate Hume’s use of
‘unintelligible’ with ‘incoherent’ (see, for
example, P. Millican, ‘Hume, Causal Realism,
and Causal Science’, Mind 118 (2009),
pp. 647–712, pp. 647–8). See also Strawson,
The Secret Connexion, pp. 49–58 and
Strawson, The Evident Connexion, sect. 2.4.
THN 1.1.4.1 / 10; emphasis added. See
Strawson, The Evident Connexion, pp. 140–1
for a discussion of the point that these mental
connections involve only a ‘gentle force’, and
are not exceptionless.
THN 1.1.4.6 / 13. He follows Newton, who
states the law of gravitational attraction
while adding that ‘the cause of Gravity . . .
I do not pretend to know’ (Principia, trans.
A. Motte and F. Cajori (Berkeley: University of
California Press, 1934 [1687]), 3.240).
Garrett, Cognition and Commitment, p. 171;
emphasis added.
Notice the relative mildness of ‘obscurity and
error then begin’. Compare THN 2.3.2.4 /
409–10.
See Garrett, Cognition and Commitment, p. 171.
It’s raised as an objection to Stroud, Hume and
T. Beauchamp, ‘Self Inconsistency or Mere Self
Perplexity?’, Hume Studies 5 (1979), pp. 37–44.
Garrett has changed his view about Hume’s
problem in the Appendix since he published
Cognition and Commitment in 1997 (see
Garrett, ‘Once More into the Labyrinth’), and he
and I are in agreement in one central respect.
2/9/2012 9:35:45 PM
 HUME ON THE MIND
26
A full account of the I-Principles – the prin-
ciples according to which the Imagination
operates – must go beyond the three principles
of the association of ideas (Resemblance,
31
Contiguity and Cause and Effect) and add the
fundamental principle according to which the
Imagination is unfailingly led to posit or ‘feign’
objective continuities (persisting objects, a per-
32
sisting individual mind, and true causal con-
tinuities) on the basis of exposure to certain
sorts of sets or series of ideas. See Strawson,
The Evident Connexion, sect. 3.4.
27
Some connection which is not just an
I-principles-generated connection in the
Imagination, and is, therefore, essentially more
than the relation of cause and effect so far as
we have any empirically contentful notion
of it. (As remarked in n. 13 above, Hume’s
prime example of ‘real connexion’ is causal
necessity thought of as something that obtains
quite independently of any action of the
imagination.)
28
The clause ‘which binds them together’ can
be read in two ways. On one reading it is
about what the I-principles do: they lead
33
us to put or bind experiences together in
such a way as to take them to be parts or
rather features of a single continuing object.
34
According to the other reading it is about the
phenomenon of our experiences being actu-
ally bound together – united, connected – in
being governed by the I-principles. According
to this second reading it is only the clause
35
‘which makes us attribute to them’ that is
about what the I-principles do. Either way, the
point remains: that the thing that Hume has
to explain is the existence of ‘the principle of
connexion’, not how it does what it does.
29
THN 1.1.4.6 / 13. Hume would have loved
modern neuroscience – although not as much
as Descartes.
30
THN 1.1.4.1 / 11, 1.4.6.16 / 260. Hume
also mentions the ‘uniting principle among
our internal perceptions’ at the heart of his
198
9780826443595_Ch08_Final_txt_print.indd 198
principal discussion of causation (THN
1.3.14.29 / 169). It is a matter of unintelligible
real connection in just the same way as the
uniting principle ‘among external objects’.
For Garrett’s own account of Hume’s prob-
lem in the Appendix, in terms of ‘placeless
perceptions’, see Garrett, ‘Once More into the
Labyrinth’.
All six occur in the space of 87 words (of
which they make up 48): ‘. . . all my hopes van-
ish, when I come to explain ([P3/P4]) the prin-
ciples, that unite our successive perceptions in
our thought or consciousness. I cannot discover
any theory, which gives me satisfaction on this
head. In short there are two principles, which I
cannot render consistent; nor is it in my power
to renounce either of them, viz. [P1] that all our
distinct perceptions are distinct existences, and
[P2] that the mind never perceives any real con-
nexion among distinct existences. Did [O1] our
perceptions either inhere in something simple
and individual, or [O2] did the mind perceive
some real connexion among them, there wou’d
be no difficulty in the case.’ (THN App. 20–1 /
635–6; emphasis added)
Garrett, Cognition and Commitment, p. 171.
Of course Garrett’s text long predates, and is
not a response to, the present one.
Recall again that he uses them constantly in
a way that presupposes that they do have
content and are to that extent intelligible, and
does not mean what present-day philosophers
mean by ‘unintelligible’.
This paper descends from ‘Hume on
Himself’, a paper given at the Hume
Society conference in Cork in 1999 and
published ‘too precipately’ in 2001 (in
D. Egonsson, J. Josefsson, B. Petersson and
T. Rønnow-Rasmussen (eds), Essays in
Practical Philosophy: From Action to Values
(Aldershot: Ashgate Press, 2001), pp. 69–94).
I am grateful to Don Garrett for comments
on a later (2003) version, and to Stephen
Buckle for his comments on this one.
2/9/2012 9:35:45 PM
 9
ACTION, REASON AND THE PASSIONS
Constantine Sandis
1. ACTION AND ITS CAUSES
Actions, for Hume, are external objects in the
sense of being things which we can observe
through the senses. Our knowledge of them
is therefore not a priori but empirical, medi-
ated as it is through perceptual impressions.
Accordingly, Hume believes that purported
explanations of action, be they singular or
general, are to be tested through experience,
either directly or through testimony, for ‘we
can give no reason for our most general and
most refin’d principles, beside our experience
of their reality’ (THN Intro.10 / xviii). This
does not entail that the reasons for which
we act are themselves external, observ-
able, objects.1 Rather, their existence is to be
inferred from behaviour, the power of any
given argument from analogy hanging on the
proper degree and nature of philosophical
scepticism about causal reasoning.
Reasons why people did or believed cer-
tain things figure on virtually every single
page of all six volumes of The History of
England. Hume also mentions such reasons
in his philosophical works, be it explicitly
(e.g. THN 2.2.5.4 / 358, 3.2.1.9 / 379) or
implicitly (e.g. THN 1.3.4.2 / 83). He also
describes reasons we have for acting (e.g.
THN 1.3.9.13 / 133), making no ontological
distinction between the latter and the former
199
9780826443595_Ch09_Final_txt_print.indd 199
kinds of reasons. Nonetheless, his view of
human nature is highly alert to our tendency
to over-rationalize actions, beliefs and pas-
sions that are typically a matter of habit, cus-
tom or sentiment (e.g. THN 2.2.3.9 / 351;
1.3.7.6 / 97). Indeed, Hume’s naturalist con-
cept of what contemporary philosophers call
normative reasons is proto-Wittgensteinian
in so far as it is to be explained by reference
to human propensities and practices, for
example, expectation and induction, rather
than the other way round (THN 1.3.6.3 /
88).2 According to Hume’s account, we may
acquire knowledge of another person’s rea-
sons or motives through a combination of
inductive and analogical reasoning relating
their behaviour to past instances:
[I]n judging the actions of men we must
proceed upon the same maxims, as when
we reason concerning external objects.
When any phænomena are constantly
and invariably conjoin’d together, they
acquire such a connexion in the imagin-
ation, that it passes from one to the
other, without any doubt or hesitation.
But below this there are many inferior
degrees of evidence and probability, nor
does one single contrariety of experi-
ment entirely destroy all our reasoning.
The mind balances the contrary experi-
ments, and deducting the inferior from
2/9/2012 9:35:53 PM
 ACTION, REASON AND THE PASSIONS
the superior, proceeds with that degree
of assurance or evidence, which remains.
Even when these contrary experiments are
entirely equal, we remove not the notion
of causes and necessity; but supposing
that the usual contrariety proceeds from
the operation of contrary and conceal’d
causes, we conclude, that the chance or
indifference lies only in our judgement
on account of our imperfect knowledge,
not in the things themselves, which are
in every case equally necessary, tho’ to
appearance not equally constant or cer-
tain. No union can be more constant and
certain; than that of some actions with
some motives and characters; and if in
other cases the union is uncertain, ’tis no
more than what happens in the opera-
tions of body, nor can we conclude any-
thing from the one irregularity, which
will not follow equally from the other
(THN 2.3.1.12 / 403–4).
The prediction and explanation of action
thereby forms part of the science of human
nature that Hume seeks to establish. Actions
are no different from other events in being
susceptible to scientific laws.3 As with natu-
ral science, explanation in social science
is inductive not deductive: it is a matter of
empirically informed conjectures. If there is
to be any such thing as a logic of history, then,
it is to be an inductive logic, the limitations
of which Hume famously exposed. These
conjectures may be based on patterns of rea-
soning as well as patterns of non-rational
connections. What degree of certainty any
given pattern entitles us to assume depends
on whether one emphasizes Hume’s positive
account of causal reasoning over his scepti-
cism about causal reasoning, or vice versa.
Either way, conjectures are to be confirmed
or refuted through ‘cautious observation of
human life’ (THN Intro. 10 / xix), the most
systematic form of which is historiography:
200
9780826443595_Ch09_Final_txt_print.indd 200
Mankind are so much the same, in all
times and places, that history informs
us of nothing new or strange in this par-
ticular. Its chief use is only to discover
the constant and universal principles of
human nature, by showing men in all
varieties of circumstances and situations,
and furnishing us with materials, from
which we may form our observations,
and become acquainted with the regular
springs of human action and behaviour.
(EHU 8.7 / 83)
The work of Hume as a historian reveals
the motivating influence of character. He
embraces a moderately stoic virtue epis-
temology according to which the historian
is in the emotionally privileged position to
evaluate past actions correctly. This is to be
achieved through an approximation of the
golden mean between involved empathy and
disinterested detachment:
When a man of business enters into
life and action, he is more apt to con-
sider the characters of men, as they
have relation to his interest, than as
they stand in themselves; and has his
judgement warped on every occasion
by the violence of his passion. When a
philosopher contemplates characters
and manners in his closet, the general
abstract view of these objects leaves the
mind so cold and unmoved, that the
sentiments of nature have no room to
play, and he scarce feels the difference
between vice and virtue. History keeps
in just medium betwixt these extremes,
and places the objects in their true point
of view. The writers of history, as well
as the readers, are sufficiently interested
in the characters and events, to have
a lively sentiment of blame or praise;
and, at the same time, have no particu-
lar interest or concern to pervert their
judgement (E 568).4
2/9/2012 9:35:53 PM

For this reason Hume begins The History of
England by noting that he will not concern
himself much with either distant or recent
history, for the former is too far removed
from our concerns to be of any interest, and
the latter too close for us to keep an impar-
tial distance. His History thus begins with the
Britons and ends with the last Stuarts (almost
a century before the time in which it was
written). Hume accordingly ends his essay
‘Of the Study of History’ with a quotation
from Lucretius’ De Rerum Natura, trans-
lating roughly as ‘only then are the words
of truth drawn up from the heart’ (E 568).
In Lucretius, the context points to times of
danger, peril or adversity whereas Hume
is referring to that ‘just medium’ between
lively sentiment and personal disinterest
required for impartial judgement, a lesson
more properly learned from Hume’s Stoic
(‘the man of action and virtue’, E 146n1)
than the Epicurean (‘the man of elegance
and pleasure’, E 138n1). In stark contrast
to the thought of Hume’s philosophical con-
temporaries, Hellenistic philosophy is not so
abstract as to risk not feeling the sentiments
of vice and virtue.5
Hume believed that the correct approach
to human action is that of evaluating the
character which the action reveals, it being
a blatant falsehood that ‘all characters and
actions [are] alike entitled to the affection
and regard of everyone’ (EPM 1.2 / 169–70).
This interest in character is temporarily for-
gotten in his discussion ‘of personal identity’,6
but the important role it plays in Books II
and III of the Treatise, the second Enquiry,
and Hume’s Essays and History suggest that
it would be myopic for any account of Hume
on the soul and the self to ignore the question
of character.7 Hume would have agreed, for
instance, with David Knowles’s pronounce-
ment that ‘a life is not a bundle of acts; it is
201
9780826443595_Ch09_Final_txt_print.indd 201
ACTION, REASON AND THE PASSIONS
a stream or a landscape; it is the manifest-
ation of a single mind and personality that
may grow more deformed or more beautiful
to the end’.8, 9
Pace Hume, numerous writers on his-
tory including Croce, Knowles and Carr
have claimed that it is not the business of
the historian to pass moral judgements on
individuals, but only to explain why they
acted as they did.10 For Hume, however, a
correct explanation of action will appeal
to the agents’ motives, the discernment of
which is necessarily evaluative: if you mis-
judge character you fail to explain action,
for the appreciation of character and the
understanding of action are inextricably tied
together. A proper evaluation of character,
Hume thinks, can determine the likelihood
that any given event was the result of situ-
ation or temper. History tells us much about
human nature but, conversely, the principles
of human nature can, among other causal
principles, help us to interpret history better.
The History of England thereby aims ‘to
provide an account of English history based
on empirically plausible assumptions’,11 as
opposed to those histories which fruitlessly
invoke miracles, prophecy and revealed
religion.
Hume allows that actions may accord
with more than one motive, just as Donald
Davidson would later claim that they may
accord with one or more reasons that the
agent might have for action. According to
Davidson, the criterion for determining which
of the numerous reasons an agent might have
for acting is the one he or she actually acted
upon is causal (in a way which has proved to
be highly problematic).12 By contrast Hume
maintains, more pragmatically, that the cor-
rect method for attributing motives to any
given individual is to ask which one(s) would
reveal him as acting characteristically, a fact
2/9/2012 9:35:54 PM
 ACTION, REASON AND THE PASSIONS
to be determined on the purely empirical
grounds of past regularity:
[A]s the union betwixt motives and
actions has the same constancy, as that in
any natural operations, so its influence
on the understanding is also the same,
in determining us to infer the existence
of one from that of another. If this shall
appear, there is no known circumstance,
that enters into the connexions and pro-
ductions of the actions of matter, that is
not to be found in all the operations of
the mind; and consequently we cannot,
without a manifest absurdity, attribute
necessity to the one, and refuse it to the
other . . . (THN 2.3.1.14 / 404)
a spectator can commonly infer our
actions from our motives and character;
and even where he cannot, he concludes
in general, that he might, were he per-
fectly acquainted with every circum-
stance of our situation and temper . . .
(THN 2.3.2.2 / 408–9)
in judging the actions of men we must
proceed upon the same maxims, as when
we reason concerning external objects . . .
(THN 2.3.1.12 / 403)
No union can be more constant and cer-
tain; than that of some actions with some
motives and characters. (THN 2.3.1.12 /
404)
The most irregular and unexpected
resolutions of men may frequently be
accounted for by those who know every
particular circumstance of their character
and situation. (EHU 8.15 / 88; cf. E 16)
What neither reason nor human nature
can explain is thereby attributed to char-
acter, which divides human beings into
sorts.13 Christine Korsgaard has objected
that the suggestion that agent-causation
may be achieved ‘when the person’s char-
acter serves as a kind of filter in the causal
202
9780826443595_Ch09_Final_txt_print.indd 202
chain, making the outcome turn one way
rather than another’ seems to ‘lose track’ of
the ‘fact’ that nothing counts as an ‘action’
unless a person ‘is the cause of an intentional
movement, or something of that sort’.14 It is
true that as a philosopher Hume says nothing
of agents causally determining intentional
movements. As a historian, however, he takes
it for granted that a strong character may
determine the course of history, and he may
well have been open to the Sellars-Davidson
thesis that an action may be caused without
its agent being caused to perform it.15
None of this leaves Hume oblivious to
competing non-psychological causes of
human action, as made clear in the fol-
lowing remark on political life and human
nature: ‘So great is the force of laws, and
of particular forms of government, and so
little dependence have they on the humours
and tempers of men, that consequences
almost as general and certain may some-
times be deduced from them, as any which
the mathematical sciences afford us’ (E 16).
The case of law and government renders
political events as close as human behav-
iour can come to naturally approximate
events observed in controlled experiments.
However, Hume’s deterministic science of
behaviour can only be understood in the
light of his understanding of causation and
necessary connexion. This is not the place
to try and settle ongoing interpretational
disputes relating to the extent, if any, to
which he espouses a ‘regularity’ theory of
causation,16 but we might nonetheless note
just how weak his definitions of ‘cause’ and
‘necessity’ actually are: ‘I define necessity in
two ways, conformable to the two defini-
tions of cause, of which it makes an essen-
tial part. I place it either in the constant
union and conjunction of like objects, or in
the inference of the mind from one to the
2/9/2012 9:35:54 PM

other’ (THN 2.3.2.4 / 409). Hume’s rejec-
tion of the doctrine that we possess a lib-
erty of indifference (which he thinks of as
the illusion that one’s actions are not caus-
ally necessitated by one’s motives) is thus
more or less tantamount to a trivial truth.
He writes:
After we have perform’d any action; tho’
we confess we were influenc’d by par-
ticular views and motives; ’tis difficult
for us to perswade ourselves we were
govern’d by necessity, and that ’twas
utterly impossible for us to have acted
otherwise; the idea of necessity seem-
ing to imply something of force, and
violence, and constraint, of which we
are not sensible . . . We may imagine we
feel liberty within ourselves; but a spec-
tator can commonly infer our actions
from our motives and character; and
even when he cannot, he concludes in
general, that he might, were he perfectly
acquainted with every circumstance of
our situation and temper, and the most
secret springs of our complexion and
disposition. Now this is the very essence
of necessity, according to the foregoing
doctrine (THN 2.3.2.1–2 / 407–9).17
Human behaviour is as much the product
of an unobservable causal necessity as any
other natural event. The only difference
between them is epistemic: our knowledge
of the principles of human nature that bind
motion to action is less precise than that of
the ‘universally allowed’ deterministic laws
which bind physical force to motion. This
is partly due to the fact that the former laws
are considerably more complicated, but it
is equally a result of the fact that it is all
but impossible to perform extensive con-
trolled experiments involving human action
(though Hume would have certainly been
interested in the work of Benjamin Libet).
203
9780826443595_Ch09_Final_txt_print.indd 203
ACTION, REASON AND THE PASSIONS
Be all this as it may, our imperfect psycho-
physical knowledge is nonetheless suffi-
cient to enable us to predict individual and
social behaviour in an indefinite number of
situations.
None of this prevents Hume from pursu-
ing his ‘reconciling project’ of demonstrating
that necessity (as he has defined it) is com-
patible with free will, which he equates to the
liberty of spontaneity to do as one desires.
Far from being an obstacle to moral respon-
sibility, the necessity which binds character
to action is required for its existence, at least
given his account of the virtues according to
which the viciousness or virtue of any given
act arises from ‘some cause in the character
and disposition of the person who performed
them’ (EHU 8.29 / 98). Another corollary of
Hume’s position is that freedom increases in
proportion to madness:
’Tis commonly allow’d that mad-men
have no liberty. But were we to judge
by their actions, these have less regular-
ity and constancy than the actions of
wise-men and consequently are further
remov’d from necessity. Our way of
thinking in this particular is, therefore,
absolutely inconsistent; but it is a natu-
ral consequence of these confus’d ideas
and undefin’d terms, which we so com-
monly make use of in our reasonings,
especially on the present subject. (THN
2.3.1.13 / 404)
Given Hume’s definitions, the claim that
free will and morality are compatible with
causal necessity is unobjectionable. Hume
asserts that ‘if anyone alters the definitions,
I cannot pretend to argue with him, till I
know the meaning he assigns to these terms’
(THN 2.3.1.18 / 407). Whether it is he or his
opponents who are playing with words is, of
course, another matter.
2/9/2012 9:35:54 PM
 ACTION, REASON AND THE PASSIONS
2. MOTIVATION, REASON AND BELIEF
Hume famously claims that ‘[r]eason is, and
ought only to be the slave of the passions, and
can never pretend to any other office than to
serve and obey them’ (THN 2.2.3.4 / 414).
This remark, in tandem with his ‘influence
argument’ to the conclusion that the rules
of morality ‘are not the conclusion of our
reason’ (THN 3.1.1.7 / 457), has spawned a
hideous number of theses in moral psychol-
ogy as diverse (and incompatible) as error
theory, quasi-realism, expressivism, emotiv-
ism, prescriptivism, projectivism, non-cogni-
tivism, reasons internalism, instrumentalism,
hypotheticalism, contextualism, scepticism,
egoism, relativism, subjectivism, motivation
internalism, sentimentalism and the Humean
theory of motivation. In what follows I focus
on the last of these, only touching upon the
rest (which are primarily concerned with
issues relating to what has come to be called
‘the nature of moral judgement’)18 as and
when they relate to everyday motivation.
Hume repeatedly emphasizes the limita-
tions of reason as a motivating power:19
Reason alone can never be a motive to any
action of the will (THN 2.3.3.1 / 413).
Abstract or demonstrative reasoning,
therefore, never influences any of our
actions, but only as it directs our judge-
ment concerning causes and effects
(THN 2.3.3.2 / 414).
[I]mpulse arises not from reason but is
only directed by it (THN 2.3.3.3 / 414).
[R]eason alone can never produce any
action, or give rise to volition . . . the
same faculty is as incapable of prevent-
ing volition (THN 2.3.3.4 / 414–5).
[R]eason has no influence on our pas-
sions (THN 3.1.1.7 / 457).
204
9780826443595_Ch09_Final_txt_print.indd 204
I have prov’d, that reason is perfectly
inert, and can never either prevent or
produce any action or affection (THN
3.1.1.8 / 458).
[R]eason can never immediately prevent
or produce any action by contradicting
or approving it . . . Reason is wholly
inactive (THN 3.1.1.10 / 458).20
Passages such as these have inspired the
Humean theory of motivation (from here
onwards HTM) according to which an agent
cannot be motivated by belief alone, but only
by a belief-desire pair. More particularly,
HTM states that an agent is motivated to act
if and only if she ‘has an intrinsic desire for
the world to be a certain way and a belief that
her acting in the relevant way, a way which
represents an option available to her, will
result in the world’s being the way she intrin-
sically desires it to be’.21 Humeans about
motivation thereby claim that their theory is
presupposed by ‘all of the other explanations
that we commonsensically give’,22 maintain-
ing that the only difference between actions
and (mere) bodily movements is that the
former may always be explained in Humean
terms that reveal the agent’s intention.23
Hume’s science of humanity, outlined
in section 1 above, gives a far more central
role to character than HTM might have us
imagine.24 Indeed, Hume’s historical expla-
nations are so unlike those produced by
HTM that it would be charitable to question
whether the latter is really to be found at all
in Hume’s philosophical work. One way of
resolving this question is that of examining
specific explanations that Hume offers in
The History of England to see whether they
can easily be reconstructed into a Humean
story. Baier does just this, reaching a sceptical
conclusion.25 But even if the results had been
positive, this would lend support to HTM
2/9/2012 9:35:54 PM

itself rather than the supposition that Hume
held such a theory.
One might try to argue that there is sim-
ply a deep inconsistency between Hume’s
practical work as a historian and his philo-
sophical theories, one which reflects his own
meta-philosophical outlook (THN 1.4.7.2
/ 264). But this would be premature for,
as we have already seen, Hume reflects on
the nature of human action in his writings
on historiography. Moreover, none of his
philosophical views lend any direct sup-
port to HTM; Hume scholars (as recently as
Radcliffe, 1999)26 are wrong to suggest that
he claimed that beliefs could never motivate
alone.
For one, the passages quoted at the start of
this section are a red herring, the term ‘belief’
being conspicuous by its absence. More to
the point, it is clear from his writings that
Hume does not equate belief with reason. He
uses the term ‘reason’ in a number of interre-
lated senses, describing it as a faculty of dis-
covery (e.g. THN 3.1.1.9 / 458; cf. EHU 4.7
/ 28), an instinct (e.g. THN 1.3.16.9 / 179),
an equivalent to the general properties of the
imagination (THN 1.4.7.7 / 267) and ‘an
affectation of the very same kind as passion’
(THN 2.3.8.13 / 437).27 The last quotation
derives from a passage in which he makes the
following subtle distinction between reason
and passion:
What we commonly understand by
passion is a violent and sensible emo-
tion of mind, when any good or evil is
presented, or any object, which, by the
original formation of our faculties, is
fitted to excite an appetite. By reason
we mean affectations of the very same
kind with the former; but such as oper-
ate more calmly, and cause no disorder
in the temper: Which tranquillity leads
us into a mistake concerning them, and
205
9780826443595_Ch09_Final_txt_print.indd 205
ACTION, REASON AND THE PASSIONS
causes us to regard them as conclusions
only of our intellectual faculties (THN
2.3.8.13 / 437).
In other places reason is contrasted with
experience (e.g. EHU 4.7 / 28), sentiment (e.g.
EPM 1.3 / 170), and imagination (e.g. THN
1.3.9.19n22 / 117n1; cf. EHU 5.12 / 49),
but never desire. Reason, it would seem, is
too calm an affectation to be called an emo-
tion, but hardly a product of the intellectual
faculties.
It might be better, then, to view reason as
a specific (but by no means the only) source
of belief, namely one capable of discoveries,
as limited by Hume’s fork:
Reason is the discovery of truth or false-
hood. Truth or falsehood consists in an
agreement or disagreement either to the
real relations of ideas, or to real exist-
ence and matter of fact.28 Whatever,
therefore, is not susceptible of this
agreement or disagreement, is incapable
of being true or false, and can never be
an object of our reason . . . our passions,
volitions, and actions . . . are original
facts and realities, compleat in them-
selves, and implying no reference to
other passions, volitions, and actions.
’Tis impossible, therefore, they can be
pronounced either true or false, and be
either contrary or conformable to rea-
son (THN 3.1.1.9 / 458; cf. THN 2.3.3
/ 413–8).29
Such discoveries may result in belief, but they
are not its only source. Mutatis mutandis, not
all beliefs are inert in Hume’s view. After all,
he believes that ‘any thing may produce any
thing’ (THN 1.3.15.1 / 173). At most, it is only
those beliefs reached through reason alone that
cannot motivate.30 But even this will prove to
be a misleading way of putting things, since
2/9/2012 9:35:54 PM
 ACTION, REASON AND THE PASSIONS
(as we shall see) Hume cannot even allow that
reason can alone produce beliefs of any kind.
According to Hume, beliefs and/or opin-
ions are lively ideas: ‘An opinion or belief is
nothing but a strong and lively idea deriv’d
from a present impression related to it.’
(THN 1.3.8.16 / 119; cf. THN 1.1.1.1 / 1–2;
1.3.7.5 / 96) Given that ideas and impressions
differ only in their degree of vivacity, Hume
naturally supposes that lively ideas such as
beliefs exhibit the same effects as impressions,
to an appropriately fainter degree (cf. THN
1.3.5.7 / 8). In fact, without the influence of
belief all of our actions would be at the com-
plete mercy of our impressions.31 In a section
entitled ‘Of the Influence of Belief’ he writes:
[T]he ideas of those objects, which we
believe either are or will be existent, pro-
duce in a lesser degree the same effect with
those impressions, which are immediately
present to the senses and perception. The
effect, then, of belief is to raise up a simple
idea to an equality with our impressions,
and bestow on it a like influence on the
passions (THN 1.3.10.3 / 119).
The context makes it clear that by ‘like influ-
ence’ Hume means ‘brings about the same
effect to a lesser degree’, the degree in ques-
tion being proportionate to the degree to
which the idea in question is fainter to the
impression of which it is a copy, beliefs being
the most lively of all ideas. As Annette Baier
has recently put it, ‘Hume does not exactly
subscribe to a “belief + desire” analysis of
motivation, since desires are only among the
passions and sentiments which lead to action,
and for him a main role for belief is to cause
passions, as well as to instruct us on how to
satisfy them’.32
In addition, while Hume contrasts reason
with sentiment he explicitly identifies belief
with it:
206
9780826443595_Ch09_Final_txt_print.indd 206
. . . belief is more properly an act of the
sensitive, than of the cogitative part of
our natures (THN 1.4.1.8 / 183).
Belief is nothing but a peculiar feeling
. . . or sentiment. . . . ’Tis felt rather than
conceived, and approaches the impres-
sion, from which it is deriv’d, in its force
and influence. (THN App. 3–9 / 624–7)
So conceived, belief is an act of mind con-
sisting of a ‘strong and steady conception of
any idea’ (THN 1.3.7.5n20 / 96–7n1). To
conceive of an idea in such a way, one which
‘approaches in some measure to an imme-
diate impression’, is to be ‘perswaded of
the truth of what we conceive’ (ibid.). Such
persuasion does not merely accompany the
simple conception (that would render the
persuasion equivalent to an impression), but
is a modification of it into something firmer.
Hume felt that the account of belief
outlined in the Treatise had been misun-
derstood, dedicating the first half of his
Appendix to clarifying his notion of belief
(whose previous expressions had ‘not been
so well chosen’) in the hope of ‘guarding
against all mistakes in readers’ (THN App.1
/ 623). Here he distinguishes more explicitly
between simple conceptions (viz. ideas) and
firm conceptions (viz. the feeling that an idea
is true). Beliefs are neither new ideas, nor
impressions accompanying simple concep-
tions but, rather, firm conceptions of the very
same idea which one may have previously
conceived simply:
if belief consisted merely in a new idea,
annex’d to the conception, it wou’d be in
a man’s power to believe what he pleas’d.
We may, therefore, conclude, that belief
consists merely in a certain feeling or sen-
timent . . . When we are convinc’d of any
matter of fact, we do not but conceive
2/9/2012 9:35:54 PM

it, along with a certain feeling, different
from what attends the mere reveries of
the imagination. And when we express
our incredulity concerning any fact, we
mean, that the arguments for that fact
produce not that feeling. Did not the
belief consist in a sentiment different
from our mere conception, whatever
objects were presented by the wild-
est imagination, wou’d be on an equal
footing with the most establish’d truths
founded on history and experience.
There is nothing but the feeling, or sen-
timent, to distinguish the one from the
other . . . there is a greater firmness and
solidarity in the conceptions, which are
the objects of conviction and assurance,
than in the loose and indolent reveries of
a castle-builder . . . they strike upon us
with more force; they are more present
to us; the mind has a firmer hold of them,
and is more actuated and mov’d by them
. . . In short, they approach nearer to
the impressions, which are immediately
present to us. (THN App. 2–3 / 623–5)
Rejecting the view that a simple belief might
be ‘annex’d’ to a conception without modi-
fying it ‘after the manner that will and desire
are annex’d to particular conceptions of
good and pleasure’ (THN App. 4 / 625),
Hume concludes that what distinguishes
beliefs from simple conceptions is the pres-
ence of some feeling or sentiment. Simply
to conceive something is not to hold it true,
but merely to have a possible truth present in
one’s mind. By contrast, to believe that x is
true is to feel that it is true. Tito Magri puts
it nicely when he writes that to believe is ‘to
have an idea present to the mind as if it were
an impression’.33 By so modifying simple
conceptions beliefs have the power to influ-
ence action: ‘The effects of belief, in influen-
cing the passions and imagination, can all be
explain’d from the firm conception.’ (THN
207
9780826443595_Ch09_Final_txt_print.indd 207
ACTION, REASON AND THE PASSIONS
App. 7 / 626) What moves us to act, then, is
not a simple conception but a belief. Reason
alone cannot produce such a feeling. It may
give rise to conceptions of matters of fact
or the relation of ideas, but it cannot prod-
uce belief, let alone passion or action.34 A
judgement may result in either knowledge
or (mere) opinion or belief, but reason alone
cannot cause it to do so, which is not to say
that it cannot play an important role in the
production of our beliefs and actions.35
Just as Lewis Carroll was later to demon-
strate that entailment and inference are not
the same thing,36 so Hume shows that infer-
ential judgement does not amount to a belief;
in each case the latter may result from the
former, but it need not do so (nor ought it to,
Hume would say). Pari Passu, one can judge
that the truth of q follows from the truth of
p, without coming to form the belief that q is
true, even if one believes that p. Conversely,
one can believe q to be true, without judging
that its truth follows from that of anything
else one believes. Such possibilities under-
lie the compatibility of sceptical judgements
with non-sceptical beliefs (and vice versa,
albeit less frequently). This is the sense in
which human nature is stronger than rea-
son.37 Hume goes to great lengths to show
that this does not make our beliefs unreason-
able.38 On the contrary, what is unreasonable
is the thought that all our beliefs are the con-
clusions of reason alone.
The above all seems to suggest that Hume
does not equate beliefs and opinions with
judgements. Unlike the latter, the former are
not judged or conceived but felt. More to the
point, while beliefs can alone cause action,
judgements can only do so in combination
with a passion:39
The action may cause a judgement, or
may be obliquely caused by one, when
2/9/2012 9:35:55 PM
 ACTION, REASON AND THE PASSIONS
the judgement concurs with a passion
. . . reason, in a strict and philosophical
sense, can have an influence on our con-
duct only after two ways: Either when it
excites a passion by informing us of the
existence of something which is a proper
object of it; or when it discovers the con-
nexions of causes and effects, so as to
afford us means of exerting any passion.
These are the only kinds of judgement,
which can accompany our actions, or can
be said to produce them in any manner;
and it must be allowed that these judge-
ments may often be false and erroneous.
(THN 3.1.1.11–12 / 459)
Hume may not be a Humean about motiv-
ation but, pace Korsgaard, he does not quite
maintain that ‘an action essentially is nothing
more than a movement caused by a judge-
ment or idea that regularly has an effect on
the will’,40 for no idea could have an effect
on the will unless it was sufficiently vivid to
qualify as a belief and beliefs do not arise
from pure reason but are typically explained
by our character.41
On the other hand, it is worth recalling that
beliefs, being ideas rather than impressions,
are not ‘compleat in themselves’, thus remain-
ing susceptible to the cogitations of reasons
and capable of truth or falsehood; though,
here as elsewhere, things are not helped by the
fact that Hume makes no attempt to distin-
guish between one’s believing something and
what one believes. To complicate things more,
recall that for Hume, reason is itself an affec-
tion, differing from passion only in its degree
of tranquillity. He accordingly also character-
izes the ‘ideas of the judgement’ as he does
sentiments: ‘it is something felt by the mind
which distinguishes the ideas of the judge-
ment from the fictions of the imagination . . .
and renders them the governing principles of
all our actions’ (THN 1.3.7.7 / 97). Norton
208
9780826443595_Ch09_Final_txt_print.indd 208
and Norton parse ‘ideas of the judgement’
as ‘ideas believed’,42 but this is somewhat
rash. The only reason to suppose that Hume
might use the words ‘judgement’ and ‘belief’
interchangeably is that he does not make
a song and dance about their differences.
In fact, he never offers a proper account of
judgement,43 thereby forcing the reader either
(a) to assume it is to be identified with belief
or (b) to reconstruct the notion out of Hume’s
distinction between simple and firm concep-
tions of ideas. Judging by his footnotes and
appended clarifications, it is certainly plaus-
ible that judgements fall somewhere between
the two:
The error consists in the vulgar division
of the acts of the understanding into
conception, judgement, and reasoning,
and in the definitions we give of them.
Conception is defined to be the simple
survey of one or more ideas; Judgement
to be the separating or uniting of differ-
ent ideas: Reasoning to be the separating
or uniting of different ideas by the inter-
position of others, which show the rela-
tion they bear to each other. But these
distinctions and definitions are faulty in
very considerable articles . . . these three
acts of the understanding . . . all resolve
themselves into the first, and are noth-
ing but particular ways of conceiving our
objects. (THN 1.3.7.5n20 / 96–7n1)
. . . the mind has a firmer hold, or more
steady conception of what it takes to be
a matter of fact, than of fictions (THN
App. 5 / 626).
A judgement is neither a simple conception
nor a feeling or sentiment but the thought that
something is true. Hume tells us precious little
about what it is to judge that something is a
matter of fact but the conception in question
is arguably more vivid than imagining and
2/9/2012 9:35:55 PM

fainter than belief. Such an outlook would
not only allow for the possibility (valuable to
Hume’s scepticism) of judging that something
is the case without believing it to be so, but it
would also help to explain his controversial
account of human morals.
Numerous books and articles have been
devoted to Hume’s account of the nature of
moral judgement44 yet he never actually men-
tions moral judgement and, if anything like
the picture outlined above is correct, he takes
morals to be not judgements but beliefs. For
morals, like beliefs but unlike judgements,
have great motivational influence: ‘Morals
excite passions and produce or prevent
actions. Reason of itself is utterly impotent in
this matter. The rules of morality, therefore,
are not conclusions of our reason.’ (THN
3.1.1.6 / 457) This allows Hume to assert
that it is more correct to speak of moral sen-
timents than of moral judgements:
Morality, therefore, is more properly
felt than judg’d of; though this feeling
or sentiment is commonly so soft and
gentle, that we are apt to confound it
with an idea, according to our common
custom of taking all things for the same,
which have any near semblance to each
other . . . To have a sense of virtue is
nothing but to feel a satisfaction of a
particular kind from the contemplation
of character. The very feeling consti-
tutes our praise and admiration . . . We
do not infer a character to be virtuous
because it pleases: But in feeling that
it pleases after such a particular man-
ner, we in effect feel that it is virtuous
(THN 3.1.2.1–3 / 470–1).
Given his idiosyncratic philosophy of mind,
the question of whether or not Hume is a non-
cognitivist is fatefully anachronistic. What
is clear is that there are no textual reasons
209
9780826443595_Ch09_Final_txt_print.indd 209
ACTION, REASON AND THE PASSIONS
to suppose that when he states that belief
includes a motivating capacity, he is work-
ing with a different notion of belief to that
explored so far. Nor is he interested in meta-
ethical questions about the nature of morality.
Rather, Hume’s investigation focuses on how
we come to reach the moral opinions and per-
suasions that we each have. His answer is that
this is not a matter of discovery (be it through
intuition or demonstration), but of feeling
or sentiment. Commentators have equated
moral sentiments with indirect passions,45 but
this cannot be right, for the moral sentiments
are ideas whereas all passions are impressions
of a particular kind.
3. RULING PASSIONS AND THE WILL
Hume presents his theory of the passions
in Book II of the Treatise, eventually trans-
formed into A Dissertation of the Passions
(1757; originally published as an essay ‘of
the passions’). The former divides into parts
on (i) pride and humility, (ii) love and hatred
(including benevolence, anger, malice, envy
and lust), and (iii) the relation of passions to
the will.
Passions, for Hume, are secondary impres-
sions of reflection as opposed to original
impressions of sensation such as bodily pain
and pleasure (THN 2.1.1.1 / 275). Secondary
impressions arise either from original impres-
sions of immediate sense-perception or from
their ideas, for example, the memory of past
sensation or the expectation of a future
one. I may feel sad because I directly per-
ceive something distressing or, just as often,
because I recall (or merely believe or imag-
ine) this had been the case. Both direct and
indirect passions are ‘founded on pain and
pleasure’ (THN 2.3.9.1 / 438). The former
2/9/2012 9:35:55 PM
 ACTION, REASON AND THE PASSIONS
require only this cause, whilst the latter also
require a related object.
Hume’s notion of a reflective impres-
sion is inspired by the Hellenistic thought
that emotions may contain – or be closely
related to – a cognition that is not discover-
able by reason, though his own conservative
stance is that ‘passions can be contrary to
reason only so far as they are accompany’d
with some judgement or opinion’, a thought
which swiftly leads to the infamous remark
about it not being contrary to reason ‘to pre-
fer the destruction of the whole world to
the scratching of my finger’ (THN 2.3.3.6
/ 416). Hume further divides such passions
into calm and violent ones. The former are
said to include ‘the sense of beauty and
deformity in action, composition, and exter-
nal objects’. By contrast ‘the passions of love
and hatred, grief and joy, pride and humil-
ity’ are all of the latter, violent, kind (THN
2.1.1.3 / 276). He is careful, however, to note
that this ‘vulgar and specious’ division ‘is far
from being exact’, noting that ‘the raptures
of poetry and music frequently rise to the
greatest height’ while ‘other impressions,
properly call’d passions, may decay into so
soft an emotion, as to become, in a manner,
imperceptible’ (ibid.).
Hume’s final division is between direct
passions, which ‘arise immediately from
good or evil . . . pain or pleasure’ (THN
2.1.2.3 / 276) and indirect passions which
‘proceed from the same principles, but with
the conjunction of other qualities’ (ibid.).
Parts I and II of Book II of the Treatise, and
much of the Dissertation on the Passions,
focus on the latter. These passions involve a
reciprocal relation between sentiments and
ideas and include such vices and virtues as
pride, humility, ambition, vanity, love, hat-
red, envy, pity, malice and generosity (though
Hume sees passions as the effects of vice and
210
9780826443595_Ch09_Final_txt_print.indd 210
virtue, see, for example, THN 2.1.7.2 / 295).
The third part relates the will to direct pas-
sions, it being somewhat of a puzzle why
Hume does not also allow for indirect pas-
sions to influence action on their own. The
answer lies in his notion that a person desires
to act (or omit from acting) in relation to
perceived good and evil (THN 2.3.9.7 / 439)
which he seems to equate with pleasure and
pain (THN 2.3.3.3 / 414, 3.1.1.12 / 459).46
As Rachel Cohon persuasively argues, indir-
ect passions cannot be motives to the will
because they are not expectations of pleasure
or pain but, rather, evaluative responses of
people that do not directly relate to ‘the good
or the absence of the evil that may be attain’d
by any action of the mind or body’ (THN
2.3.9.7 / 439).47 If so, then not all moral sen-
timents are motives to the will either (it being
the very same evaluative features that make
them easy to conflate with certain indirect
passions).
Direct passions, such as desire, aversion,
grief, joy, hope, fear, despair and security, are
akin to – yet distinct from – the will. Whilst
Hume labels the will the most remarkable
of the ‘immediate effects of pain and pleas-
ure’, it remains a mere effect and possesses
none of the remarkable causal qualities that
philosophers have traditionally associated
with it: ‘by the will, I mean nothing but the
internal impression we feel and are conscious
of, when we knowingly give rise to a new
motion of our body, or perception of our
mind’ (THN 2.3.1.2 / 399). We have already
seen that Hume believes that the will can-
not be moved by reason alone, without the
assistance of sentiments or passions. Of the
influencing motives of the will, the greatest
are the violent passions, although it remains
true that ‘the calm ones, when corroborated
by reflection, and seconded by resolution, are
able to controul them in their most furious
2/9/2012 9:35:55 PM

movements’ (THN 2.3.8.13 / 437–8). As
with all other aspects of the nature of will
and the direct passions, these causes are the
same in animals as they are in humans (see
THN 2.3.9.32 / 448).
Hume’s account of the virtues is
Aristotelian insofar as it recognizes, pace
Stoicism, that morality is largely a matter of
having the right passions, at the right time,
and to the right degree. His principle of asso-
ciation entails that certain impressions will
invoke particular passions, be they direct or
indirect, and that there may also be associ-
ations between passions of either kind.48
Whilst this does not guarantee a unity of the
virtues, it suggests that they are very closely
connected. In the Treatise, Hume declares
that the principle of sympathy is ‘the chief
source of moral distinctions’ (THN 3.3.6.1 /
618), allowing us to ‘enter into the sentiments
of the rich and poor, and partake of their
pleasure and uneasiness’ (THN 2.2.5.14 /
362). However, the principle plays no com-
parable role in the Enquiry concerning the
Principles of Morals (1751), a drastically
modified version of Book III of the Treatise,
which places a greater emphasis on the senti-
ment of approbation.
NOTES
1
See A. Baier, Death and Character: Further
Reflections on Hume (Cambridge, MA:
Harvard University Press, 2008), p. 230 for
10
the trouble this creates in relation to both of
the Treatise definitions of a ‘cause’. Baier also
notes that Don Garrett and Peter Millican have
suggested that the dropping of the principle of
contiguity from the definitions in the Enquiry
11
might have occurred to allow for mental caus-
ation. Its absence would also seem to allow for
action at a distance.
12
2
See also P. F. Strawson, Scepticism and
Naturalism (London: Routledge, 1985),
211
9780826443595_Ch09_Final_txt_print.indd 211
ACTION, REASON AND THE PASSIONS
pp. 11ff.; D. Garrett, ‘Reasons to Act and
Believe: Naturalism and Rational Justification
in Hume’s Philosophical Project’, Philosophical
Studies, 132(1) (2007), pp. 1–16, pp. 4–5;
C. Sandis, ‘Hume and the Debate on “Motivating
Reasons”’, in C. Pigden (ed.), Hume on
Motivation and Virtue (Basingstoke: Palgrave
Macmillan, 2009), pp. 142–54, pp. 151–2; and
C. Sandis, ‘Induction’, in J. Hyman and H-J.
Glock (eds), A Companion to Wittgenstein
(Oxford: Wiley-Blackwell, forthcoming 2012).
3
Hume does not use the word ‘event’ often,
least of all in relation to actions. Also, it is not
implausible to suggest that he often uses it in
a sense that would make it interchangeable
with ‘fact’ (cf. J.L. Austin, Philosophical Papers,
Oxford: Clarendon Press, 1961, p. 156).
4
See also C. Sandis, ‘A Just Medium: Empathy
and Detachment in Historical Understanding’,
in Journal of the Philosophy of History 5(1)
(2011), pp. 179–200.
5
Hume’s notion of sentiment being weak
enough to be embraced by Stoicism, which he
contrasts with apathy (E 151).
6
See Baier, 2008, Death and Character, pp. 4–5.
7
For a helpful exploration of Hume’s notion of
character see P. Russell, Freedom and Moral
Sentiment: Hume’s Way of Naturalizing
Responsibility (Oxford: Oxford University
Press, 1995), pp. 95ff.
8
D. Knowles, The Historian and Character
(Cambridge: Cambridge University Press,
1963), p. 10.
9
One way out would be to follow Christine
Korsgaard’s suggestion that perhaps Hume’s
‘notion of the person as the object of pride or love
is not the same as the notion of the person as a
bundle of successive perceptions’ (C. Korsgaard,
The Constitution of Agency: Essays on Practical
Reason and Moral Psychology (Oxford: Oxford
University Press, 2008), p. 290).
Cf. B. Croce, History as the Story of Liberty,
trans. S. Sprigge (London: Allen & Unwin, 1941),
p. 47, and Knowles, The Historian and Character,
pp. 4–5; both quoted in E.H. Carr, What is
History? (London: Penguin, 1961), p. 77.
K.R. Merrill, Historical Dictionary of Hume’s
Philosophy (Lanham, MD: The Scarecrow
Press, 2008), p. 138.
I am thinking here of deviant causal chains,
whose challenge Davidson conceded to be
2/9/2012 9:35:55 PM
 ACTION, REASON AND THE PASSIONS
problematic (see, for example, D. Davidson,
Essays on Actions and Events (Oxford:
Clarendon Press, 1980), p. 87).
13
Baier, Death and Character, p. 12.
14
Korsgaard, The Constitution of Agency, p. 292.
15
See W. Sellars, ‘Volitions Reaffirmed’, in
M. Brand and D. Walton (eds), Action Theory
(Dordrecht: D. Reidel), pp. 47–63 and
Davidson, Essays on Actions and Events, p. 65.
16
For a helpful overview see H. Beebee, Hume
on Causation (London: Routledge, 2006).
My own view is defended in ‘Hume on the
Meaning of “Necessity”’, in K. Allen and
T. Stoneham (eds), Causation and Modern
Philosophy (London: Routledge, 2011),
pp. 166-87.
17
Cf. Baier, Death and Character, pp. 226–7.
18
See further below for whether Hume actually
takes morals to be judgements.
19
Hume never talks of motivation per se, but
only ‘motives’ which he takes to ‘produce’ or
‘influence’ action. Contemporary technical
jargon is Humean in so far as motivation
is understood as a causal notion, but the
motivation (or influence) of action is differ-
ent from its production; we are frequently
motivated to perform actions that never
take place. (See C. Sandis, ‘Gods and Mental
States: The Causation of Action in Ancient
Tragedy and Modern Philosophy of Mind’,
in New Essays on the Explanation of Action
(Basingstoke: Palgrave Macmillan, 2008),
pp. 356–83).
20
Note that, unlike a number of his interpreters,
Hume only ever capitalizes the word ‘reason’
when beginning a new sentence.
21
M. Smith, ‘Humeanism about Motivation’,
in T. O’Connor and C. Sandis (eds), A
Companion to the Philosophy of Action
(Oxford: Wiley-Blackwell, 2010), pp. 153–8,
p. 153; cf. his The Moral Problem (Oxford:
Blackwell, 1994), pp. 12, 92ff.
22
M. Smith, ‘The Possibility of Philosophy of
Action’, repr. in his Ethics and the A Priori
(Cambridge: Cambridge University Press,
2004), p.157.
23
Cf. Davidson, On Actions and Events, pp. 7–8.
24
Cf. Baier, Death and Character; ‘Hume [on
Action]’ in T. O’Connor and C. Sandis (eds),
A Companion to the Philosophy of Action
(Oxford: Wiley-Blackwell, 2010), pp. 513–20;
212
9780826443595_Ch09_Final_txt_print.indd 212
and Sandis, ‘Hume and the Debate on
“Motivating Reasons”’, pp. 142–54.
25
See Baier, Death and Character and ‘Hume [on
Action]’.
26
E. Radcliffe, ‘Hume on the Generation of
Motives: Why Beliefs Alone Never Motivate’,
Hume Studies 25(1) (1999), pp. 101–22.
27
David Owen writes that ‘Locke is happy to
use the same term for a faculty, the charac-
teristic activity of that faculty, and the result
of that faculty’ (Hume’s Reason (Oxford:
Oxford University Press, 1999), p. 48). The
same might be said of Hume’s use of terms like
‘reason’, ‘judgement’ and ‘passion’, although,
as Owen points out, reason is, for Hume, at
most a ‘sub-class of the imagination’, and even
this characterization is problematically loose
(ibid., pp. 75–6).
28
Hume notes in his Appendix that ‘an inference
concerning a matter of fact is nothing but the
idea of an object, that has been frequently
conjoin’d, or is associated with a present
impression’ (THN App. 6 / 626).
29
As shall become evident in section 3, this is
not obviously true of the reflective impres-
sions, particularly passions that are indirect
(cf. N. Kemp Smith (The Philosophy of David
Hume, London: Macmillan, 1941), p. 166).
30
See C. Pigden, ‘If not Non-Cognitivism, then
What?’ in his Hume on Motivation and Virtue,
pp. 80–104 and Sandis, ‘Hume and the Debate
on “Motivating Reasons”’.
31
See also Owen, Hume’s Reason, p. 165.
32
Baier, ‘Hume [on Action]’, pp. 514–15.
Cf. C. Korsgaard, Self-Constitution: Agency,
Identity, and Integrity (Oxford: Oxford
University Press, 2009), p. 64n6.
33
T. Magri, ‘Hume on the Direct Passions
and Motivation’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Blackwell,
2008), pp. 185–200, p. 191.
34
It is worth noting that this is not a view
about the ontology of so-called motivat-
ing reasons. Contrary to what is assumed
by both sides of the contemporary debate
between psychologistic and non-psychologis-
tic accounts of so-called ‘motivating reasons’,
the consideration I act upon, and the belief
that motivates me to act upon it, are not one
and the same thing (for a detailed argu-
ment see Sandis, ‘Hume and the Debate on
2/9/2012 9:35:55 PM

“Motivating Reasons”’; cf. M.M. Karlsson,
‘The Influencing Motives of the Will’, in
S. Traiger (ed.), The Blackwell Guide to
Hume’s Treatise (Oxford: Blackwell, 2006),
pp. 246–54).
35
Karl Schafer (‘Review of C. Pigden (ed.),
‘Hume on Motivation and Virtue’, Notre
Dame Philosophical Reviews, May 2010, sect.
1) has argued that Hume’s claim that reason
is inert must be understood as a claim about
the inability of the faculty itself to generate
new volitions, passions or actions. The view
presented here is sympathetic to this insight,
but would add beliefs to the list of things that
reason cannot produce alone. This is incom-
patible with Schafer’s claim that ‘beliefs about
pleasure and pain’ may be the product of
‘abstract reasoning’ as well as his further sug-
gestion that some other faculty (viz. a moral
one) is required for motivation.
36
C. Dodgson, ‘What the Tortoise Said to
Achilles’, Mind 4 (1895), pp. 278–80.
37
The early modern distinction between demon-
strative and probabilistic reasoning (relating
to knowledge and belief respectively) should
not be conflated with the distinction between
deductive and inductive reasoning (Owen,
Hume’s Reasons, pp. 30ff., 83ff.).
38
See Owen, Hume’s Reasons, pp. 144–6.
39
This only amounts to the view that judgements
cannot motivate alone if we conflate causa-
tion with motivation, and there is no reason to
213
9780826443595_Ch09_Final_txt_print.indd 213
ACTION, REASON AND THE PASSIONS
suppose that Hume did so (see Sandis, ‘Gods
and Mental States’ for why the conflation
should be avoided).
40
Korsgaard, Self-Constitution, pp. 63–4.
41
We saw in section 1 that Korsgaard can be
considerably more sensitive to this aspect of
Hume’s account of action, though not per-
suaded by it.
42
D.F. Norton and M.J. Norton, A Treatise of
Human Nature – Oxford Philosophical Texts
edn (Oxford: Oxford University Press, 2000),
p. 454 (annotations to THN 1.3.7, note 7).
43
See B. Stroud, ‘“Gilding or Staining” the World
with “Sentiments” and “Phantasms”’, Hume
Studies 19(2) (1993), pp. 253–72, p. 268.
44
E.g. P. Foot, ‘Hume on Moral Judgement’, in
D. Pears (ed.), Hume: A Symposium (London:
Macmillan, 1963) and W. Brand, Hume’s
Theory of Moral Judgement: A Study in
the Unity of A Treatise of Human Nature
(Dordrecht: Kluwer, 1992).
45
See, for example, R. Cohon, ‘Hume’s Indirect
Passions’, in E.S. Radcliffe (ed.), A Companion
to Hume (Oxford: Wiley-Blackwell, 2008),
pp. 160ff., 174–9.
46
See also Karlson, ‘The Influencing Motives of
the Will’, pp. 246–7.
47
See Cohon, ‘Hume’s Indirect Passions’,
pp. 172–3.
48
See L. Alanen, ‘The Powers and Mechanisms of
the Passions’, in Traiger, The Blackwell Guide
to Hume’s Treatise, pp. 188–92 for detail.
2/9/2012 9:35:56 PM
 10
FREE WILL
James A. Harris
Hume made two contributions to what he
and his time generally called the problem of
liberty and necessity. The first was in Book
Two of A Treatise of Human Nature (1739);
the second was in Philosophical Essays con-
cerning Human Understanding (1748), later
retitled An Enquiry concerning Human
Understanding. Much the same line of argu-
ment was presented in both places, but the
change of context served to alter the larger
significance of that line of argument. Both of
Hume’s treatments of the problem of liberty
and necessity will be considered here. The
most obvious alteration that Hume made
was one of tone: where, in the Treatise, he
adopted an aggressive stance towards what
at one point he termed ‘this fantastical sys-
tem of liberty’ (THN 2.3.1.15 / 404), in the
Enquiry he described his approach as a ‘rec-
onciling project’ (EHU 8.23 / 95). It is thus
the Enquiry’s treatment of the problem, in
particular, that makes it plausible to under-
stand Hume as an early exponent of what
has come to be called ‘compatibilism’ with
respect to moral responsibility. It should be
kept in mind, however, that Hume’s primary
concern is not with showing how the neces-
sitation of human action might be reconciled
with the legitimacy of holding human beings
morally responsible for their actions. Much
more prominent is an interest in action
214
9780826443595_Ch10_Final_txt_print.indd 214
as such, and whether or not there are rea-
sons to regard human action as, in general,
necessitated, and whether those reasons
are also reasons to regard human action as
unfree. These are taken by Hume to be ques-
tions concerning a matter of experiential
fact. In this respect, if in few others, Hume
was typical of his age, which had, for bet-
ter or worse, little interest in distinguish-
ing between ‘philosophical’ and ‘empirical’
questions. For Hume and most of his British
contemporaries, the limits of human experi-
ence were the limits of intelligible thought
and meaningful language. There was no
realm in which a priori metaphysical inves-
tigations might be conducted. In his manner
of dealing with the problem of liberty and
necessity, Hume exploits the dependence of
philosophy upon experience in an argument
for the need of redefinitions of both freedom
and necessitation.
There was of course a considerable amount
of discussion of the problem of liberty and
necessity in the decades prior to the publi-
cation of the Treatise. Hobbes had debated
the problem with Bishop Thomas Bramhall
in the 1640s and 1650s.1 Locke had explored
the nature of human freedom in the Essay
concerning Human Understanding’s chapter
‘Of Power’, and had significantly altered his
account in successive editions of the Essay,
2/9/2012 9:36:03 PM

though without making it clear exactly
where he stood when it came to the ques-
tion of whether liberty as he understood it is
compatible with the necessitation of action.
In 1717 the deist Anthony Collins had pub-
lished A Philosophical Inquiry concern-
ing Human Liberty, in which he adopted a
markedly Hobbesian approach to the prob-
lem. Collins was in turn criticized by Samuel
Clarke, he responded to the criticism, Clarke
responded to the response, and a number of
now obscure figures carried the debate on
into the late 1720s.2 An unusual, strongly
voluntarist, position had been taken up by
the Irish Bishop William King in his De
Origine Mali of 1702. King’s position had
been critically examined by Leibniz in Essais
de Théodicée (1717). Leibniz’s objections
had been answered by Edmund Law in notes
to his English translation of King, which
was published in 1731. There is, however,
no means of being sure that Hume read very
much of this literature as he developed his
own solution to the problem. He makes no
explicit reference to any other author in the
discussions of liberty and necessity in either
the Treatise or the Enquiry. Nevertheless,
he does seem to have approached the prob-
lem with previous solutions to it in mind. In
the Abstract, the summary of the argument
of Books One and Two of the Treatise that
Hume published in order to boost sales in
the spring of 1740, he claims that ‘what our
author says concerning free-will’ ‘puts the
whole controversy in a new light, by giving
a new definition of necessity’ (Abs. 31 / 660,
34 / 661). It seems reasonable then to suppose
that Hume intended to take the discussion of
liberty and necessity in a new direction, and
that he took himself to be doing something
other than merely reiterating the kinds of
‘compatibilist’ arguments already proposed
by Hobbes, Collins and Leibniz.
215
9780826443595_Ch10_Final_txt_print.indd 215
FREE WILL
When one considers Book Two of the
Treatise as a whole, it is not obvious what
the function is of the two sections of Part
III, ‘Of liberty and necessity’ and ‘The same
subject continu’d’.3 Book Two is structured
by a distinction between ‘indirect’ and ‘dir-
ect’ passions, and as Hume admits, the will
is not, properly speaking, ‘comprehended
among the passions’ (THN 2.3.1.1 / 399).
Even so, he continues, a full understanding
of the nature and properties of the will is
necessary to the explanation of the passions.
What he means by this is only apparent
when one considers the larger argument of
Book Two, and indeed also the larger argu-
ment of the first instalment of the Treatise
taken as a whole.4 Having in Book One
demolished the idea of reason as an autono-
mous faculty able to govern and control the
economy of the human mind, Hume in Book
Two presents an account of the realm of the
passions as essentially self-regulating. The
passions of each of us, that is to say, are reg-
ulated, by means of the operations of sym-
pathy, by the opinions and feelings of those
around us. Sociability, and with it the possi-
bility of human society, is guaranteed by the
extent to which, as Hume puts it earlier on in
Book Two, ‘the minds of men are mirrors to
one another’ (THN 2.2.5.21 / 365). We are
acutely aware of how others, especially our
equals and near-equals, see us, and the pains
and pleasures arising out of that awareness
prompt us to modify our passions, to make
them less violent, and to acquire the self-
command that on other accounts of the pas-
sions is supposed to be made possible only
by means of the imposition of the order of
reason upon the chaos of affection and emo-
tion. Part III begins with what is in effect no
more than a corollary of this view. The will
is defined almost out of existence, as ‘noth-
ing but the internal impression we feel and
2/9/2012 9:36:03 PM
 FREE WILL
are conscious of, when we knowingly give
rise to any new motion of our body, or new
perception of our mind’ (THN 2.3.1.2 /
399).5 The will, as traditionally understood,
has no work to do in the theory of human
nature that Hume has delineated in the earl-
ier parts of the Treatise. By the same token, it
has already been made obvious that human
actions will be necessitated by the passions.
Hume’s explicit argument for necessity
forces the point home by showing that the
doctrine of necessity, properly understood,
neither has nor needs metaphysical or theo-
logical foundations. It has its warrant, rather,
in everyday experience.
Hume’s point of departure in both Treatise
and Enquiry discussions of liberty and neces-
sity is an examination of the operations of
‘body’, the material world which, so he says,
is universally acknowledged to be neces-
sitated in its every aspect. When, and only
when, we know what we mean when we
say that the actions of material objects are
‘necessary’ will we be in a position to know
what it might mean to say that the actions
of human beings are necessary as well. And,
needless to say, Hume believes it is he who
has at last made clear what it does mean to
say that the actions of material objects are
‘necessary’. An understanding of his argu-
ment to the conclusion that human beings
are subject to necessity depends intimately
upon an understanding of his revisionist def-
inition of causal necessity in general. The
problem is that it is a matter of considerable
debate what, exactly, Hume means to say
about causal necessity in general. For present
purposes, a host of contentious issues in this
area must be ignored, and bald assertions will
have to take the place of reasoned interpret-
ative argument. Hume, it seems to the present
writer, draws a distinction between causality
as it might, or might not, operate deep in the
216
9780826443595_Ch10_Final_txt_print.indd 216
essence of material objects, and the causality
that we have ideas of and speak about both
in scientific pursuits and in ordinary life.
Following Locke, he believes that we have no
conception of the causal powers that things
have in virtue of their essences. Essences –
real essences, as distinct from nominal
ones – lie outwith our cognitive grasp. The
idea that we have of necessity is the product,
not of perception of the operation of essen-
tial powers (for we have no such perception),
but rather of associative propensities of the
mind. Given sufficient exposure to constant
conjunctions of types of events, those pro-
pensities issue in habits of prediction, and
the feeling of confidence that goes along with
those predictions is, according to Hume, the
‘impression’-source for the idea of causal
necessity. What we really mean, therefore,
when we call a connection between events
‘necessary’ and ‘causal’ (rather than acciden-
tal) is that there has been observed to be a
regular conjunction of events of those types,
and that we are disposed to make predictions
as to future instances of the same conjunction
in the future. That is all we mean. As Hume
puts the point in the Treatise: ‘Here then are
two particulars, which we are to consider as
essential to necessity, viz. the constant union
and the inference of the mind; and wherever
we discover these we must acknowledge a
necessity’ (THN 2.3.1.4 / 400).
Hume proceeds to consider human actions
in terms of these two ‘particulars’. First, he
seeks to ‘prove from experience, that our
actions have a constant union with our
motives, tempers, and circumstances’ (THN
2.3.1.4 / 401). He says he believes that only
‘a very slight and general view of the com-
mon course of human affairs’ is needed for
such a proof to be conclusive (THN 2.3.1.5 /
401). Hume’s confidence about the common
course of human affairs is remarkable:
2/9/2012 9:36:03 PM

There is no light, in which we can take
them, that does not confirm this prin-
ciple. Whether we consider mankind
according to the difference of sexes, ages,
governments, conditions, or methods
of education; the same uniformity and
regular operation of natural principles
are discernible. Like causes still produce
like effects; in the same manner as in the
mutual action of the elements and pow-
ers of nature. (THN 2.3.1.5 / 400)
The point is not that all human beings in all
times and places behave in the same way. It
is rather that the various ways we have of
distinguishing between people – in terms
of sex, age, race, social rank – distinguish
people into groups which are defined by
constant and reliable patterns of behaviour.
Hume provides no evidence for this claim.
It is hard to believe that he means it to be
taken as the result of anthropological and
historical research that he has himself under-
taken. He asserts the universal uniformity of
human behaviour as if it were a truism. And
to a considerable extent, that is indeed what
it was at the time when Hume was writ-
ing – and for some time afterwards. Hume
and his contemporaries were keenly inter-
ested in human diversity, across time as well
as across space. They believed that human
beings have a capacity for self-improvement,
and that that capacity manifests itself in the
many kinds of ‘progress’ in which eighteenth-
century writers, perhaps especially Scottish
ones, interested themselves in displaying and
explaining. But variety and progress could
only be made intelligible, they all assumed,
as the products of interactions between
environment and a uniform human nature.6
Confidence in the existence of a uniform
human nature is implicit in the very project
(and title) of A Treatise of Human Nature,
and at no point in the book does Hume give
217
9780826443595_Ch10_Final_txt_print.indd 217
FREE WILL
an argument to provide that confidence with
reasoned evidential support. It is in fact hard
to imagine what such an argument might
look like.
Having shown that, as he puts it, ‘the
union betwixt motives and actions has the
same constancy, as that in any natural oper-
ation’, Hume moves on to the second ‘par-
ticular’ essential to necessity, proof that the
influence of that union ‘is also the same, in
determining us to infer the existence of the
one from that of another’ (THN 2.3.1.14 /
404). Again, Hume’s view is that there is no
doubt that this is so. His evidence is the gen-
eral acceptance of ‘moral evidence’ in delib-
erations and decisions concerning human
affairs. This was a semi-technical term of art,
widely used in the seventeenth and eighteenth
centuries, to denote a kind of probability
arising from observed tendencies of human
nature. When something was ‘morally’ cer-
tain, it was regarded as being extremely
likely, so likely as not to admit of reasonable
doubt, while at the same time not being so
absolutely certain that doubt was impos-
sible. Such certainty was taken to be all that
we can pretend to in what we might learn
from human words and actions – and was
often contrasted with the perfect certainty
we can have with respect to what is com-
municated to us directly by God. In this part
of his argument Hume takes himself to be
doing no more than reminding the reader of
a familiar feature of human experience: that
we generally are certain, albeit only ‘morally’
so, when it comes to what other people will
do in the future, and that we manifest such
certainty in all of the common affairs in life,
in politics, in war, in business, in household
affairs. For the most part we are not per-
turbed by the fact that it is never a matter
of absolute certainty how another person
will behave. We take advantage of our, and
2/9/2012 9:36:03 PM
 FREE WILL
others’, experience of the world regardless.
And, Hume says, ‘whoever reasons after this
manner, does ipso facto believe the actions of
the will to arise from necessity, and that he
knows not what he means, when he denies it’
(THN 2.3.1.15 / 405).
It would appear that in this argument
for the necessitation of human actions by
motives, tempers and circumstances, Hume
is being careful to stay, as it were, on the
empirical surface of things, and to withhold
all judgement about the metaphysical ques-
tion of whether, in any one case, it is pos-
sible for someone to act otherwise than he or
she in fact does. Here again there are com-
plicated and vexed interpretative issues, but
there is reason to believe that, in fact, Hume’s
own doctrines, doctrines elaborated in Part
III of Book One of the Treatise, make it
impossible for Hume to be claiming anything
stronger than that, as a matter of empirical
fact, human behaviour is regular, and that, as
another matter of empirical fact, we use that
regularity as the basis for the assumptions
and predictions we make in our dealings
with other people. For Hume has argued in
Part III of Book One that there is no reason
to believe that it is impossible that the laws
of nature, including of course laws of human
nature, could not alter at some point in the
future. Even if, as is not the case, we had
knowledge of the powers by means of which
causes necessitate their effects, we could not
be certain that those powers might not dif-
fer in their operations from one moment to
another. There is, therefore, no metaphysical
straitjacket, so to speak, to ensure that the
laws responsible for how things have hap-
pened in the past must obtain in all times
and all places. The doctrine that, since the
nineteenth century, has gone by the name
of ‘determinism’ has, and can have, nothing
other than experience to back it up, but the
218
9780826443595_Ch10_Final_txt_print.indd 218
truth is that experience does not, and can-
not, back it up. Some philosophers of the
early modern period, including Hobbes and
Collins, argued that it is literally impossible
for anything to have happened differently
from how it did happen. They argued for this
conclusion from the premise that, in the case
of any and every event, it is impossible for
an event not to have a cause, and from the
further premise that a cause, in so far as it
is a sufficient condition of the effect’s tak-
ing place, makes it impossible for the effect
not to take place. Hume has rejected both of
these premises. He has argued that there is no
good reason to believe that every event must
have a cause, and that it is always conceiv-
able that a particular cause might not have
its usual effect. If determinism involves con-
fidence that there is something in the meta-
physical order of things that ensures that
the laws of nature cannot change, Hume is
not a determinist. It was the fact that he had
drained the doctrine of necessity of meta-
physics that, it may be presumed, prompted
him to claim, in the Abstract, that he had put
the whole free will question in a new, purely
empirical, light.
As already mentioned, in the Treatise
Hume writes as if his argument for the doc-
trine of necessity is at the same time an argu-
ment against the doctrine of liberty. ‘There
is no philosopher,’ he writes, ‘whose judge-
ment is so riveted to this fantastical system
of liberty, as not to acknowledge the force
of moral evidence, and both in speculation
and practice proceed upon it, as upon a rea-
sonable foundation’ (THN 2.3.1.15 / 399).
He asserts also that to hold to the liberty
of human action is to accept that human
actions are uncaused, that therefore liberty
‘is the very same thing with chance’, and
that, ‘[a]s chance is commonly thought to
imply a contradiction, and is at least directly
2/9/2012 9:36:04 PM

contrary to experience, there are always the
same arguments against liberty or free-will’
(THN 2.3.1.18 / 407). The claim that the
idea that the will is free entails that the will
is uncaused in its determinations was a com-
mon one among necessitarians in the early
modern period. Libertarians replied that,
on the contrary, choices and decisions were
indeed caused, but by the agent him or her-
self, not by motive, temper or circumstance.
Hume does not engage with that way of con-
struing the doctrine of liberty. He imagines
that the defender of liberty will have to rest
his case on a conception of human actions
as unpredictable because arbitrary. For rea-
sons adumbrated in the previous paragraph,
it was not in fact possible for Hume to claim
that chance – in the form of an uncaused
event – implies a contradiction. The weight
of his case against ‘liberty or free-will’ lies
in the claim that chance, or arbitrariness
and randomness, is contrary to experience.
He has already disarmed the objections of
those who dwell upon the capriciousness and
inconstancy of human beings. It is not always
the case that we can explain and predict the
behaviour of material objects, he has pointed
out, but we do not conclude from that that
those objects are not subject to causal neces-
sity. There is a hint, however, in the second
part of the Treatise discussion of liberty and
necessity, that Hume is not so much arguing
against liberty as seeking to redefine it. One
of the reasons why people object to the doc-
trine of necessity, he observes, is the fact that
they imagine that it implies ‘something of
force, and violence, and constraint, of which
we are not sensible’ (THN 2.3.2.1 / 407).
Hume’s answer is that it is a mistake to imag-
ine that necessity implies force, violence and
constraint. The philosophy of ‘the schools’ –
that is, the philosophy taught in universities
through the late Middle Ages and on into
219
9780826443595_Ch10_Final_txt_print.indd 219
FREE WILL
the early modern period, even into the early
eighteenth century – makes a useful distinc-
tion between ‘the liberty of spontaneity’ and
‘the liberty of indifference’. The latter kind
of liberty involves a negation of necessity
and an equation of freedom with an absence
of causation. But the former kind of liberty
involves merely an absence of constraint and
coercion, and is, therefore, as Hobbes, Locke
and others had argued, compatible with the
determination of action by motives. Hume
does not note that others before him had
argued that liberty in this sense is compatible
with necessity. He says only that when peo-
ple talk of liberty, they usually mean only an
absence of constraint and coercion, and that
this is the kind of freedom ‘which it concerns
us to preserve’ (THN 2.3.2.1 / 407–8).
A second reason why people tend to be
repelled by the doctrine of necessity is to be
found in what Hume terms ‘a false sensation
or experience even of the liberty of indiffer-
ence; which is regarded as an argument for
its real existence’ (THN 2.3.2.2 / 408). By
‘false’ here Hume means deceptive. It was
indeed a common claim of opponents of the
doctrine of liberty that we have an intro-
spective awareness of ourselves, both at the
time of action and in retrospective reflection
upon actions of the past, as radically free
in our choices and decisions. When we are
deliberating what to do, libertarians pointed
out, it feels as if what will happen next is
completely up to us. When we act on our
deliberations, we have a sense of being unde-
termined and uncaused. When we look back
on what we have done, we often have a feel-
ing, and sometimes a very troubling one, of
its having been possible for us to have acted
differently from how we did. The eighteenth
century, as was noted at the beginning of this
chapter, was a self-consciously empirical, or
‘experimental’, period in philosophy, and it
2/9/2012 9:36:04 PM
 FREE WILL
was open to libertarians to ask why this vivid
and common experience of freedom should
be ignored in debates concerning liberty and
necessity.7 Hume, in general, was not a phi-
losopher especially attuned to or interested
in experience from the first-person stand-
point. In his moral philosophy, for example,
his focus is resolutely upon the judgements
we make of other people, and he has very
little to say about judgements we make of
ourselves. His theory of the passions, as we
have seen, has at its heart the idea that we
gain our sense of ourselves from awareness
of how others see us. And when faced by
what appears to be experience of a radical
freedom to choose to act on any one of our
motives, or indeed on no motive at all, Hume
has no hesitation in judging that that experi-
ence is deceptive, and that the truth lies in
the observer’s third-person perspective on
human action. From that perspective, human
behaviour looks regular and predictable. And
such regularity and predictability license the
conclusion that human behaviour is necessi-
tated. The supposed first-person experience
of a liberty of indifference can, therefore, be
explained away as an illusion, of no signifi-
cance when it comes to an examination of
the experience relevant to making a deci-
sion about the doctrine of necessity. Later,
more aggressive opponents of liberty, such as
David Hartley and Joseph Priestley, were to
go further and simply deny that there is any
first-person experience of the freedom of the
will.
The third and final reason for the preva-
lence of the doctrine of liberty, Hume says,
‘proceeds from religion’ (THN 2.3.2.3 /
409). Religion teaches us that we are mor-
ally responsible for what we do, in the sense
that our actions will be judged by God and
used as a basis for either salvation or damna-
tion. We are all under God’s law, and in order
220
9780826443595_Ch10_Final_txt_print.indd 220
for God’s law to be just, we must be free in
how we act. We are of course also under
human laws, and, for the same reasons, in
order for those laws to be just, we must be
free in how we act. Hume’s response is that,
on the contrary, in order for laws, whether
human or divine to be just, and also effica-
cious, our actions must be presumed to be
determined by our motives and character.
Consider the efficaciousness of law first. If
the function of law is to prevent crime, then
it must be assumed that laws of certain kinds
will motivate people not to act in criminal
ways, and that assumption contains the fur-
ther assumption that human behaviour is
regular and predictable. If it were not regular
and predictable, law-makers could have no
idea which laws to make, and no confidence
that any laws could be relied upon to modify
human behaviour. When it comes to the just-
ness of the rewards and (more pertinently)
the punishments that are decreed to go along
with the keeping and violation of laws, it is
necessary that there be some way of deciding
whether a particular action was an accident
or in some other way not one for which the
agent is liable, or whether it was intended,
caused by that intention, and in every other
respect one for which the agent is genuinely
liable. Hume’s position is that it is, in effect,
the doctrine of necessity that is used to differ-
entiate between liability and non-liability. We
do not punish actions. We punish people in
respect of their actions and, roughly speak-
ing, we take people to be responsible for
their actions to the extent that those actions
are, as we say, in character. What it is for an
action to be in character is, simply, for there
to have been experience of that person, or
that kind of person, doing that kind of thing
before. It is, in other words, for the action to
be subject to necessity as Hume has defined
necessity. ‘Actions are by their very nature
2/9/2012 9:36:04 PM

temporary and perishing,’ Hume observes;
‘and where they proceed not from some
cause in the characters and disposition of
the person, who performed them, they infix
not themselves upon him, and can neither
redound to his honour, if good, nor infamy, if
evil’ (THN 2.3.2.6 / 411). If the libertarian –
where ‘libertarian’ means a defender of the
liberty of indifference – were right, it would
not be possible to attach actions to charac-
ter, since every action is radically free, and
proceeds from nothing durable or constant
in the agent. ‘’Tis only upon the principles
of necessity,’ Hume concludes, ‘that a per-
son acquires any merit or demerit from his
actions, however the common opinion may
incline to the contrary’ (THN 2.3.2.6 / 411).
In Book Three of the Treatise Hume
returns to this line of thought and uses it
to motivate one of the most controversial
aspects of his moral philosophy: the claim
that there is no difference in kind between
‘moral virtues’ and ‘natural abilities’ (such
as intelligence and knowledge, wit and good
humour, industry and perseverance).8 Those
who insist upon a deep difference here tend
to lay great stress upon the fact that nat-
ural abilities are, as Hume puts it, ‘entirely
involuntary, and have therefore no merit
attending them, as having no dependance on
liberty and free-will’ (THN 3.3.4.3 / 608).
Hume has three things to say in reply. First
he points out that the picture of moral vir-
tue common in ancient philosophy includes
many things, including constancy, fortitude
and magnanimity, which are not under the
control of the will. Either, then, the ancients
were wrong about the nature of virtue, or
voluntariness is not the all-important fea-
ture of virtue that some people now say it is.
Hume made this point knowing that many
of his contemporaries took ancient thought
just as seriously as he did, if not more so:
221
9780826443595_Ch10_Final_txt_print.indd 221
FREE WILL
he was forcing them to choose between their
respect for the ancients and their attachment
to the will-based morality of Christianity.
Secondly, Hume points out that if you accept
that moral sentiments ‘arise from the nat-
ural distinctions of pain and pleasure’, you
will struggle to make a principled difference
between moral virtues and natural abilities,
since it is plain that natural abilities excite
pain and pleasure in the observer just as
moral virtues do (THN 3.3.4.3 / 608–9).
Here Hume has in his sights Hutcheson’s
commitment to the distinction between vir-
tues and abilities – though it was, of course,
perfectly possible for the opponent of moral
sense theory to see Hume’s argument as a
reductio ad absurdum of the whole pos-
ition. And: ‘Thirdly, as to free will, we have
shown that it has no place with regard to the
actions, no more than the qualities of men. It
is not a just consequence, that what is volun-
tary is free. Our actions are more voluntary
than our judgements; but we have not more
liberty in the one than in the other’ (THN
3.3.4.3 / 609). Hume goes on to explain, in
Mandevillean manner, how the distinction
between the voluntary and involuntary has
been used by ‘moralists’, and also by legisla-
tors and divines, as a basis for inventing this
spurious distinction between the moral and
the natural. The truth is that the latter dis-
tinction is a matter of words alone.9
Hume’s discussion of liberty and necessity
in the Treatise is thus part of a larger argu-
ment concerning the place of the passions in
human life, and concerning the basis of moral
distinctions. His decision during the recast-
ing of the doctrines of the Treatise into more
digestible essay form to include ‘Of Liberty
and Necessity’ in the new treatment of the
understanding alters how one reads Hume
on free will. He could have included it in
either An Enquiry concerning the Principles
2/9/2012 9:36:04 PM
 FREE WILL
of Morals (1751) or the ‘Dissertation on the
Passions’ (1757), but he did not, and so the
question arises what place it has in the first
Enquiry considered as a whole. In the first
six sections of the Enquiry Hume strips his
theory of the understanding down to its
essentials, and presents a single line of argu-
ment concerning the nature of probabilistic
reasoning. In the second six sections Hume
concentrates on the sceptical consequences
that line of argument has for a variety of con-
troversial philosophical questions – in par-
ticular, questions concerning the nature of
the necessary connection between causes and
effects, the freedom of the will, the rational
basis of revealed religion and the origins of
the universe. He also seeks to clarify the pre-
cise nature of the scepticism he is advocat-
ing, and to make it clear that it is a purely
speculative kind of scepticism, with no
implications for everyday life. He describes
the problem of liberty and necessity as ‘the
most contentious question of metaphysics,
the most contentious science’ (EHU 8.23 /
95); but the main aim of his discussion of the
problem here in the Enquiry is to show that
all that has been thought to be difficult when
it comes to free will is the product of philo-
sophical confusions and fallacies. His target
is metaphysics and metaphysicians, and not,
as in the Treatise, an aversion to the doc-
trine of necessity that is shared by ordinary
people and philosophers alike. He is much
more explicit in presenting the doctrine of
necessity as something that we all already
believe, did we but know it. He foregrounds
the claim, present in the Treatise but not
prominent, that our ordinary conception of
liberty is perfectly compatible with the doc-
trine of necessity. In calling his approach to
liberty and necessity a ‘reconciling project’,
he is suggesting that once the claims of meta-
physicians are shown to be groundless, the
222
9780826443595_Ch10_Final_txt_print.indd 222
whole dispute will cease, because there will
be nothing left to argue about.
The metaphysical claims that have bedev-
illed the free will problem are claims that
are generally thought (by both advocates of
necessity and their opponents) to be contained
within the doctrine of necessity. These are
claims concerning the existence and nature of
powers that causes have, considered in them-
selves – powers that make effects happen, and
that are common to both material and mental
causes. The effect of juxtaposing ‘Of Liberty
and Necessity’ with ‘Of the Idea of Necessary
Connexion’ is to emphasize the fact that no
such claims are made by the necessitarian
who understands what our idea of necessity
really amounts to. A prominent theme in
‘Of the Idea of Necessary Connexion’ is, as
Hume puts it, ‘the weakness and narrow lim-
its of human reason and capacity’ (EHU 7.28
/ 76). In the context of the free will debate, the
limits of human reason mean that the neces-
sitarian claim that human actions are no dif-
ferent from the operations of material objects
in their causal determination is robbed of all
threatening or subversive implications. All
that is being said in either case is that there is
experience of constant conjunctions of types
of events, and that that experience is taken as
a ground for the making of inferences as to the
causes and effects of observed events. A pas-
sage from the Treatise that is retained almost
without alteration in the Enquiry acquires a
particular importance in the later text:
We may here be mistaken in assert-
ing that there is no idea of any other
necessity or connexion in the actions of
body: But surely we ascribe nothing to
the actions of the mind, but what every-
one does, and must readily allow of. We
change no circumstance in the received
orthodox system with regard to the will,
but only in that with regard to material
2/9/2012 9:36:04 PM

objects and causes. Nothing, therefore,
can be more innocent, at least, than this
doctrine. (EHU 8.27 / 97)
It is the changes that Hume has made to
the received orthodox system with regard
to material objects and causes that enable
his project of reconciliation. Hume dwells
on the implications of these changes for the
understanding of necessitarianism for rather
more time than he spends on the doctrine of
liberty. There is in this way a marked differ-
ence between Hume’s approach to the ques-
tion of liberty and necessity, as restated in
the first Enquiry, and, for example, Hobbes’s
approach: where Hobbes had argued that
what needed to be revised was our under-
standing of liberty, Hume holds that it is
our – that is, philosophers’ – understanding
of necessity that has to alter.
The tone of the treatment of the free will
question in the Enquiry is set when Hume
declares early on that he hopes ‘to make it
appear that all men have ever agreed in the
doctrine both of necessity and of liberty,
according to any reasonable sense, which can
be put on these terms; and that the whole
controversy has hitherto turned merely
upon words’ (EHU 8.4 / 81). With respect to
necessity, the argument proceeds in the same
manner as the argument of the Treatise: first
necessity as usually applied to the opera-
tions of matter is defined, in terms of regu-
larities and inferences made on the basis of
those regularities; then it is shown that the
same kinds of regularities and inferences are
to be found in the realm of human behav-
iour. There does appear to be a slight change
of tack, however: what Hume is concerned
to show is that it is generally believed that
human behaviour is regular and predictable,
and there is less of a sense that Hume him-
self holds to the doctrine of the uniformity
223
9780826443595_Ch10_Final_txt_print.indd 223
FREE WILL
of human nature as a matter of unexamined
dogma. He spends more time, too, dealing
with the worry that there are ‘some actions,
which seem to have no regular connexion
with any known motives, and are exceptions
to all the measures of conduct which have
been established for the government of men’
(EHU 8.12 / 86). In response, he develops a
distinction between how ‘the vulgar’ think
about the uncertainty of events and how
‘philosophers’ do, a distinction that is con-
sonant with the rather more careful attention
to the practice of natural philosophy that is
characteristic of the first Enquiry in general.10
There is more attention also to the practices
of historians and of what we might now call
anthropologists and sociologists. It is perhaps
a consequence of this that even less interest
is shown in the deliverances of first-person
introspection than in the Treatise: only in a
footnote is mention made of ‘a false sensa-
tion or seeming experience which we have, or
may have, of liberty or indifference, in many
of our actions’ (EHU 8.22 / 94). And when
Hume finally turns to showing that ‘all men
have ever agreed in the doctrine of liberty as
well as in that of necessity’ (EHU 8.23 / 95),
he is clearer than he was in the Treatise, argu-
ing straightforwardly that experience suggests
that all we can mean by liberty is ‘a power of
acting or not acting, according to the deter-
minations of the will; that is, if we choose to
remain at rest, we may; if we choose to move,
we also may’; such liberty, he continues, tak-
ing his cue from Locke, ‘is universally allowed
to belong to every one who is not a prisoner
and in chains’ (EHU 8.23 / 95).
A further significant difference between
the Treatise and Enquiry accounts is in what
is said concerning the interest of religion
in the question of liberty and necessity. In
the Enquiry, but not in the Treatise, Hume
acknowledges the fact that the doctrine of
2/9/2012 9:36:04 PM
 FREE WILL
necessity had appeared to many to have the
consequence that it is God, and not indi-
vidual human agents, who is responsible for
the moral evil that there is in the world. The
only alternative to this conclusion, as Hume
mentions, had been taken to be a denial of
the existence of moral evil, and the concomi-
tant claim that evil is a matter of appearance
only, and that the truth of the matter is that,
as Alexander Pope had put it in the Essay on
Man, ‘whatever is, is right’. But, it had been
said by some, both conclusions are obviously
absurd: God cannot be both good and a doer
of evil; and the reality of moral evil is too pal-
pable to be denied. And therefore, the same
people had concluded, the doctrine of neces-
sity is shown to be false. Hume makes it very
plain that he agrees with those who find it
impossible to deny the existence of moral evil.
Just as physical ills, in the form of pain and
suffering, impress themselves on the mind with
a potency that makes a mockery of the claims
of those who believe, or say they believe, that
there is goodness to everything in the order of
nature, so also moral harm generates painful
feelings of outrage and resentment that are
not diminished by reflection upon the pos-
sible contribution of crimes to the good of
the universe considered as a whole. ‘Are such
remote and uncertain speculations,’ Hume
asks, ‘able to counterbalance the sentiments
which arise from the natural and immediate
view of the objects?’ (EHU 8.35 / 102). The
reality of the difference between good and
evil is ‘founded in the natural sentiments of
the human mind’, sentiments which ‘are not
to be controuled or altered by any philosoph-
ical theory or speculation whatsoever’ (EHU
8.35 / 103). Hume introduces the problem
of evil as a further element of the Enquiry’s
attempt to limit the pretentions of speculative
reason. Reason’s weakness makes impossible
not only traditional metaphysics, but also
224
9780826443595_Ch10_Final_txt_print.indd 224
traditional rational religion. Hume ends the
discussion by noting that it has been found to
exceed the power of philosophy to reconcile
the freedom of human action with theologi-
cal notions of prescience and predestination.
Philosophy, therefore, should ‘return, with
suitable modesty, to her true and proper
province, the examination of common life’
(EHU 8.36 / 103).11
Despite Hume’s having, in the Abstract,
drawn attention to the novelty of his way of
treating the question of liberty and necessity,
it went for the most part unnoticed in what
reviews the Treatise and the first Enquiry did
receive. A review of Books One and Two of the
Treatise in the periodical Common Sense in
July 1740 acknowledged the ‘moral evidence’
of inferences concerning human behaviour, but
asserted that, even so, ‘every Man must be con-
vinced from what he feels within himself, that
this Influence is not absolute and necessary;
and Self-Conviction is a much stronger Proof
than any we can have from our Observation
of external Objects, because we cannot know
their Tempers and Circumstances, and much
less the Motives they are governed by, so well
as we do our own’.12 Joseph Highmore, in his
Essays, Moral, Religious and Miscellaneous
(1766), claimed that in the Enquiry Hume
‘maintains absolute uncontroulable necessity,
in the moral as well as the natural world’,
and, like the reviewer in Common Sense, was
particularly troubled by the tension between
such necessity and the existence, which he
took Hume to allow, of ‘an universal feeling
of liberty, and free agency in man’.13 ‘Surely’,
Highmore wrote,
either man is free, feels and knows him-
self to be so, and thence arises remorse
or self-blame, on his having acted against
his judgement and conscience; - Or, he
is absolutely necessary, finds and knows
2/9/2012 9:36:05 PM

himself to be so; and therefore cannot
consistently blame himself, or suffer
remorse, for any past action; - Or, lastly,
it is a matter of doubt with him, whether
he is free or not; in which state of mind,
he cannot be wholly without anxiety, nor
perhaps without some kind or degree of
remorse.14
In 1795 John Allen, signing himself ‘a
Necessitarian’, published a pamphlet defend-
ing Hume’s account of necessity in the Enquiry
from criticisms made by James Gregory in his
Philosophical and Literary Essays (1792).
But it could not be said that Hume’s version
of the doctrine of necessity was central to
the free will debate in the second half of the
eighteenth century. In his Essays on the Active
Powers of Man (1788), for example, the liber-
tarian Thomas Reid takes as his target Joseph
Priestley’s The Doctrine of Philosophical
Necessity Illustrated (1777). The dispute
between Reid and Priestley determined the
contours of disputations about human free-
dom far into the nineteenth century.
The distinctive features of Hume’s
approach to the question of liberty and neces-
sity went ignored even as much of the rest of
his philosophy received renewed attention in
the second half of the twentieth century. This
was because it was assumed that Hume said
nothing on the topic that Hobbes and Locke
had not said before him. That assumption
met with sustained criticism in Paul Russell’s
Freedom and Moral Sentiment: Hume’s
Way of Naturalizing Responsibility, where
it is argued that Hume’s reconciliation of
freedom, in the form of moral responsibil-
ity, with necessity depends to a significant
extent upon his naturalistic analysis of the
moral sentiments.15 It will be obvious to the
reader that the present author also disagrees
with the view that Hume is merely restating
225
9780826443595_Ch10_Final_txt_print.indd 225
FREE WILL
old arguments in his writings on liberty
and necessity.16 It has been argued here that
Hume can only be understood to be a ‘deter-
minist’ in an extremely qualified sense.17
Recently John P. Wright and Peter Millican
have criticized this view, and have reasserted
the contrary view that Hume’s determinism
is full-blooded and unequivocal.18 Another
recent development in the study of Hume on
free will has been that it has come to seem
to some – understandably enough, given
the way Hume inserted ‘Of Liberty and
Necessity’ into the first Enquiry immediately
after ‘Of the Idea of Necessary Connexion’ –
that Hume’s way of treating the free will
question might have some relevance to the
so-called ‘New Hume debate’: that is, the
debate as to whether Hume is a ‘regularity
theorist’ about causation or, instead, a kind
of realist about the existence of unobserved
and unobservable causal powers.19 On the
interpretation offered here, what Hume
brings to both the question of causation
and the question of liberty and necessity is
a kind of agnosticism about fundamental
metaphysical issues. All we have to go on
when it comes to such questions, according
to Hume, is experience; and what experi-
ence tells us is, as Hume puts it, ‘that we can
draw inferences concerning human actions,
and that those inferences are founded on the
experienced union of like actions, with like
motives, inclinations, and circumstances’
(EHU 8.27 / 97).20
NOTES
1
For a modern edition of representative samples
of the Hobbes-Bramhall debate, see V. Chappell,
Hobbes and Bramhall on Liberty and Necessity
(Cambridge: Cambridge University Press, 1999).
2
For an account of the Collins-Clarke debate,
see J.A. Harris, Of Liberty and Necessity: The
2/9/2012 9:36:05 PM
 FREE WILL
11
Free Will Debate in Eighteenth-Century British
Philosophy (Oxford: Clarendon Press, 2005),
chap. 2.
3
For some suggestions, see P. Russell, The
Riddle of Hume’s Treatise: Skepticism,
Naturalism, and Irreligion (New York: Oxford
12
University Press, 2008), chap. 16.
4
For a fuller development of this reading
of Books One and Two of the Treatise, see
J.A. Harris, ‘“A compleat chain of reasoning”:
Hume’s Project in A Treatise of Human Nature,
Books One and Two’, Proceedings of the
Aristotelian Society 109 (2009), pp. 129–48.
5 13
For a useful account of Hume’s definition of
14
the will, see R.F. Stalley, ‘The Will in Hume’s
15
Treatise’, Journal of the History of Philosophy
24 (1986), pp. 41–53.
6 16
See C.J. Berry, Social Theory of the Scottish
Enlightenment (Edinburgh: Edinburgh
University Press, 1997), chap. 3.
7
The importance to the free will debate of first-
person experience is a major theme of Harris,
Of Liberty and Necessity, passim.
8
For a detailed consideration of this aspect of
17
Hume’s argument, see P. Russell, Freedom and
Moral Sentiment: Hume’s Way of Naturalizing
Responsibility (New York: Oxford University
Press), chaps 6–12.
9 18
Users of the new Clarendon edition of the
Treatise should be aware that, without any text-
ual authority, David and Mary Norton change
the title of Section Five of Part III of Book Three
from ‘Some farther reflections concerning the
natural virtues’ to ‘Some farther reflections
19
concerning the natural abilities’. Hume intended
‘natural virtues’, it may be presumed, as part of
his strategy of unsettling the reader’s grip on the
distinction between virtues and abilities.
10
For more on relevant differences between the
first Enquiry and the Treatise, see P. Millican,
‘Hume’s Sceptical Doubts concerning Induction’,
in P. Millican (ed.), Reading Hume on Human
20
Understanding (Oxford: Oxford University
Press, 2002), pp. 107–73, esp. pp. 144–5.
226
9780826443595_Ch10_Final_txt_print.indd 226
For further discussion of the religious
dimension of Hume’s discussion of lib-
erty and necessity see T. Holden, Spectres
of False Divinity: Hume’s Moral Atheism
(Oxford: Oxford University Press, 2010),
chap. 7.
This and many other contemporary responses
to Hume on liberty and necessity are reprinted
in J. Fieser (ed.), Early Responses to Hume’s
Metaphysical and Epistemological Writings,
2 vols, 2nd rev. edn (Bristol: Thoemmes
Continuum, 2005); for the passage quoted, see
vol. 1, p. 89.
Fieser (ed.), Early Responses, vol. 1, p. 172.
Ibid., p. 173.
See especially Russell, Freedom and Moral
Sentiment, Introduction.
For further development of the reading offered
here, see Harris, Of Liberty and Necessity,
chap. 3, and J.A. Harris, ‘Hume’s Reconciling
Project and “the common distinction betwixt
moral and physical necessity”’, British Journal
for the History of Philosophy 11 (2003),
pp. 451–71.
For a similar reading, see D. Garrett,
Cognition and Commitment in Hume’s
Philosophy (New York: Oxford University
Press, 1997), chap. 6.
See J.P. Wright, Hume’s A Treatise of Human
Nature: An Introduction (Cambridge:
Cambridge University Press, 2009), chap. 5;
and P. Millican, ‘Hume’s Determinism’,
Canadian Journal of Philosophy 40 (2010),
pp. 611–42.
For this debate, see R. Read and K. Richman
(eds), The New Hume Debate, rev. edn
(London: Routledge, 2007). The relevance of
Hume’s discussions of ‘liberty and neces-
sity’ to the debate is stressed especially by
P. Millican, ‘Hume, Causal Realism, and
Causal Science’, Mind 118 (2009), pp.
647–712, sect. 8.
I am grateful to Sean Greenberg for comments
on a draft of this chapter.
2/9/2012 9:36:05 PM
 11
HUME ON MIRACLES
Duncan Pritchard and Alasdair Richmond
1. HUME’S ARGUMENT: READINGS
AND MISREADINGS
Hume’s seminal ‘Of Miracles’ (EHU 10 /
109–31) still attracts heated discussion and
different interpretations of his argument
persist. What is clear, however, is that he held
that rational beliefs in miracles based purely
on testimony are (at very least) highly prob-
lematic. In particular, anyone whose belief
in a particular religious hypothesis is due
purely to testimony recorded hitherto that
a miracle has occurred (‘miracle-testimony’)
has failed to form their beliefs as Hume
prescribes.
Despite its enduring popularity as an
anthology piece, ‘Of Miracles’ is not a stand-
alone work but one closely tied to Hume’s
overall philosophical project. In particular,
‘Of Miracles’ should be considered against
the background of his regularity theory of
induction. He famously did not believe that
expectations that the future will resemble the
past (or that the unobserved must resemble
the observed) could be justified in a way that
was both rational and non-circular. First,
the principle of induction does not express
any ‘relation of ideas’. The contradiction
of a relation of ideas is inconceivable (or
nonsensical), whereas the contrary of any
induction, no matter how well-supported,
227
9780826443595_Ch11_Final_txt_print.indd 227
is always conceivable: ‘That the sun will not
rise tomorrow is no less intelligible a propo-
sition, and implies no more contradiction,
than the affirmation, that it will rise’ (EHU
4.2 / 25–6). However, neither is the princi-
ple of induction a ‘matter of fact’.1 All infer-
ences from experience must presuppose the
principle of induction; hence that principle
is too fundamental to be justified by appeal
to experience: ‘All inferences from experi-
ence suppose, as their foundation, that the
future will resemble the past’ (EHU 4.21 /
37). Likewise, attempts to justify induction
by reference to the uniformity of nature face
the insuperable obstacle that any belief in
such uniformity can itself only be justified by
induction. So while the principle of induction
and the uniformity of nature are certainly not
meaningless, neither are they susceptible to
rational, non-circular justification. However,
Hume did not think all inductions were on
a par; still less did he advocate facing the
future with inductive paralysis. Creatures
like us possess a kind of mental inertia that
preserves our mental states in being.2 Hence
we instinctively project those regularities we
observe. Furthermore, our expectation that
such regularities will continue should be
proportional to the evidence in their favour:
‘The creature expects from the present object
the same consequences, which it has always
2/9/2012 9:36:14 PM
 HUME ON MIRACLES
found in its observation to result from simi-
lar objects’ (EHU 9.4 / 106).
Thus, Hume’s pragmatic solution to induc-
tive scepticism urges the maintenance of due
proportion between degree of belief and evi-
dence. However, even ideally well-confirmed
inductions may fail. While proportioning
degree of belief to available evidence is the
best we can do, nature need not conform
to our inductions or vindicate our predic-
tions. Inductive failure is always conceivable
and even a small degree of non-uniformity
makes us hesitant in projecting regulari-
ties. Therefore, there is a sharp distinction
between unbroken regularities and those that
have admitted of exceptions. The projection
of very widely observed, uniform regularities
is the most powerful epistemic norm there is.
So Hume is not recommending mere sceptical
rejection of inductive practices, but instead
arguing that unaided reason is not the pri-
mary regulator of our inductions. Equally
importantly, he prescribes how induction
should be regulated: while widely supported,
hitherto unbroken uniformities can always
fail; we should have the strongest degree of
personal conviction that they will continue –
interruptions are always possible but, prior
to experience, they should be incredible.
Inductively wise creatures obey their projec-
tive instincts.
Crucially, Hume had a three-fold distinc-
tion between demonstrations, proofs and
probabilities. A demonstration is essen-
tially a deductive argument, one appealing
to relations between ideas. (So Pythagoras’
Theorem would be amenable to demonstra-
tion.) Proofs and probabilities are inductive
arguments, differing only in their evidential
support. A proof is the strongest possible
argument from experience – i.e., the projec-
tion of a well-confirmed, hitherto exception-
less regularity.3 (We have a proof that the
228
9780826443595_Ch11_Final_txt_print.indd 228
sun will rise tomorrow – no exceptions are
recorded and many positive instances are.)
However, a probability involves projecting a
well-supported but not exceptionless regular-
ity. (Thus, if your car has started on the first
try some 99 previous mornings out of 100,
you should have a strong probability for its
starting first try tomorrow, but not a proof.)
So, odd though it may sound to modern ears,
Humean proofs are defeasible and may be
opposed by greater proofs.4 When proofs
conflict, we should incline our beliefs to that
proof with the better evidential support both
quantitatively and qualitatively.
Key to Hume’s argument is that a miracle
is characterized as ‘a violation of the laws of
nature’ (EHU 10.12 / 114) – i.e. an excep-
tion to one of the best-confirmed regularities
we possess. At times he seems to take this to
be a complete characterization.5 There is a
popular but misleading interpretation of ‘Of
Miracles’ which we shall call the strong read-
ing. On this view, Hume argues a priori (i.e.
by appealing to general principles) that tes-
timony can never rationally ground belief in
miracles.6 That is, however strong the testi-
mony, it cannot provide compelling evidence
for a miracle because any epistemic force the
testimony has will inevitably be outweighed
by the evidence one has against the possibil-
ity of miracles. Taken in isolation, parts of
‘Of Miracles’ may favour this reading by
suggesting there will always be more reason
to suppose testimony to a miracle should be
rejected than that a miracle has occurred.
Consider this passage, for example: ‘no tes-
timony is sufficient to establish a miracle,
unless the testimony be of such a kind, that
its falsehood would be more miraculous,
than the fact, which it endeavours to estab-
lish’ (EHU 10.13 / 115–6). On the strong
reading, Hume claims that if we hear testi-
mony to an event contrary to natural law,
2/9/2012 9:36:15 PM

then rationality obliges us to reject the tes-
timony as unreliable. So if someone tells you
Jesus miraculously turned water into wine,
you necessarily have more reason to reject
the testimony than to suppose this miracle
occurred, and hence this testimony provides
you with no rational basis on which to believe
this miracle took place. However, the strong
reading is multiply suspect. For one thing, it
fails to account for passages where Hume
explicitly allows that testimony can make
belief in a miracle rationally compelling.
Another popular misreading trivializes
Hume’s case by making him claim that the
very concept of a miracle makes it a contra-
dictory or impossible event. After all, if laws
of nature are defined as inviolable, then it
trivially follows that transgressions of laws
of nature are impossible, and hence miracles
must be impossible too. No wonder, then,
that it would not be rational to believe in
miracles on a testimonial (or any other) basis!
By ruling out miracles by definition, this
strategy would give opponents of miracles
a suspiciously quick victory. However, this
strategy is demonstrably not Hume’s. Hume
never argues that a miracle is (conceptually
or otherwise) impossible – indeed he explic-
itly allows not only that miracles may occur
but that they can (at least in theory) be made
the object of compelling testimony. Even the
strong reading of Hume does not reject the
very idea of miracles, but appeals only to
the balance of evidence against their occur-
rence, given one’s overwhelming grounds in
support of the uniformity of nature and the
slender opposing grounds in favour of the
miracle supplied by testimony. Consider this
passage:
Hume is not arguing that the wise reject
testimony for miracles because they rec-
ognize that miracles are impossible by
229
9780826443595_Ch11_Final_txt_print.indd 229
HUME ON MIRACLES
definition. Nor is he claiming that no-
one could ever observe a miracle. He is
not missing the point by defining ‘mir-
acle’ in such a way that any event that
actually occurs or is observed, no matter
how bizarre, would fail to be a miracle.7
It is important to recognize that what mira-
cles oppose is, for Hume, not an inviolable
law of nature, but rather the widely observed
regularity of a natural uniformity. Thus,
miracles on this conception are certainly not
impossible, but they are problematic objects
of inference and things which we should be
at the very least reluctant to believe in.8 A
useful summary is this:
Hume is speaking in descriptive mode, in
which a law of nature is a type of phe-
nomenon that has so often been found to
occur that, our innate expectations being
of the kind they are, we cannot help but
assign it a maximal or near-maximal
probability of always being repeated in
the same circumstances. There is, how-
ever, no implication that we are correct
to do so.9
As an example of correct proportionality
between belief and evidence, consider Hume’s
(EHU 10.10 / 113–14) Indian prince. Having
lived all his life in warm climes and never
observed water freeze, the prince receives
testimony that water can turn into ice. The
prince rejects this testimony because ice runs
so counter to his own observations and thus
to what he believes are the prevailing natural
laws for water. Ice to the Indian prince seems
an impossible extrapolation from water’s
observed behaviour. Hume argues that the
prince has greater reason to reject the testi-
mony than to alter his beliefs in the light of
this testimony, and hence he has ‘reasoned
justly’ in rejecting such testimony (EHU 10.10
2/9/2012 9:36:15 PM
 HUME ON MIRACLES
/ 114). Of course, the prince would not be
correct in clinging dogmatically to his rejec-
tion of ice if his evidential basis changed – for
example, if he were transported to Muscovy
and saw water become solid with his own
eyes. However, when the prince’s sole eviden-
tial support for the existence of ice is testi-
monial, Hume clearly thinks rejecting such
testimony is more rational for the prince
than acceptance. When evaluating testimony,
rationality does not so much prescribe which
beliefs are to be accepted, but rather how to
change degrees of belief in the face of new
evidence. While the prince reasoned to a false
conclusion in rejecting the existence of ice, he
nonetheless updated his beliefs as rationally
as his evidence permits:
A wise man, therefore, proportions his
belief to the evidence. In such conclusions
as are found on an infallible experience,
he expects the event with the last degree
of assurance, and regards his past experi-
ence as a full proof of the future exist-
ence of that event. (EHU 10.4 / 110)
Note that the prince is not being offered
testimony to a miracle, since the testimony
claims only that water behaves differently
in colder climes.10 In contrast, when some-
one testifies that a miracle has occurred –
for example, that someone has walked on
water – they explicitly claim that an event
contrary to natural law has taken place,
and this makes it even harder to accept the
testimony:
On the one side we have wide and
unproblematic testimony to the effect
that when people step into water they do
not remain on its surface. On the other
side we have isolated reports of people
walking on the surface of water. Given
230
9780826443595_Ch11_Final_txt_print.indd 230
testimony of the first kind, how are we
to evaluate testimony of the second sort?
The testimony of the first sort does not
show that the testimony of the second
sort is false; it does, however, create a
strong presumption – unless countered,
a decisively strong presumption – in
favour of its falsehood.11
In opposing natural law, testimony to a mir-
acle incurs a greater testimonial opposition
than would be the case if the event were
merely unusual. The expectation that nat-
ural laws hold and will continue to hold,
although by no means infallible, is none-
theless among the most powerful rational
expectations there are, and any testimony
that aims to overturn such expectations faces
an uphill task. Barring remarkably power-
ful testimonial evidence to the contrary, the
rational expectation must always be that
natural laws are obeyed and that our reason-
ing from experience must proceed analogic-
ally. For the prince to believe in the existence
of ice on the basis of testimony alone would
offend against correct analogical reasoning:
‘The operations of cold upon water are not
gradual, according to the degrees of cold;
but whenever it comes to the freezing point,
the water passes in a moment, from the
utmost liquidity to perfect hardness’ (EHU
10.10n22 / 114n). It is not merely that the
prince has not heard testimony that water
does not turn to ice in cold climates; rather
the process attested to implies a failure of
natural analogical reasoning.
Presumably the ice-testimony the prince
receives is not very detailed – closer to ‘trav-
ellers’ tales’ in effect. But what if a miracle
is extremely well attested to? Suppose hun-
dreds of putative, mutually consistent wit-
nesses testify. Would it not then be entirely
unreasonable to dismiss the possibility that
2/9/2012 9:36:16 PM

the miracle occurred? Hume is fully aware
of this problem: ‘suppose, that all historians
who treat of England, should agree, that, on
the first of January 1600, Queen Elizabeth
died . . . and that, after being interred a month,
she again appeared, resumed her throne, and
governed England for three years . . .’ (EHU
10.37 / 128).
On the strong reading, Hume’s way of
responding to such cases is not to weaken
his stance in the context of the clear wealth
of testimonial support in play. Instead, the
‘strong reader’ would have Hume argue that
the extent of testimonial evidence on offer in
these cases means only that we should seek
out the natural causes which gave rise to
this extraordinary event and thereby deter-
mine why so many people are under the
false impression that a genuine miracle took
place. This would be akin to the prince mov-
ing to ‘Muscovy during the winter’ (EHU
10.10n22 / 114n) in order to see if water does
indeed turn to ice as has been alleged. What
the prince will discover then is that there is
nothing miraculous about this event at all,
and hence no rational bar on this score for
supposing that the target event occurs. The
additional strength of testimonial support
for the putatively miraculous event thus only
has a bearing on the extent to which we are
rationally obliged to discover a natural cause
for that event, as opposed to simply dis-
missing the testimony tout court. Crucially,
additional testimonial support cannot on the
strong reading of Hume’s view ever suffice
to give us reason to think that a miracle has
occurred.
While the strong reading has been (and
remains) popular among Hume critics, it faces
multiple problems. The fundamental prob-
lem with the strong reading is simply that it
relies on partial readings of key passages in
‘Of Miracles’. Hume explicitly countenances
231
9780826443595_Ch11_Final_txt_print.indd 231
HUME ON MIRACLES
at least the possibility of a rationally formed,
testimony-based, belief in the occurrence of
a miracle:
I beg the limitations here made may be
remarked, when I say, that a miracle can
never be proved, so as to be the founda-
tion of a system of religion. For I own,
that otherwise, there may possibly be
miracles, or violations of the usual course
of nature, of such a kind as to admit of
proof from human testimony; though,
perhaps, it will be impossible to find any
such in all the records of history. (EHU
10.36 / 127)
So clearly Hume believed (i) that miracles can
occur and, crucially, (ii) that they can (at least
in theory) be the subject of rationally com-
pelling testimony. Hume is concerned partic-
ularly with those miraculous events that are
capable of acting as support for a system of
religion – that is, ‘testimonial’ miracles, mira-
cles that testify to the divine mission, inspi-
ration or guidance of a miracle-worker. So
what is at issue is whether miracle-testimony
could ever be good enough to make a mira-
cle the foundation of a religious hypothesis
(a ‘foundational miracle’). Hume does not
merely mention the abstract possibility of
acceptance-compelling miracle-testimony, he
goes on to offer a detailed proposal for what
an evidentially compelling miracle might
look like:
Thus, suppose, all authors, in all lan-
guages, agree, that, from the first of
January 1600, there was a total darkness
over the whole earth for eight days: sup-
pose that the tradition of this extraordi-
nary event is still strong and lively among
the people: that all travellers, who return
from foreign countries, bring us accounts
of the same tradition, without the least
variation or contradiction: it is evident,
2/9/2012 9:36:16 PM
 HUME ON MIRACLES
that our present philosophers, instead
of doubting the fact, ought to receive it
as certain, and ought to search for the
causes whence it might be derived. The
decay, corruption, and dissolution of
nature, is an event rendered probable
by so many analogies, that any phenom-
enon, which seems to have a tendency
towards that catastrophe, comes within
the reach of human testimony, if that tes-
timony be very extensive and uniform.
(EHU 10.36 / 127–8)
Thus while Hume clearly thinks rational
compelling testimony to miracles may be
hard to come by, nonetheless it is clearly
not impossible, contra the strong reading.
For example, testimony can be compelling if
it is ‘very extensive and uniform’, as it is in
the ‘eight days of darkness’ case in contrast
to the relatively parochial ‘Queen Elizabeth’
case. (The latter miracle would be too spatio-
temporally localized to lend itself to the sort
of testimony which would render it worthy
of belief.) Likewise, the testimony can be
compelling if it comes from widely separated
witnesses who otherwise have little in com-
mon and who are not partisans for a particu-
lar explanation of the occurrence testified to,
as opposed to (for example) historians with a
particular wish to glorify Queen Elizabeth or
to testify to divine endorsement of her poli-
cies. Even if Elizabeth’s resurrection was uni-
versally endorsed by historians, the balance
of proof would still favour scepticism.
Thus, the strong reading fails. Provided the
above restrictions are met, testimony can be
sufficient to support a rationally held belief
in miracles.12 Nonetheless, while interrup-
tions to the ordinary course of nature are not
impossible, our epistemic nature being what
it is, such interruptions will be (and ought to
be) incredible, all else being equal. In most
conceivable, and maybe all recorded, cases,
232
9780826443595_Ch11_Final_txt_print.indd 232
we would do better to reject the testimony
than believe a miracle has occurred. Hume
thus believes that testimony to a miracle will
almost certainly not be of sufficient quality
and quantity to make belief rationally com-
pelling. Furthermore, even granted belief in
the miracle itself is compelling, it is exceed-
ingly unlikely, if not impossible, for the testi-
mony to be so powerful as to make it rational
to believe that the miracle in question testifies
uniquely to the truth of a particular religious
hypothesis. Call this the weak reading.13
While the strong reading does seem to have
some textual support in its favour, it fails to
pay due attention to two key, and related,
features of Hume’s approach. The first is the
essentially empirical nature of Hume’s dis-
cussion. Hume is at least as much concerned
to offer an adequate description of our epis-
temic and evidential practices as he is to
legislate for those practices. The second con-
cerns the need to distinguish sharply between
Hume’s treatment of miracles per se, and his
treatment of specifically religious miracles –
viz., a miracle involving a divine intervention
into the natural order, such as might serve as
the ‘foundation of a system of religion’ (EHU
10.35 / 127). This last restriction – that is, to
miracle-testimony aimed at supporting a par-
ticular religious hypothesis – is very impor-
tant and, contrary to Hume’s own cautions,
one often ignored by critics.
It is tempting to regard Hume as aiming to
offer some sort of a priori limitation on what
it is reasonable to believe regarding miracles.
After all, a good deal of what he says, particu-
larly in Part 1, is seeming general reflection on
the nature of the epistemology of testimony
as it applies to miracles. Crucially, however, it
would be utterly contrary to the general spirit
of Hume’s philosophy to establish a matter
of fact in any a priori fashion, and hence we
should be immediately suspicious of this way
2/9/2012 9:36:16 PM

of reading him. The example of the eight days
of darkness bears this suspicion out, since
Hume is clearly not ruling out the possibil-
ity of a testimony-based, rational belief in the
existence of a miracle tout court. Instead, in
offering general remarks on the nature of tes-
timony, Hume argues that the rational bar is
set extremely high for testimony-based beliefs
about miracles. He then appeals to empirical
facts about cases where miracles are testi-
fied to and argues that it would be exceed-
ingly hard for any testimony-based belief in
miracles to clear this bar. Hume does not
forbid miracles a priori or think such occur-
rences could never under any conceivable cir-
cumstances be made the object of rationally
compelling testimony. Rather, reflection on
correct epistemic practice reveals that the hur-
dle facing miracle-testimony is so high that in
practice there is a negligible likelihood that
real miracle-testimony can be found which is
compelling.
This brings us to the second point, for
Hume is quite clear that there are empirical
grounds for treating our testimony-based
beliefs in the existence of specifically reli-
gious miracles as even more epistemically
dubious than testimony-based beliefs in the
existence of non-religious miracles. It is one
thing to come to believe, on the extensive tes-
timonial basis described, that the Earth was
plunged into darkness for eight days, con-
trary to one’s current understanding of the
natural order. It is another thing to believe
that this event was a religious miracle, since
this will impose an additional epistemic bur-
den. As Robert Fogelin puts it:
With respect to miracles, Hume’s strategy
is to use the canons of causal reasoning
to evaluate testimony brought forward
in their behalf. Because, for him, no mat-
ter of fact can be established a priori, it
233
9780826443595_Ch11_Final_txt_print.indd 233
HUME ON MIRACLES
remains an open, though remote, pos-
sibility that testimony could establish
the occurrence of a miracle. That is the
point of the discussion of the eight days
of darkness. With respect to miracles
intended to serve as the foundation of
a religion, the situation, according to
Hume, is factually different. When we
examine the testimony brought forward
on their behalf, we see that it has uni-
formly failed to meet appropriate stand-
ards of acceptability.14
When dealing with testimony-based beliefs
regarding (mere) miracles versus those regard-
ing specifically religious miracles, the issue is
fundamentally an empirical one related to
whether there is sufficient epistemic support
for the testimony-based belief in question to
clear the bar Hume lays down for it. Hume’s
claim, however, is that any testimony-based
belief in a miracle which is also saddled with
the burden of a religious explanation would
require an even greater degree of epistemic
support to clear this bar, and so faces an even
stiffer challenge (one which it is hard, if not
impossible, to see any actual belief being able
to satisfy in practice).
Another difficulty for the strong read-
ing of Hume’s thesis is that it implies that
the business of ‘Of Miracles’ is effectively
completed by the end of Part 1. In Part 1,
Hume proposes a criterion for what counts
as a miracle and points out that testimony
to a miracle is of its very nature testimony to
an occurrence that violates one of our best-
supported regularities. However, while Part
1 argues that the evidential benchmark for
miracles is set very high, it does not estab-
lish that this benchmark is impossibly high
or unattainable in practice. Part 1 at least
leaves open the theoretical possibility of
compelling testimony to the occurrence of
miracles. Thus, Part 1 draws out a conflict
2/9/2012 9:36:16 PM
 HUME ON MIRACLES
between two intuitions: one that testimony
is ordinarily worthy of high (nay decisive)
epistemic credit and another that scepticism
about the continuance of laws of nature is
hard to sustain. These two intuitions ordi-
narily command our assent without coming
into conflict. Thus, the problem posed by
miracle-testimony is that of how to adjudi-
cate between two ordinarily compelling intu-
itions forced into stark confrontation.
Fogelin argues that Part 1 aims only to
emphasize this tension and does not pro-
nounce on how it should be resolved. Part
2 then develops the case against miracles
by invoking empirical considerations about
observed testimony to extraordinary events.
In Part 2, Hume offers empirical grounds
why testimony to miracles is likely to be espe-
cially untrustworthy. For example, human
beings want to believe that miracles occur
because this leads to ‘agreeable emotion[s]’
(EHU 10.16 / 117), and this may explain
why history is littered with stories of ‘forged
miracles’:
The many instances of forged miracles,
and prophesies and supernatural events,
which, in all ages, have either been
detected by contrary evidence, or which
detect themselves by their absurdity,
prove sufficiently the strong propensity
of mankind to the extraordinary and
the marvellous, and ought reasonably to
beget a suspicion against all relations of
this kind. (EHU 10.19 / 118)
Hume also argues that ‘civilised’ societies
move away from supernatural and/or mirac-
ulous explanations:
It forms a strong presumption against
all supernatural and miraculous rela-
tions, that they are observed chiefly to
abound among ignorant and barbarous
234
9780826443595_Ch11_Final_txt_print.indd 234
nations; or if a civilised people has ever
given admission to any of them, that
people will be found to have received
them from ignorant and barbarous
ancestors, who transmitted them with
that inviolable sanction and authority,
which always attend received opinions.
(EHU 10.20 / 119)
Hume is thus arguing that the development of
civilization – and thus the rejection of ‘igno-
rant and barbarous’ traditions – involves
curtailing the natural impulse towards treat-
ing events as miracles. Hume then allows
himself a little sarcasm: ‘It is strange, a judi-
cious reader is apt to say, . . . that such pro-
digious events never happen in our days. But
it is nothing strange, I hope, that men should
lie in all ages.’ (EHU 10.21 / 119–20)
Thus far, of course, these empirical grounds
against accepting testimony to miracles will
apply to any miracle. Hume clearly thinks,
however, that empirical concerns about tes-
timony are especially acute when it comes to
specifically religious miracles. Of their very
nature, religious miracles are often promoted
to favour a particular cause, and thus we can
reasonably attribute substantial self-interest
to those putting forward the initial testimony.
But as Hume (EHU 10.21–3 / 120–1) makes
clear, where self-interest is involved in testify-
ing to miracles, one should be wary of that
testimony. Moreover, because miracles are
often associated with particular religions (or
sects) it is unsurprising that there is in fact a
great deal of conflict in the kind of miracle
reports that are found in the great religions
(EHU 10.24 / 121–2). Contrariety to natural
laws already places a stiff burden of proof on
miracle enthusiasts, but applying our knowl-
edge of human psychology to such testimony
makes the evidential situation still worse. We
know people tend to embellish stories likely
2/9/2012 9:36:16 PM

to incite wonder. Miracle-testimony is usu-
ally at several removes from witnesses – that
is, it testifies to miracles far away and/or long
ago.15 Furthermore, the diversity of miracle-
reports from different religious traditions
reduces the force of miracle-testimony in
any particular case. Finally, there are natural
explanations for why people want to believe
in specifically religious miracles:
The wise lend a very academic faith to
every report which favours the passion
of the reporter; whether it magnifies his
country, his family, or himself, or in any
other way strikes in with his natural
inclinations and propensities. But what
greater temptation than to appear as a
missionary, a prophet, an ambassador
from heaven? Who would not encounter
many dangers and difficulties, in order
to attain so sublime a character? Or if,
by the help of vanity and a heated imagi-
nation, a man has first made a convert
of himself, and entered seriously into the
delusion; who ever scruples to make use
of pious frauds, in support of so holy and
meritorious a cause? (EHU 10.29 / 125)
Religious believers face temptations that
testifiers to merely secular extraordinary
occurrences do not face. To appear as the
medium of divine revelation might tempt
otherwise utterly scrupulous witnesses to
embroider, distort, omit or invent. The point
here is two-fold. On the one hand, noth-
ing would exercise the passions as much as
being able to present oneself as a witness
to a religious miracle. On the other hand, if
one is already inflamed with religious pas-
sion, one has reason to accept testimony in
support of a religious miracle, regardless of
any epistemic basis for supposing this tes-
timony reliable. Both considerations offer
specific empirical grounds against trusting
235
9780826443595_Ch11_Final_txt_print.indd 235
HUME ON MIRACLES
testimony concerning the occurrence of reli-
gious miracles.
There is another reason why Hume thinks
that any actual miracle-testimony thus far
recorded cannot support religious hypothe-
ses, and this is that miracle-traditions are live
parts of the traditions of a host of competing
and mutually exclusive religious faiths: ‘all
the prodigies of different religions are to be
regarded as contrary facts, and the evidences
of these prodigies, whether weak or strong,
as opposite to each other’ (EHU 10.24 / 122).
Inevitably, different faiths will make different
claims about the doctrines to which their mir-
acles testify. Of course, it is conceivable that
the divinity might depute miracle-working
power to several miracle-workers. (After all,
Acts 8: 18–20 never describes Simon Magus,
the sorcerer who tried to buy spiritual power
from Peter, as attempting the logically impos-
sible.) However, the possibility of disparate
miracle-workers is something Hume does
not need to rule out a priori. Hume might
retort that several miracle-workers perform-
ing miracles in support of different faiths is
of itself enough to undercut the claim that
performing a miracle uniquely testifies to one
particular faith. It is in this sense that Hume
clearly regards testimony to miracles in sup-
port of different faiths as mutually destruc-
tive. However, it is not clear that reports of
miracles from different traditions are logi-
cally in conflict, even if the religions thereby
supported make conflicting claims.
So despite some textual support for the
strong reading, the evidence actually sup-
ports the weaker reading. Crucially, however,
the weak reading is just as challenging to the
idea that one should have a rationally held,
testimony-based belief in the occurrence of
a miracle, much less a religious miracle. For
even if it is possible for such a belief to be
rationally held, the epistemic bar that must
2/9/2012 9:36:17 PM
 HUME ON MIRACLES
be cleared in order for this to be the case is
so high that it is hard to see how, practically
speaking, any belief could clear it nor that
any belief actually has cleared it.16
In what follows we will explore some
further issues of relevance in this regard,
beginning with the supposed import Hume’s
remarks on the epistemology of testimony-
based belief in miracles have been thought
to have for the epistemology of testimony in
general.
2. HUME AND THE EPISTEMOLOGY
OF TESTIMONY
In the contemporary epistemological lit-
erature Hume is often characterized as
holding a rather radical view about the
epistemology of testimony. According to
this interpretation – popularized by C.A.J.
Coady17 – Hume’s remarks about testimony-
based belief concerning miracles reveal a
general reductionist view about the episte-
mology of testimony-based belief. There are
a number of competing ways of drawing the
reductionism/anti-reductionism distinction
in the contemporary epistemological litera-
ture on testimony, but very roughly reduc-
tionism holds that the epistemic standing of
a testimony-based belief must ultimately be
completely traced back to, and hence in this
sense reduced to, non-testimonial sources,
while anti-reductionism denies this claim.18
In practice, reductionism means that agents
can never simply rely on testimony if they
want to have beliefs which enjoy appropriate
epistemic support. Instead, they must always
seek independent grounds in favour of the
target belief. So, for example, in order to form
an epistemically sound belief on the basis of
testimony, one requires such independent
236
9780826443595_Ch11_Final_txt_print.indd 236
grounds as, for instance, previous first-hand
experience of the reliability of this informant
on this subject matter and collateral informa-
tion regarding the plausibility of the proposi-
tion testified to.
There is a good rationale that can be
offered for reductionism. For one thing, it is
widely held that testimony is only a trans-
missive, and thus not a generative, source
of knowledge. That is, testimony can only
at best transfer knowledge that has already
been acquired, but cannot be used to acquire
new knowledge (testimony is held to be simi-
lar to memory on this score).19 If that is right,
then it seems that the epistemic status of a
testimony-based belief must be ultimately
derived from a non-testimonial source.
Moreover, as Elizabeth Fricker has argued,
to reject reductionism is, it seems, to put
an epistemic premium on gullibility.20 Why
should the mere fact that someone testifies
that p give you any epistemic basis, however
modest, for believing that p?
There also seems good reason for thinking
that Hume’s argument against the rationality
of testimony-based beliefs in the existence of
miracles draws upon a reductionist account
of the epistemology of testimony. Consider,
for example, the following oft-cited passage:
‘our assurance in [testimony] is derived from
no other principle than our observation of
the veracity of human testimony, and of the
usual conformity of facts to the reports of
witnesses’ (EHU 10.5 / 111). Here Hume’s
sympathy with reductionism appears per-
fectly clear: we gain epistemic support from
testimony only when we have an independ-
ent basis for the belief so formed, such as
from our first-hand observations about the
‘veracity of human testimony’.
The key problem with reductionism, how-
ever, is that it appears to entail that not only are
we unable to gain epistemically well-grounded
2/9/2012 9:36:17 PM

beliefs about the occurrence of miracles, but
also about much else besides. After all, a great
deal of our beliefs were acquired through
testimony,21 and yet in a large number of cases
not only do we have no practical independent
(i.e. non-testimonial) means of verifying the
target proposition but we also cannot even
remember what the particular testimonial
source of our belief was. Accordingly, it seems
that reductionism entails a fairly widespread
scepticism about the epistemology of testimo-
nial belief.
Right now, for example, I believe – indeed,
I think I know – that the Orinoco River flows
through Venezuela, but although I know this
belief must have been acquired via testimony,
I have not the slightest recollection of the cir-
cumstances in which it was acquired. As such,
I can hardly have any independent epistemic
basis for trusting the testimonial source for
this belief. Moreover, although I have various
means at my disposal for checking this belief,
unless I actually take the trouble to travel to
South America to determine the matter in
person, I will almost certainly need to depend
on further instances of testimony, such as an
atlas or the testimony of a colleague, in car-
rying out these checks. Hence, it is hard to
see how I could rationally hold this belief by
reductionist lights. But what applies to this
belief will also apply to many other beliefs
that I hold. The upshot is that if reduction-
ism is true then we have far less epistemically
well-grounded, testimony-based beliefs than
we tend to suppose.
In itself, of course, this is a fairly indeci-
sive strike against reductionism. After all,
perhaps we do know a lot less than we think
we do. Still, reductionism is a contentious
thesis in the literature on the epistemology
of testimony, with the dominant camp being
by far anti-reductionism.22 Accordingly, inso-
far as we treat Hume’s argument against the
237
9780826443595_Ch11_Final_txt_print.indd 237
HUME ON MIRACLES
rationality of testimony-based beliefs in the
existence of miracles as trading on this the-
sis then that might be thought to be a pretty
serious count against it. Fortunately for
Hume, however, it is clear on closer inspec-
tion that not only does his argument against
the rationality of testimony-based belief in
miracles not trade on reductionism, but it is
also questionable whether he assents to this
general thesis about the epistemology of tes-
timony anyway. We will take these points in
turn.
The first point can be established by not-
ing that the anti-reductionist could quite
consistently endorse Hume’s argument
against the rationality of a testimony-based
belief in the existence of miracles (on either
the strong or the weak reading). After all,
what is key to anti-reductionism is only the
claim that in epistemically suitable circum-
stances one can gain an epistemically well-
founded testimony-based belief even while
lacking the full, independent, epistemic sup-
port demanded by reductionism. That is,
while the reductionist demands that inde-
pendent epistemic support is always required
for a testimony-based belief to be rationally
held, the anti-reductionist demurs and argues
that at least sometimes – that is, in epistemi-
cally propitious circumstances – such a belief
can be rationally held even in the absence
of independent epistemic support. It is thus
open to the anti-reductionist to claim that
the testimonial beliefs at issue when it comes
to miracles are not formed in the required
conditions and hence cannot benefit from the
greater epistemic permissiveness of the anti-
reductionist thesis.23
Indeed, there is in fact every reason for
thinking that an anti-reductionist would be
sensible to take this line when it comes to
testimony-based beliefs in the existence of
miracles. For given that miracles are by their
2/9/2012 9:36:17 PM
 HUME ON MIRACLES
nature the kind of event that does not nor-
mally occur – indeed, as we noted above, on
a certain conception of miracles (though not
the one that Hume had in mind) they may
well be the kind of event that cannot occur – it
follows that there is always a standing reason
to doubt the veracity of testimony regarding
miracles. The kind of testimonial cases that
the anti-reductionist wants to protect from
the zealous epistemic strictures of reduction-
ism, however, are precisely those where there
is no standing basis for doubt. It is one thing
to argue that in good epistemic conditions
one can gain an adequate epistemic basis for
one’s testimony-based belief even while lack-
ing any independent support for that belief,
and quite another to claim that an adequate
epistemic basis can be had without independ-
ent support where the epistemic conditions
are problematic in some way. The case of tes-
timony regarding the occurrence of a mira-
cle clearly falls into the second, problematic,
category, on account of there being in such a
case a standing doubt about the veracity of
the testimony in question. In such instances
the anti-reductionist might well demand –
indeed, would be wise to demand – that the
agents concerned should seek independent
grounds for their belief.
Moreover, note that making the reduc-
tionist demand for independent epistemic
support in the specific case of testimony
regarding miracles would not settle the mat-
ter of whether a belief in this regard could
be rationally held. For as the ‘eight days
of darkness’ example illustrates, once the
wealth of testimonial evidence becomes par-
ticularly extensive, it is far from clear that
one does lack an independent basis for tak-
ing the testimony at face-value. After all, one
can determine on a priori grounds that where
there is a uniform convergence in the testi-
mony regarding a significant event that is
238
9780826443595_Ch11_Final_txt_print.indd 238
offered by all people of all civilized nations,
then that is a good epistemic basis on which
also to form a belief in this proposition, even
if one cannot verify the truth of that propo-
sition oneself. Accordingly, even if we were
to apply reductionist standards to testimony-
based belief in the existence of miracles, it
is still not obvious that this would thereby
necessarily prevent such a belief from being
rationally held.
This brings us to our second point, which
is that it is questionable whether Hume
would in any case assent to the general
reductionist thesis regarding the epistemol-
ogy of testimony. The foregoing should make
it clear why. For if Hume’s argument against
the rationality of testimony-based belief in
the existence of miracles is compatible with
anti-reductionism, then clearly we have no
grounds for concluding, on the basis that he
appears to apply reductionist standards in
this particular case, that he is in general in
favour of reductionism. Indeed, we could just
as well conclude on this basis that he is an
anti-reductionist.
3. BAYESIAN VERSIONS OF HUME’S
ARGUMENT
It remains disputed whether Hume’s argu-
ment embodies correct probabilistic rea-
soning and whether it can be reconstructed
in Bayesian terms. Responses here can be
sharply polarized.24 Bayesianism takes its
name from Hume’s near-contemporary, the
Reverend Thomas Bayes (1702–61), whose
ideas about probability were first published
posthumously in 1763 by Hume’s corre-
spondent Richard Price (1723–91).25 Bayes’
Theorem aims to quantify how degrees of
belief in a hypothesis should vary in the light
2/9/2012 9:36:17 PM

of new evidence (e.g. that better-supported
theories should receive greater belief). In its
simplest form, Bayes’ Theorem says that the
probability a hypothesis receives conditional
upon evidence equals the likelihood of the
evidence if the hypothesis is true multiplied
by the hypothesis’s prior probability and
divided by the evidence’s prior probability.
More formally:
So, we want to determine the posterior prob-
ability testimony (‘t’) confers on a miracle
(‘M’) – i.e. Pr(M࿽t). Probabilities range
between 0 and 1, 0 corresponding to certain
falsehood and 1 to certain truth. One popular
misreading sets Hume’s prior probability for
a miracle – i.e. Pr(M) – at 0, from which nor-
mal Bayesian conditionalization inevitably
dictates that Pr(M࿽t) = 0.26 However, setting
Pr(M) = 0 treats Hume’s proofs as indefeasi-
ble and conflicts with the ‘eight days of dark-
ness’ example. Setting Pr(M) = 0 foists onto
Hume principles he did not hold. Consider
this quotation, for example:
To understand the structure of Hume’s
argument, it is helpful to try to specify
the form that Hume thinks inductive rea-
soning follows. As a starting point, recall
Reichenbach’s straight rule of induction:
if n As have been examined and m have
been found to be Bs, then the probabil-
ity that the next A examined will be B is
m/n. Corollary: If m = n, then the prob-
ability that the next A will be a B is 1. . . .
A (Hume) miracle is a violation of a
presumptive law of nature. By Hume’s
straight rule of induction, experience con-
fers a probability of 1 on a presumptive
law. Hence the probability of a miracle is
flatly zero. Very simple. And very crude.27
239
9780826443595_Ch11_Final_txt_print.indd 239
HUME ON MIRACLES
And very hard to square with Hume’s text.28
Hume cannot accord extremal (0 or 1) prob-
abilities to any empirical outcome and the
‘straight rule’ is not Humean.29
Although Hume talks of testimony ‘estab-
lishing’ a miracle, this does not mean testi-
mony must support a miracle beyond all
(reasonable) doubt – i.e. such that Pr(M࿽t) =
1. Rather, the occurrence of M is confirmed
if t makes it more probable than not that M
occurred – i.e. if the ‘balance of probabili-
ties’ favours M rather than ¬M. Therefore,
M is ‘established’ by testimony if Pr(M࿽t)
> 0.5. In turn, Pr(M࿽t) > 0.5 if Pr(t & ¬M)
< Pr(M) Pr(t࿽M). We can safely assume
Pr(t࿽M) = 1 – i.e. had M occurred ‘then tes-
timony to that effect would certainly have
been forthcoming’.30 If M is to support a
particular religious hypothesis, then presum-
ably testimony to M must be forthcoming,
given plausible assumptions about the raison
d’être of miracles. Therefore, following Peter
Millican, Hume’s rule is:31
Pr(M࿽t) > 0.5 → Pr(M࿽t) > Pr(~M࿽t)32
‘Of Miracles’ requires two distinct arguments:
first, that the prior probability of a mir-
acle will be low, and second, that testimony
to the occurrence of a miracle is likely to be
forthcoming even if no miracle occurred33
(conclusions Hume derives in Parts 1 and 2
respectively). Therefore, Pr(M) should be very
low and Pr(t & ¬M) very high; hence it is very
unlikely Pr(M࿽t) > 0.5. Hume’s argument
would be incomplete had he argued only for
low Pr(M) – he also needs reasons for high
Pr(t࿽¬M). For example, if we have very low
Pr(M) yet think it very unlikely t would be
forthcoming if M had not happened, Pr(M࿽t)
can still be as close to 1 as desired.
Critics of Hume often object that testi-
mony might come from so many independent
2/9/2012 9:36:18 PM
 HUME ON MIRACLES
observers that the possibility of collusion or
honest mistake can be made vanishingly low
and therefore Pr(M࿽t) raised vanishingly
close to 1.34 However, this Bayesian point is
neither damaging to Hume nor one Hume
need query. As noted above, he held that
miracles can be objects of compelling testi-
mony. The crucial point is whether we think
he thought testimony can never establish a
foundational miracle or (a weaker conclu-
sion) that testimony has never established a
foundational miracle. Even if we allow that
miracles might (theoretically) be made the
subject of compelling testimony in a way that
supports a particular religious hypothesis,
we might then look at the historical creden-
tials of (e.g.) biblical miracles. Imagine that
all nations around the Red Sea recorded the
miraculous destruction of Pharaoh’s armies,
that all nations recorded that the sun had
halted over Gibeon,35 or that Roman histori-
ans recorded that Christ resurrected Lazarus
before Tiberius. In such cases, the evidential
position would be far better quantitatively
and qualitatively than in any actual biblical
case. Whether or not the historical David
Hume would have become a believer had
such (counterfactual) testimony existed, a
Humean could set the evidential benchmark
at this level without in any way contradict-
ing Hume’s principles. Believers in miracles
might wonder why actual miracle-testimony
is not better. Any finite amount of testimony
could (theoretically) be improved upon.
However, the question is not ‘Why does
actual miracle-testimony admit any possibil-
ity of doubt?’, but rather ‘Why is actual mir-
acle-testimony so impoverished?’. Remember
that Hume’s main conclusion is not that
miracles cannot be established by testimony
but that they (almost certainly) cannot be so
established as to support a system of religion.
From responses to criticism, clearly Hume
240
9780826443595_Ch11_Final_txt_print.indd 240
regarded compelling miracle-testimony as
a hypothetical possibility, yet far from an
actuality:
There is no contradiction in saying, that
all the testimony which ever was really
given for any miracle, or ever will be
given, is a subject of derision; and yet
forming a fiction or supposition of a tes-
timony for a particular miracle, which
might not only merit attention, but
amount to a full proof of it. (LDH 1.349,
188: Letter to Hugh Blair, 1761)
4. CONCLUDING REMARKS
The above aims to situate Hume’s account of
testimony regarding miracles in the context
of his views of correct inductive practice, to
guard against some popular misunderstand-
ings, to explore how far he can be regarded as
a reductionist about testimony, and to offer a
few remarks on the relevance of Bayesianism
to his argument. The history of ‘Of Miracles’
suggests that certain misreadings of Hume
are too tempting to stay buried for very long,
and no doubt claims will continue to be made
that his account of miracles contradicts his
philosophy of induction, or that he achieves
a cheap victory by defining miracles out of
existence. However, these claims are proving
harder to sustain as time passes.
Besides its continuing relevance to phil-
osophy of religion and the epistemology
of testimony, ‘Of Miracles’ remains a key
source for Hume’s positive account of induc-
tion and a useful counter to any views that
would dismiss him as an unreconstructed
sceptic or unreflective scoffer. He anticipates
in important ways the increasing trend to
naturalistic, even evolutionary, explanations
of our cognitive strategies and their success.
2/9/2012 9:36:23 PM

Although by no means immune to criticism,
his ‘Of Miracles’ is in no danger of triviality
or obsolescence.36
NOTES
1
While Hume’s ‘relations of ideas’ and ‘matters
of fact’ resemble analytic and empirical truths
respectively, the pairs of terms are not syn-
onymous. See W. Ott, Causation and Laws of
Nature in Early Modern Philosophy (Oxford:
Oxford University Press, 2009), pp. 200–3.
2
Hume sought to extend Newtonian mechani-
cal description of physical phenomena into the
science of the mind. Thus, continuity of mental
habits may not merely resemble inertial persist-
ence of motion, but may be another instance of
the same phenomenon – see S. Buckle, ‘Marvels,
Miracles, and Mundane Order’, Australasian
Journal of Philosophy 79 (2001), pp. 1–31.
3
As he puts it, proofs are ‘arguments from expe-
rience as leave no room for doubt or opposi-
tion’ (EHU 6n / 56n).
4
‘[I]t may well seem strange that Hume should
use the word ‘proof’ for an argument which
he acknowledges to be fallible, i.e. non-
demonstrative. However, his training and later
employment as a lawyer would make it much
more natural for him to use the word in its
original and legal sense rather than in its now
familiar (to philosophers anyway) mathemat-
ical and logical sense. In a legal context, proof
is a matter of degree and a probable argument,
in Hume’s sense, can be thought of as yielding a
partial proof’ (B. Gower, ‘Hume on Probability’,
British Journal for the Philosophy of Science
42 (1991), pp. 1–19, p. 4n). In any case, Hume
here arguably follows our ordinary language
usage of ‘proof’: we would normally think it
an entirely adequate ‘proof’ that there are at
least three misprints on a page simply to point
out three separate misprints (cf. G.E. Moore,
‘Proof of an External World’, Proceedings of
the British Academy 25 (1939), pp. 273–300,
p. 275). According to Buckle, ‘Marvels,
Miracles, and Mundane Order’, p. 7, for Hume:
‘A proof is not a demonstration, not a deduct-
ively valid argument – and therefore not what
we would call a proof’. Buckle delineates two
241
9780826443595_Ch11_Final_txt_print.indd 241
HUME ON MIRACLES
Humean kinds of ‘proofs’: (i) where no doubt is
possible, such as G.E. Moore’s external world
proof; and (ii) ‘the limit case of probable argu-
ment, since it is that conclusion drawn from
experience where the uniformity of experience
is unalloyed’ (ibid., p. 8). For more on Hume’s
‘proofs’, see R. Fogelin, A Defense of Hume on
Miracles (Princeton: Princeton University Press,
2003), pp. 13–17.
5
At one point Hume (EHU 10.12n23 / 115n)
does suggest that the putative transgression of
natural law must be brought about by a super-
natural agent: ‘A miracle may accurately be
defined, a transgression of a law of nature by a
particular volition of the Deity, or by the inter-
position of some invisible agent.’ According to
this definition, a transgression of natural law
brought about in some arbitrary fashion would
not count as a miracle. We set this complica-
tion aside henceforth, not least because Hume
describes alleged miracles which involve no
supernatural agency, appear perfectly secular
and have no apparent volitional origin – e.g.
Queen Elizabeth’s imaginary resurrection (EHU
10.37 / 128).
6
See, for example, D. Coleman, ‘Hume, Miracles,
and Lotteries’, Hume Studies 14 (1988), pp.
328–46 and R.J. Fogelin, ‘What Hume Really
Said about Miracles’, Hume Studies 16 (1990),
pp. 81–7. Note that in more recent work
Fogelin (A Defense of Hume on Miracles)
rejects this reading.
7
D. Garrett, Cognition and Commitment in
Hume’s Philosophy (Oxford: Oxford University
Press, 1997), p. 152.
8
For more on the notion of a miracle, see
R.F. Holland, ‘The Miraculous’, American
Philosophical Quarterly 2 (1965), pp. 43–51;
R. Young, ‘Miracles and Epistemology’,
Religious Studies 8 (1972), pp. 115–26; M.
Levine, ‘Miracles’, in E. Zalta (ed.), The
Stanford Encyclopedia of Philosophy (online
at: http://plato.stanford.edu/entries/miracles/);
and D. Corner, ‘Miracles’, in B. Dowden and
J. Feiser (eds), The Internet Encyclopedia of
Philosophy (online at: www.iep.utm.edu/M/
miracles.htm).
9
C. Howson, Hume’s Problem: Induction and
the Justification of Belief (Oxford: Oxford
University Press, 2000), pp. 241–2.
2/9/2012 9:36:23 PM
 HUME ON MIRACLES
10
One might say that ice for the prince is marvel-
lous (or extraordinary), and not miraculous –
i.e. an exception to, not contradictory of, a
law of nature. With ice: ‘Contrariety would
lie in its going solid at a tropical tempera-
ture’, whereas ‘[r]esurrection involves more
than an extension of the complex processes
of nature: it involves something close to a
reversal of those processes’ (M.A. Stewart,
‘Hume’s Historical View of Miracles’, in M.A.
Stewart and J.P. Wright (eds), Hume and
Hume’s Connexions (Edinburgh: Edinburgh
University Press, 1994), pp. 171–200,
pp. 194–5). Likewise, ‘[o]ther miracle stories,
such as turning one substance into another,
do not present the same shock to the system,
the same undoing of the past, that resurrec-
tion stories do’ (ibid., p. 195). However, how
far Hume can justifiably draw a miracles/
marvels distinction is controversial. For views
con and pro see (for example) C.A.J. Coady,
Testimony: A Philosophical Study (Oxford:
Oxford University Press, 1992), chap. 10 and
D. Coleman, ‘Baconian Probability and Hume’s
Theory of Testimony’, Hume Studies 27
(2001), pp. 195–226, respectively.
11
Fogelin, A Defense of Hume on Miracles, p. 20.
12
Indeed, in the first of the two passages just
cited Hume explicitly suggests that testimony
can provide a ‘proof’ of the target event. To
the modern ear it might sound as though
Hume intended a deductive, and thus indefea-
sible, epistemic basis for the belief so formed,
but as we have already noted above, this is not
what Hume had in mind when he talked of
‘proofs’.
13
The locus classicus for this reading is Fogelin,
A Defense of Hume on Miracles. See also
A. Flew, ‘Hume’s Check’, The Philosophical
Quarterly 9 (1959), pp. 1–18 and Hume’s
Philosophy of Belief: A Study of His First
Inquiry, 2nd edn (Chicago, IL: St. Augustine’s
Press, 1997). Textual evidence suggests Hume
originally thought testimony to a miracle could
never make belief in a particular religious
hypothesis compelling but later adopted the
‘softer’ view that testimony to a miracle has
never made belief in a particular religious
hypothesis compelling. A crucial shift in
wording comes in the third (1756) edition
of the first Enquiry. In 1756 Hume says ‘no
242
9780826443595_Ch11_Final_txt_print.indd 242
testimony for any kind of miracle has ever
amounted to a probability, much less to a
proof’ (EHU 10.35 / 127), rather than using
the earlier (1748 and 1750) formulation ‘no
testimony for any kind of miracle can ever
possibly amount to a probability’ (emphasis
added).
14
Fogelin, A Defense of Hume on Miracles, p. 62.
15
In this regard, Hume implicitly refers to the
argument found in John Craig’s Theologiae
Christianae Principia Mathematica (1699),
that ‘the longer the chain of testimony
the less assurance it gives of the fact that
a miracle had occurred’ (F. Wilson, ‘The
Logic of Probabilities in Hume’s Argument
against Miracles’, Hume Studies 15 (1989),
pp. 255–75, p. 256). Craig calculated that any
testimony-derived probability for Christianity
would reach zero in c. 3150, and hence that
this was a plausible date for the Second
Coming. See S.M. Stigler, ‘John Craig and the
Probability of History: From the Death of
Christ to the Birth of Laplace’, Journal of the
American Statistical Association 81 (1986),
pp. 879–87.
16
For a nice overview of Hume’s argument
against the reasonableness of testimony-based
beliefs regarding the existence of miracles,
see P. Russell, ‘Hume on Religion’, in E. Zalta
(ed.), The Stanford Encyclopedia of Philosophy
(online at: http://plato.stanford.edu/entries/
hume-religion, sect. 6).
17
See C.A.J. Coady, ‘Testimony and
Observation’, American Philosophical
Quarterly 10 (1973), pp. 149–55 and
Testimony: A Philosophical Study.
18
For more on the reductionism/anti-reduction-
ism distinction in the epistemology of testi-
mony, see D.H. Pritchard, ‘The Epistemology
of Testimony’, Philosophical Issues 14 (2004),
pp. 326–48 and J. Lackey, ‘Testimonial
Knowledge’, in S. Bernecker and D.H.
Pritchard (eds), The Routledge Companion to
Epistemology (New York: Routledge, 2010).
19
See J. Lackey, ‘Learning from Words’,
Philosophy and Phenomenological Research
73 (2006), pp. 77–101 for a critique of this
view of testimony.
20
See E. Fricker, ‘The Epistemology of
Testimony’, Proceedings of the Aristotelian
Society, suppl. vol. 61 (1987), pp. 57–83;
2/9/2012 9:36:23 PM

‘Against Gullibility’, in B.K. Matilal and
A. Chakrabarti (eds), Knowing from Words
(Dordrecht: Kluwer Academic Publishers,
1994), pp. 125–61; ‘Telling and Trusting:
Reductionism and Anti-Reductionism in
the Epistemology of Testimony’, Mind 104,
pp. 393–411.
21
As Hume (EHU 10.5 / 111) says: ‘there is no
species of reasoning more common, more use-
ful, and even necessary to human life, than that
which is derived from the testimony of men,
and the reports of eye-witnesses and specta-
tors.’ All else being equal, testimony has, and
should have, exceptionally powerful epistemic
force.
22
For some of the main defences of anti-
reductionism, see C.A.J. Coady, ‘Testimony and
Observation’ and Testimony: A Philosophical
Study; T. Burge, ‘Content Preservation’, The
Philosophical Review 102 (1993), pp. 457–88;
27
R. Foley, ‘Egoism in Epistemology’, in
28
F. Schmitt (ed.), Socializing Epistemology: The
Social Dimensions of Knowledge, (Lanham,
29
MD: Rowman and Littlefield, 1994); and
J. McDowell, ‘Knowledge by Hearsay’,
in B.K. Matilal and A. Chakrabarti (eds),
Knowing from Words (Dordrecht: Kluwer,
1994), pp. 125–61. For a recent discussion of
how best to draw this contrast, see Pritchard,
‘The Epistemology of Testimony’. For an excel-
lent and up-to-date overview of the literature
on the epistemology of testimony, see Lackey,
‘Testimonial Knowledge’.
23
Fogelin seems to have this point in mind
when he claims that Hume’s reflections on the
30
epistemology of testimony-based beliefs in the
31
existence of miracles can be presented equally
well within both a reductionist and an anti-
reductionist account of the epistemology of
testimony. See Fogelin, A Defense of Hume on
Miracles, pp. 4–10, esp. p. 6n3).
24
Witness the title of Earman (2000): Hume’s
Abject Failure: The Argument Against Miracles
(Oxford: Oxford University Press, 2000).
25
T. Bayes; R. Price, ‘An Essay towards Solving
a Problem in the Doctrine of Chances’,
Philosophical Transactions of the Royal
Society of London 53 (1763), pp. 370–418.
26
J.H. Sobel, ‘On the Evidence of Testimony for
Miracles: A Bayesian Interpretaion of David
Hume’s Analysis’, Philosophical Quarterly
243
9780826443595_Ch11_Final_txt_print.indd 243
HUME ON MIRACLES
37 (1987), pp. 166–86, p. 176, suggests that
Pr(M࿽t) = i /(i + Pr(~M & t)), where ‘i’ is a
positive infinitesimal. However:
Sobel treats our assurance in a law of
nature as having probability infinitely close
to one, and the corresponding violation as
infinitely close to zero. I prefer to think of
Hume’s notion of proof as being simply an
argument with very high probability indeed.
This allows there to be such a thing as a
superior proof without resort to the un-
Humean notion of “infinitely close to”, and
thus allows one to treat seriously Hume’s
important example of the real possibility of
being convinced that the earth was covered
in darkness for eight days. (D. Owen,
‘Hume versus Price on Miracles and Prior
Probabilities: Testimony and the Bayesian
Calculation’, Philosophical Quarterly 37
(1987), pp. 187–202, p. 189n).
Earman, Hume’s Abject Failure, pp. 22–3.
For a countervailing view to Earman, see
Fogelin, A Defense of Hume on Miracles.
Cf. ‘Projection’: ‘when all the observed experi-
ments lead to the same outcome O, there is
probability 1 that an unobserved experiment
leads to O’ (A. Mura, ‘Hume’s Inductive
Logic’, Synthese 115 (1998), pp. 303–31,
p. 311). Hume’s view of projection seems close
to that of Carnap, who ‘did not view projec-
tion as an axiom of inductive logic, because
in his view it is completely counterintuitive.
Indeed, Carnap referred to it only as an
illustration of the severe disadvantages of the
straight rule’ (ibid.).
Howson, Hume’s Problem, p. 244.
P. Millican, ‘“Hume’s Theorem” concerning
Miracles’, Philosophical Quarterly 43 (1993),
pp. 489–95, p. 490. Note, however, that
Millican expresses reservations about this
reading (pp. 490–1, 495n8), and has recently
developed an alternative ‘type’ interpretation in
the context of a broader discussion of Hume’s
essay: see P. Millican, ‘Twenty Questions
about Hume’s “Of Miracles”’, in A. O’Hear
(ed.), Philosophy and Religion (Cambridge:
Cambridge University Press, 2011), pp. 151–92,
sects 4–9. But as Millican points out (sect. 19),
this alternative interpretation makes Hume’s rule
strictly incorrect, though potentially revisable in
a Humean spirit.
2/9/2012 9:36:23 PM
 HUME ON MIRACLES
32
Sobel, ‘Hume’s Theorem on Testimony Sufficient
to Establish a Miracle’, p. 232, summarizes
Hume thus: Pr(t) > 0 & Pr(M࿽t) > ½ → Pr(M)
> Pr(t & ~M) (substituting ‘M’ and ‘t’ for
Millican’s and Sobel’s ‘A’ and ‘α’ respectively).
33
A consequence of Bayes’s theorem is that
Pr(M/t) = Pr(M & t)/Pr(t) – i.e. the posterior
probability conferred on M by t equals the
prior probability of the joint occurrence of M
and t, divided by the prior probability of t.
34
Howson, Hume’s Problem, p. 245, credits this
insight to Charles Babbage (1791–1871). See
the extracts from Babbage’s Ninth Bridgewater
Treatise in Earman, Hume’s Abject Failure,
pp. 203–12.
244
9780826443595_Ch11_Final_txt_print.indd 244
35
‘The sun stayed in the midst of heaven,
and did not hasten to go down for about
a whole day’ (Joshua 10: 13, Revised
Standard Version). Although Hume does
not say so explicitly, this planetary-scale
miracle presumably stands to the ‘eight days
of darkness’ rather as the resurrections of
Lazarus and Jesus do to the resurrection of
Elizabeth I.
36
Many thanks to Peter Millican for discussions
and advice. We are grateful to the editors of
this volume for inviting us to contribute a
piece on this topic. This paper was writ-
ten while DHP was in receipt of a Philip
Leverhulme Prize.
2/9/2012 9:36:23 PM
 12
DAVID HUME AND
THE ARGUMENT TO DESIGN
Andrew Pyle
1. BACKGROUND
This proof always deserves to be men-
tioned with respect. It is the oldest, the
clearest, and the most accordant with
the common reason of mankind. It enli-
vens the study of nature, just as it itself
derives its existence and gains ever new
vigour from that source. It suggests ends
and purposes, where our observation
would not have detected them by itself,
and extends our knowledge of nature by
means of the guiding concept of a special
unity, the principle of which is outside
nature. This knowledge again reacts on
its cause, namely upon the idea which has
led to it, and so strengthens the belief in a
supreme Author of Nature that the belief
acquires the force of an irresistible convic-
tion. (I. Kant, Critique of Pure Reason)1
The argument to design, often misleadingly
called ‘the argument from design’, is indeed,
as Kant says, distinguished both by its great
antiquity and its manifest psychological
appeal to minds like ours. From time imme-
morial, it seems, men have contemplated the
beauty and order of the heavens, or admired
the intricate functional contrivances of the
parts of organisms, and seen such order and
contrivance as clear evidence of the workings
245
9780826443595_Ch12_Final_txt_print.indd 245
of an intelligent cause. Behind the more for-
mal statements of the argument always lurks
the same rhetorical question: how could such
things have come to be without the operation
of an intelligent cause?
Once we attempt to spell out this reason-
ing in more detail, we immediately see that
the argument to design is not a single argu-
ment but a family of related arguments. The
distinguishing feature of such arguments is
their strict empiricism. The premises must be
independently verifiable matters of empirical
fact such as the regular planetary orbits of
the Newtonian system, or the adaptation of
structure to function in the parts of animals.
No atheist will dispute the fact that the plan-
ets move in mathematically precise orbits, or
that the parts of the human eye are admirably
arranged to make vision possible. The infer-
ence from these facts is intended to conform
to our best canons of inductive reasoning,
making the argument to design – so we are
told – a truly ‘scientific’ proof of the exist-
ence of God. It is this strict adherence to the
principles of empiricism that distinguishes the
argument to design from most of the other
attempted proofs of the existence of God,
and explains its widespread appeal to think-
ers suspicious of the metaphysical subtleties
of the ontological and cosmological proofs.
2/9/2012 9:36:30 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
Arguments to design, I have claimed, are
best thought of as a family of related infer-
ences. Here are three common and familiar
ways of setting up the argument:
(1) As an inference to best explanation,
where the intelligent designer is posited
as the best explanation of the facts to be
explained (e.g. the functional contriv-
ance of the parts of animals).
(2) As an argument from improbability,
where the claim is that the facts to be
explained (various aspects of the order
of the universe) would be staggeringly
improbable if no intelligent cause were
involved; if our cosmos were simply the
result, for example, of the random colli-
sions of atoms.
(3) As an argument from analogy, in which
the resemblance between organisms and
artefacts (which are known to be prod-
ucts of intelligent design) grounds an
inference to a similar intelligent cause for
the former.
Clear statements of what we would recog-
nize as versions of the argument to design
can be found throughout the history of
human thought. In antiquity, the great phys-
ician Galen wrote a lengthy treatise On the
Usefulness of the Parts of the Body, not-
ing the manifestations of intelligent design
throughout human anatomy.2 In the famous
dialogue On the Nature of the Gods by the
Roman philosopher and orator Cicero, the
Stoic Balbus spells out the ‘infinite improb-
ability’ version of the design argument
against his Epicurean opponent Velleius.3
In the Middle Ages, another version of the
design argument, from the goal-directedness
of the parts of nature, makes an appearance
as the fifth of the famous ‘Five Ways’ of St
Thomas Aquinas.4 In seventeenth-century
Britain, scientists of the calibre of Robert
Boyle5 and Isaac Newton6 gave it their seal
246
9780826443595_Ch12_Final_txt_print.indd 246
of approval, and their work was in turn cited
by the Boyle Lecturers.7 In early eighteenth-
century Scotland, the argument was cham-
pioned by the celebrated Newtonian Colin
MacLaurin (1698–1746), whose Account of
Sir Isaac Newton’s Philosophical Discoveries
(1748) endorses the ‘plain argument’ for the
existence of a deity ‘from the evident contriv-
ance and fitness of things for one another’.8
This ‘Newtonian’ theism was very much part
of the intellectual background of the young
David Hume – one commentator has sug-
gested MacLaurin as a possible model for
the character of Cleanthes, the advocate
of natural theology in Hume’s Dialogues.9
And of course in our own time, the hypoth-
esis of ‘intelligent design’ is still defended, by
a variety of advocates, as a ‘scientific’ rival
to Darwin’s theory of evolution by natural
selection.
2. READING THE DIALOGUES
[Hume] Some of my friends flatter me,
that it [the Dialogues] is the best thing
I ever wrote. I have hitherto forborne to
publish it, because I was of late desirous
to live quietly and keep remote from all
clamour: For though it be not more excep-
tionable than some things I had formerly
published; yet you know some of these
were thought very exceptionable; and in
prudence, perhaps, I ought to have sup-
pressed them. I there introduce a sceptic,
who is indeed refuted, and at last gives up
the argument, nay confesses that he was
only amusing himself by all his cavils; yet
before he is silenced, he advances several
topics, which will give umbrage, and will
be deemed very bold and free, as well as
much out of the common road.10
[Hume] On revising them [the Dialogues]
(which I have not done these fifteen
2/9/2012 9:36:30 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
years) I find that nothing can be more
cautiously and more artfully written.11
Although the Dialogues were published
only posthumously, we know from Hume’s
correspondence that he intended to pub-
lish them during his lifetime. With this in
mind, we need to take seriously his state-
ment, in the above-quoted letter to Adam
Smith, that the work was ‘cautiously’ and
even ‘artfully’ written. Irreligious and anti-
clerical writers had to use, at this period, a
variety of literary devices to get their views
into print. One might resort to irony or to
misdirection,12 one might report the views
of real or imaginary doubters, or of course
one might write in dialogue form. The obvi-
ous advantage of the dialogue form is that
it enables the author to distance himself
from the views of any particular character.
We know that Hume, who had a deep and
abiding love of the classics, used Cicero’s
De Natura Deorum as his model in writing
the Dialogues.13 If asked for his own opin-
ion, he might have simply repeated Cicero’s
own warning:
Those who ask for my own opinion
on every question merely show exces-
sive curiosity. In a discussion of this
kind our interest should be centred not
on the weight of the authority but on
the weight of the argument. Indeed the
authority of those who set out to teach is
often an impediment to those who wish
to learn.14
Despite such warnings, readers of a work of
philosophy written in dialogue form will still,
almost inevitably, ask the obvious question:
‘Who speaks for the author?’ Hume him-
self says, in a well-known letter to his friend
Gilbert Elliot, that ‘I make Cleanthes the hero
of the dialogue’,15 and of course the Dialogues
247
9780826443595_Ch12_Final_txt_print.indd 247
close with Philo’s notorious ‘U-turn’ and a
summing-up by the narrator Pamphilus in
Cleanthes’ favour. Contemporary reviewers
were not taken in for one moment by these
literary devices – the testimony of Hume’s
earliest critics is solidly in favour of the ‘Philo
is Hume’ reading of the work.16 Following
Dugald Stewart,17 however, a number of
commentators came to endorse the rival
‘Cleanthes is Hume’ reading, which in turn
was flatly rejected by Norman Kemp Smith
in the introduction to his 1935 edition of the
Dialogues. Kemp Smith’s own verdict was
very clear:
I shall contend that Philo, from start to fin-
ish, represents Hume; and that Cleanthes
can be regarded as Hume’s mouthpiece
only in those passages in which he is
explicitly agreeing with Philo, or in those
other passages in which, while refuting
Demea, he is also being used to prepare
the way for one or other of Philo’s inde-
pendent conclusions.18
After Kemp Smith’s attack, the ‘Cleanthes is
Hume’ reading of the Dialogues has found
few defenders. This interpretation seems deaf
to the wit and irony of the work, and blind to
the ‘cautious’ and ‘artful’ way in which it was
presented. But if we reject the ‘Cleanthes is
Hume’ reading, does it follow that we must
accept Kemp Smith’s ‘Philo is Hume’ read-
ing? By no means. A number of recent com-
mentators have argued that the Dialogues
should be read simply as dialogues, without
seeking for a single authorial voice or a sim-
ple authorial message.19 Modern opinion is
divided between the ‘Philo is Hume’ reading
and the ‘No single authorial voice’ reading.
But if Philo has by far the best of the argu-
ments, then this may be, in the final analysis,
a distinction without a difference. We set out
to read the Dialogues simply as dialogues,
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
and end up being convinced that Philo was
right on most or all of the key points at
issue.
3. CLEANTHES’ FIRST STATEMENT OF
THE ARGUMENT TO DESIGN
[Cleanthes] Look around the world:
Contemplate the whole and every part of
it: You will find it to be nothing but one
great machine, subdivided into an infi-
nite number of lesser machines, which
again admit of subdivisions, to a degree
beyond what human senses can trace
and explain. All these various machines,
and even their most minute parts, are
adjusted to each other with an accu-
racy, which ravishes into admiration all
men, who have ever contemplated them.
The curious adapting of means to ends,
throughout all nature, resembles exactly,
though it much exceeds, the productions
of human contrivance; of human design,
thought, wisdom, and intelligence. Since
therefore the effects resemble each other,
we are led to infer, by all the rules of
analogy, that the causes also resemble;
and that the Author of nature is some-
what similar to the mind of man; though
possessed of much larger faculties, pro-
portioned to the grandeur of the work,
which he has executed. By this argument
a posteriori, and by this argument alone,
do we prove at once the existence of a
deity and his similarity to human mind
and intelligence. (DNR 2.143)
The three characters in the Dialogues,
Demea, Cleanthes and Philo are all in at
least verbal agreement that there is a God,
previously defined as the cause of the exist-
ence and order of our universe. Setting aside
the literally incredible doubts of the most
248
9780826443595_Ch12_Final_txt_print.indd 248
extreme sceptics, all men agree both that
there is an orderly physical universe, and
that there must be some cause of its existence
and order.20 So Cleanthes’ first statement of
the design argument is intended to prove to
Demea and Philo not that ‘there is a God’
(which neither of them ever denies), but that
‘[t]he relation between God and the world is
like that between a human engineer and his
creations’. It is by this analogical argument,
says Cleanthes, that we prove God’s natural
attributes such as wisdom and forethought,
and (hopefully) His moral attributes such as
justice and benevolence.
Cleanthes invites us to ‘[l]ook around the
world’, and to contemplate both the whole
and its parts. The universe as a whole, he
says, is ‘one great machine’; its parts are ‘an
infinite number of lesser machines’. Two dif-
ferent aspects of order are being invoked
here. In describing the physical universe as
‘one great machine’, Cleanthes is presum-
ably referring to the Newtonian model of
the solar system. This manifests order in the
sense of perfect regularity. The Newtonian
achievement of explaining the orbits of plan-
ets, moons and even comets in terms of a
few simple mathematical laws was one of
the most spectacular intellectual feats of
all time, and made a deep impression on
eighteenth-century thinkers. But the parts
of the solar system show no clear indica-
tions of being machine-like in the sense of
having their parts designed for a purpose.
We can contemplate the rings of Saturn, or
the moons of Jupiter, and have no notion of
what they might be for, and indeed no great
assurance that they are for anything at all. If
we seek for evidence of intelligent contriv-
ance we must turn our gaze from the whole
to the parts, to the ‘lesser machines’ of which
the ‘great machine’ is composed. The argu-
ment shifts from cosmology to biology.
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
That the parts of organisms serve definite
functions, and can usefully and fruitfully be
regarded as existing for the functions they
serve, is one of the central assumptions of the
biological sciences.21 Teleological assump-
tions of this kind have been the staple of bio-
logical reasoning from Galen on the kidney
to Harvey on the valves in the veins. Exactly
as in a well-designed machine, each part or
organ has the structure it needs in order to
serve its given function. It is here in biology
that the argument to design finds its clearest
evidence and its strongest grounds. Here we
take ourselves to know the ends (survival and
of course reproduction) that must be served by
the parts of organisms, and are thus in a pos-
ition to make tolerably well-informed judge-
ments about excellence of design. A cat needs
its acute senses, sharp claws, swift reflexes
and superb agility in order to be a success-
ful hunter of mice. This ‘curious adapting of
means to ends’ in nature, says Cleanthes, is a
precise analogue of the similar (though mark-
edly inferior) adaptation of means to ends in
human engineering. By the rules of analogy
we are entitled to infer that as the known
cause of the watch is to the watch, so is the
unknown cause of the cat to the cat.
Philo responds to Cleanthes’ first state-
ment of the argument to design with two
objections. The first is straightforward, and
simply illustrates an obvious weakness of
arguments by analogy. The strength of an
analogical argument, he reminds us, depends
on the degree of similarity of the cases. If you
exaggerate the resemblances and overlook or
play down real differences, you will fall into
errors of over-generalization. He illustrates
his point with a well-chosen example from
the life sciences. Once the circulation of the
blood had been established in animals, an
analogous ‘circulation of the sap’ was sup-
posed to provide the key to the physiology of
249
9780826443595_Ch12_Final_txt_print.indd 249
plants. But the analogy was spurious – plants
turn out to have striking differences from
animals. This particular argument from anal-
ogy yielded only errors (DNR 2.144).
Philo’s second objection goes deeper. Our
inferences from effects to causes are, he insists,
always based on our past experience, and on
the patterns and regularities we have observed
in our previous experience of the world.
If we see a house, Cleanthes, we con-
clude, with the greatest certainty, that it
had an architect or builder, because this
is precisely that species of effect, which
we have experienced to proceed from
that species of cause. But surely you will
not affirm, that the universe bears such a
resemblance to a house, that we can with
the same certainty infer a similar cause,
or that the analogy is here entire and per-
fect. The dissimilitude is so striking, that
the utmost you can pretend to is a guess,
a conjecture, a presumption concerning
a similar cause . . . . (DNR 2.144)
When Cleanthes protests at Philo’s use of
terms like ‘guess’ and ‘conjecture’ to character-
ize his argument, Philo responds by insisting
that his objection is simply the consequence of
taking the empiricist theory of knowledge ser-
iously. As far as a priori reasoning goes, says
Philo, anything can cause anything:
order, arrangement, or the adjustment
of final causes is not, in itself, any proof
of design; but only so far as it has been
experienced to proceed from that prin-
ciple. For aught we can know a priori,
matter may contain the source or spring
of order originally, within itself, as well
as mind does . . . . (DNR 2.146)
To restate his argument in a manner consist-
ent with empiricist principles, Philo explains,
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
Cleanthes must claim that experience teaches
us that ‘there is an original principle of order
in mind, not in matter’ (DNR 2.146). But
this new version of the design argument suf-
fers from two glaring defects. It rests on little
more than anthropomorphic prejudice, and
it rushes far too quickly from the part to the
whole.
What peculiar privilege has this little agi-
tation of the brain which we call thought,
that we must thus make it the model of
the whole universe? Our partiality in our
own favour does indeed present it on all
occasions: But sound philosophy ought
carefully to guard against so natural an
illusion (DNR 2.148).
Stone, wood, iron, brass, have not, at this
time, in this minute globe of earth, an
order or arrangement without human art
and contrivance: Therefore the universe
could not originally attain its order and
arrangement, without something similar
to human art. But is a part of nature a
rule for another part very wide of the
former? Is it a rule for the whole? (DNR
2.149)
Since we have no experience whatsoever
of the origin of worlds, Philo concludes,
we can draw no inferences at all regarding
their causes. The application of the design
argument to the universe as a whole seems
entirely precarious, resting on no empirical
basis in human experience. Cleanthes might
conceivably retreat at this point, and restrict
the design argument to the ‘lesser machines’
of which the ‘great machine’ is composed,
i.e. to the evidence of design manifest in the
parts of plants and animals. Here we have
a large class of instances of apparent con-
trivance (in artefacts and organisms alike),
and certain knowledge that in one subset of
these cases (artefacts) the contrivance is the
250
9780826443595_Ch12_Final_txt_print.indd 250
product of intelligent design. Analogy would
then give us grounds for thinking that intel-
ligence is present in the other cases (organ-
isms) too. Philo might retort that there are
rather striking differences between organ-
isms and artefacts, most notably in their
manner of coming-to-be, and that the ana-
logical inference from artefacts to organisms
still strikes him as an instance of anthropo-
morphic prejudice.
4. SECOND STATEMENT OF THE
ARGUMENT TO DESIGN: THE
‘IRREGULAR’ INFERENCE
[Maclaurin] It strikes us like a sensation;
and artful reasonings against it may puz-
zle us, but it is without shaking our belief.
No person, for example, that knows the
principles of optics and the structure of
the eye, can believe that it was formed
without skill in that science; or that male
and female in animals were not formed
for each other, and for continuing the
species.22
[Cleanthes] Consider, anatomize the
eye: Survey its structure and contriv-
ance; and tell me, from your own feel-
ing, if the idea of a contriver does not
immediately flow in upon you with a
force like that of sensation. The most
obvious conclusion surely is in favour
of design; and it requires time, reflec-
tion, and study, to summon up those
frivolous, though abstruse, objections
which can support infidelity. Who can
behold the male and female of each
species, the correspondence of their
parts and instincts, their passions and
their whole course of life before and
after generation, but must be sensible,
that the propagation of the species is
intended by nature? (DNR 3.154)
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
Part III of the Dialogues begins with
Cleanthes regrouping, and marshalling
arguments for a counter-attack against
Philo’s scepticism. Just as some works of
art please us although they break the rules,
so too, Cleanthes insists, some arguments
will persuade us, even if cast in ‘irregular’
form (DNR 3.155). The similarity between
the works of nature and those of art is ‘self-
evident and undeniable’, the objections of
the sceptics mere quibbles, like those of the
ancient Greek philosophers who famously
denied the existence of motion. Doubts
of this kind are to be met not by counter-
argument but by the presentation of striking
examples. Cleanthes launches Part III with
his two famous thought-experiments: the
voice from the clouds and the living library.
Even without the support of a generaliza-
tion from experience, Cleanthes insists, we
would all hear the voice from the clouds as
conveying a message, and the volumes of the
living library as an indication of an autho-
rial intelligence.
When Cleanthes, here following Maclaurin
almost to the letter, claims that the idea of a
contriver for the eye strikes us ‘with a force
like that of sensation’, he is directing our
attention to two aspects of the resulting belief:
passivity and immediacy. Certain objects,
he thinks, simply strike us as products of
design, even without any background of rel-
evant experience or any conscious processes
of inductive or analogical reasoning. If this
is correct, the argument to design may rest
on the secure foundation of a natural belief,
and be as such immune to sceptical doubts.
Cleanthes, it appears, thinks that the belief in
design may be natural to humans. But could
this also have been Hume’s opinion?
There is no doubt at all that Hume accepts
the existence of some natural beliefs. Key
examples are our belief in the reality of
251
9780826443595_Ch12_Final_txt_print.indd 251
the external world and our assumption of
the uniformity of nature. Our confidence is
entirely unshaken when we come to realize
that neither reason nor experience provides
adequate grounds for these assumptions.
Sceptical arguments against external world
realism and the uniformity of nature can puz-
zle us, but they can never convince us. Why
not? Because, as Hume says in a famous line
from the Treatise, ‘Nature, by an absolute
and uncontrollable necessity has determin’d
us to judge as well as to breathe and feel’
(THN 1.4.1.7 / 183).23 If natural judgements
are involuntary, they cannot be subject to the
usual norms of belief-formation. The sceptic
tells me that I ought to suspend judgement
regarding the truth or falsehood of the prop-
osition ‘This ball will fall if I release it’, but
experience tells me that I do in fact form the
strong and confident expectation that it will
fall. Natural judgements thus enjoy a privi-
leged position in our network of beliefs, and
a special immunity to sceptical doubts.
But could Hume have thought that our
belief in design is another natural belief, on a
par with our beliefs in the external world, the
regularity of nature, and the existence of other
minds? This ‘natural belief’ interpretation of
the Dialogues was first proposed in a famous
paper by Ronald Butler,24 and has recently
been championed by Stanley Tweyman.25 The
‘natural belief’ interpretation has two signifi-
cant virtues. It fits precisely with Cleanthes’
strategy in Part III: there can be little doubt
that Cleanthes is encouraging us to regard
belief in an intelligent designer of our world
as so deeply engrained in human nature that
it would be folly to attempt to doubt it. On
this view, the sceptic’s doubts are akin to
Zeno’s doubts about the reality of motion,
mere artefacts of the philosopher’s study, not
objections that demand serious study and
detailed point-by-point rebuttal. And the
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
‘natural belief’ interpretation explains neatly
Philo’s famous U-turn in Part XII, where he
admits that his earlier doubts were mere ‘cav-
ils’ and that he too accepts the existence of
intelligent design in nature. If belief in design
is a natural belief, the sceptic can consist-
ently reject the arguments but accept their
conclusion. So Philo’s notorious U-turn is no
problem for the ‘natural belief’ reading of the
Dialogues.
But could Hume himself have regarded the
belief in an intelligent designer of nature as
another natural belief? The evidence, both
textual and philosophical, is negative. In
the introduction to the Natural History of
Religion, Hume notes that theism is a very
widespread belief among human societies,
but he then adds this crucial qualification:
The belief of invisible, intelligent power
has been very generally diffused over the
human race, in all places and all ages; but
it has neither perhaps been so universal
as to admit of no exception, nor has it
been, in any degree, uniform in the ideas,
which it has suggested. Some nations
have been discovered, who entertained
no sentiments of Religion, if travellers
and historians are to be credited; and no
two nations, and scarce any two men,
ever agreed precisely in the same senti-
ments. It would appear, therefore, that
this preconception springs not from an
original instinct or primary impression
of Nature . . . since every instinct of this
kind has been found absolutely univer-
sal in all nations and ages. . . . The first
religious principles must be secondary
(NHR 134).
In 1751, while composing the Dialogues,
Hume found himself engaged in correspond-
ence on precisely this issue with his friend
Gilbert Elliot. The textual evidence provided
252
9780826443595_Ch12_Final_txt_print.indd 252
by this well-known letter against the ‘natural
belief’ reading of the Dialogues is very pow-
erful. Hume here admits a ‘propensity’ of the
human mind to see design in nature, but adds
that:
. . . unless that Propensity towards it [the
design argument] were as strong and
universal as that to believe in our Senses
and Experience, [it] will still, I am afraid,
be esteem’d a suspicious Foundation.
We must endeavour to prove that this
Propensity is somewhat different from
our Inclination to find our own Figures
in the Clouds, our Face in the Moon, our
Passions and Sentiments even in inani-
mate Matter. Such an inclination may, &
ought to be control’d, & can never be a
legitimate Ground of Assent.26
Our tendency to anthropomorphize aspects
of nature is a staple of the poetic imagina-
tion, but in the natural sciences it has been a
fertile source of errors such as the medieval
belief that ‘Nature abhors a vacuum’. There
may be a natural tendency here, but Hume
insists that it is only a natural weakness that
can and should be controlled, and could
never provide a legitimate ground of assent.
The classic study of this question is that of
John Gaskin.27 He lists four marks or criteria
of a Humean natural belief. They must be:
(1) Ordinary beliefs of common life.
(2) Incapable of rational justification in the
face of sceptical doubts.
(3) Indispensable in practice for everyday
life.
(4) Universally held and culturally invariant
among humans.
Our beliefs in the external world, the uni-
formity of nature, and the existence of other
minds fit all four of these criteria extremely
2/9/2012 9:36:31 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
well. Theism fares much worse: it manifestly
fails criteria (3) and (4), and arguably fails
(1) and (2) as well. We should conclude, with
Gaskin, that the belief in design cannot rea-
sonably be construed as a ‘natural belief’ in
the strong Humean sense. It may be a ‘natu-
ral belief’ in some weaker sense, in that we
humans do seem to have a strong propensity
to read nature in anthropomorphic terms,
but this may be a mere natural weakness of
significance for anthropology but not for
metaphysics or epistemology.
5. THE DESIGN ARGUMENT AND
THEISM
[Philo] In a word, Cleanthes, a man, who
follows your hypothesis, is able, perhaps,
to assert, or conjecture, that the universe,
sometime, arose from something like
design: But beyond that position he can-
not ascertain one single circumstance,
and is left afterwards to fix every point
of his theology, by the utmost licence of
fancy and hypothesis. (DNR 5.168–9)
[Kant] Now no one, I trust, will be so bold
as to profess that he comprehends the
relation of the magnitude of the world as
he has observed it (alike as regards both
extent and content) to omnipotence, of
the world order to supreme wisdom, of
the world unity to the absolute unity
of its Author, etc. Physico-theology is
therefore unable to give any determin-
ate concept of the supreme cause of the
world, and cannot therefore serve as the
foundation of a theology which is itself
in turn to form the basis of religion.28
In many respects, Part V of the Dialogues is
the most straightforward and unproblematic.
Here the sceptical Philo launches a series of
253
9780826443595_Ch12_Final_txt_print.indd 253
arguments for the conclusion that natural
theology is unable to provide the sort of sup-
port that the defenders of established reli-
gions such as Christianity demanded from
it. If we simply start from the phenomena of
nature, and argue without bias and precon-
ception, we have no way of establishing the
traditional list of attributes of the God of the
great monotheist tradition. Indeed, if we take
the argument from analogy seriously, it may
in fact provide better support for a variety of
heterodox views.
The first and most obvious problem con-
cerns God’s supposed infinitude. Since we are
inferring the cause from the effect, and the
effect (at least ‘so far as it falls under our cog-
nisance’) is finite, we ought to infer that the
cause is finite (DNR 5.166). Since an infin-
ite mind – if such a thing exists – would be
radically unlike ours, the analogy is much
stronger and clearer for a finite designer
than for an infinite one. Cleanthes does not
respond to this objection in Part V, but later
in the Dialogues, in the discussion of evil in
Part XI, he shows himself quite prepared
to drop all talk of infinity in our discussion
of the divine attributes, and embrace the
hypothesis of a finite god. ‘I have been apt to
suspect,’ Cleanthes admits, that ‘the frequent
repetition of the word infinite, which we
meet with in all theological writers, to savour
more of panegyric than of philosophy, and
that any purposes of reasoning, and even of
religion, would be better served, were we to
rest contented with more accurate and more
moderate expressions’ (DNR 11.203).
But Philo is only warming up. Along with
God’s infinity must go his supposed perfec-
tion. If we are arguing from the phenom-
ena of nature without bias or prejudice,
our world might well strike us as mani-
festly imperfect in any number of respects.
And even if we were to judge the design of
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
our world to be perfect (which would fly in
the face of experience) we would have no
grounds for assuming a corresponding per-
fection in the designer. In human technology,
perfection of design is generally the result of
a long, slow process of trial and error, with
each generation of designers building on the
work of its predecessor. The marks of design
in our world might reflect a similar history:
‘Many worlds might have been botched and
bungled, throughout an eternity, ’ere this
system was struck out’ (DNR 5.167). We
have no way, Philo concludes, of assigning
‘where the probability lies’ between the rival
hypotheses: excellent first-time design versus
a long story of world-building by trial and
error. But if we take the argument from anal-
ogy seriously, the probability must lie with
world-building by trial and error rather than
perfect first-time design.
The unity of God is obviously a central
pillar of the theology of the great mono-
theist tradition running through Judaism,
Christianity and Islam. But, Philo asks
Cleanthes, ‘what shadow of an argument . . .
can you produce, from your hypothesis, to
prove the unity of the Deity?’ (DNR 5.167).
In human affairs, many men frequently col-
laborate to produce great feats of architec-
ture and engineering. If I come across such a
work, I might naturally assume that the req-
uisite knowledge and skills were distributed
around the design team rather than being
concentrated in a single head. Unity of design
need not indicate unity of designer. Here we
need to note that in his Natural History of
Religion, Hume had dismissed polytheism as
an ‘arbitrary supposition’, and had endorsed
the argument from unity of design to unity
of designer (NHR 2.138). So what are we to
make of the apparent contradiction between
the two texts? If we deny that Philo speaks
for Hume, we could see his flirtation with
254
9780826443595_Ch12_Final_txt_print.indd 254
polytheism as a mere jeu d’esprit, but it is
noteworthy that Cleanthes is given no seri-
ous reply in defence of monotheism. I think
the Philo of the Dialogues is very close to
Hume’s own views, in which case the explicit
monotheism of the Natural History must be
taken with a large pinch of salt.
Philo’s final point takes us right into the
heart of Christian theology. A man who rea-
sons in accordance with your principles, he
warns Cleanthes, has no reason to believe in
divine providence:
This world, for ought he knows, is very
faulty and imperfect, compared to a
superior standard; and was only the first
rude essay of some infant Deity, who
afterwards abandoned it, ashamed of his
lame performance; it is the work only of
some dependent, inferior Deity; and is the
object of derision to his superiors: it is the
production of old age and dotage in some
superannuated Deity; and ever since his
death, has run on at adventures, from the
first impulse and active force, which it
received from him . . . . (DNR 5.169)
The scholastic philosophers of the Middle
Ages drew a distinction between the cause of
the coming-to-be of a thing (causa secundum
fieri) and the sustaining cause of its ongoing
existence (causa secundum esse). The causes
of the coming-to-be of a child are its parents,
and the cause of the coming-to-be of a build-
ing is a builder. But the parents can desert or
disown the child, or even die, and the child
will continue to exist. By contrast, accord-
ing to the schoolmen, the cause of the very
being of sunlight is the sun: if the sun were to
cease to exist, so too would its light. (A mod-
ern example might be the relation between
a magnet and its magnetic field.) The prob-
lem for Cleanthes is that he represents the
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
relation between God and his Creation as
akin to that between a human designer and
his creations, that is, as the relation of causa
secundum fieri. But the theologian wants
the relation between God and his works to
be that of causa secundum esse – in Him,
as Saint Paul so memorably puts it, ‘we live
and move and have our being’.29 So beneath
the surface of this light-hearted gibe about a
‘superannuated deity’, Philo is making a deep
point about the inadequacy of the design
argument to serve Christian theology.
Part V ends with Cleanthes’ response to
Philo’s objections. Astonishingly, he claims a
sort of victory. Your flights of imagination,
he tells Philo, give me pleasure,
when I see, that, by the utmost indulgence
of your imagination, you never get rid of
the hypothesis of design in the universe,
but are obliged, at every turn, to have
recourse to it. To this concession I adhere
steadily; and this I regard as a sufficient
foundation for religion. (DNR 5.169)
It is the pious Demea who responds with
signs of horror, as he sees the implications
of this sort of natural theology for religious
faith and practice:
While we are uncertain, whether there is
one Deity or many; whether the Deity or
Deities, to whom we owe our existence,
be perfect or imperfect, subordinate or
supreme, dead or alive; what trust or
confidence can we repose in them? What
veneration or obedience pay them? To
all the purposes of life, the theory of
religion becomes altogether useless . . . .
(DNR 6.170)
Can we do any better for Cleanthes? With
regard to God’s infinity, he does not want
255
9780826443595_Ch12_Final_txt_print.indd 255
our assistance – he tells us quite clearly that
he does not think it helpful or enlighten-
ing to insist on the infinitude of the divine
attributes. With regard to unity, we must try
to find some marks or criteria – within the
products of human arts and crafts – which
indicate the unity of the designing mind.
This seems unpromising, not to say deeply
problematic. With regard to perfection, we
will find ourselves mired in the problems
of evil – of which more anon. But the most
fundamental problem for empirical natural
theology is Philo’s last and deepest objec-
tion. Cleanthes’ argument rests squarely on
the analogy between the products of human
craft (machines) and the products of sup-
posed divine craft (organisms). But the rela-
tion between human craftsmen and their
products is always that of causa secundum
fieri, and this does not and cannot give theo-
logians the relation between God and man
that they want. Even switching from a craft
metaphor to a biological one (‘Our Father’)
does not help. Our parents will inevitably
become old and feeble and eventually die,
leaving us to our own resources. As Demea
sees clearly, Cleanthes’ style of natural theol-
ogy is all but useless as a support for ortho-
dox religious belief and practice.
6. ALTERNATIVE POSSIBILITIES
[Philo] A continual circulation of matter
in it [the world] produces no disorder:
A continual waste in every part is inces-
santly repaired: The closest sympathy is
perceived throughout the entire system:
And each part or member, in perform-
ing its proper offices, operates both to its
own preservation and that of the whole.
The world, therefore, is an animal, and
the Deity is the soul of the world,
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
actuating it, and actuated by it. (DNR
6.170–1)
[Philo] But if we must needs fix on some
hypothesis; by what rule, pray, ought we
to determine our choice? Is there any
other rule than the greater similarity of
the objects compared? And does not a
plant or an animal, which springs from
vegetation or generation, bear a stronger
resemblance to the world, than does any
artificial machine, which arises from rea-
son and design? (DNR 7.177)
At the end of Part V, Cleanthes thinks he is
winning the argument, since his sceptical
opponent Philo never gets rid of ‘the hypoth-
esis of design in the universe’. In Parts VI
to VIII, Philo takes his sceptical critique of
the argument to design still further, arguing
that arguments from analogy can support
naturalistic as well as supernatural accounts
of the order of our world. The world, says
Cleanthes, resembles a machine in various
respects, so analogy suggests that its cause is
an intelligent designer. But the world, Philo
retorts, also resembles an animal or a veg-
etable, in which case analogy suggests that
it owes its origin to generation or vegeta-
tion. We favour the hypothesis of intelligent
design because it mirrors our human way of
designing and making artefacts. But this par-
tiality for the design hypothesis may be mere
human prejudice. A race of super-intelligent
spiders might favour the hypothesis that the
world was spun from the bowels of a vast
spider. ‘Why an orderly universe may not be
spun from the belly as well as from the brain,
it will be difficult for him [Cleanthes] to give
a satisfactory reason.’ (DNR 7.180–1) Within
our experience, objects showing functional
contrivance of parts arise from four distinct
sources: Reason (the pocket watch), Instinct
(the spider’s web), Generation (the cat) and
256
9780826443595_Ch12_Final_txt_print.indd 256
Vegetation (the oak tree). To insist that
Reason must somehow be involved behind
the scenes in all cases in which order arises
from Instinct, Generation and Vegetation is
merely question-begging.
Philo goes on, in Part VIII, to revisit the
old Epicurean hypothesis, albeit with a few
important modifications. If we assume the
existence of a finite stock of atoms, recom-
bining endlessly over infinite periods of time,
then all possible combinations of atoms will
be tried out sooner or later (DNR 8.182).
Among those possible combinations of
atoms, some will be relatively stable, able
to maintain themselves in their respective
environments. If all possible combinations
have been tried out over the ages, and only
viable ones have survived and reproduced,
then, says Philo, we have the beginnings of a
reductive and non-teleological account of the
apparent teleology of nature:
It is in vain, therefore, to insist upon
the uses of the parts in animals or veg-
etables, and their curious adjustment to
each other. I would fain know how an
animal could subsist, unless its parts
were so adjusted? Do we not find, that
it immediately perishes whenever this
adjustment ceases, and that its matter
corrupting tries out some new form?
(DNR 8.185)
Here Philo is simply repeating Book Five of
Lucretius’ De Rerum Natura, that classic
statement of Epicurean Atomism.30 When
Cleanthes attacks the Epicurean hypoth-
esis, Philo is quick to respond that he is not
in fact championing it but merely raising it
as an alternative hypothesis to Cleanthes’
theism. The same warning holds, of course,
for the world-animal and world-vegetable
hypotheses. As a sceptic, says Philo, it is my
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
business to come up with alternative scenar-
ios to the claims of the dogmatist. My dog-
matic opponent, he says, claims to know that
hypothesis h1 is true, on the basis of evidence
e. In a purely sceptical spirit, says Philo, I
raise the suggestion that the same evidence e
is equally compatible with rival hypotheses
h2, h3, or h4. For the sceptic, the more rival
hypotheses in play the better. And if we are
rational, we will have to take the ‘catch-all’
hypothesis (that the truth is not even among
the hypotheses we are currently considering)
very seriously. As Philo puts it, with perhaps
a hint of exaggeration:
Without any great effort of thought, I
believe that I could, in an instant, pro-
pose other systems of cosmogony, which
would have some faint appearance of
truth; though it is a thousand, a million
to one, if either yours or any of mine be
the true system (DNR 8.182).
Is Philo’s final position one of strict neutral-
ity between the rival hypotheses, leading to
a simple suspension of judgement and the
conclusion that the whole subject (the origin
of worlds) is simply beyond our faculties?
Or are there hints and suggestions in the text
that would lead us to think that, although
all our hypotheses are admittedly precari-
ous, some have better grounds in experience
than others? In his recent study Spectres of
False Divinity, Thomas Holden argues that
Philo (who he takes to be Hume’s spokes-
man throughout the Dialogues) accepts only
what he calls a ‘liminal’ natural theology,
in which we know no more of the intrinsic
nature of God (defined in relational terms as
the cause of our world) than of any unknown
object X.31 On this view, Philo’s scepticism
leads him to eschew entirely what Holden
calls ‘core’ natural theology, that is, attempts
257
9780826443595_Ch12_Final_txt_print.indd 257
to say anything positive about the specific
nature and distinguishing properties of God.
This is, of course, a perfectly natural reading
of much of what Philo says throughout the
Dialogues, and is perhaps also what Philo
should say, assuming as Holden does that he
is Hume’s spokesman, arguing from Hume’s
own epistemological principles.
There remain, however, clear hints and
suggestions throughout the Dialogues of an
anti-theistic agenda that goes beyond the
mere sceptical suspension of judgement. In
Part II Philo betrays a clear materialist bias
in describing thought as ‘this little agita-
tion of the brain’ (DNR 2.148). In Part IV
he suggests that we should cease our search
for causes with the material world (DNR
4.161–2), rather than seeking to explain
its order in terms of the order of a suppos-
edly prior ideal world (the divine mind). In
Part VI he argues that the great advantage
of the world-animal and world-vegetable
hypotheses is that they rid us of the notion,
‘repugnant to common experience’, of a
mind existing without a body (DNR 6.171).
In Part VII we are told that the evidence of
experience is that Reason (intelligence) often
arises from Generation (sex), but never vice
versa (DNR 7.179). In all these cases, the
naturalistic counter-hypothesis to Cleanthes’
theism is raised within the context of a scep-
tical agenda, allowing Philo plenty of room
for evasion and retreat. If challenged, he can
say that he is not positively advocating these
naturalistic and materialistic hypotheses –
he is merely proposing them as a corrective
to the widespread and systematic prejudice
in favour of theism, and hence as an aid to
the suspension of judgement he is advocat-
ing. But the arguments can easily take on a
life of their own, and provide a cumulative
case for a naturalistic metaphysics. We can
readily grant that we do not know the nature
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
and properties of the cause or causes of our
world. We can grant that all of our hypoth-
eses have very low probabilities given the
evidence. But Philo does seem to be claiming
that, if we view the evidence of experience
without bias or preconception, naturalistic
hypotheses will seem more plausible (better
grounded in experience) than theistic ones.
This comparative judgement (the probability
of h2 given e is greater than the probability of
h1 given e) is of course perfectly compatible
with the assignment of very low probabilities
to both of the rival hypotheses. On this view,
Philo can be regarded as a sceptic with a defi-
nite leaning towards naturalism and atheism,
rather than a pure sceptic who is entirely
neutral between the rival hypotheses.
7. THE INFERENCE PROBLEM OF EVIL
[Philo] Look around this universe. What
an immense profusion of beings, ani-
mated and organised, sensible and active!
You admire this prodigious variety and
fecundity. But inspect a little more nar-
rowly these living existences, the only
beings worth regarding. How hostile and
destructive to each other! How insuffi-
cient all of them for their own happiness!
How contemptible or odious to the spec-
tator! The whole presents nothing but
the idea of a blind nature, impregnated
by a great vivifying principle, and pour-
ing forth from her lap, without discern-
ment or parental care, her maimed and
abortive children (DNR 11.211).
Parts X and XI of the Dialogues deal with
the issue of moral and physical evil, and fol-
low thematically from Part V, in which Philo
has raised the problem for Cleanthes of our
inability to infer the traditional attributes
258
9780826443595_Ch12_Final_txt_print.indd 258
of God from the argument to design. If
Cleanthes’ type of natural theology is to serve
a religion such as Christianity, it ought to
provide evidence from experience for God’s
moral attributes such as benevolence and
justice. A God without benevolence (who did
not intend the happiness of his creatures), or
a God without justice (who did not reward
the virtuous and punish the vicious) could
hardly be a suitable object of worship for us.
So if experience does not provide evidence of
benevolence and justice built into the very
fabric of the natural world, Cleanthes has a
problem.
Philosophical discussions of the so-called
‘problem of evil’ often focus on the compati-
bility or otherwise of the various evils (moral
and physical) of our world with the assump-
tion of an all-powerful, all-knowing, and
benevolent God. But this is only a ‘problem’
for those people already committed to a par-
ticular variety of theism. The problem facing
a natural theologian such as Cleanthes is the
quite distinct inference problem. Examine
the natural and human worlds without bias
or preconception, and ask what moral prop-
erties, if any, one can infer about its designer
and creator. In the discussion of evil in the
Dialogues, Philo raises the consistency prob-
lem, but the heart of his disagreement with
Cleanthes concerns – quite properly – the
inference problem. This is the real problem
facing the empirical natural theologian. In
Section 11 of the first Enquiry, Hume had
introduced a ‘friend who loves sceptical para-
doxes’ (EHU 11.1 / 132), and who wondered
whether experience testifies to a God or gods
motivated by justice. Here in the Dialogues,
the focus is on the question of whether the
phenomena of the natural world give us rea-
sons for inferring benevolence in its designer
and creator. The pious Demea and the scepti-
cal Philo agree that the sufferings of men and
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
beasts outweigh their pleasures in this world.
Demea thinks that this world is merely a
porch or antechamber to the next, and that
our sufferings in this world are merely trials,
to be amply compensated for in the afterlife
(DNR 10.199). Cleanthes, to his credit, sees
that this stratagem will simply not do: we are,
after all, trying to infer the moral properties
of God (the gods) from experience, not to fit
the facts of experience to an arbitrarily cho-
sen hypothesis (DNR 10.199–200). But, says
Philo, taking our world at face value, you
would never infer benevolence in its creator:
I will allow, that pain or misery in man is
compatible with infinite power and good-
ness in the Deity, even in your sense of
these attributes: What are you advanced
by all these concessions? A mere possible
compatibility is not enough. You must
prove these pure, unmixed, and uncon-
trollable attributes from the present
mixed and confused phenomena, and
from these alone. A hopeful undertak-
ing! (DNR 10.201)
Philo goes on to list a variety of natural phe-
nomena that seem, at least on the face of it,
difficult to reconcile with the hypothesis of
a God or gods motivated by benevolence.
At this point the Manichean theory of two
warring gods, one good and the other evil,
presents itself as a serious possibility.32 This
hypothesis, says Philo, has ‘more probabil-
ity than the common hypothesis’, but seems
inconsistent with the apparent unity of
design manifest in our world. The more natu-
ral conclusion is that God (the gods) is (are)
indifferent to human well-being:
There may four hypotheses be framed
concerning the first causes of the uni-
verse: that they are endowed with perfect
259
9780826443595_Ch12_Final_txt_print.indd 259
goodness, that they have perfect malice,
that they are opposite and have both
goodness and malice, that they have nei-
ther goodness nor malice. Mixed phe-
nomena can never prove the two former
unmixed principles. And the uniformity
and steadiness of general laws seems to
oppose the third. The fourth, therefore,
seems by far the most probable (DNR
11.212).
How seriously are we meant to take this
argument? Remember that it is Philo who is
speaking here, and Philo’s overall position is
sceptical, arguing throughout the Dialogues
that we lack the data to establish any sys-
tem of cosmology. Is he here abandoning his
sceptical principles? Thomas Holden thinks
that Philo cannot intend this argument to be
taken at face value, and that it is best read as a
‘parody’ of Cleanthes.33 Holden’s judgement
is in part determined by his own distinction
between ‘core’ and ‘liminal’ natural theology,
and his insistence that Philo/Hume utterly
rejects the former. If we regard Holden’s dis-
tinction as un-Humean, we can allow Philo
to advance hypotheses in the domain of
‘core’ natural theology, albeit with the usual
sceptical disclaimers. Philo’s message, as best
we can read it, seems to be the following.
If we were perfectly wise, we would realize
that theology is simply beyond our powers,
and abandon the subject altogether. But we
are not perfectly wise, and will always find
ourselves speculating about what, if any-
thing, lies behind the world of experience. If
forced to assess the relative merits of com-
peting hypotheses about God (defined as ‘the
cause of the existence and/or order of the
universe’), what else can we do other than to
ask which of our rival hypotheses seems best
to fit with the course of experience? On these
grounds, we find ourselves favouring the
2/9/2012 9:36:32 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
hypothesis of God as the world-soul against
the hypothesis of an immaterial God separate
from the world, because experience testifies
clearly against the hypothesis of disembodied
minds. Such a thing has no analogy within
the world of experience. Similarly, Philo can
say, our experience of the mixture of pleas-
ures and pains that life presents seems most
easily reconciled with the hypothesis that
God (the gods) is (are) indifferent to our weal
or woe. There is simply less argumentative
work to be done to reconcile this hypothesis
with experience than there is for the others.
The chief weakness in Holden’s account
is his failure even to address the clear – and
very marked – difference in tone between
Philo’s discussion of the inference from
nature to God’s intelligence and the corre-
sponding inference to God’s benevolence. In
our past disputes, says Philo to Cleanthes, I
had to exercise my ‘sceptical and metaphysi-
cal subtilty’, and even to invent ‘mere cav-
ils and sophisms’ to elude the force of the
inference from the adaptations of the parts
of animals and vegetables to the existence
of an intelligent designer (DNR 10.202–3).
A neutral observer might have concluded
that Cleanthes was right, and that experi-
ence does present strong reasons for belief in
intelligent design in nature. On the issue of
God’s moral properties, however, the sceptic
triumphs, and the would-be natural theolo-
gian is defeated:
[Philo] But there is no view of human
life, or of the condition of mankind,
from which, without the greatest vio-
lence, we can infer the moral attributes,
or learn that infinite benevolence, con-
joined with infinite power and infinite
wisdom, which we must discover by the
eyes of Faith alone. It is your turn now
to tug the labouring oar, and to support
260
9780826443595_Ch12_Final_txt_print.indd 260
your philosophical subtilties against the
dictates of plain reason and experience
(DNR 10.202).
I conclude that we have no reason to accept
Holden’s characterization of Philo’s argument
for the moral indifference of God (the gods)
as mere parody. There are of course gaps and
weaknesses in the argument, obvious short-
cuts and overlooked objections. But Philo is
always giving us two judgements, not one.
There is the over-arching sceptical agenda,
with its clear message that we just do not
know enough to make confident judgements
in matters of theology. But there is also the
equally clear insistence that, if we are forced
to rank a variety of competing hypotheses, we
can do so only on the basis of the testimony
of experience. And unless we want to ‘tug the
labouring oar’, we must admit that ‘the gods
don’t care about us’ has better grounds in
experience than ‘the gods love us’.
8. CONCLUSION
[Philo] If the whole of Natural Theology,
as some people seem to maintain, resolves
itself into one simple, though somewhat
ambiguous, at least undefined proposi-
tion, that the cause or causes of order in
the universe probably bear some remote
analogy to human intelligence: If this
proposition be not capable of extension,
variation, or more particular explica-
tion. If it afford no inference that affects
human life, or can be the source of any
action or forbearance: And if the anal-
ogy, imperfect as it is, can be carried no
farther than to the human intelligence;
and cannot be transferred, with any
appearance of probability, to the other
qualities of the mind: If this really be the
2/9/2012 9:36:33 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
case, what can the most inquisitive, con-
templative, and religious man do more
than to give a plain, philosophical assent
to the proposition, as often as it occurs;
and believe that the arguments, on which
it is established, exceed the objections
which lie against it? (DNR 12.227)
In the final part of the Dialogues, Philo
performs his notorious ‘U-turn’ and tells
Cleanthes that, for all his stated doubts and
objections, he too has a firm sense of the
evidence of divine intelligence provided by
‘the inexplicable contrivance and artifice
of nature’ (DNR 12.214). Galen’s inference
from the functional complexity of the parts
of animals to the existence of an intelligent
designer is, Philo now tells us, all but irresist-
ible, and only strengthened by subsequent
advances in anatomy and physiology (DNR
12.215). My doubts and objections, he tells
Cleanthes, were prompted by ‘my love of sin-
gular arguments’, and were never intended to
be taken seriously (DNR 12.214).
What are we to make of Philo’s U-turn?
If the ‘natural belief’ interpretation of the
Dialogues were correct, it would resolve
the problem and dispel any air of mystery
about Philo’s (and Hume’s?) position. On
this reading, Philo demolishes the arguments
for intelligent design but holds fast to the
belief in intelligent design, because that belief
is grounded in human nature rather than
in rational argument. (Likewise, of course,
Hume persuades us that we have no rational
grounds for believing in the uniformity of
nature, but is emphatically not trying to per-
suade us to become inductive sceptics.) The
U-turn, on this reading, would be simply the
reaffirmation of the natural belief in the face
of sceptical doubts that are dismissed as liter-
ally incredible. Instead of meeting the scep-
tic’s arguments, we just look at the functional
261
9780826443595_Ch12_Final_txt_print.indd 261
contrivance of the parts of the eye and allow
the resulting belief in design to strike us ‘with
a force like that of sensation’ (DNR 3.154).
But we have already seen both textual and
philosophical reasons to reject the ‘natural
belief’ interpretation. So how else are we to
explain Philo’s U-turn?
One reading that stays close to the text is
to take Philo’s words as perfectly sincere, and
to see Philo and Cleanthes as converging, in
Part 12 of the Dialogues, on a position that
might best be described as a sort of ‘weak
deism’. On this reading, some of Philo’s
doubts are, as he himself says, mere ‘cavils
and sophisms’, not intended to furnish seri-
ous objections to the argument to design. If
Philo’s doubts were never 100 per cent seri-
ous, then talk of a ‘U-turn’ is exaggerated.
Our overall subjective probability for the
proposition that ‘Nature manifests intelligent
design’ is somewhat reduced when we sur-
vey Philo’s ingenious list of alternative pos-
sibilities, but remains high enough to count
as firm assent. This reading has been champi-
oned by John Gaskin in his important study,
Hume’s Philosophy of Religion.34 It fits with
the text not only of the Dialogues, but also
of the Natural History of Religion, where the
argument to design is given in plain unvar-
nished form. ‘The whole frame of nature,’ we
are there told, ‘bespeaks an intelligent author;
and no rational enquirer can, after serious
reflection, suspend his belief a moment with
regard to the primary principles of genuine
Theism and Religion.’ (NHR 134)
Although Gaskin’s interpretation remains
perfectly permissible as a reading of the
text – as no doubt Hume intended – it does
overlook some of the textual evidence, and
in particular fails to do justice to the ‘art-
ful’ way in which the Dialogues were writ-
ten. When Philo concludes that ‘the cause or
causes of order in the universe probably bear
2/9/2012 9:36:33 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
some remote analogy to human intelligence’
(DNR 12.227), we naturally think that he is
making a significant concession to Cleanthes.
But Philo has already told us that there is ‘a
certain degree of analogy among all the oper-
ations of nature’, including ‘the rotting of a
turnip, the generation of an animal, and the
structure of human thought’ (DNR 12.218).
So the final tentative conclusion of a prob-
able case for a ‘remote’ analogy between the
cause of the world and human intelligence is
in fact no concession at all. The ‘U-turn’ is
exaggerated not (as Gaskin thinks) because
Philo’s initial doubts were never serious, but
because his eventual conversion is merely
verbal.
Two other lines of evidence support this
mildly ‘atheistic’ reading of the Dialogues.
In presenting his case against Cleanthes’ the-
ism, Philo advances and defends a variety of
naturalistic hypotheses regarding the cause
of order in the natural world. Officially,
these naturalistic hypotheses are put forward
within a sceptical agenda, as alternative pos-
sibilities intended not to elicit our assent but
to induce us to suspend judgement. But the
arguments can easily take on a life of their
own, and provide a cumulative case for the
conclusion that some form of naturalism
is more probable – on the evidence – than
theism.35 Contemporaries of Hume such as
Joseph Priestley were not fooled by Philo’s
supposed U-turn. In his Examination of Mr
Hume’s Dialogues (1780) he sums up Hume’s
intentions as atheistic:
although Philo . . . advances nothing but
common-place objections against the
belief of a God, and hackneyed declama-
tion against the plan of Providence, his
antagonists are seldom represented as
making any satisfactory reply. And when,
at the last, evidently to save appearances,
262
9780826443595_Ch12_Final_txt_print.indd 262
he relinquishes the argument, on which
he had expatiated with so much triumph,
it is without alleging any sufficient rea-
son; so that the arguments are left, as no
doubt the writer intended, to have their
full effect in the mind of the reader. And
although the debate seemingly closes in
favour of the theist, the victory is clearly
on the side of the atheist.36
The final piece of evidence against Philo’s
supposed theism comes from a key passage
from the end of Part IV of the Dialogues.37
The argument to design is presented as an
inference to best explanation, giving an
account of the material order of nature in
terms of the ideal order of the contents of
the divine mind. But, says Philo, ideal order
requires explanation every bit as much as
material order (DNR 4.160). Neither Reason
nor Experience provides any grounds for our
prejudice in favour of ideal order over mate-
rial. If we demand an explanation for func-
tional complexity as such, we must demand
such an explanation for the contents of the
divine mind, which will launch us into a
regress. But if this is the case, why not stop
with the material world rather than search-
ing for the cause of its order in a supposedly
prior ideal world? ‘An ideal system, arranged
of itself, without a precedent design, is not
a whit more explicable than a material one,
which attains its order in like manner; nor is
there any more difficulty in the latter suppo-
sition than in the former’ (DNR 4.164). Since
Cleanthes gives only a feeble response, this
unanswered objection to the theistic hypoth-
esis is left – as no doubt Hume intended – to
influence our eventual understanding of the
Dialogues as a whole.
Whether the Dialogues teach a weak
deism, or agnosticism,38 or, as I have argued,
show evidence of clear leanings towards
2/9/2012 9:36:33 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
naturalism and atheism, one point remains
completely clear. The argument to design,
whatever its persuasive power, affords ‘no
inference that affects human life, or can be the
source of any action or forbearance’ (DNR
12.227). This often-overlooked qualification
to the conclusion is in many respects the key
to the whole work. The all-important conclu-
sion that Hume clearly wants us to take from
his critical examination of the argument to
design is that the argument gives no support
whatsoever to any established religion, and
can provide no shred of a basis for any reli-
gious claims about our moral duties and obli-
gations. As Thomas Holden forcefully argues
in his impressive new book, Hume was def-
initely a moral atheist,39 believing that mor-
ality was a purely human affair, grounded in
our human sentiments and affections, and
inapplicable for deep reasons of principle to
the cause of the existence and order of the
universe.
NOTES
1
I. Kant, Critique of Pure Reason, trans. N. Kemp
Smith (London: Macmillan, 1976), p. 520.
2
Galen, On the Usefulness of the Parts of the
Body, trans. Margaret Talladge May (New
York: Cornell University Press, 1968).
3
Cicero, On the Nature of the Gods,
trans. H.C.P. McGregor, intro. J.M. Ross
(Harmondsworth: Penguin, 1972), p. 161.
Cicero’s work was of course Hume’s model for
his own Dialogues.
4
For the ‘Five Ways’, see Aquinas, Summa
Theologiae, Questions on God, ed. B. Davies
and B. Leftow (Cambridge: Cambridge
University Press, 2006), ‘Does God Exist?’,
pp. 20–7.
5
See, for example, R. Boyle’s Usefulness of
Natural Philosophy (Works, 6 vols, ed. T. Birch,
London, 1772), vol. 5, p. 427.
6
I. Newton, Mathematical Principles of Natural
Philosophy, trans. Motte, revised F. Cajori,
263
9780826443595_Ch12_Final_txt_print.indd 263
(Berkeley, Los Angeles and London: University
of California Press, 1934), 2 vols, General
Scholium, vol. 2, p. 544: ‘This most beautiful
system of the sun, planets, and comets, could
only proceed from the counsel and dominion of
an intelligent being’. For the argument from the
functional contrivance of the parts of animals,
see Opticks (New York, Dover, 1952), Query
28, pp. 369–70: ‘How come the Bodies of
Animals to be contrived with so much Art, and
for what ends were their several Parts? Was the
Eye contrived without Skill in Opticks, or the
Ear without Knowledge of Sounds?’
7
For the place of the design argument in the
Boyle Lectures, see my Introduction to a new
reprint of G. Burnet’s A Defence of Natural
and Revealed Religion, being an Abridgement
of the Sermons Preached at the Lecture
Founded by Robert Boyle, 4 vols (Bristol:
Thoemmes, 2000), vol. 1, pp. xxvii–xxx.
8
C. MacLaurin, An Account of Sir Isaac
Newton’s Philosophical Discoveries [1748]
(New York and London: Johnson Reprint
Corporation, 1968), p. 381.
9
See R.H. Hurlbutt III, Hume, Newton, and the
Design Argument (Lincoln, NB: University of
Nebraska Press, 1965).
10
Hume to William Strahan (his publisher), 8
June 1776 (LDH 2.322, 525).
11
Hume to Adam Smith, 15 August 1776 (LDH
2.334, 538).
12
An obvious example would be Voltaire dedi-
cating his play about the prophet Mohammed
to the Pope.
13
See C. Battersby, ‘The Dialogues as Original
Imitation: Cicero and the Nature of Hume’s
Scepticism’, in D.F. Norton, N. Capaldi and
W. Robison (eds), McGill Hume Studies (San
Diego: Austin Hill Press, 1979), pp. 239–52.
14
Cicero, On the Nature of the Gods, p. 73n3.
15
Hume to Gilbert Elliot, 10 March 1751 (LDH
1.154n10, 72).
16
For a brief summary of this evidence, see
my Hume’s Dialogues concerning Natural
Religion: A Reader’s Guide (London:
Continuum, 2006), pp 135–6. For the original
reviews, see S. Tweyman (ed.), Hume on
Natural Religion (Bristol: Thoemmes, 1996).
17
D. Stewart, Collected Works, ed. Sir W.
Hamilton, 11 vols (Edinburgh: Constable,
1854–60), vol. 1, p. 605.
2/9/2012 9:36:33 PM
 DAVID HUME AND THE ARGUMENT TO DESIGN
18
N. Kemp Smith, introduction to Hume’s
Dialogues Concerning Natural Religion
(Indianapolis: Bobbs-Merrill, 1947), p. 59.
19
See, for example, W.L. Sessions, Reading
Hume’s Dialogues: A Veneration for True
Religion (Bloomington and Indianapolis:
Indiana University Press, 2002), p. 212, and
Michel Malherbe, ‘Hume and the Art of
Dialogue’, in M.A. Stewart and J.P. Wright
(eds), Hume and Hume’s Connexions
(Philadelphia: Pennsylvania University Press,
1995), pp. 201–23.
20
In the Treatise, of course, external world
scepticism is dismissed in the following
famous passage. ‘We may well ask, What
causes induce us to believe in the existence of
body? But ’tis in vain to ask, Whether there
be body or not? That is a point, which we
must take for granted in all our reasonings.’
(THN 1.4.2.1 / 187; ‘Of Scepticism with
Regard to the Senses’).
21
Kant claims that this is a regulative assumption
of the life sciences. We cannot, he says, prove
that it is correct, but biologists must proceed as
if it were known to be correct. See the Critique
of Judgment, trans. W.S. Pluhar (Indianapolis:
Hackett, 1987), pp. 280–3.
22
Maclaurin, An Account of Sir Isaac Newton’s
Philosophical Discoveries; cited by Hurlbutt,
Hume, Newton, and the Design Argument, p. 42.
23
THN 1.4.1.7 / 183, ‘Of Scepticism with
Regard to Reason’.
24
R.J. Butler, ‘Natural Belief and the Enigma
of Hume’, Archiv für die Geschichte der
Philosophie 42 (1960), pp. 73–100.
25
S. Tweyman, Scepticism and Belief in Hume’s
Dialogues concerning Natural Religion
(Dordrecht: Kluwer, 1986), chap. 8, pp. 121–56.
26
LDH 1.155, 72.
27
J.C.A. Gaskin, Hume’s Philosophy of Religion,
2nd edn (London: Macmillan, 1988), p. 109.
264
9780826443595_Ch12_Final_txt_print.indd 264
28
Kant, Critique of Pure Reason, p. 523.
29
Acts 17: 28.
30
Titus Lucretius Carus, On the Nature of
Things, trans. Sir R. Melville, with an
Introduction by D. and P. Fowler (Oxford:
Oxford University Press, 1999), Bk 5, lines
837–48.
31
T. Holden, Spectres of False Divinity: Hume’s
Moral Atheism (Oxford: Oxford University
Press, 2010), chap. 2, pp. 19–47.
32
According to Pierre Bayle in his famous
Dictionnaire historique et critique, the consist-
ency problem of evil cannot be solved by
philosophy or rational theology. (See especially
the articles ‘Manicheans’ and ‘Paulicans’.) It
is only the eye of Faith that can reconcile the
manifest evils of our world with the belief in
a perfectly good God. Whether Bayle’s fideism
was sincere or merely tactical continues to
divide scholarly opinion.
33
T. Holden, Spectres of Divinity, p. 169n30.
For more traditional readings, see Gaskin,
Hume’s Philosophy of Religion, p 72n27,
and D. O’Connor, ‘Scepticism and Hume’s
Atheistic Preference’, Hume Studies 29 (2003),
pp. 267–82.
34
Gaskin, Hume’s Philosophy of Religion, chap.
7, pp 120–31.
35
I present a sketch of this case in my Reader’s
Guide to Hume’s Dialogues (London:
Continuum, 2006), pp. 130–2.
36
Joseph Priestley, Letters to a Philosophical
Unbeliever, quoted from Tweyman (ed.), Hume
on Natural Religion, p. 81.
37
Philo’s argument is repeated – without any
acknowledgment of Hume – in Richard
Dawkins’s The Blind Watchmaker (London:
Penguin, 1988), p. 141.
38
J. Noxon, ‘Hume’s Agnosticism’, The
Philosophical Review 73 (1964), pp. 248–61.
39
Holden, Spectres of False Divinity, pp. 4–9.
2/9/2012 9:36:33 PM
 13
PSYCHOLOGICAL EXPLANATIONS
OF RELIGIOUS BELIEF
David O’Connor
In his ‘Introduction’ to The Natural History
of Religion, Hume tells us that two questions
about religion ‘challenge our attention’(NHR
134). One is about religion’s ‘foundation
in reason’, the other is about its ‘origin in
human nature’. Taken together, Hume’s
investigations of the two questions make a
strong case against the reasonableness of reli-
gious belief.
Hume, however, does not claim this as his
aim in either investigation, as the outcome of
either investigation, or as the aim or outcome
of the two together, and there are passages in
his work that, taken at face value, suggest the
opposite, namely, that the reasonableness of
religious belief is not in question for him at
all. For instance, taking up the first question
in Dialogues concerning Natural Religion,
Hume frames it as a conversation just
about the nature of God, not the existence
of God, and has the sceptical Philo mouth-
ing agreement with the pious Demea that,
to ‘reasonable men’, the existence of God is
‘unquestionable and self-evident’ and proved
by the cosmological argument (DNR 2.142).
And when he refers to that first question in
The Natural History, Hume’s words are that
no reasonable person could doubt the evi-
dence of design in nature or the perfection of
265
9780826443595_Ch13_Final_txt_print.indd 265
the designer. But the idea that the existence
of God is not at issue in Hume’s discussion
of design in nature or of the nature of the
supposed designer is undercut by his actual
investigation of those things in the Dialogues
and in Section XI of An Enquiry concerning
Human Understanding. The result is that
Hume’s praise of the design argument is
mere praise1 and Philo’s professed agreement
with Demea comes to a good deal less than
it seems to promise, reaching at best to what
J.C.A. Gaskin calls an ‘attenuated deism’,
something far short of the concept of God
in Demea’s mind and in the monotheistic
religions.2
Taking up the second question in The
Natural History, Hume never states that
either the existence of God or the reasona-
bleness of believing that God exists is a tar-
get of his inquiry into the original causes of
religion. But those things are at issue in the
investigation nonetheless. This is because the
investigation provides good reason to think,
in Annette Baier’s words, that ‘. . . gods are
our minds’ progeny . . . .’3
The full subversive implication of Hume’s
analysis of religion comes with the con-
junction of his two investigations and their
respective findings. For the implication of the
2/9/2012 9:36:40 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
two together is that ‘. . . gods are [only] our
minds’ progeny . . . .’ Read in conjunction
with the Dialogues and its associated texts,
then, the message of The Natural History is
not just that understanding certain features of
the human mind, together with certain facts
about the living conditions of our long-ago
ancestors, is necessary in order to grasp the
essence of religion but that it is sufficient for
it. The former point is not especially subver-
sive or even controversial, but the latter is.
In addition to its implications for religion
specifically, Hume’s theory of the origin of
religion shows us The Natural History mak-
ing an important contribution to his central
philosophical project, the development of
a comprehensive theory of human nature.
However, I shall not go any farther into that
aspect of the book here.
1. HUME’S PHILOSOPHICAL
QUESTION ABOUT THE ORIGIN OF
RELIGION
Hume’s question about the origin of religion
is ostensibly psychological, as well as histori-
cal and anthropological. Without disputing
that, I approach it here as a philosophical
question and as no less central in Hume’s
philosophy of religion than his examina-
tion of the design argument, miracle reports,
and the other topics addressed in his inves-
tigation of religion’s foundation in reason.4
Specifically, I approach it as an investiga-
tion of the feasibility of a wholly naturalistic
theory of the origin, content and objects of
religious belief.5
Although Gaskin emphasizes The Natural
History’s important bearing on Hume’s
philosophical thinking about religion, he
represents a consensus when he characterizes
266
9780826443595_Ch13_Final_txt_print.indd 266
the book as ‘not itself philosophy’.6 I agree
that Hume’s theory of the origin of religion
is of philosophical importance, but I suggest
that its importance is in Hume’s philosophy
of religion, not just for his philosophy of reli-
gion. In this I side both with Keith Yandell,
who maintains that The Natural History is
‘the key to Hume’s philosophy of religion’7
and that Hume himself saw it the same
way, and with P.J.E. Kail, who also sees The
Natural History as an indispensible part of
Hume’s specifically philosophical thinking
about religion.8 But I do not subscribe to
Yandell’s claim that ‘even if Hume’s explan-
ation of religious belief and experience is
correct, nothing negative follows concern-
ing religious belief or religious believers, or
about religious experiences or experiencers’,9
and I think my argument here will vindicate
my disagreement.
At face value, neither question identified
in the ‘Introduction’ to The Natural History
is especially philosophical or more or less
philosophical than the other. After all, just
from the concept of religion’s foundation in
reason, we do not know that the foundation
is not conclusive evidence accessible to all.
And if it is, then there is no particular reason
to suppose an investigation of it would be a
philosophical investigation. But the investi-
gation of religion’s basis in reason is a philo-
sophical investigation. The reason is that
the issue is vexed and disputed in particular
kinds of ways, by its turning, for instance, on
the kind and strength of evidence that would
outweigh our confidence in a law of nature,
on whether and to what extent occurrences
in the natural world resemble things done on
purpose, as well as on whether, why and to
what extent a good being would justifiably
tolerate or permit the occurrence of evils that
seem both pointless and preventable, and so
on. Clearly, philosophical issues all.
2/9/2012 9:36:40 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
Hume’s question about the origin of reli-
gion seems to be a factual question aimed
at psychological, or historical, or anthropo-
logical discovery, not a conceptual question
oriented to philosophical reflection. Aimed at
finding an explanation of religion’s origin in
certain developments in the past, The Natural
History seems to promise a story of the devel-
opment of religion more or less on a par with
accounts of the origin of tool use among pri-
mates or the development of art, and so on.
Two points are worth noting. First,
Hume’s question about the origin of religion
does not suggest a psychological investiga-
tion to the exclusion of philosophy. After all,
Hume’s first philosophical book is A Treatise
of Human Nature. Thus, for Hume at least,
investigations into human nature can be as
much philosophical as psychological. And,
second, an investigation framed to seek the
origin, content and objects of religious belief
in human nature is self-consciously natur-
alistic. And considering that a concept of
‘invisible, intelligent power’ is at the heart of
religion, and that a concept of supernatural,
invisible, intelligent, providential power is at
the heart of the monotheistic religion that
especially interests Hume, such a naturalizing
project as he sets for himself in The Natural
History quickly raises questions about the
epistemic status of those concepts and about
what, if anything, their conjunction denotes.
My point here is not that Hume’s investiga-
tion turns up evidence that religion issues
from certain dispositions of human beings
and that the discovery has philosophical sig-
nificance, roughly the consensus reflected in
Gaskin’s view. It is that Hume’s project, being
an attempt fully to describe and explain the
core concepts of religion in naturalistic terms,
is philosophical from the start.
Seen this way, Hume’s project seems analo-
gous to the project to explain consciousness
267
9780826443595_Ch13_Final_txt_print.indd 267
without recourse to any Cartesian-style con-
cept of mind. That project is not just philo-
sophical, of course, and likewise Hume’s is
not just philosophical either. But, like the
project to explain consciousness without
bringing in substance dualism, it is no less
philosophical than anything else.
There are two stages in Hume’s project.
The first gives an account of the nature,
origin and development of religion, which,
while omitting nothing essential to religion,
refers to nothing supernatural.10 The second
gives an account of monotheism, which,
while omitting nothing essential to monothe-
ism, depicts it as an evolutionary product of
polytheism.
My argument here is that Hume’s project
is feasible in both of these respects and
seriously challenges the reasonableness of
religious belief. But the second of these con-
clusions depends on a point for which I have
no space here to argue. So I stipulate it. It is
that Hume’s investigation of religion’s foun-
dation in reason shows that the principal sup-
porting arguments for the existence of God
do not make a strong case.11 This stipulates
a large and much-contested proposition. But,
pleading space restrictions, I ask the reader
to go along with it for the sake of argument.
2. PROPENSITIES AND PROJECTION
The main philosophical question about
deities is whether any exist. Hume’s theory
that the idea of them originates in ourselves
does not, in itself, answer that question. After
all, human beings are the source of the ideas
of germs, electrons, photosynthesis and so
on, and there is good evidence in each case
that those ideas are representative of things
or processes existing outside the mind. We
2/9/2012 9:36:40 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
ourselves are also the source of the ideas of
animal spirits, the abominable snowman,
Limbo and so on, and there is good evidence
in each case that those ideas are not repre-
sentative of real things or places at all. What,
then, is it about Hume’s theory that, prima
facie, challenges the reasonableness of belief
in either gods or God? The answer is, his the-
ory of the particular ways that both the idea
of deities and the belief in them come about.
In this section I will describe his theory of
those ways, and in the two following sections
evaluate it.
The essence of the theory is that it is
various dispositions or ‘propensities’ inher-
ent in our psychological make-up that are
the source of our ideas of gods and God.
For present purposes, the most important
of those propensities is our disposition to
anthropomorphize. We talk of the sun trying
to break through the clouds, of the sea’s being
angry, of the wind moaning and, thanks to
William Wordsworth, of daffodils dancing in
the breeze. In Hume’s words, there is a ‘uni-
versal tendency among mankind to conceive
all beings like themselves, and to transfer to
every object, those qualities, with which they
are familiarly acquainted . . . We find human
faces in the moon, armies in the clouds . . . .’
(NHR 3.141). We nowadays in the so-called
developed world know that these things are
not really so. We know that in such cases
language is being used in non-literal ways.
But our knowing this reflects both our hav-
ing some decent measure of knowledge of
the actual causes of such things as the move-
ments of clouds and waves and daffodils, and
because we know well enough that different
kinds of language-use fit different kinds of
situations.
But our long-ago ancestors in primitive
societies did not know these things, and nei-
ther do many people today in various parts of
268
9780826443595_Ch13_Final_txt_print.indd 268
the world. Anthropomorphizing on purpose
takes both knowledge and a fair amount of
intellectual and linguistic sophistication,
while literal-minded anthropomorphism –
call it naive anthropomorphism – reflects a
lack of those things. Hume’s theory of the
origin of the idea of deities is that its source
is naive anthropomorphism, in conjunction
with feelings of dread, anxiety, fear and so
on, in the harsh and dangerous environment
of early human life. As Hume puts the latter
point in Part XII of the Dialogues, ‘. . . terror
is the primary principle of religion . . .’ (DNR
12.225–6).12 On this theory, the essence of
which is naive anthropomorphism plus fear
and anxiety, there is nothing religious in or
about the origin of religion.
In Hume’s description, life for our long-ago
ancestors meant a hard struggle for survival,
a short life-expectancy, and constant danger
of harm and death from disease, the weather,
predators, enemies and so on. The life of ‘a
barbarous, necessitous animal (such as a man
is on the first origin of society)’(NHR 1.136)
may not have been solitary, but it was poor,
nasty, brutish and short. True, the natural
environment also facilitates various life-en-
hancing activities, but, then and now, primi-
tive people (especially) are still largely at its
mercy insofar as coping with diseases, floods,
droughts, forest fires, predators of many
kinds, birth deformities, early death from
unexplained causes and so on is concerned.
The distinction between the beneficial
and the destructive sides of nature is import-
ant in Hume’s theory. For the theory empha-
sizes that our long-ago ancestors took the
former for granted as the natural order of
things, such that they had no particular
curiosity about such things and required
no explanation of them (NHR 1.136), no
heightened curiosity, for instance, about a
piece of land’s continuing to support crops
2/9/2012 9:36:40 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
of various kinds. Thus taking the benefi-
cial patterns of natural occurrences more
or less in their stride, it was for deviations
from those patterns that explanations were
sought. But this curiosity about the sources
of storms, thunder, lightning, floods, dis-
eases, droughts and so on – not everyday,
but common enough, occurrences – reflected
no desire for understanding in its own right.
Instead, it reflected the urgent practical
need to cope with difficult and inscrutable
circumstances and with the anxiety and ter-
ror they caused (NHR 2.139).
Vulnerability to predators and to enemies
is vulnerability to beings acting with pur-
pose. To us nowadays that kind of vulnera-
bility contrasts with vulnerability to disease,
floods, natural disasters and so on, for we
distinguish as a matter of course between
animate and inanimate nature. That we do
enables us to employ anthropomorphism
intentionally. But, in Hume’s theory, our
long-ago ancestors’ immediate and natural
response to thunder, lightning, floods and so
on was no different in kind from their imme-
diate response to being attacked by preda-
tors or enemies, to finding their livestock
slaughtered or stolen, their crops burned and
dwellings plundered, and so on. They saw
agency equally and equally immediately in
all instances of danger, harm and destruc-
tion. The stress on agency and immediacy is
important in Hume’s theory; in a threatening
environment rife with ignorance and anxiety,
it is ‘a natural propensity, if not corrected by
experience and reflection, [to] ascribe mal-
ice or good-will to every thing, that hurts or
pleases us’ (NHR 3.141).
This automatic, pre-reflective transfer-
ence of qualities of our own minds – intent,
anger, jealousy and so on – to the causes of
lightning, droughts, storms and so on is fre-
quently characterized in the literature on
269
9780826443595_Ch13_Final_txt_print.indd 269
Hume as projection, as are other important
concepts elsewhere in his work, the concept
of causality being the most obvious. This is
perhaps the aspect of anthropomorphism
that comes most readily to mind. Kail use-
fully calls it ‘feature projection’.13
Sometimes the agents responsible for some
harm are nowhere to be seen. Sometimes this
is because they have fled before they could
be discovered. And, to the naive anthropo-
morphist, sometimes it is because they are
invisible. These projected, invisible powers
are seen as being formidable, unpredict-
able and beyond the control of mere human
beings. Thus, to the fear and anxiety regard-
ing the destructive floods, fires, diseases and
so on against which early people were largely
powerless, there was added fear and anxiety
regarding the capricious, invisible agents seen
as behind those things.
In Hume’s theory, this idea of invisible,
intelligent power is the essential component
in the idea of the gods and of God, and his
theory of the origin of that idea is the essen-
tial point in his theory of the origin of reli-
gion (NHR 134, 3.142).14 It is a theory in
which there is nothing religious in the core
idea in religion.
This anthropomorphic idea of invisible,
intelligent power is not a one-size-fits-all
idea. It is the idea of invisible, intelligent
powers – plural, not singular. Consistent with
the perceived fact of various skills and apti-
tudes among people and animals, the invis-
ible, intelligent powers seen as being behind
storms, floods and so on are seen as having
various competencies and specializations.
Thus the source of a lightning storm result-
ing in a forest fire might not be supposed to
be the same as the source of an infant’s birth
deformity, but the latter source might well be
supposed to be behind stillborn livestock. In
short, different kinds of harmful departures
2/9/2012 9:36:40 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
from the more common, beneficial patterns
in the natural world are seen as being brought
about on purpose by various hidden powers,
on the model of the different kinds of haz-
ards and harms issuing from such different
kinds of discernible sources as insects, tigers,
powerful enemies, cunning enemies and so
on. In Hume’s words: ‘Such limited beings,
though masters of human fate, being, each
of them, incapable of extending his influence
every where, must be vastly multiplied, in
order to answer that variety of events, which
happen over the whole face of nature’ (NHR
3.142–3).
An important point underscored in this
description is that these projected, invis-
ible powers are seen as being both just as
real as visible agents of dangers and just as
much part of the natural world. The princi-
pal differences between them and their vis-
ible counterparts are their hiddenness and,
in many but not necessarily all cases, their
comparatively greater powers. The two kinds
of intelligent sources of danger are continu-
ous in the minds of our long-ago ancestors,
then, not just because both kinds operate in
the natural world, but because both are of
the natural world. Thus there is nothing oth-
erworldly or supernatural in or about these
invisible agents, as Hume describes our long-
ago ancestors’ idea of them.
If naive anthropomorphic transference or
feature projection accounts for the origin of
the idea of invisible, intelligent powers, the
core idea in all religions centered on deities,
what accounts for early humans believing in
such things? After all, having the idea of a
thing and believing in its existence are not
the same, and the latter is not entailed by the
former. I have the idea of mermaids, but no
belief that there are any.
Two features are central in Hume’s account
of early humans’ belief in invisible, intelligent
270
9780826443595_Ch13_Final_txt_print.indd 270
powers, namely, the felt seamlessness to them
of the two-part phenomenon, seeing-as and
believing-in, and the psychological helpful-
ness of belief in invisible, intelligent powers
in coping with a largely inscrutable and dan-
gerous world.
As an example of the first feature, take a
person’s seeing or feeling a particular place
as being haunted. Seeing the place this way,
the person does not want to be there, goes
out of his or her way to avoid it, and so on.
That is, the idea, the belief and the behav-
ioural disposition are experientially seamless;
typically, a person does not first think that a
place is haunted, then subsequently believe
that it is, and then be disposed to avoid it.
What explains this seamless phenomenon of
thinking, believing and inclining to behave a
certain way? The short answer is something
in the mind of the beholder, triggered by cer-
tain features of the place.
The second feature in Hume’s account of
the origin of belief in invisible, intelligent
powers is the belief’s helpfulness to those
early humans in coping with the anxiety
occasioned by the largely inscrutable natural
world. Believing that there are minds behind
the harmful occurrences in the natural world
opens up a possibility of influencing those
minds and thereby the possibility of mitigat-
ing the harms.15
To our long-ago ancestors, the powers of
these supposed, invisible agents are evident
everywhere, and so there is no escaping them.
Added to this, there is the unpredictability of
these supposed agents, as well as the inscrut-
ability of their motives and intentions in
causing the various destructions attributed to
them. The best coping strategy, then, is sub-
mission. Making offerings and sacrifices to
these projected beings, engaging in rituals of
praise and self-abnegation, ‘groveling’ (NHR
1.135), ‘trembling . . . sacrificing . . . praying’
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
(NHR 1.136), and appeasement (NHR
3.143), seem like prudent ways of coping
with such volatile and powerful beings, since
neither fighting back nor flight would be of
any use. The accommodationist pattern that
develops, then, is akin to a protection racket
voluntarily entered into. But what about the
obvious fact that the storms and fires and
inexplicable deaths and deformities con-
tinue? Perhaps those early people supposed
that things might have been much worse
without the offerings and so on. And any-
way, what is the alternative strategy, given
the invisibility, the unpredictability and the
destructive powers of the supposed beings in
question?
As noted, just as all enemies are not equal
or equally a threat, all the projected, invisible,
intelligent powers are not seen as the same.
So, to gain the most benefit from the protec-
tion racket dictates ingratiating oneself, one’s
family and one’s community with the more
rather than the less powerful among the sup-
posed, invisible agents. Such seeking of addi-
tional, and ultimately maximum, security in
this way reflects a process of inflation in the
powers that these supposed, invisible beings
are seen as possessing. As Hume puts the
point,
In proportion as men’s fears or distresses
become more urgent, they still invent
new strains of adulation; and even he
who outdoes his predecessor in swelling
up the titles of his divinity, is sure to be
outdone by his successor in newer and
more pompous epithets of praise. Thus
they proceed; till at last they arrive at
infinity itself, beyond which there is no
farther progress . . . (NHR 6.155).
Thus, over time, the logic of this inflation
leads to the idea of, the belief in and the rela-
tionship of submission to and dependence on
271
9780826443595_Ch13_Final_txt_print.indd 271
the projected, maximally powerful, invisible
agent. In this way, Hume conjectures that the
idea of a single agent behind the workings
of nature, the idea of a being with limitless
power and knowledge, develops, gradually
replacing the idea of a hierarchy of invisible
agents and its corresponding dispersal of
unseen powers and talents, and that the same
anxiety, fear, desire to curry favour and so
on that cemented belief in the lesser deities
of polytheism now cement belief in this uber-
deity of monotheism. Furthermore, on the
inflation model, the limitations on the pow-
ers of the various supposed, invisible agents
by dint of size, form, situation in time and
place and so on give way to the idea of, belief
in and devotion to a supposed, limitless, all-
powerful, immaterial, invisible agent, not
subject to any of the constraints of the natu-
ral world. Thus, according to Hume’s theory,
the original version of the idea of invisible,
intelligent powers as products of nature just
like the rest of us becomes replaced by the
inflated idea of a being beyond nature alto-
gether, a supernatural being. This, in out-
line, is Hume’s account of the evolution, in
response to various kinds of desperate cir-
cumstances, of the idea of, the belief in and
the dedication to a being supposed to be
beyond the natural world and subject to no
limitations whatsoever (NHR 8.159). An
important addition to the package of benefits
that comes with seeing oneself as under the
protection of such a being is protection from,
or at least the hope of protection from, the
ultimate threat of death itself.16
There is an irony in this evolutionary tale.
It is that the greater the inflation in the idea
of this single, all-powerful, all-knowing being,
the more remote and inaccessible that pro-
jected being comes to seem. After all, how is
a person to think of, to visualize, to relate to
a supposedly infinite and unchanging being,
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
not physical in any way, thus without body,
surface, shape, size, location and so on? And
the more remote and inaccessible it comes to
seem, the harder it becomes to believe that my
well-being and protection could matter to it
or that my pleas and entreaties could influence
it. According to Hume’s theory, the solution
which evolves is to surround this projected,
infinite, supernatural being with a variety
of supposed intermediaries and intercessors,
thereby giving the ordinary believer a way of
coping with this in itself largely unimaginable,
supposed being. The irony, then, is that while
the idea of a single, limitless, invisible power
evolves from the idea of a hierarchy of limited,
invisible powers, in its turn a version of the
latter subsequently evolves from the former as
a mechanism to make it seem accessible, thus
for the hope of protection to continue seem-
ing viable. Think of the ideas of angels, saints,
cherubim, of the idea of the supposed, infinite
being itself temporarily entering into human
form, and so on (NHR 8.159).
Earlier, I suggested that religious belief
is prima facie vulnerable to being under-
mined by Hume’s naturalistic project. To see
this, recall the example of the place seen as
haunted. Suppose you manage to persuade
the person that the place he saw as haunted
is really not, that it was all only in his mind.
His now believing that undercuts his belief
that the place is haunted. And Hume would
say that, with the exception of our beliefs in
the external world and induction, the point
holds in general, namely, our coming to see
a belief as reflecting only subconscious pro-
jection undercuts believing that it represents
anything outside the mind. The two excep-
tions that Hume makes are beliefs that, while
akin to religious belief in coming about by
projection, are unavoidable for any extended
period since, unlike religious belief, they are
instinctual or natural.17
272
9780826443595_Ch13_Final_txt_print.indd 272
Now suppose you manage to persuade a
religious believer that Hume is right about
the nature and origin of religious belief. True,
this is probably easier said than done. But if a
believer becomes persuaded that the gods are
only our minds’ progeny,18 we may expect
the same result as in the haunted place exam-
ple, since Hume is right that religious belief
is not instinctual.19
The key question, then, is whether Hume’s
naturalistic project is feasible as an expla-
nation of the nature and original causes of
religion. If it is not, perhaps there is some-
thing in religion that is irreducible to human
nature. If Hume’s reductionist project is fea-
sible for polytheism but not for monotheism,
it is plausible to look to monotheism’s super-
naturalism for the reason. But if the project
is feasible for both, then a full explanation
of religion can in principle be given with no
need to posit invisible, intelligent entities,
natural or supernatural. Then, if there is no
good reason to think that gods or God exist
anyway, that explanation represents a serious
challenge to the reasonableness of believing
in them or it.
3. OBJECTIONS TO HUME’S
NATURALISTIC PROJECT
Let us test the feasibility of Hume’s project
against the following five objections.
For the sake of argument, the first objec-
tion grants the following three things:
Hume’s propensity theory, his evolution
story of the development of the idea of a
single, supreme deity from the idea of a var-
iety of limited deities, and the improbabil-
ity of polytheism. Granting these things, the
objection challenges a conclusion we might
then draw, namely, that undercutting the
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
reasonableness of believing in gods under-
cuts the reasonableness of believing in God.
In essence, the objection is that it is the gen-
etic fallacy to suppose that monotheistic
belief, by virtue of evolving from polythe-
istic belief, suffers the same epistemic fate
as it.
Consider modern science and technology.
Both trace to origins no less humble than
those of the monotheistic religions. Modern
technology is a sophisticated form of primi-
tive tool use. And the ancestry of mod-
ern science includes magic and witchcraft.
Nowadays, in the industrially developed
parts of the world, we would judge it silly at
best to prefer magic to medicine, for instance,
or to think the one is just as good (or bad) as
the other in treating the sick and injured. So,
the objection goes, primitive origins do not
entail primitiveness later on.
Clearly, that is right. The question, then,
is whether the general rule covers the case
at hand. To test it, let us see if monothe-
ism’s relationship to its polytheistic origin is
(closely enough) analogous to the relation-
ship of modern science to its origin.
Where is the success in monotheism, anal-
ogous to the success of modern science? The
success in science being essentially epistemic
success, the question becomes: What is the
epistemic success of monotheism? As Hume
himself emphasizes, monotheism includes
a theory about the origin of the universe,
whereas polytheism explains only certain
kinds of occurrences within it (NHR 2.139,
4.147). So the question comes down to the
merits of the monotheistic theory of the ori-
gin of the universe, which is to say, in large
measure, to the case for an invisible, intel-
ligent, providential designer and cause of the
universe. And this is where Hume’s project
in the Dialogues and its associated texts, via
the conclusion we stipulated earlier, comes
273
9780826443595_Ch13_Final_txt_print.indd 273
into its full power in the present context. On
that stipulation, monotheism can point to
no record of epistemic success even remotely
like that of natural science. The first objec-
tion, then, fails. To think that undercutting
the core polytheistic belief in invisible, intelli-
gent powers undercuts the descendant, mon-
otheistic version too is not shown to commit
the genetic fallacy.
The analogy at the heart of the objection
is weakened also by Hume’s argument in
Part V of the Dialogues that such evidence
of natural design as there is points at least
as plausibly to designers as to a designer.20
In that argument, monotheism, as a theory
of the universe, is no advance over some ver-
sions of polytheism.
The second objection targets a supposi-
tion implicit in Hume’s theory. It is the sup-
position that religious belief, by originating
in the conjunction of our propensities, first,
to project certain features of our minds onto
the natural world and, second, to believe
that they are features of the natural world
because so believing psychologically benefits
us, falls short of some minimum standard of
epistemically acceptable belief. Like the first
objection, this objection grants for the sake
of argument that religious belief originates in
those propensities.
The objection can be developed in two
ways. One is to demand a full account,
including justification, of this supposed
standard of epistemically acceptable belief,
as a pre-condition of considering the suppos-
ition that religious belief fails to meet it. The
other emphasizes that, implicit in the suppos-
ition that religious belief falls short of this
supposed standard, there is the idea of some
other belief or beliefs not falling short, and
in comparison with which religious belief is
epistemically deficient. This way of develop-
ing the objection targets both that implicit
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
comparison and the supposed standard itself,
the latter indirectly. I will address this second
form of the objection.
Our belief in the external world is perhaps
our most basic substantive, as opposed to
formal, belief. So, if any non-formal belief
meets the supposed standard, surely it is
that belief. But Hume’s theory of the origin
of belief in the external world parallels his
theory of the origin of religious belief; both
originate in propensities of our minds, and
neither is justified by evidence or argument.21
Thus, the objection goes, with even our most
basic substantive belief failing to meet the
supposed standard, religious belief suffers no
epistemic downgrade by comparison and the
standard itself is suspect.
But the objection fails. The epistemic par-
ity on which it depends, namely, Hume’s
own point that neither belief in the external
world nor in God is justified by evidence or
argument, does not support the objection.
To see this, take the core content of reli-
gious belief, the belief in invisible, intelli-
gent power. That belief depends on believing
both in cause-effect relations and in caus-
ally linked things and events. That is, belief
in invisible, intelligent power presupposes
belief in the external world. The latter belief
is a necessary condition of the former, thus
of religious belief. But the dependence is not
reciprocated; belief in invisible, intelligent
power is not a necessary condition of belief
in the external world.
True, as the objection emphasizes, Hume
thinks that belief in the external world, like
religious belief, is not justified by evidence.
But the counterpoint to the objection is that,
nonetheless, belief in the external world is
epistemically more basic than belief in invis-
ible, intelligent power, being presupposed by
it. Thus, the epistemic standing of belief in
invisible, intelligent power, thus of religious
274
9780826443595_Ch13_Final_txt_print.indd 274
belief, can be measured relative to the epi-
stemic standing of belief in the external
world. Thus, the objection fails.
The third objection enlists the Pyrrhonian
point that, for any (philosophical) argument
that seems convincing now, we cannot dis-
miss the possibility that a successful counter-
argument to it will be developed. For the sake
of the objection, let us suppose that Hume’s
argument about the origin of religion is prima
facie convincing. But, for all we know, there
is a compelling counterpoint to it which, one
of these days, some philosopher will discover.
And, given that, we are within our epistemic
rights in not going along with Hume’s argu-
ment. An alternative, and more modest, for-
mulation of this objection emphasizes that,
for all we know, his naturalistic theory of the
development of religion may turn out to be
a special case of some wider theory. And if it
does, it may come to be seen in a quite dif-
ferent light.
In one respect, the objection is right. Even
if a naturalistic theory of the origin of reli-
gion is today unblocked or undefeated, for
all we know there is a successful limitation,
rebuttal or even refutation awaiting discov-
ery. But, absent specific evidence that those
are live, and not merely theoretical, possibili-
ties, the point is an idle one, applying to all
empirical theories. In another respect, how-
ever, the objection fails.
The key to the failure is the monothe-
ist’s disbelief in the polytheistic deities. The
believer in God typically disbelieves in Zeus
and Jupiter and all the pagan deities, as well
as in lesser supposed beings such as elves
and fairies, and plausibly will atribute both
the concept of and the belief in those sup-
posed beings to ignorance, superstition and
a tendency to naive anthropomorphism. But
Hume’s theory emphasizes that the sources of
the ideas of and belief in the pagan gods cover
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
the God of monotheism too. Consequently,
agreeing that Hume’s account of the origin of
polytheism is plausible and given the failure
of the first objection, this third objection will
be hard-pressed to accept polytheism as due
only to anthropomorphism while rejecting
that monotheism is.
The fourth objection reflects a point that
sometimes comes up in responses on religious
grounds to the theory of evolution by natural
selection. It is that the initial conditions at
the Big Bang and the laws by which natural
selection works are due to God. Applied here,
this becomes the claim that the propensities
in human nature to which the idea of invis-
ible, intelligent power traces exist by design
of God.22
Possibly, this is true. But, according to
our stipulation about religion’s foundation
in reason, we have on balance no good rea-
son to think there is a designer of human
nature or of nature at large. Thus, again, we
see the cumulative power of the two sides of
Hume’s philosophical investigation of reli-
gious belief.
The fifth objection refuses to restrict phil-
osophical investigation of religion to the two
questions set out in the ‘Introduction’ to The
Natural History. This objection emphasizes
that no philosophical investigation of religion
can be adequate if it omits consideration of
revelation and religious experience and that
Hume discusses neither.
But Hume is not entirely silent on these
topics.23 First, the issue of the reliability of
scriptural revelation seems closely analogous
to the reliability of miracle-testimony, which
he discusses in Section X of the Enquiry
concerning Human Understanding, while he
speculates about us ourselves having a puta-
tive miracle-experience through hearing ‘an
articulate voice . . . in the clouds’ in Part III of
the Dialogues (DNR 3.152). Second, Hume’s
275
9780826443595_Ch13_Final_txt_print.indd 275
discussion in The Natural History of people
seeing occurrences in nature as things brought
about on purpose seems analogous to at least
some kinds of religious experience.24
Given these prima facie proximities, we can
say with reasonable confidence how Hume
might respond to the objection. On analogy
with his analysis of miracle-claims, the gist of
what he might say about claims to base reli-
gion on revelation is that, while revelation is
possible in principle, there is no evidence for
it that withstands impartial scrutiny. And in
so far as basing religion on religious experi-
ence is concerned, he would emphasize that
religious experience could plausibly be traced
to the same projective propensities that, on
his theory, are the sources of the idea of,
and the belief in, invisible, intelligent power.
Furthermore, insofar as reports of religious
experience are concerned, and again on anal-
ogy with his analysis of miracle-claims, he
could be expected to emphasize that there is
no evidence that the reported experiences are
encounters with either a divine person or a
supernatural intermediary that would be per-
suasive to an impartial judge.
4. THE FEASIBILITY OF HUME’S
NATURALISTIC PROJECT
With none of the foregoing objections suc-
ceeding, our verdict is that Hume’s project to
give a wholly naturalistic account of religion
is feasible25 and that, together with his inves-
tigation of religion’s foundation in reason, it
gives us good reason to think that ‘. . . gods
are [only] our minds’ progeny . . .’. True, this
does not entail that there are no gods, since it
remains possible that our ideas of them could
have developed in the ways Hume describes
and such entities exist anyway.26 Apart from
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
contradictions, our ideas do not guarantee
what cannot exist. And the idea of an invis-
ible, intelligent agent is not self-contradictory.
Nonetheless, in the spirit of Ockham and
granting our stipulation about the various
arguments for the existence of God, Hume’s
naturalistic theory of religion, being feasible,
calls into serious question the reasonableness
of supposing that there is any more to the
idea of invisible, intelligent agency, including
the idea of invisible, intelligent, providential
agency, than what Hume proposes.
Hume’s contemporary, William Warburton,
in a letter to the publisher Andrew Millar in
1757, seems to see something like this clearly.
In Warburton’s interpretation, Hume’s ‘design’
in The Natural History is ‘to establish natural-
ism, a species of atheism, instead of religion
. . . .’27 While this may or may not be Hume’s
design, and while The Natural History does
not strictly entail atheism, Warburton is right
that a viable naturalistic account of the idea
of God and of belief in God calls the reasona-
bleness of the belief into serious question.28
It does so, not by highlighting a gap between
belief and evidence as, for instance, Hume’s
examination of the design argument does, but
by reducing the idea of the providential God
of monotheism to nothing but our own men-
tal projection.
Its target being the reasonableness of belief
in a providential God itself, not the reasons
that may be thought to give evidential sup-
port to the belief, Hume’s naturalistic theory
is well positioned to challenge a kind of fide-
ism that may be attractive to some believers,
namely, what Kail calls ‘rational fideism’.29
Rational fideism – roughly the idea that
belief in God is reasonable apart from argu-
ments or evidence that God exists and not-
withstanding arguments or evidence that
God does not exist – contrasts with a more
permissive kind of fideism which would be
276
9780826443595_Ch13_Final_txt_print.indd 276
satisfied that there is no contradiction in
either the idea of God or the proposition
that God exists. This more permissive fide-
ism would seem to be immune to a successful
reduction of the idea of God to nothing more
than a mental projection, as would forms of
fideism even more permissive than it.
Finally, Hume’s naturalistic account of
religion at the level of psychology could
fail, or only partly succeed, without thereby
undermining naturalism about religion, since
other naturalistic theories of religion could
be formulated independently of Hume’s. For
instance, perhaps early (and other) humans’
propensity to anthropomorphize the sources
of danger in the natural world gives a sur-
vival advantage passed on in the genes, per-
haps by enhancing the ability to recognize
danger from temporarily invisible, intelligent
agents like snakes and alligators, or perhaps
it enhances the bonds of community among
those who see intervention in nature by invis-
ible agents. If something along these lines is
true, then such versions of a naturalistic the-
ory go deeper than Hume’s. If so, then the
naturalizing project could endure even if
Hume’s particular theory were superseded.
For analogy, think of how the project to
develop a non-Cartesian theory of the mind
continues to fare well despite serious short-
comings in the ‘simple materialism’30 of early
versions. That said, Hume’s theory is bet-
ter supported than simple materialism and
seems at least as feasible as any rival.31
NOTES
1
T. Penelhum, ‘Comments and Responses’, in
J.J. MacIntosh and H.A. Meynell (eds), Faith,
Scepticism and Personal Identity (Calgary:
University of Calgary Press, 1994), pp. 235–90,
p. 253. In ‘Natural Belief and Religious Belief
in Hume’s Philosophy’, in his Themes in Hume
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
(New York: Oxford University Press, 2003),
p. 211, Penelhum describes the praise as ‘only
lip service’.
2
J.C.A. Gaskin, Hume’s Philosophy of Religion,
2nd edn (Atlantic Highlands, NJ: Humanities
Press, 1988), p. 7. When Hume praises the
design argument in The Natural History, he
means, at most, that its conclusion is attenu-
ated deism, while realizing, and surely intend-
ing, that many readers will suppose he means
much more. Henceforth, by ‘God’ I mean the
God of the monotheistic religions.
3
A.C. Baier, ‘Hume on Religion: Stopping the
Ocean with a Bull-Rush?’, in MacIntosh and
Meynell (eds), Faith, Scepticism and Personal
Identity, pp. 61–82, p. 64.
4
Compared to the design argument, the problem
of evil, the credibility of miracle reports, and
other topics much discussed in studies of Hume’s
philosophy of religion, his examination of the
origin of religion is neglected and marginalized.
Having myself subscribed to this consensus
about the philosophical status of The Natural
History relative to Hume’s other work on reli-
gion, I will cite my own Hume on Religion (New
York: Routledge, 2001) as a case in point.
5
As I use the term ‘naturalism’, Hume’s project
is naturalistic by virtue of attempting to
explain the development of religion without
reference to supernatural entities.
6
Gaskin, Hume’s Philosophy of Religion, p. 183.
Also, J.C.A. Gaskin, ‘Introduction’, in J.C.A.
Gaskin (ed.), Dialogues and Natural History
of Religion (Oxford: Oxford University Press,
1998), p. xxv.
7
K. Yandell, Hume’s ‘Inexplicable Mystery’:
His Views on Religion (Philadelphia: Temple
University Press, 1990), p. xiv.
8
P.J.E. Kail, ‘Understanding Hume’s Natural
History of Religion’, The Philosophical
Quarterly 57 (2007), pp. 190–211, p. 191.
9
Yandell, Hume’s ‘Inexplicable Mystery’,
p. 340. If we take Yandell’s ‘follows’ to mean
strict logical entailment, then he is right. But,
taken in the sense that Hume’s account of the
causes of religious belief ‘rationally destabi-
lizes’ it, that is, provides a reason to ‘suspend
[religious] belief if nothing else speaks in
favour of it’, my argument is that something
negative does follow
277
9780826443595_Ch13_Final_txt_print.indd 277
from Hume’s account. Gaskin makes a similar
point: J.C.A. Gaskin, The Quest for Eternity
(Harmondsworth: Penguin Books, 1984), p.
33. The concept of rational destabilization of
belief is introduced and developed by Kail:
P.J.E. Kail, Projection and Realism in Hume’s
Philosophy (Oxford: Oxford University Press,
2007), pp. 21–4.
10
Here I disagree with Mackie: J.L. Mackie,
The Miracle of Theism (Oxford: Clarendon
Press, 1982), p. 190.
11
The ‘foundation’ arguments discussed by
Hume are the cosmological argument, the
design argument (including its triggering
of what Gaskin usefully calls ‘the inference
problem of evil’, Hume’s Philosophy of
Religion, pp. 53–8), and miracle-testimony. If,
as stipulated, these arguments, either individu-
ally or together, fail to give strong evidential
support to religious belief, then, as Gaskin
aptly puts it, theories of the origin of religion
such as Hume’s ‘provide a well worked out
. . . answer to the question “How is it that so
many people believe in God . . . if there are no
good reasons for [it]?”’ (Gaskin, The Quest
for Eternity, p. 33).
12
Hume’s emphasis on feelings such as terror
and anxiety in the emergence of the idea of
invisible, intelligent power finds an echo and a
corroboration in a claim sometimes made by
monotheists to support their religious outlook
or to discredit non-believers. It is their claim
that there are no atheists in foxholes.
13
Kail, Projection, p. xxix, and thereafter.
14
Kail puts Hume’s point about the status of the
idea of invisible, intelligent power in the devel-
opment of the idea of deities this way: ‘the
notion of invisible, intelligent power is treated
as a necessary and sufficient condition for any
belief to be a religious belief.’ (Kail, Projection,
p. 7) In addition to Kail’s own footnoted
qualification of Hume’s point, we might
emphasize that religions such as those forms
of Buddhism which involve no belief in deities
are not covered, or intended to be covered, by
Hume’s point.
15
Kail describes this aspect of Hume’s account
as ‘explanatory projection’ (Kail, Projection,
p. xxix, and thereafter).
16
However, as Baier notes, hope of surviving
death or of being reconstituted after death
2/9/2012 9:36:41 PM
 PSYCHOLOGICAL EXPLANATIONS OF RELIGIOUS BELIEF
is not part of Hume’s account of our long-
ago ancestors’ outlook. See Baier, ‘Hume on
Religion’, p. 81n12.
17
See note 21 below.
18
By the expression ‘being persuaded that the gods
are only our minds’ progeny’ I mean accepting
both Hume’s (or some other such) account of
the original causes of religious belief and that
there is no persuasive evidence that there are
gods.
19
With no space to argue for this point, I stipulate it.
20
That argument in Part V of the Dialogues
receives indirect support from the ‘weighing
scale’ argument in Section XI of An Enquiry
concerning Human Understanding.
21
Two points are worth noting: (1) Hume’s
thinking on instinctual or natural beliefs is
both the subject of much debate and relevant
to this paper. But I have no space to examine
it. In his view, our belief in the external world
is an instinctual belief, whereas religious
belief, as he seems to make pretty clear in his
‘Introduction’ to The Natural History, is not
(NHR 134). Gaskin’s argument supporting
that distinction in Hume is persuasive (Hume’s
Philosophy of Religion, pp. 116–19). (2)
Although the objection I am now considering
is in the immediate vicinity of the debate on
religious belief’s being or not being instinctual,
it is independent of that debate.
22
This objection may also be viewed as an
instance of the previous one. For, essentially,
its point is that the propensity theory of the
origin of religion is a special case of the design
argument.
278
9780826443595_Ch13_Final_txt_print.indd 278
23
L. Falkenstein, ‘Hume’s Project in The Natural
History of Religion’, Religious Studies 39
(2003), pp. 1–21, p. 8.
24
W. James, The Varieties of Religious
Experience (New York: Modern Library,
1902), p. 71, for instance.
25
Others seeing Hume’s project as feasible are
Gaskin, The Quest for Eternity, pp. 33–6;
Kail, ‘Understanding Hume’s Natural
History’, pp. 202–3, p. 205; and M. Webb,
‘The Argument of the Natural History’, Hume
Studies 17 (1991), pp. 141–59, pp. 141, 146.
26
See note 9. Likewise, to go back to an earlier
example, you could persuade someone that his
belief in a place’s being haunted is all only in his
mind, and possibly the place is haunted anyway.
27
Quoted in E.C. Mossner, The Life of David
Hume, 2nd edn (Oxford: Clarendon Press,
1980), p. 325.
28
Opposition to the point that our idea of God
could be an idea of our own devising is taken
to an extreme length by Descartes in the third
Meditation, where he insists that he himself
could not even possibly be the source of his
idea of God.
29
Kail, ‘Understanding Hume’s Natural
History’, p. 191. See Gaskin, The Quest for
Eternity, p. 34.
30
R. Taylor, Metaphysics (Englewood Cliffs, NJ:
Prentice-Hall, 1965), p. 25.
31
I thank the provost of Seton Hall University,
the dean of its College of Arts and Sciences,
and my colleagues in the Department of
Philosophy for sabbatical leave during which I
worked on this essay.
2/9/2012 9:36:42 PM
 14
HUME’S SENTIMENTALIST ACCOUNT
OF MORAL JUDGEMENT
Julia Driver
The end of all moral speculations is to
teach us our duty; and, by proper rep-
resentations of the deformity of vice
and beauty of virtue, beget correspond-
ing habits, and engage us to avoid the
one, and embrace the other . . . What is
honorable, what is fair, what is becom-
ing, what is noble, what is generous,
takes possession of heart, and animates
us to embrace and maintain it. What
is intelligible, what is evident, what is
probable, what is true, procures only
the cool assent of the understanding[.]
(EPM 1.7 / 172)
David Hume’s account of moral judgement
falls squarely in the sentimentalist tradition.
This tradition holds, very broadly, that moral
judgement involves affect, or a character-
istic feeling or attitude such as approval or
disapproval, like or dislike. This is extremely
vague, and how the account is fleshed out in
more detail varies amongst sentimentalists.
In Hume’s work he spends a good deal of
time refining his version of sentimentalism as
it relates to moral judgement. In doing so, he
has two main contrasts in mind: the contrast
between his approach and egoism, and the
contrast between his approach and ration-
alism. Hume differs from the egoists in his
279
9780826443595_Ch14_Final_txt_print.indd 279
belief that sympathetic engagement with oth-
ers is an aspect of human nature. We are not
merely self-interested creatures. He differs
from the rationalists in that he believes moral
judgement – which certainly is capable of
motivating people to act – must involve emo-
tion. Just discovering that giving to Oxfam
helps the needy is insufficient to motivate
a person to give. The person must also care
about others, and have a desire to help allevi-
ate their suffering.1 However, Hume clearly
believes that reason plays a role in refining
our emotional reactions.
There is some controversy over David
Hume’s account of moral judgement. He
clearly is a sentimentalist, believing that
when we make moral judgements of persons
and their character we do so as the result of
having certain feelings, that is, we become
sympathetically engaged either with the
agents themselves or the others they interact
with, or both. There is some disagreement
over whether or not this commits him to a
kind of anti-realism. When we make moral
judgements, that is, are they about anything
in the world? Further, if moral judgement
involves emotion in some crucial way, can
a complete account of moral judgement be
reduced to expression of emotion? These are
2/9/2012 9:36:47 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
some of the main controversies surrounding
Hume’s account.
1. HUMAN NATURE
Although Hume disagreed with Hobbes on
the fundamental nature of human beings,
he agreed with him on at least one impor-
tant, and radical, theoretical point. Whatever
morality is, its content does not depend on a
divine will. Rather, morality depends (in some
way) on human nature. In Hume’s case, the
crucial feature of human nature that underlies
morality is sympathy. There is some unclarity
about what Hume meant to pick out by ‘sym-
pathy’. This is because Hume makes different
points about sympathy at different places in
his writing. Árdal notes that he makes a dis-
tinction between sympathy and benevolence.
Benevolence, too, is a feature of human nature
that underwrites virtues like charity towards
others. Sympathy, on the other hand, is a
capacity that human beings have to pick up
on the feelings of others and be affected by
those feelings. It is not just that we are affected
by the emotions of others, however – we tend
to be affected in concert with those emo-
tions. When one person, through sympathetic
engagement with another, detects that person’s
distress, there is a tendency for him to become
distressed himself. This account of sympathy
is owed to Hume’s discussion of it in Book II
of the Treatise. There Hume views sympathy
as a kind of ‘communication’ which allows us
to receive the emotions of others. The proc-
ess works roughly as follows: we observe
the behaviour or aspect of another person
and from those infer the person’s feelings or
opinions. Then this idea we obtain from our
observation and inference is itself converted –
via sympathy – into an impression, and ‘. . .
280
9780826443595_Ch14_Final_txt_print.indd 280
acquires such a degree of force and vivacity,
as to become the very passion itself, and pro-
duce an equal emotion, as any original affec-
tion’ (THN 2.1.11.3 / 317). This is facilitated
by resemblance. We know that other persons
resemble us in relevant respects, and so our
imagination moves between the other person
to ourselves quite easily.
However, in the Enquiry, Hume uses ‘sym-
pathy’ along with ‘benevolence’ to refer to
our capacity or tendency to care about the
suffering of others, to have a desire to alle-
viate it, for example. In this view, sympathy
does not simply allow us to pick up on the
feelings and views of others, it underwrites
our concern for others. It, like benevolence, is
a response to the egoism of Hobbes.
Hume clearly believed that, as part of our
nature, we care about other human beings.
This sympathy, in actual practice, may be
weak, and it may be weaker than self-inter-
est. But it exists in all normal human beings.
Without this we would not be motivated to
engage in moral behaviour. We would not
make moral judgements.
Hume viewed sympathy as a universal fea-
ture of human nature:
The notion of morals implies some sen-
timent common to all mankind, which
recommends the same object to general
approbation, and makes every man, or
most men, agree in the same opinion
concerning it. It also implies some senti-
ment, so universal and comprehensive as
to extend to all mankind, and render the
actions and conduct, even of the persons
the most remote, an object of applause
or censure, according as they agree or
disagree with that rule of right which is
established. (EPM 9.5 / 272)
All normal human beings will experience
sympathy. They will be happy when others
2/9/2012 9:36:47 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
are happy, and sad when others are sad. This
is true even when the persons being regarded
are far away geographically, relationally and
temporally. However, Hume goes on to note
that distance does affect the intensity of the
sympathetic response. This will mean that,
in order for us to reach the agreement he
knows we do reach on moral issues, some
corrective process takes place in which we
reflect on the sympathetic responses we feel
and modulate them on the basis of informa-
tion that we have or by weeding out various
other factors that might be distorting our
emotions.
Because Hume ties moral judgement to
emotion rather than the agent’s picking up
on objective moral properties, some writers
view Hume as a moral anti-realist. However,
whether or not we regard Hume as an anti-
realist largely depends on what we mean by
the term. There are some senses of the term
in which it is clear that he is anti-realist, but
for weaker understandings of ‘anti-realist’ it
is not so clear. This issue will be explored in
the next section.
But Hume’s sentimentalism raises other
worries for many people. If it is true that
morality depends upon features of human
nature such as our capacity for sympathy,
does that not make morality itself contingent
in a way that conflicts with moral phenom-
enology? This is a feature of the approach
that Kant would find problematic. Morality
should bind us of necessity, regardless of our
desires and emotions. Indeed, such elements
of our nature are fickle and unstable. While
Hume can account for universality of moral
judgement amongst human beings (though
not a more robust universality), the necessity
claim cannot be accommodated in his view,
or any sentimentalist view. However, he can
provide a plausible error theory. That will be
reviewed in a later section.
281
9780826443595_Ch14_Final_txt_print.indd 281
2. THE CASE AGAINST REALISM
There are passages in Hume where he holds,
clearly, that there are no moral properties
that are intrinsic properties of actions. The
most famous passage in which this point is
made occurs in the Treatise:
(AR) Take any action allow’d to be
vicious: Wilful murder, for instance.
Examine it in all lights, and see if you can
find that matter of fact, or real existence,
which you call vice. In which-ever way
you take it, you find only certain pas-
sions, motives, volitions, and thoughts.
There is no other matter of fact in the
case. The vice entirely escapes you, as
long as you consider the object. You can
never find it, till you turn your reflection
to your own breast, and find a sentiment
of disapprobation, which arises in you,
towards this action. Here is a matter of
fact; but ’tis the object of feeling, not
of reason. It lies in yourself, not in the
object. (THN 3.1.1.26 / 468–9).
As other writers have noted, this makes it
look as though Hume were holding a very
subjectivist view about morality – whenever
one makes a moral judgement, the judge-
ment is tantamount to simply stating some-
thing about oneself, or of expressing one’s
own emotions.2
However, in this passage Hume is describ-
ing how we go about making moral judge-
ments in our real, everyday lives. In his
account of what virtues are he provides a
standard for virtue that relies on this proced-
ure, but one that is employed from the ‘gen-
eral point of view’.
Whether or not one considers David Hume
to be a moral anti-realist largely depends
on what one means by ‘anti-realist’. It is
clear that he does not believe that there are
2/9/2012 9:36:47 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
stance-independent moral facts or properties
in the world. In his defence of moral realism,
Russ Shafer-Landau defines it as ‘the theory
that moral judgements enjoy a special sort of
objectivity: such judgements, when true, are
so independently of what any human being,
anywhere, in any circumstance whatever,
thinks of them’.3 Therefore, if a moral claim
is true, its truth is not something owed to me
as an individual. But forms of this view are
compatible with (AR). It is entirely possible
to hold a view that moral truths are objective
in that their truth conditions are not supplied
by the individual’s beliefs or attitudes, and yet
also deny that the conditions are supplied by
intrinsic properties of actions. This is the line
I should like to pursue. It renders Hume anti-
realist in a very robust sense, and yet accom-
modates a moderate objectivity or realism in
that there are moral truths that are independ-
ent of individual reactions to actions and
character traits. There are degrees, or levels,
of stance-independence.
Since many philosophers characterize
moral realism as the view that there are moral
facts that exist independently of us, Hume is a
moral anti-realist. However, it does not follow
from this that he believes there are no moral
facts. If there are moral facts, however, the
Humean believes that they are the product of
an affective and reflective process. For exam-
ple, the claim that John is generous is a posi-
tive moral evaluation of John and, putatively,
states a fact about John, that he possesses
a trait that is approved of from the general
point of view. One could hold that the claim:
(1) John possesses the virtue of generosity
is true, and true in virtue of correspond-
ence to facts about John having to do with
approval from the general point of view.
Further, a claim such as
282
9780826443595_Ch14_Final_txt_print.indd 282
(2) Robert possesses the vice of self-
deprecation,
if true, would also be true due to facts about
Robert, and the disapproval of Robert’s trait
from the general point of view. But what is
the general point of view that plays such
a crucial role in refining our sympathetic
reactions?
3. THE GENERAL POINT OF VIEW
There is a variety of passages in which Hume
discusses the general point of view, or the
common point of view.
Our situation, with regard both to per-
sons and things, is in constant fluctua-
tion; and a man that lies at a distance
from us may, in a little time, become a
familiar acquaintance. Besides, every
particular man has a peculiar position
with regard to others; and ’tis impossi-
ble we could ever converse together on
any reasonable terms, were each of us to
consider characters and persons, only as
they appear from his peculiar point of
view. In order, therefore, to prevent those
continual contradictions, and arrive at
a more stable judgement of things, we
fix on some steady and general points
of view; and always, in our thoughts,
place ourselves in them, whatever may
be our present situation. (THN 3.3.1.15
/ 581–2)
Because sympathy is at the root of moral
judgement in his view, we need a method for
correcting that sympathetic response, oth-
erwise we end up with contradictions. The
general point of view provides that method
of correction:
2/9/2012 9:36:47 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
Sympathy, we shall allow, is much fainter
than our concern for ourselves, and sym-
pathy with persons remote from us much
fainter than that with persons near and
contiguous; but for this very reason it
is necessary for us, in our calm judge-
ments and discourse concerning the
characters of men, to neglect all these
differences, and render our sentiments
more public and social. Besides, that we
ourselves often change our situation in
this particular, we every day meet with
persons who are in a situation different
from us, and who could never converse
with us were we to remain constantly in
that position and point of view, which is
peculiar to ourselves. The intercourse of
sentiments, therefore, in society and con-
versation, makes us form some general
unalterable standard, by which we may
approve or disapprove of characters and
manners. (EPM 5.42 / 229)
When making a moral judgement, an agent
will perceive behaviour, infer the presence
of a certain character trait, and then decide
whether it is a virtue or a vice (or neither) by
judging it from the steady, general, point of
view. Because our approval, based on our sym-
pathy, is so variable, we need this fixed point
in order to reach agreement, at the very least.
But this is not an arbitrary point that is sim-
ply selected for mere practical convenience.
The general point of view requires impartial-
ity, lack of bias and prejudice and other fac-
tors that distort sympathetic responses.
From the above passages we can see that
the sorts of factor Hume held to be distort-
ing were those such as the proximity to us of
the person being judged. Proximity might be
geographical or personal. Another distorting
factor might be how much the person being
judged resembles me, or someone I love. My
self-interest can distort, as can beliefs I have
283
9780826443595_Ch14_Final_txt_print.indd 283
that are idiosyncratic. One also needs to con-
sider luck factors – did the person succeed or
fail through sheer bad luck – and discount for
those. To make a proper moral judgement we
need to extrapolate away from these factors.
There is a good deal of controversy in the
literature about the general point of view
itself. Clearly Hume intends this point of
view to weed out various distortions. But the
different passages in which he discusses the
employment of this process provide differ-
ent ways of understanding it. For example,
should the agent be concerned to weed out
any distortion, or just his or her own distor-
tions? In the Treatise he writes the following:
In general, all sentiments of blame or praise
are variable, according to our situation of
nearness or remoteness, with regard to
the person blam’d or prais’d, and accord-
ing to the present disposition of our mind.
But these variations we regard not in our
general decisions, but still apply the terms
expressive of our liking or dislike, in the
same manner, as if we remain’d in one
point of view . . . . ’Tis therefore from
the influence of characters and qualities,
upon those who have an intercourse with
any person, that we blame or praise him
. . . . (THN 3.3.1.16–17 / 582)
This passage implies that, though we cer-
tainly correct for distortions of our own
perspective we do so by considering not the
influence of the character trait in general, but
its influence on those around the agent, the
character, that we are evaluating. This leaves
open the possibility that one simply trades
one distorting set of influences for another.
Other passages provide a rather different
view of the process.
When a man denominates another his
enemy, his rival, his antagonist, his
2/9/2012 9:36:47 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
adversary, he is understood to speak
the language of self-love, and to express
sentiments, peculiar to himself, and aris-
ing from his particular circumstances
and situation. But when he bestows on
any man the epithets of vicious or odi-
ous or depraved, he then speaks another
language, and expresses sentiments, in
which, he expects, all his audience are
to concur with him. He must here, there-
fore, depart from his private and particu-
lar situation, and must choose a point of
view, common to him with others: He
must move some universal principle of
the human frame, and touch a string, to
which all mankind have an accord and
symphony. (EPM 9.6 / 272)
In this passage in the Enquiry the process
involves taking the point of view that is uni-
versally accessible. This could be taking the
point of view of those affected by the charac-
ter in question. It could also be taking a com-
pletely disinterested point of view. But Geoff
Sayre-McCord notes a reason why we should
look at this point of view as non-idealized.
When we read passages such as the above it
seems fairly clear that Hume thought of the
general point of view as accessible to ordi-
nary moral reasoners. And yet, an idealized
point of view is not accessible.
I have argued elsewhere that we can
understand the process better by consider-
ing what Hume has to say about making
mistakes in our virtue judgements.4 He fam-
ously maintains that our passions themselves
can neither be true nor false (more on this
below). However, those passions are influ-
enced and directed by reason in that they
are discounted if based on false beliefs, or if
‘in exerting any passion in action, we choose
means insufficient for the design’d end, and
deceive ourselves in our judgement of causes
and effects’ (THN 2.3.3.6 / 416). Thus the
284
9780826443595_Ch14_Final_txt_print.indd 284
truth of a virtue judgement will depend upon
whether or not the trait in question actually
leads to good results (utility), which can be
understood in terms of social utility or local
utility, or in terms of immediate agreeable-
ness which generates pleasure in the specta-
tor independent of considerations of social
and local utility. It is clear that Hume does
think that a person’s reaction to a character
trait can be mistaken. In the Enquiry he dis-
cusses the notorious ‘monkish’ virtues which
are not, in fact, virtues at all:
as every quality, which is useful or agree-
able to ourselves or others, is, in com-
mon life, allowed to be a part of personal
merit; so no other will ever be received,
where men judge of things by their natu-
ral, unprejudiced reason, without the
delusive glosses of superstition and false
religion. Celibacy, fasting, penance, mor-
tification, self-denial, humility, silence,
solitude, and the whole train of monkish
virtues; for what reason are they every
where rejected by men of sense, but
because they serve to no manner of pur-
pose; neither advance a man’s fortune in
the world, nor render him a more valu-
able member of society, neither qualify
him for the entertainment of company,
nor encrease his power of self-enjoy-
ment? (EPM 9.3 / 270)
Those who view self-denial as a virtue are
mistaken. It is not a virtue. It is not a virtue
because it fails to have features that we – at
least from an impartial standpoint – fail to
find pleasing.5
Hume’s numerous claims against the role
of reason in making moral judgements, and
his emphasis on emotion in moral judge-
ment, have led to other interpretations of his
work. One of the more common is that he
is really a non-cognitivist. That is, he does
2/9/2012 9:36:48 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
not think that moral claims are the sorts of
things that even have truth value. In this kind
of view, Hume could be saying that when
one makes a judgement of the sort that ‘John
is a generous man’ one is engaged in posi-
tive evaluation of John, since generosity is
a moral virtue, and that this does not mean
that one is stating a fact. One’s judgement is
normatively loaded, so to speak, and as such
one is expressing one’s feelings of approval
for John’s behaviour when one attributes a
virtue to him. Thus, even though the surface
grammar of the utterance makes it look like
an assertion, it is more like an exclamation,
such as ‘Yay!’. It has no truth value.
4. THE CASE FOR NON-COGNITIVISM
Francis Snare is the most recent author
known for having endorsed this line in
Hume’s account of moral judgement.6 Hume
famously maintained that moral ‘distinc-
tions’ are not ‘derived from reason’. Snare
focuses on what has been dubbed Hume’s
influence argument. Hume holds that reason
has no influence over the passions, and yet
our moral judgements do. It follows, then,
that moral distinctions do not derive from
reason.
As Snare argues, the first premise in the
argument is a form of Hume’s view that
things like actions, passions and willings are
neither true nor false. In the Treatise Hume
writes:
A passion is an original existence, or, if
you will, modification of existence, and
contains not any representative qual-
ity, which renders it a copy of any other
existence or modification. When I am
angry, I am actually possest with the
285
9780826443595_Ch14_Final_txt_print.indd 285
passion, and in that emotion have no
more reference to any other object, than
when I am thirsty, or sick, or more than
five foot high. ’Tis impossible, therefore,
that this passion can be oppos’d by, or be
contradictory to, truth and reason . . . .
(THN 2.3.3.5 / 415)
However, it is possible to interpret Hume’s
claim as restricted to motivation, in which
case there is scope for truth or falsity with
moral claims.
Some have argued that, in Hume’s view,
the judgement that ‘X is virtuous’ simply
means that someone – that is, the utterer of
the claim – approves of X. Thus such judge-
ments tell us something about the speaker,
that he or she approves of X. There are very
obvious problems with this sort of subject-
ivism about moral judgement, to the extent
that Páll S. Árdal believed that it was clear
that Hume did not hold such a view.7
Reason may be the slave of the passions
in that what our judgement tells us will not
move us without the passions, but this does
not mean that reason plays no significant
role in both making moral judgements and
offering a standard for evaluating the moral
judgements themselves. It is not that moral
judgements lack truth value. It is, rather,
that their truth or falsity itself, lacking moti-
vational force, is irrelevant practically, it is
insufficient to take ‘possession of the heart’
(EPM 1.7 / 172).
This possibility can be expanded so that
another interpretation of Hume is possible
when we focus on his discussions of how
virtue traits elicit approval. That interpret-
ation is that when we view a character as
good – consider generosity, again – we are
attributing to the person a kind of property
that is dispositional in nature – that is, it is
the sort of property that, via the right kind of
2/9/2012 9:36:48 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
sympathetic engagement, produces approval
in the spectator. Thus the claim that a trait
such as generosity is good has truth-value via
correspondence with its disposition (or lack
thereof) to produce the approval. This then
raises the very deep issue of what it is that
persons approve of ‘from the general point
of view’. My own view is that Hume focuses
on various forms of utility in answering that
question.8 As Sayre-McCord notes, however,
this is not at all to say that Hume is a utilitar-
ian.9 Hume makes no appeal to maximizing
utility. Of course, his account of evaluation
focuses on character evaluation rather than
act evaluation.
This does not make Hume a realist in the
robust sense. His standard is not stance-
independent. His account of human nature,
rather, offers a substantive constraint on our
judgement. But it is a standard that does not
leave it up to the individual and his own set
of desires. In this way, the standard is stance-
independent, if what is meant is ‘independent
from an individual’s perspective’.
Many other writers reject the non-cog-
nitivist line. Don Garrett, for example, also
holds the view that for Hume at least some
moral evaluations were beliefs that could
be either true or false.10 Rachel Cohon also
believes the non-cognitive interpretation to
be false. In her view, when we sense virtue
and vice, or good and evil, via our moral
sentiments, we form ideas based on the cor-
responding impressions. ‘We then form ideas
that are copies of these impressions and
that can be combined into various sorts of
(believed) judgements, some of which are
true because they are accurate copies of their
originals.’11 She further notes that ‘the moral
properties themselves (not as we conceive of
them but as they really are, if it is possible
to talk about that) may well simply be ten-
dencies to cause approval and disapproval in
286
9780826443595_Ch14_Final_txt_print.indd 286
human beings who contemplate an object in
a sufficiently disinterested way.’12 But Cohon
also resists using utility as a standard for
virtue. She argues that this runs up against
the anti-realism Hume is famous for embrac-
ing. I disagree. There is a variety of ways in
which we understand anti-realism, and as I
have argued, one can hold that utility is a
standard for virtue and yet also hold that it
is such a standard because from an impartial
point of view it is what is approved of, what
is found pleasing, and so forth. This renders
the standard stance-independent in the very
important senses that the standard does not
depend on the individual’s views, or actual
group views, for example. Individuals can be
wrong about virtue, as can groups. In this
theory, even the whole of humanity can be
mistaken. What makes utility a standard is
our (impartial) response. But there is more
to the story. Again, Sayre-McCord notes the
many ways in which Hume appeals to utility
in discussing virtue. He also correctly notes
that this does not, by itself, commit Hume
to being a utilitarian. But what exactly do
we mean by ‘utility’? I believe the idea is
appealed to in a variety of guises in Hume.
There is local utility, social utility, and simple
raw pleasure. At the bottom of it all is pleas-
ure – local and social utility are valued by us
because of their tendency to generate pleas-
urable responses in the observer which in
turn is due to the fact that – via sympathetic
engagement – we desire the happiness of oth-
ers. All of these sorts of things come under
the rubric of ‘utility’, although when Hume
uses the word it is usually in discussing social
utility. Annette Baier approves of the fact that
Hume distinguishes agreeableness from util-
ity, but this does not mean that what Hume
considers ‘agreeableness’ does not fall into
the more modern category of utility.13 That
something is immediately agreeable to us is a
2/9/2012 9:36:48 PM
 HUME’S SENTIMENTALIST ACCOUNT OF MORAL JUDGEMENT
way for it to generate pleasure, and pleasure
is at least a kind of utility.
6. CONCLUSION
Hume was neither a non-cognitivist nor a
robust moral realist. When human beings
make moral judgements, they are making
true or false claims, but not about intrinsic
properties of actions. Rather, the truth con-
ditions for those claims are set by a standard
of what would be approved or disapproved
of from the general point of view, where
the general point of view is understood
as idealized in ways that remove bias and
distortion.
NOTES
1
This motivational thesis – that reason alone
cannot motivate – is often seen as a core
Humean tenet. But Hume’s views on this issue
are highly complex, and Rachel Cohon has
done a nice job of arguing that it is not clear
that Hume held this view, at least if what is
meant by it is that beliefs by themselves can-
not motivate. See her discussion in chap. 3
of Hume’s Morality: Feeling and Fabrication
(New York: Oxford University Press, 2008).
287
9780826443595_Ch14_Final_txt_print.indd 287
2
See A.E. Pitson, ‘Projectionism, Realism, and
Hume’s Moral Sense Theory’, Hume Studies
15 (1989), pp. 69–92.
3
R. Shafer-Landau, Moral Realism: A Defence
(Oxford: Oxford University Press, 2005), p. 2.
4
J. Driver, ‘Pleasure as the Standard of Virtue
in Hume’s Moral Philosophy’, Philosophical
Quarterly 85 (2004), pp. 173–94.
5
I also argue in ‘Pleasure as the Standard of
Virtue in Hume’s Moral Philosophy’ that
these sorts of passages show that Hume was
offering a metaphysics of virtue as well as
an epistemology of virtue, or as well as an
account of how we go about making good
virtue judgements.
6
F. Snare, Morals, Motivation and Convention
(New York: Cambridge University Press, 1991).
7
See his Passion and Value in Hume’s Treatise
(Edinburgh: University of Edinburgh Press,
1989), pp. 198ff.
8
Driver, ‘Pleasure as the Standard of Virtue in
Hume’s Moral Philosophy’.
9
G. Sayre-McCord, ‘Hume and the Bauhaus
Theory of Ethics’, Midwest Studies in
Philosophy 20 (1995), pp. 280–98.
10
D. Garrett, Cognition and Commitment in
Hume’s Philosophy (New York: Oxford
University Press, 1997).
11
Cohon, Hume’s Morality, p. 124.
12
Ibid.
13
In A Progress of Sentiments (Cambridge, MA:
Harvard University Press, 1991), pp. 204–5 she
holds that Hume, in contrast to Bentham, does
not reduce virtue evaluation to ‘cash utility’
and the ‘accountant’s style’ of thinking about
the good.
2/9/2012 9:36:48 PM
 15
HUME AND THE VIRTUES
Dan O’Brien
1. VIRTUE
Human beings possess traits that are benefi-
cial to themselves and to others, traits that
have a ‘tendency to the good of mankind’
(THN 3.3.1.10 / 578). Hume calls such traits
virtues, and examples appear throughout his
writings: ‘activity, vigilance, application, . . .
resolution’ (THN 3.3.4.7 / 610), ‘cleanliness’
(THN 3.3.4.10 / 611), ‘[m]eekness, benefi-
cence, charity, generosity, clemency, modera-
tion, [and] enquiry’ (THN 3.3.1.11 / 578).
Besides discretion, caution, enterprise,
industry, assiduity, frugality, œconomy,
good-sense, prudence, discernment; besides
these endowments, I say, whose very names
force an avowal of their merit, there are
many others, to which the most deter-
mined scepticism cannot, for a moment,
refuse the tribute of praise and appro-
bation. Temperance, sobriety, patience,
constancy, perseverance, forethought, con-
siderateness, secrecy, order, insinuation,
address, presence of mind, quickness of
conception, facility of expression; these,
and a thousand more of the same kind, no
man will ever deny to be excellences and
perfections. (EPM 6.21 / 242–3)
He offers a four-way classification of
such traits. There are those that are useful
288
9780826443595_Ch15_Final_txt_print.indd 288
to others, such as generosity, humanity,
bravery and beneficence, and those that
are useful to oneself, such as good sense,
‘prudence, temperance, frugality, industry,
assiduity, enterprise, [and] dexterity’ (THN
3.3.1.24 / 587). There are also those that
are agreeable to others – scalability, wit,
ingenuity, cheerfulness – and those that
are agreeable to oneself, such as courage,
good humour and philosophical tranquil-
ity. Certain virtues fit into more than one
category. Benevolence, for example, is both
useful to others and agreeable to oneself;
certain traits are both useful and agreeable
to ourselves: ‘courage, intrepidity, ambi-
tion, love of glory’ (THN 3.3.2.13 / 599),
‘[c]onstancy, fortitude, and magnanimity’
(THN 3.3.4.3 / 608); and various traits are
both useful and agreeable to others: ‘gen-
erosity, humanity, compassion, friendship,
fidelity, zeal, disinterestedness, [and] liber-
ality’ (THN 3.3.3.3 / 603).1
Hume imagines a group of friends discuss-
ing the qualities of an ideal son-in-law.
You are very happy, we shall suppose
one to say . . . that you have given your
daughter to Cleanthes. He is a man of
honour and humanity. Every one, who
has any intercourse with him, is sure of
fair and kind treatment. (EPM 9.2 / 269)
2/9/2012 9:36:54 PM

I congratulate you too, says another, on
the promising expectations of this son-
in-law; whose assiduous application to
the study of laws, whose quick penetra-
tion and early knowledge both of men
and business, prognosticate the greatest
honours and advancement. (ibid.)
Another adds further praise:
I met him lately in a circle of the gay-
est company, and he was the very life
and soul of our conversation: So much
with good manners; so much gallantry
without affectation; so much ingenious
knowledge so genteely delivered, I have
never before observed in any one. (ibid.)
And another:
You would admire him still more . . .
if you knew him more familiarly. That
cheerfulness, which you might remark
in him, is not a sudden flash struck out
by company: It runs through the whole
tenor of his life, and preserves a per-
petual serenity on his countenance, and
tranquillity in his soul. (ibid.)
Thus, ‘[a] philosopher might select this char-
acter as a model of perfect virtue’ (EPM 2.9
/ 270).
Hume also notes that history has already
supplied us with such a model – Alfred, a
king who ‘deservedly attained the appella-
tion of Alfred the Great’ (H 1.74).
The merit of this prince, both in private
and public life, may with advantage be
set in opposition to that of any mon-
arch or citizen, which the annals of any
age or any nation can present to us. He
seems indeed to be the model of that
289
9780826443595_Ch15_Final_txt_print.indd 289
HUME AND THE VIRTUES
perfect character . . . So happily were all
his virtues tempered together; so justly
were they blended; and so powerfully
did each prevent the other from exceed-
ing its proper boundaries! He knew how
to reconcile the most enterprising spirit
with the coolest moderation; the most
obstinate perseverance with the easiest
flexibility; the most severe justice with
the gentlest lenity; the greatest vigour in
commanding with the most shining tal-
ents for action. (H 1.74–5)
History, however, also showcases vice. Henry
VIII’s ‘catalogue of . . . vices would comprehend
many of the worst qualities incident to human
nature: Violence, cruelty, profusion, rapacity,
injustice, obstinacy, arrogance, bigotry, presump-
tion, [and] caprice’ (H 3.322). Acknowledged
elsewhere as vicious are ‘[i]ndolence, negligence,
want of order and method, . . . fickleness, rash-
ness, credulity’ (EPM 6.1 / 233), ‘prodigality,
luxury, irresolution, [and] uncertainty’ (THN
3.3.4.7 / 611). Virtues can also have associated
vices when overplayed: King James had ‘[m]any
virtues . . . ; but scarce any of them pure, or free
from the contagion of neighbouring vices. His
generosity bordered on profusion, his learning
on pedantry, his pacific disposition on pusilla-
nimity, his wisdom on cunning, his friendship
on light fancy and boyish fondness’ (H 5.121).
Thus, ‘[a] due medium . . . is the characteristic of
virtue’ (EPM 6.2 / 233), and this is determined by
utility and how agreeable the varying forms of
our character traits are to oneself and to others –
a certain amount of cheerfulness, for example, is
good for everyone; too much can cloy.
In the eighteenth century there was a con-
troversial debate concerning whether virtues
were ‘original instincts of the human mind’
(THN 3.3.6.3 / 619) – that is, innate – or
whether they merely described forms of behav-
iour that ultimately were pursued for reasons
of self-interest. Francis Hutcheson supported
2/9/2012 9:36:54 PM
 HUME AND THE VIRTUES
the former view, seeing benevolence as innate
and as the fundamental virtue from which
a list of virtues similar to that of Hume can
be derived. In contrast, Bernard Mandeville
in The Fable of the Bees (1714) sees human
beings as driven by purely selfish ends. There
is no natural benevolence. We may pursue
seemingly benevolent projects, but these are in
the end to satisfy ourselves and are calculated
to do so – either by ourselves or, more usually,
through ‘artifice and education, when politi-
cians endeavour . . . to restrain the turbulent
passions of men, and make them operate to
the public good, by the notions of honour and
shame’ (THN 3.3.1.11 / 578).
Hume’s position combines aspects of both
approaches. Hume sides with Mandeville over
the question of whether virtue is innate or
acquired. There may be ‘some spark of friend-
ship for human kind, some particle of the
dove, kneaded into our frame, along with the
elements of the wolf and the serpent’ (EPM
9.4 / 271), but it seems unlikely, given our
experience of human nature, that mankind
is innately disposed to manifest the extensive
range of virtues that Hume cites. And, as we
shall see below, there is a category of virtues –
the artificial virtues – that Hume explicitly
denies are natural to us. There is, however, an
aspect of human nature that is, in a certain
sense, natural, and it is this that enables us to
see certain traits as virtuous, those that are
useful and agreeable to ourselves and, in con-
trast to Mandeville’s self-centred perspective,
those that are useful and agreeable to others.
This aspect of human nature is sympathy.
2. VIRTUE AND SYMPATHY
Hume’s account of the virtues is grounded in
his naturalistic account of the passions: ‘To
290
9780826443595_Ch15_Final_txt_print.indd 290
approve of a character is to feel an original
delight upon its appearance. To disapprove
of it is to be sensible of an uneasiness. The
pain and pleasure, therefore, being the pri-
mary causes of vice and virtue’ (THN 2.1.7.5
/ 296).2 Learning about or observing cer-
tain acts leads us to feel approval, whereas
other acts lead us to feel disapproval and our
reactions to many kinds of cases are almost
universal: acts that result in needless and
avoidable suffering, for example, are disap-
proved of by all. Further, when considering
actions in terms of virtue, we do not think of
them in isolation from the agent performing
them; we think of them as being caused by
him or her and by aspects of his or her char-
acter. Those aspects of an agent’s character
towards which we feel approval are virtues
and those towards which we feel disapproval
are vices.
Empathy or what Hume calls ‘sympathy’
plays a crucial role here. Our responses to
the actions of others can be biased. I may
approve of Eric’s arrogance – feel ‘delight’
at his actions – because I am dazzled by his
charm. Judgements, however, concerning
his character should be independent of my
own weaknesses, interests, mood or circum-
stances. They should not be made solely from
my own point of view, but from ‘some com-
mon point of view’ (THN 3.3.1.30 / 591).
In judging virtue a thinker should ‘depart
from his private and particular situation, and
must choose a point of view, common to him
with others’ (EPM 9.6 / 272). We must con-
template an agent’s character, not just from
our own perspective, but from the perspec-
tives of others relevant to the action being
appraised. Hume does not demand that we
take up an ideal point of view, that of a ‘spec-
tator who is fully informed and unsullied
by prejudice’, a view from nowhere, or an
‘angelic equi-sympathetic engagement with
2/9/2012 9:36:54 PM

all of humanity’.3 Rather, we should ‘confine
our view to that narrow circle, in which any
person moves, in order to form a judgement
of his moral character’ (THN 3.3.3.2 / 602).
I consider those who have a ‘particular con-
nexion’ (ibid.) with Eric, those who have an
‘immediate connexion or intercourse’ (THN
3.3.3.2 / 603) with him.
When we enumerate the good qualities
of any person, we always mention those
parts of his character, which render him
a safe companion, an easy friend, a gen-
tle master, an agreeable husband, or an
indulgent father. We consider him with
all his relations in society; and love or
hate him, according as he affects those,
who have any immediate intercourse
with him (THN 3.3.3.9 / 606).
I may feel delight at Eric’s arrogance, but
others disapprove and in sympathizing with
their disapproval I can come to judge Eric’s
arrogance as a character flaw. Via empathy I
‘receive by communication’ (THN 2.1.11.2
/ 316) and ‘enter into’ and ‘embrace’ (THN
2.1.11.5 / 318) the sentiments of others.4 In
judging virtue one ‘must move some univer-
sal principle of the human frame, and touch
a string to which all mankind have an accord
and symphony’ (EPM 9.6 / 272). ‘The minds
of men are mirrors to one another.’ (THN
2.2.6.21 / 365) It is thus via sympathy that we
are alive to the effects of virtue and vice. Simon
Blackburn summarizes Hume’s position:
A social virtue, such as benevolence,
is then a trait which would have done
them good, and ourselves feeling that,
we admire the subject for it. The mecha-
nism here is one of sympathy or empa-
thy, imitated, however, by an exercise of
imagination: feeling for ourselves what
291
9780826443595_Ch15_Final_txt_print.indd 291
HUME AND THE VIRTUES
it would be like to be thrown in with
this character, we recognize the pleasure
it could have brought, if it is virtuous,
or the trouble and pain it would have
brought, if it is vicious. And ‘internal-
izing’ those outcomes, we admire or
despise accordingly.5
Hume’s focus is on virtuous character traits
and not on virtuous acts. A person’s charac-
ter is constituted by their prevailing pas-
sions and dispositions. Sometimes, however,
through no fault of their own, a person’s
character may not be able to be expressed. In
such cases that person is not judged accord-
ing to the actual consequences of his actions,
but by the general tendencies of his char-
acter. ‘Virtue in rags is still virtue; and the
love, which it procures, attends a man into
a dungeon or desart, where the virtue can no
longer be exerted in action, and is lost to the
world’ (THN 3.3.1.18 / 584).
Where a character is, in every respect, fit-
ted to be beneficial to society, the imagi-
nation passes easily from the cause to
the effect, without considering that there
are still some circumstances wanting to
render the cause a compleat one (THN
3.3.1.20 / 585).6
Hume’s account of virtue is grounded in sen-
timents and feelings, and to judge virtue we
must follow an imaginative ‘progress of the
sentiments’ (THN 3.2.2.25 / 500) in order to
attain the common point of view from where
the usefulness and agreeability of the effects
of character traits can be assessed. Sometimes,
however, ‘[t]he heart does not always take
part with those general notions, or regulate
its love and hatred by them’ (THN 3.3.3.2
/ 603) – I may, for example, not be able to
see past Eric’s charm. In such cases, however,
we follow rules. Sympathy ‘soon teaches us
2/9/2012 9:36:54 PM
 HUME AND THE VIRTUES
this method of correcting our sentiments, or
at least, of correcting our language, where
the sentiments are more stubborn and inal-
terable’ (THN 3.3.1.16 / 582). Learnt gen-
eralizations concerning virtuous and vicious
behaviour enable us to express what we
would feel from the common point of view
if, on occasion, we do not have such feelings
or are unable to adopt this perspective.7
3. THE MORAL VIRTUES
We praise a variety of human qualities –
qualities with a clear moral dimension and
those without – and on Hume’s account such
praise has its source in our sympathy with
others. ‘It must . . . be allowed that every qual-
ity of the mind, which is useful or agreeable
to the person himself or to others, communi-
cates a pleasure to the spectator, engages his
esteem, and is admitted under the honour-
able denomination of virtue or merit.’ (EPM
9.12 / 277) The witty raconteur is therefore
virtuous, as is the courageous knight, the
compassionate careworker and the patient
teacher. Hume also notes the virtuous nature
of certain ‘intellectual endowments’ (EPM
App. 4.2 / 313) such as ‘prudence, penetra-
tion, discernment, [and] discretion’ (ibid.),
and he is alive to the agreeableness and util-
ity of bodily attributes such as beauty, ‘broad
shoulders’ and a ‘lank belly’ (THN 3.3.5.3
/ 615).8 Where, then, is Hume’s account of
morality? What, if anything, is special about
the moral virtues?
In the four-way classification that was dis-
cussed in section 1, Hume sees the traits in
the first two categories as ‘form[ing] the most
considerable part of morality’ (THN 3.3.6.2
/ 619). These are the traits that are useful to
others and oneself. This criterion may rightly
292
9780826443595_Ch15_Final_txt_print.indd 292
rule out wit and cheerfulness, for example,
from moral consideration, but there are also,
however, characteristics in these first two
categories that are not traditionally seen as
constituting morality, such as frugality and
dexterity.
Hume also canvasses the suggestion
that the moral virtues can be distinguished
in virtue of their voluntary nature: we can
choose to act benevolently in a certain situ-
ation, whereas we cannot decide whether or
not to have broad shoulders. Hume notes,
however, that many traditional moral quali-
ties, those that ‘all moralists . . . comprehend
under the title of moral virtues’ (EPM 3.3.4.3
/ 608) – ‘courage, equanimity, patience, self-
command’ (EPM App. 4.2 / 313), ‘constancy,
fortitude, magnamity; and, in short, all the
qualities which form the great man’ (THN
3.3.4.3 / 608) – are involuntary; they are not
under our control. Further, the voluntary
exercise of certain actions would not seem to
delineate the moral from the non-moral. The
witty joke and the elegant pass in football
can be voluntarily performed to bring pleas-
ure yet one would not like to claim that such
actions are ‘moral’.
The problem here – that is, if one sees it as
such – is that Hume provides a naturalistic
account of virtue, one grounded in sympathy
with utility and pleasure, and people just do
find utility and pleasure in a set of character
traits wider than that circumscribed by tradi-
tional morality. We have vocabulary to mark
various distinctions between such traits –
talents, abilities, the virtues (the latter often
reserved for the moral sphere) – and we may
feel deeper levels of sentiment towards certain
kinds of trait. But Hume does not find the
‘boundaries are exactly fixed between virtues
and talents, vices and defects’ (EPM App. 4.2
/ 313), and any investigation into this distinc-
tion is ‘only a grammatical enquiry’ (ibid.),
2/9/2012 9:36:55 PM

not, that is, an investigation into the distinc-
tive nature of virtue.9, 10 ‘Whether natural
Abilitys be Virtues is a Dispute of Words.’
(LDH 1.33, 13)
If . . . the sentiments are similar, which
arise from these endowments and from
the social virtues; is there any reason
for being extremely scrupulous about
a word, or disputing whether they be
entitled to the denomination of virtues?
(EPM App. 4.6 / 316)
The title of Book III of the Treatise may be
‘Of Morals’ but it discusses much else – all
agreeable and useful human qualities: from
the businessman who has ‘dexterity in busi-
ness’ (THN 3.3.1.25 / 588), to the military
hero ‘so dazzling in his character’ (THN
3.3.2.15 / 601), the attentive friend (THN
3.3.3.5 / 604–5) and to
good women’s men, who have either
signalized themselves by their amorous
exploits, or whose make of body prom-
ises an extraordinary vigour of that
kind, [such] are well received by the fair
sex, and naturally engage the affections
even of those, whose virtue prevents any
design of ever giving employment to
those talents. (THN 3.3.5.2 / 614–5)
Human life is better for the existence of such
people – better for them themselves, and bet-
ter for others.
Hume’s account of the virtues and of
morality was influenced by Hutcheson’s
theory of the moral sentiments. In the 1748
and 1750 editions of the first Enquiry Hume
claims that:
[Hutcheson] has taught us, by the most
convincing Arguments, that Morality is
293
9780826443595_Ch15_Final_txt_print.indd 293
HUME AND THE VIRTUES
nothing in the abstract Nature of Things,
but is entirely relative to the Sentiment
or mental Taste of each particular being;
in the same Manner as the Distinctions
of sweet and bitter, hot and cold, arise
from the particular Feeling of each sense
or Organ. Moral Perceptions there-
fore, ought not to be class’d with the
Operations of the Understanding, but
with the Tastes or Sentiments. (EHU,
p. xxi)
Moral distinctions are not based on, nor
discoverable through, reason or argument;11
they are, rather, grounded in human nature
and in our natural, sympathetic responses to
the actions of our fellows.12
Reason, however, is not completely inert
with respect to morality. Probabilistic or
causal reasoning is involved in gauging the
likely effects of character traits on others:
‘reason must enter for a considerable share’
in all moral judgements, reason ‘to instruct
us in the tendency of qualities and actions,
and point out their beneficial consequences
to society and to their possessor’ (EPM App.
1.2 / 285). Moral judgements are also often
corrected by ‘argument and reflection’: just
as with ‘many orders of beauty . . . it is req-
uisite to employ much reasoning, in order to
feel the proper sentiment’, so too with ‘moral
beauty . . . [this] demands the assistance of
our intellectual faculties, in order to give
it a suitable influence on the human mind’
(EPM 1.9 / 173). As we saw earlier, rules are
required for times when we do not feel the
‘correct’ moral sentiments, when, for exam-
ple, Eric’s charm blinds me to his arrogance.
[A]s the advantage to society results only
from observances of the general rule, and
from the concurrence and combination of
several persons in the same equitable con-
duct; the case here becomes more intricate
2/9/2012 9:36:55 PM
 HUME AND THE VIRTUES
and involved . . . And a very accurate rea-
son or judgement is often requisite, to
give the true determination, amidst such
intricate doubts arising from obscure or
opposite utilities. (EPM App. 1.2 / 286)
But in order to pave the way for such
a sentiment, and give a proper discern-
ment of its objects, it is often necessary,
we find, that much reasoning should pre-
cede, that nice distinctions be drawn, dis-
tant comparisons formed, complicated
relations examined, and general facts
fixed and ascertained. (EPM 1.9 / 173)
Reasoning is thus involved in various ways
in moral judgement, but only in order to aid
the necessary moral sentiments to be felt or
to remind us what should be felt if disturb-
ing influences were not present and we could
adopt the common point of view. Reasoning
alone cannot provide us with knowledge of
morality.13
4. THE ARTIFICIAL VIRTUES
In the Treatise Hume draws a distinction
between natural virtues and artificial vir-
tues (although, as we shall see below, this
distinction is thinner than it first appears).
The natural virtues are manifest in forms of
behaviour that come naturally to us, irre-
spective of the influence of society, whereas
the artificial virtues are those aspects of
character that are encouraged in society by
the ‘artifice and contrivance of men’ (THN
3.3.1.1 / 574) because we have discovered
that they lead to utility and pleasure, even
though they may not come naturally to us
as individuals. ‘It is through experience and
observation that we discover the contingent
circumstances of external objects and human
294
9780826443595_Ch15_Final_txt_print.indd 294
psychology which require us to invent the
artificial virtues.’ (THN 3.2.1.9 / 474–5) The
artificial virtues include politeness, chival-
rousness, loyalty to government, chastity,14
promise-keeping and justice.
‘Artificial’ and talk of ‘invention’ are redo-
lent of the language of Mandeville and his
claim that moral distinctions are arbitrary
inventions for the preservation of man and
for his selfish pleasure and utility. A sense
of justice is not hardwired in virtuous men,
as Hutcheson claimed; justice, rather, is an
invention of human societies and it manifests
itself in the calculations of self-interested
creatures. And Hume does, with Mandeville,
reject the Hutchesonian picture: ‘we must
allow, that the sense of justice and injustice
is not deriv’d from nature’ (THN 3.2.1.17
/ 483). Hume, however, also has a role for
sympathy within his account of justice and of
the artificial virtues in general. Societies – and
the artificial virtues that they encourage – are
not planned, Mandevillean creations; rather,
[t]wo men, who pull the oars of a boat,
do it by an agreement or convention,
tho’ they have never given promises to
each other. Nor is the rule concerning
the stability of possession the less deriv’d
from human conventions, that it arises
gradually, and acquires force by a slow
progression, and by our repeated experi-
ence of the inconveniences of transgress-
ing it. (THN 3.2.2.10 / 490)
The convergence of such useful, virtuous
behaviour – or ‘conventions’ – arises through
our sympathetic engagement with our fel-
lows. If we were unrestrained, we would be
likely to break promises and steal the pos-
sessions of others. Such behaviour, however,
would lead to social problems and thus
rules of justice have developed. Theft is not
tempting if one adopts the common point of
2/9/2012 9:36:55 PM

view because from there the pain caused by
the loss of possessions would override one’s
own selfish motivations. Justice systems have
evolved, not through explicit forward plan-
ning or contracts between individuals, but
rather, through a long, gradual adoption of
co-operative behaviour guided by our natu-
ral sympathetic responses to others.15 At the
heart of Hume’s account is an altruistic sense
of the feelings of others, rather than, as for
Mandeville, calculated self-interest.
Such conventions, as one would expect,
can change over time. At times in history
virtues associated with military prowess can
impress; at other times such virtues recede
into the background.
The martial temper of the Romans,
enflamed by continual wars, had raised
their esteem of courage so high, that,
in their language, it was called virtue,
by way of excellence and of distinction
from all other moral qualities. . . . The
Scythians, according to Herodotus,
after scalping their enemies, dressed the
skin like leather, and used it as a towel;
and whoever had the most of those tow-
els was most esteemed among them. So
much had martial bravery, in that nation,
as well as in many others, destroyed the
sentiments of humanity; a virtue surely
much more useful and engaging.
It is indeed observable, that, among all
uncultivated nations, who have not, as
yet, had full experience of the advantages
attending beneficence, justice and the
social virtues, courage is the predomi-
nant excellence (EPM 7.13–15 / 254–5).
What is seen as virtuous changes over histori-
cal time and over the lifetime of individuals:
The decorum or indecorum of a qual-
ity, with regard to the age, or character,
295
9780826443595_Ch15_Final_txt_print.indd 295
HUME AND THE VIRTUES
or station, contributes also to its praise
or blame. . . . ’Tis usual to see men lose
their levity, as they advance in years.
Such a degree of gravity, therefore, and
such years, are connected together in
our thoughts. When we observe them
separated in any person’s character,
this imposes a kind of violence in our
imagination, and is disagreeable. (THN
3.3.4.12 / 612)
– ‘men’s virtues have their season, as fruits
do’.16
Changes occur because different traits
become useful to society and different things
are seen as agreeable.
I only meant to represent the uncertainty
of all these judgements concerning char-
acters; and to convince you, that fashion,
vogue, custom, and law, were the chief
foundation of all moral determinations.
The Athenians surely, were a civilized,
intelligent people, if ever there were one;
and yet their man of merit might, in this
age, be held in horror and execration.
The French are also, without doubt, a
very civilized, intelligent people; and
yet their man of merit might, with the
Athenians, be an object of the highest
contempt and ridicule, and even hatred.
(EPM Dial. 24 / 333)
All, however, is not in flux: there is much
that is consistent across time: both Athenians
and the French would look kindly on ‘good
sense, knowledge, wit, eloquence, humanity,
fidelity, truth, justice, courage, temperance,
constancy, [and] dignity of mind’ (EPM Dial.
27 / 334).
Hume’s distinction between the artificial
and natural virtues is not a rigid one. First,
artificial virtues are not less valuable than
natural virtues and it is just as laudable to
manifest the former. Second, both the natural
2/9/2012 9:36:55 PM
 HUME AND THE VIRTUES
and the artificial virtues depend on our sym-
pathy with those affected by the actions of
those who are virtuous. Both, also, can be
seen as constituting the character of an indi-
vidual: one can be an honest, polite and just
person (artificial virtues) just as one can be a
benevolent person (a natural virtue). Honest
people may not be innately honest, and they
may not be honest because they have ‘calcu-
lated’ that dishonesty would be bad for them;
honesty, rather, is a common trait – in a cer-
tain sense, a ‘natural’ trait17 – because peo-
ple have established conventions to inculcate
such behaviour via sympathizing with those,
for example, from which possessions have
been stolen. Conventions are artificial in the
sense that their roots lie in societal interac-
tion, but such interaction involves our innate
capacity for fellow-feeling and resultant con-
ventions can become second nature.
The natural virtues also depend on our
hardwired sympathetic responses. A man
who does not care for his children is seen as
not virtuous because we are able to sympa-
thize with his children and with others in his
narrow circle whom his negligence affects.
Given that both the natural and the artifi-
cial virtues involve sympathy, and that they
both play a constitutive role in character,
the distinction between the natural and the
artificial virtues all but collapses. Hume sug-
gests that:
The only difference betwixt the natural
virtues and justice lies in this, that the
good, which results from the former,
arises from every single act, and is the
object of some natural passion: Whereas
a single act of justice, consider’d in itself,
may often be contrary to the public
good; and ’tis only the concurrence of
mankind, in a general scheme or system
of action, which is advantageous. (THN
3.3.1.12 / 579)
296
9780826443595_Ch15_Final_txt_print.indd 296
The virtue of politeness, for example, is good
because the institution as a whole is ben-
eficial to society, even if no one is edified by
forced politeness to some officious or brutish
acquaintance.18 In contrast, though, a natu-
ral virtue such as benevolence is always ben-
eficial to others and to the virtuous person.
Any hard distinction between the natural
and the artificial virtues is explicitly rejected
in the second Enquiry, where all disputes
concerning the naturalness of justice are seen
as ‘merely verbal’ (EPM App. 3.9n64 / 308n).
Hume’s account of virtue – whether ‘natu-
ral’ or ‘artificial’ – depends on his account
of human nature and particularly on the fact
that we are naturally social creatures tuned
in, via sympathy, to the pleasures and pains of
our fellows. Such a sympathy-based picture
allows Hume a distinctive account of virtue
and of morality, one opposed to the natural,
God-given benevolence of Hutcheson and to
the calculating self-interest of Mandeville.
5. THE MONKISH VIRTUES
In the History of England we see royalty,
politicians and noblemen exemplifying a
wide range of virtues, including prudence,
discretion, generosity, affability, bravery,
patience, politeness, openness and sincerity.
Note, however, that faith, hope and charity
do not appear, nor other traits typically seen
as virtuous by the religious. Hume, ‘Christian
baiting’ as Baier puts it,19 claims that we
may even ‘transfer them [the virtues of the
religious] to the opposite column, and place
them in the catalogue of vices’ (EPM 9.3 /
270). In the Natural History Hume targets
‘the monkish virtues of mortification, pen-
ance, humility, and passive suffering’ (NHR
10.163), and in the second Enquiry he also
2/9/2012 9:36:55 PM

mentions ‘celibacy, fasting . . . , self-denial,
silence, [and] solitude’ (EPM 9.3 / 270).
He, in the voice of Philo, criticizes religion
for ‘raising up a new and frivolous species
of merit’ (DNR 12.222). The monkish vir-
tues ‘cross all . . . desirable ends; stupefy the
understanding and harden the heart, obscure
the fancy and sour the temper’ (EPM 9.3 /
270). Such displays do not, except in the heat
of religious fervour, fill us with approba-
tion and, if they do, this may be because we
are not considering them from the common
point of view. ‘A gloomy, hair-brained enthu-
siast, after his death, may have a place in the
calendar; but will scarcely ever be admitted
when alive, into intimacy and society, except
by those who are delirious and dismal as
himself.’ (EPM 9.3 / 270)
Suffering, for example, has a large part
to play in the lives of some Christians, not
only as something that good, compassion-
ate Christians should try to alleviate in oth-
ers, but also as something that it is good
for one to endure. The philosopher Marilyn
Adams sees personal suffering as a posi-
tive experience since it enables one to have
some appreciation of the suffering of Christ
and therefore some insight into the mind of
God.20 We should be grateful that we have
such insight and therefore such suffering.
Hume would see such thinking as unnatural
and perverse, something of a fetish.
Christianity also promotes humility, pride
being the first of the seven cardinal sins. This
is often applauded as a positive injunction,
even by critics of religion. Not so, however,
for Hume: for him, humility can be a vice and
pride is his first natural virtue. His account
of pride, though, is rather subtler than this
Christian-baiting headline might suggest.
We take pride from ‘qualities of the mind;
wit, good-sense, learning, courage, integrity:
from those of the body; beauty, strength,
297
9780826443595_Ch15_Final_txt_print.indd 297
HUME AND THE VIRTUES
agility, good mein, address in dancing, riding,
fencing: from external advantages; country,
family, children, relations, riches, houses, gar-
dens, horses, dogs, [and] cloaths’ (Abs. 30 /
659–60), and it is the imagination – and the
principles of association – that allows us to do
this. Thinking of my house leads me to think
of myself and the pleasure I take in my new
curtains leads to pleasure in my self, or pride.
Pride is good. It is a type of natural vir-
tue that is ‘always agreeable to ourselves’
(THN 3.3.2.9 / 597); it ‘capacitates us for
business’ (THN 3.3.2.14 / 600). It is ‘noth-
ing but a steady and well-establish’d pride
and self-esteem’ (THN 3.3.2.13 / 599) that
makes a ‘great’ person. Pride, and the self-
confidence that stems from it, leads to cour-
age, magnaminity and ambition, all qualities
of greatness.
[N]othing can be more laudable, than
to have a value for ourselves, where we
really have qualities that are valuable.
The utility and advantage of any qual-
ity to ourselves is a source of virtue, as
well as its agreeableness to others; and,
’tis certain, that nothing is more useful to
us . . . than a due degree of pride, which
makes us sensible of our own merit, and
gives us a confidence and assurance in
all our projects and enterprises (THN
3.3.2.8 / 596–7).
In The whole Duty of Man (1658), a popular
Protestant devotional work usually attrib-
uted to Richard Allestree, humility is seen as
the prime virtue, of which there are two sorts,
both virtuous: ‘the first is the having a mean
and low opinion of our selves, the second is
the being content that others should have so
of us.’ We need humility ‘to keep us from any
high conceits of our performances, which
if we once entertain, it will blast the best of
2/9/2012 9:36:55 PM
 HUME AND THE VIRTUES
them, and make them utterly unacceptable
to God’.21 Such an account is anathema to
Hume.22 The Church sees vice as bringing
(illicit) pleasure; Hume, however, sees virtues
as grounded in pleasure and utility.
There are, however, limits to Hume’s sub-
versive reversal of the moral status of pride
and humility: self-esteem may be laudable,
but vanity is not.
Nothing is more disagreeable than a
man’s over-weaning conceit of himself:
Every one almost has a strong propen-
sity to that vice: No one can well dis-
tinguish in himself betwixt the vice and
virtue, or be certain, that his esteem of
his own merit is well-founded: For these
reasons, all direct expressions of this
passion are condemn’d; . . . that imper-
tinent, and almost universal propensity
of men, to over-value themselves, has
given us such a prejudice against self-
applause, that we are apt to condemn
it, by a general rule, wherever we meet
with it; and ’tis with some difficulty we
give a privelige to men of sense, even
in their most secret thoughts. (THN
3.3.2.10 / 597–8)
Pride can have a destabilizing effect on soci-
ety since ‘when we compare the sentiments of
others to our own, we feel a sensation directly
opposite to the original one’ (THN 2.2.9.1
/ 381). Pride in another leads to humility in
ourselves, and humility is vicious: ‘firm per-
swasion he has of his own merit, takes hold
of the imagination, and diminishes us in our
own eyes’ (THN 3.3.2.6 / 595).23 Pride may
be pleasant to those who feel it, but some-
times it is unpleasant to witness it in others.
Society, however, has found a solution to this
destabilizing quality of pride. It has come to
be seen as polite to conceal one’s self-love
and vanity.
298
9780826443595_Ch15_Final_txt_print.indd 298
’Tis, however, certain, that good-
breeding and decency require that we
shou’d avoid all signs and expressions,
which tend directly to show that pas-
sion . . . . In like manner, therefore, as
we establish the laws of nature, in order
to secure property in society, and pre-
vent the opposition of self-interest; we
establish the rules of good-breeding, in
order to prevent the opposition of men’s
pride, and render conversation agreeable
and inoffensive. . . . At least, it must be
own’d, that some disguise in this particu-
lar is absolutely requisite; and that if we
harbour pride in our breasts, we must
carry a fair outside, and have the appear-
ance of modesty and mutual deference
in all our conduct and behaviour. (THN
3.3.2.10 / 597–8)24
Again, this solution is not the result of our
natural modesty, nor of a self-interested
calculation of how to get the most pleas-
ure out of life; it is, rather, the result of the
gradual development of, in this case, rules of
decency – the artificial virtue of politeness –
through our sympathetic interaction with
each other.
Hume thus rejects the monkish virtues
and religious morality in general; instead
he offers an alternative, secular, naturalistic
account of virtue, one based on social and
personal interests and grounded in our sym-
pathetic responses to others. For Hume, reli-
gious morality can be vicious, perverse and
even comical.
Some ornaments, which the ladies at that
time wore upon their petticoats, excited
mightily the indignation of the preach-
ers; and they affirmed, that such vanity
would provoke God’s vengeance, not
only against these foolish women, but
against the whole realm. (H 4.42)25
2/9/2012 9:36:56 PM

It is because we have such natural moral sen-
timents that we are able to see such indig-
nation towards petticoat ornaments as silly.
We find, however, that Hume is rather less
scathing about early polytheistic religions,
claiming that they are morally superior to
monotheism and Christianity. The monkish
virtues are a peculiarly Christian develop-
ment and they ‘sink the human mind into the
lowest submission and abasement’ (NHR
10.163). ‘The doctrines of the Christian
religion . . . recommend only passive cour-
age and suffering, had subdued the spirit of
mankind, and had fitted them for slavery
and subjection.’ (NHR 10.164) Polytheists,
however, can ‘aspire to a rivalship and emu-
lation of them [their gods]. Hence activity,
spirit, courage, magnanimity, love of liberty,
and all the virtues which aggrandize a peo-
ple.’ (NHR 10.163–4) The move from poly-
theism to monotheism has been a retrograde
step.
Instead of the destruction of monsters, the
subduing of tyrants, the defence of our
native country; whippings and fastings,
cowardice and humility, abject submis-
sion and slavish obedience, are become
the means of obtaining celestial honours
among mankind. (NHR 10.164)
Monotheistic religion, then, has led to a
perverted sense of virtue. Hume is hos-
tile towards religion and this hostility is
grounded, to a great extent, in what he sees
as the moral failings of religion.
Annette Baier approvingly sums up Hume’s
account of virtue:
Human happiness is the touchstone, and
Hume takes it to be obvious that hap-
piness requires fellowship, commerce,
intercourse. Consistently with this test,
299
9780826443595_Ch15_Final_txt_print.indd 299
HUME AND THE VIRTUES
he includes in his catalogue of virtues all
and only the qualities of head, heart and
expressive body that he believes we will
agree do make a person a welcome fellow,
whether in ‘that narrow circle, in which
any person moves’ or in ‘the greater soci-
ety or confederacy of mankind’.26
NOTES
1
Richard Dees, in ‘Hume on the Characters of
Virtue’, Journal of the History of Philosophy
35(1) (1997), pp. 45–64, discusses various ten-
sions between these virtues, where, for example,
a trait may be useful to oneself yet disagreeable
to others.
2
Such delight and uneasiness are ‘peculiar
sentiments of pain and pleasure’ (THN 3.3.1.3
/ 574), distinguishable from the kinds of senti-
ments we may feel when, say, we are uneasy at
the prospect of a daunting journey.
3
G. Sayre-McCord, ‘On Why Hume’s General
Point of View Isn’t Ideal – and Shouldn’t Be’,
Social Philosophy and Policy 11(1) (1994),
pp. 202–28, p. 202.
4
In the Treatise Hume provides an associationist
account of the mechanics of sympathy. We can
embrace the sentiments of others – feel them –
through our idea of their sentiments becoming
enlivened. In sympathy, the source of this extra
vivacity is the impression of the self. ‘’Tis evident,
that the idea, or rather impression of ourselves
is always intimately present with us, and that
our consciousness gives us so lively a concep-
tion of our own person, that ’tis not possible to
imagine, that any thing can in this particular go
beyond it. Whatever object, therefore, is related
to ourselves must be conceived with a like vivac-
ity of conception.’ (THN 2.2.11.4 / 317) An
idea of a passion is enlivened by the vivacity of
the impression of the self, thus becoming itself a
passion. Such an account, however, is problem-
atic given Hume’s scepticism concerning the
self (THN 1.4.6 / 251–63). It is perhaps for this
reason that Hume drops this particular account
of sympathy in his later works. In the first and
second Enquiries sympathy is seen as just a brute
fact about human thought.
2/9/2012 9:36:56 PM
 HUME AND THE VIRTUES
Some, however, have argued that Hume is
only sceptical with regard to our having the
impression, and thus idea, of a ‘simple and
continu’d’ self (THN 1.4.6.3 / 252), a self with
‘perfect identity and simplicity’ (THN 1.4.6.1
/ 251) – a persisting, Cartesian self. We can,
however, have impressions and ideas of our
social, embodied selves (see A. Baier, A Progress
of Sentiments (Cambridge, MA: Harvard
University Press, 1991), p. 130).
8
Herdt notes that Hume talks of our aware-
ness of our personal identity with respect to
the passions: ‘’tis evident, that . . . we are at all
times intimately conscious of ourselves, our
sentiments and passions’ (THN 2.2.2.15 / 339).
It is similarities in our passions through time
that create the impression of a self. Hume high-
lights the role of the passion of pride: ‘nature
has given to the organs of the human mind, a
certain disposition fitted to produce a peculiar
impression or emotion, which we call pride:
To this emotion she has assign’d a certain idea,
viz. that of self, which it never fails to produce’
(THN 2.1.5.6 / 287).
5
S. Blackburn, How to Read Hume (London:
Granta, 2008), p. 61.
6 9
See also THN 3.2.1.3 / 477–8: ‘If we find,
upon enquiry, that the virtuous motive was
still powerful over his breast, tho’ check’d in
its operation by some circumstances unknown
to us, we retract our blame, and have the same
esteem for him, as if he had actually perform’d
10
the action, which we require of him.’
7
I shall not discuss here whether Hume intends
the standard set by the common point of view
to be normative and, if so, whether his natu-
ralistic account of normativity is successful.
R. Hardin, David Hume: Moral and Political
Theorist (Princeton: Princeton University
Press, 2007), chap. 2, argues that Hume
only offers a descriptive account of human
psychology, ‘except in the occasional “sally of
panegyric”’ (ibid., p. 200) where he seemingly
recommends certain virtues. Bentham has a
11
similar interpretation, claiming that Hume
only intends ‘to account for that which is,
[whereas] I shew what ought to be.’ (‘Bentham
to Étienne Dumont’, 6 September 1822, in C.
Fuller (ed.), The Correspondence of Jeremy
Bentham (Oxford: Clarendon Press, 2000),
vol. 11, p. 149; quoted in J.P. Wright, Hume’s
300
9780826443595_Ch15_Final_txt_print.indd 300
A Treatise of Human Nature: An Introduction
(Cambridge: Cambridge University Press,
2009), p. xiv). For normative interpretations of
Hume see R. Cohon, Hume’s Morality (New
York: Oxford University Press, 2008) and L.
Loeb, Stability and Justification in Hume’s
Treatise (Oxford: Oxford University Press,
2002), pp. 119–23). Also see the chapters by
Hardin and Driver in this volume for further
discussion.
As James Harris notes in this volume: ‘without
any textual authority, David and Mary Norton
change the title of Section Five of Part III of
Book Three from “Some farther reflections con-
cerning the natural virtues” to “Some farther
reflections concerning the natural abilities”.
Hume intended “natural virtues”, it may be
presumed, as part of his strategy of unsettling
the reader’s grip on the distinction between
virtues and abilities.’ (p. 226n9) Baier talks
of Hume’s ‘defiant inclusion’ of such abilities
in his catalogue of virtues (A. Baier, ‘Kinds
of Virtue Theorist: A Response to Christine
Swanton’, in C. Pigden (ed.), Hume on
Motivation and Virtue (Basingstoke: Palgrave
Macmillan, 2009), pp. 249–58, p. 251).
Or, for that matter, of vice: ‘A blemish, a fault,
a vice, a crime; these expressions seem to
denote different degrees of censure and disap-
probation; which are, however, all of them, at
the bottom, pretty nearly of the same kind of
species.’ (EPM App. 4.22 / 322)
Hume notes that the ancient moralists did
not distinguish between virtues and talents;
‘Cicero, in imitation of all the ancient moral-
ists, . . . enlarges very much his ideas of virtue,
and comprehends every laudable quality or
endowment of the mind, under that honour-
able appellation [“virtue”]’; ‘the ancient moral-
ists . . . made no material distinction among
the different species of mental endowments
and defects, but treated all alike under the
appellation of virtues and vices’ (EPM App.
4.11 / 318).
If moral distinctions are discoverable through
reason, Hume argues that we should be
able to discover vice in thinking about the
objective relations involved in, for example,
incest and matricide. Such objective distinc-
tions are not, however, to be found. Incest
in animals does not have ‘the smallest moral
2/9/2012 9:36:56 PM

turpitude or deformity’ (THN 3.1.1.25 / 467)
and there is no apparent moral wrong-doing
involved when an oak sapling ‘overtops and
destroys the parent tree’ (THN 3.1.1.24 / 467).
For discussion of Hume’s other arguments
against rationalist conceptions of morality,
see J. Baillie, Hume on Morality (London:
Routledge, 2000), chap. 5.
12
Hutcheson takes us to have a ‘moral sense’
that enables us to judge vice and virtue with-
out acts of reasoning, and associated with such
‘perception’ is a particular kind of aesthetic
pleasure, that associated with ‘moral beauty’.
Hume also refers to virtue as ‘moral beauty’
(THN 2.1.8.3 / 300) or ‘beauty of . . . the
mind’ (EPM 6.23 / 244), but I shall not discuss
here whether he should be seen as a moral
sense theorist or how far such talk of moral
beauty commits him to an anti-realist account
of morality. See M. Gill, ‘Moral Rationalism
vs. Moral Sentimentalism: Is Morality More
Like Math or Beauty?’, Philosophy Compass
2(1) (2007), pp. 16–30.
13
Sentimentalist theories, including that of
Hume, influenced early utilitarian accounts of
morality. Hardin (David Hume, p. 231) says
that ‘[b]y focusing on utility and putting moral
and political theory into a single coherent
frame, he opens the door for the grand entry
of utilitarianism’. On reading Hume’s account
of justice – see below – Jeremy Bentham felt
as though ‘scales had fallen from [his] eyes’ in
discovering that ‘the foundations of all virtue
are laid in utility’ (Fragment on Government,
in J. Bowring (ed.), Works of Jeremy Bentham,
10 vols (New York: Russell and Russell, 1962),
vol. 1, p. 149, note H). Hume, though, is not
a utilitarian. According to him, actions are
not good according to whether they maximize
utility; rather, a person is virtuous if she has
the sorts of character traits that lead to me
sympathizing with the likely approbation
of her narrow circle. Hume’s moral theory
is character-based rather than action-based.
The primary focus of moral verdicts is on the
character and virtues of those we are judging,
although – and here is the root of utilitarian
thinking – according to such a virtue theory,
virtues are not good in themselves; they are
only good because they aid utility in some
way. For virtue ethicist readings of Hume, see
301
9780826443595_Ch15_Final_txt_print.indd 301
HUME AND THE VIRTUES
C. Swanton, ‘What Kind of Virtue Theorist
Is Hume?’, in C. Pigden (ed.), Hume on
Motivation and Virtue (Basingstoke: Palgrave
Macmillan, 2009), pp. 226–48 and G. Sayre-
McCord, ‘Hume and the Bauhaus Theory of
Ethics’, in P.A. French, T.E. Uehling and H.K.
Wettstein (eds), Midwest Studies in Philosophy
20 (1996), pp. 280–98.
14
‘The long and helpless infancy of man requires
the combination of parents for the subsistence
of their young; and that combination requires
the virtue of chastity or fidelity to the mar-
riage bed. Without such a utility, it will readily
be owned, that such a virtue would never have
been thought of’ (EPM 4.5 / 206–7).
15
Although our natural sympathetic responses
can be reinforced and encouraged by the
actions of politicians and parents. See THN
3.2.2.25–6 / 500–1.
16
La Rochefoucauld, Collected Maxims and
Other Reflections (Oxford: Oxford University
Press, 2007), p. 83.
17
See THN 3.2.2.19 / 484: ‘To avoid giving
offence, I must here observe, that when I deny
justice to be a natural virtue, I make use of
the word, natural, only as oppos’d to artifi-
cial. In another sense of the word; as no prin-
ciple of the human mind is more natural than
a sense of virtue; so no virtue is more natural
than justice. Mankind is an inventive spe-
cies; and where an invention is obvious and
absolutely necessary, it may as properly be
said to be natural as any thing that proceeds
immediately from original principles, without
the intervention of thought or reflection. Tho’
the rules of justice be artificial, they are not
arbitrary. Nor is the expression improper
to call them laws of nature; if by natural we
understand what is common to any species, or
even if we confine it to mean what is insepa-
rable from the species.’ And EPM App. 3.9 /
307: ‘If self-love, if benevolence be natural to
man; if reason and forethought be also natu-
ral; then may the same epithet be applied to
justice . . . the sentiment of justice, throughout
all ages, has infallibly and certainly had place,
to some degree or other, in every individual of
the human species. In so sagacious an animal,
what necessarily arises from the exertion
of his intellectual faculties, may justly be
esteemed natural.’
2/9/2012 9:36:56 PM
 HUME AND THE VIRTUES
18
See also THN 3.2.2.22 / 497: ‘When a man of
merit, of a beneficent disposition, restores a
great fortune to a miser, or a seditious bigot, he
has acted justly and laudably; but the public is
a real sufferer.’
19
Baier, A Progress of Sentiments, p. 207.
20
M. McCord Adams, Horrendous Evils and
the Goodness of God (Ithaca, NY: Cornell
University Press, 1989).
21
R. Allestree, The Whole Duty of Man (1658),
6.3, 2.4; quoted in J. Herdt, Religion and Faction
in Hume’s Moral Philosophy (Cambridge:
Cambridge University Press, 1997), p. 71.
22
‘[I]f Cicero were now alive, it would be found
difficult to fetter his moral sentiments by nar-
row systems; or persuade him, that no qualities
were to be admitted as virtues, or acknowl-
edged to be a part of personal merit, but what
were recommended by The Whole Duty of
Man’ (EPM 4.11n72 / 319n); ‘Upon the whole,
I desire to take my Catalogue of Virtues from
Cicero’s Offices, not from the Whole Duty of
Man.’ (LDH 1.34, 13)
23
No one puts it better than Gore Vidal:
‘whenever a friend succeeds, a little something
in me dies.’ (The Sunday Times Magazine,
16 September 1973)
302
9780826443595_Ch15_Final_txt_print.indd 302
24
See also E 132–3 and EPM 8.1 / 261.
25
Also see H 1.241–2 for Hume’s mocking
treatment of the Church’s vehement condem-
nation of a certain mode of footwear in the
eleventh century: ‘Though the clergy, at that
time, could overturn thones, and had author-
ity sufficient to send above a million of men
on their errand to the desarts of Asia, they
could never prevail against these long-pointed
shoes.’ (H 1.242)
26
Baier, A Progress of Sentiments, p. 219. See
also A. Baier, ‘Civilizing Practices’, in Postures
of the Mind (London: Methuen & Co., 1985),
pp. 246–62, 258–9: ‘Hume was a thorn in
the flesh for the defenders of religion in his
own day not just because of the case against
religion he made patiently and thoroughly in
all his writings, but because he so obviously
in his own life achieved the sort of calm, and
grace of spirit, that his opponents thought
only supernatural powers, self-denial, and
floggings of the flesh could achieve. He man-
aged to live, and live well, and live without
any apparent sense of purposelessness, within
the very secular practices and tradition he
examined, cultivating the virtues he had
analysed.’
2/9/2012 9:36:56 PM
 16
HUME’S HUMAN NATURE
Russell Hardin
David Hume invokes human nature to
explain individual behaviour and belief. His
account of human nature is partly explicit
and partly implicit, so that we usually see it in
use rather than enunciated, whereas one cen-
tral part of it, his assumption of self-interest,
self-love, etc., is frequently stated and illus-
trated. He addresses various aspects of our
nature, including our strong bias towards our
self-interest, our capacity for empathy, and
our capacity for thinking strategically. Self-
interest was arguably the centrally most div-
isive issue in the efflorescence of Scottish and
English moral theorizing during the eight-
eenth century. Almost universally among
these thinkers self-interest was disparaged
and their theorizing was directed at defeat-
ing or overriding it in order to enable us to
be moral. Almost uniquely among these writ-
ers Hume attempts to deal positively with
self-interest. One of his precursors, Bernard
Mandeville, exulted in the supposed good
effects of self-interest. Hobbes dealt with it
by elevating a powerful sovereign to keep
it in check and even to crush it. Hume took
it fully into account in trying to explain how
we could successfully create institutions to
accomplish good ends despite erecting the
institutions on self-interested individuals.
Two other central elements of Hume’s
thought are more complex and will be treated
303
9780826443595_Ch16_Final_txt_print.indd 303
at greater length below. In brief, empathy
comes to us through nature in the psycho-
logical phenomenon of mirroring. When
I see you smile or wince, I tend to smile or
wince in return and to have the feelings
that normally accompany these expressions.
Others have taken note of this curious trait
but Hume makes it an important part of the
explanation of our often sociable behaviour.
Next, Hume is the greatest strategic thinker
in all the history of philosophy and he thinks
that this capacity is part of human nature.
As a strategic thinker he is rivalled, if at all,
only by Thomas Hobbes and by later econo-
mists such as Adam Smith, who was his great
friend, and Thomas Schelling in our time.
That is to say that, when Hume addresses
any major topic, he typically frames it in a
social, not an abstract, context. This is espe-
cially true for his account of, and many refer-
ences to, human nature.
Finally, Hume resolves a problem in vir-
tue theory that perplexed other philosophers:
a seeming difference in the motivations for
individual and institutional actions.
Among the most important and, in his
time, unusual aspects of Hume’s approach to
morality is his claim to investigate this and
other topics as a matter of science. According
to the subtitle to the Treatise he makes An
Attempt to Introduce the Experimental
2/9/2012 9:39:24 PM
 HUME’S HUMAN NATURE
Method of Reasoning into Moral Subjects.
Hence his account should be read as explan-
atory rather than as normative. He gives us
a science and not a philosophy of human
nature. He sees himself as a follower of
Newton, although it is fair to say that his
experiments were generally thought experi-
ments, very nearly puzzles rather than data.1
He assumes that humans are fundamentally
social animals in the sense of depending on
each other.2 Human nature is a social con-
struction, not an innately determinate mat-
ter. His strategic account here is surely too
demanding for most people consciously to
follow his vision in their daily lives, although
they might commonly do so without full
understanding of the strategic burdens they
face. Thinkers who do not understand stra-
tegic thinking might also not follow Hume’s
accounts of many phenomena.
1. SELF-INTEREST
Hume says, ‘[n]othing is more certain, than
that men are, in a great measure, govern’d by
interest, and that even when they extend their
concern beyond themselves, ’tis not to any
great distance’ (THN 3.2.7.1 / 534). This is
Hume’s dominant claim about human nature
in the Treatise. Holding this view seemingly
must make his programme to explain moral
outcomes in the world of human and insti-
tutional action implausible. But that is his
audacious programme, to explain why it
is that a motley collection of substantially
self-interested people could achieve justice
and other beneficial social results, all with-
out a primary or a religious commitment to
acting morally. Some moralists in Hume’s
time ran these two together, supposing their
preferred god instills moral character in us.
304
9780826443595_Ch16_Final_txt_print.indd 304
This assumption is patently beyond belief for
Hume, who thinks, on the basis of pervasive
evidence, that our nature is not fundamen-
tally moral; and he further doubts all argu-
ments for the truth of religious claims. Still
we often act as though we were moral and
we establish social rules, norms and laws evi-
dently without necessary moral commitment
to their working. How do we turn this trick?
Explaining our unintended success in doing it
is the major programme of the Treatise, Book
III. A clue to Hume’s argument is that, if we
had such characters as the religious, moralis-
tic and pure-reason schools of thought sup-
pose, we would not be much concerned with
law because we would be well ordered with-
out its additional incentives.
To begin an account of self-interested
human nature and the ways it affects social
and political practices and institutions, let us
canvas a few direct arguments Hume makes
for a virtually necessary connection between
self-interested human nature and the form of
such institutions as those for securing justice
in Hume’s sense of justice as order (not dis-
tributive justice):
(1) The proper business of municipal laws
is to fix what the principles of human
nature have left undetermined (THN
3.2.3.11n76 / 513).
(2) If we follow our natural bent, we will
perform few actions for the advantage
of others from disinterested views (THN
3.2.5.8 / 519–21). This runs against the
views of many other moral philosophers
of Hume’s time, many of whom think our
brains are somehow committed to moral
action, perhaps because god has planted
morality in us or, in current jargon,
because we are hardwired to be moral.
(3) Interest is the first motive for keeping
promises (THN 3.2.5.12–13 / 523–4).
There is a vast literature that moralizes
2/9/2012 9:39:24 PM

promise-keeping. Hume says that such
views are mistaken; that it is typically in
our interest to keep our promises because
doing so enables us to enter into other
promises and contributes to our benefi-
cial reputation. Generally there need be
no further consideration to motivate us
to keep our promises.
(4) The general system comprehending
the interest of each individual is also
advantageous to the public though it
not be intended for that purpose by the
inventors (THN 3.2.6.6 / 528–9). I will
expand on this important passage below.
It can rightly be seen as Hume’s central
formula in the use of self-interest to
accomplish general good.
(5) Justice handles defects of man that can
be remedied by society. For example,
it manages the division of labour to
enhance collective productivity (THN
3.2.2.3 / 485, 3.2.4.1 / 514).
(6) It is ‘only from the selfishness and
confin’d generosity of men, along with
the scanty provision nature has made
for his wants, that justice derives its ori-
gin.’ Therefore, justice cannot be either a
prior or a universal principle. It is clearly
not abstractly or otherwise deducible but
is entirely contingent, not a priori ratio-
nal (THN 3.2.2.18 / 495; also 3.2.6.6 /
528–9). Public ‘utility is the sole origin
of justice, and . . . reflections on the ben-
eficial consequences of this virtue are the
sole foundation of its merit’ (EPM 3.1 /
183); ‘self-interest is the original moti-
vation to the establishment of justice’
(THN 3.2.2.24 / 499).
(7) In ‘order to establish laws for the regula-
tion of property, we must be acquainted
with the nature and situation of man’
(EPM 3. 27 / 194).
Hume makes this final point, that justice as
order makes no sense in those states in which
conditions are misfitted in many ways. For
305
9780826443595_Ch16_Final_txt_print.indd 305
HUME’S HUMAN NATURE
example, he says that the fictional state of
nature or a golden age can be useful in show-
ing that justice is contingent because it would
be irrelevant in some kinds of society or
among some kinds of people, whose charac-
ter must differ from ours. His great achieve-
ment is to make collective outcomes very
different from the human material going into
that result. This is his unique answer to the
eighteenth-century puzzle of how to explain
or engrain seemingly moral behaviour.
Hume largely shares Thomas Hobbes’s
basic assumption of what we now call homo
economicus, of individuals as largely self-
interested, although Hume’s position here is
considerably milder than Hobbes’s, whose
position is milder than Bernard Mandeville’s.
Hume may have largely relented from this
focus on self-interest in his works after
the Treatise. In any case, one must qualify
the homo economicus claim with ‘largely’
because both Hobbes and Hume allow for
concern for others. For example, Hume
enunciates a principle for altruism or benef-
icence. He says, ‘[a] man naturally loves his
children better than his nephews, his neph-
ews better than his cousins, his cousins bet-
ter than strangers, where everything else is
equal’ (THN 3.2.1.18 / 483–4). In response
to claims that altruism must be genetically
programmed into us through natural selec-
tion, the great population geneticist, J.B.S.
Haldane, perhaps under the influence of the
pint he was drinking, is supposed to have
quipped that he would lay down his life for
two of his brothers or eight of his cousins.3
Hume is the original Haldane of beneficence.
He elaborates the point to show that it is an
important part of human nature and moral-
ity, attributing this partiality to ‘the original
frame of our mind’ in which ‘our strongest
attention is confin’d to ourselves; our next
is extended to relations and acquaintance;
2/9/2012 9:39:24 PM
 HUME’S HUMAN NATURE
and ’tis only the weakest which reaches to
strangers and indifferent persons’ (THN
3.2.2.8 / 488).
Hume has two extraordinary insights that
have been slighted, even ignored in the vast
body of commentary on his work. One of
these is the psychological phenomenon of
mirroring, and the second is the sociologi-
cal account of convention in the spontane-
ous, iterated co-ordination of large numbers
of people in mutually beneficial ways. After
brief discussion of Hume’s naturalism, I will
therefore begin with these two issues. There
are, incidentally, no entries for convention
or for mirroring in the index of the generally
superb new Oxford edition of Hume’s Treatise
although Hume discusses both phenom-
ena there (for example, respectively at THN
3.2.2.9–12 / 489–92 and 2.2.5.21 / 365).4
2. NATURALISM
Because Hume does not present a moral the-
ory but only a naturalist account of why we
have the moral views and the political princi-
ples that we have, we cannot judge his theory
to be morally right or wrong in principle any
more than we can say physics is morally right
or wrong. Hence, advocates of quite varied
moral theories can, if they share his psycho-
logical and sociological views, accept virtu-
ally all of Hume’s account and could use it to
enrich their own theories. For example, his
explanations could fit the views of Hobbes
or John Rawls if the sociological possibili-
ties are fitted to their times. The implication
for Hobbes is a draconian ruler to control
very self-centred subjects, some of whom
might even be potential thieves. For Rawls
it is to inculcate moral commitment to fair-
ness through education and encouragement.
306
9780826443595_Ch16_Final_txt_print.indd 306
Contemporary theorists face a much broader
range of possibilities than Hobbes or Hume
faced because our resources and theoretical
understandings far exceed those of earlier
times. Consider one example. Rawls does the
political philosophy of distributive justice in
our time. Much of his argument would be
well beyond the possibilities in Hume’s time,
as Hume de facto argues in his discussion
of efforts to achieve ‘perfect equality’ (EPM
3.26 / 194). He might also have noted the
irrelevance of concern with distributive jus-
tice in the tiny, essentially separate and dis-
connected towns of medieval England.
In essence, Hume’s first guiding insight is
recognition that our moral views are natu-
ral phenomena and they are therefore to be
explained just as other natural phenomena
are. They are a subject for scientific exami-
nation and explanation, not for moral judge-
ment. And their content is most certainly
not subject to claims of truth or falsity. They
just are. They might be the product of the
brain at work in speculative moments or a
déjà vu of views that have been drummed
into us in childhood and later. This view
would make eminently good sense of the
diversity of moral claims and, especially, of
so-called intuitions. Indeed, the force with
which many people (especially intuitionist
moral philosophers) evidently believe and
assert their moral intuitions is itself no more
than another natural phenomenon in need
of explanation rather than of affirmation,
which must be circular.
Hume’s naturalism implies that he has a
moral theory (to which I have been referring)
only in the sense that it is an explanatory,
not a normative theory. He is and means
to be a social theorist. There is no way to
demonstrate the truth of a normative theory
and social science must leave such a question
aside.
2/9/2012 9:39:24 PM

Occasionally, Hume notes that he finds
some action or character very appealing
and worthy of praise. He then often imme-
diately depreciates his judgement by calling
it panegyric, as though to explain it psycho-
logically and thereby to dismiss it as morally
irrelevant, as enthusiasm gone awry. In one
passage he says, ‘it is difficult to abstain from
some sally of panegyric’ (EPM 2.5 / 177–8).
He does not attempt an explanation of such
‘moral’ intuitions. He implicitly grants that
his own psychology is subject to such flaws,
which seemingly derive from the pleasure we
take from observing virtuous acts. He con-
tinues in the suggestion that we mirror the
sensations of those we praise.
3. MIRRORING
Hume is a discoverer of the phenomenon of
mirroring, no doubt along with many others
who are independently observant in daily
life. He appears, however, to be unique in
recognizing its significance and putting the
phenomenon to work in explaining certain
sensations and actions. ‘The human counte-
nance, says Horace, borrows smiles or tears
from the human countenance’ (EPM 5.18 /
220). That is to say, if I see you face to face
when you are in great sadness, I will mirror
your sadness to the extent of being sad myself.
To some degree, this means that we can mas-
ter the so-called problem of other minds.
Hume cites Horace as an expert observer
who sees the phenomenon but who other-
wise makes little of it. It is typical of Hume’s
scientific bent that he goes on to use his grasp
of the phenomenon first to spell it out in a
perspicuous way and then to make compel-
ling use of it as a core part of human nature.
As John Laird says, Hume is a keen observer
307
9780826443595_Ch16_Final_txt_print.indd 307
HUME’S HUMAN NATURE
as well as a deft psychologist.5 Nowhere does
he show this so clearly as in his nearly unique
appreciation of mirroring.
Mirroring underlies sympathy, or what
we might today sooner call empathy. In this
view, sympathy is genetically wired into us. It
is not simply a normative commitment that
we have chosen for ourselves. This is possi-
bly the most miscast issue in commentaries
on Hume’s social and political philosophy.
For Hume, empathy is a scientific, psycho-
logical matter and is not normative but is,
in today’s vocabulary, genetic. Empathy (or
sympathy) is a centrally important part of
human nature.6 This descriptive fact has no
moral content. In Hume’s view, it is simply
wrong to make of it a moral issue.
Hume’s most compelling example of mir-
roring is of a ship sinking offshore. Onshore
observers might vaguely see or surmise that
there are people on the ship and that they
are likely to drown. Their interest in and this
account of the sinking ship are ones of sim-
ple curiosity. Hume then supposes the ship
is near enough for us to see the agony and
terror in the faces of those about to drown.
Seeing their emotions being played out, we
mirror these emotions and we suffer with
those who are about to drown (THN 3.3.2.5
/ 594–5; see also 2.1.11.6 / 318).
It is sad that Hume’s insight on mirroring
was neglected for well more than a century
before it was introduced into psychology
about a century ago. Since then, research sug-
gests that there is a specific class of ‘mirror’
neurons in the brain that perform Hume’s
trick. We might expect soon to see accounts
of the evolution of these neurons driven by
their benefits to us in causing us to empathize
with others, especially infants and young
children who depend on generous caretak-
ers for their survival and flourishing. This
is one of the most important stages in the
2/9/2012 9:39:24 PM
 HUME’S HUMAN NATURE
evolutionary creation of humans. It will be of
interest to discover which other animals have
any of this capacity, as many appear to have.
Newborns soon smile at the adults around
them and the adults mirror the smiles, thus
reinforcing the newborns’ connection to the
world. Virtually all parents of wanted chil-
dren must have had such experiences and
must have enjoyed their nearly automatic
connection with their children.
Hume is sufficiently perceptive that, once
he has noticed the phenomenon, he observes
it carefully, and then makes it the central
element of his account of the psychology of
sympathy. Therefore it is for him part of dis-
tinctively moral psychology, or of morality
psychologized, as discussed below. Mirroring
makes Hume’s explanation of morality psy-
chologically richer than all of the contem-
porary moral sense and sentiments theories,
which are inherently psychological in their
foundations. Their proponents are generally
content to stop their inquiries at the point of
asserting that we just do know right from
wrong, that reason can determine these,
or that God has given us such knowledge.
Hume has empirically observed – and sup-
poses we can all observe – the phenomena of
mirroring and sympathy.
There is a further task that must be daunt-
ingly difficult: evoking an emotional response
of a similar kind in response to the interests
of society. We cannot mirror society – where
is its face? – but can only reason about it.
Perhaps the best we can do is some version
of a federated or hierarchical set of commit-
ments. The result or clear implication will be
a much weaker set of commitments to the
society than we have to the individuals with
whom we interact closely enough to mirror
each other. We interact directly with very
many individuals, but still with only a tiny
fraction of all those in our society. But we
308
9780826443595_Ch16_Final_txt_print.indd 308
might often recognize that we will be enticed
to co-ordinate with many others whose com-
mitments can be read from the general ten-
dency of those with whom we do interact.
Any group with whom I interact may, how-
ever, be very different from various other
groups in our society.
The algebra of our mirrored commitments
to a group and to all individuals with whom
we do meaningfully interact enough to mir-
ror each other suggests many variations in
the nature of one individual’s and another’s
commitments. For an obvious example, one
who is a relative recluse cannot have as
strong a federated commitment to the soci-
ety as someone who is richly involved in
exchanges and other interactions with fellow
citizens. Their individual-level interactions
may cover people in more than one society,
so that their patriotism or group loyalty, for
example, might be weaker than that of citi-
zens with a narrower range of interactions.
In sum, mirroring is a fundamentally
important part of Hume’s human nature.
It defines and enables sympathy (empathy),
without which we would not have given rise
to any of the institutions of order and jus-
tice. Without it, our lives would fit Hobbes’s
characterization as nasty, brutish and short.
Mirroring is the cue for humanity.
4. CONVENTION
If many of us repeatedly face a co-ordination
problem, we are apt eventually to co-ordinate
on one of the possible outcomes of our inter-
action. Once we do, we are then given incen-
tive to resolve our problem that same way
thereafter. At some point in our sequence of
repeated actions we may simply stop analys-
ing our former problem and stick with our
2/9/2012 9:39:24 PM

resolution of it. At that point it becomes
a convention in this formal sense.7 We face
remarkably many such interactions in our
daily lives. The ideal example is the rule of
the road: drive right in North America and
left in the British Isles. If there were only one
feasible choice, we would have no problem
selecting it. Game theorists call such a sim-
ple interaction a game of harmony. The driv-
ing rule is problematic because there are two
equally attractive choices available so that
we must somehow co-ordinate on one rather
than the other. Thomas Schelling presents a
highly varied array of such interactions.8
Again, for Hume the capacity for thinking
strategically is part of human nature. As is
true of other parts as well, this one is dif-
ferentially rich in people’s capacities. Hume
has astonishing capacity for strategic think-
ing, so much so that he can do strategic
analyses without the modern vocabulary of
game theory that makes the task radically
easier, so that ordinarily bright undergradu-
ates can do such analyses today. Hume dis-
cussed conventions before modern traffic
patterns presented us with what is now seen
as the definitive example of a convention:
drive right in North America and drive left
in the British Isles. Yet he grasped the phe-
nomenon clearly enough to apply it to many
forms of social interaction. He identifies at
least 22 conventions in a remarkable variety
of contexts.9 He obviously did not need the
matrixes of game theory to recognize iterated
co-ordination interactions and to grasp their
strategic significance.
Another sense of normative theory is as in
a theory of norms. Hume’s strategic account
of conventions is de facto an explanation of
the formation and working of many norms,
both positive and negative. Co-operativeness
is a norm in many contexts, but so is racial
discrimination. These have normative force
309
9780826443595_Ch16_Final_txt_print.indd 309
HUME’S HUMAN NATURE
only in the sense that they get us to act in
ways that are mutually advantageous, albeit
sometimes at great cost to those somehow
excluded from the norm. The force of a con-
vention that makes a norm comes from its
securing our compliance with it by making it
in our interest to comply, not from the moral
rightness of it. Hence we can count following
such a norm as individually rational, because
it eliminates any individual costs even while
it produces a collective benefit.
That a particular norm can benefit some
while harming others means, again, that the
spontaneous creation of a norm through vol-
untary actions of members of a group does
not make it normatively good. Hume does
not violate his own stricture against infer-
ring an ‘ought’ from an ‘is’, although this is
easier for him to avoid than for most of us
because he allows very few ‘oughts’ in any
case. A convention can be represented game
theoretically, but this is not at all necessary.
Indeed, Hume, the inventor of the category
and its strategic character was two centuries
before game theory and he analysed conven-
tions entirely verbally.
Within his own assumptions, Hume some-
times fails to consider that if conventions can
be unintended then they might be contrary
to the interests of those under their sway. He
says, wrongly of many cases, that conven-
tions are ‘intended’ (THN 3.2.2.16 / 494–5).
This can be true but typically only when a
government imposes a convention already
extant elsewhere that came into being there
through spontaneous, unplanned develop-
ment. For a famous example, in 1967 Sweden
switched its convention from driving left, as
in the British Isles, to driving right as in most
of Europe. Only a government could have
organized such a change.
To break a convention often requires
putting a new one in its place. This can be
2/9/2012 9:39:25 PM
 HUME’S HUMAN NATURE
done with law as it was in Sweden and is
done regularly in common law decisions and
in legislation. It can be virtually impossible
to change one spontaneously, as it surely
would have been for the Swedish driving
convention even though that convention
probably arose at least partly through spon-
taneous co-ordination. A joke in Sweden at
the time of the switch was that the conserva-
tive, unhurried Swedes would change the rule
gradually, with trucks switching to the right
one week, buses the next and cars last of all.
A massive police presence, heavy advertis-
ing and instant changing of traffic signs and
lights in the middle of the night helped make
the move remarkably safe.
Hume implicitly raises the suggestion of
functional explanation for conventions. We
need a rule for some context but there might
be many possible, different rules that might
be equally good, as in the driving convention
and in Hume’s example of the details of the
civil law. These can vary greatly even though
they might address very similar problems. As
Hume says, sometimes ‘society may require a
rule of justice in a particular case, but may not
determine any particular rule, among several,
which are all equally beneficial’ (EPM 3.31
/ 195). For example, the many societies in
the European Union have widely varied legal
systems to govern contractual relations. This
diversity of laws poses an obstacle to easier
economic relations and development. Any of
these systems would most likely be a good
one for all of Europe today. Writing a new
European body of contract law or switching
to any of the extant systems would, however,
be very difficult.
Convention arguments may be the clear-
est application of Hume’s assumption of self-
interest. I adhere to a convention not because
it is best for all but because it is best for me.
It is, of course, best for all that I adhere to
310
9780826443595_Ch16_Final_txt_print.indd 310
the local driving convention, but that is not
part of my motivation. Nevertheless, that is
what happens. It would be odd to say that
co-ordination is part of human nature. But
the motivation de facto to go along with a
co-ordination convention is part of human
nature because in this case that selfish nature
produces a collective benefit. This conver-
sion of action from self-interest into action
for collective benefit is a compelling case of
Hume’s standard insight that self-interest
can be mobilized for public good. In this
case it happens at essentially no cost but only
benefit.
5. MORALITY PSYCHOLOGIZED
It would be wrong to say that Hume is a
utilitarian in the sense of advocating that
position in moral theory. He rather presents
an explanatory psychological account of
how people obtain their moral views or, as
Rawls says, he presents an account of ‘moral-
ity psychologized’.10 This morality is part of
human nature, which is a matter for scien-
tific investigation. In essence, Hume explains
our seeming morals. He tries to understand
why we have the commitments, motivations
and beliefs that we have, he tries to explain
behaviour and actions from these, and he
attempts an explanation of the moral insti-
tutions, such as those for justice. This is a
theory of what we do rather than of what
we ought to do. Standard moral theorists
could object to this programme, saying that
it misses the normativity of morality by mak-
ing it merely a matter of scientific psycho-
logical motivation. Hume, of course, would
agree in important respects, but he would
also challenge the claim for normativity. His
central claim is simply that individuals and
2/9/2012 9:39:25 PM

institutions do commonly act for utilitar-
ian or welfarist reasons. It is part of human
nature to be welfarist, a fact that subjects
anti-welfarist moral theorists to great strain
and, often, to very artful rejections of conse-
quentialist arguments.11
The central issue in eighteenth-century
debates on moral theory is how to master
self-interest and how to get people to act
morally. Apart from Hume, all the major
writers, many of them Scots, invoke some-
thing else to override or displace self-interest:
common sense, a moral sense, morality as
deduced from reason, religious constraints
or the goodness of innate human nature.
Again, apart from Hume, all of them moral-
ize or even virtually preach. Thomas Reid of
the common-sense school says justice implies
abstaining from all injury and humanity
implies that we do all the good in our pow-
er.12 Self-interest seems likely to win the
day in this conflict of unequal psychologi-
cal forces so we are left with underpinnings
that are too weak for morality’s triumph. All
of these devices are intended to trump self-
interest with a stronger commitment to the
public good than to the self.
Given that Hume hopes to explain why we
have many of these moral views despite the
tendency to trump them with self-interest,
what can he argue? In brief, he turns self-
interest loose on these beliefs. The other the-
orists want to override self-interest; Hume
wishes to use it in his explanations of our
creation of good public institutions and of
decent behaviour by individuals. This may
be the smartest and cleverest move in moral
theory up to his time. Hume says:
if men had been endow’d with such
a strong regard for public good, they
wou’d never have restrain’d themselves
by these rules [of justice]; so that the laws
311
9780826443595_Ch16_Final_txt_print.indd 311
HUME’S HUMAN NATURE
of justice arise from natural principles
in a manner still more oblique and arti-
ficial [than in the argument from com-
monsense or moral sense]. ’Tis self-love
which is their real origin; and as the self-
love of one person is naturally contrary
to that of another, these several interested
passions are oblig’d to adjust themselves
after such a manner as to concur in
some system of conduct and behaviour.
This system, therefore, comprehend-
ing the interest of each individual, is of
course advantageous to the public; tho’
it be not intended for that purpose by the
inventors. (THN 3.2.6.6 / 529; see also
3.2.2.21 / 496–7)
This is a compelling argument for Hume’s
social scientific analysis of the individually
selfish pattern of actions that join to pro-
duce an estimable public good for all. In a
short statement this is Hume’s explanatory
theory of public institutions and of how they
produce public good out of the crude self-
ishness of individual actions. Self-interested
human nature yields this artful, surprising
result. This argument appears nowhere to be
noticed by Hume’s often dismissive critics in
his own time or in subsequent writings from
the eighteenth-century profusion of moral
systems. It is hard to suppress the thought
that they simply did not get the argument.
Instead, many of them dismissed Hume for
his supposed atheism and did not seriously
enough reflect on his writings on moral
theory. Concern with his scepticism was
the main focus of Hume scholarship until
Norman Kemp Smith took up the cause of
Hume’s naturalism in our less religious age.13
Adam Smith similarly makes sense of
economic relations as the unintended good
consequences of narrowly self-interested
individual action by the self-seeking actions
of the butcher, the baker and the rest of us.
2/9/2012 9:39:25 PM
 HUME’S HUMAN NATURE
Bernard Mandeville preceded both Hume
and Smith in applying this insight to eco-
nomic relations. In our time one might excus-
ably think that the goal of rational choice
theorists is to discover similar results that
strike us in the first instance as perverse or
paradoxical. Hume’s vision here might seem
perverse in logic, but it is compelling in prac-
tice. It settles his claims to explain our seem-
ing morals.
One of Hume’s great inventions is in vir-
tue theory. Hume divides this into two con-
texts. The natural duties are those that we
can follow individually with the confidence
that our action will produce a good result
whose goodness we immediately recognize.
For example, suppose you see a small child
floundering desperately in a shallow pond
and suppose you can immediately rescue the
child. If you do, you know with virtual cer-
tainty that you have done a good thing. The
natural duties are the focus of the bulk of vir-
tue theory, which often has difficulty dealing
with collectivities and institutions.
The other category is, under Hume’s label,
artificial duties. He defines artificial as ‘being
purposely contriv’d and directed to a cer-
tain end’ (THN 3.2.6.6 / 528–9). He later
effectively extends this definition through
examples to include creation as an unin-
tended consequence. You cannot by yourself
achieve justice in dealing with a supposed
criminal. Justice requires the creation of
institutions and roles. It is the artificial virtue
of the law as made by humans that enables
a large society to achieve social order and
justice. Hobbes thought he must explain the
rise of a state with an authoritative leader.
Hume explains the rise of the state over time
in his philosophical history of developments
from primitive society to tribal society and
on to urban civilization (THN 3.2.7 / 534–9
and 3.2.2.13–14 / 492–3). In his account
312
9780826443595_Ch16_Final_txt_print.indd 312
conventions, not dictators or contracts, gov-
ern social and political relations.
6. INSTITUTIONAL UTILITARIANISM
Hume repeatedly mentions utility, pleasure,
absence of pain, well-being, happiness and
interests and all of these have welfarist impli-
cations when he mentions them. In a title that
rightly promises a strong stance on this issue,
‘Why Utility Pleases’ (EPM 5 / 212), Hume
lays down his charge. And this response to
utility is merely human nature, not a com-
mitment to any moral theory, none of which
would be known to vast numbers of people
who readily act from utility. Hume perva-
sively claims ‘usefulness has, in general, the
strongest energy, and most entire command
over our sentiments’ (EPM 3.48 / 204). My
central defence of Hume’s project here is the
trivial empirical observation that people and
institutions commonly do in fact act on utili-
tarian or welfarist values, especially when
making collective decisions or decisions
affecting others. But Hume seems too opti-
mistic about the goodness of government. In
the contemporary US congress it would be
false to claim that the public interest is in the
forefront of many legislators’ commitments.
Normative utilitarianism has been sub-
jected to massive criticism, as is perhaps una-
voidable for the leading theory in contest, a
role that Rawls’s theory of justice has now
inherited. Most often, perhaps, it is attacked
for its consequentialism. Taken seriously,
this complaint would disable much of moral
theory in specific applications and certain
variants of moral theory entirely. It would
also block normative consideration of insti-
tutions, which hardly make sense if they are
not to have deliberately chosen consequences
2/9/2012 9:39:25 PM

and if they are not to be judged primarily by
their consequences.
The twentieth-century effort to read ear-
lier utilitarians, such as John Stuart Mill and
Hume, as really rule-utilitarians14 misses the
central point of their concern. That point is
about institutional arrangements that would
secure good consequences and not with
individual-level rule following. Mill and oth-
ers do not have the confused deontological
leanings that deontologically confused critics
assert. They are ‘institutional utilitarians’. It
is true that we cannot imagine living without
rules in our daily lives but the main body of
such rules is institutionally promulgated and
enforced.15 Following many of these rules is
in our mutual interest (again, driving conven-
tions). In any modern nation the main body
of such rules is the body of law, but there are
countless others that are specific to particu-
lar professions or organizations. Institutional
utilitarianism is about the structure of law
and other institutions, such as systems of
rights and of practices. It is a second-order
theory of how first-order choices are to be
made.
To make clear sense of this position, con-
sider a standard criticism of Hume, one that
would upset him if it were true, because it
would imply that Hume violates his own
vision to see moral and political philosophy
as a single, coherent enterprise. Barry Stroud
and others argue that Hume’s move from
moral to political theory violates his own
vision in just this way.16 Supposedly, it seems
not to be coherent in that the devices that
secure moral action do not work to secure
compliance with prescriptions of institu-
tions. As John Mackie frames the complaint,
‘does Hume show that self-interest together
with [primarily strategic] understanding is
sufficient to make each man accept the sys-
tem of justice . . . ?’ 17 Mackie goes on to
313
9780826443595_Ch16_Final_txt_print.indd 313
HUME’S HUMAN NATURE
give a compelling reason for our compliance
with the law – it can punish and its threat
of punishment will usually motivate us.18
Mackie sees Stroud’s complaint but rightly
does not think it is a problem. He says that
Hume’s ‘explanation of the artificial virtues
is essentially sociological, while his explan-
ation of the natural virtues is essentially
psychological’.19 Why? These are the natures
of the two categories of the issues Hume
addresses: sociological and psychological.
It is the real world that forces the sepa-
rate analyses of these two categories. Hume
wishes to understand these strategically
different categories because he recognizes
them as strategically different. Keeping your
promise is fully rational in the sense of being
in your interest. Adhering to a convention, a
large-number co-ordination, is also rational
in this sense. Contributing to a collective
action that is the large-number analogue of
dyadic promising commonly is not rational
in this sense, so that some freeride on the
efforts (if any) of others.20 This bifurcation of
the large-number category may be a source
of misunderstandings. In essence, when we
design institutions of justice, we include
within them devices to give strong incentives
for compliance that, without the incentives,
would fail. Such incentives are de facto built
into promise-keeping, but they must com-
monly be added on to the incentives for con-
tributing to a collective action.
6. PROMISE-KEEPING AND JUSTICE
The issue of the difference between small-
number and large-number interactions and
of any moral distinction between them is
especially clear and acute for the compari-
son of promise-keeping and justice. Hume’s
2/9/2012 9:39:25 PM
 HUME’S HUMAN NATURE
central argument for justice is grounded in
its small-number analogue, promise-keeping,
which is one of Hume’s three fundamental
laws of nature, by which he seems to mean
that they are causal sociological laws that
we must follow if we are to prosper. They
are not analogous to inescapable physical
laws but they are also not normative. Many
moral theorists simply assume that promise-
keeping is morally required, as Kant does.
Hume says, however, that the first motive to
keeping promises is interest (THN 3.2.5.11 /
522–3). If I fail to fulfill my promise to you,
my reputation will suffer and I will have
difficulties entering into other co-operative
relations, especially with you. Hume does
not often cite others’ arguments in order to
criticize, but here he does because the point
seems to be a hard one to accept. He says:
Justice is commonly defin’d to be a con-
stant and perpetual will of giving every
one his due. In this definition ’tis sup-
posed, that there are such things as right
and property, independent of justice, and
antecedent to it; and that they wou’d
have subsisted, tho’ men had never
dreamt of practicing such a virtue (THN
3.2.6.2 / 526–7).
He thinks such claims, abstracted from expe-
rience, are wrong and of no interest. He simi-
larly disposes of the fictional consent theory
of government (THN 3.2.8.3 / 541–2). People
have never dreamt of asserting their consent
to government. Only a philosopher would
propose such an idea.
Considerations of human nature are at the
foundation of all aspects of social life. For
example, when Hume addresses justice or
law, he recurs to his views of human nature.
In our time much of the discussion of justice
takes the subjects – people – for granted and
314
9780826443595_Ch16_Final_txt_print.indd 314
does not explicitly craft the theory of just-
ice for the particular kind of people we are
or for the social and economic conditions
we face. Hume gives a richly varied set of
possibilities for the circumstances of just-
ice to show that justice would be irrelevant
under certain circumstances (EPM 3.1–21 /
183–92). He supposes that experience must
have taught earlier people that they would
benefit from greater societal interaction. It is
only from experience that we come to know
this. For example, the division of labour, or
Hume’s ‘partition of employments’, makes
us more productive. As Hume says, there are
defects of man that can be remedied by soci-
ety (THN 3.2.2.3 / 485).
Our self-interest runs against common
or collective interest. This natural partiality
affects not only our relations and our actions
but also our conception of virtue. Hence we
come to think that a transgression of our
interests is immoral even as we act from
partiality in transgressing another’s inter-
ests. Hume supposes that this bias is a fairly
systematic distortion. We therefore need a
system of justice to maintain good relations.
The justice that Hume has in mind is jus-
tice as order: maintaining certain rules of
behaviour, property ownership and social
order. He implicitly rejects the idea of natu-
ral rights as somehow time immemorial,
preceding humans and their institutions.
It is human nature and actions that deter-
mine property rights, not some authoriza-
tion from a god or from pure reason (THN
3.2.2.20 / 496). Before there is law, there
is no ownership of property and, indeed,
no property. This view unites Hume and
Hobbes in opposition to Lockean and all
religious and other justifications of public
order grounded on some notion of natu-
ral or a priori rights (THN 3.2.2.20 / 496;
3.2.6.6 / 528–9).
2/9/2012 9:39:25 PM

Promise-keeping can be fully or at least
substantially self-regulated. As quoted earlier,
Hume holds that ‘it is in our interest to keep
our promises’ (THN 3.2.5.12–13 / 523–4).
This follows because, if you and I have a suf-
ficient relationship for me even to make and
for you to rely on my promise, then I will
want to maintain good relations with you
and will therefore have strong incentive to
keep my promise. Moreover, I will want to
maintain a good reputation for reliability,
and if I default with you, I may have trouble
finding other partners for future co-opera-
tion and promising.
‘But tho’ it be possible for men to maintain
a small uncultivated society without govern-
ment, ’tis impossible they shou’d maintain a
society of any kind without justice’ (THN
3.2.8.3 / 541). In the small society, spontan-
eous conventions could lead to rules govern-
ing property and promise-keeping. On the
first formation of society interest alone can
motivate us to beneficial action. ‘But when
society has become numerous, and has
encreas’d to a tribe or nation, this interest
is more remote’ (THN 3.2.2.24 / 499). Here
Hume grasps the logic of collective action.
In the sixteenth century and earlier English
village life was on a very small scale. R.H.
Tawney observes that then most people in
England ‘have never seen more than a hun-
dred separate individuals in the course of
their whole lives, where most households live
by tilling their great-grandfather’s fields with
their great-grandfather’s plough’.21 One can
imagine that in such a close society, every-
one knows everyone else extremely well and
everyone is in a position to sanction every-
one else for any misbehaviour. Suppose no
one goes far out of line. Can anyone be called
immoral? Does morality even have meaning
in this context? The rise of moral theory must
be associated with the rise of urban society
315
9780826443595_Ch16_Final_txt_print.indd 315
HUME’S HUMAN NATURE
that goes beyond Tawney’s vision of life five
or six centuries ago.22
7. LEGAL THEORY
Although he discusses law, Hume does not
give us a legal theory. We may construct
what seems likely to have been his legal the-
ory from various discussions. Some elements
follow easily enough from human nature.
For example, in our conflicts over access to
limited goods, we suffer from the ‘jealousy
of interest, which justice supposes’, without
which we would see no point in justice (THN
3.2.2.16 / 494). We therefore want property
laws that stabilize expectations, letting all
of us keep our property unless we agree to
transfer it to another, and that let you win
one time and me perhaps the next when our
conflicts require settlement by courts. Hume
has been criticized for holding that justice is
a jealous virtue. What this seems to mean is
that law as order has arisen, perhaps from
customs, in large part because it regulates
our jealousies of interest when we do come
into conflict, as we often will because our
nature is to be partial to our own interests
even when we claim impartiality and genu-
inely believe we are impartial. We conflict,
we therefore go to court, you win, and I
can only hope a better day will come even
while I begrudge your victory today, a vic-
tory that, because of my natural partiality, I
might think was falsely won. But the stabil-
ity that comes from having a court system
to adjudicate and settle our claims when we
are in conflict is generally beneficial to all
of us. It is mutually advantageous (THN
3.2.6.6 / 528–9).
Hume’s legal theory must begin from its
being an account of the most important body
2/9/2012 9:39:25 PM
 HUME’S HUMAN NATURE
of artificial duties. It is therefore a two-stage
theory. First we design institutions and then
the individual role-holders and the subjects
of the institutions do what the institutions
require of them. The general quality of the
system is likely to be relatively utilitarian.
For example, Hume supposes that the regime
of private property is utilitarian. This does
not mean that any particular instance of
property ownership can be overridden by a
direct appeal to utility (THN 3.2.3.2 / 502).
Having such actions available would desta-
bilize property and wreck its productivity.
This result would harm everyone. An obvi-
ous fact of our lives is that we often require
help from various social institutions to
accomplish the things we wish to do and to
regulate our interactions. As a general back-
ground for virtually our entire lives and all
of our projects we need institutions to main-
tain social order so that we may go about our
activities without concern for watching our
backs at all times.
One of the most important things our
institutions can do for us is to override
indeterminacies to allow us greater predict-
ability in our actions. Here we cannot put
an institution in place and then expect it
to fulfill its purpose as though there were
an overall designer of the institution, a
designer whose plan for the institution is to
be carried out as designed. Any significant
social or governmental institution is sure to
violate such a hope. Standard institutions
evolve over time to the extent of radically
altering their structures and even their pur-
poses to some degree. At some point we
must say that some institution is grown
rather than designed, so much so that we
can say its later structure is an unintended
consequence; there is no one in control of
its development. Hume seems to suppose
that the changes along the way tend to
316
9780826443595_Ch16_Final_txt_print.indd 316
be utilitarian. It is unclear how one could
show this in general. He argues that the
institutions of justice, for example, have a
purpose, to protect our prosperity and well-
being. Without good institutions of justice,
society would be far less orderly and far
less likely to enable us to have good lives.
Hume sometimes gives the impression that
these good things will happen. We know all
too well that there can be radically unjust
institutions, as there were under the Nazis
and under Stalin. Hume’s sanguine view
here might be merely relative, as in a com-
parison of Scotland in his day and in the
harsher period of the Tudors. A causal the-
ory of the rise of institutions cannot be sure
to produce ‘good’ as opposed to ‘bad’ insti-
tutions. Hume can argue that poorer insti-
tutions that do not serve you well might
provoke you to struggle to change them for
the better for your interests and that such
changes might also benefit others, but this
is an empirical assertion that is not easy to
turn into a demonstrable general claim.
Hume’s legal theory was displaced by
one of the most antithetical possibilities:
Jeremy Bentham’s rigidly designed legal
system of positive law. Hobbes thinks there
must be a draconian sovereign who is above
the law and who has very nearly absolute
power. Hume demolishes this assumption.
He presents a nascent legal philosophy that
starts from the assumption that the system of
law or justice is an unintended consequence.
In particular, he supposes that conventions
can control our governors enough to obvi-
ate the need for a Hobbesian sovereign.
Bentham reverts to something like Hobbes’s
view with his principle that there must be
a lawgiver who is above the law. This odd
vision dominated Anglo-Saxon legal theory
until the 1960s. Its last major exponent was
H.L.A. Hart.
2/9/2012 9:39:26 PM

8. CONCLUDING REMARKS
Any system as encompassing and detailed as
the system of justice must be largely unin-
tended in any substantial modern society. For
this and other reasons, the social world must
be indeterminate to a large extent. We are
not all co-ordinated in our social organiza-
tion, and we may often come into conflict. As
quoted earlier, Hume holds that the general
system comprehending the interest of each
individual – such as, for example, in pro-
tecting a contract – is also beneficial to the
general public, although it is not intended for
that purpose by its creators (THN 3.2.6.6 /
528–9); they want to resolve their own prob-
lem, not the public’s problems.
Contemporary multiculturalists use the
fact of the social construction of human
nature to drive arguments for so-called
group or cultural rights. For example, the
Yanamamo tribe of the Amazon is com-
monly thought to have a fundamentally dif-
ferent human nature from that (or those) of
contemporary heirs of Hume’s Scotland. One
can imagine that cultural differences between
these two groups are so great as to evoke
Willard Quine’s thesis of the impossibility
of translation. The language of rights here is,
of course, meant to be normatively binding,
thus making a failure to effect and protect
cultural rights a moral failing. Hume would
reject this claim, because he would reject the
argument from normativity in the abstract.
The only rights he would credit are positive
legal rights. Natural rights, human rights and
so-called cultural rights make no sense in his
system of thought. One can, however, argue
that welfarist considerations might justify
policies on behalf of the members of such
groups. This follows from considerations of
charity or equality rather than of rights.
317
9780826443595_Ch16_Final_txt_print.indd 317
HUME’S HUMAN NATURE
Hume is psychologically utilitarian, but
he makes no claim that utilitarianism is the
true morality, but nor is any other moral
theory. Truth and moral theory do not go
together. He does not do moral theory in
a normative or prescriptive manner. It is a
category mistake to speak of the truth of
a moral theory. His is almost entirely the
voice of a social scientist, almost always
explaining, almost never justifying. When
we create laws and systems of social order
that might win moral approbation, gen-
erally we do this as an unintended conse-
quence of self-interested actions. Insofar as
we are self-interested, we can be motivated
to action by incentives. If this device works,
we do not need to be in internal conflict
between interests and other motivations,
including ostensibly moral motivations.
With his remarkable strategic sense, Hume
is able to put self-interest to work in sup-
porting social institutions, most especially
the institutions of justice.
Hume does not invoke or require moral
motivations at the individual level to
achieve good results at the collective or soci-
etal level. This might be seen as his signature
move, as his trope. It is also Mandeville’s
main argument, as in his subtitle, Private
Vices, Publick Benefits. Mandeville’s nar-
rowly economic argument is that of the
later Smith, that self-seeking leads us col-
lectively to be prosperous. Many findings
in contemporary social science imply that
individual self-interest commonly wrecks
the hope for good collective outcomes, even
though we might all favour the collective
outcome. This is the implication of the logic
of collective action and of game theory’s
prisoner’s dilemma. Hume’s argument cuts
the other way and is far more subtle and
more broadly relevant.
2/9/2012 9:39:26 PM
 HUME’S HUMAN NATURE
NOTES
1
See, for example, his experiments ‘to confirm
his explanations of love and hatred, pride and
humility’ (THN 2.2.2 / 332–47).
2
This term sometimes implies general
co-operativeness, as in an ant colony.
3
Robert J. Richards, Darwin and the
Emergence of Evolutionary Theories of Mind
and Behavior (Chicago: University of Chicago
Press, 1987), p. 541.
4
Convention but not mirroring is included in
the index of the new Oxford EPM.
5
John Laird, Hume’s Philosophy of Human
Nature (London: Methuen, 1932), p. 188.
6
For a brief, accessible account, see Greg
Miller, ‘Reflecting on Another’s Mind’, Science
(13 May 2005), pp. 945–7.
7
David K. Lewis, Convention (Cambridge,
MA: Harvard University Press, 1960); Russell
Hardin, Collective Action (Baltimore: Johns
Hopkins University Press, 1981), chaps 9–10.
8
Thomas C. Schelling, The Strategy of Conflict
(Cambridge, MA: Harvard University Press,
1960).
9
See the list in Russell Hardin, David Hume:
Moral and Political Theorist (Oxford, Oxford
University Press, 2009), p. 85.
10
John Rawls, Lectures on the History of
Moral Philosophy (Cambridge, MA: Harvard
University Press, 2000); See further, Hardin,
David Hume, pp. 32–3.
318
9780826443595_Ch16_Final_txt_print.indd 318
11
Gilbert Plumer, ‘Kant’s Neglected
Argument against Consequentialism’,
Southern Journal of Philosophy 29(4) (1991),
pp. 501–20.
12
Thomas Reid, Practical Ethics, ed. Knud
Haakonssen (Princeton: Princeton University
Press, 1990), p. 138.
13
Norman Kemp Smith, The Philosophy of
David Hume (London: Macmillan, 1941).
14
For example, see J.O. Urmson, ‘The
Interpretation of the Moral Theory of J.S. Mill’,
Philosophical Quarterly 3 (1953), pp. 33–9.
15
Hardin, Morality within the Limits of Reason
(Chicago: University of Chicago Press, 1988),
chaps 3 and 4.
16
Barry Stroud, Hume (London: Routledge &
Kegan Paul, 1977).
17
J.L. Mackie, Hume’s Moral Theory (London:
Routledge, 1980), p. 86.
18
Mackie, ibid., also canvasses less compelling
devices.
19
Ibid., p. 123.
20
Mancur Olson, Jr., The Logic of Collective
Action (Cambridge, MA: Harvard University
Press, 1965); Hardin, Collective Action.
21
Quoted in Alan Macfarlane, The Origins
of English Individualism (Cambridge:
Cambridge University Press, 1978/1979),
pp. 53–4.
22
See Russell Hardin, ‘From Bodo Ethics to
Distributive Justice’, Ethical Theory and Moral
Practice 2 (1999), pp. 399–413.
2/9/2012 9:39:26 PM
 17
HUME AND FEMINISM
Lívia Guimarães
1. INTRODUCTION
In this chapter, I intend to put on display
Hume’s depiction of gender difference, his
transgression, and finally his abolition of it.
In my view, when he thinks both within and
without the boundaries of gender, by exalt-
ing the feminine he may be said, in a way, to
present us not only with a feminist philoso-
phy, but also with a feminist utopia. My main
focus is on ‘tenderness’, a feminine excellence
according to Hume. Once freed from gender
restrictions, enlarged ‘tenderness’ benefits
personal, social and political life.
After examining Hume, the man, I proceed
to Hume, the thinker, by addressing four
different topics, which are, in my opinion,
fundamental in Hume’s philosophy, and to
which feminist theory is sensitive. They are:
circumstance, embodiment, gendering and
non-gendering. My thoughts are guided by
intuitions originally stemming from Annette
Baier’s work, and may be taken as a starting
point for further conversation.
While being aware of the many theoreti-
cal achievements of modern feminist think-
ing, and of its transformative role, I believe
Hume’s thought represents a peculiar chal-
lenge to it. When facing the philosophical
tradition of the past, at times one meets
criticism and even rejection of authors who
319
9780826443595_Ch17_Final_txt_print.indd 319
manifest gender biases in the very categories
of their systems; at other times, one meets
a disposition to engage in dialogue, albeit
usually tempered by prudent wariness. But
within Hume scholarship, neither approach
seems to be appropriate.
It is arguable that Hume does not ever
hold that reason, sense, virtue and a place in
the public sphere are a man’s birthright and
privilege. His heroes and heroines are equal
in their excellence, for excellence itself is not
gender-specific. His fascination with the psy-
chology of human passions and occasional
attempt at humour before the sometimes-
absurd circumstances of human life place
him in a position unique in its openness and
inquisitiveness, which translate to a com-
plete lack of sexism. Sensibility and sensitiv-
ity, rather than gender, command the best of
Hume’s attention. One might thus be as bold
as to say that Hume creates his own eight-
eenth-century feminist Scottish haven.
2. ANNETTE BAIER
Hence, possibly, what distances a Humean
a little from feminist theorizing is the fact
that one already finds in Hume much of
what twentieth-century authors worked
2/9/2012 9:37:02 PM
 HUME AND FEMINISM
so hard to establish. With the help of
Annette Baier’s ‘Hume, the Women’s Moral
Theorist?’, ‘Hume, the Reflective Women’s
Epistemologist?’, A Progress of Sentiments
and her extensive work on trust, this suspi-
cion gains strength.1 With beauty of style and
depth of analysis, she shows us that Hume
was or, at least, might have been a feminist.
A reflection on Hume and feminism must
by necessity start with the help of Annette
Baier.
Take, for example, her work on the con-
cept of trust, in all its inspired simplicity
and perceptiveness. As she notes, many phi-
losophers, while acknowledging that trust is
what keeps a society together, fail to notice
that trust begins in the mother or caregiver-
child relation. Thus the mighty supporter of
political society is shown to rest on early,
ordinary human experience – a beginning
soon forgotten by climbers to the heights of
philosophical, political and scientific specu-
lation. In a rare, unexpected sense, Baier
gives us proof that the personal is political.
From this same source, she derives other
core elements of Hume’s philosophy: causa-
tion, for example, she associates with the ties
among family members. From her comes the
view that Hume’s was a social and cultural
epistemology, analogous, I think, to feminist
claims for an epistemology sensitive to situ-
ation, context and point of view. And from
her comes too an incisive apology of embodi-
ment, against rarefied abstraction. One of
her strongest and most remarkable points is
that, for Hume, we are mammals, carrying in
ourselves all the possibilities and limitations
constitutive of this condition.
Intuitions coming from Annette Baier
guide the thoughts that follow below. I
should like, because it seems to me some-
how fit, to progress from Hume’s life to his
writings.
320
9780826443595_Ch17_Final_txt_print.indd 320
3. LIFE
Some of the most significant characters in
Hume’s life and (why not?) adventures are
women. He was raised by his mother, peace-
fully shared a home with his sister and had
in Margaret Irvine a housekeeper able to
learn to perfection how to cook some of his
favourite French dishes. Among his chance
encounters, biographers tell of a fishwife
who acquiesced to his call for help when he
fell into a bog on his way to the New Town
only if he submitted to reciting the Lord’s
Prayer. Good humouredly, and not without
some reason, he declared her the best theolo-
gian he ever met – a bow to feminine sagacity.
Another episode tells of the wife of a candle-
maker visiting him with the firm intention
of converting and rescuing him from infidel-
ity. The interview, it is said, ended happily
for both parties, with Hume placing a large
order of moulded candles from her husband.
In these tales, there is a feeling of easy and
unconstrained exchange between Hume and
ordinary women.
As regards gentle women, we have a legal
clue in Hume’s testament – he left to his sister
a legacy of money, of course, but of a hun-
dred books of her choice too – and books,
it is needless to say, must have been among
Hume’s most prized possessions. A woman
was, for him, naturally entitled to them. In
France and in Edinburgh, he made many
sincere women friends and correspondents,
with whom he confided, and from whom
he sought advice. When deciding on how
to act in Rousseau’s crisis, had he followed
his women friends’ counsel, he would have
been less dissatisfied with the outcome. With
young Nancy Orde, his last love, we still
find playfulness and freedom. But it is with
the Comtesse de Boufflers, the object of his
deepest love, that we discover particularly
2/9/2012 9:37:02 PM

revealing incidents. A passage in one of his
letters, I think profoundly touching and
extremely telling, says:
Among other obligations, which I owe
you, without number, you have saved
me from a total indifference towards
everything in human life. I was falling
very fast into that state of mind, and it is
perhaps worse than even the inquietudes
of the most unfortunate passion: how
much, then, is it inferior to the sweetness
of your commerce and friendship! (LDH
1.451, 244)
Lionized in France, happy in his friendships
and in his intellectual pursuits, Hume makes
in this passage the most passionate declara-
tion: ‘before meeting you,’ he says, ‘my life
was meaningless.’
Hume is the fiery lover:
You may cut me to pieces, limb by limb;
but like those pertinacious animals in
my country, I shall expire still attached
to you, and you will in vain attempt to
get free. (LDH 1.457, 247)
He is the humbled, conquered lover:
Good god, how much am I fallen from
the airs which I at first gave myself! . . .
now, I throw myself at your feet, and
give you nothing, but marks of patience
and long suffering and submission.
(LDH 1.459, 247)
Have you ever had any experience of
the situation of your mind, when we are
very angry with the person whom we
passionately love? You have, surely; can
anything be more tormenting and more
absurd . . . but I then reflected, is this the
person for whose welfare I would sacri-
fice my existence, and can I now think of
321
9780826443595_Ch17_Final_txt_print.indd 321
HUME AND FEMINISM
taking pleasure in her pain and uneasi-
ness? . . . the very sight of your handwrit-
ing, I own began the cure. (LDH 1.462,
249)
Softness, I beseech you, dear madam,
continue to like me a little, for other-
wise I shall not be able in a little time to
endure myself. (LDH 1.463, 249)
He is the lover willing to commit himself
completely to the beloved:
I shall never, I hope, be obliged to leave
the place where you dwell . . . This long
absence convinces me more fully than
ever before, that no society can make me
compensation for the loss of yours, and
that my attachment to you is not of a little
or common nature. (LDH 1.475, 257)
And he is the dreaming lover, fancying the
beloved, who is away, engaged in the same
pastoral amusements as his:
If you have been so happy, as to execute
your purpose, you are almost in the same
state as myself, and are at present wan-
dering along the banks of the same beau-
tiful river, perhaps with the same books
in your hand, a Racine, I suppose, or a
Virgil, and despise all other pleasure and
amusement. Alas. Why am I not so near
you, that I could see you for half an hour
a day, and confer with you on such sub-
jects? (LDH 1.449, 243)
Lastly, his description of the Comtesse
endows her with excellences that include, as
would be expected, graces and charms, but
also with character and understanding – his
perfect ideal:
Should I meet with one in the future, in
any time future, for, to be sure, I know of
2/9/2012 9:37:02 PM
 HUME AND FEMINISM
none such at present, who was endowed
with graces and charms beyond all expres-
sion, whose character and understanding
were equally an object of esteem, as her
person was of tenderness; I ought to fly
all company, to avoid all connexion with
her, even such as might bear the name
of friendship; and to endeavour to forget
her as soon as possible . . . I know not
if it would be prudent even to bid her
adieu; surely, it would be highly impru-
dent to receive from her any testimonies
of friendship and regard, but who, in
that situation, could have resolution to
reject them? Who would not drink up
the poison with joy and satisfaction?
(LDH 1.451, 244)
The fragments above portray a man’s sincere
and heartfelt sentiments: in a romantic mood,
in the throes of passion, in fear of loss. After
the change of affairs in the Comtesse’s life,
Hume remained her loyal friend and adviser.
On 20 August 1776 he wrote her a farewell
letter. It was one of his last letters, and the
only one of them not concerning business.
The much-maligned Comtesse, in her
turn, proved a loyal friend to him too, and
a thoughtful and caring one. It is believed
that it was she who made the effort to obtain
for him an appointment in France. After his
return to Britain, she desired him back in
France, and fancied providing him with an
environment to his liking, congenial to his
studious disposition. In the Rousseau affair,
she was adamantly against the publication
of Hume’s account, and sent Rousseau an
awfully severe letter, very disapproving of his
conduct. She was not, and was not seen by
Hume as, a stereotypical woman. When their
love cooled (did it ever?), warm tenderness,
respect and admiration remained.
After this glimpse of Hume, the man, how
to proceed to Hume, the thinker? I propose
322
9780826443595_Ch17_Final_txt_print.indd 322
by addressing the topics mentioned above:
circumstance, embodiment, gendering and
non-gendering.
4. CIRCUMSTANCE: METHOD AND
CONTENT – EMBODIMENT AND
EMBEDDEDNESS
Having reminisced about some of Hume’s
real encounters with women, it may be fit-
ting to make an initial approach to his
thought from the perspective of its treatment
of situation, context and social conditions, in
a word, from what in his vocabulary he calls
‘circumstances’. By this I mean the notion
that human judgements, beliefs, passions,
likes and dislikes depend, in a great measure,
on the particular and contingent circum-
stances of human life.
As we know, in the Introduction to the
Treatise, Hume sets himself the task of
founding the science of man, which itself is
the foundation of all the other sciences. He
aims to discover the essence of the mind –
its powers and qualities – by the method
of observation and experience. The model
is Newtonian. Some feminist thinkers tend,
however, to vilify this method, in its positiv-
istic variant, which warrants objective, nec-
essary and universal truth, and occludes the
knowing subject and her concrete situation. I
believe this is neither Hume’s science, nor his
method at all.
His science is fallibilist and probabilistic,
ambitious for no more than general princi-
ples, tinted with moderate scepticism. As for
the knowing subject, in the Introduction,
he already marks an important difference
between his and Newton’s science. In moral
philosophy, contrarily to natural philosophy,
it is not possible to carry out experiments
2/9/2012 9:37:02 PM

with premeditation and exact control of
the variables. And since the subject of the
experiment is also its object, the mere aware-
ness of this condition changes the observed
behaviour.
In addition, his method consists in a set
of diverse investigative strategies that much
surpass the summary description Hume gives
of them. To experiment, for him, is to have
a rich and varied practice. In some cases,
experiments are actually performed, while
in others they are simply imagined. Some of
them are impossible to perform effectively,
and are just inconceivable – which is in itself
instructive. Hume’s moral enquiry indiscrim-
inately welcomes contents from the most
diverse sources: the poetical and literary
traditions – the poetry of Ovid, Horace and
Homer, Euripides’ tragedies and even fables
are not lesser contributors to his science
than the works of philosophers, and among
the latter, the ancients are no less valuable
than the moderns. Further instances are fam-
ily life, historical facts, conceptual analyses,
analogies, thought experiments, counterfac-
tuals and personal memories. There is not
one unified method in the experimental sci-
ence of man. One could almost say that the
choice of its components obeys only personal
taste and manners.
What is true of the method applies to the
findings as well. Hume is not always certain
that politics, for example, admits of general
truths and can be reduced to a science. But
he knows a few things: the long and help-
less infancy of man requires the combination
of parents for the subsistence of their young;
had the conditions of human life been dif-
ferent, the practice of justice may not have
been necessary; to the general principle
affirming that human nature cannot subsist
without the association of individuals, and
that in order to co-ordinate their actions to
323
9780826443595_Ch17_Final_txt_print.indd 323
HUME AND FEMINISM
the common good, government is necessary,
Hume adds the observation that humans
nonetheless constitute many different soci-
eties, with different ends (EPM 4 / 205–11).
The examples proliferate: republics favour
the progress of the sciences, and monarchies,
of the arts (EPM 6.35 / 248–9; E 111–37,
‘Of the Rise and Progress of the Arts and
Sciences’); England, where domestic life pre-
vails, better values the useful, while France,
where social life has prevalence, better values
the agreeable (EPM 8.4 / 262).
Also part of Hume’s point of view is the
notion that things are rarely judged by their
real and intrinsic worth and merit. This prin-
ciple applies to objects (more valued if rare),
pain and sickness (which, for being common
to all, does not cause humility in the young,
but which does cause it in the elderly and
those suffering from contagious diseases),
objects of esteem (less valued if they fall short
of a degree of perfection one is used to), one’s
estimate of one’s share of happiness or mis-
ery (which is susceptible to comparison with
others), the strength and liveliness of an idea
(having an advantage over what appears in
an obscure light), the superior influence of
contiguity, and so on.
Awareness of the influence of concrete
conditions is not always a matter of intrinsic
value being obscured by circumstances. The
virtue of courage, Hume observes, is more
highly appreciated in uncultivated societies,
whereas beneficence, justice and the social
virtues stand above it in cultivated ones
(EPM 7.13–15 / 254–5). That is exactly how
it ought to be.
In the second Enquiry, Hume asks, what
is best?
When it is asked, ‘Whether a quick
or a slow apprehension be most valu-
able? Whether one, that, at first view,
2/9/2012 9:37:03 PM
 HUME AND FEMINISM
penetrates far into a subject, but can
perform nothing upon study; or a contrary
character, which must work out everything
by dint of application? Whether a clear
head or a copious invention? Whether
a profound genius or a sure judgement?
In short, what character, or peculiar turn
of understanding, is more excellent than
another?’ (EPM 6.17 / 240–1)
His reply is: ‘It is evident, that we can answer
none of these questions, without considering
which of those qualities capacitates a man
best for the world, and carries him farthest in
any undertaking.’ (ibid.) Customs and man-
ners, situations and accidents truly alter the
usefulness and merit of qualities. In Hume’s
words:
He will always be more esteemed, who
possesses those talents and accomplish-
ments, which suit his station and profes-
sion, than he whom fortune has misplaced
in the part which she has assigned him.
The private or selfish virtues are, in this
respect, more arbitrary than the public
and social. In other respects they are,
perhaps, less liable to doubt and contro-
versy. (EPM 6.20 / 241–2)
And, yet more radically, some defects, when
allied to some qualities, cease to be defects.
Conversely some qualities, in the wrong cir-
cumstances, lose their value. Here are two
examples from the History of England, where
such complex valuations and devaluations
take place in detailed analyses: constancy,
sometimes, gives weight to wrong measures
(H 2.29); avarice is a sign of largeness of
mind, when its acquisitions are intended as
the ‘instruments for attaining farther power
and grandeur’ (H 2.54). Finally, in quite a
few matters, it is impossible to aspire to a
universal standard: taste will always differ
324
9780826443595_Ch17_Final_txt_print.indd 324
on account of the different humours of par-
ticular persons, and of the different particular
manners of our age and country (E 226–49,
‘Of the Standard of Taste’).
There are exceptions, however. According to
Hume, ‘discretion, caution, enterprise, industry,
assiduity, frugality, economy, good sense, pru-
dence, discernment’ cannot ever be refused ‘the
tribute of praise and approbation’. Likewise,
‘[t]emperance, sobriety, patience, constancy,
perseverence, forethought, considerateness,
secrecy, order, insinuation, address, presence
of mind, quickness of conception, facility of
expression’ will never be denied to be ‘excel-
lences and perfections’ (EPM 6.21 / 242–3).
For him, much is relative indeed, but not all.
In taking circumstances into account,
obviously, by necessity, Hume’s thinking
must move in the midst of men, women, rul-
ers, ruled, children, youth, the elderly, town
and countryside dwellers, masters, servants,
slaves, nationals, foreigners, relatives, par-
ents, friends, lovers, etc. That means, as, once
again, already vividly pointed out by Annette
Baier, that his thinking is naturally commit-
ted to embodiment.
Women are among the main interests and
principal actors in the world that Hume
draws and that includes their roles, their
cares and their concerns. Romantic love is
the last passion viewed in Book II, Part 2 of
the Treatise. Sexual love and the love of off-
spring are innate – in Hume’s words, ‘nature
has infused into all animals’ a ‘general appe-
tite between the sexes’ (E 162). Most inter-
estingly, he does not fear women’s sexuality.
Instead, he delights in it, and even pauses to
reflect on the infelicity of impotence:
What derision and contempt, with both
sexes, attend impotence; while the unhappy
object is regarded as one deprived of so
capital a pleasure in life, and at the same
2/9/2012 9:37:03 PM

time, as disabled from communicating it to
others. Barrenness in women, being also a
species of inutility, is a reproach, but not
in the same degree: of which the reason
is very obvious, according to the present
theory. (EPM 6.27 / 245)
As Hume proceeds, he draws circles within
circles of women’s embodied and embedded
belonging: sex, love, family, nation, place,
time and humankind.
5. GENDERING: FEMININE TRAITS
If we take embodiment and sexual difference
in one hand, and social conditions or embed-
dedness in the other, we shall discover gender-
ing in Hume’s text in both hands – a text not
prey to the dichotomy of nature and nurture.
In numerous passages, there is great
emphasis on difference. Women’s main
traits are softness and tenderness. They lack
strength ‘for the extrenuous efforts of the most
abstruse philosophy’ (note that the expression
‘most abstruse philosophy’ is never rid of an
ambivalent, if not outright suspicious, status),
they are tender-hearted, to the point of being
more prone than men to sympathize (even
with thieves, if handsome); they love intrigue
and romance, they are physically inferior to
men. More often than not they appear under
a becoming light and positive appreciation.
They are Hume’s chosen students of history
(which is, after all, the true laboratory of the
science of man), rulers of taste, sovereigns of
the world of conversation.
Difference, when it exists, does not impede
equality. In an eloquent passage, Hume says:
In many nations, the female sex are
reduced to like slavery, and are rendered
325
9780826443595_Ch17_Final_txt_print.indd 325
HUME AND FEMINISM
incapable of all property, in opposition
to their lordly masters. But though the
males, when united, have in all countries
bodily force sufficient to maintain this
severe tyranny, yet such are the insinu-
ation, address, and charms of their fair
companions, that women are commonly
able to break the confederacy, and share
with the other sex in all the rights and
privileges of society. (EPM 3.19 / 191)
As we can see, the gender-specificity of
‘insinuation, address, and charms’ is not here
judgementally perceived; actually, such traits
serve well in achieving their ends. I would
risk saying that they are much preferable to
other means – they are qualities admired by
Hume, qualities that should be part of the
polite and enlightened society of his dreams.
6. CONCEPTS
Hume, probably more frequently than most
philosophers, has feminine metaphors for
core concepts of his thought. Virtue, for him,
is a woman, and a humane, gentle, benefi-
cent, affable and even, at times, frolicsome
woman:
But what philosophical truths can be more
advantageous to society, than those here
delivered, which represent virtue in all
her genuine and most engaging charms,
and makes us approach her with ease,
familiarity, and affection? The dismal
dress falls off, with which many divines,
and some philosophers, have covered
her; and nothing appears but gentleness,
humanity, beneficence, affability; nay,
even at proper intervals, play, frolic, and
gaiety. She talks not of useless austerities
and rigours, suffering and self-denial.
2/9/2012 9:37:03 PM
 HUME AND FEMINISM
She declares that her sole purpose is to
make her votaries and all mankind, dur-
ing every instant of their existence, if pos-
sible, cheerful and happy; nor does she
ever willingly part with any pleasure but
in hopes of ample compensation in some
other period of their lives. The sole trou-
ble which she demands, is that of just cal-
culation, and a steady preference of the
greater happiness. And if any austere pre-
tenders approach her, enemies to joy and
pleasure, she either rejects them as hypo-
crites and deceivers; or, if she admit them
in her train, they are ranked, however,
among the least favoured of her votaries.
(EPM 9.15 / 279–80)
Beyond metaphor, the concepts of Hume’s
philosophy truly assimilate features related
to the feminine gender; either elevated to
or merged into general principles of human
nature. Reason, sympathy, sentiment are
exemplary. Ancillary concepts are often mir-
rored on feminine qualities as well. Such is
the case of modesty, complaisance, delicacy
and tenderness. Similarly, to trust as under-
stood by Baier, they all share this earthy
and mundane origin – womanly, feminine
excellence.
Thus, regarding conceptual content, a
first lesson of Hume’s approach is the trans-
gression of the binary ascription of gender
traits. Traditional (or, should we say, clichéd)
gendering recites masculine and feminine
as even/odd, sun/moon, light/dark, dry/wet,
active/passive, mind/body, culture/nature,
reason/passion.
It happens that for Hume, in the causal
beliefs concerning matters of fact, reason
includes sentiment – the reflective impres-
sion or sentiment of necessity that arises
from customary conjunction, and which is
projected back onto the world in the idea of
necessary connection. Conversely, moral and
326
9780826443595_Ch17_Final_txt_print.indd 326
aesthetic judgements of taste are expressions
of sentiments, but sentiments that aggregate
rational, reflexive or cognitive skills and
operations.
In the Treatise, Hume points out that natu-
ral philosophers take the ‘liberty’ of consider-
ing ‘any motion as compounded and consisting
of two parts separate from each other, tho’ at
the same time they acknowledge it to be in
itself uncompounded and inseparable’ (THN,
3.2.2.14 / 493). If Hume’s observation is true,
the distinction between the passions and the
understanding is reduced to an expository
device in the study of moral philosophy. Thus,
however much a description may make use of
the instrumental distinction between reason
and sentiment, a truer account of the same
phenomena implies the ultimate dismissal of
such a distinction and, of the two categories,
the latter, i.e. sentiment, seems to be the more
fundamental.
In addition to transgressive fusions,
such as the one of sentiment and reason
cited above, we find in Hume’s thinking, as
already mentioned, the expansion of restrict-
ive, feminine-gendered concepts into larger
spheres of meaning and dominance. Hume
says of delicacy of taste: ‘The very sensibility
to these beauties, or a delicacy of taste, is
itself a beauty in any character; as convey-
ing the purest, the most durable, and most
innocent of all enjoyments’ (EPM 7.28 /
260). Modesty, in its turn, begins in feminine,
chaste behaviour, but develops into the social
virtue that counterbalances pride (EPM 8.8
/ 263–4, THN 3.3.2.10 / 587–8) and into
a philosophical virtue that counterbalances
dogmatism.
Thus the second lesson in content, from
a gendered perspective, is that if one gen-
der’s qualities are to prevail, they are going
to be the feminine ones. There is no barrier
preventing the free movement of women
2/9/2012 9:37:03 PM

between the domestic and public spheres.
And, as hinted above, feminine virtues are
especially appropriate to civilized times. It
is certain that Hume values social civil vir-
tues above the virile virtues of a warrior. In
politics, Hume recommends mildness and
moderation. Hence manners enter politics.
In society, Hume argues that the presence
of women and the free intercourse between
the sexes enables livelier, more polished,
and refined relations. Finally, at the per-
sonal level, the aim to be sought, according
to him, is softened tempers and cultivated
minds.
In a slightly modified paraphrase, we may
say that Hume’s is the age of sentiment, and
his is the sentimental nation – of refined,
reflective, elaborated, sophisticated, simple
and natural sentiment; of sentiment assimi-
lated into morals, aesthetics and knowledge.
Therefore, when there is gender difference at
work in Hume’s philosophy, it works mostly
in favour of women, conferring them great
eminence. Not, observe, in a casual way, but
in a way very much sustained by the rigour
of Hume’s analysis and enquiry.
7. UNGENDERING
From another perspective, and that would
be Hume’s third lesson, gender categoriza-
tion ends up unnecessary. In one sense, what
I mean here is that in many passages Hume
is keener on resemblance than on difference.
In the families he conceives, there is no dis-
tinction of property. His women, just like
his men, are fond of power and, if frustrated
by too strict and domineering a husband,
become household tyrants. In love, they are
to be companions in a relation of equality,
where sexual passion, friendship, kindness
327
9780826443595_Ch17_Final_txt_print.indd 327
HUME AND FEMINISM
and tenderness are cemented by mutual
interests, and by interest in the well-being of
the offspring. Hume says:
Love between the sexes begets a com-
placency and good-will, very distinct
from the gratification of an appetite.
Tenderness to their offspring, in all sen-
sible beings, is commonly able alone to
counter-balance the strongest motives of
self-love, and has no manner of depend-
ence on that affection. What interest can
a fond mother have in view, who loses
her health by assiduous attendance on
her sick child, and afterwards languishes
and dies of grief, when freed, by its
death, from the slavery of that attend-
ance? (EPM App. 2.9 / 300)
Women are free and able to rule outside
the home also. And, for Hume, denying them
this right is clear evidence of barbarism:
But though positive law seems wanting
among the French for the exclusion of
females, the practice had taken place;
and the rule was established beyond
controversy on some ancient as well as
some modern precedents. During the
first race of the monarchy, the Franks
were so rude and barbarous a people,
that they were incapable of submitting
to a female reign; and in that period
of their history there were frequent
instances of kings advanced to roy-
alty in prejudice of females, who were
related to the crown by nearer degrees
of consanguinity. These precedents,
joined to like causes, had also estab-
lished the male succession in the second
race; and though the instances were nei-
ther so frequent nor so certain during
that period, the principle of excluding
the female line seems still to have pre-
vailed, and to have directed the conduct
of the nation. (H 2.197)
2/9/2012 9:37:03 PM
 HUME AND FEMINISM
In the same vein, when Hume treats of the
moral sublime, which consists in greatness
of mind, spirit and dignity of character, he
places side-by-side Alexander, Ajax, Vitellius
and a woman, Medea – whose magnanimity
consists in self-reliance (EPM 7.7 / 252–3).
This sort of ungendering, by deconstruct-
ing dichotomies, is especially visible in the
History of England. Women, no less than
men, are protagonists of the history. And the
actors’ qualities do not usually present them-
selves with gender labels. As many kings are
led astray by their too warm affections and
soft tempers, as by impulsive and violent
tempers; as many by their immoderate love
of glory, as by their weakness and indecision;
as many by their gentle, humane and merci-
ful spirits, as by levity and fickleness. Female
valour can show itself in the field of war and
in the cabinet, as much as in kneeling and
pleading for a good cause. A personage can
at once possess courage and grace, vigilance
and affableness. Thus Hume erases the pre-
sumed gender-specificity of particular pas-
sions and dispositions.
In another sense, ungendering alludes to
the fact that Hume seeks and finds in his
science general knowledge of humankind.
Above and beyond the differences, he aspires
to viewing a commonality inclusive of all
humans, extensive to all sentient beings. And
in yet another sense, which I intend to sug-
gest only later on, at the end of this essay, his
embracing and exalting the feminine some-
how marks the abolition of gender.
For Hume, all humans agree over some
general preferences and distinctions. Without
them, language would not be comprehensi-
ble, translation would be impossible. Hume
admits we enter more readily into sentiments
that resemble our own. But this is not, for
him, an unsurpassable obstacle. In his words:
‘no passion, when well represented, can be
328
9780826443595_Ch17_Final_txt_print.indd 328
entirely indifferent to us; because there is
none, of which every man has not, within
him, at least the seeds and first principles’
(EPM 5.30 / 222).
We all have a propensity to sympathy,
even if in a slight degree (e.g. the miser, EPM
6.3n26 / 234–5n1). By sympathy we are sus-
ceptible to the emotions of other persons;
sympathy extends our concern beyond our
immediate circle of relations; it causes benev-
olence, compassion and other social virtues;
and, therefore, it is a cause of our very exist-
ence in society, since the source of social rela-
tions resides in shared passions. Sympathy
creates society by establishing affective ties,
standards of behaviour, common concepts
and experiences among its members. A natu-
ral operation of the human mind, it is also a
necessary condition of morals. Moral senti-
ments of approbation depend on sympathy
exempt from fluctuations due to proximity
and distance. And that we attain by assuming
general points of view, from which we take
positions imaginatively close to an individual
and those around her, so as to feel by sympa-
thy the effects of her actions over them. The
pleasure she causes induces our approba-
tion, and the pain induces disapprobation. In
our moral distinctions, we reach a common
standard by the intercourse of sentiments –
and this intercourse is, in principle, ungen-
dered. In short, morality ‘depends on some
internal sense or feeling, which nature has
made universal in the whole species’ (EPM
1.9 / 173).
8. UTOPIAS
I began this essay recollecting the contri-
butions of modern feminist thinking. One
curious effect of that movement has been
2/9/2012 9:37:03 PM

the production of utopian and dystopian
novels. Interestingly, this genre dates back at
least to the early moderns. In early modern-
ity, proto-feminists would contribute to the
presses either with treatises on the education
of women or with utopian fictions. Two
examples that come readily to mind are
Margaret Cavendish, of course, and Sarah
Fielding.
When I think of Sarah Fielding, first I
remember The Governess; or, Little Female
Academy, a little pastoral where young
pupils in their conversations, by each telling
a fairy tale or her own history, and listening
to the others, progress in the path of virtue,
teach and are taught goodness and the right
ruling of the passions. In this small world
of theirs, they are in the process of building
characters exemplary to all women. In The
Adventures of David Simple, and its sequel,
Fielding contemplates a more ambitious
plan: a utopian society in which the ties
of love and friendship bring together two
couples (David and Camilla, Cynthia and
Valentine), one of their parents, and their
children. All is shared in this happy society,
ruled by trust, tenderness and selflessness.
The small circle is incorruptible. Its princi-
ples remain upstanding throughout many
trials (such as poverty, illness, separation)
before its ultimate downfall. It ends cruelly
shattered by the outside world, which is full
of envy, treachery, jealousy, callous indiffer-
ence, and the cost is high: the life of almost
everyone. Fielding creates a utopian niche
within and under the pressure of a dys-
topian larger society. There is real suffering
in David Simple, even tragedy. But Fielding’s
is a moral narrative, which also envisages a
society, still to come, where soft, feminine
virtues rule over men and women alike.2
When I think of Margaret Cavendish,
I remember that not content with a room
329
9780826443595_Ch17_Final_txt_print.indd 329
HUME AND FEMINISM
of her own she creates The Blazing World.
In this fantastic tale, a woman transported
from this to another world is taken to the
emperor, who marries her and gives her
‘absolute power to rule and govern that
world as she please[s]’.3 She reveals her-
self to be a blazing empress. Of a ‘generous
spirit and ready wit’, she governs wisely,
commands scientific investigations into all
matters, and is the supreme judge of their
results, which sometimes appear inconclu-
sive, or contradictory, and always bizarre.
She extinguishes an entire field of research in
mathematics, delivers orators from grandilo-
quent but empty speeches, and confines logi-
cians to very narrow bounds. She fluently
debates theology with spiritual beings and,
when her old world is under threat, she sets
sail and defeats its enemies in naval combat.
At the end of the book, having befriended
Cavendish, she tells her that she too can rule
absolute over a world of her creation. This
bold and spirited fantasy trespasses or, bet-
ter yet, subverts all the boundaries of gender.
The result is amazement, bewilderment and,
yes, great amusement for the reader.
I have no evidence that Hume read any
of these authors. He probably did not. But
if we consider his literary likes and dislikes,
it is certain that he would have a hearty
laugh, and one or two mischievous ideas, at
the fantastic tale of Cavendish, and would
be touched by the softness of Fielding’s.
The former would possibly remind him of
the not so modest, but intriguing women
who had crossed paths with him. The lat-
ter’s imaginary world is similar to his own,
as it must have appeared to him, both in his
confident, and in his diffident and sceptical
moods.
Now Hume does not speak of man as
the universal human. He does not single
out the faculty of reason as the essential
2/9/2012 9:37:03 PM
 HUME AND FEMINISM
and distinctive human quality. He does not
attribute a subordinate role to women. He
is far above the demeaning stereotypes that
so worried twentieth-century feminists.
He embraces diversity and rejects dualism.
When he keeps himself within the limits of
gender difference, the beneficiary is the femi-
nine gender, with its more humane qualities.
When he transgresses these limits, by consid-
ering an impression of reflection, or senti-
ment, to be the determinant of knowledge,
morals and aesthetics, and hence turning rea-
son, so to speak, into a ‘slave of the passions’,
once again he exalts the feminine gender. And
when he does away with gender altogether,
speaks of and to the whole of humankind,
discovering similar features in each and all,
we may say that he puts the accent on the
feminine, softer and more tender affections.
Moral approbation is a tender sympathy for
others, a generous concern with those of our
kind and species. It is a species of love and
esteem. The first of social virtues, benevo-
lence, Hume describes as humane and tender.
According to the second Enquiry, the social
virtues bring order to society, happiness to
family and mankind, mutual support among
friends and a gentle dominion over the hearts
of men.
Therefore I should like, in this concluding
paragraph, to add another member to the
list of early feminist utopians – Hume. His
utopia, as already insinuated: tenderness. Its
main source? Passages on justice where he
contemplates scenarios that would make its
practice unnecessary; if nature were provi-
dent and plentiful, there would be no need
to distinguish between ‘mine and thyne’. Or
else, if human nature were only tenderness
and concern for others, there would be no
need of the distinction either. Outside of uto-
pia, as the example of benevolence shows,
tenderness is everywhere that matters for
330
9780826443595_Ch17_Final_txt_print.indd 330
Hume. It is implicit even in the founding
principles of his philosophy, for one may well
raise the question: how can one be receptive
of impressions and ideas and responsive to
them if not endowed with tenderness? 4
NOTES
1
A. Baier, ‘Hume, the Women’s Moral Theorist?’,
in Moral Prejudices (Cambridge, MA: Harvard
University Press, 1994), pp. 51–75; ‘Hume, the
Reflective Women’s Epistemologist?’, in Moral
Prejudices, pp. 76–94; A Progress of Sentiments
(Cambridge, MA: Harvard University Press,
1991).
2
Sarah Fielding, The Governess; or, Little Female
Academy (Lenox, MA: Hard Press); The
Adventures of David Simple and Volume the
Last, ed. Peter Sabor (Lexington: The University
Press of Kentucky, 1998).
3
Margaret Cavendish, The Blazing World &
Other Writings, ed. Kate Lilley (London:
Penguin Books, 1994), p. 132.
4
I am very grateful to Dan O’Brien, and to
Alan Bailey and Continuum Press as well,
for the kind invitation to participate in this
volume – a gift of pure delight. This paper
was first given to the 37th International Hume
Conference in Antwerp, Belgium, 6–10 July
2010, in a panel on ‘Hume and Feminist
Philosophy’, to which I was invited by
Jacqueline Taylor. For this unique opportunity
of amiable dialogue, I am especially grateful
to her. I am also immensely grateful to Amy
Schmitter and Anne Jaap Jacobson (my
generous co-panelists), to Donald Ainslie
and Willem Lemmens (Conference Directors,
together with Jacqueline Taylor), Petra Van
Brabandt, Catharina Paxman and Livia
Verbrugge (the local Organizing Committee),
and to all participants in the event. The paper
was given to the seminar ‘Fellows’ Hume
Conversations’ at the IASH (The Institute
for Advanced Studies in the Humanities)
at The University of Edinburgh. For their
congenial fellowship, I am grateful to Susan
Manning, Pauline Phemister, Anthea Taylor,
Donald Ferguson (director, deputy director,
2/9/2012 9:37:04 PM

administrator, secretary), fellow seminarian Jo
Clifford, and the whole supportive community
at the Institute. As ever, I owe a debt of
gratitude to Annette Baier and also to Don
Garrett, who first brought to my attention
Hume’s concept of tenderness. For kindly
helping me with a critical, yet complaisant
331
9780826443595_Ch17_Final_txt_print.indd 331
HUME AND FEMINISM
reading of the text, I am grateful to Gabriela
Guimarães Gazzinelli. The paper is part of
a research project sponsored by Conselho
Nacional para o Desenvolvimento Científico
e Tecnológico (CNPq) and Fundação de
Amparo à Pesquisa do Estado de Minas Gerais
(FAPEMIG), Brazil.
2/9/2012 9:37:04 PM
 18
HUME ON ECONOMIC WELL-BEING
Margaret Schabas
‘The great end of all human industry
is the attainment of happiness. For this
were arts invented, sciences cultivated,
laws ordained, and societies modelled . . . .
Even the lonely savage . . . forgets not,
for a moment, this grand object of his
being’ (E 148).
Although David Hume’s Political Discourses
(1752) was the first of his works to win
immediate recognition, Hume scholars have
for the most part sidestepped his economic
thought.1 Yet a perusal of Hume’s corre-
spondence indicates a sustained engage-
ment with economic debates on money and
banking, public finance and economic devel-
opment, and not only with the first tier of
economist-theorists of his day, Adam Smith,
James Steuart and A.R.J. Turgot, but also
with the second tier, Abbé Morellet, Josiah
Tucker and Isaac de Pinto, to name just a
few. Hume was true to his proclamation of
pursuing ‘a mixed kind of life’, one that has
‘direct relevance to action and society’ and
at several key junctures in his life, he delved
into the world of commerce (EHU 1.25 / 7).
In 1734 he apprenticed briefly with a Bristol
sugar merchant and in 1737–9, while in
London, his primary social circle was com-
posed of displaced Scots there on government
business, including Archibald Campbell,
332
9780826443595_Ch18_Final_txt_print.indd 332
founder of the Royal Bank of Scotland in
1727 and Scotland’s main improver up until
his death in 1761. From 1746 to 1748, while
serving as personal secretary to General St
Clair, Hume kept careful accounts of the
army’s financial operations and made par-
ticular note of the range of fiscal policies that
he witnessed on their European tour. Upon
returning to Scotland, Hume forged close
ties with the leading bankers and merchants
of Edinburgh, for example John Coutts and
Adam Fairholm.2 In 1765, during his term
in Paris as British chargé d’affaires, he dis-
mantled the anomalous paper currency of
Lower Canada, something that had intrigued
him since the 1740s.3 In 1767, while Under-
Secretary for the Northern Department in
London, his quotidian duties included the
settlement of contracts with organizations
such as the East India Company.4
More substantially, consider Hume’s
scholarly record. His Treatise of Human
Nature (1739–40) offers many contributions
to economic theory: the promotion of the
scientific standing of the moral sciences in
Book One, the insights about human moti-
vation and rationality in Book Two, and the
analyses of contracts, money and property in
Book Three.5 There are also many important
insights on economics in Hume’s earlier set
of essays (1741–2), especially ‘That Politics
2/9/2012 9:37:10 PM

May Be Reduced to a Science’, ‘The Rise and
Progress of the Arts and Sciences’, ‘Of the
Standard of Taste’, ‘Of National Characters’,
and his quartet on Greek and Roman phi-
losophers, the so-called ‘Happiness Essays’.
Hume’s two Enquiries (1748; 1751) offer
several mature reflections on economic top-
ics, such as on the nomotheticity of human
behaviour and on the institutional arrange-
ments for a system of property. Hume’s long
narrative celebrating commercial develop-
ment over barbarian ways in his History of
England (1754–62) is a seminal contribution
to economic history, both for its empirical
details and its command of the long durée.6
Even the Natural History of Religion (1757)
and the posthumous Dialogues concerning
Natural Religion (1779) are congruent with
Hume’s engagement with the science of com-
merce. They advance similar methodologies,
the treatment of social facts as robust pat-
terns and a proto-Bayesian appeal to eviden-
tial support, and they are all in the service of
promoting secularism.7
Nevertheless, it is Hume’s Political
Discourses that established him as a major
voice in economics, then and now. Several
of the essays explicitly address the key phe-
nomena of commerce and trade, money, the
interest rate, public finance, population and
economic growth. By the time Hume died he
had overseen ten editions of his essays (the
eleventh appeared in 1777) and he knew of
at least ten more in several other European
languages, including six in French.8 Hume
left his mark on his contemporaries, François
Quesnay (1758), Turgot (1769) and Smith
(1776), on the American Federalists, as well
as on the two most prominent economists of
the early nineteenth century, Jean-Baptiste
Say (1803) and David Ricardo (1817).9
Among twentieth-century economists,
Hume’s legacy is synonymous with his
333
9780826443595_Ch18_Final_txt_print.indd 333
HUME ON ECONOMIC WELL-BEING
monetary theory, both the specie-flow mech-
anism that commits to a global equilibrium
of money and prices, and the insight that
an unanticipated and localized surge in the
money supply will stimulate genuine eco-
nomic growth. When Milton Friedman was
asked in 1975 to assess the past 25 years of
monetary theory, he responded by question-
ing whether we had learned anything in the
200 years since Hume. Very little, he replied.
‘We now have a more secure grasp on the
quantitative magnitudes involved; second we
have gone one derivative beyond Hume.’ 10
While something of an oversimplification,
Hume’s monetary theory put the capstone on
a century or more of debate from Jean Bodin
to John Law, and thereby relegated money
to a backseat in mainstream economic dis-
course until the early twentieth century.
The scientific discourse of economics, par-
ticularly analyses of money, trade, and usury,
reaches back to Xenophon and Aristotle, but
was only systematized into specific theories
in the seventeenth century with the forma-
tion of several schools, notably the mercantil-
ists, cameralists, and political arithmeticians.
Hume may have read some of this literature
over the course of his two-year sojourn in
France, where he had access to several good
collections, including the Jesuit’s library of
some 40,000 volumes at La Flèche.11 Hume
has traditionally been portrayed as aligned
with ‘liberal economics’, in the sense of pro-
moting freer trade and the individual right
to the pursuit of wealth.12 Although this
anti-mercantilist movement commenced in
the 1690s with the writings of Josiah Child,
John Locke, Dudley North, and Pierre de
Boisguilbert, there was still much work to be
done, by Hume and by others, to convince
policy-makers to dismantle barriers to trade.
In a letter to James Oswald, 1 November
1750, Hume wrote that ‘to prohibit the
2/9/2012 9:37:10 PM
 HUME ON ECONOMIC WELL-BEING
exportation of money, or the importation
of commodities, is mistaken policy; and I
have the pleasure of seeing you agree with
me’ (LDH 1.144, 67). Appeals to the liber-
alization of trade pepper Hume’s essays of
1752, but discerning that the argument for
the gains from international trade had not
been made sufficiently clear, Hume issued
a subsequent short essay, in 1757, entitled
‘Of the Jealousy of Trade’, that concludes
with a prayer for the flourishing commerce
of Germany, Spain, Italy and even Britain’s
arch-rival, France (E 331).
By the eighteenth century there was a ver-
itable explosion of publications in economics.
In France alone, there were over 3,000 titles
issued by over 1,500 different authors in the
80 or so years leading up to the Revolution
of 1789.13 Some of the more prominent
groups were the neo-Colbertists such as Jean-
François Melon, the circle formed around
J.-C.-M. Vincent de Gournay and, most fam-
ously, the physiocrats founded by Quesnay
and Victor Mirabeau in 1757. To this list
could be added the writings of Charles-Louis
de Montesquieu, Jean-Jacques Rousseau,
Claude Adrien Helvetius, Etienne Bonnot de
Condillac, Turgot and many more. Hume’s
theoretical reach has been most aptly com-
pared to the inspirational text by Richard
Cantillon, whose Essai sur la nature du com-
merce en générale survived his mysterious
death in London in 1734 and was brought
into print in 1755, most probably by the lead-
ing member of the Gournay circle, François
Véron de Forbannais. Alas, there is no con-
crete evidence that Hume knew the draft of
the Essai that circulated in the interim but
there is some likelihood that he did read it,
either while living in Reims in 1734–5 and
using the library of Noël-Antoine Pluche
(who also wrote on economics) or while in
London shepherding his Treatise into print.14
334
9780826443595_Ch18_Final_txt_print.indd 334
Hume reacted directly to the ideas of
Melon, Montesquieu and Charles de Ferrère
Du Tot, and his initial translators into
French effectively wedded him with the
Gournay circle.15 He also drew inspiration
from the British writers of his day, John Law,
Bernard Mandeville, Isaac Gervaise, Jacob
Vanderlint, Joshua Gee, Joseph Massie and
Josiah Tucker as well as the proto-sociolog-
ical accounts by Daniel Defoe and Jonathan
Swift. George Berkeley’s delightful economic
rant, the Querist (1735–7), and the early eco-
nomic writings by Benjamin Franklin, both
of which endorsed the shift to a paper cur-
rency, may also have figured in Hume’s intel-
lectual genesis. A recent study has identified
some striking parallels between Hume and
the Italian economists who were active in
the 1750s, Ferdinando Galiano and Antonio
Genovesi, and argues that the flourishing of
economic discourse in the relative backwa-
ters of Scotland and Naples was due in part
to the keen sense of underdevelopment with
respect to the ruling nations of England and
Spain respectively.16
One of the first works that Quesnay read
when he decided at age 63 to take up eco-
nomics was Hume’s Political Discourses. But
the favourable regard was not returned. In
a 1769 letter to Morellet, Hume denounced
the physiocrats as ‘the set of men the most
chimerical and most arrogant that now
exist’ (LDH 2.205, 431). Hume disliked
their doctrine of a single tax and privileg-
ing of the agrarian sector and expressed the
hope that Morellet’s Dictionnaire du com-
merce will ‘crush them [the physiocrats], and
pound them, and reduce them to dust and
ashes’ (ibid.). An immense aura surrounded
Quesnay, even though he wrote very little
apart from a few encyclopedia entries and
a brief exposition for his celebrated tableau
économique.17 His many disciples, however,
2/9/2012 9:37:11 PM

wrote books and established four periodicals
and in that respect, the physiocrats were the
dominant school of the 1760s and 1770s
with a reach as far as Russia and America.
Because their ideology was so closely tied to
the ancien régime, however, their ideas held
little sway in France after 1789, and had
already faded considerably with Quesnay’s
death in 1774.18 If one looks to the rapid and
extensive circulation of Hume’s economic
essays and their dominance of the field from
1752 right up until the 1790s, it is reasonable
to view Hume as of comparable prominence
during the second half of the eighteenth cen-
tury. He would be overshadowed by Adam
Smith, to be sure, but not until the 1790s,
as Hume himself predicted on reading Adam
Smith’s hefty tome.19
The long friendship between Hume and
Smith is remarkable for the potential rifts
along the way, not least Smith’s failure to sup-
port Hume’s candidacy at the University of
Glasgow in 1752. There is reason to believe
that their friendship was most fervent in the
early 1750s, while living near one another
in Scotland, and there is evidence that their
friendship transmuted quite profoundly first
in the wake of Hume’s criticisms of Smith’s
Theory of Moral Sentiments (1759) and then
with the Rousseau debacle of 1766, and may
never have been fully repaired. Unfortunately,
towards the end of their respective lives,
Hume and Smith burned some of their man-
uscripts and, in the latter case, it seems, most
of his letters as well. What little we have to
go by suggests that there were many disa-
greements and that over time Smith became
more guarded, holding back drafts of the
Wealth of Nations despite Hume’s intermit-
tent requests to read them. Smith wrote a
moving tribute on the occasion of Hume’s
death, praising his courage and equanimity
as he faced the end and emphasizing Hume’s
335
9780826443595_Ch18_Final_txt_print.indd 335
HUME ON ECONOMIC WELL-BEING
mission to combat superstition. Smith’s trib-
ute feebly masks his residual guilt at having
reneged on his more than 20-year commit-
ment as Hume’s literary executor, motivated
by the worry that the publication of Hume’s
Dialogues would undercut his reputation.20
From student lecture notes, we know that
Smith had developed many of the core ideas
for his Wealth of Nations in the early 1750s.
From Hume’s periodic promptings, we also
know that Smith had delayed publishing
his book for up to 20 years. In short, Smith
played Brahms to Hume’s Beethoven.21
David Hume was part of a chorus of
economists in the eighteenth century who
recognized the dramatic rise of wealth in
Western Europe since the sixteenth century.
Hume is particularly astute for the pains he
took to document this with careful empirical
research and proto-econometric methods
and, most brilliantly, to locate the source of
this in both the division of labour and the
influx of silver from the New World. Prices,
he argued, had come down significantly in
real terms. For while nominal prices had risen
by a factor close to four, the quantity of new
specie (gold and silver coins) had risen by at
least a factor of seven. Even then, metal coins
accounted for only one-third of commercial
transactions; the rest were serviced by fidu-
ciary notes of one kind or another (E 320).
This significant discrepancy, Hume asserted,
could only be explained by economic growth,
as the increased stock of money spread itself
over more goods and services rather than
result in inflation. It is a brilliant argument,
analogous to the one advanced by William
Harvey a century before to demonstrate the
circulation of the blood. Both had to appeal
to very imprecise figures for the production
and velocity of blood or money, but could
then forcefully derive the desired conclu-
sion because of the significant differential
2/9/2012 9:37:11 PM
 HUME ON ECONOMIC WELL-BEING
between the two variables (in Hume’s case,
inflation and money supply). Hume also
marshalled data on French grain prices that
showed quite convincingly that while the
nominal price had stayed relatively constant,
the real price had fallen by about 40 per cent
since the 1680s.22 Indeed, Hume was one
of the first to emphasize that the price level
of many key goods brought to urban mar-
kets had fallen over the past few decades, as
wholesale methods developed and brought
about price uniformity.23
In his essay ‘Of Commerce’, Hume
viewed foreign trade and manufacturing as
the motive force for the economic growth
of Western Europe. Agricultural develop-
ment came in the wake of manufactur-
ing and international trade, since the latter
alone provided the incentives for peasants to
generate a surplus from the land (E 260–1;
419–20). This was well illustrated by Spain,
where considerable tracts of land lay barren.
Although early modern Spain had imported
vast quantities of silver from South America,
the death penalty that governed the export
of money severely limited foreign trade and
brought about economic ruin, especially in
the agrarian sector (E 312). Indeed, the rapid
demise of Spain served as the paradigm
among Enlightenment economists for mis-
guided policies.
A burgeoning population was also a criti-
cal component of the growth account and
Hume took great pains to argue that the
population of Western Europe had surpassed
that of ancient Rome. His lengthy essay ‘Of
the Populousness of Ancient Nations’ is a
gem of systematic inductive reasoning.24
Here too he lacked concrete and precise data
(the first British census was in 1801), and
he had to resort to qualitative comparisons,
drawing on his knowledge of classical texts
as well as parish records that the political
336
9780826443595_Ch18_Final_txt_print.indd 336
arithmeticians had marshalled. Insofar as
marriage was restricted for Roman slaves,
and insofar as slavery itself was less produc-
tive than wage labour, Hume motivated the
conclusion that contemporary population
must be larger on both legal and economic
grounds. A comparison of standards of liv-
ing in ancient Rome and modern Europe also
demonstrated that the latter is considerably
better off and thus likely to sustain a higher
population as well (E 383–401).
Hume thus advanced the link between
the recent expansion of national wealth and
population in Western Europe, and while he
was not unique in making these claims, his
specific arguments were without precedent in
terms of their force and abductive skill. Also
commonplace among eighteenth-century
savants was the belief that these phenomena
were conjoined with well-being or happi-
ness.25 ‘If everything else be equal, it seems
natural to expect, that, wherever there are
most happiness and virtue, and the wisest
institutions, there will also be most people’
(E 382). In the opening pages of his Political
Discourses, Hume observes that ‘the great-
ness of a state, and the happiness of its sub-
jects, how independent soever they may be
supposed in some respects, are commonly
allowed to be inseparable with regard to
commerce’ (E 255). But he then notes that
since ‘Man is a very variable being’, this
maxim ‘may possibly admit of exceptions,
and that we often establish it with too little
reserve and limitation’ (E 255).
It is in this space of ‘exceptions’ that I pro-
ceed here to unpack his concept of economic
well-being. Almost nothing is straightfor-
ward with Hume, especially given his his-
torical sensibility, and there are reasons to
believe that he favoured the Stoic disjunctive
between wealth and happiness, albeit with
some qualifications. As he remarks in his
2/9/2012 9:37:11 PM

essay ‘Of the Delicacy of Taste and Passion’,
it is impossible to ‘render happiness entirely
independent of every thing external’, but that
‘every wise man will endeavour to place his
happiness on such objects chiefly as depend
upon himself, . . . and receives more enjoy-
ment from a poem or a piece of reasoning
than the most expensive luxury can afford’
(E 5). In this respect, the positioning of
individual mental well-being over material
well-being, Hume and Smith had more in
common. They were also sceptical about col-
lective well-being in the long term, and alerted
their readers to the many respects in which
the commercial era prolonged our myopia to
full human flourishing. In this respect, they
shared kinship with John Maynard Keynes,
who also admired the wealth-creating effects
of the capitalist system while deploring its
dehumanizing tendencies.
Both Hume and Smith set foot on the
hedonic highway that stretched from
Shaftesbury to Bentham and on to Mill and
Sidgwick, although to what variety of utili-
tarianism they subscribed, respectively, is
much more contested.26 Their efforts to trade
pleasures and pains also addressed many non-
pecuniary goods, such as friendship, civility,
security, learning or spirituality.27 Smith puts
much weight on the value of a good night’s
sleep which he deemed inversely correlated
with wealth, while Hume opines in his auto-
biographical essay ‘My Own Life’ that he
would not trade his disposition towards opti-
mism for an estate of 10,000 [pounds] a year
(E xxxvi). Both put emphasis on self-com-
mand and equanimity but note just how diffi-
cult this is to achieve given the high likelihood
of some misfortune in the course of a long life
(E 4). Hume remains sunnier than Smith, to be
sure, but clouds are forming on the horizon.
Adam Smith embraced a dark and dyspep-
tic vision of the human condition. As humans
337
9780826443595_Ch18_Final_txt_print.indd 337
HUME ON ECONOMIC WELL-BEING
advanced beyond the first two stages of
hunter-gatherers then nomads, and became
cultivators of the land, the additional wealth
only increased the inequality among people.28
There was a fundamental injustice between
those who labour and those who reap where
they do not sow, motivated by avarice and
vanity. The commercial era brought even
greater inequalities, since manufacturers
always have the means to outlive labour dis-
putes and generally conspire to keep wages
down. ‘Rent and profit eat up wages and the
two superior orders of the people oppress the
inferior one’ (emphasis added).29 But even
those in the superior position took more
pleasure from ‘the parade of riches’ than
from the consumption itself.30 The world is
driven by vanity, greed and a hefty dose of
self-deception, motivated solely by the desire
to gain the approval of others. No one, not
even the wealthy, is protected from the afflic-
tions of sorrow, disease, danger or anxiety.
Smith observed with characteristic stoicism:
wealth and power ‘keep off the summer
shower, not the winter storm’.31
True, the commercial state improved
material well-being, as wealth trickled down,
and the English commoner enjoyed more
conveniences than the kings in Africa, or so
they believed. Adam Smith famously wrote
that no country can be deemed happy ‘if
the far greater part of its members are poor
and miserable’ and, as a result, he advocated
high wages.32 Certainly in comparison to the
Roman Empire, modern Europeans were
for Smith considerably better off in mate-
rial terms, and those in the middle station,
through industry and probity, were more
likely to happen upon the path to virtue. But
woe betide the man who chose to rise any
higher, only to see at the end of his life that
the costs greatly outweighed the benefits,
that the pleasures of a fine carriage or house
2/9/2012 9:37:11 PM
 HUME ON ECONOMIC WELL-BEING
were vastly outweighed by even one month of
the insults and grovelling suffered along the
way.33 As the division of labour takes hold in
manufacturing and those who once enjoyed
the sunshine of the field are driven indoors
to a life of repetitive tasks hammering at the
head of a pin, their humanity is depleted till
they become ‘as stupid and ignorant as it is
possible for a human creature to become’.34
Furthermore, Western Europe’s transition
into a mature commercial stage and the super-
ficial pursuit of trinkets and gadgets would
only expedite the decline of the martial spirit,
and increase foppery and effeminacy, discord
and animosity. For Smith, the highest virtue
was self-command, especially in the face of
death, and this was best exemplified among
the peoples most oppressed by the Europeans,
namely the African slaves and American abo-
riginals. But the tide was to turn eventually
as opportunities for profitable investments
became scarce and the stationary state set
in. As Smith warned on the closing page of
his great tome, Britain had best ‘accommo-
date her future views and designs to the real
mediocrity of her circumstances’.35
All of the above is faithful to Smith’s texts,
but it is not of course the whole Smith or even
the Smith that is best known to posterity.
There is also an optimistic voice and indeed
many voices woven through his two books.36
But by extracting this set of most pessimistic
claims we can form a sense of that extreme
pole to which, arguably, he was drawn, even
if he equivocated along the way. The fact that
there is so much pessimism about the modern
European predicament, and so much praise
bestowed on the groups most exploited by
the European colonizers, is also very telling.
In one interesting passage, Smith also sug-
gests that among the Caribbean islands, the
magistrates of the nations considered most
liberal (that is, Britain and Holland), are
338
9780826443595_Ch18_Final_txt_print.indd 338
the harshest in their treatment of slaves.37
Whether empirically sound or not, Smith uses
this as a way to expose the shortcomings of
liberal government for its propensity to pan-
der to public interest. In short, Smith takes
many opportunities to expose the underbelly
of the modern commercial state.
Hume struck what appears to be a very
different stance, painting the modern com-
mercial world in a more favourable light. He
saw the new chapter in human history as one
that tended to enhance civility and learning,
freedom and equality, not just among the dif-
ferent ranks but between men and women as
well.38 One vision entertained the hope that
everyone might come to ‘enjoy the privilege
of rational creatures, to think as well as to
act, to cultivate the pleasures of the mind as
well as those of the body’ (E 271). Hume can
readily be cast as an enthusiast for manufac-
turing and overseas trade. He deemed mer-
chants ‘one of the most useful races of men’,
and admired their ingenuities, skill and dili-
gence in manufacturing, banking and trading
(E 300). He believed they were fully entitled
to their wealth, not to mention their acquisi-
tion of aristocratic titles. Indeed, his praise
for manufacturers and merchants is of such a
high pitch that one wonders why it was not
Hume rather than Smith who became the
poster boy for free-market capitalism.
For the most part, historical microscopes
draw these two thinkers apart; while we
know Hume was Smith’s sounding board for
the Wealth of Nations, they appear to have
differed on many of the core elements of eco-
nomic theory.39 Certainly Smith was more
inclined, like the physiocrats, to give priority
to the agrarian sector and to sing the song
of the rustic idyll, whereas Hume favoured
urbanization. Hume was more inclined to
sustain the inequalities between economic
classes than Smith, while Smith was more
2/9/2012 9:37:11 PM

inclined to sustain gender inequality. And yet,
on the question of economic well-being, both
individual and collective, Hume had laid out
many of the key elements that paved the way
to Smith’s gloomy view of the commercial
world. In this respect, they were more closely
aligned.
Although Hume sought international
peace and stability, and although he con-
demned slavery, he showed none of the
admiration that Smith expressed for the
oppressed, either for the lower orders of
Europe or the enslaved and conquered in the
Americas. Of course, it was Smith who was
the exception for that age, not Hume, and
historians still grasp at straws when it comes
to locating the source of Smith’s enlight-
enment.40 Hume believed that historically
the poor have a deeply ingrained ‘habit of
indolence’ and never thought to better their
condition (E 261). Even with the advent of
manufacturing and the increased standard
of living, labourers for Hume remain pas-
sive and subject to the illusory properties of
money, those ‘little yellow or white pieces’
of metal (E 286). Only the deceptive ring of
full-bodied coins jangling in their pockets on
payday, prior to the rise of prices and wages,
could inspire them to work, briefly, at full
intensity and alacrity.41
Hume’s specie-flow mechanism implied
the grander narrative of the migration of
economic opportunity around the globe, as
capital shifted to nations with lower wages,
only to boost that recipient nation into one
of high-priced exports and prompt the flight
of capital once more. There was, for Hume,
a dynamic process of global justice: ‘Nature,
by giving a diversity of geniuses, climates,
and soils, to different nations, has secured
their mutual intercourse and commerce, as
long as they all remain industrious and civi-
lized’ (E 329). Moreover, for Hume no single
339
9780826443595_Ch18_Final_txt_print.indd 339
HUME ON ECONOMIC WELL-BEING
nation could ever engross the art and indus-
try of the entire globe. Quite the opposite,
once a given nation had reached its height,
a relative downturn was not only inevita-
ble but the restoration of its height virtually
impossible.42
In this sense, Hume paved the way for
Smith’s closing sentence of the Wealth of
Nations regarding Britain’s gradual dete-
rioration. Like the Dutch, the British might
sustain their market share in absolute terms
because of a comparative advantage (pro-
duction of cloth) but sooner or later, another
region – Hume looks to America and then
China – would supersede Britain as the hege-
monic economic power. These long-term
shifts could also be exacerbated by war.
Although Hume, more than Smith, sub-
scribed to the ‘Montesquieu-Steuart vision’,
whereby trade was seen to be the best cure for
war, Hume was not sanguine that war could
be altogether avoided.43 For most of Hume’s
adult life, peace was the striking exception,
but the wars tended to be ‘off-shore’, rather
than intruding on the capitals of Europe.44
Nevertheless, as a young man, Hume saw
combat in Brittany and in the Netherlands
and doubts regarding the diminution of war
lingered on. As he observed in a letter to
Turgot, in 1768, regarding Ministers of the
State:
They tend to great Mischiefs [that] . . .
proceed from the still imperfect State
of our Knowledge . . . but will Men
ever reach a much more perfect State;
while the rich have so many more allur-
ing Appetites to gratify than that for
Knowledge, and the poor are occupyed
in daily Labour, and Industry. I men-
tion not the Disturbances arising from
foreign Wars, an incurable Evil, which
often springs from the greatest & most
unexpected Absurdity, and discourages
2/9/2012 9:37:11 PM
 HUME ON ECONOMIC WELL-BEING
every Project from serving or improving
human Society (LDH 2.181, 417).
The path to greater knowledge and peace
clearly lay with those in the middle station of
life, but economic inequality was inevitable
and, moreover, bred discontent. In his essay
‘Of Commerce’, Hume avers that ‘too great
a disproportion among the citizens weakens
any state. Every person, if possible, ought to
enjoy the fruits of his labour, in a full posses-
sion of all the necessaries, and many of the
conveniences of life’ (E 265). While conserva-
tive in temperament, he discerned that, at the
margin, higher incomes brought more util-
ity to the poor than to the rich. ‘No one can
doubt, but such an equality is most suitable
to human nature, and diminishes much less
from the happiness of the rich than it adds to
that of the poor’ (E 265). Hume thus cham-
pioned higher wages, arguing that a healthy
remuneration was the best incentive for dili-
gence and ingenuity and that higher levels of
consumption tend toward greater happiness,
at least up to a point.
He also believed that the development of
the manufacturing sector would cultivate
what we would now call human or social
capital: ‘We cannot reasonably expect, that
a piece of woollen cloth will be wrought to
perfection in a nation, which is ignorant of
astronomy, or where ethics are neglected’
(E 270–1). The ‘indissoluble chain’ of ‘indus-
try, knowledge, and humanity’ was at the core
of Hume’s economic vision, as the conjunc-
tion of modern commerce and urbanization
increased civility and sociability and hence
our capacity for honesty, probity and polite-
ness in both the private and public spheres.
The overall stock of trust had increased, as
evident in the issuance of fiduciary notes,
the shift to market contracts, uniform prices,
and the formation of cartels.45 The expansion
340
9780826443595_Ch18_Final_txt_print.indd 340
of markets also meant a greater reliance on
strangers and Hume’s cosmopolitanism was
part and parcel of this effort to reach out
to members of our entire species.46 Modern
commerce was also conducive to fostering
greater liberty, in the form of representa-
tive government, in the form of mobility and
expression, and in the form of unregulated
markets. All of these features of the modern
commercial era were cause for celebration.
Hume remarked on the potential for the
scientific study of agriculture to result in
higher yields, but his linkage of industry
and knowledge was a pale shadow of what
was to come. In short, he had no inkling of
the mechanical ingenuity and steam power
that would galvanize British manufacturing
after his death. His appeals to manufactur-
ing were highly traditional, the production
of cloth (wool and linen), ships and luxury
goods, ‘steel, lace, silk, books, coaches,
watches, furniture, fashions’ (LDH 1.143–4,
67). But there was considerable knowledge
acquired in banking and merchandising
in his day, and he discerned some of these
benefits. He praised bankers for new credit
instruments, daily-interest accounts and
the rapid liquidation of stocks and bonds.
While many blamed the manifest inflation
of quotidian prices on the rise of middle-
men, Hume showed that such specialization
and economies of scale had in fact reduced
real prices and made everyone wealthier. He
notes the concentrated economic activity of
the 200-mile radius around London, which
due to its ‘united extensive commerce and
middling empire, has, perhaps, arrived at a
greatness, which no city will ever be able to
exceed’ (E 448).47
Hume’s optimistic voice also shines
through in his attention to those in the mid-
dle station, a subset of which, lawyers, phi-
losophers and clergymen, Smith would later
2/9/2012 9:37:12 PM

deem unproductive. This group was more
inclined to happiness, not only because of
their higher level of material well-being from
those in the lower orders, but because they
could derive a kind of schadenfreude by vir-
tue of their position, looking down on the
poor for their deprivation, and up to the rich,
who suffered from the weight of both rank
and obligation (E 546). Hume also believed
that those in the middle station were more
likely to land on the path to virtue, that they
could cultivate the virtues of both the poor
(industry) and the rich (benevolence) and not
succumb as deeply as the latter to libertine
manners that so readily corrupted the moral
fabric. For Hume, material well-being, on
average, had an upper bound and, as a result,
the aggregate happiness of a country was best
located in fattening the middle, so to speak.
By contrast, ‘poverty and hard labour debase
the minds of the common people’ (E 198).
One way to make sense of the differ-
ent voices and ascriptions in Hume’s four
‘Happiness’ essays is to accept that he
intended ‘the reader to not only understand,
but to experience each of the four different
visions of happiness’.48 The set taken as a
whole emphasizes the fragility of human
life and its fleeting nature, and the intract-
ability of character reform, much as do the
central chapters of his Enquiry concerning
Human Understanding that rob us of little
more than a soupçon of free will. If Hume
remains enigmatic on the subject of happi-
ness in the essays most directed to the topic,
he is more inclined to unpack the notion
when it comes to the subject of luxury, or
what he calls ‘refinement in the arts’. Hume
identifies three components to personal hap-
piness: repose, pleasure and action. The first
two are essential but the true anchor of hap-
piness lies in action; ‘the most constant and
insatiable craving or demand is for exercise
341
9780826443595_Ch18_Final_txt_print.indd 341
HUME ON ECONOMIC WELL-BEING
and employment’ (E 269–7). Moreover, such
action must engage the mind, enhancing ‘its
powers and faculties’. Indolence or repose
is needed to refresh our energies, but these
become destructive if prolonged, as is often
the case among the idle rich. ‘Deprive men
both of action and of pleasure; and leaving
nothing but indolence in their place, you
even destroy the relish of indolence, which
never is agreeable, but when it succeeds to
labour’ (E 270).
There is reason, then, to question the verity
linking material well-being and happiness at
the individual level. Excessive wealth, unless
sustained by mental labour, will only bring
despair. Nevertheless, one must have pleas-
ures, luxuries of some sort, and as manufac-
turing takes hold and increases, this is likely
to provide those in the middle station with
more pleasure. Hume famously cast the
consumption of luxury goods in a favour-
able light, noting that ‘they add to the hap-
piness of the state; since they afford to many
the opportunity of receiving enjoyments,
with which they would otherwise have been
unacquainted’ (E 256). He adopted the early
eighteenth-century view that luxury goods
served to banish sloth and idleness and thus
urged the dismantling of sumptuary laws.
He was keen to see domestic manufactur-
ers imitate and adopt the lines of produc-
tion from abroad, as exemplified by the silk
production in Spitalfields. But there were
different types of luxuries, new and old,
innocent and vicious.49 He favoured the
new and innocent luxuries, not the extrava-
gances of feudal lords, nor those that would
induce ‘feverish, empty amusements’ that
‘draw ruin upon us, and incapacitate us for
business and action’ (THN 3.3.4.7 / 611).
As we saw, even the ‘most expensive luxury’
is inferior to the enjoyment of ‘a poem or a
piece of reasoning’ (E 5).
2/9/2012 9:37:12 PM
 HUME ON ECONOMIC WELL-BEING
Bought pleasures become hollow with
time. Hume heaps scorn on the pursuit of
‘worthless toys and gewgaws’ (EPM 9.25 /
283). It is the set of non-pecuniary goods,
‘conversation, society, study, even health and
the common beauties of nature, but above all
the peaceful reflection on one’s own conduct’
that yield the highest pleasures, both qualita-
tively and quantitatively (EPM 9.25 / 283–4).
And yet, as we become more learned and
reflective, we find society increasingly intol-
erable and friendships wane. For the man of
refined taste, Hume asserts, only a few per-
sons meet his standards for friendship and,
as a result, he may be more likely to despair.
These afflictions may beset only a minority
of persons (those wise and refined), but inso-
far as the contemplative life is the paragon of
virtue, a liability worth noting. Smith would
later propose that merchants are also inclined
to cultivate limited numbers of friends and to
refrain from frequent amusements and other
indulgences.
For Hume, ‘the greatest part of mankind
float between vice and virtue’, and are sub-
ject to the persuasions of the mob (E 594).
In the political sphere, as Knud Haakonsen
has argued, Hume was one of the first to
alert us to the sway of public interest and
the difficulties of fulfilling our full indi-
viduality.50 The fact that Hume draws these
observations is indeed a sign that he puts
considerable value on countering that trend.
Underlying economic forces set up strong
currents towards mob-like behaviour, how-
ever, against which bold individuals must
vigilantly fight. Nevertheless, Hume is clear
that luxury consumption is to be promoted
provided it does not undercut other virtues
and non-pecuniary ends, such as benevo-
lence, paternal duties, friendship or one’s
reputation.51 The right kinds of luxuries will
eventually become conveniences, and even
342
9780826443595_Ch18_Final_txt_print.indd 342
possibly necessities as wealth expands, since
these stimulate industry, prudence and the
accumulation of property.
Hume believed that the on-going march
towards increased consumption of luxuries
(of the beneficial kind) would actually lead
to more refinement and the transcendence of
excess. Here are intimations of the enlighten-
ment virtues of self-command and the culti-
vation of good character. Nevertheless, Hume
emphasized the many temporal and political
contingencies that undergird the pursuit of
luxury and hence the overarching tendency
to instability.52 Ever the conservative, Hume
worried that the distance between the very
wealthy and those in the middle might lessen
and thus undercut the system of ranks so
essential, he thought, to stability. Like Smith,
he put much weight on the importance of the
ostentatious display of rulers, and the admi-
ration it instilled in those beneath.
For Hume, avarice was not only a wide-
spread trait but perhaps the most difficult to
eradicate (E 571). Most merchants, he avers,
are also misers and when coupled with ava-
rice, tend only to drive away pleasure and
friendship, thus weakening sociability. ‘This
avidity alone, of acquiring goods and posses-
sions for ourselves and our nearest friends,
is insatiable, perpetual, universal, and
directly destructive of society. There scarce
is any one, who is not actuated by it’ (THN
3.2.2.12 / 491–2). He was a firm believer
in the ancient caveat of generational decay;
sons rarely shared the virtues of their fathers.
Hume relates a fable by Antoine Houdar de
la Motte, in which Charon bestows the great-
est punishment on a miser after his death,
namely that he be ‘sent back to the earth, to
see the use his heirs are making of his riches’
(E 572).
Hume reminds us again and again that eve-
rything is fleeting and ephemeral: ‘Nothing
2/9/2012 9:37:12 PM

in this world is perpetual. Every being, how-
ever seemingly firm, is in continual flux and
change’ (E 597). As a result, he expresses
humility in making predictions. Given the
dramatic rise to economic dominance, first
of Holland then of Britain in the past two
centuries, he admits that ‘such mighty revo-
lutions have happened in human affairs, and
so many events have arisen contrary to the
expectation of the ancients, that they are suf-
ficient to beget the suspicion of still further
changes’ (E 89). Wealth does not simply aug-
ment over time but rather contains within
it the seeds of its own transformation, and
it is these epigenetic processes that distin-
guish Hume’s economic vision from that of
his predecessors. Modes of production and
patterns of consumption evolve in step with
the migration of economic opportunities, not
only between households and nations, but
also between classes. In 500 years, Hume
conjectures, servants and masters will have
changed places (E 357).
One main source of change was the bur-
geoning public debt.53 In the 1720s Sir
Robert Walpole initiated the practice of issu-
ing government bonds and Hume feared that,
without some inherent system of checks and
balances, public debts were to undo us all. ‘It
would scarcely be more imprudent to give a
prodigal son a credit in every banker’s shop
in London, than to impower a statesman to
draw bills, in this manner, upon posterity’
(E 352). The greater injustice is located in
the fact that creditors are a minority, about
17,000 individuals, but the safety of millions
is sacrificed to their pecuniary interests as
long as the system of public credit endures.
Hume provides a thought experiment,
whereby taxes are ‘screwed up to the utmost
which the nation can bear’, eighteen or nine-
teen shillings on the pound. ‘The seeds of ruin
are here scattered with such profusion’ that
343
9780826443595_Ch18_Final_txt_print.indd 343
HUME ON ECONOMIC WELL-BEING
tyranny is inevitable (E 357). Hume predicts
the ‘death of public credit’, both through nat-
ural and violent means. He resists assigning a
specific date to his prediction, just an interval
of 50 to 500 years (E 365n).
What makes this dire outcome all the more
inevitable for Hume is the ever-increasing
cost of national defence. As the martial spirit
wanes, standing armies must rely on size and
subordination to compensate. Britain’s navy
relied upon the system of the press-gangs,
which Hume likened to a Hobbesian state of
nature for its violence, disorder and unpre-
dictability (E 374–6). The Dutch were devoid
of martial spirit altogether and thus drained
their coffers by hiring Swiss mercenaries. As
personal secretary to General St Clair in 1746,
Hume had witnessed the absurdity of mod-
ern combat, an aborted invasion of Lorient
whereby more lost their lives through deser-
tion than on the battlefield. As Montesquieu
had famously observed, war and conquest
had hardly changed the face of Europe, not-
withstanding immense bloodshed in the sev-
enteenth century. Both he and Hume believed
in the robustness of national character and
thus the limited success of imperialism.
To provide empirical support, Hume calcu-
lated the cost of mounting the Roman army
to show that contemporary European mili-
tary expenditure had grown considerably in
real terms (E 282–3n). He knew from Tacitus
that the daily pay of a soldier was one denar-
ius, and he knew something about this con-
sumption bundle, food and raiment. Hume
thus estimated the equivalent of eight-pence,
not a princely sum. In a different essay, he
notes that ‘every person in England is com-
puted by some to spend six-pence a day: Yet
is he esteemed but poor who has five times
that sum’ (E 429). There were 25 Roman
legions, each with 5,000 men, so the over-
all bill, per annum, came to 1.5 million, but
2/9/2012 9:37:12 PM
 HUME ON ECONOMIC WELL-BEING
because officers were paid double, Hume
rounded up to 1.6 million pounds. By com-
parison, he noted that the British navy alone
cost over 2.5 million pounds.54 He failed to
note that it was also assisting traders by pro-
tecting them from pirates on the open seas,
but the overall comparison between modern
and ancient expenditures was aptly made.
More than his predecessors and contem-
poraries in economic thought, Hume’s ana-
lysis is grounded in a historical sensibility. He
sweeps over hundreds of years, noting prices
and practices from antiquity, the Middle
Ages and the early modern period. He asserts
that the world is quite young, that the advent
of the commercial era, but a century or so
ago, means that we are still not in a position
of sufficient temporal distance to make sense
of it. And in his assessment of the value of
historical scholarship, he puts much weight
on the value of surveying a thousand years,
and the knowledge one garners from the
accumulated data. But insofar as we are on a
developmental path – and for Hume I think
that is never in doubt – we cannot yet know
‘what may be expected of mankind’ in terms
of ‘degree of refinement, either in virtue or
vice’ (E 87). Humans are at best middle-aged
as a species, and there is thus much time left
to run (E 378). Wealth will continue to cycle
around the globe, reaching its height in one
region only to decline and move elsewhere
a century or two later.55 Happiness, or more
generally well-being, might sustain itself if
the proper institutions foster the right condi-
tions, such as the growth of the middle class,
the promotion of industry conjoined with
knowledge and sociability, the perfect bal-
ance between labour and leisure and, above
all, the promotion of peace rather than war.
The probability of such a confluence sustain-
ing itself in any one place over the long term,
however, was low. Despite glimmers of hope,
344
9780826443595_Ch18_Final_txt_print.indd 344
Hume seems all too inclined to envision our
eventual downturn as the commercial era
unfolds in the centuries ahead.
NOTES
1
The introduction by Eugene Rotwein to
Hume’s economic writings provides a
thorough overview of Hume’s economics in the
English language. See Eugene Rotwein (ed.),
‘Introduction’, Writings on Economics: David
Hume (New Brunswick, NJ: Transaction,
2007; originally pub. 1955), pp. ix–cxi. For
more recent assessments, see Andrew Skinner,
‘Hume’s Principles of Political Economy’, in
David Fate Norton and Jacqueline Taylor (eds),
The Cambridge Companion to Hume, 2nd edn
(Cambridge: Cambridge University Press, 2009)
and the collection of essays in Carl Wennerlind
and Margaret Schabas (eds), David Hume’s
Political Economy (London and New York:
Routledge, 2008).
2
The most detailed account of Hume’s life,
including his first sojourn in London, remains
Ernest C. Mossner, The Life of David Hume,
2nd edn (Oxford: Clarendon Press, 1980). On
Hume’s interactions with leading economists of
his day, see Ian Simpson Ross, ‘The Emergence
of David Hume as a Political Economist’,
in Wennerlind and Schabas (eds), Hume’s
Political Economy, pp. 31–48. Ross also points
out the sketch of economic topics in Hume’s
‘Early Memoranda’ based on recent findings
by Tatsuya Sakamoto (ibid., pp. 41–2). For
Hume as an accountant, see the archival papers
of Lieutenant General James St Clair in the
National Library of Scotland. For an interesting
account of Hume’s extensive connections with
the Scottish elite and his efforts to influence
government policies on the Highlanders, see
Roger Emerson, ‘The Scottish Contexts for
David Hume’s Political-Economic Thinking’, in
Wennerlind and Schabas (eds), Hume’s Political
Economy, pp. 10–30.
3
Playing cards had been used as currency in
Canada since 1685. Hume had remarked on
Canadian paper money in correspondence
to Montesquieu and to Oswald, respectively,
in 1749. See Robert Dimand, ‘David Hume
2/9/2012 9:37:12 PM

on Canadian Paper Money’, Wennerlind and
Schabas (eds), Hume’s Political Economy,
pp. 168–80.
4
See Mossner, Life of Hume, p. 539. For an
additional account of Hume’s prolonged
interest in the commerce and trade in the
North Atlantic, see Emma Rothschild, ‘The
Atlantic Worlds of David Hume’, in Bernard
Bailyn and Patricia Denault (eds), Soundings
in Atlantic History: Latent Structures and
Intellectual Currents, 1500–1830 (Cambridge,
MA: Harvard University Press, 2009),
pp. 405–50.
5
See, for example, Carl Wennerlind, ‘The Link
between David Hume’s A Treatise of Human
Nature and His Fiduciary Theory of Money’,
History of Political Economy 33(1) (2001),
pp. 139–60; Till Grüne-Yanoff and Edward F.
McClennen, ‘Hume’s Framework for a Natural
History of the Passions’, in Wennerlind and
Schabas (eds), Hume’s Political Economy,
pp. 86–104; and Robert Sugden, ‘Hume’s Non-
instrumental and Non-propositional Decision
Theory’, Economics and Philosophy 22(2)
(2006), pp. 365–91.
6
See Carl Wennerlind, ‘David Hume’s Political
Philosophy: A Theory of Commercial
10
Modernization’, Hume Studies 28(2) (2002),
pp. 247–70.
7
On economics at the vanguard of secular-
ism, see John Robertson, The Case for the
Enlightenment: Scotland and Naples 1680–
1760 (Cambridge: Cambridge University Press,
2005). On Hume as a proto-Bayesian, see Sally
Ferguson, ‘Bayesianism, Analogy, and Hume’s
Dialogues concerning Natural Religion’, Hume
Studies 28(1) (2002), pp. 113–30.
8 11
These three French translations were all
issued promptly, in 1754 and 1755. See Loïc
Charles, ‘French “New Politics” and the
12
Dissemination of David Hume’s Political
Discourses on the Continent, 1750–70’,
in Wennerlind and Schabas (eds), Hume’s
Political Economy, pp. 181–202. He notes
(p. 181) that for the period 1752 to 1767,
13
Hume’s ‘Discourses was republished no
fewer than seventeen times in five languages’,
including six in French, two in German, two
in Italian and one in Swedish.
9
See F. Quesnay, Tableau Économique [1758] in
M. Kuczynsi and R.L. Meek (eds), Quesnay’s
345
9780826443595_Ch18_Final_txt_print.indd 345
HUME ON ECONOMIC WELL-BEING
Tableau Économique (New York: A.M. Kelley,
1972); A.-R.-J. Turgot, Valeurs et Monnaies
[1769] in J.-T. Ravix and P.-M. Romani (eds),
Turgot: Formation et distribution des Richesses
(Paris: Flammarion, 1997); and A. Smith,
An Inquiry into the Nature and Causes of
the Wealth of Nations, ed. E. Cannan, 5th
edn (London: Methuen, 1904). For Hume’s
American legacy see, for example, John G.A.
Pocock, ‘Hume and the American Revolution’,
in Virtue, Commerce, and History: Essays on
Political Thought and History, Chiefly in the
Eighteenth Century (Cambridge: Cambridge
University Press, 1985), pp. 125–41. Ricardo
and Say, the most widely read economists in
France and the United States in the first half
of the nineteenth century, dominated British
economics until John Stuart Mill (1848). See
J.-B. Say, A Treatise on Political Economy; or
the Production, Distribution, and Consumption
of Wealth [1803], ed. C.C. Biddle, trans. C.R.
Prinsep (Philadelphia: Lippincott, Grambo,
1855); D. Ricardo, On the Principles of
Political Economy and Taxation [1817] in Piero
Sraffa (ed.), The Works and Correspondence of
David Ricardo, 11 vols (Indianapolis: Liberty
Fund, 2005), vol. 1.
See Milton Friedman, ‘Comment on Empirical
Monetary Macroeconomics: What Have We
Learned in the Last 25 Years?’, American
Economic Review 65(2) (1975), pp. 176–9.
For a general overview of Hume’s monetary
theory, see Carl Wennerlind, ‘David Hume’s
Monetary Theory Revisited: Was He Really
a Quantity Theorist and an Inflationist?’,
Journal of Political Economy 113(1) (2005),
pp. 223–37.
See Ross, ‘Hume as a Political Economist’,
in Wennerlind and Schabas (eds), Hume’s
Political Economy, p. 40.
For Hume on trade, see Skinner, ‘Hume’s
Political Economy’, and John Berdell,
‘Innovation and Trade: David Hume and the
Case for Freer Trade’, History of Political
Economy 28(1) (1996), pp. 107–26.
Christine Thère has done a masterful job of
sorting the titles into subfields such as trade
or public finance, and not double-counting
subsequent editions of the same work. Her
precise count for the period of 1700 to 1784
for French publications comes to 3,563 titles
2/9/2012 9:37:12 PM
 HUME ON ECONOMIC WELL-BEING
by 1,856 different authors. See Christine
Thère, ‘Economic Publishing and Authors,
1566–1789’, in Gilbert Faccarello (ed.), Studies
in the History of French Political Economy:
From Bodin to Walras (London and New
York: Routledge, 1998), pp. 1–56. For a
broader study of French economic thought, see
Catherine Larrère, L’Invention de l’Économie
au XVIIIe Siècle (Paris: Léviathan, 1992).
14
For two recent judgements on the possibility
of Hume knowing Cantillon prior to 1752, see
Carl Wennerlind, ‘An Artificial Virtue and the
Oil of Commerce’, in Wennerlind and Schabas
(eds), Hume’s Political Economy, pp. 121–2;
and Istvan Hont, ‘The “Rich Country-Poor
Country” Debate Revisited’, in Wennerlind
and Schabas (eds), Hume’s Political Economy,
p. 319n.
15
See Charles, ‘Dissemination on the Continent’.
For more about du Tot, see Antoin E.
Murphy, ‘The Enigmatic Monsieur du Tot’,
in Faccarello, French Political Economy,
pp. 57–77.
16
See Robertson, Case for the Enlightenment.
17
See Ronald Meek (ed.), The Economics of
Physiocracy: Essays and Translations (London:
George Allen & Unwin, 1962). For Hume on
Quesnay’s deficient empirical accuracy, see
Margaret Schabas, ‘Hume’s Monetary Thought
Experiments’, Studies for the History and
Philosophy of Sciences 39 (2008), p. 165.
18
Larrère argues that the sudden demise of physi-
ocracy after Quesnay died is incorrect and that
physiocratic concepts of natural rights lived on
with the political theories of Emmauel-Joseph
Sieyès. See Larrère, L’Invention de l’Economie,
pp. 269–72.
19
Hume wrote to Smith in 1776, ‘I shall
still doubt for some time of its [Wealth of
Nations] being at first very popular’ (LDH
2.311, 517).
20
For a discerning account of the peculiar ties
between Hume and Smith, see Eric Schliesser,
‘The Obituary of a Vain Philosopher: Adam
Smith’s Reflections on Hume’s Life’, Hume
Studies 29(2) (2003), pp. 327–62.
21
Hume lorded his eminence in economics over
Smith. In a letter to Smith in 1767, wonder-
ing why his book (Wealth of Nations) was
delayed, Hume wrote: ‘I am positive you are
in the wrong in many of your Speculations,
346
9780826443595_Ch18_Final_txt_print.indd 346
especially where you have the Misfortune to
differ from me’ (LDH 2.207, 432).
22
Hume took his figures from a well-known text
by Charles Ferrère Du Tot. Silver coins in 1683
were at 30 livres the mark and had risen to 50
due to significant debasement (E 287–8). The
silver price of corn was also adopted by Adam
Smith as the best index for century-by-century
comparisons.
23
See M. Schabas, ‘Market Contracts in the
Age of Hume’, in N. De Marchi and M.S.
Morgan (eds), Higgling: Transactors and
Their Markets in the History of Economics
(Durham, NC: Duke University Press, 1994),
pp. 117–34.
24
This was a response to the Reverend Robert
Wallace. A draft of his Dissertation on the
Numbers of Mankind in Ancient and Modern
Times (1753) circulated in the mid-1740s.
25
The linkage of happiness with economic pros-
perity was perhaps most developed among
Italian economists. See Luigino Bruni and Pier
Luigi Porta, ‘Economia civile and pubblica felic-
tia in the Italian Enlightenment’, in Margaret
Schabas and Neil De Marchi (eds), Oeconomies
in the Age of Newton (Durham and London:
Duke University Press, 2003), pp. 361–85.
26
For Hume as an institutionalist utilitarian,
see Russell Hardin, David Hume: Moral and
Political Theorist (Oxford: Oxford University
Press, 2007).
27
It is worth noting that both rendered markets
more capacious by incorporating ‘public goods’
such as military protection and, in Smith’s case,
the provision of education and religion. See James
R. Otteson, Adam Smith’s Marketplace of Life
(Cambridge: Cambridge University Press, 2002).
28
This ‘paradox of wealth’ is laid out in
full in Istvan Hont, ‘The “Rich Country-
Poor Country” Debate in the Scottish
Enlightenment’, in I. Hont (ed.) Jealousy of
Trade (Cambridge, MA: Belknap Press, 2005),
pp. 267–322.
29
Adam Smith, An Inquiry into the Nature
and Causes of the Wealth of Nations, 2 vols
(Oxford: Oxford University Press, 1976),
vol. 2, p. 565 (4.7.3).
30
Ibid., vol. 1, p. 190 (1.9.31).
31
Adam Smith, The Theory of Moral Sentiments
(Oxford: Oxford University Press, 1776),
p. 183 (4.1.8).
2/9/2012 9:37:12 PM

32
Smith, Wealth of Nations, vol. 1, p. 96 (1.8.36).
33
Smith, Moral Sentiments, p. 181 (4.1.8).
34
Smith, Wealth of Nations, vol. 2, p. 782
(5.1.50).
35
Ibid., vol. 2, p. 613 (4.7.63); p. 947 (5.3.92).
36
For some excellent assessments of Adam
Smith’s vocal counterpoint see Vivienne
Brown, Adam Smith’s Discourse: Canonicity,
Commerce and Conscience (London:
Routledge, 1994); Charles L. Griswold, Adam
Smith and the Virtues of Enlightenment
(Cambridge: Cambridge University Press,
1999); and Samuel Fleischacker, On Adam
Smith’s Wealth of Nations (Princeton:
Princeton University Press, 2004).
37
‘The law, so far as it gives some weak
protection to the slave against the violence of
his master, is likely to be better executed in
a colony where the government is in a great
measure arbitrary, than in one where it is
altogether free’ (Smith, Wealth of Nations,
vol. 2, p. 587 (4.7.54)).
38
On Hume’s high but complicated regard for
women, see Lívia Guimarães, ‘Hume and
Feminism’, chap. 17 of this volume, and ‘The
Gallant and the Philosopher’, Hume Studies
30(1), (2004), pp. 127–47.
39
To give but one example, they held differ-
ent views on the formation of prices. Both
focused on the cost of production and
positioned wages as the key factor, but Hume,
contrary to Smith, believed that the return to
land or rent did not form a part of the price
(LDH 2.312, 517).
40
Ian Ross points to the influence of three
books : Joseph-François Lafitau, Mœurs des
sauvages ameriquains, compareés aux mœurs
des premiers temps (1724); Cadwallader
Colden, History of the Five Nations of Canada
(1727, 1747); and Pierre-François Xavier de
Charlevoix, Histoire et description générale de
la Nouvelle France (1744). But the question as
to why Smith and not others was so moved by
these accounts still remains partly unanswered.
See Ian Simpson Ross, The Life of Adam Smith
(Oxford: Clarendon Press, 1995), p. 169.
Also see Leonidas Montes, Adam Smith in
Context (Basingstoke and New York: Palgrave
Macmillan, 2004).
41
For more on Hume’s mechanism by which a
singular rise in the money stock stimulated
347
9780826443595_Ch18_Final_txt_print.indd 347
HUME ON ECONOMIC WELL-BEING
economic productivity, see Margaret Schabas,
‘Temporal Dimensions in Hume’s Monetary
Theory’, in Wennerlind and Schabas (eds),
Hume’s Political Economy, pp. 127–45. Smith,
by contrast, was more inclined to ascribe
inventiveness to ordinary workers and deemed
universal the propensity to better one’s con-
dition. He did, however, acknowledge that if
piecework is well paid, workers will supply
more hours and possibly intensify their effort
by the hour.
42
Hume wrote: ‘when the arts and sciences come
to perfection in any state, from that moment
they naturally, or rather necessarily decline,
and seldom or never revive in that nation,
where they formerly flourished’ (E 135).
43
See Albert O. Hirschman, The Passions and
the Interests (Princeton: Princeton University
Press, 1977).
44
If one takes the period from the War of the
Austrian Succession in 1739 to the end of the
War of American Independence in 1783 as a
‘single conflict’, the so-called ‘forty years’ war
of the eighteenth century’, then it is reasonable
to view this as ‘the dominating circumstance
of Hume’s public life’. See Rothschild, ‘Atlantic
Worlds’, p. 419.
45
See Schabas, ‘Market Contracts’, pp. 117–34;
and Richard Boyd, ‘Manners and Morals:
David Hume on Civility, Commerce, and
the Social Construction of Difference’, in
Wennerlind and Schabas (eds), Hume’s
Political Economy, pp. 65–85.
46
See Michael Ignatieff, The Needs of Strangers
(London: Chatto and Windus, 1984).
47
For more on Hume on the division of labour
both in production and marketing, see
Schabas, ‘Market Contracts’. For an excellent
overview of the industrialization of Britian,
see Joel Mokyr, The Enlightened Economy:
An Economic History of Britain 1700–1850
(New Haven, CT: Yale University Press,
2009).
48
The four essays are ‘The Epicurean’, ‘The
Stoic’, ‘The Platonist’ and ‘The Sceptic’. See
Colin Heydt, ‘Relations of Literary Form and
Philosophical Purpose in Hume’s Four Essays
on Happiness’, Hume Studies 33(1) (2007),
pp. 3–19, p. 7.
49
See Christopher J. Berry, ‘Hume and Superfluous
Value (or the Problem with Epictetus’ Slippers)’,
2/9/2012 9:37:13 PM
 HUME ON ECONOMIC WELL-BEING
53
in Wennerlind and Schabas (eds), Hume’s
Political Economy, pp. 49–64.
50
See Knud Haakonssen (ed.), ‘Introduction’
to Hume: Political Essays (Cambridge:
54
Cambridge University Press, 1994), pp. xi–xxx.
I have argued that when it comes to his politi-
cal economy, Hume was a methodological
holist. See Margaret Schabas, ‘Groups versus
Individuals in Hume’s Political Economy’, The
Monist 90(2) (2007), pp. 200–12.
51
For an excellent probing into this topic,
see Ryu Susato, ‘Hume’s Nuanced Defense
55
of Luxury’, Hume Studies 32(1) (2006),
pp. 167–86, p. 171.
52
See Susato, ‘Hume’s Nuanced Defense of
Luxury’, p. 181.
348
9780826443595_Ch18_Final_txt_print.indd 348
See Istvan Hont, ‘The Rhapsody of Public
Debt: David Hume and Voluntary State
Bankruptcy’, in I. Hont (ed.), Jealousy of
Trade, pp. 325–53.
There is reason to believe that Hume knew
the last figure with considerable accuracy. In
1744 he wrote to his friend William Mure
that James Oswald ‘had shown me the whole
Oeconomy of the Navy, the Source of the Navy
Debt; with many other Branches of Public
Business’ (LDH 1.58, 24). Also see Rothschild,
‘Atlantic Worlds’, p. 409.
For more on the evolutionary context of
Hume’s economics, see Margaret Schabas,
The Natural Origins of Economics (Chicago:
University of Chicago Press, 2005), chap. 4.
2/9/2012 9:37:13 PM
 19
‘OF THE STANDARD OF TASTE’:
DECISIONS, RULES AND
CRITICAL ARGUMENT
M.W. Rowe
Near the beginning of his essay, ‘Of the
Standard of Taste’ (E 226–49), Hume offers
two suggestions as to what a standard of
aesthetic taste might look like: on the one
hand it might be a rule, on the other, a deci-
sion: ‘It is natural for us to seek a Standard
of Taste; a rule, by which the various sen-
timents of men may be reconciled; at least,
a decision, afforded, confirming one senti-
ment, and condemning another.’ (E 229) In
the essay as a whole, he devotes more space
to the idea of a decision: he clearly thinks
that this is the more important notion, and
he will eventually define the standard of
taste in terms of decisions alone. However,
there are passages which suggest he thinks
rules play a significant role in forming and
confirming decisions, and that the two
notions are not only compatible but mutu-
ally supportive.
In this article I propose to argue that Hume
does not establish the standard of taste with
reference to any kind of decision, and that he
does not establish the place of rules in aes-
thetic decision-making.
I shall argue for these positions in the fol-
lowing way. The first section reconstructs
Hume’s qualified subjectivism – which I take
349
9780826443595_Ch19_Final_txt_print.indd 349
to be the most valuable part of his essay –
without mentioning the notion of rules. The
next four sections show that his introduc-
tion of rules, far from supporting his case for
subjectivism, is actually in tension with it.1
The last main section shows that his attempt
to establish the standard of taste with ref-
erence to the joint verdict of qualified crit-
ics fails, and fails for the same reason as his
attempt to claim that rules play a role in
critical argument. In the course of these dis-
cussions, I will of course show what ‘rule’,
‘decision’, ‘standard of taste’ and ‘qualified
subjectivism’ mean, all of which are cur-
rently unclear.
1. QUALIFIED SUBJECTIVISM
The best place to begin an exposition of
Hume’s subjectivism is with the section which
immediately follows the short paragraph
about rules and decisions quoted above. This
section – in the manner of an antinomy – out-
lines two apparently incompatible intuitions
about the objectivity of taste, both of which
are found in common sense.
2/9/2012 9:37:20 PM
 ‘OF THE STANDARD OF TASTE’
On the one hand, he argues, there seems to
be no right or wrong in matters of aesthetic
judgement: the ascription of aesthetic qual-
ities seems to depend on sentiment not opin-
ion; and the qualities themselves are projected
onto the world rather than discovered in it:
Among a thousand different opinions
which men may entertain of the same
subject, there is one, and but one, that
is just and true; and the only difficulty is
to fix and ascertain it. On the contrary, a
thousand different sentiments excited by
the same object are all right: Because no
sentiment represents what is really in the
object. It only marks a certain conformity
or relation between the organs or faculties
of the mind; and if that conformity did
not really exist, the sentiment could never
possibly have being. Beauty is no quality
in things themselves: It exists merely in
the mind which contemplates them; and
each mind perceives a different beauty.
One person may even perceive deform-
ity, where another is sensible of beauty;
and every individual ought to acquiesce
in his own sentiment, without pretending
to regulate those of others. To seek real
beauty or real deformity, is as fruitless an
inquiry, as to pretend to ascertain the real
sweet or the real bitter. (E 230)
On the other hand, he argues, we feel that in
some cases a critic just is right or wrong, and
there really is a fact of the matter which eve-
ryone ought to acknowledge. If this is so, then
aesthetic predicates do not seem to be merely
projected onto the world by the sentiments;
they seem to be real qualities in the world
which can be the objects of opinion and belief:
Whoever would assert an equality of
genius and elegance between Ogilby
and Milton, or Bunyan and Addison,
would be thought to defend no less an
350
9780826443595_Ch19_Final_txt_print.indd 350
extravagance, than if he had maintained
a mole-hill to be as high as Teneriffe,
or a pond as extensive as the ocean.
Though there may be found persons,
who give the preference to the former
authors; no one pays attention to such a
taste; and we pronounce without scruple
the sentiments of these pretend critics to
be absurd and ridiculous. The principle
of the natural equality of tastes is then
totally forgot . . . . (E 230–1)
As Hume’s sympathy remains with the first
thesis of this antinomy, he realizes he needs
to explain away the intuition underlying the
second. He does this by pointing out that the
colours we see and the flavours we detect are
undoubtedly the result of the external world
impinging on our internal constitution, but this
does not mean that just any judgements about
colour or flavour are to be accepted as true:
A man in a fever would not insist on
his palate as able to decide concerning
flavours; nor would one, affected with
jaundice, pretend to give a verdict with
regard to colours. In each creature, there
is a sound and a defective state; and the
former alone can be supposed to afford
us a true standard of taste and sentiment,
. . . [just as] the appearance of objects in
daylight, to the eye of the man in health,
is denominated their true and real col-
our, even while colour is allowed to be a
phantasm of the senses. (E 233–4)
Thus, even if we acknowledge that colour is
just a reaction of a certain kind of body to a
certain kind of stimulation, and therefore a
‘phantasm of the senses’, we can still rule out
certain colour judgements as illegitimate. In
the same way, even if we agree that aesthetic
reactions are just reactions of certain kinds
of body to certain kinds of stimulation, we
can still rule out some aesthetic reactions as
2/9/2012 9:37:20 PM

equally ill-founded. Consequently, the anti-
nomy of taste begins to look less sharp and
absolute than once it did: the fact that taste is
subjective does not mean the concept of error
cannot be applied to it.
If various kinds of illness, injury and
deformity rule out some judgements about
flavour and colour, what rules out certain
judgements about beauty and ugliness?
Hume allows that such judgements are more
easily disturbed than judgements about col-
our and flavour:
Those finer emotions of the mind are
of a very tender and delicate nature,
and require the concurrence of many
favourable circumstances to make them
play with felicity and exactness . . . . The
least exterior hindrance to such small
springs, or the least internal disorder,
disturbs their motion, and confounds the
operation of the whole machine. When
we would make an experiment of this
nature, and would try the force of any
beauty or deformity, we must choose
with care a proper time and place, and
bring the fancy to a suitable situation
and disposition. (E 232)
Hume now begins to outline the kind of
person who is acknowledged to have well-
grounded aesthetic reactions. He emphasizes
five necessary qualifications – I shall call
them ‘judgement conditions’ – in particular:
i. DELICACY OF MENTAL TASTE
This allows fine discriminations so that noth-
ing relevant is allowed to escape notice and
observation. The smallest objects can only
be discerned by the best eye, and the smallest
amounts of flavour can only be detected by
the best palate. ‘In a like manner, a quick and
accurate perception of beauty and deformity
351
9780826443595_Ch19_Final_txt_print.indd 351
‘OF THE STANDARD OF TASTE’
must be the perfection of our mental taste; nor
can a man be satisfied with himself while he
suspects, that any excellence or blemish in a
discourse has passed him unobserved.’ (E 236)
ii. EXTENSIVE PRACTICE
Nothing better promotes delicacy of taste
than practice.
When objects of any kind are first pre-
sented to the eye or imagination, the sen-
timent, which attends them, is obscure
and confused; and the mind is, in great
measure, incapable of pronouncing con-
cerning their merits or defects. . . . But
allow [a man] to acquire experience in
these objects, his feeling becomes more
exact and nice. . . . The mist dissipates,
which seemed formerly to hang over the
object: The organ acquires greater perfec-
tion in its operations; and can pronounce,
without danger or mistake, concern-
ing the merits of each performance. In
a word, the same address and dexterity,
which practice gives to the execution of
any work, is also acquired by the same
means, in the judging of it. (E 237)
iii. WIDE EXPERIENCE OF COMPARISON
Evaluating a work correctly means placing
it on the continuum of all other such works,
and in order to do this, we need to know the
full range of quality, from the highest to the
lowest:
By comparison alone we fix the epithets
of praise or blame, and learn how to
assign the due degree of each. The coars-
est daubing contains a certain lustre
of colours and exactness of imitation,
which are so far beauties, and would
affect the mind of a peasant or Indian
2/9/2012 9:37:20 PM
 ‘OF THE STANDARD OF TASTE’
with the highest admiration. . . . One
accustomed to see, and examine, and
weigh the several performances, admired
in different ages and nations, can alone
rate the merits of work exhibited to his
view, and assign its proper ranks among
the productions of genius. (E 238)
iv. IMPARTIALITY
If your reaction is clouded by the fact that you
are ‘a rival to’ or have ‘a friendship or enmity
with the author’ (E 239), then your judge-
ment must be set aside as biased or partial.
v. GOOD SENSE
This is necessary, not only to detect and erad-
icate prejudice, but to see how – and how
well – the parts of a work cohere and make
their effect:
In all the nobler productions of genius,
there is a mutual relation and corre-
spondence of parts; nor can either the
beauties or the blemishes be perceived
by him, whose thought is not capacious
enough to comprehend all those parts,
and compare them with each other, in
order to perceive the consistence and
uniformity of the whole. (E 240)
Hume summarizes the position I have called
‘qualified subjectivism’ by saying that true
judges or critics have all the five qualifica-
tions outlined above. ‘Strong sense, united to
delicate sentiment, improved by practice, per-
fected by comparison, and cleared of all preju-
dice, can alone entitle critics to this valuable
character [of the true judge]; and the joint
verdict of each, wherever they are found, is
the true standard of taste and beauty.’ (E 241)
Hume admits that we seem to have simply
352
9780826443595_Ch19_Final_txt_print.indd 352
replaced the problem of how to recognize
great works of art with the problem of how
to recognize great critics, but is unapologetic:
‘It is sufficient for our present purpose, if we
have proved that the taste of all individuals
is not upon an equal footing, and that some
men in general, however difficult to be par-
ticularly pitched upon, will be acknowledged
by universal sentiment to have preference
above others.’ (E 242) Thus Hume resolves his
antimony of taste, and in so doing arrives at
how a standard of taste is to be determined. It
is the joint verdict or decision of true judges.
However, because human beings who
satisfy the five conditions still vary greatly,
a wide diversity of aesthetic opinion is still
not only possible but likely. Hume thinks
that two kinds of variation in particular
are ineradicable. The first is the ‘different
humours of particular men’ (E 243), and this
is itself broken down into two subdivisions.
The first subdivision is age:
A young man, whose passions are warm,
will be more sensibly touched with amo-
rous and tender images, than a man more
advanced in years, who takes pleasure in
wise, philosophical reflections concern-
ing the conduct of life and moderation of
the passions. At twenty, Ovid may be the
favourite author; Horace at forty; and
perhaps Tacitus at fifty. Vainly would
we, in such cases, endeavour to enter
into the sentiments of others, and divest
ourselves of those propensities, which
are natural to us. (E 244)
The second subdivision is character-type:
One person is more pleased with the
sublime, another with the tender; a third
with raillery. One has a strong sensibility
to blemishes, and is extremely studious
2/9/2012 9:37:20 PM

of correctness: Another has a more lively
feeling for beauties, and pardons twenty
absurdities and defects for one elevated
and pathetic stroke. . . . [It] is almost
impossible not to feel a predilection for
that which suits our particular turn and
disposition. Such preferences are inno-
cent and unavoidable, and it can never
reasonably be the object of dispute,
because there is no standard, by which it
can be decided. (E 244)
The second main division of human diver-
sity is the ‘particular manners and opinions
of our age and country’ (E 243). This Hume
explicates as follows:
[We] are more pleased in the course of
our reading, with pictures and charac-
ters, that resemble objects which are
found in our own age or country, than
with those which describe a different set
of customs. . . . For this reason, comedy
is not easily transferred from one age
or nation to another. A Frenchman
or Englishman is not pleased with
the Andria of Terence, or Clita of
Machiavel; where the fine lady, upon
whom all the play turns, never once
appears to the spectators, but is always
kept behind the scenes, suitable to the
reserved humour of the ancient Greeks
and the modern Italians. (E 244–5)
Thus Hume’s overall position is that even
though some eccentric opinions can be ruled
out, we must expect some explicable diversity
and disagreement amongst the remainder.2
2. RULES: A CRITIQUE
Hume introduces the idea of ‘rules of com-
position’ immediately after the paragraph
353
9780826443595_Ch19_Final_txt_print.indd 353
‘OF THE STANDARD OF TASTE’
about Ogilby and Milton. The way he men-
tions these rules suggests they have been dis-
cussed in the essay before, whereas this is in
fact their first appearance:
It is evident that none of the rules of com-
position are fixed by reasonings a priori,
or can be esteemed abstract conclusions of
the understanding, from comparing those
habitudes and relations of ideas, which are
eternal and immutable. Their foundation
is the same with that of all the practical
sciences, experience; nor are they anything
but general observations, concerning what
has been universally found to please in all
countries in all ages. (E 231)
Hume gives no examples of these inductive
rules at all, but fortunately a certain amount
can be deduced from the continuation of the
paragraph:
But though poetry can never submit to
exact truth, it must be confined by rules
of art, discovered to the author either by
genius or by observation. If some neg-
ligent or irregular writers have pleased,
they have not pleased by their transgres-
sions of rule and order, but in spite of
these transgressions: They have possessed
other beauties, which are conformable
to just criticism; and the force of these
beauties have been able to overpower
censure, and give the mind a satisfaction
superior to the disgust arising from the
blemishes. Ariosto pleases: but not by
his monstrous and improbable fictions,
by his bizarre mixture of the serious and
comic styles, by the want of coherence in
his stories, or by the continual interrup-
tions of his narrative. (E 231–2)
On the basis of this passage, it would be rea-
sonable to think that the following are rules
of composition:
2/9/2012 9:37:21 PM
 ‘OF THE STANDARD OF TASTE’
• A work should be written with care and
attention
• A work should be regular
• Fictions should not be monstrous
• Fictions should be probable
• The serious and comic should not be mixed
(or should not be bizarrely mixed)
• Stories should be coherent
• Stories should not be continually
interrupted
In all these assertions, Hume is unconsciously
constrained by the taste of his time, and none
of them now seems compelling. Lack of care
and attention (rather than just, shall we say,
an intention to convey a sense of improvisa-
tory freedom) seems to be part of the aes-
thetic value of the stanza Byron scrawled
out on the back of Don Juan’s manuscript.
It ends:
And for the future – (but I write this
reeling
Having got drunk exceedingly today
So that I seem to stand upon the ceiling)
I say – the future is a serious matter –
And so – for Godsake – Hock and Soda
water!3
Lack of regularity is one of the central fas-
cinations of Beethoven’s piano Phantasie in
G minor op. 77. Hugh Macdonald describes
it as ‘the most violently disconcerting of any
[of Beethoven’s compositions]’: ‘It begins in
G minor and ends in B major; in between
it passes through D minor, Ab major and
many other unrelated keys. It has no the-
matic skeleton, the tempo fluctuates wildly.
Pauses, cadenzas, flourishes and violent
changes of dynamic and direction abound.’4
And yet it is quite clear that violently irregu-
lar and disconcerting changes are part of
what the critic praises: ‘[Beethoven’s] genius
consists precisely in his capacity to astonish
and surprise . . . .’5
354
9780826443595_Ch19_Final_txt_print.indd 354
It is just as easy to find widely held criti-
cal judgements that undermine Hume’s
other rules. Caliban and Satan – two obvi-
ously monstrous elements – play essential
roles in the greatness of The Tempest and
Paradise Lost; the improbabilities of Alice in
Wonderland play their part in its profundity
and charm; Joyce’s Ulysses derives part of
its power from the bizarre mix of the tragic
and comic; lack of coherence makes the
story of Kubrick’s 2001 wonderfully com-
pelling; and it is the continual and fantas-
tic interruptions which make both Tristram
Shandy and many of Billy Connolly’s ram-
bling anecdotes so funny. In sum, the advent
of both Romanticism and Modernism have
swept away all of Hume’s proposed ‘rules of
composition’.
The central difficulty with any ‘rules of
composition’, not just Hume’s, is that they
are not rules at all; they are at best inductive
generalizations. And if they are generaliza-
tions, it is reasonable to ask how we know
they are correct, how we know we are in
possession of all the relevant generalizations,
and how any generalization of this kind can
have normative force.
To make these points clearer, let us sup-
pose – contrary to the argument given
above – that every work of art hitherto
enjoyed by human beings has been regular,
coherent and probable. Now suppose you
find yourself gripped by a work which is
not, by any stretch of the imagination, regu-
lar, coherent and probable. Why should your
interest in the current work die just because
your attention is now sustained by qualities
which have never sustained anyone’s interest
before? Why should you not have discovered,
contrary to all expectation, that human inter-
est can be sustained by at least one irregular
and improbable work? In the same way, if
you find yourself enjoying some particularly
2/9/2012 9:37:21 PM

bitter food, why should you be disposed to
stop enjoying it just because someone points
out that no one has enjoyed such bitter food
in the past? Both the objections about regu-
larity and bitterness may give you pause, they
may make you reflect on the possibility that
you may not have fulfilled one or more of the
judgement conditions (or made some other
kind of error), but if, after an appropriately
thorough self-interrogation, you decide this
is not the case, why should you not continue
with your pleasure? Why indeed should you
not start evangelizing on its behalf?
The objection can be put in another way.
How do we know that a principle of compo-
sition we are operating with is correct? Only
because it picks out works of art, or aspects
of works, we consider meritorious. But this
of course means that we can judge the value
of works of art independently of any such
principle. The point is reinforced when we
ask ourselves how the first person to decide
on the merit of a work of art came to his
conclusion: clearly not by the use of induc-
tive principles founded on his experience of
previous works of art. And if he can judge
the first work of art without the benefit of
principles, why can he not judge the second,
third and fourth by the same method? Both
arguments in this paragraph show that we
can judge the value of any work of art with-
out the use of a principle; indeed, the use
and value of the principle is entirely unclear.
Hume, in one passage, appears to concede
this point:
Did our pleasure really arise from
those parts of [Ariosto’s] poem, which
we denominate faults, this would be
no objection to criticism in general: It
would only be an objection to those par-
ticular rules of criticism, which would
establish such circumstances to be faults,
355
9780826443595_Ch19_Final_txt_print.indd 355
‘OF THE STANDARD OF TASTE’
and would represent them as universally
blameable. If they are found to please,
they cannot be faults; let the pleasure,
which they produce be ever so unex-
pected and unaccountable. (E 232)
If one thinks that all criticism must involve
rules, then one reading of this passage leaves
open the possibility that a correct rule can
always be substituted for an incorrect rule;
but a more likely interpretation is that aes-
thetic responses and criticism of them can be
conducted in the absence of any rules at all.
Of course, Hume can claim that if your
hypothetical work is regular, coherent and
probable it is more likely to be of interest
than a work which is irregular, incoherent
and improbable. But it is quite clear that
Hume intends his rules of composition to be
of use in evaluating the very work which now
stands in front of us. In these circumstances,
what is the point of saying a work which
conforms to his rules is more likely to give
pleasure and interest when all parties agree
that this particular work, which conforms
to his rules, manifestly does not? Naturally
enough, there can be critical argument about
whether the work before us is interesting and
absorbing, or about why it is interesting and
absorbing, but it is equally clear that this
argument will not involve rules of the kind
which Hume invokes.
Hume tries to defend the idea that ‘gen-
eral rules of art’ (E 232) underlie (or at least
describe) well-grounded aesthetic reactions,
by considering a further example:
The same Homer, who pleased at Athens
and Rome two thousand years ago, is
still admired at Paris and at London.
All the changes of climate, government,
religion, and language, have not been
able to obscure his glory. Authority or
2/9/2012 9:37:21 PM
 ‘OF THE STANDARD OF TASTE’
prejudice may give a temporary vogue to
a bad poet or orator; but his reputation
will never be durable or general. When
his compositions are examined by pos-
terity or foreigners, the enchantment is
dissipated, and his faults appear in their
true colours. (E 233)
‘It appears then,’ concludes Hume, that ‘there
are certain general principles of approbation
or blame, whose influence a careful eye may
trace in all operations of the mind.’ (ibid.)
But his Homer example is not a case of a rule
or general principle always holding good,
or of a certain kind of work always being
admired; it is an example of two particular
works – the Iliad and the Odyssey – always
being admired. These examples do no more
than show that, in some cases, human beings
value the very same work over a long period;
they do nothing to show that the works’
achievements depend on conforming to a
certain number of rules or principles.
It is also worth pointing out that, even if
value-judgements do need to be supported
by general reasons, it is not at all clear that
the same general reasons are needed to sup-
port similar value-judgements. Shakespeare
has always been held in high esteem, but
the reasons for this judgement have changed
substantially over the course of time. In the
seventeenth century he was often praised for
being able to point out a valuable moral les-
son; in the eighteenth for holding a mirror up
to nature; in the nineteenth for psychological
insight into character; and in the early to
mid-twentieth for his ability to weave dense
metaphoric patterns throughout his plays. In
the same way, one can imagine a work being
held in high regard across all cultures, but for
completely different reasons. Thus, a univer-
sally high valuation may be explained, not by
uniformity in the principles people apply, but
356
9780826443595_Ch19_Final_txt_print.indd 356
by the application of different or continually
changing principles.
3. SANCHO’S KINSMEN: A
MISLEADING ANALOGY
Hume makes a more thoroughgoing attempt
to show the value and role of rules in critical
discussion when he discusses the story about
Sancho Panza’s kinsmen.6 In this incident
from Cervantes, two of Sancho’s relatives
are asked for their opinion on a hogshead
of wine. One pronounces the wine excel-
lent apart from a slight taste of leather; the
other pronounces it excellent except for a
faint taste of iron. Both are heartily laughed
at. But when the hogshead is emptied, an old
key with a leather thong is found at the bot-
tom of it. Their delicacy of taste is then uni-
versally acknowledged. (E 234–5)
Hume retells the anecdote to illustrate
what delicacy of physical taste is, and to sug-
gest what delicacy of mental taste might be.
However, he also uses it to show the role and
value of rules in critical discussion: ‘Here then
the general rules of beauty are of use; . . . To
produce these general rules or avowed pat-
terns of composition is like finding the key
with the leathern thong; which justified the
verdict of Sancho’s kinsmen, and confounded
those pretended judges who had condemned
them.’ (E 235) In fact, the anecdote about
Sancho’s kinsmen does not provide a help-
ful model of the critic’s activity, but, before
this can be demonstrated, two adjustments
need to be made to the story. First, in the
original tale, the kinsmen’s delicacy of taste
could have been confirmed by the emergence
of something other than a key with a leather
thong: it could have been confirmed by the
emergence of an all-metal penknife and a
wallet, or any number of other objects. To
2/9/2012 9:37:21 PM

streamline my discussion, I shall from now
on assume that they said the wine was key-
flavoured. Second, we need to assume that
the kinsmen really did detect a peculiar taste
in the wine, the emergence of the key is not
a lucky accident, and they do not have some
other method for knowing there is a key in
the hogshead (one of them recalling he put it
there earlier, for example).
Now, suppose that further experiments
with objects in hogsheads are carried out.
Sancho’s kinsmen posit all kinds of foreign
objects in the wine – tobacco quids, dead
mice, pickled eggs and so forth – which,
when the hogsheads are emptied, always turn
out to be present. Once this ability has been
demonstrated any number of times, then the
other men in the room who have tasted wine
from the contaminated hogsheads, but have
been unable to tell it apart from uncontam-
inated wine, may well conclude that the kins-
men’s taste is well-nigh infallible. These other
drinkers may also decide that, in future, they
will rely entirely on the kinsmen to determine
when and how wine has been adulterated. If
the room also contains ‘by-standers’ (E 236),
who have not tasted the wine, then they too
may come to the same decisions as the drink-
ers. The behaviour of both groups would be
entirely rational, and we can see how the
various discovered objects would furnish
proofs of the kinsmen’s claims.
If Hume’s account of criticism is to be on
all fours with this model, then we need to
imagine a situation where a group of people
are asked to assess poems; some of them read
the poems, some of them do not. Suppose the
readers consider a poem, decide that it is hope-
less, and then ask a convenient critic for his
opinion. He declares that, on the contrary, the
poem is excellent and then goes on to show
that it conforms to a well-established rule of
composition. The critic’s opinion is, therefore,
357
9780826443595_Ch19_Final_txt_print.indd 357
‘OF THE STANDARD OF TASTE’
held to be right, and the readers’ wrong. After
this, and a good deal more trying and failing,
the group may simply give up trying to assess
poems, and rely on the critic and his rules to
determine which poems are good and bad. The
non-readers may also be impressed by these
outcomes, and decide that they too will rely on
the critic rather than the readers to determine
which poems are good and which are bad. But
in this story, the behaviour of both groups is
entirely irrational. If the readers satisfy Hume’s
five judgement conditions, one would want to
ask them: why do you believe that this critic is
as expert as he claims to be? Why should you
take his opinion to be better than yours? In
what way do these rules and principles act as
proofs? Who decides on these rules? And how
do you know that these rules are correct? One
would want to ask the non-readers at least
two additional questions: how can you claim
to know that a poem is good or bad without
having read it? Is doing this not akin to lying?
4. SANCHO’S KINSMEN: THE
FUNCTIONS OF CRITICAL
ARGUMENT
To move the Cervantes example rather closer
to art criticism, consider another variation
on the story: the key-taste is present but the
key is not. This could be either because the
key has been removed, or because some unre-
lated chemical process made the wine taste
of key. Let us also assume that there is no
objective evidence of any kind for the pres-
ence of the key-taste in the wine (no signed
statements from a person who removed a
key from the hogshead, no CCTV footage of
the key’s removal, no positive scientific tests
for the relevant chemicals, etc.). How in this
case do Sancho’s kinsmen demonstrate to the
other men that the wine is key-flavoured?
2/9/2012 9:37:21 PM
 ‘OF THE STANDARD OF TASTE’
The only method is for the kinsmen to edu-
cate the others’ taste so that they too come to
taste the inappropriate flavour in the wine.
This is a much more ambitious and labour-
intensive undertaking than merely emptying
the hogshead, because in this case the kinsmen
have to ensure that the others acquire (some
of) their own expertise. It is quite likely that
this attempt to teach the others will fail; the
outcome will certainly be less decisive than the
original story’s emptying of the hogshead; and
the new demonstration will be effective, not on
mere ‘bystanders’, but only on people prepared
to taste the wine for themselves. Although
this method is unreliable and exhausting, it is
also the only one that can prove conclusive.
To other possible ‘proofs’ someone can say:
‘I heard you say you took a key out of the
wine, I’ve seen the CCTV footage of it being
removed, I’ve witnessed the positive chemical
tests for key-flavour – even ‘I’ve seen the key
in the hogshead’ – but the wine still doesn’t
taste of key.’ Only making him taste the key-
flavour will lead to a sincere withdrawal of the
last clause in this statement.
A good wine critic does not merely inform
or predict; he gets you to taste things in the
wine you could not taste before. Similarly, the
good music critic does not just inform or pre-
dict; he gets you to hear things in pieces you
had not noticed hitherto. They will use the
same kind of methods: asking questions (‘Can
you taste the slightly musty after-taste?’, ’Can
you hear the octaves in the bass?’); pointing
(‘Notice the hints of citrus’, ‘Listen to the
strings’ harmonics’); using analogies (‘There’s
something dark and earthy . . .’, ‘That wood-
wind figure is a little turbine of happiness’);
making comparisons (‘This is more like a
Merlot’, ‘This reminds me of fugue-subject
in the Magic Flute overture’). Sometimes the
critic will move beyond words and ask us to
listen to certain other compositions, try other
358
9780826443595_Ch19_Final_txt_print.indd 358
drinks, read critical books, contemplate cer-
tain paintings as we listen, and so on.
Despite his initial statement to the con-
trary, Hume realizes that producing ‘general
rules or avowed patterns of composition’
does not and should not have the same effect
as ‘finding the key with the leather thong;
which confounded those pretended judges’.
We can see this when he is describing how
these general rules must be used to show an
interlocutor he is in the wrong.
But when we show him an avowed
principle of art; when we illustrate this
principle by examples, whose opera-
tion, from his own particular taste, he
acknowledges to be conformable to the
principle; when we prove, that the same
principle may be applied to the present
case, where he did not perceive or feel its
influence: He must conclude, upon the
whole, that the fault lies in himself . . . .
(E 236; emphasis added)
Hume realizes that the interlocutor’s inter-
pretation and evaluation cannot change
while his experience remains unaltered:
hence Hume’s emphasis on ‘[the interlocu-
tor’s] own particular taste’ and factors which
the interlocutor ‘acknowledges’. It also
explains Hume’s sudden plunge into the past
tense ‘where he did not perceive or feel [the
principle’s] influence’; if the interlocutor still
does not perceive or feel the principle’s influ-
ence (or as I would prefer to say, the critic’s
influence) but merely believes it to be true,
then his assent to the new interpretation and
evaluation can only be insincere. This pas-
sage shows that Hume has at least a dim
intimation that critical remarks convince in
an entirely different way from the discovery
of the key in the original Cervantes story. In
the latter case, as we have seen, both drinkers
2/9/2012 9:37:21 PM

and non-drinkers could have been convinced
by the kinsmen’s expertise without having
their own experience altered in any way.
5. THE RATIONALITY OF CRITICAL
ARGUMENT
If Hume’s account of the role of rules par-
tially occludes the truth of his subjectivism,
why does he introduce them into his account
of aesthetic judgement? One reason is merely
historical (he lived in a society where the
neo-classical rules of Rymer and Dubos were
much discussed); others are merely psycho-
logical or sociological (it is comforting to
know that one’s own taste, or the taste of
one’s class, is not self-sustaining, but exter-
nally sanctioned by widely acknowledged
rules and principles).
There may, however, be a deeper reason.
In his essay, ‘The Sceptic’, Hume observes:
‘There is something approaching to principles
in mental taste: and critics can reason and
dispute more plausibly than cooks or perfum-
ers.’ (E 163) It seems reasonable to conclude
from this, although Hume does not quite say
as much, that he believes critics can reason
plausibly about art because there is some-
thing approaching to principles in this case,
whereas cooks and perfumers cannot reason
about the phenomenology of food or per-
fume because such principles are lacking. In
other words, because Hume does not have a
clear picture of what critical reasoning looks
like, he assumes it must follow the pattern of
deductive and inductive argument, and this
requires the existence of general principles.
Hume is quite right to think critics can
engage in rational disputes about one
another’s views, but wrong to think that
this requires the existence of principles.
359
9780826443595_Ch19_Final_txt_print.indd 359
‘OF THE STANDARD OF TASTE’
Philosophers and logicians are concerned
with abstract thought. They operate with
principles and inferences because they want
to help us arrive at true or probable propo-
sitions. Critics, on the other hand, are con-
cerned with imagination and imagination
as it is embedded in perception. They rede-
scribe, point and compare because they want
to massage, adjust or explain our experience.
A critic has not done his work if you merely
come to believe with a high degree of cer-
tainty that a certain proposition is true; he
has only succeeded when your experience has
been adjusted or explained. (Although some
of these changes and explanations can then
be reported in propositions.) It is not enough
to believe that a piece of music is melan-
choly; you have to hear its melancholy. It is
not enough to know you ought to like it; you
have to like it. This makes critical arguments
both more ambitious and more likely to fail
than inductive or deductive arguments.
Critical arguments are also constrained by
rationality, but it is not the kind of rationality
we associate with induction and deduction.
Consider a number of suggestions as to what
an object can be seen as, and the methods
which might be used to make us see the rele-
vant aspects. We have no difficulty in detect-
ing which are sane and sensible, which are
strained, and which are mad and impossible.
Thus I have no difficulty seeing Wittgenstein’s
duck/rabbit7 as a duck or a rabbit; I have a
little more difficulty in seeing it as a hare or
a goose; I cannot see it as a harp or a foot-
ball at all. At a more metaphorical level, I can
imagine agreeing with a critic who describes
Dawn from Britten’s Four Sea Interludes as
‘calm’ and ‘serene’, possibly as ‘tense’ and
‘expectant’, but not as ‘raging’ or ‘jolly’. There
is irrationality in sincerely asserting that the
duck/rabbit can be seen as a harp, or that
Britten’s Dawn can be heard as jolly, but it is
2/9/2012 9:37:22 PM
 ‘OF THE STANDARD OF TASTE’
not the irrationality of shaky principles and
invalid inferences. Indeed, because criticism
deals with a deeper level of the psyche than
abstract thought, the two cases of inappropri-
ate perceiving-as come much closer to mad-
ness. In sum, and contra Hume, we can see
that the existence of rational critical disputes
gives no support whatever to the idea that
there are or must be critical principles.8
I have tackled the question of why there can
be effective rhetoric about sights and sounds
but not about smells and tastes elsewhere, and
I do not have space to rehearse my arguments
here.9 Accordingly, I shall just state my earlier
conclusion. If the object of a sense has distinct
temporal or spatio-temporal parts then this
has two relevant consequences. First, these
parts will be capable of depicting – rather than
simply representing – other objects. Second,
the imagination will be able to arrange these
parts into different patterns that a critic can
prompt us to see. Consequently, the objects of
sight, hearing and touch, which do have dis-
tinct parts, can both depict and be effectively
rearranged by critical rhetoric; the objects of
smell and taste, which do not have distinct
parts, cannot depict, and cannot be effect-
ively rearranged by critical rhetoric. This is
why perfumers and cooks cannot reason as
effectually about the phenomenology of their
products as painters and musicians can about
theirs. Once again, we see that the existence
of criticism and disputation does not imply or
require principles.
6. TRUE JUDGES AND EXCELLENT
CRITICS
As I mentioned in Section 1, Hume eventually
defines his standard of taste with reference to
the decisions of qualified critics and without
360
9780826443595_Ch19_Final_txt_print.indd 360
reference to rules: ‘Strong sense, united to
delicate sentiment, improved by practice, per-
fected by comparison, and cleared of all preju-
dice, can alone entitle critics to this valuable
character [of the true judge]; and the joint
verdict of each, wherever they are found, is
the true standard of taste and beauty.’ (E 241)
How well does this method work?
Hume does not distinguish between a true
judge and a critic but it is useful to do so
now. Clearly, there could be someone who
fulfils all five of Hume’s judgement condi-
tions to an eminent or outstanding degree,
but is completely hopeless at accounting for
his own taste or persuading other people to
share it. I think we would be quite happy to
describe such a person as being in a position
to judge works of art, as possibly even a man
of taste, but we would not be happy to call
him a critic. To merit this title, someone must
usually be in a position to articulate why he
feels as he does, and be able to persuade oth-
ers to share his taste and outlook.
One way to determine the standard of taste
might be to observe what the true judges do,
which works of art they like to contemplate
and avoid, and then decide that these pref-
erences determine what is worth looking at
and what is not. Hume states that it is ‘the
joint verdict’ of true judges (and this, I take
it, can be determined by behaviour other than
linguistic behaviour) which determines what
the standard of taste is. But the same objec-
tion can be brought against treating such
a verdict as the standard of taste that has
already been brought against treating criti-
cal principles as determining the standard of
taste: if you are as well qualified as these
judges, why should you feel obliged to make
your opinion fall in line with theirs? Should
you start training yourself to dislike works
you love, and liking works you detest? Why
should you not claim that it is the majority of
2/9/2012 9:37:22 PM

true judges who have got the matter wrong
and not you? After all, the majority of true
judges in the nineteenth century disparaged
Donne and Pope but admired Meyerbeer,
and we are now inclined to feel that figures
such as Browning, Byron and Wagner who
set their faces against these majority tastes
were actually in the right.
Of course, the unanimity of the true judges’
opinion might make you pause for thought.
You may ask yourself what they are seeing in
works you dislike; you might wonder whether
you have sufficiently fulfilled the judgement
conditions; you might question whether
your mind is in an appropriately receptive
state. But if at the end of a self-examination
and self-education programme you still find
yourself repelled by works which the judges
value highly, you may well come to the con-
clusion that they are in the wrong.
Hume also seems remarkably confident
that the verdicts of individual true judges will
form a consensus. But his five conditions, as he
himself points out, lay down some very mini-
mal conditions for judgement. We can thus
expect a highly diverse set of people to meet
them, and also expect that these people will
hold a very wide range of legitimate aesthetic
opinions. We may find that qualified critics
break into a number of competing schools
(as in the dispute in mid-nineteenth-century
Germany over ‘the Music of the Future’), or
we may only find a scattering of individual
opinions which are not homogeneous enough
to be classified into groups at all. In these
cases there is no ‘joint verdict’ to follow.
If there is no majority opinion or consen-
sus, how should we decide which true judge
or judges to follow? We could try discover-
ing which judges fulfil the judgement criteria
to the most eminent or outstanding degree.
But consider a case of two judges, one of
whom has 30 years’ experience of evaluating
361
9780826443595_Ch19_Final_txt_print.indd 361
‘OF THE STANDARD OF TASTE’
Baroque keyboard suites, the other 25. It
would be slightly absurd to favour the first
judge just because he has five more years’
experience; after a certain saturation point,
we are rightly inclined to treat all judges as
being on the same footing. In any case, even if
we select a judge as being outstandingly well
qualified, a familiar problem arises when we
find ourselves disagreeing with him: cannot
a person better qualified than us make mis-
takes and have his incapacities and blind-
spots? Why do we have to feel that he is right
and we are wrong? Might we not be able to
persuade him? And so on.
If we want to judge which works of art
are truly great, perhaps we should listen to
other people’s explanations and arguments
as to why they have chosen one work over
another. This will involve withdrawing our
attention from people who are just true
judges, and turning it towards those who are
both true judges and excellent critics.
Hume says that identifying a true judge is
a matter ‘of fact, not of sentiment’ (E 242).
It is certainly true that there are objective
tests for how many books someone has read,
whether he has motives for bias, whether he
can notice small differences, and so on. But
how does one recognize an excellent critic?
Hume offers only the following observation:
Though men of delicate taste be rare,
they are easily to be distinguished in
society, by the soundness of their under-
standing and the superiority of their
faculties above the rest of mankind. The
ascendant [i.e. ascendency], which they
acquire, gives a prevalence to the lively
approbation, with which they receive
any production of genius, and renders it
generally predominant. (E 243)
The emphasis on society, achieving ascend-
ency and making a view predominant,
2/9/2012 9:37:22 PM
 ‘OF THE STANDARD OF TASTE’
suggests that we are here concerned with
identifying good critics rather than people
who are just true judges, but Hume’s way
of identifying the former is deeply unsatis-
factory. There are plenty of people of sound
understanding and superior faculties who
have no gift for criticizing or appreciating
art. In addition, it is easy to say that great
critics respond with lively approbation to
‘productions of genius’, but what of those
whose response to inferior works is equally
lively and enthusiastic and who manage to
persuade others to share their opinion? What
of those who were distinctly unenthusiastic
about works which were later acknowledged
to be masterpieces? Just as it is possible to
feel that a certain work of art – for example,
Hesse’s The Glass Bead Game in the 1960s
and 1970s – acquired unwarranted popu-
larity and praise, so it is possible to feel that
a critic – for example, Leavis in England in
the 1960s and 1970s – attained a degree of
influence that was out of all proportion to
his merits. How do we tell the difference
between a work which is good rather than
merely popular, or a critic who is truly great
rather than merely influential?
The test of a work of art is how well and
how long it holds, absorbs and interests me;
and the test of a critic is how profoundly
he alters my responses to a work of art, or
makes me better understand the responses
I already have. For example, I did not par-
ticularly enjoy my first readings of Austen’s
Mansfield Park (it seemed pious and wor-
thy) and Flaubert’s Bouvard and Pecuchet
(it seemed flat and inert). I then read Lionel
Trilling’s essays on these novels, and they sud-
denly came to seem interestingly problematic
or touchingly human.10 On the other hand,
several years ago I read through a good deal
of the music criticism of J.W. Davison – the
once-famous mid-nineteenth-century critic
362
9780826443595_Ch19_Final_txt_print.indd 362
of The Times – and found that, at the end of
the process, he had said nothing which either
explained or altered my feelings about the
music discussed.11 Consequently, although
both critics were admired and immensely
influential in their day, I am inclined to
think that Trilling deserved (and deserves)
his reputation and Davison did not. This not
only means that there is an element of sub-
jectivity in determining who the best critics
are, it means that the subjective response
that identifies great critics is parasitic on
the subjective response that identifies great
works of art.
7. CONCLUSION
Hume shows that some people are quali-
fied to speak about art and some are not,
and some are better qualified to speak than
others. But the idea that there is a standard
of taste over and above the various opinions
of qualified people – their joint verdict, for
example, or acknowledged principles – is
illusory and contrary to the spirit of Hume’s
own subjectivism. Any attempt to cut the
Gordian Knot of critical disputation puts an
end to the distinctive profundity, expressiv-
ity and richness of aesthetic experience, and
transforms it into intellectual acquiescence to
received and groundless opinion.12
NOTES
1
For a fine article focusing on the tension
between rules and decisions in Hume’s essay,
see Jonathan Friday, ‘Hume’s Sceptical Standard
of Taste’, Journal of the History of Philosophy
36(4) (1998), pp. 545–66.
2
There are a number of contentious points
in Hume’s discussion of subjectivism which
2/9/2012 9:37:22 PM

8
I do not have space to examine here. For
example, his understanding of the aesthetic,
the analogy between aesthetic taste and colour
perception, the link between aesthetics and
morality, and whether his five judgement
conditions require supplementation and
9
refinement.
3
Byron, Don Juan, ed. B. Lee (London and
Glasgow: Collins, 1969), p. 172.
4
Hugh Macdonald, ‘Beethoven’s Game of Cat
and Mouse’, in Beethoven’s Century: Essays
on Composers and Themes (Rochester, NY:
10
Rochester University Press, 2008), pp. 3–15, p. 6.
5
Ibid., p. 8.
6
Sections 3 and 4 expand on two compressed
and brilliant paragraphs (pp. 87–8) in Stanley
Cavell’s ‘Aesthetic Problems of Modern
11
Philosophy’, in his Must We Mean What We
Say: A Book of Essays (Cambridge: Cambridge
University Press, 1976 [1969]), pp. 73–96.
7
See L. Wittgenstein, Philosophical
Investigations, trans. G.E.M. Anscombe, 3rd
12
edn (Cambridge: Cambridge University Press,
2001 [1953]) p. 166.
363
9780826443595_Ch19_Final_txt_print.indd 363
‘OF THE STANDARD OF TASTE’
For a fuller account of the nature of critical
argument, and why there are no critical
principles, see M.W. Rowe, ‘Criticism without
Theory’, in his Philosophy and Literature: A
Book of Essays (Aldershot: Ashgate, 2004),
pp. 22–45.
See M.W. Rowe, ‘The Objectivity of Aesthetic
Judgements’, British Journal of Aesthetics
39(1) (1999), pp. 40–52. The present article is
more sympathetic to the notion of intelligent
and effective wine criticism than the 1999
paper.
Trilling’s essays are: ‘Flaubert’s Last
Testament’ and ‘Mansfield Park’, both in his
The Opposing Self: Nine Essays in Criticism
(Oxford: Oxford University Press, 1980
[1955]), pp. 152–80, 181–202.
A selection of Davison’s more vituperative
criticism can be found in Charles Reid, The
Music Monster: A Biography of James William
Davison (London: Quartet Books, 1984),
pp. 141–234.
I would like to thank Marie McGinn for
helpful discussions of this article.
2/9/2012 9:37:22 PM
 20
HUME ON HISTORY
Timothy M. Costelloe
1. INTRODUCTION
Hume’s verdict that the Treatise suffered
from the exuberances of youth is well
known, as is his lament that it ‘fell dead-born
from the press’, a fate that led him in sub-
sequent years to recast his philosophy in the
form of inquiry, dialogue and essay, literary
forms more conducive to a wider and – as the
century progressed – increasingly educated
middle-class readership. While this redirec-
tion brought Hume both literary fame and
financial independence (he reports that by
the 1760s he had ‘become not only independ-
ent, but opulent’ (E xxxviii)), it remains true
that for much of his lifetime his reputation
rested primarily on the History of England.
In his concise autobiography written shortly
before his death, Hume indicates that he
only conceived the project after becoming
Librarian to the Faculty of Advocates in
1752; the two volumes on the Stuarts were
published in 1754 and 1756 (dated 1757),
the two on the Tudor period in 1759, and the
medieval volumes covering the period from
the invasion of Julius Ceasar to the acces-
sion of Henry VII in 1761 (dated 1762). It is
likely, however, that Hume had thought long
about the work: many of his other writings,
including the Treatise, Enquiry concerning
the Principles of Morals, Natural History of
364
9780826443595_Ch20_Final_txt_print.indd 364
Religion and a good number of the essays
show Hume’s command of historical mate-
rial as well as his interest in the practical and
theoretical issues confronted by the practis-
ing historian. It is not an exaggeration to
say, as Donald Livingston puts it, that ‘from
the beginning and throughout his career as
a writer, [Hume] was engaged in historical
work as well as in the philosophical prob-
lems to which such work gives rise’.1
Hume’s long-standing interest notwith-
standing, some interpreters have tended to
regard the History as a turn away from phil-
osophy, and even judged the approach of the
work to be ‘anti-historical’, especially given
the focus of the Treatise, which in anatom-
izing the fundamental features of human
nature seems to preclude the possibility of
appreciating manners, customs and institu-
tions as the outcome of historical develop-
ment.2 Other commentators, by contrast,
have argued for the fundamental continuity
in Hume’s concern with philosophy and his-
tory, understanding them as mutually sup-
porting components of his overall approach
to the ‘science of man’. Ernest Mossner, for
example, writes that for Hume philosophy
and history ‘are closely akin because the
development of the human mind, which it
is the historian’s task to trace, provides the
materials from which the philosopher derives
2/9/2012 9:37:28 PM

the very principles of thinking and conduct’;3
Livingston emphasizes that ‘Hume consid-
ered his historical writings as an application
and extension of this philosophical work’,4
and in a similar vein, Gregory Moses argues
that for Hume the ‘roles of historian and phil-
osopher compliment each other and in some
places even overlap’, so that any differences
between the two come down to a difference
in emphasis rather than one of kind. Both
concern ‘historical events and principles’,
Moses urges, pursued as a result of natural
curiosity and ‘what counts as an explanation
in one or the other is also the same’.5
In what follows, I want to pursue this lat-
ter view by focusing on what Hume terms
‘philosophical’ or ‘true’ history. This concept
has three distinct components, which I shall
discuss in turn. First, Hume emphasizes that
history should be beholden to evidence and
thus avoid (as far as possible) corruptions of
the record based on prejudice and partiality.
Historical writing reflects the power of mem-
ory to present and preserve matter of fact in
contrast to the imagination that represents
and subverts the same. Hume thus juxtaposes
the memorial nature of history, which reflects
events as they happened, with the inventive
character of poetry that creates ideas of fic-
tional objects; when history ceases to be philo-
sophical it is because it becomes poetic by
mixing fiction with fact. Hume is also aware,
second, that reflecting the past accurately is
complicated by the nature of the evidence
upon which the historian draws, viz., past
events are remote and the testimony bear-
ing witness to them is itself an interpretation
of the record. As a result, while historians
must distance themselves from the fictions of
poetry, a successful narrative is unavoidably a
reconstruction or representation of the past.
Historians are thus obliged to draw on the
very faculty – the imagination – that threatens
365
9780826443595_Ch20_Final_txt_print.indd 365
HUME ON HISTORY
to undermine their enterprise and as such are
successful only to the degree that they prod-
uce ideas with sufficient force and vivacity to
move the reader and inspire belief: they must
transform ordinary experience, arouse agree-
able ideas in an audience, and do so delib-
erately according to a plan or design. Third,
while it is difficult to manage this demand
for balancing an accurate (memorial) presen-
tation of the past with a vivacious (imagina-
tive) representation of it, when historians do
so effectively they open up the past as a rich
source of knowledge and wisdom that can
guide present conduct: the historian is at once
a moral scientist uniquely positioned to dis-
cover the principles of human nature, frame
rules to be instantiated in pursuit of a well-
ordered society, and paint models of virtuous
conduct for emulation.
2. PHILOSOPHICAL HISTORY AND
MATTER OF FACT
The first aspect of philosophical history is
an application of Hume’s empiricist prin-
ciple that all knowledge in the form of ideas
can be traced to the experience of matter of
fact, which in the case of history is manifest
as evidence of past events.6 The philosoph-
ical historian reflects the power of memory,
the ‘faculty, by which we raise up the images
of past perceptions’ (THN 1.4.6.18 / 260)
and in so doing ‘preserves the original form,
in which its objects were presented’ (THN
1.1.3.3 / 9). As individuals form beliefs on the
basis of inference from experience, so the his-
torian applies probabilistic reasoning to avail-
able evidence and draws conclusions about
the likelihood of events having happened
one way rather than another. Historians
thus memorialize the past by presenting and
2/9/2012 9:37:28 PM
 HUME ON HISTORY
preserving events, as far as possible, in their
original form, according to the due pos-
ition and temporal sequence in which they
occurred. This contrasts with the effects of
imagination, the faculty in which ‘perception
is faint and languid, and cannot without dif-
ficulty be preserv’d by the mind steady and
uniform for any considerable time’ (THN
1.1.3.1 / 9). Hume identifies this faculty with
the creative activity of artists, through which
they compound, transpose, augment and
diminish materials gained from experience
to produce fictions that do not correspond to
any real existence, and with aesthetic recep-
tivity, which explains how ideas, albeit of
fictional objects, warm and please artist and
audience alike, tempting both to regard what
is false as having real referents in experience.
Imagination, then, subverts experience and
undermines belief, and when historians err, it
is because they allow the power of imagin-
ation to encroach on the domain of memory;
history becomes poetic when practitioners
either unwittingly or by design allow fictions
to substitute for matter of fact. The same is
true of philosophers, Hume points out, for
when they foster some ‘predominant inclin-
ation’ (EHU 5.1 / 40), prejudice, or bias they
depart from experience and ‘impos[e] con-
jectures and hypotheses on the world’ (THN
Intro. 9 / xviii). ‘When a philosopher has once
laid hold of a favourite principle, which per-
haps accounts for many natural effects,’ as
Hume expresses the same thought elsewhere,
‘he extends the same principle over the whole
creation, and reduces to it every phænome-
non, though by the most violent and absurd
reasoning.’ (E 159)7
When historians free the imagination
in this way, their narratives are corrupted
and their conclusions are false, reflected in
Hume’s time by the shadow of the tradition
of ‘theological’ history that had regarded
366
9780826443595_Ch20_Final_txt_print.indd 366
even the Bible as a historical document.8
Hume acknowledges that some inaccuracies
can be explained as honest mistakes due to
the fact that events are too recent to afford
proper perspective (an explanation for the
blame heaped on the unfortunate Edward II
(H 2.173)), so complex as to have precluded
accurate depiction (as in reports of naval
battles (H 6.277)), or when the ambiguity
of the record prevents definitive clarification
(as with the English Civil War (H 2.469)).
Hume also observes that individuals tend to
admire the past and value it above the pre-
sent, a tendency ‘strongly rooted in human
nature’ and having ‘an influence even on per-
sons endued with the profoundest judgement
and most extensive learning’ (E 464). Other
corruptions of the record, by contrast, are
deliberate distortions, originating in national
prejudice (H 1.84; 6.277; EPM App. 4.17
/ 320, 4.20 / 321–2), the sway of party or
faction (H 1.227), and the power of passion
(H 4.40). He is especially critical of ‘monk-
ish’ historians (H 2.82, 2.88, 2.328 passim),
those medieval chroniclers who turned the
past to their own ends by subordinating
civil transactions to the ecclesiastical and
paraded bombast, inaccuracy, exaggeration,
‘spurious erudition’ (H 2.88, 2.328, 2.477)
and outright ‘invention and artifice’ (H 1.85,
1.132, 1.241) as historical truth. Even great
historians are not immune to such influ-
ences: Tacitus, Hume insists, is hardly cred-
ible when he reports that Vespasian cured a
blind man through his spittle and a lame man
by the touch of his foot (EHU 10.25 / 122),
as Quintus Curtius is ‘much to be suspected’
when he claims to document the ‘supernat-
ural courage’ of Alexander the Great (EHU
8.8 / 84). Even Hume, it might be noted, can
be thought to err when he repeats, apparently
without irony, Herodotus’ account of the
Scythians who ‘after scalping their enemies,
2/9/2012 9:37:28 PM

dressed the skin like leather, and used it as
a towel, and whoever had the most of those
towels was most esteemed among them’
(EPM 7.14 / 255).
Deliberate distortions of this kind, Hume
laments, have involved the ‘history of remote
ages . . . in obscurity, uncertainty, and contra-
diction’ (H 1.3), a state of affairs that can
be rectified, however, when historians ‘con-
fine themselves to strict truth and reality’
(EHU 3.10), matter of fact (E 564), ‘annals’
and ‘records’ (E 123, 204). This is possible
because, at its heart, historical reasoning is
an expression of the general capacity pos-
sessed by human beings to take the ‘gen-
eral point of view’. This enables historians
to distance themselves from events, actions
and characters, and regard the past with a
disinterested, unprejudiced eye, bearing wit-
ness to events as they happened, stripping
away falsehood and allowing ‘all human
race, from the beginning of time, pass as it
were, in review before us; appearing in their
true colours, without any of those disguises,
which, during their life-time, so much per-
plexed the judgements of the beholders’
(E 567). Historians distinguish fact from fic-
tion, as ideas of memory are separated from
those of imagination; they discern the real
shape of events under the clutter with which
contemporary reports and time have effect-
ively masked them. Accordingly, as Hume
emphasizes in the course of discussing the
reported feats of Joan of Arc, history should
‘distinguish between the miraculous and the
marvelous; to reject the first in all narrations
merely profane and human; to doubt the
second; and when obliged by unquestion-
able testimony, to receive as little of it as is
consistent with the known facts and circum-
stances’ (H 2.398). For this reason Hume
says that the ‘first page of Thucydides is
. . . the commencement of real history. All
367
9780826443595_Ch20_Final_txt_print.indd 367
HUME ON HISTORY
preceding narrations are so intermixed with
fable, that philosophers ought to abandon
them, in great measure, to the embellishment
of poets and orators’ (E 422). History should
provide an objective or impartial view of its
subject matter and even has a ‘duty’ to do so
(H 6.140).9
History that approaches a true narrative
also has the advantage of correcting errors
made by other historians. The veracity of the
aforementioned report of Quintus Curtius on
Alexander’s supernatural courage is exploded
as a ‘forgery in history’ when we realize that
the actions ascribed to him ‘are directly con-
trary to the course of nature, and that no
human motives, in such circumstances, could
ever induce him to such a conduct’ (EHU 8.8
/ 84). Where contradictory evidence exists
the benefit of doubt should be given to con-
temporary reports – taking the account of
Scottish historians of Robert the Bruce, for
example (H 2.137) – since those with a stake
in the events reported are more likely to be
better informed and more accurately reflect
the facts of the matter than others whose
interest is of a superficial sort. Of course,
the proximity of events might prejudice the
judgement of contemporary writers, but the
conclusions they draw can be rectified with
the wisdom of hindsight and the perspective
of distance. It is ‘a shameful delusion in mod-
ern historians,’ Hume charges for example,
‘to imagine, that all the ancient princes, who
were unfortunate in their government, were
also tyrannical in their conduct’ (H 2.173–4),
but this is a conclusion that the philosophical
historian is able to condemn and correct.10
Other conclusions can be confirmed accord-
ing to the same principles. The ‘whole tenor
of English history,’ Hume writes for instance,
clarifies the once disputed fact that ‘it would
be difficult to find in all history a revolution
more destructive, or attended with a more
2/9/2012 9:37:28 PM
 HUME ON HISTORY
complete subjection of the ancient inhabit-
ants’ than the Norman Conquest of Britain,
and the vicious character of King John dis-
played in his actions confirms that the ‘dis-
agreeable picture’ of him painted by ancient
historians is not overblown (H 1.453). Even
where there is no direct evidence available,
a general knowledge of historical circum-
stances will support the likelihood of some-
thing having been the case. Hume reasons in
this manner when considering whether well-
born thanes resisted the rise of merchants or
ceorles through the ranks of medieval soci-
ety. ‘Though we are not informed of any of
these circumstances by ancient historians,’ he
suggests, ‘they are so much founded on the
nature of things, that we may admit them as
a necessary and infallible consequence of the
situation of the kingdom during those ages’
(H 1.170).
3. PHILOSOPHICAL HISTORY AND THE
NATURE OF HISTORICAL EVIDENCE
While philosophical history should conform
to matter of fact in the manner urged above,
Hume is also aware that the nature of those
facts throws up barriers to their own inter-
pretation. For a ‘historical fact,’ Hume writes
in the Natural History,
while it passes by oral tradition from
eyewitnesses and contemporaries, is dis-
guised in every successive narration, and
may at least retain but very small, if any,
resemblance to the original truth, on
which it was founded. The frail memo-
ries of men, their love of exaggeration,
their supine carelessness; these principles,
if not corrected by books and writing,
soon pervert the account of historical
events; where argument or reasoning has
368
9780826443595_Ch20_Final_txt_print.indd 368
little or no place, nor can ever recal the
truth which has once escaped those nar-
rations (NHR 1.137).
The peculiarity of history lies in the indirect
access it has to its subject matter. The histo-
rian’s originals are events long gone, tem-
porally and spatially remote and available
only in and through a historical narrative.
Historians thus have to bring the past to life
by performing the function Hume attributes
to the memory when it recalls past impres-
sions by reinvigorating them with something
approaching their original force and vivac-
ity. In so doing, the historian ‘extends our
experience to all past ages, and, to the most
distant nations; making them contribute as
much to our improvement in wisdom, as if
they had actually lain under our observation’
(E 566–7; emphasis added). In performing
this task, the historian effectively annihilates
the distance imposed by past time and inac-
cessible space, and brings events into a vivid
present in which the reader can experience
them. Historians project themselves into the
past, reviving events that compose it and
giving reality to what would otherwise be
inaccessible.
When historians undertake this task of
recalling and making the past present, how-
ever, the originals with which they deal differ
significantly from the impressions of sense
and reflection that are copied by memory,
and this threatens the basis on which beliefs
are ordinarily formed. When the memory
recollects past impressions, it refers dir-
ectly to the immediacy of lived experience,
but the originals of the historian are beyond
living memory; they are reports in the form
of written record that are already images,
or even images of images, copied by others
and available as testimony of events they
never experienced. The difficulty this poses
2/9/2012 9:37:28 PM

is that, according to the copy principle, belief
diminishes as the distance from the original
impression increases, and reasoning over a
long period of time and through many transi-
tions requires having ‘a very strong and firm
imagination to preserve the evidence to the
end, where it passes thro’ so many stages’
(THN 1.3.13.3 / 144). ‘Perhaps, therefore,’
Hume points out, ‘it may be concluded . . .
that the evidence of all antient history must
now be lost; or at least, will be lost in time,
as the chain of causes encreases, and runs
on to a greater length . . . If belief consisted
only in a certain vivacity, convey’d from an
original impression, it wou’d decay by the
length of the transition, and must at last be
utterly extinguish’d: And vice versa, if belief
on some occasions be not capable of such
an extinction; it must be something different
from that vivacity’ (THN 1.3.13.4 / 145).
True history overcomes this disadvantage,
Hume maintains, because its narrative is
secured ultimately by some impression or real
existence that lies at the end of the ‘chain’ it
traces, anchored in the bed of past lived experi-
ence revived and represented to the reader in
the present. A historian, as Hume writes:
traces the series of actions according to
their natural order, remounts to their
secret springs and principles, and deline-
ates their most remote consequences. He
chooses for his subject a certain portion
of the great chain of events, which com-
pose the history of mankind: Each link
in this chain he endeavours to touch in
his narration . . . And always, he is sens-
ible, that the more unbroken the chain is,
which he presents to his reader, the more
perfect is his production (EHU 3.9).
Thus while each link in the historical chain
tends to decrease in vivacity, the historian
369
9780826443595_Ch20_Final_txt_print.indd 369
HUME ON HISTORY
takes advantage of the fact that connections
between cause and effect resemble each other
in the same way that a copy of a book resem-
bles the original:
One edition passes into another, and that
into a third, and so on, till we come to
that volume we peruse at present. There
is no variation in the steps. After we
know one, we know all of them; and
after we have made one, we can have
no scruple as to the rest. This circum-
stance alone preserves the evidence of
history, and will perpetuate the memory
of the present age to the latest posterity
. . . as most of these [historical] proofs
are perfectly resembling, the mind runs
easily along them, jumps from one part
to another with facility, and forms but
a confus’d and general notion of each
link. By this means a long chain of argu-
ment, has as little effect in diminishing
the original vivacity, as a much shorter
wou’d have, if compos’d of parts, which
were different from each other, and of
which each requir’d a distinct considera-
tion (THN 1.3.13.6 / 146).
This gives to history its own mode of reason-
ing. ‘[W]e may choose any point in history,’
Hume writes in a passage worth quoting in
full,
and consider for what reason we either
believe or reject it. Thus we believe that
Caesar was kill’d in the senate-house
on the ides of March; and that because
this fact is establish’d on the unanimous
testimony of historians, who agree to
assign this precise time and place to that
event. Here are certain characters we
likewise remember to have been us’d as
certain signs of certain ideas; and these
ideas were either in the minds of such
as were immediately present at that
2/9/2012 9:37:29 PM
 HUME ON HISTORY
action, and receiv’d the ideas directly
from its existence; or they were deriv’d
from the testimony of others, and that
again from another testimony, by a vis-
ible gradation, till we arrive at those who
were eye-witnesses and spectators of the
event. ’Tis obvious all this chain of argu-
ment or connexion of causes and effects,
is at first founded on those characters or
letters, which are seen or remember’d,
and that without the authority either
of the memory or sense our whole rea-
soning wou’d be chimerical and with-
out foundation. Every link of the chain
wou’d in that case hang upon another;
but there wou’d not be any thing fix’d to
one end of it, capable of sustaining the
whole; and consequently there wou’d be
no belief or evidence. And this actually
is the case with all hypothetical argu-
ments, or reasoning upon a supposition;
there being in them, neither any present
impression, nor belief of a real existence
(THN 1.3.4.2 / 83).
At bottom, there is some past present impres-
sion or real existence that anchors a chain of
historical narrative in the bed of lived expe-
rience, an original scene of action ‘beyond
which there is no room for doubt or enquiry’
that the historian can reach (THN 1.3.4.1 /
83). The idea of Rome, for example, is gained
neither from sense nor memory, but is derived
from ‘conversation and books of travellers
and historians’, and the idea acquires the req-
uisite force and vivacity to distinguish it from
the mere fictions of the imagination. As such,
Rome is an object of true belief because we
assent to its existence (THN 1.3.10.4 / 108).
These features of historical evidence –
the remoteness of past events and the inter-
preted nature of testimony – stamp history
with a character that stands in contrast to
Hume’s description of it as a straightfor-
wardly empirical discipline. He emphasizes
370
9780826443595_Ch20_Final_txt_print.indd 370
the ideally philosophical character of his-
torical writing, but he acknowledges that it
is a species of ‘invention’ (E 567) and thus
depends upon the very faculty – the imagin-
ation – that threatens to turn true history in
a false direction. Without the temporal and
spatial reach of the imagination, historical
time and knowledge of events that compose
it would be impossible. The original cannot
be displayed by cutting the past at its joints;
the past is not an open book to be read, but
a scene of interpretation to be reconstructed.
There is a chain, to follow Hume’s metaphor,
which leads from the present into remote
regions of the past, but the links are images of
events and the natural connections between
them shadows to be illuminated. Historians
might abjure the creative fictions of litera-
ture, but, like poets, they at once craft scenes
in a way that places them in a historical sys-
tem of things that confers a certain status on
the ideas involved; this does not give them
real existence, but serves as a sufficient foun-
dation for any historical fact. The success of
a historical narrative depends upon raising
ideas in an audience with sufficient force and
vivacity to move and convince them of its
truth, for which reason the historian’s craft
is marked by three criteria to which it must
conform.
First, historians transform the past by
highlighting certain aspects of it and cast-
ing others into shadow; they embellish the
world by ‘adorning’ the facts in the process
of selecting what is relevant for the narrative
(H 3.82). Historians approach the past as
the sedimentation of action and events, from
which the heavier elements are dredged and
placed before the reader as relevant facts.
History memorializes the past, but copying
it faithfully would make a history as insipid
as a comedy that simply repeated the ‘chit-
chat of the tea-table, copied faithfully and
2/9/2012 9:37:29 PM

at full length’ (E 191–2). ‘History,’ Hume
writes, in a preamble to considering the
reign of Henry III,
being a collection of facts which are
multiplying without end, is obliged to
adopt such arts of abridgment, to retain
the more material events, and to drop
all the minute circumstances, which are
only interesting during the time, or to the
persons engaged in the transactions . . .
What mortal could have the patience to
write or read a long detail of frivolous
events as those with which it [the reign of
Henry III] is filled, or attend to a tedious
narrative which would follow, through
a series of fifty-six years, the caprice
and weaknesses of so mean a prince as
Henry? (H 2.3–4).
From a mass of mundane details, the drama
of history is revealed as true by way of events
that are deemed important to the matter at
hand. The historical lens brings events closer,
but in a way that focuses attention on a select
number, pushing others into the background.
To some degree, the narrative depends on the
writer and the kind of history being written
and, as a result, there might be more than
one true history or correct representation of
the past depending on what is deemed rel-
evant. The same applies to time itself, which
the historian represents in a more or less
foreshortened way to suit the purposes of
historical narration: in the History, Hume’s
account of Anglo-Saxon England from the
Roman Empire circa 55 ad to the death of
Harold and the Norman Conquest in 1066
is compressed into some 160 pages of nar-
rative, and he does not begin dating events
systematically until the reign of Egbert in
827–38 (H 1.55ff.). More significantly, nar-
rative time reflects the importance attached
by the author to the events under discussion.
371
9780826443595_Ch20_Final_txt_print.indd 371
HUME ON HISTORY
As Hume examines events of increasing
importance, detail and drama – those leading
up to the dissolution of the Long Parliament
and the Restoration of Charles II, for exam-
ple (H 6.111ff.) – the narrative grows pro-
tracted and the temporal sequence slows,
sometimes abruptly, from years to months, to
days and even, as in the events preceding the
murder of Thomas à Becket, to a matter of
hours (H 2.328ff.).11
Second, the dramatic representation the
historian achieves by transforming and
embellishing the past must produce ideas that
are agreeable to an audience; if historians are
to convey conviction in the reader, they must
affect the imagination in such a way that
gives ideas raised sufficient force and vivac-
ity to constitute belief. Hume emphasizes
that by opening up the past and bringing the
past into the present, history is a source of
‘entertainment to the fancy’ (H 1.4) because
the ‘occupation or agitation of the mind’ it
affords ‘is commonly agreeable and amusing’
(THN 3.3.4.14 / 613; see H 1.4). History
entertains because it opens up the world in a
unique way, so that ‘those, who consider the
periods and revolutions of human kind, as
represented in history, are entertained with
a spectacle full of pleasure and variety, and
see, with surprise, the manners, customs, and
opinions of the same species susceptible of
such prodigious changes in different periods
of time’ (E 97). ‘In reality,’ Hume writes,
what more agreeable entertainment to
the mind, than to be transported into
the remotest ages of the world, and to
observe human society, in its infancy,
making the first faint essays towards the
arts and sciences . . . What amusement,
either of the senses or imagination, can
be compared with it? Shall those trifling
pastimes, which engross so much of our
time, be preferred as more satisfactory,
2/9/2012 9:37:29 PM
 HUME ON HISTORY
and more fit to engage our attention?
How perverse must that taste be, which
is capable of so wrong a choice of pleas-
ures (E 565–6).
Like poets, historians thus rely upon the
receptivity of an audience and the mechanism
of sympathy. ‘The perusal of a history seems
a calm entertainment,’ Hume writes, ‘but
would be no entertainment at all, did not our
hearts beat with correspondent movements
to those which are described by the histo-
rian’ (EPM 5.32 / 223). In this respect, the
historian might be compared to the tragedian
whose literary depiction of a scene provides
occasion through sympathy for the conver-
sion of pain experienced by those at the scene
into pleasure for the reader.
Third, historians do not produce such
agreeable sentiments by accident, but delib-
erately bring about a response in the reader
by employing certain skills and techniques.
Hume is inclined to see poetry as an inher-
ently corrupt enterprise; practitioners are
‘liars by profession,’ he remarks, who ‘always
endeavour to give an air of truth to their fic-
tions’ (THN 1.3.10.5 / 121). Historians do
not suffer from this problem so acutely, since
they do not transform reality into fiction
but represent matter of fact as accurately as
possible, but they too must manipulate the
reader if a historical narrative is to achieve its
effects by striking a balance between repre-
senting events to bring about agreeable ideas
and achieving ‘true and establish’d judge-
ment’ (THN 1.3.9.14 / 115). In Hume’s view,
this balance is possible precisely because his-
tory refers ultimately to matter of fact and
aims at bringing about belief that reaches to
conviction. The historian appeals to evidence
based on inferences from cause to effect and
on this, Hume says, ‘is founded all our belief
in history’ (Abs. 10 / 650). In recollecting
372
9780826443595_Ch20_Final_txt_print.indd 372
events of a historical story there is a differ-
ent feeling from that which accompanies the
effects of literature. ‘Every particular fact is
there the object of belief. Its idea is modify’d
differently from the loose reveries of a castle-
builder’ (THN App. 4 / 625). ‘If one person
sits down to read a book as a romance, and
another as a true history,’ Hume observes,
they plainly receive the same ideas, and in
the same order; nor does the incredulity
of one, and the belief of the other hinder
them from putting the very same sense
upon their author. His ideas produce the
same ideas in both; tho’ his testimony
has not the same influence on them. The
latter has a more lively conception of
all the incidents. He enters deeper into
the concerns of the persons: Represents
to himself their actions, and characters,
and friendships, and enmities: He even
goes so far as to form a notion of their
features, and air, and person. While the
former, who gives no credit to the testi-
mony of the author, has a more faint and
languid conception of all these particu-
lars; and except on account of the style
and ingenuity of the composition, can
receive little entertainment from it (THN
1.3.7.8 / 97–8).
Hume makes the same point when he recalls
sending a copy of Plutarch’s Lives to a female
admirer ‘assuring her . . . that there was not a
word of truth in them from beginning to end.
She perused them very attentively,’ he contin-
ues, ‘’till she came to the lives of Alexander
and Cæsar, whose names she had heard by
accident; and then returned me the book, with
many reproaches for deceiving her’ (E 564).12
In these instances, a poetic romance is dis-
tinguished from a historical narrative by the
degree of belief each inspires, which depends
not directly on the degree of force and vivac-
ity they contain, but on the way they strike the
2/9/2012 9:37:29 PM

imagination, their ‘manner of appearance’. ‘A
poetical description may have a more sensible
effect on the fancy, than an historical narra-
tion,’ Hume maintains.
It may collect more of those circum-
stances, that form a compleat image or
picture. It may seem to set the object
before us in more lively colours. But still
the ideas it presents are different to the
feelings from those which arise from
memory and the judgement. There is
something weak and imperfect amidst all
that seeming vehemence of thought and
sentiment, which attends the fictions of
poetry (THN 1.3.10.10 / 631).
If the objects ‘presented by the wildest
imagination’ were accompanied by the
same sentiment as the facts of history, Hume
observes, they ‘wou’d be on the same footing
with the most establish’d truths founded on
history and experience’ (THN App. 2 / 624).
4. HISTORICAL WISDOM
While, as noted in the previous section,
Hume considers history a source of enter-
tainment, he also emphasizes that knowledge
of the past furnished by philosophical history
is a source of ‘erudition’ or ‘wisdom’ (E 15):
the ‘object of . . . history is to instruct,’ Hume
writes, as it is of eloquence ‘to persuade’ and
‘of poetry to please by means of the passions
and the imagination’ (E 240; see H 1.244
and 5.67). This aspect of history reflects the
natural proclivity of any individual to cast a
glance backwards at the ‘history of the human
race’ and ‘forward to see the influence of his
actions upon posterity, and the judgements
which will be formed of his character a thou-
sand years hence’ (E. 82). History provides a
373
9780826443595_Ch20_Final_txt_print.indd 373
HUME ON HISTORY
unique, transcendent point of view, an eksta-
sis that effectively stops time and lays out
the past as a landscape viewed from a high
vantage point. Thus expanding our ordinar-
ily limited perspective it becomes possible to
transfer lessons learned from past experience
into the present and project them into the
future. Given the ‘shortness of life, and our
limited knowledge,’ as Hume expresses the
idea,
we should be for ever children in under-
standing, were it not for this invention,
which extends our experience to all past
ages, and to the most distant nations;
making them contribute as much to our
improvement in wisdom, as they had
actually lain under our observation. A
man acquainted with history may, in some
respect, be said to have lived from the
beginning of the world, and to have been
making continual additions to his stock of
knowledge in every century (E 567).
With this character of history in mind, there
are two principal advantages that its study
provides. First, Hume emphasizes that the
past is an archive within and on which the
historian-cum-moral scientist works to deci-
pher truths about human nature and ‘con-
firm [sic] the reasonings of true philosophy’
(THN 3.2.10.15 / 562).13 The ‘chief use’ of
history, he observes in the first Enquiry,
is only to discover the constant and uni-
versal principles of human nature, by
showing men in all varieties of circum-
stances and situations, and furnishing
us with materials, from which we may
form our observations, and become
acquainted with the regular springs of
human action and behaviour. These
records of wars, intrigues, factions, and
revolutions, are so many collections of
experiments, by which the politician or
2/9/2012 9:37:29 PM
 HUME ON HISTORY
moral philosopher fixes the principle of
his science. (EHU 8.7 / 83)
In particular, history is a source of instruc-
tion for the art of politics. Since ‘each inci-
dent has a reference to our present manners
and situation, instructive lessons occur every
moment during the course of the narration’
(H 3.82). For this reason, as Hume remarks
of the English people looking back on their
history, an
acquaintance with the ancient periods
of their government is chiefly useful by
instructing them to cherish their present
constitution, from a comparison or con-
trast with the condition of those distant
times . . . by instructing them in the
great mixture of accident, which com-
monly concurs with a small ingredient
of wisdom and foresight, in erecting the
complicated fabric of the most perfect
government. (H 2.525)
Specific events may also be a source of such
wisdom. From the history of English revolu-
tion, Hume is convinced, ‘we may naturally
deduce the same useful lesson, which Charles
himself, in his later years, inferred: that it is
dangerous for princes, even from the appear-
ance of necessity, to assume more authority,
than the laws have allowed them’, as well as
‘another instruction, no less natural, and no
less useful, concerning the madness of the
people, the furies of fanaticism, and the dan-
ger of mercenary armies’ (H 5.545–6).
Second, history is potentially a source
of models of virtue for study and emula-
tion. ‘History, the great mistress of wis-
dom,’ Hume observes, ‘furnishes examples
of all kinds; and every prudential, as well as
moral precept, may be authorized by those
events, which her enlarged mirror is able to
present to us’ (H 5.545–6).14 The study of
374
9780826443595_Ch20_Final_txt_print.indd 374
history – as Hume’s metaphor suggests – is
one mode of reflection, that ‘constant habit
of surveying ourselves,’ as he describes it,
that ‘keeps alive all the sentiments of right
and wrong, and begets, in noble natures, a
certain reverence for themselves as well as
others; which is the surest guardian of every
virtue’ (EPM 9.10 / 276). It is in this spirit
that Hume concludes the account of each
reign narrated in the History with a charac-
ter sketch of the monarch in question: one
might compare Hume’s evaluation of John,
whose character is ‘nothing but a complica-
tion of vices, equally mean and odious; ruin-
ous to himself, and destructive to his people’
(H 1.452), with the ‘model of that perfect
character’, Alfred the Great: ‘So happily
were all his virtues tempered together; so
justly were they blended; and so powerfully
did each prevent the other from exceeding
its proper boundaries!’ (H 1.74–5).
One potential difficulty in this regard,
however – and one that returns us to the
empirical character of history with which
we began – is whether history can strengthen
virtue without taking some necessarily par-
tial view of the past, and becoming poetic,
fabulous or dogmatic, cease to be philosoph-
ical in the sense Hume urges. One answer
to this question is to be found in a principle
underlying Hume’s moral philosophy more
generally, namely, that there is ‘no spectacle
so fair and beautiful as noble and generous
action, nor any which gives us more abhor-
rence than one that is cruel and treacherous’
(THN 3.1.2.2 / 470).15 In the mirror held up
to the past, actions and characters appear
in their true colours and as they give rise to
pleasure or pain in the spectator stand to be
praised as virtuous or condemned as vicious.
History presents action and characters in
their ‘true colours’ and the impressions aris-
ing from its portraits produce agreeable and
2/9/2012 9:37:29 PM

disagreeable sentiments that constitute virtue
and vice, respectively. Actions and characters
display a moral beauty or deformity, which
correspond to our capacity to be affected in
a particular way. Like a work of art, history
presents certain features to which we natu-
rally assent or from which we recoil. ‘There
is no rule in painting more reasonable than
that of ballancing [sic] the figures,’ Hume
observes, ‘and placing them with the greatest
exactness on their proper centers of gravity.
A figure which is not justly ballanc’d [sic],
is disagreeable; and that because it conveys
the ideas of its fall, of harm, and of pain:
Which ideas are painful, when by sympathy
they acquire any degree of force and vivac-
ity’ (THN 2.2.5.19 / 364–5). History that
produces a similar kind of balance ‘fits’ our
capacity for moral sentiments just as figure
and form in an art-work ‘fit’ our capacity to
experience an object as beautiful.
In the final analysis, however, there is some-
thing peculiar to historical as opposed to philo-
sophical inquiry that allows it to produce such
an effect. Hume himself explains this differ-
ence by pointing to the fact that history frees
the inquirer from passion and prejudice, while
at the same time moving the sentiments in a
way that philosophy cannot. ‘When a man of
business enters into life and action,’ he writes,
he is more apt to consider the characters
of men, as they have relation to his inter-
est, than as they stand in themselves; and
has judgement warped on every occa-
sion by the violence of passion. When
a philosopher contemplates characters
and manners in his closet, the general
abstract view of the objects leaves the
mind so cold and unmoved, that the
sentiments of nature have no room to
play, and he scarce feels the difference
between vice and virtue. History keeps
in a just medium betwixt these extremes,
375
9780826443595_Ch20_Final_txt_print.indd 375
HUME ON HISTORY
and places the objects in their true point
of view (E 567–8).
History, Hume suggests, is precisely the sort
of work that can discover both the secret
springs and principles of human nature
and at the same time describe the grace and
beauty of their actions. History views the
body of history, frozen in time, and dissects
it to discover the causal chain of events in a
convincing and true narrative, but at once
breaths life into it by colouring those events
in a certain way. It thus paints a picture of
virtue and vice, and thus recommends one
action and condemns another.
NOTES
1
Donald. W. Livingston, Hume’s Philosophy of
Common Life (Chicago: University of Chicago
Press, 1984), p. 214. See also Victor G. Wexler,
‘David Hume’s Discovery of a New Scene of
Thought’, Eighteenth Century Studies 10(2)
(1976–7), pp. 185–202, esp. pp. 185–6, 189.
2
See Livingston, Hume’s Philosophy of Common
Life, pp. 210ff., and Laird Okie, ‘Ideology and
Partiality in David Hume’s History of England’,
Hume Studies 11(1) (1985), pp. 1–32.
3
Ernest C. Mossner, The Life of David Hume,
2nd edn (Oxford: Clarendon Press, 1980
[1954]), p. 301.
4
Livingston, Hume’s Philosophy of Common
Life, p. 214.
5
Gregory Moses, ‘David Hume as Philosophical
Historian’, Australian Journal of Politics and
History 35.1 (1989), pp. 81, 83–5. See also
David Fate Norton, ‘History and Philosophy
in Hume’s Thought’, in David Fate Norton
and Richard Popkin (eds), David Hume:
Philosophical Historian (New York: Bobbs-
Merrill, 1965), pp. xxxii–l; Stephen Paul Foster,
‘Different Religions and the Difference They
Make: Hume on the Political Effects of Religious
Ideology’, The Modern Schoolman 66(4) (1989),
pp. 253–74, and Donald Siebert, The Moral
Animus of David Hume (Newark, DE: The
University of Delaware Press, 1990), pp. 119–20.
2/9/2012 9:37:29 PM
 HUME ON HISTORY
6
See David Wootton, ‘Hume, “The Historian” ’,
in David Fate Norton (ed.), The Cambridge
Companion to Hume (Cambridge: Cambridge
University Press, 1993), pp. 281–312, who
11
emphasizes the historical context of Hume’s
concept of philosophical history, especially the
influence of The Port-Royal Logic of Antoine
Arnauld and Pierre Nicole that ushered in a
‘downgrading of the reliability of oral tradi-
tion’ and the rise of ‘source criticism’ that bore
a close relation to the desire to understand
contemporary events (p. 286).
7
For a critique of Hume’s distinction between
memory and imagination and its purportedly
failed application to history, see James Noxon,
‘Remembering and Imagining the Past’, in
Donald W. Livingston and James T. King (eds),
Hume: A Re-evaluation (New York: Fordham
University Press, 1976), pp. 270–95.
8
Richard H. Popkin, ‘David Hume: Philosophical
versus Prophetic Historian’, in Kenneth R.
12
Merrill and Robert W. Shahan (eds), David
Hume: Many-sided Genius (Norman, OK:
University of Oklahoma Press, 1976),
pp. 83–95, highlights how in dismissing ‘provi-
dential and prophetic history totally [Hume]
. . . set a pattern for purely secular history
and the secular examination of man . . . [He]
constructed a different kind of historical world
for man to live in . . . a world in which the pro-
phetic and miraculous were so unlikely that all
13
that one could profitably study was the actual
normal course of events’ (pp. 83, 89–90, 92).
9 14
See Ernest C. Mossner, ‘Was Hume a Tory
Historian? Facts and Considerations’, Journal
of the History of Ideas 2 (1941), pp. 225–36,
repr. in Donald Livingston and Marie Martin
(eds), Hume as Philosopher of Society, Politics,
and History. Library of the History of Ideas
IV (Rochester: University of Rochester
Press, 1991), pp. 106–17, esp. pp. 113ff.;
Duncan Forbes, Hume’s Philosophical Politics
(Cambridge: Cambridge University Press,
1975), chap. 8; Christopher J. Berry, ‘Hume on
15
Rationality in History and Social Life’, History
and Theory 21(2) (1982), pp. 234–47, p. 244,
and Okie, ‘Ideology and Partiality in David
Hume’s History of England’, pp. 5–6.
10
See ‘Of the Populousness of Ancient Nations’
(E 377–464), which is an extended correction
376
9780826443595_Ch20_Final_txt_print.indd 376
of reports and conclusions drawn by other
historians who claimed that ‘antiquity was
so much more populous’ than modernity
(p. 380).
See Siebert, The Moral Animus of David
Hume, who remarks that ‘Hume’s way of
writing history is to shape the historical fact
for a desired instructive or emotional impact’,
which, on occasion, leads him to ‘abandon his
normal care in establishing historical truth to
tell a morally invigorating story’ (pp. 55, 58).
Siebert considers Hume’s treatment of Mary,
Queen of Scots, Charles I and the Marquess
of Montrose to show how he accomplishes
this end. See also Donald T. Siebert, ‘The
Sentimental Sublime in Hume’s History of
England’, The Review of English Studies, n.s.
40 (1989), pp. 352–72, and Karen O’Brien,
Narratives of Enlightenment. Cosmopolitan
History from Voltaire to Gibbon (Cambridge:
Cambridge University Press, 1997), chap. 3.
Dugald Stewart, one might note, saw this as a
mark of Hume’s success as a historian. ‘There
are few books more interesting that Hume’s
History of England’, Stewart writes, ‘but,
if we conceived the events to be fictitous, it
would make a very indifferent romance.’ The
Collected Works of Dugald Stewart, vol. 5, ed.
Sir William Hamilton (Edinburgh: Thomas
Constable and Co., 1885–60; repr., Bristol:
Thoemmes Press, 1994), p. 273.
See Norton, ‘History and Philosophy in
Hume’s Thought’, p. xxxiv.
See Siebert, The Moral Animus of David
Hume, who remarks that ‘Hume’s History
projects a moral vision by its ability to reshape
the past, to impose meanings on the past,
creating patterns that imply a corresponding
beauty in human nature – all too seldom
instantiated in human life, it is true, but
nonetheless capable of being discovered,
indeed created in the fiat of narrative, by the
historian’s moral imagination’ (p. 21; see also
p. 44).
In this context, see Siebert, The Moral Animus
of David Hume, pp. 44ff.; S.K. Wertz, ‘Moral
Judgments in History: Hume’s Position’,
Hume Studies 12(2) (1996), pp. 339–86,
esp. p. 342ff., and Moses, ‘David Hume as
Philosophical Historian’, p. 81.
2/9/2012 9:37:30 PM
 21
HUME’S LEGACY AND
THE IDEA OF BRITISH EMPIRICISM
Paul Russell
David Hume (1711–76) is one of the most
important among philosophers because
he developed to its logical conclusion
the empirical philosophy of Locke and
Berkeley, and by making it self-consist-
ent made it incredible. (Bertrand Russell,
A History of Western Philosophy)1
1. INTERPRETATION, LEGACY AND
COLLIGATORY CONCEPTS
It is a familiar feature of historical studies of
various kinds that an understanding of a given
activity or event requires that it be placed in
some wider scheme or framework that will
enable us to say something intelligible about
its nature and significance. Generally speak-
ing, this requires us to look beyond the inten-
tions and thoughts of the agents involved, so
that we can appreciate and recognize their
acts and activities in relation to both what
went before and what came after. It is a legiti-
mate and indeed crucial responsibility of the
historian to be able to identify and describe
the larger movements, processes and develop-
ments that hold agents and events together
and represent and relate them in some coher-
ent fashion. W.H. Walsh has coined the term
377
9780826443595_Ch21_Final_txt_print.indd 377
‘colligation’ to cover the activity by which
historians arrange and gather different events
together ‘under appropriate conceptions’.2
The activity of colligation is also plainly
essential to the history of philosophy and it
presents us with some similar challenges and
difficulties arising from our efforts to make
sense of the larger sweep and developments
that shape the course of the history of phi-
losophy. In this chapter I am concerned with
these issues as they relate, more specifically,
to describing and evaluating the philosophi-
cal legacy of David Hume.3
The general notion of a philosopher’s leg-
acy is, of course, intimately bound-up with the
way in which the writings and contributions
of that philosopher have been read and inter-
preted – both by his own contemporaries and
by those who have followed. Clearly, however,
interpretation and legacy are not the same
thing. A philosopher’s legacy reaches well into
the future and is inevitably shaped, not only
by various factors and features that the author
has no control or influence over, but also by
critical responses and creative contributions
of others which could not be anticipated and
which may not even be intelligible or mean-
ingful to the author or thinker concerned.
Philosophers, like other historical agents, may
2/9/2012 9:37:36 PM
 HUME’S LEGACY
have their own aims and ambitions about
how their work is received and what impact it
may have. Nevertheless, every thinker who is
party to the on-going debate is aware that his
or her ideas and arguments will be considered
and assessed, not just with a view to what has
happened prior to presentation, but also with
a view to what will happen long after he or
she is no longer around to answer for those
ideas and arguments. Their contributions are,
in this sense, open-ended and will inevitably
be transformed and amended in the subse-
quent flow of philosophical discussion and
debate. In the light of these considerations,
the historian of philosophy must always allow
for the possibility that a ‘gap’ may open up
between the philosopher’s original intentions
and objectives and that philosopher’s actual
legacy, understood in terms of the impact and
reception of his ideas over time. Related to this
point, there is no guarantee that an accurate
or full interpretation of the author’s original
intent will properly or adequately explain the
overall impact of his or her contribution on
the subsequent unfolding of the philosoph-
ical conversation. The potential of this gap
is, indeed, crucial if philosophical ideas are to
prove fertile and creative for later generations
of thinkers.
While an important and crucial gap
between interpretation and legacy must be
recognized and accepted as a given of the his-
tory of philosophy, there remains an intimate
and complex relation between them. In the
first place, the actual influence and impact
that a philosophical thinker and his ideas
will have – understood in terms of the kind
of critical reception and creative development
that they are subject to and receive – is itself,
as we have noted, conditioned by how his
contemporaries and those who follow inter-
pret the arguments and works in question.
The interests, prejudices and perspectives of
378
9780826443595_Ch21_Final_txt_print.indd 378
this audience will, naturally, affect the man-
ner of the interpretation provided – what is
emphasized, what is found significant and
worthwhile, and so on – and this will shape
the trajectory and impact of the philosophical
contributions under consideration. However
much a philosopher’s legacy takes on a life
of its own, after publication, interpretation
continues fundamentally to constrain and
direct the perceived worth and value of his
work – either for better or for worse. To this
extent, the ongoing debates concerning inter-
pretation themselves become an integral com-
ponent of a philosopher’s legacy. There are,
plainly, issues and concerns to be noted in this
regard. For example, a good, accurate inter-
pretation may prove to be a philosophical
dead-end or limited in its future creative pos-
sibilities. Similarly, an incomplete and inad-
equate interpretation, one that we may judge
partial and piece-meal, may nevertheless turn
out to be highly fruitful, even though it may
plainly distort or misrepresent the original
intentions of the thinker who is the source of
these (later) developments. It cannot be said,
therefore, that there is any simple or neat cor-
respondence between accurate and reliable
interpretation and productive and fruitful
critical philosophy following in its wake. Part
of the task of the historian of philosophy is to
keep a keen eye open for these points of diver-
gence between interpretation and legacy.
A further important complication to be
noted concerning the interpretation/legacy
relationship is that they have a dynamic and
reciprocal relationship. That is to say, it is a
mistake to treat interpretation as essentially
a static condition of the subject matter (e.g. a
fixed text with an established, rigid meaning)
and the legacy or critical value of the work
as involving a process of building upon this
in a manner that prevents or precludes chal-
lenging or questioning the (established or
2/9/2012 9:37:36 PM

recognized) interpretation. On the contrary,
as a work or text is examined and challenged
over time, later audiences and readers may
return to the question of interpretation and
ask, again, if an adequate and accurate read-
ing has been provided. A negative answer to
this question will invite new readings and
new interpretations, which will, in turn, pro-
vide new opportunities and perspectives for
criticisms and evaluation. A dynamic and
reciprocal process of this kind, between inter-
pretation and critical response, allows for the
possibility and potential of retrieval of ori-
ginal intent (i.e. an alternative reading of the
author’s aims and ambitions) that will reori-
ent and redirect the trajectory of the phil-
osopher’s legacy – changing the impact and
significance of his or her contribution in the
eyes of contemporary and future audiences.
The importance of these observations is
that precision and accuracy of interpretation
are themselves a part of the seamless proc-
ess of criticizing and evaluating a thinker’s
contribution or text for its (living) relevance
and interest. It is a mistake to erect a sharp
dichotomy between ‘scholars’ concerned
with getting the interpretation right and
critical philosophers who are concerned with
the value and worth of the arguments and
ideas as presented. Good scholarship and
good philosophy are more intimately fused
together than this picture of things allows
for. With all this in mind, let us turn to the
question of understanding Hume’s legacy in
the history of philosophy.
2. HUME AND THE IDEA OF BRITISH
EMPIRICISM
The familiar and established way of present-
ing Hume in almost all standard histories of
379
9780826443595_Ch21_Final_txt_print.indd 379
HUME’S LEGACY
philosophy is as the third and last member
of the triumvirate of great British empiricists:
Locke, Berkeley and Hume. This picture of
Hume has indeed shaped his legacy in the his-
tory of philosophy over the two centuries and
more that have followed his death in 1776.
The view that Hume should be understood
primarily in terms of his prominent place in
the tradition of ‘British empiricism’ owes its
secure status in part to the dominant classical
sceptical interpretation of Hume’s fundamen-
tal philosophical intentions and also, in part,
to a wider view of the history of philosophy
which gives Immanuel Kant’s project in The
Critique of Pure Reason (1781) prominence
as the crowning and culminating achieve-
ment of early modern philosophy, in relation
to which all subsequent philosophical work
needs to position itself and be measured.
These two perspectives on Hume’s philoso-
phy and its place in the history of philoso-
phy are themselves closely related, insofar as
Kant’s own reading of Hume’s philosophy is
itself largely informed by the classical scepti-
cal interpretation.4
As it is generally described, the sceptical
reading of Hume’s philosophy dates back
to its early reception provided by Hume’s
own Scottish contemporaries, most nota-
bly Thomas Reid and James Beattie. The
‘Reid-Beattie’ account of Hume’s philosophy
locates his fundamental philosophical contri-
butions as following directly in the tracks that
had been laid down by Locke and Berkeley,
constituted by the ‘theory of ideas’ around
which their own systems of philosophy had
been constructed.5 From this perspective
Hume’s teachings are regarded as essentially
‘destructive’ in character. Hume is taken
to be an extreme, systematic sceptic whose
principal aim is to show that our most basic
common sense beliefs (i.e. concerning causal-
ity and induction, the material world, self,
2/9/2012 9:37:37 PM
 HUME’S LEGACY
free will, etc.) lack any foundation in reason.
Interpreted in this way, Hume’s project is
read as serving to show that when the theory
of ideas is embraced as the starting point of
our philosophical investigations, as suggested
by his empiricist predecessors (i.e. Locke and
Berkeley), then radical and extreme sceptical
consequences will directly follow.
This view of Hume’s philosophy is suc-
cinctly summed up by James Seth in his his-
tory of English Philosophers and Schools of
Philosophy (first published in 1912).6 Seth
suggests that it was Hume’s achievement to
follow relentlessly ‘the logical consequences
of the empirical point of view’ to their inevi-
table sceptical conclusion.7 According to
Seth, Hume’s relationship with Locke and
Berkeley can be explained in these terms:
It would be unjust to both Locke and
Berkeley to say that they stopped short
of these [sceptical] conclusions from
theological or other prejudices. The
truth is that empiricism was only a part
of their philosophy, the other part being
. . . of a rationalistic type; so that we can-
not describe the sceptical philosophy of
Hume as the complete logical develop-
ment of the Lockean and Berkeleyean
philosophy, but only as the logical com-
pletion of the empirical element in the
philosophy of his predecessors. That
which had for them been a part becomes
for Hume the whole: he is an empiricist
pure and simple, and he shows us with
singular insight the ultimate meaning
and consequences of pure empiricism.8
Hume was, Seth claims, ‘fully conscious of
the novel and revolutionary character of his
views, as substituting scepticism, the result
of a thorough-going empiricism, for the mix-
ture of empiricism and rationalism which
he found in Locke and Berkeley . . .’.9 Seth’s
380
9780826443595_Ch21_Final_txt_print.indd 380
sceptical reading of Hume’s empiricist princi-
ples more or less codifies what was, until well
into the twentieth century, the orthodox view
of Hume’s basic intentions and ambitions.10
The classical empiricist-sceptical interpre-
tation has certain prominent features which
have shaped the reception of Hume’s philos-
ophy up until the present time. Hume’s major
text, on this view, is his first and most sub-
stantial work, his Treatise of Human Nature
(1739–40). While his Enquiries and later
writings on religion (e.g. his posthumously
published Dialogues concerning Natural
Religion) are essential for an understanding
of his complete philosophy, and its modifi-
cations and development over time, the real
core of Hume’s philosophical achievement
and insight is still taken to reside with the
central arguments and aims of the Treatise.
From this general perspective, the primary
interest of the Treatise rests with the episte-
mological and metaphysical topics raised in
Book One, ‘Of the Understanding’. Although
Hume’s arguments on these topics were
further refined and modified in the first
Enquiry, his most powerful and original con-
tributions, containing his ambitious sceptical
assault on the foundations of human under-
standing, are presented in their most potent
form in the Treatise. While he, no doubt, had
interesting and worthwhile things to say on
other matters relating to morals, politics and
religion, the sceptical reading gives pride of
place to the question of the scope and limits
of human understanding.11
According to this general account, Hume’s
philosophical significance in the history of
philosophy rests squarely with the funda-
mental sceptical challenge that he has posed
and that all subsequent generations of phi-
losophers must address and respond to. This
challenge takes the form of asking to what
extent human understanding and human
2/9/2012 9:37:37 PM

knowledge can be vindicated against the
(extreme sceptical) conclusions that Hume
has advanced on the basis of his empiri-
cist principles. One line of reply has been
to repudiate and reject Hume’s empiricist
assumptions, as we find in the philosophical
views of Reid and, above all, Kant. Another
alternative is to accept and embrace his
empiricist commitments and to reconstruct
the whole edifice of human knowledge in
the light of Hume’s sceptical constraints and
the limits they impose on our metaphysical
investigations and ambitions. This is the
route that was taken, in various forms, by
later generations of ‘British empiricists’, such
as John Stuart Mill, Bertrand Russell and A.J.
Ayer.12 The primary lesson to be learned from
Hume’s philosophy, from the perspective of
his empiricist followers in this tradition, is
that all a priori metaphysical system-building
and speculations are worthless. All genuine
knowledge, as it concerns our understanding
of the nature of reality, must take the form of
either empirical science or mathematical and
logical investigations. Hume’s significance,
so interpreted, is that he reorients philoso-
phy to play the only role it is truly capable
of, which is to provide a clear logical and
psychological framework and foundation for
the empirical study of nature by means of the
natural sciences. When philosophy attempts
to step outside these boundaries the inevit-
able result is ‘sophistry and illusion’ (EHU
12.34 / 165).13
Interpreted in terms of his empiricist-
sceptical programme, Hume’s legacy has
been to direct the philosophical energies of
the British tradition (along with its American
and continental adherents) in the direction
of an empiricist understanding of our scien-
tific practices and procedures. The central
problems of philosophy, so understood, are
constituted by the relevant set of problems
381
9780826443595_Ch21_Final_txt_print.indd 381
HUME’S LEGACY
generated by this core programme: induction
and causation, knowledge and belief, per-
ception and the external world, the nature
of mind and self, and questions of meaning
and language.14 Understood in these terms,
the task of empiricist philosophy is to iden-
tify and describe the basic building blocks
of human knowledge and explain them in a
manner that is consistent with the empiricist
commitments. The basic elements for this
project are provided by Hume’s philosophy,
beginning with his account of impressions
and ideas (or ‘sense data’ in the idiom of a
later generation of empiricist thinkers). One
way in which later generations of empiri-
cist thinkers – particularly in the twentieth
century – amended and altered the focus of
Hume’s empiricist programme, was in the
emphasis they gave to problems of language
and logic. (This trend is especially apparent
in the work of Russell and Ayer.) Hume’s
philosophy, in contrast with earlier think-
ers in the empiricist tradition (i.e. Locke and
Berkeley, but most notably Hobbes), pays
rather scant attention to problems of lan-
guage and meaning and manifests a stronger
interest in the psychological processes of
human understanding than in the logical ana-
lysis of its forms and structure.15 However,
although this is a clear point of contrast
between Hume and the major representa-
tives of twentieth-century British empiricism,
Hume’s contributions continued to play a
crucial role as a source of many of the key
components and distinguishing features
of their systems. Perhaps the most striking
example of this is the way in which ‘Hume’s
fork’ – his distinction between ‘relations of
ideas’ and ‘matters of fact’ (EHU 4.1 / 25;
cf. THN 3.1.1.9 / 458) – serves as the basis
of the verification principle of logical positiv-
ism. According to the logical positivists, the
meaning (or significance) of any statement
2/9/2012 9:37:37 PM
 HUME’S LEGACY
depends on how (and whether) it can be
verified. This principle requires that a prop-
osition must be either analytic (i.e. trivially
true or a tautology) or empirically or obser-
vationally verifiable.16 In general, the project
of eliminating ‘metaphysics’ by employing an
empiricist standard of meaning is one that is
certainly traceable to certain prominent fea-
tures of Hume’s philosophy and it is plainly
consistent with his own attitude of hostility
and scepticism in regard to the ambitions and
claims of theology and metaphysical ration-
alism in its various forms.17
Although the empiricist-sceptical interpre-
tation of Hume’s philosophy held sway well
into the twentieth century, and as such was the
predominant force in shaping Hume’s philo-
sophical legacy throughout this period, this
way of reading Hume’s philosophical inten-
tions was challenged and brought into ques-
tion by Norman Kemp Smith’s enormously
influential study The Philosophy of David
Hume.18 According to Kemp Smith, what is
central to Hume’s philosophical system ‘is
not Locke’s or Berkeley’s “ideal” theory and
the negative consequences which flow from
it, . . . but the doctrine that the determining
influence in human, as in other forms of ani-
mal life, is feeling, not reason’.19 According to
this reading, the ‘main thesis’ of Hume’s phi-
losophy, as presented in the Treatise and the
Enquiry concerning Human Understanding,
is his claim ‘that belief is more properly an
act of the sensitive, than of the cognitive part
of our natures’.20 On this basis, Kemp Smith
argues that Hume’s philosophy ‘can be more
adequately described as naturalistic than as
sceptical, and that its main governing princi-
ple is the thorough subordination of reason
to the feelings and instincts’.21 Understood
in these terms, Hume’s basic philosophical
strategy is essentially an extension of his
views on morals and aesthetics. With respect
382
9780826443595_Ch21_Final_txt_print.indd 382
to this (fundamental) aspect of Hume’s phil-
osophical system the key influence, Kemp
Smith maintains, was Francis Hutcheson.22
Hume’s insight, it is claimed, was to have
recognized that Hutcheson’s account of the
role of feeling in the sphere of morals could
be applied to ‘several of the chief problems to
which Locke and Berkeley had drawn atten-
tion, but to which they had not been able to
give a satisfactory answer’.23 One particu-
larly significant feature of Kemp Smith’s nat-
uralistic interpretation, so described, is that it
restores a balance between Hume’s concern
with metaphysics and epistemology, on one
side, and morals, on the other – avoiding the
one-sided emphasis on the former, which is
a pronounced feature of the classical empiri-
cist-sceptical interpretation.
Although Kemp Smith’s naturalistic read-
ing places heavy emphasis on the influence of
Hutcheson and the role of feeling as opposed
to reason in human life, Kemp Smith is
himself clear that there is more to Hume’s
‘constructive’ or ‘positive’ programme than
this form of naturalism. More specifically,
in addition to the influence of Hutcheson,
Hume was also inspired by the example of
Newton and his teachings concerning the
proper methods of scientific-inquiry.24 It was,
in particular, Hume’s ambition to develop a
scientific account of the operations of the
human mind modelled after the pattern of
Newtonian physics.25 This way of looking
at Hume’s project of a ‘science of man’ and
his ambition to become ‘the Newton of the
moral sciences’ is now itself a familiar and
standard theme of most general histories of
philosophy.26 This scientific side of Hume’s
naturalism (i.e. his application of the ‘experi-
mental method’ to ‘moral subjects’) has, in
fact, become the more dominant feature of
most naturalistic accounts of Hume’s phil-
osophy.27 There is, however, a fundamental
2/9/2012 9:37:37 PM

difficulty presented by any account that aims
to accommodate Hume’s (putative) ambi-
tion to become ‘the Newton of the moral sci-
ences’. How can this ambition be reconciled,
not only with his (strong) sceptical principles,
but with a form of ‘naturalism’ that maintains
‘that reason, as traditionally understood, has
no role in human life’?28 Claims of this kind
are difficult to square with an understanding
of Hume’s philosophy as a contribution to
the ‘science of man’.
The basic concern here is not simply that
Hume’s aims and objectives are plural, com-
plex and multi-faceted. The real concern
is that his most basic commitments – his
sceptical principles and his scientific ambi-
tions – are in direct conflict with each other,
rendering his entire philosophical system
broken-backed. This was an issue that was
clearly understood by Thomas Reid, who
is generally recognized as the most percep-
tive and penetrating of Hume’s early critics.
Speaking of Hume’s Treatise Reid says:
It seems to be a peculiar strain of humour
in this author, to set out in his introduc-
tion by promising, with a grave face, no
less than a complete system of the sci-
ences, upon a foundation entirely new –
to wit, that of human nature – when the
intention of the whole work is to show,
that there is neither human nature nor
science in the world.29
These observations concerning the appar-
ently fractured and conflicting nature of
Hume’s basic intentions in his major philo-
sophical writings are indicative of the general
problematic that has occupied Hume schol-
ars for more than a century. The difficulty
has been to provide a coherent, consistent
account of Hume’s philosophy in a manner
that fully acknowledges the existence of both
383
9780826443595_Ch21_Final_txt_print.indd 383
HUME’S LEGACY
his sceptical and naturalistic commitments.
While it is tempting to emphasize one side or
the other of this divide, or simply to set aside
or overlook their opposition, a more satisfy-
ing approach must tackle this difficulty more
directly.
These (ongoing) fundamental problems of
interpretation are not without relevance for
our understanding of Hume’s legacy. In par-
ticular, the problem we are presented with is
that these difficulties of interpretation, and the
lack of consensus and agreement concerning
the character of his philosophical aims and
achievements, encourage the thought that his
philosophical legacy, erected as it is on the
foundation of the (problematic) scepticism/
naturalism dichotomy, may itself rely upon
a faulty or incomplete understanding of his
philosophy. That is to say, while there can be
no doubt that his philosophy has established
itself as a main pillar in the larger edifice
of ‘British empiricism’, the difficulties and
doubts of interpretation that we continue to
encounter suggest that this entire edifice, in
so far as it is supposed to help us understand
Hume’s own philosophical contribution, is
itself unstable and liable to collapse under
critical scrutiny. Another way of putting this
general point is to say that the dominant col-
ligatory concept in terms of which Hume’s
legacy has reached us (i.e. in the early twen-
ty-first century) is that of the idea of ‘British
empiricism’. It is under this general rubric
that Hume is portrayed as having played
the pivotal role of developing the arguments
and ideas of his great predecessors (Locke
and Berkeley) and laying the foundations for
subsequent developments in the same trad-
ition, by thinkers who were operating with
the same basic set of empiricist commitments
and constraints found in Hume’s system (i.e.
Mill, Russell, Ayer, et al.). The fact that inter-
pretations of Hume’s philosophy constructed
2/9/2012 9:37:37 PM
 HUME’S LEGACY
around this core colligatory concept render
Hume’s philosophy hopelessly fractured and
incoherent suggests that this concept and
the interpretation(s) associated with it need
to be radically revised, if not abandoned
altogether.
3. HUME, IRRELIGION AND
THE MYTH OF BRITISH EMPIRICISM
Before we consider what alternative inter-
pretations may be made available to us, were
we to abandon the colligatory framework
of ‘British empiricism’ and the scepticism/
naturalism dichotomy associated with it,
we should first consider our methodological
situation. There is no ‘going back’ on Hume’s
established legacy or impact as understood
over the previous century or more, consid-
ered as a pivotal figure in the British empiri-
cist tradition. It is a (historical) given that this
has indeed been the dominant perspective in
which his philosophy has acquired influence
and secured a prominent place in the history
of philosophy. Nevertheless, as we noted,
these facts relating to Hume’s established
and existing legacy do not themselves serve
to guarantee the adequacy or reliability of
the interpretations on which this legacy has
been built. The internal, persisting problems
of interpretation force us to reconsider these
issues and remain open to the possibility
of ‘retrieving’ a better and more adequate
account of how Hume’s fundamental philo-
sophical aims and ambitions can best be rep-
resented and articulated. On the assumption
that a project of retrieval and revision along
these lines is realized, we will inevitably open
up the possibility that Hume’s significance
(i.e. in light of the revised interpretation)
will take his future legacy in a wholly new
384
9780826443595_Ch21_Final_txt_print.indd 384
direction – propelling it into a quite new tra-
jectory. Whether that future trajectory proves
philosophically fruitful or not, judged from
the perspective of later generations of critical
philosophical activity, is an assessment we
are not now in a position to make. With this
observation in place, let us consider an alter-
native perspective on Hume’s philosophy
that largely abandons the colligatory concept
of ‘British empiricism’ and the general histor-
ical framework associated with it.
In a recent study I have argued that the
key to a proper understanding of Hume’s
philosophy as a whole rests with a more
plausible interpretation of his project in
A Treatise of Human Nature.30 Granted
that the Treatise serves as the platform
from which Hume’s overall philosophical
achievement has been erected, it is especially
important that we arrive at some acceptable
solution to the ‘riddle’ of the Treatise – the
(apparent) opposition between his sceptical
and naturalist commitments. How might this
be achieved? The crucial move required to
resolve this interpretive impasse is to chal-
lenge the deeply entrenched assumption that
his Treatise has little or nothing of a direct or
substantial kind to do with problems of reli-
gion. It has been a long established assump-
tion – indeed, a dogma – that his substantial
contributions on the subject of religion are all
to be found in his later writings, most nota-
bly in his posthumously published Dialogues
concerning Natural Religion. According to
this account of things, although Hume origi-
nally intended to include irreligious material
in the Treatise, these passages were removed
so as to avoid causing the orthodox any
‘offence’.31 Contrary to this view, the irre-
ligious interpretation maintains that it is
problems of religion, broadly conceived, that
hold the contents of the Treatise together
as a unified work. More specifically, the
2/9/2012 9:37:37 PM

structure and flow of Hume’s thought in the
Treatise is shaped, on one side, by his attack
on Christian metaphysics and morals and,
on the other, by his efforts to construct in its
place a secular, scientific account of moral-
ity. The constructive or positive side of his
thought – his ‘science of man’ – begins with a
detailed examination of human thought and
motivation based on a naturalistic and neces-
sitarian understanding of human beings. The
model for this project – after which it was
both planned and structured – was the work
of Thomas Hobbes, the most infamous ‘athe-
ist’ thinker of the seventeenth century. The
destructive or critical side of the philoso-
phy of the Treatise is simply the other side
of the same anti-Christian coin. In order to
build the edifice of a secular morality, Hume
had to clear the ground and provide a sys-
tematic sceptical attack on those theologi-
cal doctrines and principles that constitute
an obstacle to this project. The varied and
seemingly disparate sceptical arguments that
are advanced in the Treatise are, in fact, very
largely held together by his overarching aim
to discredit and refute Christian metaphys-
ics and morals. Prominent among the most
obvious and significant of Hume’s sceptical
targets in the Treatise was Samuel Clarke, an
influential Christian rationalist who aimed
to refute demonstrably the ‘atheistic’ philos-
ophy of Hobbes.
Understood in these terms, the irreligious
interpretation provides a fundamentally dif-
ferent account of the nature and character of
Hume’s philosophical project in the Treatise
and the way it is rooted in its relevant his-
torical context. The irreligious interpretation
not only makes it possible to understand the
specific arguments and positions that Hume
takes up on various particular issues and top-
ics (i.e. causation, induction, external world,
etc.), it also enables us to explain how his
385
9780826443595_Ch21_Final_txt_print.indd 385
HUME’S LEGACY
more radical sceptical arguments are sup-
posed to cohere with his ambitions to con-
tribute to the ‘science of man’. Beyond this,
the irreligious interpretation enables us to
account for not only the unity and coherence
of his thought in the Treatise, it also pro-
vides a clear and consistent account of the
unity of Hume’s philosophical thought as a
whole. From this perspective, we no longer
have a serious discontinuity between his
earlier and later works as they concern the
subject of religion. On the contrary, the irre-
ligious interpretation of the Treatise suggests
there is a close and intimate link between this
work and his elaboration of these irreligious
themes and arguments in his later writings.
Granted that irreligious aims and objec-
tives serve as the key to understanding the
core motivation and unity lying behind
Hume’s philosophical work, what is the sig-
nificance of this for our assessment of his
legacy? For reasons that have already been
mentioned, it is evident that Hume’s legacy
has been built upon the foundation of read-
ings that rely on very different assumptions
about both his context and his primary con-
cerns. Although there has been some disagree-
ment about these matters for the established
interpretations (i.e. lying on either side of
the scepticism/naturalism divide), there has
nevertheless been a general acceptance of the
framework of locating Hume’s thought in
the tradition and context of ‘British empiri-
cism’. The irreligious interpretation strongly
suggests that readings of this kind, however
influential (and philosophically fruitful) they
have been, are wholly suspect and mislead-
ing. As we have noted, according to the
classical empiricist-sceptical interpretation,
Hume’s basic achievement throughout his
philosophical work is to have drawn out the
full implications and logical consequences
of empiricist principles as they relate to the
2/9/2012 9:37:38 PM
 HUME’S LEGACY
scope and limits of human understanding.
On this account, Hume is following closely
in the footsteps, and furthering the argu-
ments, of his British empiricist predecessors
Locke and Berkeley – primarily in opposition
to the great triumvirate of continental ration-
alism (Descartes, Spinoza and Leibniz). The
irreligious interpretation makes clear that
this picture of things cannot be a correct or
accurate account of how Hume (or his own
contemporaries) understood the significance
of his work.
From the perspective of the irreligious
interpretation, the primary context in which
Hume’s philosophical works and concerns
must be understood is in terms of the wider
opposition between ‘religious philosophers’
and ‘speculative atheists’ (EHU 12.1 / 149).
When the philosophical lines are drawn this
way, Hume belongs squarely in a tradition
of ‘atheistic’ thought that is represented most
obviously and prominently by Hobbes and
Spinoza, who were closely linked and asso-
ciated in Hume’s early eighteenth-century
context (for example, as we find in the work
of Samuel Clarke, George Berkeley, et al.).32
Although Hobbes is at times included among
the tradition of ‘British empiricists’, he is
frequently dropped from this grouping on
account of his rationalist methodological
commitments (which are deemed at odds
with empiricist methodologies). Spinoza is
plainly a thinker on the ‘wrong side of the
divide’ when judged in terms of the empiricist/
rationalist contrast. Similarly, when we con-
sider Hume’s philosophical ambitions from
the perspective of the irreligious framework,
it is also evident that he stands in direct and
deep opposition to both Locke and Berkeley,
both of whom employed their ‘empiricist’
philosophies with the aim of defending
(Christian) religion and refuting the ‘scepti-
cal and atheistic’ philosophies of the kind
386
9780826443595_Ch21_Final_txt_print.indd 386
that Hobbes and Spinoza had advanced and
argued for. Samuel Clarke, an enormously
important and influential figure in Hume’s
context, is a complete anomaly when consid-
ered in terms of the (continental) rationalist
versus (British) empiricist schema. In con-
trast with this, the irreligious interpretation
places him comfortably and squarely on the
side of ‘religious philosophers’ and also notes
the considerable points of resemblance and
affinities between Clarke’s demonstrative
proof of the Christian religion and key com-
ponents of Locke’s system (e.g. with respect
to the cosmological argument). The upshot
of all this is that the irreligious interpretation
of Hume’s philosophical intentions firmly
rejects the entire colligatory structure built
around the idea of ‘British empiricism’ as a
suitable basis for understanding and describ-
ing his basic aims and objectives throughout
his philosophical writings. Any approach of
this kind is, according to the irreligious inter-
pretation, a historical fabrication that distorts
and misrepresents Hume’s core philosophical
concerns, as both he and his own contempo-
raries would have understood them.33
These general observations regarding the
significance of the irreligious interpretation
as it concerns Hume’s status as a main pillar
of the ‘British empiricist tradition’ are plainly
relevant to our assessment and understand-
ing of his philosophical legacy. Indeed, on the
face of it, the irreligious interpretation may
be taken to discredit thoroughly the basis
of Hume’s legacy, showing that it relies on
interpretative assumptions that cannot be
sustained or supported after critical scrutiny.
We must, however, be careful about the sort
of claims that are made regarding his philo-
sophical legacy in the light of any revisionist
project of the kind that the irreligious inter-
pretation provides. More specifically, as we
have already noted, the relationship between
2/9/2012 9:37:38 PM

interpretation and legacy is not so straight-
forward that we may simply conclude that we
should repudiate Hume’s (established) legacy
on the ground that it has been found to rely
on a faulty or inadequate set of interpretive
assumptions. In the first place, faulty or not,
the legacy of Hume’s thought constructed
around the idea of ‘British empiricism’, and
the scepticism/naturalism divide associated
with it, has its own (independent) genu-
ine history – which itself requires detailed
description and analysis when considering its
various modes and manifestations over the
past two centuries. There is no ‘going back’
or ‘retreating’ from this established legacy,
as it has evidently played a central role in
directing the thoughts and creative energies
of several generations of philosophers who
have taken Hume to be a source of inspir-
ation for their own work.
Clearly, then, we have every reason to
resist the suggestion that we should simply
dismiss or abandon this legacy. Even if we
judge, as we do according to the irreligious
interpretation, that all efforts to straight-
jacket Hume’s philosophical work along
the narrow tracks of the ‘British empiricist’
framework and its associated worries about
the epistemological challenge of scepticism
seriously distort and misrepresent his phil-
osophy, we can hardly fail to acknowledge
that the work done on this basis has itself
proved to be of considerable value and inter-
est. The irony here may be that misinterpret-
ation and distortion of Hume’s core concerns
and historical situation have served us well –
bringing us, among other notable contribu-
tions, nothing less than Kant’s Critique of
Pure Reason. No responsible historian of
philosophy will want to deny either that this
is indeed a genuine aspect of Hume’s philo-
sophical legacy, nor should anyone want to
deny that, considered in terms of the ideas
387
9780826443595_Ch21_Final_txt_print.indd 387
HUME’S LEGACY
and arguments generated, this is a legacy that
is worthy of respect and that ought to com-
mand our (philosophical) appreciation and
attention. Be this as it may, however, we may,
at the same time, hold that the classical scep-
tical interpretation which has spawned this
legacy is itself flawed and deeply unsatisfac-
tory as judged by the standards of historical
and interpretative accuracy.
If it is a mistake to suppose that Hume’s
(established) legacy is worthless because it
rests with the suspect materials supplied by
the idea of British empiricism, it is no less a
mistake to suppose that we have no reason
to question or challenge the received empir-
icist-sceptical interpretation because it has
proved philosophically fruitful and fertile in
the light of later developments. Two consid-
erations are especially important here. First,
accuracy of interpretation is itself a worthy
and valid aim of the historian of philoso-
phy, considered in terms of his or her role
and activity as a historian. It matters, for
the purposes of our own self-interpretation
and the narrative coherence of the history of
philosophy itself, that we are able to offer a
plausible and consistent account of a think-
er’s work in a manner that satisfies our need
to comprehend the ideas concerned and
the context in which his thought emerged
and evolved.34 Second, and perhaps more
importantly, if we abandon the project of
accuracy and adequacy of interpretation we
give up on the possibility that the alternative,
revised interpretation may itself prove philo-
sophically fertile and laden with its own dis-
tinct potential for future generations. If we
remain rooted or locked into an established
interpretation solely on the ground that it
has already (i.e. in the past) proved itself
philosophically fertile, then we fail to allow
for the possibility that its failings, prejudices
and narrowness of focus may prove limiting
2/9/2012 9:37:38 PM
 HUME’S LEGACY
and deadening in the light of (unknown)
future developments or that the alternative
readings might themselves offer significant
critical and imaginative benefits.
With respect to the latter consideration, let
me explain how the project of interpretative
retrieval, as I have described it, may prove
valuable from the point of view of under-
standing the future potential of Hume’s
philosophical legacy. When his Treatise and
his later writings are read in terms of the
traditional schema of the idea of ‘British
empiricism’, his contributions are generally
presented as fragmented and disconnected –
a series of disjointed, sceptical conundrums
relating to the implications of empiricist prin-
ciples. As has been explained, this approach
has, despite these failings, proved enormously
fruitful in terms of generating significant and
substantial contributions in response to this
reading of Hume (especially in the areas of
metaphysics and epistemology as they con-
cern our understanding of scientific prac-
tice). Nevertheless, from the perspective of
the irreligious interpretation, this general
approach to Hume’s philosophy involves
significant losses and costs – and these are
not simply or solely a matter of the partial
and distorted understanding of Hume’s own
aims and ambitions (i.e. understood from a
perspective that places value on our histor-
ical concerns and interests). On the contrary,
what is lost is an appreciation of Hume’s
Treatise considered as a complete system
of irreligion or ‘atheism’ and, as such, as a
possible contribution to our own ongoing
philosophical concerns and interests. When
Hume’s work is read within the structure
suggested by the irreligious interpretation, it
provides us with an ambitious and coherent
world-view. So considered, the value and sig-
nificance of this work, and the key compo-
nents falling within it, rest with its proposals
388
9780826443595_Ch21_Final_txt_print.indd 388
for dealing with some of the more pressing
and disturbing issues that we are currently
facing – namely, our general difficulties in
responding to the claims, dogmas and prac-
tices of religion (especially in their dominant
monotheistic forms). These are issues that
are, for us, matters of immediate and deep
concern. The irreligious interpretation of
the Treatise, along with the wider irreligious
understanding of Hume’s fundamental inten-
tions throughout his philosophy, reaches out
to a very different audience with a different
set of priorities and interests. This audience
is not restricted to circles in academic (pro-
fessional) philosophy with their relatively
arcane worries about the sceptical implica-
tions of empiricist principles as applied to
our common beliefs and practices. The rele-
vant audience for the irreligious reading of
Hume’s thought – a reading that requires
that we abandon, or at least substantially
amend, the perspective of Hume encouraged
by the label of ‘British empiricism’ – extends
well beyond these confines into the general
educated public, embracing all those who
are concerned with the relevance of Hume’s
philosophical system in as far as it speaks to
them as ‘global citizens’.
It is important to acknowledge, and indeed
emphasize, that this alternative way of read-
ing Hume’s intentions and concerns is one
that is, in some degree, already embedded
in his established legacy, despite the way in
which the empiricist-sceptical framework has
obscured these core concerns and features
of his thought. That is to say, according to
the traditional reading, Hume is understood
to have turned his sceptical and naturalistic
attention to matters of religion, in his later
writings. These contributions have, in turn,
become an important part of Hume’s legacy
among subsequent generations of ‘British
empiricists’ who have presented themselves
2/9/2012 9:37:38 PM

as sharing and defending Hume’s anti-reli-
gious attitudes and outlook (e.g. Russell and
Ayer).35 Viewed in this way, the established
or classical interpretations of his philosophy
have aimed to accommodate his irreligious
concerns as an important and significant
aspect or dimension of his wider ‘anti-met-
aphysical’ outlook and stance. The problem
with this approach, from the perspective of
the irreligious interpretation, is that accounts
of this kind not only fail to organize and
arrange properly the structure and develop-
ment of Hume’s views concerning religion
(i.e. by presenting them as peripheral and
derivative in relation to his more fundamen-
tal aims and ambitions), they also distort and
neglect key features and arguments in his
system. The irreligious interpretation takes
its task, therefore, to be one of retrieving or
restoring the integrity of Hume’s thought
with a view to repositioning it in the con-
text of our own current (and future) circum-
stances, in the expectation that this will open
up new avenues and alternative possibilities
for a creative, critical engagement with the
ideas and works concerned.
4. MYTH AND REALITY IN HUME’S
LEGACY
The description we have offered of Hume’s
legacy as it relates to issues of interpret-
ation makes clear that there is an intimate
but complex (dynamic) relationship between
interpretation and philosophical legacy. No
interpretation, whether it be well founded
or not, can by itself ‘fix’ the trajectory of a
philosopher’s legacy and reputation. This
depends, as we have noted, on many fac-
tors and variables that are external to and
independent of the interpretation provided.
389
9780826443595_Ch21_Final_txt_print.indd 389
HUME’S LEGACY
Nevertheless, for obvious reasons, the spe-
cific way in which a thinker and his major
works secure a reputation and influence over
later generations of philosophers and read-
ers, and what they do with the arguments and
ideas conveyed to them, will depend crucially
on the dominant and established modes and
patterns of interpretation. It is in this sense
that the (ongoing) activity of interpretation
is integral to the developing and evolving
legacy of the thinker. The historian of phil-
osophy plays the essential role of integrating
and co-ordinating the activities of interpret-
ation, criticism and creative development
as manifest in the reception that a thinker
receives from later generations. It is a part of
the responsibility of the historian of philoso-
phy to make all those involved in this process
and these activities self-conscious about the
way in which a philosopher’s legacy is itself
embedded and dependent upon given modes
of interpretation and thus subject to all the
limitations and prejudices that this interpret-
ation may bring with it. Related to this, it is
the task of the historian of philosophy to keep
challenging and questioning the adequacy of
the (established) interpretations with a view
not only to maintaining the integrity of our
historical understanding, but also to preserv-
ing and expanding possible avenues of criti-
cism and illumination that might otherwise
be closed off (i.e. in the absence of any activ-
ity of ‘retrieval’ or ‘revision’). There is, as we
noted, no guarantee that these activities will
themselves bear philosophical fruit in the
light of future developments. However, part
of the art of creative, critical scholarship is to
find avenues of interpretation that encourage
confidence that there exists some (philosoph-
ical) potential and value in the alternative
readings when they are considered from the
perspective of our current and future inter-
ests and concerns.
2/9/2012 9:37:38 PM
 HUME’S LEGACY
In general, this understanding of the rela-
tionship between interpretation and legacy
suggests that we should be sceptical or leery
of any account of philosophical method-
ology as it relates to the nature and role of
the history of philosophy that tends to over-
compartmentalize or too sharply separate the
activities of (‘scholarly’) interpretation and
doing (‘creative’) critical philosophy.36 There
is certainly a common picture – encouraged
by certain styles of analytic philosophy that
are not entirely comfortable historical modes
of thought and reflection – that presents the
task of interpretation as essentially back-
ward-looking and temperamentally detached
from living, critical philosophical activity and
attitudes. The relationship between activities
of interpretation and that of engaged philo-
sophical reflection and evaluation is, how-
ever, much more intimate and fluid than
any such model implies. Different styles of
investigation in the area of the history of
philosophy will, of course, give different
weight and prominence to the activities and
relationships involved (e.g. some interpretive
approaches may place heavier emphasis on
contextual matters, others on illuminating
texts and arguments with reference to later
or more contemporary work, and so on).
The crucial point remains, nevertheless, that
good scholarship must always have its eye on
the potential a given interpretation has for
current concerns and issues and, in the same
way, good critical philosophy must always
have an eye on historical self-understanding
of a kind that leaves itself open to looking
for new possibilities and avenues of investi-
gation that may be opened up by (creative,
critical) scholarship.37 On either side of these
two dimensions of philosophical activity the
task may be done in a dull or illuminating
manner. What matters, for our purposes, is to
recognize the seamlessness of the processes
390
9780826443595_Ch21_Final_txt_print.indd 390
and activities involved and recognize that
their relationship is both reciprocal and
dynamic (i.e. interpretation will shape and
inform critical reception, and critical recep-
tion will encourage new and alternative
patterns of interpretation). Interpretation
lacking all reference to critical significance
and interest is (philosophically) wooden and
dead; critical response without any reference
to interpretive alternatives and accuracy will
limit creative possibilities and distort histor-
ical understanding. An integration of these
activities is essential if either is to flourish.
How, then, does all this relate to our ear-
lier observations concerning Hume’s philo-
sophical legacy? I have argued that Hume’s
legacy over the past two centuries has been
structured primarily around the colliga-
tory concept provided by the idea of ‘British
empiricism’. It is this way of categorizing
Hume’s philosophy, in terms of his place as
the third member of the great triumvirate of
British empiricists, which has shaped the way
in which Hume’s key problems and contribu-
tions have been received and criticized. This
perspective on his philosophy has without
doubt been hugely influential, not only in
encouraging later thinkers broadly to self-
identify their own contributions as belong-
ing to this legacy, so interpreted, but also in
generating critical responses from outside
this tradition (e.g. from various schools of
anti-empiricist, or anti-positivist thought
that stand in opposition to the ‘Humean phi-
losophy’, read in this manner). Historians of
philosophy, and Hume scholars in particular,
cannot properly account for the significance
of Hume’s thought over the past two centu-
ries unless they acknowledge the reality and
power of this legacy as constructed around
this reading of Hume. At the same time, for
reasons that have been outlined, it is evident
that the idea of British empiricism, around
2/9/2012 9:37:38 PM

which this legacy has developed and estab-
lished itself, is highly suspect and distorts
and misrepresents Hume’s thought. There
exists, in other words, a wide gap between
(accurate) interpretation and (actual) legacy
as it relates to his philosophical contributions
and achievements. The right response to this
situation, I have argued, is neither to capitu-
late to faulty interpretation nor to deny or
retreat from a recognition of the reality of
Hume’s established reputation understood
in terms of the idea of ‘British empiricism’.
What we need, I maintain, is to find a bal-
anced response which both challenges the
accuracy and adequacy of this dominant col-
ligatory concept and, at the same time, gives
due weight to the reception that Hume’s
philosophy has in fact received, considered
in these terms. The aim of revisionary or
alternative interpretations is, therefore, not
so much to deny Hume’s legacy as already
established but rather to provide new pos-
sibilities or directions for this legacy on the
basis of more accurate and reliable readings
of his arguments and texts.38
The worth and value of any given inter-
pretation, I have argued, cannot be judged
solely in terms of either its accuracy and
adequacy, on one side, or in terms of its pow-
ers to generate fruitful philosophical insights
and illumination for later generations, on the
other side. The reason for this, as we have
noted, is that even careful and accurate inter-
pretation may prove philosophically sterile –
there is no guarantee this will not prove the
case until the interpretation is subject to the
test of time. Similarly, suspect and inadequate
interpretations may still have their merits in
the way they go on to influence and stimu-
late later generations. In assessing and evalu-
ating interpretations, therefore, a certain
tolerance and latitude must be allowed if the
full significance of a thinker’s contribution
391
9780826443595_Ch21_Final_txt_print.indd 391
HUME’S LEGACY
is to be properly appreciated. The activities
of (critical) scholarship, as it concerns itself
with issues of interpretation, are themselves
an integral part of the general dynamics of
Hume’s legacy. It has been my concern in this
chapter to suggest that from the perspective
of interpretation, it is not possible to justify
or defend the use of the dominant colligatory
concept of the idea of ‘British empiricism’
as a basis for understanding and explaining
the essential features of Hume’s philosophy.
At the same time, from the perspective of
providing a plausible account of his legacy
throughout the past two centuries it is indeed
exactly this understanding of his philosophy
that has proved central to the reception and
influence that his philosophical ideas have
generated. While there is no paradox here, it
is, for the reasons I have outlined, import-
ant to avoid succumbing to the temptation
to emphasize or insist upon one side of this
relationship at the expense of the other.39
NOTES
1
B. Russell, History of Western Philosophy
(London: Allen & Unwin, 1947), p. 685.
2
W.H. Walsh, ‘Colligatory Concepts in History’,
repr. in P. Gardiner (ed.), The Philosophy of
History (Oxford: Oxford University Press,
1974), pp. 127–44. See also W.H. Walsh,
An Introduction to Philosophy of History, 2nd
edn (London: Hutchinson, 1958), pp. 25–6.
3
Hume was, of course, more than just a phi-
losopher, as he made major contributions in
several other fields, most notably in the field
of history. His reputation among his own con-
temporaries was, in fact, established primarily
on the basis of his hugely successful History of
England rather than on his philosophy (which
never received the attention and acclaim that
Hume had hoped for). Over time, however, this
situation has reversed itself and it is Hume’s
contributions as a philosopher for which he is
now best known.
2/9/2012 9:37:38 PM
 HUME’S LEGACY
4
Immanuel Kant, Prolegomena to Any Future
Metaphysics that Will Be Able to Present Itself
as a Science, ed. Gunter Zoller, trans. P.G. Lucas
and G. Zoller (Oxford: Oxford University Press,
2004), Preface (esp. pp. 66–7). According to
Kant, Hume’s ‘attack’ on metaphysics ‘started
in the main from a single but important concept
in metaphysics, namely that of the connection
of cause and effect . . .’ (pp. 64–5). For a recent
statement to the effect that Hume’s importance
in the history of philosophy rests primarily with
the influence his views on causation had on
Kant, see Anthony Kenny, ‘Descartes to Kant’, in
A. Kenny (ed.), The Oxford History of Western
7
Philosophy (Oxford: Oxford University Press,
2000), p. 172.
5
An influential account of the ‘Reid-Beattie
interpretation of Hume’s teaching’ is presented
in Norman Kemp Smith, The Philosophy of
8
David Hume, with a new introduction by
9
D. Garrett (Basingstoke: Palgrave, 2005),
10
pp. 3–8. See also Thomas Reid, An Inquiry
into the Human Mind (1764), repr. in Reid’s
Philosophical Works, ed. W. Hamilton, 2 vols
(Hildesheim: Olms, 1967), vol. 1, pp. 95, 104,
204–11. According to Reid, the real source
of the ‘theory of ideas’ is Descartes – which is
why Reid refers to the doctrine with which he
is concerned as ‘the Cartesian system’. Despite
this, later generations have narrowed their
attention on to Locke and Berkeley, considered
as Hume’s immediate predecessors. The most
extreme version of this version of the sceptical
reading, as it relates to Hume’s predecessors,
is presented by T.H. Grose and T.H. Green,
who edited Hume’s philosophical work in
11
the middle of the nineteenth century. Grose
claimed that Hume’s Treatise ‘from beginning
to end is the work of a solitary Scotchman,
who devoted himself to the critical study of
Locke and Berkeley’ (T.H. Grose, ‘History of
the Editions’, in David Hume, Essays: Moral,
Political and Literary, ed. with an introduction
by T.H. Green and T.H. Grose, 2 vols (London:
Longman, 1875), vol. 1, pp. 39–40. A similar
account is presented in Leslie Stephen, History
of English Thought in the Eighteenth Century,
2 vols, 3rd edn (London: Harcourt, Brace and
World, 1962 [1902]), vol. 1, pp. 36–7.
6
James Seth, English Philosophers and Schools
of Philosophy (London: Dent & Sons, 1925).
392
9780826443595_Ch21_Final_txt_print.indd 392
Hume was, of course, self-consciously Scottish
(and he could, at times, be a bit prickly about
the English). Unfortunately, his Scottish
identity has not always served him or his
reputation well: ‘It would be easy to interpret
Hume’s life as the expression of an inferiority
complex, as a writer and a Scotsman. He was
always complaining of a lack of recognition,
of prejudices, of his failure to create a stir; he
was always extremely touchy about everything
Scottish . . .’ (John H. Randall, The Career
of Philosophy: From the Middle Ages to the
Enlightenment, 2 vols (New York: Columbia
University Press, 1962), vol. 1, p. 631.)
Seth, English Philosophers, p. 150. Seth also
notes that his observations in this regard are
in line with those of Wilhelm Windelband, the
influential nineteenth-century German histo-
rian of philosophy.
Ibid., p. 150.
Ibid., p. 154.
See, e.g., Bertrand Russell, History of Western
Philosophy, chap. 7 (see, in particular, his
remarks at p. 685, as quoted at the beginning
of this essay). This general account of Hume’s
philosophy continues to have influence and
is still widely endorsed in standard histories
of the subject. See, e.g., Roger Scruton, From
Descartes to Wittgenstein: A Short History
of Modern Philosophy (London: Routledge
& Kegan Paul, 1981), p. 120: ‘Hume’s fame
rests on his scepticism. He saw, or thought he
saw, that the outlook of empiricism circum-
scribed the possibilities of human knowledge
to the extent that his predecessors had not
recognized . . .’.
We may, of course, distinguish several distinct
aspects of Hume’s empiricism. Two features
of his empiricism are particularly significant.
The first is that Hume insists all thought and
belief has its origins or source in experience
(i.e. impressions of sense and reflection). The
second is that he aims to make philosophy
scientific by way of introducing and applying
the ‘experimental method’, rather than relying
on a priori investigations. Both these aspects
of his empiricism are intimately linked to his
‘naturalism’, which is discussed further below.
For an account of this distinction between the
two main components of Hume’s empiri-
cism see Anthony Quinton, Hume (London:
2/9/2012 9:37:39 PM

Phoenix, 1998), pp. 10–11). For a more
fine-grained approach to the various compo-
nents of Hume’s empiricism see Don Garrett,
Cognition and Commitment in Hume’s
Philosophy (Oxford: Oxford University Press,
1997), pp. 29–38.
12
For a general account of some of these later
developments see Stephen Priest, The British
Empiricists, 2nd edn (London and New York:
Routledge, 2007), chaps 6–8. It should also
be noted, however, that later ‘British empiri-
cists’ such as Russell and Ayer diverge in some
significant ways from their predecessors in
respect of their ‘empiricist’ commitments. See,
for example, John Passmore’s observations
about the rationalist roots of Russell’s (early)
philosophy: A Hundred Years of Philosophy,
2nd edn (Harmondsworth: Penguin, 1966),
pp. 214–16; and the related observations
concerning Russell’s ontology in Robert G.
Meyers, Understanding Empiricism (Chesham:
Acumen, 2006), pp. 6–7.
13
In relation to this matter, the issue that sepa-
rates Kant from the empirical school is not the
rejection of ‘speculative metaphysics’, since
Kant shared this, but rather to what extent
fundamental principles of (pure) reason serve
to (a priori) structure experience, as opposed
to being derived from it (e.g. as per Hume’s
account of causation). There is an important
sense in which Kant must be judged – and
indeed viewed himself – as a follower of
Hume, and not just a critic.
14
Representative texts manifesting the priority
and salience of this set of issues can be found
in, for example, John S. Mill, A System of
Logic, 8th edn (London: Longmans, Green
& Co., 1898; first pub. 1843); Bertrand
Russell, The Problems of Philosophy (Oxford:
Oxford University Press, 1912); A.J. Ayer, The
Problems of Knowledge (Harmondsworth:
Penguin, 1956).
15
Hobbes’s place in the British empiricist trad-
ition is itself ‘problematic’ and he is frequently
dropped from the list of its representatives
(primarily on the ground that his understand-
ing of science, based on the geometric method,
is insufficiently empirical in character). This
is indicative of more general worries about
the supposed identity and boundaries of the
empiricist tradition.
393
9780826443595_Ch21_Final_txt_print.indd 393
HUME’S LEGACY
16
On this see, for example, Antony Flew, Hume’s
Philosophy of Belief (London: Routledge,
1961), chap. 3 (esp. p. 53, where Flew refers
to ‘Hume’s fork’ to describe this basic divi-
sion between two kinds of proposition). For
a more general account of the relevance of
Hume’s philosophy to the ‘positivist’ pro-
gramme see Leszek Kolakowski, Positivist
Philosophy: From Hume to the Vienna Circle
(Harmondsworth: Penguin, 1972), esp. chap. 2.
17
A.J. Ayer makes the following remarks about
the relationship between Hume and the logi-
cal positivism of the Vienna Circle: ‘Although
they [the Vienna Circle] didn’t themselves
know, or care, much about the history of
philosophy, what they said was very like what
was said by the Scottish philosopher, David
Hume, in the eighteenth century. So to that
extent they weren’t all that novel, or that rev-
olutionary. What was revolutionary was, in a
sense, their fervour, their seeing this as putting
philosophy on a new road. They thought:
“At last we’ve discovered what philosophy is
going to be! It’s going to be the handmaiden
of science.”’ (quoted in Bryan Magee, Men
of Ideas: Some Creators of Contemporary
Philosophy (Oxford: Oxford University Press,
1978), p. 97).
18
Although Kemp Smith’s The Philosophy of
David Hume did not appear until 1941, he
published two influential articles on ‘The
Naturalism of Hume’ in Mind in 1905.
These articles are reprinted in Kemp Smith,
The Credibility of Divine Existence, ed. A.J.
Porteous, R.D. MacLennan and G.E. Davie
(London: MacMillan, 1967).
19
Kemp Smith, The Philosophy of David Hume,
p. 11.
20
THN 1.4.1.8 / 183; Kemp Smith, The
Philosophy of David Hume, p. 546.
21
Kemp Smith, The Philosophy of David Hume,
p. 84.
22
Ibid., pp. 12–13.
23
Ibid., p. 13.
24
Ibid., p. 53.
25
Ibid., p. 71.
26
John Passmore, Hume’s Intentions, 3rd edn
(London: Duckworth, 1980), pp. 3–4; and
E.C. Mossner, Life of David Hume, 2nd edn
(Oxford: Oxford University Press, 1980),
pp. 73–5. See also, for example, Frederick
2/9/2012 9:37:39 PM
 HUME’S LEGACY
Copleston, Modern Philosophy – The British
Philosophers (Garden City, NY: Doubleday,
1964), p. 66 [vol. 5, pt II of Copleston’s
A History of Philosophy, 9 vols]; D.W.
Hamlyn, A History of Western Philosophy
(Harmondsworth: Penguin, 1988), pp. 188–9.
27
Two especially important and influential con-
tributions along these lines are Barry Stroud,
Hume (London: Routledge & Kegan Paul,
1977), chap. 3; and also Garrett, Cognition
and Commitment, e.g. pp. 10, 94–5, 161,
240–1. For related themes concerning Hume’s
anticipations of ‘cognitive science’ and ‘evolu-
tionary psychology’ see Jerry A. Fodor, Hume
Variations (Oxford: Clarendon, 2003); Simon
Blackburn, How to Read Hume (London:
Granta, 2008); and also Alan Bailey and Dan
O’Brien, Hume’s Enquiry Concerning Human
Understanding: Reader’s Guide (London:
Continuum, 2006), chap. 5.
28
Stroud, Hume, p. 14.
29
Reid, Works, I, 102a. Not all commentators
are convinced that there is a conflict or incom-
patibility here. Roger Scruton, for example,
suggests that there are ‘two ways of reading
Hume’. Either ‘as a sceptic who defends, from
empiricist premises, the view that the standard
claims to knowledge are untenable’, or ‘as
the proponent of a “natural philosophy” of
man, who begins from empirical observations
about the human mind and concludes that
the mind has been wrongly construed by the
metaphysicians’. Scruton goes on to claim
that these ‘two readings are not incompatible’
but that the second ‘has been emphasized in
recent commentaries, partly because it parallels
recent developments in philosophy’ (Scruton,
A Short History of Modern Philosophy: From
Descartes to Wittgenstein, 2nd edn (London:
Routledge, 1995), pp. 115–16). In the earlier
edition of this work Scruton focused, more
narrowly, on Hume’s reputation as a sceptic
(see note 10 above).
30
Paul Russell, The Riddle of Hume’s Treatise:
Skepticism, Naturalism and Irreligion (New
York and Oxford: Oxford University Press,
2008).
31
A standard account along these lines is
presented in John Laird, Hume’s Philosophy
of Human Nature (London: Methuen, 1932),
pp. 282–3. In support of the view that the
394
9780826443595_Ch21_Final_txt_print.indd 394
Treatise contains little of a significant nature
concerned with problems of religion, Laird
cites a famous 1737 letter from Hume to
Henry Home (Lord Kames) concerning the
‘castration’ of his Treatise (LDH 1.23–5, 6).
See also Mossner, Life of Hume, pp. 111–13.
32
This tradition can itself be traced back to
irreligious thinkers among the ancients, such
as Lucretius. It stretches ahead to thinkers
such as D’Holbach and Nietzsche. Plainly the
irreligious interpretation not only reconfigures
Hume’s philosophy (i.e. in terms of its overall
structure and central themes and motifs), it
also repositions his place in the history of
philosophy and, in so doing, rearranges our
dominant perspectives on early modern phi-
losophy more generally.
33
Similar observation may be made regarding
Kemp Smith’s claim about the influence of
Hutcheson and Newton as they relate to
Hume’s basic philosophical ambitions. While
both these thinkers may well, like Locke and
Berkeley, have played some constructive role
in shaping aspects of Hume’s philosophy,
they stand directly on the other side of the
main divide between ‘religious philosophers’
and ‘speculative atheists’. Hutcheson was
deeply opposed to the (anti-Christian)
philosophical system of Hobbes and was
not alone in recognizing significant elements
of this in Hume’s system. Among Hume’s
sternest and most severe early critics were
various followers and admirers of the
philosophy of Samuel Clarke – who was
himself a close collaborator and champion
of Newton’s philosophy and its associated
theology. Clearly, then, the irreligious
interpretation serves to discredit any account
of Hume’s philosophy that aims to explain
his core aims and ambitions in terms of the
(constructive) influence of Hutcheson and
Newton. Accounts along these lines fail to
identify and emphasize properly Hume’s
specific irreligious aims and objectives.
34
As already noted, Hume was a historian, as
well as a philosopher, and so would appreciate
these constraints. Contemporary philosophy is
heavily dominated by the values and methods
of science – if not scientism – in such a man-
ner that the intrinsic value and importance of
historical imagination and understanding is
2/9/2012 9:37:39 PM

not obvious to many of its (most influential)
practitioners.
35
Considered from this perspective Hume
becomes a key figure in ‘the Enlightenment
tradition’, where this is understood as a col-
37
ligatory concept that has a life of its own in
relation to Hume’s legacy. To the extent that
it is understood and described in terms of the
(concept of the) ‘Enlightenment’ there is no
obvious mismatch or lack of fit between his
legacy and the irreligious interpretation, given
that the irreligious interpretation places his
irreligious aims and ambitions squarely in the
context of the ‘Radical Enlightenment’. For
more details on this aspect of the irreligious
interpretation see Russell, Riddle of Hume’s
38
Treatise, esp. chaps 3 and 18.
36
For an illuminating discussion of some of
these methodological difficulties see Bernard
Williams, ‘Descartes and the Historiography
39
of Philosophy’, in his The Sense of the Past:
Essays in the History of Philosophy, ed.
M. Burnyeat (Princeton: Princeton University
Press, 2006), pp. 257–66. Speaking of the
history of philosophy, Williams concludes his
essay as follows: ‘What we must do is to use
the philosophical materials that we now have
to hand, together with historical understand-
ing, in order to find in, or make from, the
395
9780826443595_Ch21_Final_txt_print.indd 395
HUME’S LEGACY
philosophy of the past a philosophical struc-
ture that will be strange enough to help us to
question our present situation and the received
picture of that tradition, including those mate-
rials themselves.’
There is, of course, a constant difficulty in find-
ing a balance between being sensitive to the
historical differences between our own situa-
tion and concerns and those of the past and,
on the other side, finding some relevance and
interest in the ideas and arguments made avail-
able through the study of historical texts and
thinkers. The important point is that sensitivity
to historical difference should not prevent us
from finding contemporary (living) relevance
in the works in question.
Of course, the revised interpretations will also
allow us to look critically at the legacy itself,
with a view to evaluating its own claims to
accuracy and adequacy of interpretation.
A talk based on this chapter was given at a
Hume Workshop, hosted at Simon Fraser
University, August 2010. I am grateful to
members of the audience and to my fellow
presenters (Dario Perinetti, Lisa Shapiro and
Jackie Taylor) for their helpful comments
and suggestions. I would also like to thank
Scott Edgar for additional comments and
suggestions.
2/9/2012 9:37:39 PM
 BIBLIOGRAPHY
1. WORKS BY HUME
EARLIEST EDITIONS, ARRANGED IN ORDER OF
FIRST PUBLICATION
A Treatise of Human Nature: Being an
Attempt to Introduce the Experimental
Method of Reasoning into Moral
Subjects, 3 vols (first 2 vols – London:
John Noon, 1739; 3rd vol. – London:
Thomas Longman, 1740).
An Abstract of a Book lately Published;
Entituled, A Treatise of Human Nature
&c. Wherin the Chief Argument of that
Book is Farther Illustrated and Explained
(London: C. Borbett [misprint for C.
Corbett], 1740). [Although this work was
published anonymously, the present-day
scholarly consensus is that it was written
by Hume himself, although it has in the
past been attributed to Adam Smith.]
Essays, Moral and Political (Edinburgh: A.
Kincaid, 1741).
Essays, Moral and Political. Volume II
(Edinburgh: A. Kincaid, 1742).
A Letter from a Gentleman to His Friend
in Edinburgh: Containing Some
Observations on A Specimen of the
Principles concerning Religion and
Morality, said to be maintain’d in a Book
396
9780826443595_Ref_Final_txt_print.indd 396
latedly publish’d, intituled, A Treatise of
Human Nature, &c. (Edinburgh, 1745).
[An anonymous work that seems to have
been based on material supplied by Hume
with some editorial input from Henry
Home, later Lord Kames.]
A True Account of the Behaviour and
Conduct of Archibald Stewart, Esq; Late
Lord Provost of Edinburgh. In a Letter to
a Friend (London: M. Cooper, 1748).
Philosophical Essays concerning Human
Understanding (London: A. Millar,
1748). [This work was given its present
title of An Enquiry concerning Human
Understanding in 1758.]
Three Essays, Moral and Political. Never
before Published. Which Compleats the
Former Edition, in Two Volumes, Octavo
(London: A. Millar and A. Kincaid, 1748).
The Petition of the Grave and Venerable
Order of Bellmen, or Sextons, of the
Church of Scotland to the Hon. House of
Commons (Edinburgh, 1751).
An Enquiry concerning the Principles of
Morals (London: A. Millar, 1751).
Political Discourses (Edinburgh: A. Kincaid
and A. Donaldson, 1752).
Scotticisms (n.p. [1752]).
Essays and Treatises on Several Subjects, 4
vols (London: A. Millar, 1753–4). [The
2/9/2012 9:38:20 PM

first collected edn of previously published
works by Hume. Vol. I. Containing
Essays Moral and Politic. The Fourth
Edition corrected, with Additions. Vol.
II. Containing Philosophical Essays
concerning Human Understanding.
The Second Edition, with Additions
and Corrections. Vol. III. Containing
An Enquiry concerning the Principles
of Morals. Vol. IV. Containing Political
Discourses – either the 2nd or 3rd edn.]
The History of Great Britain, Vol. I.
Containing the Reigns of James I and
Charles I (Edinburgh, 1754).
The History of Great Britain, Vol. II.
Containing the Commonwealth, and
the Reigns of Charles II and James II
(London: A. Millar, 1757).
The History of England, Under the House of
Tudor, 2 vols (London: A. Millar, 1759).
The History of England, From the Invasion
of Julius Caesar to the Accession of
Henry VII, 2 vols (London: A. Millar,
1762).
The History of England, From the Invasion
of Julius Caesar to the Revolution in
1688, 6 vols (London: A. Millar, 1762).
[An expanded and reordered edition
incorporating the contents of The History
of Great Britain, The History of England,
Under the House of Tudor, and The History
of England, From the Invasion of Julius
Caesar to the Accession of Henry VII.]
Four Dissertations. I. The Natural History
of Religion. II. Of the Passions. III. Of
Tragedy. IV. Of the Standard of Taste
(London: A. Millar, 1757).
Exposé Succinct de la Contestation qui s’est
élevée entre M. Hume et M. Rousseau,
avec les pieces justificatives ([Paris],
1766).
A Concise and Genuine Account of the
Dispute between Mr. Hume and Mr.
397
9780826443595_Ref_Final_txt_print.indd 397
BIBLIOGRAPHY
Rousseau: With the Letters that Passed
between Them during Their Last
Controversy (London: T. Beckett and P.A.
De Hondt, 1766). [This is a translation of
the Exposé Succinct but introduces some
changes and revisions.]
The Life of David Hume, Esq. Written
by Himself (London: W. Strahan and
T. Cadell, 1777).
Two Essays (London: 1777) [Contains
the essays ‘On Suicide’ and ‘On the
Immortality of the Soul’].
Dialogues concerning Natural Religion
([London], 1779).
Letters of David Hume and Extracts from
Letters Referring to Him, ed. T. Murray
(Edinburgh: A. & C. Black, 1841).
Life and Correspondence of David Hume,
ed. J.H. Burton, 2 vols (Edinburgh:
W. Tait, 1846).
Letters to William Strahan, ed. G.B. Hill
(Oxford: Clarendon Press, 1888).
The Letters of David Hume, ed. J.Y.T. Greig,
2 vols (Oxford: Clarendon Press, 1932).
‘Hume’s Early Memoranda, 1729–1740:
The Complete Text’, ed. E.C. Mossner,
Journal of the History of Ideas (1948),
pp. 492–518.
New Letters of David Hume, ed. R.
Klibansky and E.C. Mossner (Oxford:
Clarendon Press, 1954).
2. RECOMMENDED MODERN
EDITIONS
The most authoritative modern editions of
Hume’s works are those contained in The
Clarendon Edition of the Works of David
Hume and the parallel volumes in the
Oxford Philosophical Texts editions. The
Oxford Philosophical Texts editions are
2/9/2012 9:38:20 PM
 BIBLIOGRAPHY
aimed at supporting student readers, and
they accordingly supplement the central text
with introductions, guides to supplementary
reading, and annotations that are likely to
be particularly useful to anyone reading the
work in the course of undergraduate or ini-
tial postgraduate study. The Clarendon edi-
tions, in contrast, concentrate on providing a
surrounding scholarly apparatus that is more
appropriate to the needs and interests of peo-
ple engaged in advanced research.
The works by Hume currently available
in Clarendon or Oxford Philosophical Texts
editions are as follows:
A Treatise of Human Nature, The
Clarendon Edition of the Works of David
Hume, ed. D.F. Norton and M.J. Norton,
2 vols (Oxford: Clarendon Press, 2007).
This includes the text of An Abstract of
a Book lately Published; Entituled, A
Treatise of Human Nature and A Letter
from a Gentleman to His Friend in
Edinburgh.
A Treatise of Human Nature, Oxford
Philosophical Texts, ed. D.F. Norton
and M.J. Norton (Oxford: Oxford
University Press, 2000). This includes
the text of An Abstract of a Book lately
Published; Entituled, A Treatise of
Human Nature.
An Enquiry concerning Human
Understanding, The Clarendon Edition
of the Works of David Hume, ed. T.L.
Beauchamp (Oxford: Clarendon Press,
2000).
An Enquiry concerning Human
Understanding, Oxford Philosophical
Texts, ed. T.L. Beauchamp (Oxford:
Oxford University Press, 1999).
An Enquiry concerning the Principles
of Morals, The Clarendon Edition of
the Works of David Hume, ed. T.L.
398
9780826443595_Ref_Final_txt_print.indd 398
Beauchamp (Oxford: Clarendon Press,
1998).
An Enquiry concerning the Principles of
Morals, Oxford Philosophical Texts,
ed. T.L. Beauchamp (Oxford: Oxford
University Press, 1998).
A Dissertation on the Passions; The Natural
History of Religion, The Clarendon
Edition of the Works of David Hume,
ed. T.L. Beauchamp (Oxford: Clarendon
Press, 2007).
Before the publication of the above editions,
the standard editions of the Treatise and the
Enquiries were those prepared by L.A. Selby-
Bigge and revised by P.H. Nidditch. Although
these editions must now be regarded as non-
essential, many books and articles of endur-
ing value to anyone studying Hume quote
from and refer to their version of Hume’s
text:
A Treatise of Human Nature, ed. L.A. Selby-
Bigge, 2nd edn with text rev. and notes by
P.H. Nidditch (Oxford: Clarendon Press,
1978).
Enquiries concerning Human Understanding
and concerning the Principles of Morals,
ed. L.A. Selby-Bigge, 3rd edn with text
rev. and notes by P.H. Nidditch (Oxford:
Clarendon Press, 1975).
The above edition of the Treatise also con-
tains the full text of the Abstract but does
not contain the text of A Letter from a
Gentleman. A facsimile edition of this latter
work, combined with a substantial editorial
essay, is provided by:
A Letter from a Gentleman to his Friend
in Edinburgh, ed. E.C. Mossner and J.V.
Price (Edinburgh: Edinburgh University
Press, 1967).
2/9/2012 9:38:20 PM

The Clarendon Edition will ultimately include
Hume’s Essays and the Dialogues concerning
Natural Religion, but there are apparently no
plans to extend it to the History of England.
For the foreseeable future, therefore, the best
available edition of the History will continue
to be:
The History of England, from the Invasion
of Julius Caesar to the Revolution in
1688, ed. W.B. Todd, 6 vols (Indianapolis:
Liberty Classics, 1983).
In the case of the Essays and the Dialogues
the current standard editions for those under-
taking research on Hume are:
Essays, Moral, Political, and Literary, ed.
E.F. Miller (Indianapolis: Liberty Classics,
1987).
Hume’s Dialogues concerning Natural
Religion, ed. N. Kemp Smith, 2nd edn
with supp. (London and Edinburgh:
Thomas Nelson: 1947).
Although the Kemp Smith edition offers a
superb introductory essay that constitutes a
major piece of Hume scholarship in its own
right, its availability is now increasingly lim-
ited except for those copies held by academic
libraries. Consequently the following alterna-
tive editions of the Dialogues are often used
in university classes and seminars:
Dialogues and Natural History of Religion,
Oxford World’s Classics, ed. J.C.A.
Gaskin (Oxford: Oxford University Press,
1993).
Dialogues concerning Natural Religion,
ed. R.H. Popkin, 2nd edn (Indianapolis:
Hackett Publishing Company, 1998).
Dialogues concerning Natural Religion,
and Other Writings, ed. D. Coleman
399
9780826443595_Ref_Final_txt_print.indd 399
BIBLIOGRAPHY
(Cambridge: Cambridge University Press,
2007).
The Gaskin edition has the signal advantage
of combining the Dialogues with the full text
of the Natural History. It therefore serves the
function of providing an inexpensive alterna-
tive to the Clarendon edition of the Natural
History, whereas the Coleman edition
reprints much of the Natural History but not
the entire work.
There is also an important place for edi-
tions of Hume’s writings that serve the needs
of readers who are primarily interested in
Hume as an important literary and cultural
figure alongside such eighteenth-century nov-
elists and satirists as Swift, Richardson and
Sterne. Readers falling into this category are
particularly well served in respect of Hume’s
writings on religion by the Gaskin edition
of the Dialogues and the Natural History.
And the following edition, which combines
a compact format with a judicious introduc-
tion by its editor and an internal referencing
system that makes it usable in conjunction
with references to both the Beauchamp and
Selby-Bigge editions, provides an excellent
version of Hume’s first Enquiry for general
reading purposes:
An Enquiry concerning Human
Understanding, Oxford World’s Classics,
ed. P. Millican (Oxford: Oxford
University Press, 2007).
3. HUME BIBLIOGRAPHIES
Capaldi, N., King, J.T., and Livingston,
D.W., ‘The Hume Literature of the
1970’s’, Philosophical Topics 12 (1981),
pp. 167–92.
2/9/2012 9:38:21 PM
 BIBLIOGRAPHY
Capaldi, N., King, J.T., and Livingston,
D.W., ‘The Hume Literature of the
1980’s’, American Philosophical
Quarterly 28 (1991), pp. 255–72.
Chūō Daigaku Library, David Hume and
the Eighteenth Century British Thought,
2 vols (Tokyo: Chuo University Library,
1986 and 1988).
Cunningham, I.C., ‘The Arrangement of
the Royal Society of Edinburgh’s David
Hume Collection’, Bibliotheck: A Scottish
Journal of Bibliography and Allied Topics
15 (1988), pp. 8–22.
Greig, J.Y.T., and Benyon, H., Calendar
of Hume Manuscripts in the Possession
of the Royal Society of Edinburgh
(Edinburgh: Royal Society of Edinburgh,
1932; repr. as facs. edn, Bristol:
Thoemmes Press, 1990).
Hall, R., Fifty Years of Hume Scholarship:
A Bibliographical Guide (Edinburgh:
Edinburgh University Press, 1978).
Jessop, T.E., A Bibliography of David Hume
and of Scottish Philosophy from Francis
Hutcheson to Lord Balfour (London: A.
Brown & Sons, 1938).
Millican, P., ‘Critical Survey of the Literature
on Hume and the First Enquiry’, in P.
Millican (ed.), Reading Hume on Human
Understanding (Oxford: Clarendon Press,
2002), pp. 413–74.
Todd, W.B., ‘David Hume: A Preliminary
Bibliography’, in W.B. Todd (ed.), Hume
and the Enlightenment: Essays presented
to Ernest Campbell Mossner (Edinburgh:
Edinburgh University Press, 1974),
pp. 189–205.
Tweyman, S., Secondary Sources on the
Philosophy of David Hume: A David
Hume Bibliography, 1741–2005, in Two
Volumes (Ann Arbor, MI: Caravan Books,
2006).
400
9780826443595_Ref_Final_txt_print.indd 400
Yalden-Thomson, D.C., ‘Recent Work on
Hume (A Survey of Hume Literature
1969–79)’, American Philosophical
Quarterly 20 (1983), pp. 1–22.
The journal Hume Studies (published by
the Hume Society – ISSN 0319-7336) regu-
larly includes a comprehensive listing of new
work published on Hume in the preceding
one or sometimes two years. These bibliog-
raphies normally appear in the November
issue of the journal. This practice lapsed dur-
ing the years 1989–93, but Hume Studies
20(2) (November 1994) includes a bibliog-
raphy compiled by W.E. Morris that covers
this period.
4. BIOGRAPHIES AND HUME’S LIFE
Burton, J.H., Life and Correspondence of
David Hume, 2 vols (Edinburgh: W. Tait,
1846).
Graham, R., The Great Infidel: A Life of
David Hume (East Linton: Tuckwell
Press, 2004).
Greig, J.Y.T., David Hume (London:
Jonathan Cape, 1931).
Miller, S., Three Deaths and Enlightenment
Thought: Hume, Johnson, Marat
(Lewisburg, PA: Bucknell University
Press, 2001).
Mossner, E.C., ‘Philosophy and Biography:
The Case of David Hume’, The
Philosophical Review 59 (1950),
pp. 184–201.
–––––, The Forgotten Hume: Le Bon David
(New York: Columbia University Press,
1943).
–––––, The Life of David Hume, 2nd edn
(Oxford: Clarendon Press, 1980).
2/9/2012 9:38:21 PM

Ross, I.S., ‘The Intellectual Friendship
of David Hume and Adam Smith’,
in E. Mazza and E. Ronchetti (eds),
New Essays on David Hume (Milan:
FrancoAngeli, 2007), pp. 345–64.
Schliesser, E., ‘The Obituary of a Vain
Philosopher: Adam Smith’s Reflections on
Hume’s Life’, Hume Studies 29(2) (2003),
pp. 327–62.
Stewart, M.A., The Kirk and the Infidel,
inaugural lecture (Lancaster: Lancester
University Publications Office, 1995; repr.
with corrections, 2001).
Zaretsky, R., and Scott, J.T., The
Philosophers’ Quarrel: Rousseau, Hume,
and the Limits of Human Understanding
(New Haven, CT: Yale University Press,
2009).
5. EARLY RESPONSES TO HUME
Bongie, L.L., David Hume: Prophet
of the Counter-Revolution, 2nd edn
(Indianapolis: Liberty Fund, 2000).
Burton, J.H. (ed.), Letters of Eminent
Persons addressed to David Hume
(Edinburgh: William Blackwood and
Sons, 1849).
Cohen, A.A., ‘The Making of a
Philosophical Classic: The Reception of
David Hume in Europe’, in E. Mazza and
E. Ronchetti (eds), New Essays on David
Hume (Milan: FrancoAngeli, 2007),
pp. 457–68.
Fieser, J. (ed.) Early Responses to Hume:
History of England, 2 vols, 2nd edn
(Bristol: Thoemmes Continuum, 2005).
–––––, Early Responses to Hume: Life and
Reputation, 2 vols, 2nd edn (Bristol:
Thoemmes Continuum, 2005).
401
9780826443595_Ref_Final_txt_print.indd 401
BIBLIOGRAPHY
–––––, Early Responses to Hume:
Metaphysical and Epistemological
Writings, 2 vols, 2nd edn (Bristol:
Thoemmes Continuum, 2005).
–––––, Early Responses to Hume: Moral,
Literary and Political Writings, 2 vols,
2nd edn (Bristol: Thoemmes Continuum,
2005).
–––––, Early Responses to Hume: Writings
on Religion, 2 vols, 2nd edn (Bristol:
Thoemmes Continuum, 2005).
Jones, P. (ed.), The Reception of David
Hume in Europe (London: Thoemmes
Continuum, 2005).
Mossner, E.C., ‘The Continental Reception
of Hume’s Treatise, 1739–1741’, Mind 56
(1947), pp. 31–43.
Norton, D.F., ‘Hume and His Scottish
Critics’, in D.F. Norton, N. Capaldi and
W.L. Robison (eds), McGill Hume Studies
(San Diego, CA: Austin Hill Press, 1979),
pp. 309–24.
Popkin, R.H., ‘The Early Critics of Hume’,
in his The High Road to Pyrrhonism, ed.
R.A. Watson and J.E. Force (San Diego,
CA: Austin Hill Press, 1980), pp. 197–212.
Rivers, I., ‘Responses to Hume on Religion
by Anglicans and Dissenters’, Journal
of Ecclesiastical History 52 (2001),
pp. 675–95.
Spencer, M.G. (ed.), Hume’s Reception in
Early America, 2 vols (Bristol: Thoemmes
Press, 2002).
Tweyman, S. (ed.), Hume on Miracles
(Bristol: Thoemmes, 1996).
–––––, Hume on Natural Religion (Bristol:
Thoemmes, 1996).
Wright, J.P., ‘The Treatise: Composition,
Reception, and Response’, in S. Traiger
(ed.), The Blackwell Guide to Hume’s
Treatise (Oxford: Blackwell Publishing,
2006), pp. 5–25.
2/9/2012 9:38:21 PM
 BIBLIOGRAPHY
6. GENERAL MONOGRAPHS ON
HUME
Allison, H.E., Custom and Reason in Hume:
A Kantian Reading of the First Book of
the Treatise (Oxford: Clarendon Press,
2008).
Anderson, R.F., Hume’s First Principles
(Lincoln, NE: University of Nebraska
Press, 1966).
Ayer, A.J., Hume: A Very Short Introduction
(Oxford: Oxford University Press, 2000).
Baier, A., A Progress of Sentiments:
Reflections on Hume’s Treatise
(Cambridge, MA: Harvard University
Press, 1991).
–––––, Death and Character: Further
Reflections on Hume (Boston, MA:
Harvard University Press, 2008).
Bailey, A., and O’Brien, D., Hume’s Enquiry
Concerning Human Understanding:
Reader’s Guide (London: Continuum,
2006).
Bennett, J., Learning from Six Philosophers:
Descartes, Spinoza, Leibniz, Locke,
Berkeley, Hume, 2 vols (Oxford:
Clarendon Press, 2001).
–––––, Locke, Berkeley, Hume: Central
Themes (Oxford: Oxford University
Press, 1971).
Blackburn, S., How to Read Hume
(London: Granta, 2008).
Botwinick, A., Ethics, Politics and
Epistemology: A Study in the Unity
of Hume’s Thought (Lanham, MD:
University Press of America, 1980).
Buckle, S., Hume’s Enlightenment Tract
(Oxford: Clarendon Press, 2001).
Capaldi, N., David Hume: The Newtonian
Philosopher (Boston, MA: Twayne,
1975).
Cappiello, I., Hume, Critica della ragione
(Naples: Liguori, 1969).
402
9780826443595_Ref_Final_txt_print.indd 402
Çevik, M., David Hume ve Din Felsefesi
(Istanbul: Dergah Yayinlari, 2006).
Chirpaz, F., Hume et le procès de la
métaphysique (Paris: Beauchesne, 1989).
Cléro, J.-P., Hume: une philosophie des
contradictions (Paris: J. Vrin, 1998).
Coventry, A., Hume: A Guide for the
Perplexed (London: Continuum, 2007).
Craig, E.J., David Hume: Eine Einführung
in seine Philosophie (Frankfurt am Main:
Klostermann, 1979).
–––––, The Mind of God and the Works of
Man (Oxford: Clarendon Press, 1987).
Danford, J.W., David Hume and the
Problem of Reason: Recovering the
Human Sciences (New Haven, CT: Yale
University Press, 1990).
Dal Pra, Mario, Hume e la scienza della
natura umana (Rome: Editori Laterza,
1973).
Deleuze, G., Empiricism and Subjectivity:
An Essay on Hume’s Theory of Human
Nature, trans. C.V. Boundas (New York:
Columbia University Press, 1991).
Dicker, G., Hume’s Epistemology and
Metaphysics: An Introduction (Routledge,
1998).
Finlay, C.J., Hume’s Social Philosophy:
Human Nature and Commercial
Sociability in A Treatise of Human
Nature (London: Continuum, 2007).
Flew, A., David Hume: Philosopher of
Moral Science (Oxford: Blackwell
Publishing, 1986).
–––––, Hume’s Philosophy of Belief: A Study
of His First Inquiry (London: Routledge,
1961).
Foglia, G., Immaginazione e Natura Umana:
Studio sulla Theoria della Conoscenza di
David Hume (Bologna: Il Mulino, 1998).
Garrett, D., Cognition and Commitment
in Hume’s Philosophy (Oxford: Oxford
University Press, 1997).
2/9/2012 9:38:21 PM

Hendel, C., Studies in the Philosophy
of Hume (Princeton, NJ: Princeton
University Press, 1925).
Jenkins, J., Understanding Hume
(Edinburgh: Edinburgh University Press,
1992).
Johnson, O.A., The Mind of David Hume:
A Companion to Book I of A Treatise of
Human Nature (Urbana, IL: University of
Illinois Press, 1995).
Kail, P.J.E., Projection and Realism in
Hume’s Philosophy (Oxford: Oxford
University Press, 2007).
Kemp Smith, N., The Philosophy of David
Hume (London: Macmillan, 1941).
Laing, B.M., David Hume (London: Ernest
Benn, 1932).
Laird, J., Hume’s Philosophy of Human
Nature (London: Methuen, 1932).
Leroy, A.L., David Hume (Paris: Presses
Universitaires de France, 1953).
Livingston, D.W., Hume’s Philosophy of
Common Life (Chicago, IL: University of
Chicago Press, 1984).
–––––, Philosophical Melancholy and
Delirium: Hume’s Pathology of
Philosophy (Chicago, IL: University of
Chicago Press, 1998).
Loeb, L., Stability and Justification in
Hume’s Treatise (Oxford: Oxford
University Press, 2002).
MacNabb, D.G.C., David Hume: His
Theory of Knowledge and Morality,
2nd edn (Oxford: Blackwell & Mott,
1966).
Malherbe, M., La Philosophie Empiriste de
David Hume (Paris: J. Vrin, 1984).
Merrill, K.R., Historical Dictionary of
Hume’s Philosophy (Lanham, MD: The
Scarecrow Press, 2008).
Michaud, Y., Hume et la fin de la
philosophie (Paris: Presses Universitaires
de France, 1983).
403
9780826443595_Ref_Final_txt_print.indd 403
BIBLIOGRAPHY
Monteiro, J.P., Hume e a Epistemologia
(Lisbon: Imprensa Nacional-Casa da
Moeda, 1984).
Mounce, H.O., Hume’s Naturalism
(London: Routledge, 1999).
Mwaipaya, P.A., The Foundation of
Hume’s Philosophy (Aldershot: Ashgate,
1999).
Nelson, J.O., Hume’s ‘New Scene of
Thought’ and the Several Faces of Hume
in the Dialogues concerning Natural
Religion (Lanham, MD: University Press
of America, 2009).
Noonan, H., Hume on Knowledge (London:
Routledge, 1999).
–––––, Hume (Oxford: Oneworld
Publications, 2007).
Norton, D.F., David Hume: Common-
Sense Moralist, Sceptical Metaphysician
(Princeton, NJ: Princeton University
Press, 1982).
Noxon, J., Hume’s Philosophical
Development: A Study of his Methods
(Oxford: Clarendon Press, 1973).
Owen, D., Hume’s Reason (Oxford: Oxford
University Press, 1999).
Passmore, J., Hume’s Intentions, 3rd edn
(London: Duckworth, 1980).
Pears, D., Hume’s System: An Examination
of the First Book of his Treatise (Oxford:
Oxford University Press, 1990).
Penelhum, T., David Hume: An Introduction
to his Philosophical System (West
Lafayette, IN: Purdue University Press,
1992).
–––––, Hume (London: Macmillan, 1975).
–––––, Themes in Hume: The Self, the Will,
Religion (Oxford: Clarendon Press,
2000).
Price, J.V., David Hume (New York:
Twayne, 1986).
Quinton, A., Hume (London: Phoenix,
1998).
2/9/2012 9:38:21 PM
 BIBLIOGRAPHY
Rábade Romeo, S., Hume y el fenomenismo
moderno (Madrid: Gredos, 1975).
Russell, P., The Riddle of Hume’s Treatise:
Skepticism, Naturalism, and Irreligion
(Oxford: Clarendon Press, 2008).
Santucci, A., Sistema e ricerca in David
Hume (Bari: Laterza, 1969).
Schmidt, C.M., David Hume: Reason in
History (University Park, PA: Pennsylvania
State University Press, 2003).
Stern, G., A Faculty Theory of Knowledge:
The Aim and Scope of Hume’s First
Enquiry (Lewisburg, PA: Bucknell
University Press, 1971).
Stroud, B., Hume (London: Routledge &
Kegan Paul, 1977).
Waxman, W., Hume’s Theory of
Consciousness (Cambridge: Cambridge
University Press, 1994).
Williams, C., A Cultivated Reason: An
Essay on Hume and Humeanism
(University Park, PA: Pennsylvania State
University Press, 1998).
Wright, J.P., Hume’s A Treatise of Human
Nature: An Introduction (Cambridge:
Cambridge University Press, 2009).
–––––, The Sceptical Realism of David
Hume (Manchester: Manchester
University Press, 1983).
Zabeeh, F., Hume, Precursor of Modern
Empiricism: An Analysis of His Opinions
on Meaning, Metaphysics, Logic and
Mathematics, 2nd edn (The Hague:
Martinus Nijhoff, 1973).
7. GENERAL ANTHOLOGIES
(MULTIPLE AUTHORS)
Chappell, V.C. (ed.), Hume: A Collection
of Critical Essays (London: Macmillan,
1968).
404
9780826443595_Ref_Final_txt_print.indd 404
Frasca-Spada, M., and Kail, P.J.E. (eds),
Impressions of Hume (Oxford:
Clarendon Press, 2005).
Guimarães, L. (ed.), Ensaios sobre Hume
(Belo Horizonte: Segrac Editora, 2005).
Livingston, D.W., and King, J.T. (eds),
Hume: A Re-evaluation (New York:
Fordham University Press, 1976).
Mazza, E., and Ronchetti, E. (eds),
New Essays on David Hume (Milan:
FrancoAngeli, 2007).
Merrill, K.R., and Shahan, R.W. (eds), David
Hume: Many-sided Genius (Norman, OK:
University of Oklahoma Press, 1976).
Millican, P. (ed.), Reading Hume on Human
Understanding (Oxford: Clarendon Press,
2002).
Morice, G.P. (ed.), David Hume:
Bicentenary Papers (Edinburgh:
Edinburgh University Press, 1977).
Norton, D.F. (ed.), The Cambridge
Companion to Hume (Cambridge:
Cambridge University Press, 1993).
Norton, D.F., and Taylor, J. (eds), The
Cambridge Companion to Hume, 2nd
edn (Cambridge: Cambridge University
Press, 2009).
–––––, Capaldi, N., and Robison, W.L. (eds),
McGill Hume Studies (San Diego, CA:
Austin Hill Press, 1979).
Owen, D. (ed.), Hume: General Philosophy
(Aldershot: Ashgate, 2000).
Pears, D. (ed.), David Hume: A Symposium
(London: Macmillan, 1963).
Radcliffe, E.S. (ed.), A Companion to Hume
(Oxford: Wiley-Blackwell, 2008).
Sesonske, A., and Fleming, N. (eds),
Human Understanding: Studies in the
Philosophy of David Hume (Belmont,
CA: Wadsworth, 1968).
Stewart, M.A., and Wright, J.P. (eds), Hume
and Hume’s Connexions (Edinburgh:
Edinburgh University Press, 1994).
2/9/2012 9:38:21 PM

Todd, W.B. (ed.), Hume and the
Enlightenment: Essays presented to
Ernest Campbell Mossner (Edinburgh:
Edinburgh University Press, 1974).
Traiger, S. (ed.), The Blackwell Guide to
Hume’s Treatise (Oxford: Blackwell
Publishing, 2006).
Tweyman, S. (ed.), David Hume: Critical
Assessments, 6 vols (Routledge: London:
1995).
8. THE SOURCES OF HUME’S
THOUGHT
Aarsleff, H., ‘Locke’s Influence’, in V.
Chappell (ed.), The Cambridge Companion
to Locke (Cambridge: Cambridge
University Press, 1994), pp. 252–89.
Ayers, M.R., ‘Berkeley and Hume: A Question
of Influence’, in R. Rorty, J.B. Schneewind
and Q. Skinner (eds), Philosophy in
History (Cambridge: Cambridge University
Press, 1984), pp. 303–27.
Barfoot, M., ‘Hume and the Culture of
Science in the Early Eighteenth Century’,
in M.A. Stewart (ed.), Studies in the
Philosophy of the Scottish Enlightenment
(Oxford: Clarendon Press, 1990),
pp. 151–90.
Brahami, F., Le travail du scepticisme:
Montaigne, Bayle, Hume (Paris: Presses
Universitaires de France, 2001).
Broadie, A., A History of Scottish
Philosophy (Edinburgh: Edinburgh
University Press, 2009).
–––––, The Scottish Enlightenment: The
Historical Age of the Historical Nation
(Edinburgh: Birlinn, 2001).
––––– (ed.), The Cambridge Companion to
the Scottish Enlightenment (Cambridge:
Cambridge University Press, 2003).
405
9780826443595_Ref_Final_txt_print.indd 405
BIBLIOGRAPHY
Buckle, S., Natural Law and the Theory
of Property: Grotius to Hume (Oxford:
Clarendon Press, 1991).
Coventry, A., ‘A Re-examination of Hume’s
Debt to Newton’, in L. Guimarães (ed.),
Ensaios sobre Hume (Belo Horizonte:
Segrac Editora, 2005).
De Pierris, G., ‘Hume and Locke on
Scientific Methodology: The Newtonian
Legacy’, Hume Studies 32(2) (2006),
pp. 277–329.
Drever, J., ‘A Note on Hume’s Pyrrhonism’,
The Philosophical Quarterly 3 (1953),
pp. 40–50.
Echelbarger, C., ‘Hume and the Logicians’,
in P. Easton (ed.), Logic and the Workings
of the Mind: The Logic of Ideas and
Faculty Psychology in Early Modern
Philosophy (Atascadero, CA: Ridgeview
Press, 1997), pp. 137–52.
Emerson, R.L., Essays on David Hume,
Medical Men and the Scottish
Enlightenment: Industry, Knowledge and
Humanity (Farnham: Ashgate, 2009).
Force, J., ‘Hume’s Interest in Newton and
Science’, in J. Force and R.H. Popkin
(eds), Essays on the Context, Nature and
Influence of Isaac Newton’s Theology
(Dordrecht: Kluwer, 1990), pp. 181–206.
Fosl, P.S., ‘The Bibliographical Bases of
Hume’s Understanding of Sextus Empiricus
and Pyrrhonism’, Journal of the History of
Philosophy 36 (1998), pp. 261–78.
–––––, ‘Doubt and Divinity: Cicero’s
Influence on Hume’s Religious
Skepticism’, Hume Studies 20(1) (1994),
pp. 103–20.
Gill, M., The British Moralists on Human
Nature and the Birth of Secular Ethics
(Cambridge: Cambridge University Press,
2006).
Groarke, L., and Solomon, G., ‘Some
Sources for Hume’s Account of Cause’,
2/9/2012 9:38:22 PM
 BIBLIOGRAPHY
Journal of the History of Ideas 52 (1991),
pp. 645–63.
Hall, R., ‘Hume’s Actual Use of Berkeley’s
Principles’, Philosophy 43 (1968),
pp. 278–80.
–––––, ‘Hume’s Use of Locke on Identity’, The
Locke Newsletter 5 (1974), pp. 56–75.
Hampton, J., ‘The Hobbesean Side of
Hume’, in A. Reath, B. Herman and
C.M. Korsgaard (eds), Reclaiming the
History of Ethics: Essays for John Rawls
(Cambridge: Cambridge University Press,
1997), pp. 66–101.
Hope, V. (ed.), Philosophers of the Scottish
Enlightenment (Edinburgh: Edinburgh
University Press, 1984).
Hutchison, T.W., Before Adam Smith:
The Emergence of Political Economy,
1622–1776 (Oxford: Blackwell Publishing,
1988).
James, E.D., ‘Scepticism and Religious
Belief: Pascal, Bayle, Hume’, in R.R.
Bolgar (ed.), Classical Influences on
Western Thought A.D. 1650–1870
(Cambridge: Cambridge University Press,
1979), pp. 93–104.
Jones, P., Hume’s Sentiments: Their
Ciceronian and French Context
(Edinburgh: Edinburgh University Press,
1982).
––––– (ed.), Philosophy and Science in the
Scottish Enlightenment (Edinburgh: John
Donald Publishers, 1988).
––––– (ed.), The ‘Science of Man’ in the
Scottish Enlightenment: Hume, Reid
and Their Contemporaries (Edinburgh:
Edinburgh University Press, 1989).
Kail, P.J.E., ‘On Hume’s Appropriation
of Malebranche: Causation and Self’,
European Journal of Philosophy 16
(2008), pp. 55–80.
Klever, W., ‘Hume contra Spinoza?’, Hume
Studies 16(2) (1990), pp. 89–105.
406
9780826443595_Ref_Final_txt_print.indd 406
–––––, ‘More about Hume’s Debt to
Spinoza’, Hume Studies 19(1) (1993),
pp. 55–74.
Kors, A.C., ‘The French Context of Hume’s
Philosophical Theology’, Hume Studies
21(2) (1995), pp. 221–36.
Kuhns, R., ‘Hume’s Republic and the
Universe of Newton’, in P. Gay (ed.),
Eighteenth-Century Studies, Presented to
A.M. Wilson (Hanover, NH: University
Press of New England, 1972), pp. 73–95.
Kuypers, M.S., Studies in the Eighteenth
Century Background of Hume’s
Empiricism (Minneapolis, MN:
University of Minnesota Press, 1930).
McCosh, J., The Scottish Philosophy,
Biographical, Expository, Critical, From
Hutcheson to Hamilton (New York:
Robert Carter and Brothers, 1875).
McCracken, C.J., Malebranche and British
Philosophy (Oxford: Clarendon Press,
1983).
McIntyre, J.L., ‘Hume: Second Newton of
the Moral Sciences’, Hume Studies 20(1)
(1994), pp. 1–18.
Moore, J., ‘Hume and Hutcheson’, in M.A.
Stewart and J.P. Wright (eds), Hume
and Hume’s Connexions (Edinburgh:
Edinburgh University, 1994), pp. 23–57.
Olshewsky, T.M., ‘The Classical Roots
of Hume’s Skepticism’, Journal of the
History of Ideas 52 (1991), pp. 269–87.
Ott, W., Causation and Laws of Nature
in Early Modern Philosophy (Oxford:
Oxford University Press, 2009).
Owen, D., ‘Hume and the Lockean
Background: Induction and the
Uniformity Principle’, Hume Studies
18(2) (1992), pp. 199–208.
–––––, ‘Hume’s Doubts about Probable
Reasoning: Was Locke the Target?’, in
M.A. Stewart and J.P. Wright (eds), Hume
and Hume’s Connexions (Edinburgh:
2/9/2012 9:38:22 PM

Edinburgh University Press, 1994),
pp. 140–59.
Penelhum, T., ‘Butler and Hume’, Hume
Studies 14(2) (1988), pp. 251–76; repr. in
T. Penelhum, Themes in Hume: The Self,
the Will, Religion (Oxford: Clarendon
Press, 2000), pp. 244–60.
Pittion, J.-P., ‘Hume’s Reading of Bayle: An
Inquiry into the Source and Role of the
Memoranda’, Journal of the History of
Philosophy 15 (1977), pp. 373–86.
Popkin, R.H., ‘Bayle and Hume’, in R.H.
Popkin, The High Road to Pyrrhonism,
ed. R.A. Watson and J.E. Force (San
Diego, CA: Austin Hill Press, 1980),
pp. 149–59.
–––––, ‘David Hume and the Pyrrhonian
Controversy’, The Review of Metaphysics
6 (1952), pp. 65–81; repr. in R.H. Popkin,
The High Road to Pyrrhonism, ed. R.A.
Watson and J.E. Force (San Diego, CA:
Austin Hill Press, 1980), pp. 133–47.
–––––, ‘Hume and Spinoza’, Hume Studies
5(2) (1979), pp. 65–93.
–––––, The High Road to Pyrrhonism, ed.
R.A. Watson and J.E. Force (San Diego,
CA: Austin Hill Press, 1980).
–––––, ‘Sources of Knowledge of Sextus
Empiricus in Hume’s Time’, Journal of the
History of Ideas 54 (1993), pp. 137–41.
Porter, R., Enlightenment: Britain and the
Creation of the Modern World (London:
Penguin, 2001).
Raynor, D., ‘Hume and Berkeley’s Three
Dialogues’, in M.A. Stewart (ed.),
Studies in the Philosophy of the Scottish
Enlightenment (Oxford: Clarendon Press,
1990), pp. 231–50.
Rivers, I., Reason, Grace, and Sentiment: A
Study of the Language of Religion and
Ethics in England 1660–1780, Volume
2: Shaftesbury to Hume (Cambridge:
Cambridge University Press, 2000).
407
9780826443595_Ref_Final_txt_print.indd 407
BIBLIOGRAPHY
Ross, I., Lord Kames and the Scotland of
His Day (Oxford: Clarendon Press, 1972).
Russell, P., ‘Epigram, Pantheists and
Freethought in Hume’s Treatise’, Journal
of the History of Ideas 54 (1993),
pp. 659–73.
–––––, ‘Hume’s Treatise and Hobbes’ The
Elements of Law’, Journal of the History
of Ideas 46 (1985), pp. 51–64.
–––––, ‘Hume’s Treatise and the Clarke-
Collins Controversy’, Hume Studies 21(1)
(1995), pp. 95–115.
Sapadin, E., ‘Newton, First Principles, and
Reading Hume’, Archiv für Geschichte
der Philosophie 74 (1992), pp. 74–104.
Sher, R.B., Church and University in the
Scottish Enlightenment: The Moderate
Literati of Edinburgh (Edinburgh:
Edinburgh University Press, 1985).
Stewart, M.A., ‘Hume’s Intellectual
Development 1711–1752’, in M. Frasca-
Spada and P.J.E. Kail (eds), Impressions
of Hume (Oxford: Clarendon Press,
2005), pp. 11–58.
–––––, (ed.), Studies in the Philosophy of
the Scottish Enlightenment (Oxford:
Clarendon Press, 1990).
Talmor, E., Descartes and Hume (Oxford:
Pergamon Press, 1980).
Turco, L., ‘MacLaurin, Reid, and Kemp Smith
on the Ancestry of Hume’s Philosophy’,
Reid Studies 2 (1999), pp. 71–87.
Woolhouse, R.S., The Empiricists (Oxford:
Oxford University Press, 1988).
9. LITERARY AMBITIONS AND
PRACTICE
Box, M.A., The Suasive Art of David Hume
(Princeton, NJ: Princeton University
Press, 1990).
2/9/2012 9:38:22 PM
 BIBLIOGRAPHY
Braudy, L., Narrative Form in History and
Fiction: Hume, Fielding and Gibbon
(Princeton, NJ: Princeton University
Press, 1970).
Christensen, J., Practicing Enlightenment:
Hume and the Formation of a Literary
Career (Madison: University of Wisconsin
Press, 1987).
Morrisroe, M., ‘Rhetorical Methods in
Hume’s Works on Religion’, Philosophy
and Rhetoric 2 (1969), pp. 121–38.
Loptson, P., ‘Hume, Multiperspectival
Pluralism, and Authorial Voice’, Hume
Studies 24(2) (1998), pp. 313–34.
Parker, F., Scepticism and Literature: An
Essay on Pope, Hume, Sterne, and
Johnson (Oxford: Oxford University
Press, 2003).
Price, J.V., The Ironic Hume (Austin, TX;
University of Texas Press, 1965).
Richetti, J.J., Philosophical Writing: Locke,
Berkeley, Hume (Cambridge, MA:
Harvard University Press, 1983).
10. PHILOSOPHICAL METHODOLOGY
AND INTENTIONS
Abramson, K., ‘Hume’s Distinction
between Philosophical Anatomy and
Painting’, Philosophy Compass 2 (2007),
pp. 680–98.
Árdal, P.S., ‘Some Implications of the
Virtue of Reasonableness in Hume’s
Treatise’, in D.W. Livingston and J.T.
King (eds), Hume: A Re-Evaluation
(New York: Fordham University Press,
1976).
Coleman, D.P., ‘Hume’s Dialectic’, Hume
Studies 10(2) (1984), pp. 139–55.
Dauer, F., ‘Towards a Copernican Reading
of Hume’, Noûs 9 (1975), pp. 269–93.
408
9780826443595_Ref_Final_txt_print.indd 408
Harris, J.A., ‘“A compleat chain of
reasoning”: Hume’s Project in A Treatise
of Human Nature, Books One and Two’,
Proceedings of the Aristotelian Society
109 (2009), pp. 129–48.
Henze, D., ‘The Linguistic Aspect of Hume’s
Method’, Journal of the History of Ideas
30 (1969), pp. 116–26.
Heydt, C., ‘Relations of Literary Form and
Philosophical Purpose in Hume’s Four
Essays on Happiness’, Hume Studies
33(1) (2007), pp. 3–19.
Immerwahr, J., ‘A Skeptic’s Progress: Hume’s
Preference for Enquiry 1’, in D.F. Norton,
N. Capaldi, W.L. Robison (eds), McGill
Hume Studies (San Diego, CA: Austin
Hill Press, 1979), pp. 227–38.
Millican, P., ‘The Context, Aims and
Structure of Hume’s First Enquiry’,
in P. Millican (ed.), Reading Hume
on Human Understanding (Oxford:
Clarendon Press, 2002), pp. 27–65.
Moses, G., ‘Hume’s Playful Metaphysics’,
Hume Studies 18(1) (1992), pp. 63–80.
Penelhum, T., ‘David Hume: An
Appreciation’, in his Themes in Hume:
The Self, the Will, Religion (Oxford:
Clarendon Press, 2000), pp. 1–22.
Stewart, M.A., ‘Two Species of Philosophy:
The Historical Significance of the First
Enquiry’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 67–95.
11. IMPRESSIONS AND IDEAS
Bricke, J., ‘Hume’s Associationist
Psychology’, Journal of the History
of the Behavioral Sciences 8 (1974),
pp. 397–409.
2/9/2012 9:38:22 PM

Broughton, J., ‘Impressions and Ideas’, in
S. Traiger (ed.), The Blackwell Guide
to Hume’s Treatise (Oxford: Blackwell
Publishing, 2006), pp. 43–58.
Butler, R.J., ‘Hume’s Impressions’, Royal
Institute of Philosophy Lectures 9 (1975),
pp. 122–36.
Dauer, F.W., ‘Force and Vivacity in the
Treatise and the Enquiry’, Hume
Studies 25 (1999), pp. 83–99.
Everson, S., ‘The Difference between
Feeling and Thinking’, Mind 97 (1988),
pp. 401–13.
Fodor, J.A., Hume Variations (Oxford:
Clarendon, 2003).
Frasca-Spada, M., ‘Simple Perceptions in
Hume’s Treatise’, in E. Mazza and E.
Ronchetti (eds), New Essays on David
Hume (Milan: FrancoAngeli, 2007),
pp. 37–54.
Garrett, D., ‘Hume’s Theory of Ideas’, in
E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 41–57.
Govier, T., ‘Variations on Force and Vivacity
in Hume’, The Philosophical Quarterly
22 (1972), pp. 44–52.
Landy, D., ‘Hume’s Impression/Idea
Distinction’, Hume Studies 32(1) (2006),
pp. 119–40.
Rollin, B., ‘Hume’s Blue Patch and
the Mind’s Creativity’, The Journal
of the History of Ideas 32 (1971),
pp. 119–28.
Savage, R., ‘Hume’s Missing Shade of
Blue’, History of Philosophy Quarterly 9
(1992), pp. 199–206.
Schmicking, D.A., ‘Hume’s Theory of Simple
Perceptions Reconsidered’, Hume Studies
30(1) (2004), pp. 3–31.
Traiger, S., ‘Impressions, Ideas, and
Fictions’, Hume Studies 13(2) (1987),
pp. 381–99.
409
9780826443595_Ref_Final_txt_print.indd 409
BIBLIOGRAPHY
Waxman, W., ‘Impressions and Ideas:
Vivacity as Verisimilitude’, Hume Studies
19(1) (1993), pp. 75–88.
12. IDEAS AND MEANING
Bennett, J., ‘Empiricism about Meanings’,
in P. Millican (ed.), Reading Hume
on Human Understanding (Oxford:
Clarendon Press, 2002), pp. 97–106.
Craig, E.J., ‘Hume on Thought and Belief’,
in G. Vesey (ed.), Philosophers Ancient
and Modern (Cambridge: Cambridge
University Press), pp. 93–110.
Flage, D., ‘Hume’s Relative Ideas’, Hume
Studies 7 (1981), pp. 55–73.
–––––, ‘Relative Ideas Re-viewed’, in
R. Read and K.A. Richman (eds), The
New Hume Debate (London: Routledge,
2000), pp. 138–55.
Livingston, D.W., ‘Hume’s Historical Theory
of Meaning’, in D.W. Livingston and J.T.
King (eds), Hume: A Re-evaluation (New
York: Fordham University Press, 1976),
pp. 213–38.
Losee, J., ‘Hume’s Demarcation Project’,
Hume Studies 18(1) (1992), pp. 51–62.
McRae, R., ‘Hume on Meaning’, Dialogue 8
(1969), pp. 486–91.
Ott, W., ‘Hume on Meaning’, Hume Studies
32(2) (2006), pp. 233–52.
Thomas, M.M., ‘Relative Ideas Rejected’,
Hume Studies 7(2) (1982), pp. 149–57.
Wright, J.P., ‘Hume’s Rejection of
the Theory of Ideas’, History of
Philosophy Quarterly 8 (1991),
pp. 149–62.
Zartman. J.F., ‘Hume and “the meaning
of a word”’, Philosophy and
Phenomenological Research 36 (1975),
pp. 255–60.
2/9/2012 9:38:22 PM
 BIBLIOGRAPHY
13. CONCEIVABILITY AND
POSSIBILITY
Garrett, D., ‘Simplicity and Separability
in Hume’s Empiricism’, Archiv für
Geschichte der Philosophie 67 (1985),
pp. 270–88.
Gendler, T.S., and Hawthorne, J.,
‘Introduction: Conceivability and
Possibility’, in T.S. Gendler and J.
Hawthorne (eds), Conceivability and
Possibility (Oxford: Clarendon Press,
2002), pp. 1–70.
Lightner, T., ‘Hume on Conceivability and
Inconceivability’, Hume Studies 23(1)
(1997), pp. 113–32.
Suchting, W.A., ‘Hume and Necessary
Truth’, Dialogue 5 (1966–7), pp. 47–60.
Van Woudenberg, R., ‘Conceivability and
Modal Knowledge’, Metaphilosophy 37
(2006), pp. 210–21.
14. SPACE AND TIME
Baxter, D.L.M., Hume’s Difficulty: Time
and Identity in the Treatise (London:
Routledge, 2008).
–––––, ‘Hume’s Theory of Space and Time
in Its Sceptical Context’, in D.F. Norton
and J. Taylor (eds), The Cambridge
Companion to Hume, 2nd edn
(Cambridge: Cambridge University Press,
2009), pp. 105–46.
Broad, C.D., ‘Hume’s Doctrine of Space’,
Proceedings of the British Academy 47
(1961), pp. 161–76.
Falkenstein, L., ‘Space and Time’, in S.
Traiger (ed.), The Blackwell Guide to
Hume’s Treatise (Oxford: Blackwell,
2006), pp. 59–76.
410
9780826443595_Ref_Final_txt_print.indd 410
Frasca-Spada, M., Space and the Self in
Hume’s Treatise (Cambridge: Cambridge
University Press, 1998).
Holden, T., ‘Infinite Divisibility and Actual
Parts’, Hume Studies 28(1) (2002),
pp. 3–25.
Jacquette, D., David Hume’s Critique of
Infinity (Leiden: Brill, 2001).
–––––, ‘Hume on Infinite Divisibility and
the Negative Idea of a Vacuum’, British
Journal for the History of Philosophy 10
(2002), pp. 413–35.
–––––, ‘Hume on the Infinite Divisibility of
Extension and Exact Geometrical Values’,
in E. Mazza and E. Ronchetti (eds),
New Essays on David Hume (Milan:
FrancoAngeli, 2007), pp. 81–100.
Johnson, O., ‘Time and the Idea of Time’,
Hume Studies 15(1) (1989), pp. 205–19.
Newman, R., ‘Hume on Space and
Geometry’, Hume Studies 7(1) (1981),
pp. 1–31.
Waxman, W., ‘Hume and the Origins of Our
Ideas of Space and Time’, in E.S. Radcliffe
(ed.), A Companion to Hume (Oxford:
Wiley-Blackwell, 2008), pp. 72–88.
15. CAUSATION AND CAUSAL
REASONING
Beauchamp, T.L., and Rosenberg, A., Hume
and the Problem of Causation (New
York: Oxford University Press, 1981).
Beebee, H., Hume on Causation (London:
Routledge, 2006).
Bell, M., ‘Hume on Causation’, in D.F.
Norton and J. Taylor (eds), The
Cambridge Companion to Hume, 2nd
edn (Cambridge: Cambridge University
Press, 2009), pp. 147–76.
2/9/2012 9:38:22 PM

–––––, ‘Sceptical Doubts concerning Hume’s
Causal Realism’, in R. Read and K.A.
Richman (eds), The New Hume Debate
(London: Routledge, 2000), pp. 122–37.
Blackburn, S., ‘Hume and Thick
Connexions’, in R. Read and K.A.
Richman (eds), The New Hume Debate
(London: Routledge, 2000), pp. 100–12.
Broughton, J., ‘Hume’s Ideas about
Necessary Connection’, Hume Studies
13(2) (1987), pp. 217–44.
Collier, M., ‘A New Look at Hume’s Theory
of Probabilistic Inference’, Hume Studies
31(1) (2005), pp. 21–36.
Costa, M.J., ‘Hume and Causal Realism’,
Australasian Journal of Philosophy 67
(1989), pp. 172–90.
Coventry, A., Hume’s Theory of Causation:
A Quasi-Realist Interpretation (London:
Continuum, 2006).
Craig, E.J., ‘Hume on Causality: Projectivist
and Realist’, in R. Read and K.A.
Richman (eds), The New Hume Debate
(London: Routledge, 2000), pp. 113–21.
–––––, ‘The Idea of Necessary
Connexion’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 211–57.
De Pierris, G., ‘Causation as a Philosophical
Relation in Hume’, Philosophy and
Phenomenological Research 64 (2002),
pp. 499–545.
Falkenstein, L., and Welton, D., ‘Humean
Contiguity’, History of Philosophy
Quarterly 18 (2001), pp. 279–96.
Garrett, D., ‘The Representation of
Causation and Hume’s Two Definitions
of “Cause”’, in M. Atherton (ed.), The
Empiricists: Critical Essays on Locke,
Berkeley, and Hume (Lanham, MD:
Rowman & Littlefield), pp. 153–82.
411
9780826443595_Ref_Final_txt_print.indd 411
BIBLIOGRAPHY
Gower, B., ‘Hume on Probability’, British
Journal for the Philosophy of Science 42
(1991), pp. 1–19.
Hacking, I., ‘Hume’s Species of Probability’,
Philosophical Studies 33 (1978),
pp. 21–37.
Hanfling, O., ‘Hume’s Idea of Necessary
Connexion’, Philosophy 54 (1979),
pp. 501–14.
Hearn, T., ‘General Rules in Hume’s
Treatise’, Journal of the History of
Philosophy 8 (1970), pp. 405–22.
Jacobson, A.J., ‘Does Hume Hold a
Regularity Theory?’, History of
Philosophy Quarterly 1 (1984),
pp. 75–91.
Lesher, J., ‘Hume’s Analysis of Cause and
the “Two Definitions” Dispute’, Journal
of the History of Philosophy 11 (1973),
pp. 387–92.
Mackie, J.L., ‘Causes and Conditions’,
American Philosophical Quarterly 2
(1965), pp. 245–64.
Martin, M., ‘The Rational Warrant for
Hume’s General Rules’, Journal of
the History of Philosophy 31 (1993),
pp. 245–57.
McBreen, B., ‘Realism and Empiricism
in Hume’s Account of Causality’,
Philosophy 82 (2007), pp. 421–36.
Millican, P., ‘Hume, Causal Realism, and
Causal Science’, Mind 118 (2009),
pp. 647–712.
Mura, A., Dal noto all’ignoto: causalità e
induzione nel pensiero di David Hume
(Pisa: Edizioni ETS, 1996).
–––––, ‘Hume’s Inductive Logic’, Synthese
115 (1998), pp. 303–31.
Ott, W., ‘What Can Causal Claims Mean?’,
Philosophia 37 (2009), pp. 459–70.
Phalen, A.K., ‘Hume’s Psychological
Explanation of the Idea of Causality’,
2/9/2012 9:38:23 PM
 BIBLIOGRAPHY
International Philosophical Quarterly 17
(1977), pp. 43–57.
Read, R., and Richman, K.A. (eds), The
New Hume Debate (London: Routledge,
2000).
–––––, The New Hume Debate, rev. edn
(London: Routledge, 2007).
Richards, T.J., ‘Hume’s Two Definitions of
“Cause”’, The Philosophical Quarterly
15 (1965), pp. 247–53.
Robinson, J.A., ‘Hume’s Two Definitions of
“Cause”’, The Philosophical Quarterly
12 (1962), 162–71.
Rosenberg, A., ‘Hume and the Philosophy
of Science’, in D.F. Norton (ed.), The
Cambridge Companion to Hume
(Cambridge: Cambridge University Press,
1993), pp. 64–89.
Roth, A.S., ‘Causation’, in S. Traiger (ed.),
The Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 95–113.
Strawson, G., ‘David Hume: Objects and
Power’, in R. Read and K.A. Richman
(eds), The New Hume Debate (London:
Routledge, 2000), pp. 31–51.
–––––, The Secret Connexion: Causation,
Realism, and David Hume (Oxford:
Clarendon Press, 1989).
Stroud, B., ‘“Gilding or Staining” the World
with “Sentiments” and “Phantasms”’, in
Hume Studies 29(2) (November 1993),
pp. 253–72; repr. in R. Read and K.A.
Richman (eds), The New Hume Debate
(London: Routledge, 2000), pp. 16–30.
–––––, ‘Hume and the Idea of Causal
Necessity’, Philosophical Studies 33
(1978), pp. 39–59.
Wilson, F., Hume’s Defence of Causal
Inference (Toronto: University of Toronto
Press, 1997).
Winkler, K., ‘The New Hume’, in R. Read
and K.A. Richman (eds), The New Hume
412
9780826443595_Ref_Final_txt_print.indd 412
Debate (London: Routledge, 2000),
pp. 52–87.
Wright. J.P., ‘Hume’s Causal Realism:
Recovering a Traditional Interpretation’,
in R. Read and K.A. Richman (eds), The
New Hume Debate (London: Routledge,
2000), pp. 88–99.
16. SCEPTICISM ABOUT CAUSAL
REASONING
Arnold, N.S., ‘Hume’s Skepticism about
Inductive Inference’, Journal of the
History of Philosophy 21 (1983),
pp. 31–56.
Beauchamp, T.L., and Mappes, T.A., ‘Is
Hume Really a Sceptic about Induction?’,
American Philosophical Quarterly 12
(1975), pp. 119–29.
Broughton, J., ‘Hume’s Skepticism about
Causal Inferences’, Pacific Philosophical
Quarterly 64 (1983), pp. 3–18.
Howson, C., Hume’s Problem: Induction
and the Justification of Belief (Oxford:
Clarendon Press, 2000).
Jacobson, A., ‘Inductive Scepticism and
Experimental Reasoning in Moral
Subjects in Hume’s Philosophy’, Hume
Studies 15(2) (1989), pp. 325–38.
Loeb, L., ‘Causal Inference, Associationism,
and Scepticism in Part III of Book
I of Hume’s Treatise’, in P. Easton
(ed.), Logic and the Workings of the
Mind: The Logic of Ideas and Faculty
Psychology in Early Modern Philosophy
(Atascadero, CA: Ridgeview, 1997),
pp. 283–305.
–––––, ‘Psychology, Epistemology, and
Skepticism in Hume’s Argument about
Induction’, Synthese 152 (2006),
pp. 321–38.
2/9/2012 9:38:23 PM

Millican, P., ‘Hume on Reason and
Induction: Epistemology or Cognitive
Science?’, Hume Studies 24(1) (1998),
pp. 141–59.
–––––, ‘Hume’s Sceptical Doubts concerning
Induction’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 107–74.
Morris, W.E., ‘Hume’s Refutation of
Inductive Probabilism’, in J.H. Fetzer
(ed.), Probability and Causality: Essays in
Honor of Wesley C. Salmon (Dordrecht:
D. Reidel, 1988), pp. 43–77.
Robison, W.L., ‘Hume’s Causal Scepticism’,
in G.P. Morice (ed.), David Hume:
Bicentenary Papers (Edinburgh: Edinburgh
University Press, 1977), pp. 156–66.
Stove, D.C., Probability and Hume’s
Inductive Scepticism (Oxford: Clarendon
Press, 1973).
Winkler, K., ‘Hume’s Inductive Skepticism’,
in M. Atherton (ed.), The Empiricists;
Critical Essays on Locke, Berkeley,
and Hume (Lanham, MD: Rowman &
Littlefield), pp. 183–212.
17. THE NATURE AND ORIGINS OF
BELIEF
Bell, M., ‘Belief and Instinct in Hume’s First
Enquiry’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 175–85.
Flage, D.E., ‘Hume on Memory and
Causation’, Hume Studies 10th
Anniversary Issue (1984), pp. 168–88.
Frasca-Spada, M., ‘Belief and Animal Spirits
in Hume’s Treatise’, Eighteenth-Century
Thought 1(2003), pp. 151–69.
413
9780826443595_Ref_Final_txt_print.indd 413
BIBLIOGRAPHY
Gorman, M., ‘Hume’s Theory of Belief’,
Hume Studies 21(1) (1993), pp. 89–102.
Hansen, S.J., ‘Hume’s Impression of Belief’,
Hume Studies 14(2) (1988), pp. 277–304.
Hodges, M., and Lachs, J., ‘Hume on
Belief’, Review of Metaphysics 30 (1976),
pp. 3–18.
Jones, W.E., ‘Explaining Our Own Beliefs:
Non-epistemic Believing and Doxastic
Instability’, Philosophical Studies 111
(2002), pp. 217–49.
McDonough, J.K., ‘Hume’s Account of
Memory’, British Journal for the History
of Philosophy 10 (2002), pp. 71–87.
Marušić, J.S., ‘Does Hume Hold a
Dispositional Account of Belief?’,
Canadian Journal of Philosophy 40
(2010), pp. 155–84.
Nuyen, A.T., ‘The Role of Reason in Hume’s
Theory of Belief’, Hume Studies 14(2)
(1988), pp. 372–89.
Traiger, S., ‘Hume on Memory and
Imagination’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 58–71.
18. JUSTIFICATION AND NATURALISM
Bell, M., and McGinn, M., ‘Naturalism
and Scepticism’, Philosophy 65 (1990),
pp. 399–418.
Broughton, J., ‘Hume’s Naturalism about
Cognitive Norms’, Philosophical Topics
31 (2003), pp. 1–20.
–––––, ‘Hume’s Naturalism and His
Skepticism’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 425–40.
Connon, R.W., ‘The Naturalism of Hume
Revisited’, in D.F. Norton, N. Capaldi
and W.L. Robison (eds), McGill Hume
2/9/2012 9:38:23 PM
 BIBLIOGRAPHY
Studies (San Diego, CA: Austin Hill Press,
1979), pp. 121–45.
Falkenstein, L., ‘Naturalism, Normativity,
and Scepticism in Hume’s Account of
Belief’, Hume Studies 23(1) (1997),
pp. 29–72.
Ferreira, M.J., ‘Hume’s Naturalism –
“Proof” and Practice’, The Philosophical
Quarterly 35 (1985), pp. 45–57.
Garrett, D., ‘Reasons to Act and Believe:
Naturalism and Rational Justification
in Hume’s Philosophical Project’,
Philosophical Studies 132 (2007),
pp. 1–16.
Kemp Smith, N., ‘The Naturalism of Hume’,
Mind 14 (1905), pp. 149–73, 335–47.
Loeb, L., ‘Hume on Stability, Justification,
and Unphilosophical Probability’, Journal
of the History of Philosophy 33 (1995),
pp. 101–32.
–––––, ‘Instability and Uneasiness in Hume’s
Theories of Belief and Justification’,
British Journal of the History of
Philosophy 3 (1995), pp. 301–27.
–––––, ‘Stability, Justification, and Hume’s
Propensity to Ascribe Identity to Related
Objects’, Philosophical Topics 19 (1991),
pp. 237–70.
McCormick, M., ‘Why Should We Be
Wise?’, Hume Studies 31(1) (2005),
pp. 3–19.
Michaud, Y., ‘Hume’s Naturalized
Philosophy’, Hume Studies 13(2) (1987),
pp. 360–80.
Morris, W.E., ‘Belief, Probability,
Normativity’, in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 77–94.
Owen, D., ‘Philosophy and the Good Life:
Hume’s Defence of Probable Reasoning’,
Dialogue 35 (1996), pp. 485–504.
414
9780826443595_Ref_Final_txt_print.indd 414
Pears, D., ‘The Naturalism of Book I of
Hume’s Treatise of Human Nature’,
Proceedings of the British Academy 62
(1976), pp. 249–68.
Ridge, M., ‘Epistemology Moralized: David
Hume’s Practical Epistemology’, Hume
Studies 29(2) (2003), pp. 165–204.
Singer, I., ‘Nature Breaks Down: Hume’s
Problematic Naturalism in Treatise I.iv’,
Hume Studies 26(2) (2000), pp. 225–43.
Stroud, B., ‘The Constraints of Hume’s
Naturalism’, Synthese 152 (2006),
pp. 339–51.
Weintraub, R., ‘The Naturalistic Response
to Scepticism’, Philosophy 78 (2003),
pp. 369–86.
Wilson, F., ‘Hume’s Cognitive Stoicism’,
Hume Studies 10th Anniversary Issue
(1984), pp. 52–68.
Winters, B., ‘Hume on Reason’, Hume
Studies 5(1) (1979), pp. 20–35.
Wright, J.P., ‘Kemp Smith and the Two Kinds
of Naturalism in Hume’s Philosophy’,
in E. Mazza and E. Ronchetti (eds),
New Essays on David Hume (Milan:
FrancoAngeli, 2007), pp. 17–36.
19. THE EXTERNAL WORLD
Aune, B., Knowledge of the External World
(London: Routledge, 1991).
Baxter, D.L.M., ‘Identity, Continued
Existence, and the External World’, in
S. Traiger (ed.), The Blackwell Guide
to Hume’s Treatise (Oxford: Blackwell
Publishing, 2006), pp. 114–32.
Black, T., ‘The Distinction between
Coherence and Constancy in Hume’s
Treatise I.iv.2’, British Journal for
the History of Philosophy 15 (2007),
pp. 1–25.
2/9/2012 9:38:23 PM

Brett, N., ‘Scepticism and Vain Questions’,
Dialogue 13 (1974), pp. 657–73.
Butler, A., ‘Natural Instinct, Perceptual
Relativity, and Belief in the External
World in Hume’s Enquiry’, Hume Studies
34(1) (2008), pp. 115–58.
Collier, M., ‘Filling the Gaps: Hume and
Connectionism on the Continued
Existence of Unperceived Objects’, Hume
Studies 25(1&2) (1999), pp. 155–70.
Cook, J.W., ‘Hume’s Scepticism with Regard
to the Senses’, American Philosophical
Quarterly 5 (1968), pp. 1–17.
Costa, M., ‘Hume and Belief in the Existence
of the External World’, Philosophical
Studies 32 (1988), pp. 99–112.
Dicker, G., ‘Three Questions about Treatise
1.4.2’, Hume Studies 33(1) (2007),
pp. 115–54.
Grene, M., ‘The Objects of Hume’s Treatise’,
Hume Studies 20(2) (1994), pp. 163–77.
Malherbe, M., ‘Hume on Belief in the
External World’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 126–39.
O’Shea, J.R., ‘Hume’s Reflective Return
to the Vulgar’, British Journal for
the History of Philosophy 4 (1996),
pp. 285–315.
Price, H.H., Hume’s Theory of the
External World (Oxford: Clarendon
Press, 1940).
Rocknak, S., ‘The Vulgar Conception of
Objects in ‘Of skepticism with Regard to
the Senses’, Hume Studies 33(1) (2007),
pp. 67–90.
Schnall, I.M., ‘Constancy, Coherence, and
Causality’, Hume Studies 30(1) (2004),
pp. 33–50.
Weller, C., ‘Why Hume Is a Direct Realist’,
Archiv für Geschichte der Philosophie 83
(2001), pp. 258–85.
415
9780826443595_Ref_Final_txt_print.indd 415
BIBLIOGRAPHY
Wilson, F., ‘Is Hume a Sceptic with Regard
to the Senses?’, Journal of the History of
Philosophy 27 (1989), pp. 49–73.
–––––, The External World and Our
Knowledge of It: Hume’s Critical Realism,
an Exposition and a Defence (Toronto:
University of Toronto Press, 2008).
20. SCEPTICISM AND PHILOSOPHY
Allison, H.E., ‘Hume’s Philosophical
Insouciance: A Reading of Treatise
1.4.7’, Hume Studies 31(2) (2005),
pp. 317–46.
Annas, J., ‘Hume and Ancient Scepticism’,
in J. Sihvola (ed.), Ancient Scepticism
and the Sceptical Tradition (Helsinki:
Philosophical Society of Finland, 2000),
pp. 271–85.
Campbell, S.M., ‘The Surprise Twist in
Hume’s Treatise’, Hume Studies 35(1&2)
(2009), pp. 103–34.
Castiglione, D., ‘Hume’s Two Views of
Modern Scepticism’, History of European
Ideas 32 (2006), pp. 1–27.
Cohen, A., ‘Scepticism and “Angst”: The
Case of David Hume’, Manuscrito 11
(1988), pp. 49–66.
Cummins, P.D., ‘Hume’s Diffident
Skepticism’, Hume Studies 25(1&2)
(1999), pp. 43–65.
Dauer, F.W., ‘Hume’s Scepticism with
Regard to Reason: A Reconsideration’,
Hume Studies 22(2) (1996), pp. 211–29.
De Pierris, G., ‘Hume’s Pyrrhonian
Skepticism and the Belief in Causal
Laws’, Journal of the History of
Philosophy 39 (2001), pp. 351–83.
DeWitt, R., ‘Hume’s Probability Argument
of I.iv.1’, Hume Studies 11(2) (1985),
pp. 125–40.
2/9/2012 9:38:23 PM
 BIBLIOGRAPHY
Fieser, J., ‘Hume’s Pyrrhonism: A
Developmental Interpretation’, Hume
Studies 15(1) (1989), pp. 93–119.
Fogelin, R.J., ‘Hume’s Scepticism’, in D.F.
Norton (ed.), The Cambridge Companion
to Hume (Cambridge: Cambridge
University Press, 1993), pp. 90–116.
–––––, Hume’s Skeptical Crisis: A Textual
Study (New York: Oxford University
Press, 2009).
–––––, Hume’s Skepticism in the Treatise of
Human Nature (London: Routledge &
Kegan Paul, 1985).
–––––, ‘The Tendency of Hume’s Skepticism’,
in M. Burnyeat (ed.), The Skeptical
Tradition (Berkeley, CA: University of
California Press, 1983), pp. 397–412.
Garrett, D., ‘Hume’s Conclusion in
“Conclusion of This Book”’, in S. Traiger
(ed.), The Blackwell Guide to Hume’s
Treatise (Oxford: Blackwell Publishing,
2006), pp. 151–76.
–––––, ‘Ideas, Reason, and Scepticism:
Replies to My Critics’, Hume Studies 24
(1998), pp. 171–94.
Hakkarainen, J., Hume’s Scepticism and
Realism: His Two Profound Arguments
Against the Senses (Finland: Tampere
University Press, 2007).
Imlay, R.A., ‘Hume’s “Of Scepticism
with Regard to Reason”: A Study in
Contrasting Themes’, Hume Studies 7(2)
(1981), pp. 121–36.
Karlsson, M., ‘Epistemic Leaks and
Epistemic Meltdowns: A Response to
William Morris on Scepticism with
Regard to Reason’, Hume Studies 16(2)
(1990), pp. 121–30.
Kekes, J., ‘The Case for Scepticism’, The
Philosophical Quarterly 25 (1975),
pp. 28–39.
Kuehn, M., ‘Hume’s Antinomies’, Hume
Studies 9(1) (1983), pp. 25–45.
416
9780826443595_Ref_Final_txt_print.indd 416
Laursen, J.C., The Politics of Skepticism
in the Ancients, Montaigne, Hume, and
Kant (Leiden: Brill, 1992).
LoLordo, A., ‘Probability and Skepticism
about Reason in Hume’s Treatise’, British
Journal for the History of Philosophy 8
(2000), pp. 419–46.
Loptson, P., ‘Phenomenological Skepticism
in Hume’, Southern Journal of Philosophy
28 (1990), pp. 367–88.
Lynch, M.P., ‘Hume and the Limits of
Reason’, Hume Studies 22(1) (1996),
pp. 89–104.
Mazza, E., ‘In and Out of the Well: Flux
and Reflux of Scepticism and Nature’,
in E. Mazza, and Ronchetti, E. (eds),
New Essays on David Hume (Milan:
FrancoAngeli, 2007), pp. 101–32.
Meeker, K., ‘Hume: Radical Skeptic or
Naturalized Epistemologist?’, Hume
Studies 24(1) (1998), pp. 31–52.
–––––, ‘Hume’s Iterative Probability
Argument: A Pernicious Reductio’,
Journal of the History of Philosophy 38
(2000), pp. 221–38.
Michaud, Y., ‘How to Become a
Moderate Skeptic: Hume’s Way Out of
Pyrrhonism’, Hume Studies 11(1) (1985),
pp. 33–46.
Morris, W.E., ‘Hume’s Scepticism about
Reason’, Hume Studies 15(1) (1989),
pp. 39–60.
Norton, D.F., ‘How a Sceptic May Live
Scepticism’, in J.J. MacIntosh and H.A.
Meynell (eds), Faith, Scepticism, and
Personal Identity: A Festschrift for
Terence Penelhum (Calgary: University of
Calgary Press, 1994), pp. 119–39.
–––––, ‘Of the Academical or Sceptical
Philosophy’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 371–92.
2/9/2012 9:38:23 PM

Owen, D., ‘Reason, Reflection, and
Reductios’, Hume Studies 20(2) (1994),
pp. 195–210.
Popkin, R.H., ‘Hume’s Pyrrhonism and
His Critique of Pyrrhonism’, The
Philosophical Quarterly 1 (1951),
pp. 385–407; repr. in R.H. Popkin, The
High Road to Pyrrhonism, ed. R.A.
Watson and J.E. Force (San Diego, CA:
Austin Hill Press, 1980), pp. 103–32.
Ribeiro, B., ‘Hume’s Changing Views
on the “Durability” of Scepticism’,
Journal of Scottish Philosophy 7 (2009),
pp. 215–36.
Robison, W.L., ‘David Hume: Naturalist and
Meta-Sceptic’, in D.W. Livingston and
J.T. King (eds), Hume: A Re-evaluation
(New York: Fordham University Press,
1976), pp. 23–49.
–––––, ‘Hume’s Scepticism’, Dialogue 12
(1973), pp. 87–99.
Santucci, A., L’umanismo scettico di David
Hume (Bologna: Zanichelli, 1965).
Schnall, I.M., ‘Hume on “Popular” and
“Philosophical” Skeptical Arguments’,
Hume Studies 33(1) (2007), pp. 41–66.
Singer, I., ‘Hume’s Extreme Skepticism in
Treatise I.iv.7’, Canadian Journal of
Philosophy 25 (1995), pp. 595–622.
Stanistreet, P., Hume’s Scepticism and the
Science of Human Nature (Aldershot:
Ashgate, 2002).
Stroud, B., ‘Hume’s Scepticism: Natural
Instincts and Philosophical Reflection’,
in M. Atherton (ed.), The Empiricists:
Critical Essays on Locke, Berkeley,
and Hume (Lanham, MD: Rowman &
Littlefield), pp. 229–52.
Waxman, W., ‘The Point of Hume’s
Skepticism with Regard to Reason: The
Primacy of Facility Affect in the Theory
of Human Understanding’, Hume Studies
24(2) (1998), pp. 235–73.
417
9780826443595_Ref_Final_txt_print.indd 417
BIBLIOGRAPHY
Wilson, F., ‘Hume’s Sceptical Argument
against Reason’, Hume Studies 9(2)
(1983), pp. 90–129.
Wright, J.P., ‘Hume’s Academic Scepticism:
A Reappraisal of His Philosophy of
Human Understanding’, Canadian
Journal of Philosophy 16 (1986),
pp. 407–36.
21. THE MIND AND THE SELF
Ainslie, D.C., ‘Hume on Personal Identity’,
in E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 140–56.
Baxter, D.L.M., Hume’s Difficulty: Time
and Identity in the Treatise (London:
Routledge, 2008).
Biro, J., ‘Hume on Self-identity and
Memory’, The Review of Metaphysics 30
(1976), pp. 19–38.
–––––, ‘Hume’s New Science of the Mind’,
in D.F. Norton (ed.), The Cambridge
Companion to Hume (Cambridge:
Cambridge University Press, 1993),
pp. 33–63.
Bricke, J., Hume’s Philosophy of Mind
(Edinburgh: Edinburgh University Press,
1980).
–––––, ‘Hume on Self-identity, Memory, and
Causality’, in G.P. Morice (ed.), David
Hume: Bicentenary Papers (Edinburgh:
Edinburgh University Press, 1977),
pp. 167–74.
Ellis, J., ‘The Contents of Hume’s Appendix
and the Source of His Despair’, Hume
Studies 32 (2006), pp. 195–231.
Flage, D., David Hume’s Theory of Mind
(London: Routledge, 1990).
Fogelin, R.J., ‘The Soul and the Self’, in
M. Atherton (ed.), The Empiricists;
2/9/2012 9:38:24 PM
 BIBLIOGRAPHY
Critical Essays on Locke, Berkeley,
and Hume (Lanham, MD: Rowman &
Littlefield), pp. 213–28.
Henderson, R.S., ‘David Hume on Personal
Identity and the Indirect Passions’, Hume
Studies 16(1) (1990), pp. 33–44.
Martin, R., and Barresi, J., Naturalization of
the Soul: Self and Personal Identity in the
Eighteenth Century (London: Routledge,
2000).
Penelhum, T., ‘Hume on Personal Identity’,
in V.C. Chappell (ed.), Hume: A
Collection of Critical Essays (London:
Macmillan, 1968), pp. 213–39; repr. in
T. Penelhum, Themes in Hume: The Self,
the Will, Religion (Oxford: Clarendon
Press, 2000), pp. 23–39.
–––––, ‘Hume’s Theory of the Self Revisited’,
Dialogue 14 (1975), pp. 389–409; repr. in
T. Penelhum, Themes in Hume: The Self,
the Will, Religion (Oxford: Clarendon
Press, 2000), pp. 40–60.
–––––, ‘Self-identity and Self-regard’, in
his Themes in Hume: The Self, the Will,
Religion (Oxford: Clarendon Press,
2000), pp. 61–87.
–––––, ‘Hume, Identity and Selfhood’, in
his Themes in Hume: The Self, the Will,
Religion (Oxford: Clarendon Press,
2000), pp. 99–126.
Pike, N., ‘Hume’s Bundle Theory of the Self: A
Limited Defense’, American Philosophical
Quarterly 4 (1967), pp. 159–65.
Pitson, A.E., Hume’s Philosophy of the Self
(London: Routledge, 2002).
Stevenson, G., ‘Humean Self-consciousness
Explained’, Hume Studies 24(1) (1998),
pp. 95–130.
Strawson, G., ‘Hume on Himself’, in
D. Egonsson et al. (eds), Exploring
Practical Philosophy: From Action to
Values (Aldershot: Ashgate Press, 2001),
pp. 69–94.
418
9780826443595_Ref_Final_txt_print.indd 418
–––––, The Evident Connexion: Hume
on Personal Identity (Oxford: Oxford
University Press, 2011).
Swain, C.G., ‘Personal Identity and the
Skeptical System of Philosophy’, in
S. Traiger (ed.), The Blackwell Guide
to Hume’s Treatise (Oxford: Blackwell
Publishing, 2006), pp. 133–50.
Wilbanks, J., Hume’s Theory of the
Imagination (The Hague: Martinus
Nijhoff, 1968).
Wilson, F., Body, Mind and Self in Hume’s
Critical Realism (Frankfurt: Ontos, 2008).
–––––, ‘Hume’s Theory of Mental Activity’,
in D.F. Norton, N. Capaldi and W.L.
Robison (eds), McGill Hume Studies
(San Diego, CA: Austin Hill Press, 1979),
pp. 101–20.
Wolff, R.P., ‘Hume’s Theory of Mental
Activity’, Philosophical Review 69
(1960), pp. 289–310; repr. in V.C.
Chappell (ed.), Hume: A Collection of
Critical Essays (London: Macmillan,
1968), pp. 99–128.
Yandell, K., ‘Continuity, Consciousness, and
Identity’, Hume Studies 18(2) (1992),
pp. 255–74.
22. FREE WILL AND DETERMINISM
Botterill, G., ‘Hume on Liberty and
Necessity’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 277–300.
Bricke, J., ‘Hume, Freedom to Act,
and Personal Evaluation’, History
of Philosophy Quarterly 5 (1988),
pp. 141–56.
–––––, ‘Hume on Liberty and Necessity’,
in E.S. Radcliffe (ed.), A Companion to
2/9/2012 9:38:24 PM

Hume (Oxford: Wiley-Blackwell, 2008),
pp. 201–16.
Harris, J.A., ‘Hume’s Reconciling Project
and “the Common Distinction betwixt
Moral and Physical Necessity”’, British
Journal for the History of Philosophy 11
(2003), pp. 451–71.
–––––, Of Liberty and Necessity: The Free
Will Debate in Eighteenth-Century
British Philosophy (Oxford: Clarendon
Press, 2005).
Millican, P., ‘Hume’s Determinism’,
Canadian Journal of Philosophy 40(4)
(2010), pp. 611–42.
Penelhum, T., ‘Hume and the Freedom of the
Will’, in his Themes in Hume: The Self,
the Will, Religion (Oxford: Clarendon
Press, 2000), pp. 156–76.
Pitson, A.E., ‘Liberty, Necessity, and
the Will’, in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 216–32.
Steinberg, E., ‘Hume on Liberty, Necessity
and Verbal Disputes’, Hume Studies 13(2)
(1987), pp. 113–37.
Vesey, G., ‘Hume on Liberty and Necessity’,
in G. Vesey (ed.), Philosophers Ancient
and Modern (Cambridge: Cambridge
University Press, 1986), pp. 111–27.
23. TESTIMONY
Coady, C.A.J., Testimony: A Philosophical
Study (Oxford: Clarendon Press, 1992).
–––––, ‘Testimony and Observation’,
American Philosophical Quarterly 10
(1973), pp. 149–55.
Coleman, D., ‘Baconian Probability and
Hume’s Theory of Testimony’, Hume
Studies 27(2) (2001), pp. 195–226.
419
9780826443595_Ref_Final_txt_print.indd 419
BIBLIOGRAPHY
Faulkner, P., ‘David Hume’s Reductionist
Epistemology of Testimony’, Pacific
Philosophical Quarterly 79 (1998),
pp. 302–13.
–––––, ‘On the Rationality of Our Response
to Testimony’, Synthese 131 (2002),
pp. 353–70.
Gelfert, A., ‘Hume on Testimony Revisited’,
Logical Analysis and History of
Philosophy 13 (2010), pp. 60–75.
Hribek, T., ‘Against Coady on Hume on
Testimony’, Acta Analytica 11 (1996),
pp. 189–200.
O’Brien, D., ‘A Feminist Interpretation of
Hume on Testimony’, Hypatia 25 (2010),
pp. 632–52.
Root, M., ‘Hume on the Virtues of
Testimony’, American Philosophical
Quarterly 38 (2001), pp. 19–35.
Traiger, S., ‘Experience and Testimony
in Hume’s Philosophy’, Episteme 7(1)
(2010), pp. 42–57.
–––––, ‘Humean Testimony’, Pacific
Philosophical Quarterly 74 (1993),
pp. 135–49.
Welbourne, M., ‘Is Hume Really a
Reductivist?’, Studies in History and
Philosophy of Science Part A 33 (2002),
pp. 407–23.
Wilson, F., ‘Hume and the Role of
Testimony in Knowledge’, Episteme 7(1)
(2010), pp. 49–73.
24. THE PSYCHOLOGY OF THE
PASSIONS
Alanen, L., ‘Reflection and Ideas in
Hume’s Account of the Passions’, in
J. Jenkins, J. Whiting and C. Williams
(eds), Persons and Passions: Essays in
Honor of Annette Baier (Notre Dame,
2/9/2012 9:38:24 PM
 BIBLIOGRAPHY
IN: University of Notre Dame Press,
2005), pp. 117–42.
–––––, ‘The Powers and Mechanisms of
the Passions’, in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 179–198.
Baier, A., ‘Hume’s Analysis of Pride’,
Journal of Philosophy 75 (1978),
pp. 27–40.
Bricke, J., ‘Emotion and Thought in Hume’s
Treatise’, Canadian Journal of Philosophy,
supp. vol. 1 (1974), pp. 53–71.
Burch, R.W., ‘Hume on Pride and Humility’,
The New Scholasticism 49 (1975),
pp. 177–88.
Capaldi, N., ‘Hume’s Theory of the
Passions’, in D.W. Livingston and J.T.
King (eds), Hume: A Re-evaluation (New
York: Fordham University Press, 1976),
pp. 172–90.
Cléro, J.-P., La philosophie des passions chez
David Hume (Paris: Klincksieck, 1985).
Cohon, R., ‘Hume’s Indirect Passions,’ in
E.S. Radcliffe (ed.) A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 159–84.
Davidson, D., ‘Hume’s Cognitive Theory of
Pride’, in Essays on Actions and Events
(Oxford: Oxford University Press, 1980).
Fieser, J., ‘Hume’s Classification of the
Passions and Its Precursors’, Hume
Studies 18(1) (1992), pp. 1–17.
Gardiner, P.L., ‘Hume’s Theory of the
Passions’, in D. Pears (ed.), David Hume:
A Symposium (London: Macmillan,
1963), pp. 31–42.
Grüne-Yanoff, T., and McClennen, E.F.,
‘Hume’s Framework for a Natural
History of the Passions’, in C. Wennerlind
and M. Schabas (eds), David Hume’s
Political Economy (London: Routledge,
2007), pp. 86–104.
420
9780826443595_Ref_Final_txt_print.indd 420
McIntyre, J.L., ‘Hume’s “New and
Extraordinary” Account of the Passions’,
in S. Traiger (ed.), The Blackwell Guide
to Hume’s Treatise (Oxford: Blackwell
Publishing, 2006), pp. 199–215.
25. REASON AND THE PASSIONS
Árdal, P.S., Passion and Value in Hume’s
Treatise, 2nd edn (Edinburgh: Edinburgh
University Press, 1989).
Audi, R., ‘Prospects for a Naturalization
of Practical Reason: Humean
Instrumentalism and the Normative
Authority of Desire’, International
Journal of Philosophical Studies 10
(2001), pp. 235–63.
Brett, N., and Paxman, K., ‘Reason in
Hume’s Passions’, Hume Studies 34(1)
(2008), pp. 43–60.
Cohon, R., and Owen, D., ‘Hume on
Representation, Reason, and Motivation’,
Manuscrito 20 (1997), pp. 47–76.
Falk, W.D., ‘Hume on Practical Reason’,
Philosophical Studies 27 (1975),
pp. 1–18.
Karlsson, M.M., ‘Reason, Passion, and
the Influencing Motives of the Will’, in
S. Traiger (ed.), The Blackwell Guide
to Hume’s Treatise (Oxford: Blackwell
Publishing, 2006), pp. 235–255.
Korsgaard, C., ‘Skepticism about Practical
Reason’, Journal of Philosophy 83
(1986), pp. 5–25.
Kydd, R.M., Reason and Conduct in
Hume’s Treatise (Oxford: Oxford
University Press, 1946).
Magri, T., ‘Hume on the Direct Passions
and Motivation’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 185–200.
2/9/2012 9:38:24 PM

Mason, M., ‘Hume and Humeans on
Practical Reason’, Hume Studies 31(2)
(2005), pp. 347–78.
Phillips, D., ‘Hume on Practical Reason:
Normativity and Psychology in Treatise
2.3.3’, Hume Studies 31(2) (2005),
pp. 299–316.
Radcliffe, E.S., ‘Hume on the Generation
of Motives: Why Beliefs Alone Never
Motivate’, Hume Studies 25(1&2)
(1999), pp. 101–22.
–––––, ‘The Humean Theory of Motivation
and Its Critics’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 477–92.
Sandis, C., ‘Hume and the Debate on
“Motivating Reasons”’, in C. Pigden
(ed.), Hume on Motivation and Virtue:
New Essays (Basingstoke: Palgrave
Macmillan, 2009), pp. 142–54.
Smith, M., ‘The Humean Theory of
Motivation’, Mind 96 (1987), pp. 36–61.
Stalley, R.F., ‘The Will in Hume’s Treatise’,
Journal of the History of Philosophy 24
(1986), pp. 41–53.
Tweyman, S., Reason and Conduct in
Hume and His Predecessors (The Hague:
Martinus Nijhoff, 1974).
Weller, C., ‘Scratched Fingers, Ruined Lives,
and Acknowledged Lesser Goods’, Hume
Studies 30(1) (2004), pp. 51–85.
–––––, ‘The Myth of Original Existence’,
Hume Studies 28(2) (2002),
pp. 195–230.
26. THE INTERPRETATION OF THE
DIALOGUES CONCERNING NATURAL
RELIGION
Austin, W., ‘Philo’s Reversal’, Philosophical
Topics 13 (1985), pp. 103–12.
421
9780826443595_Ref_Final_txt_print.indd 421
BIBLIOGRAPHY
Battersby, C., ‘The Dialogues as Original
Imitation: Cicero and the Nature of
Hume’s Scepticism’, in D.F. Norton, N.
Capaldi and W.L. Robison (eds), McGill
Hume Studies (San Diego, CA: Austin
Hill Press, 1979), pp. 239–52.
Bell, M., ‘The Relation between Literary Form
and Philosophical Argument in Hume’s
Dialogues concerning Natural Religion’,
Hume Studies 27 (2001), pp. 227–46.
Berman, D., A History of Atheism in
Britain: From Hobbes to Russell
(London: Routledge, 1988).
Bricke, J., ‘On the Interpretation of Hume’s
Dialogues’, Religious Studies 11 (1975),
pp. 1–18.
Carnochan, W.B., ‘The Comic Plot of
Hume’s Dialogues’, Modern Philology 85
(1988), pp. 514–22.
Dancy, J., ‘“For here the author is
annihilated”: Reflections on Philosophical
Aspects of the Use of the Dialogue
Form in Hume’s Dialogues concerning
Natural Religion’, in T.J. Smiley (ed.),
Philosophical Dialogues: Plato, Hume,
Wittgenstein (Oxford: Clarendon Press,
1995), pp. 29–60.
Davis, S., ‘Irony and Argument in Dialogues,
XII’, Religious Studies 27 (1991),
pp. 239–57.
Dendle, P., ‘Reconciling Philo and Hume:
Habits, Caprice, and Inclinations’, Texas
Studies in Literature and Language 40
(1998), pp. 26–47.
Foley, R., ‘Unnatural Religion:
Indoctrination and Philo’s Reversal in
Hume’s Dialogues concerning Natural
Religion’, Hume Studies 32(1) (2006),
pp. 83–112.
Hurlbutt, R.H., ‘The Careless Skeptic –
the “Pamphilian” Ironies in Hume’s
Dialogues’, Hume Studies 14(2) (1988),
pp. 207–50.
2/9/2012 9:38:24 PM
 BIBLIOGRAPHY
Malherbe, M., ‘Hume and the Art of
Dialogue’, in M.A. Stewart and J.P.
Wright (eds), Hume and Hume’s
Connexions (Edinburgh: Edinburgh
University, 1994), pp. 201–23.
Morrisroe, M., ‘Characterization as Rhetorical
Device in Hume’s Dialogues concerning
Natural Religion’, Enlightenment Essays 1
(1970), pp. 95–107.
–––––, ‘Hume’s Rhetorical Strategy: A
Solution to the Riddle of the Dialogues
concerning Natural Religion’, Texas
Studies in Literature and Language 11
(1969), pp. 963–74.
Mossner, E.C., ‘Hume and the Legacy of the
Dialogues’, in G.P. Morice (ed.), David
Hume: Bicentenary Papers (Edinburgh:
Edinburgh University Press, 1977),
pp. 1–22.
Pakaluk, M., ‘Philosophical “Types”
in Hume’s Dialogues’, in V. Hope
(ed.), Philosophers of the Scottish
Enlightenment (Edinburgh: Edinburgh
University Press, 1984), pp. 116–32.
Penelhum, T., ‘Hume’s Scepticism and the
Dialogues’, in D.F. Norton, N. Capaldi
and W.L. Robison (eds), McGill Hume
Studies (San Diego, CA: Austin Hill Press,
1979), pp. 253–78; repr. in T. Penelhum,
Themes in Hume: The Self, the Will,
Religion (Oxford: Clarendon Press,
2000), pp. 177–203.
Prince, M.B., ‘Hume and the End of
Religious Dialogue’, Eighteenth-Century
Studies 25 (1992), pp. 283–308.
Shapiro, G., ‘The Man of Letters and
the Author of Nature: Hume on
Philosophical Discourse’, The Eighteenth
Century 26 (1985), pp. 115–37.
Simpson, D., ‘Hume’s Intimate Voices and
the Method of Dialogue’, Texas Studies
in Literature and Language 21 (1979),
pp. 68–92.
422
9780826443595_Ref_Final_txt_print.indd 422
Wadia, P., ‘Philo Confounded’, in D.F.
Norton, N. Capaldi and W.L. Robison
(eds), McGill Hume Studies (San Diego,
CA: Austin Hill Press, 1979), pp. 279–90.
White, R., ‘Hume’s Dialogues and the
Comedy of Religion’, Hume Studies
14(2), (1988), pp. 390–407.
Wieand, J., ‘Pamphilus in Hume’s Dialogues’,
The Journal of Religion 65 (1985),
pp. 33–45.
27. RELIGION
Andre, S., ‘Was Hume an Atheist?’, Hume
Studies 19(1) (1993), pp. 141–66.
Badia Cabrera, M.A., Hume’s Reflection on
Religion (Dordrecht: Kluwer Academic,
2001).
Baier, A.C., ‘Hume on Religion: Stopping
the Ocean with a Bull-rush?’, in J.J.
MacIntosh and H.A. Meynell (eds),
Faith, Scepticism and Personal Identity
(Calgary: University of Calgary Press,
1994), pp. 61–82.
Barker, S., ‘Hume on the Logic of Design’,
Hume Studies 9(1) (1983), pp. 1–18.
–––––, ‘Reasoning by Analogy in Hume’s
Dialogues’, Informal Logic 11(3) (1989),
pp. 173–84.
Bayne, S.M., ‘Hume on Miracles: Would It
Take a Miracle to Believe in a Miracle?’,
Southern Journal of Philosophy 45
(2007), pp. 1–29.
Beckwith, F.J., David Hume’s Argument
against Miracles: A Critical Analysis
(Lanham, MD: University Press of
America, 1989).
Bell, M., ‘Hume on Superstition’, in D.Z.
Phillips and T. Tessin (eds), Religion and
Hume’s Legacy (New York: St. Martin’s
Press, 1999), pp. 153–70.
2/9/2012 9:38:24 PM

–––––, ‘Hume on the Nature and Existence
of God’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 338–52.
Bernard, C., ‘Hume and the Madness of
Religion’, in M.A. Stewart and J.P. Wright
(eds), Hume and Hume’s Connexions
(Edinburgh: Edinburgh University Press,
1994), pp. 224–38.
Buchegger, J., David Humes Argumente
gegen das Christentum (Frankfurt am
Main: Peter Lang, 1987).
Buckle, S., ‘Marvels, Miracles, and
Mundane Order’, Australasian Journal of
Philosophy 79 (2001), pp. 1–31.
Burch, R., ‘Bayesianism and Analogy in
Hume’s Dialogues’, Hume Studies 6(1)
(1980), pp. 32–44.
Burns, R.M., The Great Debate on Miracles:
From Joseph Glanvill to David Hume
(Lewisburgh, PA: Bucknell University
Press, 1981).
Butler, R.J., ‘Natural Belief and the Enigma
of Hume’, Archiv für die Geschichte der
Philosophie 42 (1960), pp. 73–100.
Campbell, J.K., ‘Hume’s Refutation of the
Cosmological Argument’, International
Journal for Philosophy of Religion 40
(1996), pp. 159–73.
Carabelli, G., Hume e la retorica
dell’ideologia: Uno studio dei Dialoghi
sulla Religione Naturale (Firenze: La
Nuova Italia Editrice, 1972).
Coleman, D., ‘Hume, Miracles, and Lotteries’,
Hume Studies 14(2) (1988), pp. 328–46.
Craig, E.J., Hume on Religion (Shimla:
Indian Institute of Advanced Study, 1997).
Dees, R.H., ‘“The Paradoxical Principle and
Salutary Practice”: Hume on Toleration’,
Hume Studies 31(1) (2005), pp. 145–64.
Earman, J., Hume’s Abject Failure: The
Argument Against Miracles (Oxford:
Oxford University Press, 2000).
423
9780826443595_Ref_Final_txt_print.indd 423
BIBLIOGRAPHY
Echelbarger, C., ‘Hume’s Tacit Atheism’,
Religious Studies 11 (1975), pp. 19–35.
Falkenstein, L., ‘Hume’s Project in The
Natural History of Religion’, Religious
Studies 39 (2003), pp. 1–21.
Ferguson, S., ‘Bayesianism, Analogy, and
Hume’s Dialogues concerning Natural
Religion’, Hume Studies 28(1) (2002),
pp. 113–30.
Fieser, J., ‘Hume’s Concealed Attack on
Religion and His Early Critics’, Journal
of Philosophical Research 20 (1995),
pp. 83–101.
Fogelin, R.J., ‘What Hume Actually Said
about Miracles’, Hume Studies 16 (1990),
pp. 81–7.
–––––, A Defense of Hume on Miracles
(Princeton, NJ: Princeton University
Press, 2003).
Foster, S.P., Melancholy Duty: The
Hume-Gibbon Attack on Christianity
(Dordrecht: Kluwer Academic, 1997).
Garrett, D., ‘Hume on Testimony concerning
Miracles’, in P. Millican (ed.), Reading
Hume on Human Understanding
(Oxford: Clarendon Press, 2002),
pp. 301–34.
Gaskin, J.C.A., ‘Hume, Atheism, and the
“Interested Objection” of Morality’,
in D.F. Norton, N. Capaldi and W.L.
Robison (eds), McGill Hume Studies
(San Diego, CA: Austin Hill Press, 1979),
pp. 147–59.
–––––, ‘Hume on Religion’, in D.F. Norton
(ed.), The Cambridge Companion to
Hume (Cambridge: Cambridge University
Press, 1993), pp. 313–44.
–––––, Hume’s Philosophy of Religion, 2nd
edn (London: Macmillan, 1988).
–––––, ‘Religion: The Useless Hypothesis’,
in P. Millican (ed.), Reading Hume
on Human Understanding (Oxford:
Clarendon Press, 2002), pp. 349–69.
2/9/2012 9:38:25 PM
 BIBLIOGRAPHY
Groulez, M., Le scepticisme de Hume: les
dialogues sur la religion naturelle (Paris:
Presses Universitaires de France, 2005).
Hanson, D.J., Fideism and Hume’s
Philosophy: Knowledge, Religion, and
Metaphysics (New York: Peter Lang, 1993).
Hester, M.B. (ed.), Hume’s Philosophy of
Religion (Winston-Salem, NC: Wake
Forest University Press, 1986).
Holden, T., ‘Religion and Moral Prohibition
in Hume’s “Of Suicide”’, Hume Studies
31(2) (2005), pp. 189–210.
–––––, Spectres of False Divinity: Hume’s
Moral Atheism (Oxford: Oxford
University Press, 2010).
Houston, J., Reported Miracles: A Critique
of Hume (Cambridge: Cambridge
University Press, 1994).
Hurlbutt, R.H., Hume, Newton, and
the Design Argument (Lincoln, NE:
University of Nebraska Press, 1965).
Jeffner, A., Butler and Hume on Religion:
A Comparative Analysis (Stockholm:
Diakonistyrelsens Bokförlag, 1966).
Johnson, D., Hume, Holism, and Miracles
(Ithaca, NY: Cornell University Press, 1999).
Jones, P., ‘Hume’s Two Concepts of God’,
Philosophy 47 (1972), pp. 322–33.
Jordan, W.R., ‘Religion in the Public Square:
A Reconsideration of David Hume and
Religious Establishment’, The Review of
Politics 64 (2002), pp. 687–713.
Kail, P.J.E., ‘Understanding Hume’s Natural
History of Religion’, The Philosophical
Quarterly 57 (2007), pp. 190–211.
Kemp Smith, N., ‘Introduction’, in N. Kemp
Smith (ed.), Hume’s Dialogues concerning
Natural Religion, 2nd edn (Edinburgh:
Thomas Nelson, 1947), pp. 1–123.
Khamara, E.J., ‘Hume versus Clarke on
the Cosmological Argument’, The
Philosophical Quarterly 42 (1992),
pp. 34–55.
424
9780826443595_Ref_Final_txt_print.indd 424
Kraay K.J., ‘Philo’s Argument for Divine
Amorality Reconsidered’, Hume Studies
29(2) (2003), pp. 283–304.
Leroy, A.L., La critique et la religion chez
David Hume (Paris: Librairie Félix Alcan,
1930).
Levine, M.P., Hume and the Problem of
Miracles: A Solution (Dordrecht: Kluwer
Academic, 1989).
–––––, ‘Hume on Miracles and Immortality’,
in E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 353–70.
Logan, B., ‘“Aiding the Ascent of Reason by
the Wings of Imagination”: The Prospect
of a Future State’, Hume Studies 25(1&2)
(1999), pp. 193–205.
–––––, ‘The Irregular Argument in Hume’s
Dialogues’, Hume Studies 18(2) (1992),
pp. 483–500.
–––––, A Religion without Talking:
Religious Belief and Natural Belief in
Hume’s Philosophy of Religion (New
York: P. Lang, 1993).
Millican, P., ‘“Hume’s Theorem” concerning
Miracles’, The Philosophical Quarterly
43 (1993), pp. 489–95.
Mitchell, T.A., David Hume’s Anti-
Theistic Views: A Critical Appraisal
(Lanham, MD: University Press of
America, 1986).
Mossner, E.C., ‘The Religion of David
Hume’, Journal of the History of Ideas 39
(1978), pp. 653–63.
Nathan, G.J., ‘Hume’s Immanent God’, in
V.C. Chappell (ed.), Hume: A Collection
of Critical Essays (London: Macmillan,
1968), pp. 396–423.
Norton, D.F., ‘Hume, Atheism, and the
Autonomy of Morals’, in M.B. Hester
(ed.), Hume’s Philosophy of Religion
(Winston-Salem, NC: Wake Forest
University Press, 1985), pp. 97–144.
2/9/2012 9:38:25 PM

Noxon, J., ‘Hume’s Agnosticism’, The
Philosophical Review 73 (1964),
pp. 248–61.
–––––, ‘Hume’s Concern with Religion’,
in K.R. Merrill and R.W. Shahan (eds),
David Hume: Many-sided Genius
(Norman, OK: University of Oklahoma
Press, 1976), pp. 58–92.
O’Connor, D., Hume on Religion (New
York: Routledge, 2001).
–––––, ‘Scepticism and Philo’s Atheistic
Preference’, Hume Studies 29(2) (2003),
pp. 267–82.
Owen, D., ‘Hume versus Price on Miracles
and Prior Probabilities: Testimony
and the Bayesian Calculation’, The
Philosophical Quarterly 37 (1987),
pp. 187–202.
Penelhum, T., God and Skepticism: A Study
in Skepticism and Fideism (Dordrecht:
Reidel, 1983).
–––––, ‘Natural Belief and Religious Belief
in Hume’s Philosophy’, The Philosophical
Quarterly 33 (1983), pp. 166–81; repr. in
T. Penelhum, Themes in Hume: The Self,
the Will, Religion (Oxford: Clarendon
Press, 2000), pp. 204–21.
–––––, ‘Religion in the Enquiry and After’,
in his Themes in Hume: The Self, the
Will, Religion (Oxford: Clarendon Press,
2000), pp. 222–43.
Phillips, D.Z., and Tessin, T. (eds), Religion
and Hume’s Legacy (New York: St.
Martin’s Press, 1999).
Pitson, A.E., ‘The Miseries of Life: Hume
and the Problem of Evil’, Hume Studies
34(1) (2008), pp. 89–114.
Pyle, A., Hume’s Dialogues concerning
Natural Religion: Reader’s Guide
(London: Continuum, 2006).
Reich, L., Hume’s Religious Naturalism
(Lanham, MD: University Press of
America, 1998).
425
9780826443595_Ref_Final_txt_print.indd 425
BIBLIOGRAPHY
Ross, I.S., ‘Hume’s Language of
Skepticism’, Hume Studies 21(2) (1995),
pp. 237–54.
Russell, P., ‘Skepticism and Natural
Religion in Hume’s Treatise’, Journal
of the History of Ideas 49 (1988),
pp. 247–65.
Salmon, W.S., ‘Religion and Science:
A New Look at Hume’s Dialogues’,
Philosophical Studies 33 (1978),
pp. 143–76.
Sessions, W.L., Reading Hume’s Dialogues:
A Veneration for True Religion
(Bloomington, IN: Indiana University
Press, 2002).
Siebert, D.T., ‘Hume on Idolatry and
Incarnation’, Journal of the History of
Ideas 45 (1984), pp. 379–96.
Sobel, J.H., ‘Hume’s Theorem on Testimony
Sufficient to Establish a Miracle’, The
Philosophical Quarterly 41 (1991),
pp. 229–37.
Soles, D.H., ‘Hume, Language and God’,
Philosophical Topics 12 (1981),
pp. 109–19.
Solon, T.P.M., and Wertz, S.K., ‘Hume’s
Argument from Evil’, The Personalist 50
(1969), pp. 383–92.
Stewart, M.A., ‘Hume’s Historical View of
Miracles’, in M.A. Stewart and J.P. Wright
(eds), Hume and Hume’s Connexions
(Edinburgh: Edinburgh University Press,
1994), pp. 171–200.
Streminger, G., ‘Religion a Threat to
Morality: An Attempt to Throw Some
New Light on Hume’s Philosophy of
Religion’, Hume Studies 15(2) (1989),
pp. 277–93.
Swinburne, R., ‘The Argument from Design’,
Philosophy 43 (1968), pp. 199–212.
Tweyman, S., ‘An Inconvenience of
Anthropomorphism’, Hume Studies 8(1)
(1982), pp. 19–42.
2/9/2012 9:38:25 PM
 BIBLIOGRAPHY
–––––, ‘Hume’s Dialogues on Evil’, Hume
Studies 13(1) (1987), pp. 74–85.
–––––, Scepticism and Belief in Hume’s
Dialogues concerning Natural
Religion (Dordrecht: Martinus Nijhoff,
1986).
Vanderburgh, W.L., ‘Of Miracles and
Evidential Probability: Hume’s “Abject
Failure” Vindicated’, Hume Studies 31(1)
(2005), pp. 37–62.
Webb, M., ‘The Argument of the Natural
History’, Hume Studies 17(2) (1991),
pp. 141–59.
Wengenroth, A., Science of Man:
Religionsphilosophie und
Religionskritik bei David Hume und
seinen Vorgängern (Frankfurt am Main:
P. Lang, 1997).
Williams, B.A.O., ‘Hume on Religion’,
in D. Pears (ed.), David Hume: A
Symposium (London: Macmillan, 1963),
pp. 77–88.
Wilson, F., ‘The Logic of Probabilities in
Hume’s Argument against Miracles’,
Hume Studies 15(2) (1989), pp. 255–75.
Wootton, D., ‘Hume’s “Of Miracles”:
Probability and Irreligion’, in M.A.
Stewart (ed.), Studies in the Philosophy
of the Scottish Enlightenment (Oxford:
Clarendon Press, 1990), pp. 191–230.
Yandell, K., Hume’s ‘Inexplicable Mystery’:
His Views on Religion (Philadelphia, PA:
Temple University Press, 1990).
Yenor, S., ‘Revealed Religion and the Politics
of Humanity in Hume’s Philosophy
of Common Life’, Polity 38 (2006),
pp. 395–415.
Yoder, T.S., Hume on God: Irony,
Deism and Genuine Theism (London:
Continuum, 2008).
Young, R, ‘Miracles and Epistemology’,
Religious Studies 8 (1972), pp. 115–26.
426
9780826443595_Ref_Final_txt_print.indd 426
28. MORALITY
Abramson, K., ‘Sympathy and Hume’s
Spectator-centred Theory of Virtue’, in
E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 240–56.
Árdal, P.S., ‘Another Look at Hume’s
Account of Moral Evaluation’, Journal
of the History of Philosophy 15 (1977),
pp. 405–21.
Ashford, E., ‘Utilitarianism with a Humean
Face’, Hume Studies 31(1) (2005),
pp. 63–92.
Baier, A., ‘Artificial Virtues and the Equally
Sensible Non-knaves: A Response to
Gauthier’, Hume Studies 18 (1992),
pp. 429–39.
Baillie, J., Hume on Morality (London:
Routledge, 2000).
Botros, S., Hume, Reason and Morality:
A Legacy of Contradiction (London:
Routledge, 2006).
Botwinick, A., ‘A Case for Hume’s
Nonutilitarianism’, Journal of the History
of Philosophy 15 (1977), pp. 423–35.
Brand, W., Hume’s Theory of Moral
Judgement: A Study in the Unity of A
Treatise of Human Nature (Dordrecht:
Kluwer Academic Publishers, 1992).
Bricke, J., ‘Hume, Motivation, and Morality’,
Hume Studies 14 (1988), pp. 1–24.
–––––, Mind and Morality: An Examination
of Hume’s Moral Psychology (Oxford:
Oxford University Press, 1996).
Brown, C.R., ‘Hume on Moral Rationalism,
Sentimentalism, and Sympathy’, in
E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 219–39.
Capaldi, N., Hume’s Place in Moral
Philosophy (New York: Peter Lang, 1989).
2/9/2012 9:38:25 PM

Cohon, R., Hume’s Morality: Feeling and
Fabrication (Oxford: Oxford University
Press, 2008).
–––––, ‘Hume’s Artificial and Natural
Virtues’, in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 256–75.
Costelloe, T.M., Aesthetics and Morals in
the Philosophy of David Hume (New
York: Routledge, 2007).
Driver, J., ‘Pleasure as the Standard of
Virtue in Hume’s Moral Philosophy’,
Philosophical Quarterly 85 (2004),
pp. 173–94.
Foot, P., ‘Hume on Moral Judgement’,
in D. Pears (ed.), David Hume: A
Symposium (London: Macmillan, 1963).
Gauthier, D., ‘Artificial Virtues and the
Sensible Knave’, Hume Studies 18(2)
(1992), pp. 401–27.
–––––, ‘David Hume, Contractarian’,
Philosophical Review, 88 (1979),
pp. 3–38.
–––––, ‘Hume and Morality’s “Useful
Purpose”’, in J. Jenkins, J. Whiting and
C. Williams (eds), Persons and Passions:
Essays in Honor of Annette Baier (Notre
Dame, IN: University of Notre Dame
Press, 2005), pp. 211–28.
–––––, ‘Three against Justice: The Foole,
the Sensible Knave, and the Lydian
Shepherd’, Midwest Studies in Philosophy
7 (1982), pp. 11–29.
Glathe, A.B., Hume’s Theory of the Passions
and of Morals: A Study of Books II and
III of the Treatise (Berkeley: University of
California Press, 1950).
Gräfrath, B., Moral Sense und Praktische
Vernunft: David Humes Ethik und
Rechtsphilosophie (Stuttgart: J.B. Metzler,
1991).
427
9780826443595_Ref_Final_txt_print.indd 427
BIBLIOGRAPHY
Greco, L., L’io morale: David Hume e
l’Etica Contemporanea (Naples: Liguori
Editore, 2008).
Haakonssen, K., The Science of a Legislator:
The Natural Jurisprudence of David
Hume and Adam Smith (Cambridge:
Cambridge University Press, 1981).
Harrison, J., Hume’s Moral Epistemology
(Oxford, Clarendon Press, 1976).
–––––, Hume’s Theory of Justice (Oxford,
Clarendon Press, 1981).
Hearn, T., ‘General Rules and the Moral
Sentiments in Hume’s Treatise’, Review of
Metaphysics 30 (1976), pp. 57–72.
King, J.T., ‘Hume’s Classical Theory of
Justice’, Hume Studies 7(1) (1981),
pp. 32–54.
–––––, ‘The Place of the Language of Morals
in Hume’s Second Enquiry’, in D.W.
Livingston and J.T. King (eds), Hume:
A Re-evaluation (New York: Fordham
University Press, 1976), pp. 343–61.
Korsgaard, C., ‘The General Point of View:
Love and Moral Approval in Hume’s
Ethics’, Hume Studies 25(1&2) (1999),
pp. 3–42.
Le Jallé, E., Hume et la régulation morale
(Paris: Presses Universitaires de France,
1999).
Lecaldano, E., Hume e la nascita dell’etica
contemporanea (Rome: Laterza, 1991).
MacIntyre, A.C., ‘Hume on “Is” and
“Ought”’, in V.C. Chappell (ed.), Hume:
A Collection of Critical Essays (London:
Macmillan, 1968), pp. 240–64.
Mackie, J.L., Hume’s Moral Theory (London:
Routledge & Kegan Paul, 1980).
Martin, M.A., ‘Hume as Classical Moralist’,
International Philosophical Quarterly 34
(1994), pp. 323–34.
Mercer, P., Sympathy and Ethics: A Study
of the Relationship between Sympathy
2/9/2012 9:38:25 PM
 BIBLIOGRAPHY
and Morality with Special Reference to
Hume’s Treatise (Oxford: Clarendon
Press, 1972).
Norton, D.F., ‘Hume, Human Nature, and
the Foundations of Morality’, in D.F.
Norton (ed.), The Cambridge Companion
to Hume (Cambridge: Cambridge
University Press, 1993), pp. 148–81.
–––––, ‘Hume’s Common Sense Morality’,
Canadian Journal of Philosophy 5 (1975),
pp. 523–43.
Penelhum, T., ‘Hume’s Moral Psychology’,
in D.F. Norton (ed.), The Cambridge
Companion to Hume (Cambridge:
Cambridge University Press, 1993),
pp. 117–47.
Pigden, C.R., ‘If not Non-cognitivism,
then What?’, in C. Pigden (ed.), Hume
on Motivation and Virtue: New Essays
(Basingstoke: Palgrave Macmillan 2009),
pp. 80–104.
––––– (ed.), Hume on Motivation and
Virtue: New Essays (Basingstoke:
Palgrave Macmillan, 2009).
––––– (ed.), Hume on Is and Ought
(Basingstoke: Palgrave Macmillan,
2010).
Pitson, A.E., ‘Projectionism, Realism and
Hume’s Moral Sense Theory’, Hume
Studies 15(1) (1989), pp. 69–92.
Platts, M., ‘Hume and Morality as a
Matter of Fact’, Mind 97 (1988),
pp. 189–204.
Postema, G.J., ‘“Cemented with Diseased
Qualities”: Sympathy and Comparison
in Hume’s Moral Psychology’, Hume
Studies 31(2) (2005), pp. 249–98.
–––––, ‘Hume’s Reply to the Sensible
Knave’, History of Philosophy Quarterly
5 (1988), pp. 23–40.
Potkay, A., The Passion for Happiness:
Samuel Johnson and David Hume (Ithaca,
NY: Cornell University Press, 2000).
428
9780826443595_Ref_Final_txt_print.indd 428
Radcliffe, E.S, ‘Moral Internalism
and Moral Cognitivism in Hume’s
Metaethics’, Synthese 152 (2006),
pp. 353–70.
Raphael, D.D., ‘Hume’s Critique of Ethical
Rationalism’, in W.B. Todd (ed.), Hume
and the Enlightenment: Essays Presented
to Ernest Campell Mossner (Edinburgh:
Edinburgh University Press, 1974),
pp. 14–29.
Russell, P., Freedom and Moral Sentiment:
Hume’s Way of Naturalizing
Responsibility (Oxford: Clarendon Press,
1995).
–––––, ‘Moral Sense and Virtue in Hume’s
Ethics’, in T. Chappell (ed.), Values and
Virtues: Aristotelianism in Contemporary
Ethics (Oxford: Clarendon Press, 2006),
pp. 158–70.
Sayre-McCord, G., ‘On Why Hume’s
“General Point of View” Isn’t Ideal – and
Shouldn’t Be’, Social Philosophy and
Policy 11 (1994), pp. 202–28.
Siebert, D.T., The Moral Animus of David
Hume (Newark, DE: The University of
Delaware Press 1990).
Snare, F., Morals, Motivation, and
Convention: Hume’s Influential Doctrines
(Cambridge: Cambridge University Press,
1991).
Sobel, J.H., Walls and Vaults: A Natural
Science of Morals (Virtue Ethics
According to David Hume) (Hoboken,
NJ: John Wiley, 2009).
Stewart, J.B., The Moral and Political
Philosophy of David Hume (New York:
Columbia University Press, 1963).
Sturgeon, N.L., ‘Moral Scepticism and
Moral Naturalism in Hume’s Treatise’,
Hume Studies 27(1) (2001), pp. 3–83.
Swanton, C., ‘Can Hume Be Read as a
Virtue Ethicist?’, Hume Studies 33(1)
(2007), pp. 91–113.
2/9/2012 9:38:26 PM

Taylor, J., ‘Hume and the Reality of
Value’, in A.J. Jacobson (ed.), Feminist
Interpretations of David Hume
(University Park, PA: Pennsylvania State
University Press, 2000), pp. 107–36.
–––––, ‘Virtue and the Evaluation of
Character’, in S. Traiger (ed.), The
Blackwell Guide to Hume’s Treatise
(Oxford: Blackwell Publishing, 2006),
pp. 276–96.
Turco, L., Lo scetticismo morale di David
Hume (Bologna: CLUEB, 1984).
Vitz, R., ‘Hume and the Limits of Benevolence’,
Hume Studies 28(2) (2002), pp. 271–96.
Wallace, K., ‘Hume on Regulating Belief and
Moral Sentiment’, Hume Studies 28(1)
(2002), pp. 83–111.
Watkins, M., ‘Humean Moral Knowledge’,
Inquiry 51 (2008), pp. 581–602.
Welchman, J., ‘Hume and the Prince of
Thieves’, Hume Studies 34(1) (2008),
pp. 3–20.
29. POLITICAL SCIENCE
Allen, W.R., ‘Scarcity and Order: The
Hobbesian Problem and the Humean
Resolution’, Social Science Quarterly 57
(1976), pp. 263–75.
Berry, C.J., David Hume (London:
Continuum, 2009).
–––––, ‘Science and Superstition: Hume
and Conservatism’, European Journal of
Political Theory 10 (2011), pp. 141–55.
Brownsey, P.F., ‘Hume and the Social
Contract’, The Philosophical Quarterly
28 (1978), pp. 132–48.
Cohon, R., ‘The Shackles of Virtue: Hume
on Allegiance to Government’, History
of Philosophy Quarterly 18 (2001),
pp. 393–413.
429
9780826443595_Ref_Final_txt_print.indd 429
BIBLIOGRAPHY
Conniff, J., ‘Hume’s Political Methodology:
A Reconsideration of “That Politics May
be Reduced to a Science”’, Review of
Politics 38 (1976), pp. 88–108.
Day, J., ‘Hume on Justice and Allegiance’,
Philosophy 40 (1965), pp. 35–56.
Dees, R.H., ‘“One of the Finest and
Most Subtile Inventions”: Hume on
Government’, in E.S. Radcliffe (ed.), A
Companion to Hume (Oxford: Wiley-
Blackwell, 2008), pp. 388–405.
Forbes, D., Hume’s Philosophical Politics
(Cambridge: Cambridge University Press,
1975).
–––––, ‘Hume’s Science of Politics’, in G.P.
Morice (ed.), David Hume: Bicentenary
Papers (Edinburgh: Edinburgh University
Press, 1977), pp. 39–50.
Giarrizzo, G., David Hume politico e storico
(Turin: Giulio Einaudi, 1962).
Haakonssen, K., ‘Hume’s Political Theory’,
in D.F. Norton (ed.), The Cambridge
Companion to Hume (Cambridge:
Cambridge University Press, 1993),
pp. 182–221.
Hardin, R., David Hume: Moral and
Political Theorist (Princeton: Princeton
University Press, 2007).
Hayek, F.A., ‘The Legal and Political
Philosophy of David Hume’, in V.C.
Chappell (ed.), Hume: A Collection of
Critical Essays (London: Macmillan,
1968), pp. 335–60.
Kolin, A., The Ethical Foundations of
Hume’s Theory of Politics (New York:
Peter Lang, 1991).
Linares, F., Das Politische Denken von
David Hume (Hildesheim: Georg Olms,
1984).
McArthur, N., David Hume’s Political
Theory: Law, Commerce and the
Constitution of Government (Toronto:
University of Toronto Press, 2007).
2/9/2012 9:38:26 PM
 BIBLIOGRAPHY
–––––, ‘Laws not Men: Hume’s Distinction
between Barbarous and Civilized
Government’, Hume Studies 31(1)
(2005), pp. 123–44.
Marshall, G., ‘David Hume and Political
Scepticism’, The Philosophical Quarterly
4 (1954), pp. 247–57.
Miller, D., ‘Hume and Possessive
Individualism’, History of Political
Thought 1 (1980), pp. 261–78.
–––––, Philosophy and Ideology in Hume’s
Political Thought (Oxford: Clarendon
Press, 1981).
Pocock, J.G.A., ‘Hume and the American
Revolution: Dying Thoughts of a North
Briton’, in his Virtue, Commerce, and
History: Essays on Political Thought
and History, Chiefly in the Eighteenth
Century (Cambridge: Cambridge
University Press, 1985), pp. 125–41.
Sabl, A., ‘When Bad Things Happen from
Good People (and Vice-Versa): Hume’s
Political Ethics of Revolution’, Polity 35
(2002), pp. 73–92.
Stewart, J.B., Opinion and Reform in
Hume’s Political Philosophy (Princeton,
NJ: Princeton University Press,1992).
Whelan, F.G., Hume and Machiavelli:
Political Realism and Liberal Thought
(Lanham, MD: Lexington Books,
2004).
–––––, Order and Artifice in Hume’s
Political Philosophy (Princeton, NJ:
Princeton University Press, 1985).
Wulf, S.J., ‘The Skeptical Life in Hume’s
Political Thought’, Polity 33 (2000),
pp. 77–99.
Wolin, S., ‘Hume and Conservatism’, in
D.W. Livingston and J.T. King (eds),
Hume: A Re-evaluation (New York:
Fordham University Press, 1976),
pp. 239–56.
430
9780826443595_Ref_Final_txt_print.indd 430
30. SOCIETY AND GENDER
Baier, A., ‘Good Men’s Women: Hume on
Chastity and Trust’, Hume Studies 5(1)
(1979), pp. 1–19.
–––––, ‘Hume on Women’s Complexion’,
in P. Jones (ed.), The ‘Science of Man’ in
the Scottish Enlightenment – Hume, Reid
and Their Contemporaries (Edinburgh:
Edinburgh University Press, 1989),
pp. 33–53.
–––––, ‘Hume: The Reflective Women’s
Epistemologist’, in A.J. Jacobson (ed.),
Feminist Interpretations of David Hume
(University Park, PA: Pennsylvania State
University Press, 2000), pp. 19–38.
Battersby, C., ‘An Enquiry concerning the
Humean Woman’, Philosophy 56 (1981),
pp. 303–12.
Burns, S., ‘The Humean Female’, Dialogue
15 (1976), pp. 415–24.
Gardner, C.V., ‘Chastity and the Practice
of the World in Hume’s Treatise’, Hume
Studies 32(2) (2006), pp. 331–45.
Garrett. D., ‘Hume as Man of Reason and
Woman’s Philosopher’, in L. Alanen and
C. Witt (eds), Feminist Reflections on
the History of Philosophy (Dordrecht:
Kluwer Academic Publishers, 2004),
pp. 171–92.
Guimarães, L., ‘The Gallant and the
Philosopher’, Hume Studies 30(1) (2004),
pp. 127–47.
Immerwahr, J., ‘David Hume, Sexism and
Sociobiology’, Southern Journal of
Philosophy 21 (1988), pp. 359–69.
Jacobson, A.J. (ed.), Feminist Interpretations
of David Hume (University Park, PA:
Pennsylvania State University Press, 2000).
Lacoste, L.M., ‘The Consistency of Hume’s
Position concerning Women’, Dialogue
15 (1976), pp. 425–40.
2/9/2012 9:38:26 PM

Levy, A., ‘Under Constraint: Chastity and
Modesty in Hume’, Hume Studies 23(2)
(1997), pp. 213–26.
Saap, V.J., ‘The Philosopher’s Seduction:
Hume and the Fair Sex’, Philosophy and
Literature 19 (1995), pp. 1–15.
31. RACE AND PREJUDICE
Bracken, H.M., ‘Essence, Accident and
Race’, Hermathena 116 (1973),
pp. 81–96.
Chamley, P.E., ‘The Conflict between
Montesquieu and Hume: A Study of the
Origins of Adam Smith’s Universalism’, in
A.S. Skinner and T. Wilson (eds), Essays
on Adam Smith (Oxford: Clarendon
Press, 1975), pp. 274–305.
Eze, E.C., ‘Hume, Race, and Human
Nature’, Journal of the History of Ideas
61 (2000), pp. 691–8.
Immerwahr, J., ‘Hume’s Revised Racism’,
Journal of the History of Ideas 53 (1992),
pp. 481–6.
Mazza, E., Falsi e cortesi: pregiudizi,
stereotipi e caratteri nazionali in
Montesquieu, Hume e Algarotti (Milan:
U. Hoepli, 2002).
Palter, R., ‘Hume and Prejudice’, Hume
Studies 21(1) (1995), pp. 3–24.
Popkin, R.H., ‘Hume’s Racism’, The
Philosophical Forum 9(2–3) (1977–8),
pp. 211–26; repr. in R.H. Popkin, The
High Road to Pyrrhonism, ed. R.A.
Watson and J.E. Force (San Diego, CA:
Austin Hill Press, 1980), pp. 251–66.
–––––, ‘Hume’s Racism Reconsidered’, in his
The Third Force in Seventeenth-Century
Thought (Leiden: E.J. Brill, 1992),
pp. 64–75.
431
9780826443595_Ref_Final_txt_print.indd 431
BIBLIOGRAPHY
Valls, A.,‘“A Lousy Empirical Scientist”:
Reconsidering Hume’s Racism’, in A.
Valls (ed.), Race and Racism in Modern
Philosophy (Ithaca, NY: Cornell
University Press, 2005), pp. 127–49.
32. ECONOMICS
Berdell, J., ‘Innovation and Trade: David
Hume and the Case for Freer Trade’,
History of Political Economy 28(1)
(1996), pp. 107–26.
Berry, C.J., ‘Hume and Superfluous Value
(or the Problem with Epictetus’ Slippers)’,
in C. Wennerlind and M. Schabas
(eds), David Hume’s Political Economy
(London: Routledge, 2008), pp. 49–64.
–––––, ‘Hume and the Customary Causes
of Industry, Knowledge, and Humanity’,
History of Political Economy 38 (2006),
pp. 291–317.
Boyd, R., ‘Manners and Morals: David
Hume on Civility, Commerce, and the
Social Construction of Difference’, in C.
Wennerlind and M. Schabas (eds), David
Hume’s Political Economy (London:
Routledge, 2008), pp. 65–85.
Deleule, D., Hume et la naissance du
libéralisme économique (Paris: Aubier
Montaigne, 1979).
Emerson, R., ‘The Scottish Contexts for
David Hume’s Political-Economic
Thinking’, in C. Wennerlind and M.
Schabas (eds), David Hume’s Political
Economy (London: Routledge, 2008),
pp. 10–30.
Grüne-Yanoff, T., and McClennen, E.F.,
‘Hume’s Framework for a Natural
History of the Passions’, in C. Wennerlind
and M. Schabas (eds), David Hume’s
2/9/2012 9:38:26 PM
 BIBLIOGRAPHY
Political Economy (London: Routledge,
2008), pp. 86–104.
Henderson, W., The Origins of David Hume’s
Economics (London: Routledge, 2010).
Hont, I., Jealousy of Trade: International
Competition and the Nation-State in
Historical Perspective (Cambridge MA:
Harvard University Press, 2005).
Nakano, T., ‘“Let Your Science Be Human”:
Hume’s Economic Methodology’,
Cambridge Journal of Economics 30
(2006), pp. 687–700.
Ross, I.S., ‘The Emergence of David
Hume as a Political Economist’, in C.
Wennerlind and M. Schabas (eds), David
Hume’s Political Economy (London:
Routledge, 2008), pp. 31–48.
Rotwein, E., ‘David Hume, Philosopher-
Economist’, in K.R. Merrill and R.W.
Shahan (eds), David Hume: Many-sided
Genius (Norman, OK: University of
Oklahoma Press, 1976), pp. 117–34.
–––––, ‘Introduction’, in E. Rotwein (ed.),
David Hume: Writings on Economics,
with new introduction by M. Schabas
(Piscataway, NJ: Transaction Publishers,
2007 [1955]), pp. ix–cxi.
Sakamoto, T., ‘Hume’s Economic Theory’,
in E.S. Radcliffe (ed.), A Companion to
Hume (Oxford: Wiley-Blackwell, 2008),
pp. 373–87.
Schabas, M., ‘Groups versus Individuals in
Hume’s Political Economy’, The Monist
90(2) (2007), pp. 200–12.
–––––, ‘Temporal Dimensions in Hume’s
Monetary Theory’, in C. Wennerlind and
M. Schabas (eds), David Hume’s Political
Economy (London: Routledge, 2008),
pp. 127–45.
–––––, ‘Hume’s Monetary Thought
Experiments’, Studies for the History and
Philosophy of Science Part A 39 (2008),
pp. 161–9.
432
9780826443595_Ref_Final_txt_print.indd 432
–––––, The Natural Origins of Economics
(Chicago, IL: University of Chicago Press,
2005).
Skinner, A., ‘Hume’s Principles of Political
Economy’, in D.F. Norton and J. Taylor
(eds), The Cambridge Companion to
Hume, 2nd edn (Cambridge: Cambridge
University Press, 2009), pp. 381–413.
Soule, E., ‘Hume on Economic Policy and
Human Nature’, Hume Studies 26(1)
(2000), pp. 143–57.
Stockton, C.N., ‘Economics and the
Mechanism of Social Progress in Hume’s
History’, in D.W. Livingston and J.T.
King (eds), Hume: A Re-evaluation (New
York: Fordham University Press, 1976),
pp. 296–320.
Sugden, R., ‘Hume’s Non-instrumental and
Non-propositional Decision Theory’,
Economics and Philosophy 22(2) (2006),
pp. 365–91.
Susato, R., ‘Hume’s Nuanced Defense of
Luxury’, Hume Studies 32(1) (2006),
pp. 167–86.
Velk. T., and Riggs, A.R., ‘David Hume’s
Practical Economics’, Hume Studies 11(2)
(1985), pp. 154–65.
Wennerlind, C., ‘An Artificial Virtue and the
Oil of Commerce’, in C. Wennerlind and
M. Schabas (eds), David Hume’s Political
Economy (London: Routledge, 2008),
pp. 105–26.
–––––, ‘David Hume’s Monetary Theory
Revisited: Was He Really a Quantity
Theorist and an Inflationist?’, Journal
of Political Economy 113 (2005),
pp. 223–37.
–––––, ‘David Hume’s Political Philosophy:
A Theory of Commercial Modernization’,
Hume Studies 28(2) (2002), pp. 247–70.
–––––, and Schabas, M. (eds), David Hume’s
Political Economy (London: Routledge,
2008).
2/9/2012 9:38:26 PM

33. ART AND CRITICISM
Bouveresse, R., Esthétique, psychologie et
musique: l’esthétique expérimentale et
son origine philosophique chez David
Hume (Paris: J. Vrin, 1995).
Brunet, O., Philosophie et esthétique chez
David Hume (Paris: Librairie A.-G. Nizet,
1965).
Brunius, T., David Hume on Criticism
(Stockholm: Almqvist & Wiksell, 1952).
Carabelli, G., On Hume and Eighteenth-
Century Aesthetics: The Philosopher on a
Swing, trans. J.K. Hall (New York: Peter
Lang, 1995).
Carroll, N., ‘Hume’s Standard of Taste’, The
Journal of Aesthetics and Art Criticism 43
(1984), pp. 181–94.
Cohen, R., ‘The Rationale of Hume’s
Literary Inquiries’, in K.R. Merrill
and R.W. Shahan (eds), David Hume:
Many-sided Genius (Norman, OK:
University of Oklahoma Press, 1976),
pp. 97–115.
Costelloe, T.M., ‘Hume’s Aesthetics: The
Literature and Directions for Research’,
Hume Studies 30(1) (2004), pp. 87–126.
Dadlez, E.M., Mirrors to One Another:
Emotion and Value in Jane Austen and
David Hume (Malden, MA: Wiley-
Blackwell, 2009).
–––––, ‘The Vicious Habits of Entirely
Fictive People: Hume on the Moral
Evaluation of Art’, Philosophy and
Literature 26 (2002), pp. 143–56.
Friday, J., ‘Hume’s Sceptical Standard
of Taste’, Journal of the History of
Philosophy 36(4) (1998), pp. 545–66.
Hipple, W.J., The Beautiful, The Sublime
and the Picturesque in Eighteenth-
Century British Aesthetic Theory
(Carbondale, IL: Southern Illinois
University Press, 1957).
433
9780826443595_Ref_Final_txt_print.indd 433
BIBLIOGRAPHY
Jones, P., ‘“Art” and “Moderation”
in Hume’s Essays’, in D.F. Norton,
N. Capaldi and W.L. Robison (eds),
McGill Hume Studies (San Diego, CA:
Austin Hill Press, 1979), pp. 161–80.
–––––, ‘Cause, Reason, and Objectivity
in Hume’s Aesthetics’, in D.W.
Livingston and J.T. King (eds), Hume:
A Re-evaluation (New York: Fordham
University Press, 1976), pp. 323–42.
–––––, ‘Hume on Art, Criticism and
Language: Debts and Premises’,
Philosophical Studies 33 (1978),
pp. 109–34.
–––––, ‘Hume’s Aesthetics Reassessed’,
The Philosophical Quarterly 26 (1976),
pp. 48–62.
–––––, ‘Hume’s Literary and Aesthetic
Theory’, in D.F. Norton (ed.), The
Cambridge Companion to Hume
(Cambridge: Cambridge University Press,
1993), pp. 255–80.
Kivy, P., ‘Hume’s Neighbour’s Wife: An
Essay on the Evolution of Hume’s
Aesthetics’, British Journal of Aesthetics
23 (1983), pp. 195–208.
–––––, ‘Hume’s Standard of Taste: Breaking
the Circle’, British Journal of Aesthetics 7
(1967), pp. 57–66.
Korsmeyer, C.W., ‘Hume and the
Foundations of Taste’, The Journal of
Aesthetics and Art Criticism 35 (1976),
pp. 201–15.
Levinson, J., ‘Hume’s Standard of Taste:
The Real Problem’, The Journal of
Aesthetics and Art Criticism 60 (2002),
pp. 227–38.
MacLachlan, C., ‘Hume and the Standard
of Taste’, Hume Studies 12(1) (1986),
pp. 18–38.
MacMillan, C., ‘Hume, Points of View and
Aesthetic Judgements’, Journal of Value
Inquiry 20 (1986), pp. 109–23.
2/9/2012 9:38:30 PM
 BIBLIOGRAPHY
Marshall, D., ‘Arguing by Analogy: Hume’s
Standard of Taste’, Eighteenth-Century
Studies 28 (1995), pp. 323–43.
Mossner, E.C., ‘Hume’s “Of Criticism”’, in
H. Anderson and J.S. Shea (eds), Studies
in Criticism and Aesthetics, 1660–1800:
Essays in Honor of Samuel Holt Monk
(Minneapolis, MN: University of
Minnesota Press, 1967), pp. 232–48.
Mothersill, M., ‘Hume and the Paradox
of Taste’, in G. Dickie, R. Sclafani and
R. Roblin (eds), Aesthetics: A Critical
Anthology, 2nd edn (New York: St.
Martin’s Press, 1989), pp. 269–86.
–––––, ‘In Defense of Hume and the
Causal Theory of Taste’, The Journal of
Aesthetics and Art Criticism 55 (1997),
pp. 312–17.
Neill, A., ‘“An Unaccountable Pleasure”:
Hume on Tragedy and the Passions’,
Hume Studies 24(2) (1998), pp. 335–54.
Noxon, J., ‘Hume’s Opinion of Critics’, The
Journal of Aesthetics and Art Criticism 20
(1961), pp. 157–62.
Osborne, H., ‘Hume’s Standard and the
Diversity of Aesthetic Taste’, British
Journal of Aesthetics 7 (1967), pp. 50–6.
Paton, M., ‘Hume on Tragedy’, British Journal
of Aesthetics 13 (1973), pp. 121–32.
Rose, M.C., ‘The Importance of Hume
in the History of Western Aesthetics’,
British Journal of Aesthetics 16 (1976),
pp. 218–29.
Saccamano, N., ‘Parting with Prejudice:
Hume, Identity, and Aesthetic
Universality’, in V. Kahn, N. Saccamano
and D. Coli (eds), Politics and the
Passions, 1500–1850 (Princeton, NJ:
Princeton University Press, 2006),
pp. 175–95.
Shiner, R.A., ‘Hume and the Causal Theory
of Taste’, The Journal of Aesthetics and
Art Criticism 54 (1996), pp. 237–49.
434
9780826443595_Ref_Final_txt_print.indd 434
Sugg, R.S., ‘Hume’s Search for the Key with
the Leathern Thong’, The Journal of
Aesthetics and Art Criticism 16 (1957),
pp. 96–102.
Townsend, D., Hume’s Aesthetic Theory:
Taste and Sentiment (London: Routledge,
2001).
34. HISTORY
Burke, J.J., ‘Hume’s History of England:
Waking the English from a Dogmatic
Slumber’, Studies in Eighteenth-Century
Culture 7 (1978), pp. 235–50.
Capaldi, N., and Livingston, D.W. (eds),
Liberty in Hume’s History of England
(Dordrecht: Kluwer Academic, 1990).
Davis, N.Z., ‘History’s Two Bodies’,
American Historical Review 93 (1988),
pp. 1–30.
Farr, J., ‘Hume, Hermeneutics, and History:
A “Sympathetic” Account’, History and
Theory 17 (1978), pp. 285–310.
Gautier, C., Hume et les savoirs de l’histoire
(Paris: Vrin, 2005).
Hicks, P.S., Neo-classical History and
English Culture: From Clarendon to
Hume (Basingstoke: Macmillan, 1996).
Hilson, J.C., ‘Hume: The Historian as a
Man of Feeling’, in J.C. Hilson, M.M.B.
Jones and J.R. Watson (eds), Augustan
Worlds (Leicester: Leicester University
Press, 1978), pp. 205–22.
Miller, J.J., ‘Neither Whig nor Tory: A
Philosophical Examination of Hume’s
Views on the Stuarts’, History of
Philosophy Quarterly 19 (2002),
pp. 275–308.
Moses, G., ‘David Hume as Philosophical
Historian’, Australian Journal of Politics
and History 35(1) (1989), pp. 80–91.
2/9/2012 9:38:34 PM

Mossner, E.C., ‘Was Hume a Tory Historian?
Facts and Reconsiderations’, Journal of
the History of Ideas 2 (1941), pp. 225–36.
Norton, D.F., ‘History and Philosophy
in Hume’s Thought’, in D.F. Norton
and R.H. Popkin (eds), David Hume:
Philosophical Historian (Indianapolis:
Bobbs-Merrill, 1965), pp. xxxii–l.
–––––, and Popkin, R.H. (eds), David
Hume: Philosophical Historian
(Indianapolis: Bobbs-Merrill, 1965).
O’Brien, K., Narratives of Enlightenment:
Cosmopolitan History from Voltaire
to Gibbon (Cambridge: Cambridge
University Press, 1997).
Okie, L., ‘Ideology and Partiality in David
Hume’s History of England’, Hume
Studies 11(1) (1985), pp. 1–32.
Phillips, M.S., Society and Sentiment:
Genres of Historical Writing in Britain,
1740–1820 (Princeton, NJ: Princeton
University Press, 2000).
–––––, ‘“The Most Illustrious Philosopher
and Historian of the Age”; Hume’s
History of England’, in E.S. Radcliffe
(ed.), A Companion to Hume (Oxford:
Wiley-Blackwell, 2008), pp. 406–22.
Phillipson, N., Hume: The Philosopher as
Historian (Harmondsworth: Penguin,
2011).
Pompa, L., Human Nature and Historical
Knowledge: Hume, Hegel and Vico
(Cambridge: Cambridge University Press,
1990).
Popkin, R.H., ‘David Hume: Philosophical
versus Prophetic Historian’, in K.R.
Merrill and R.W. Shahan (eds), David
Hume: Many-sided Genius (Norman,
OK: University of Oklahoma Press,
1976), pp. 83–95.
Potkay, A., ‘Hume’s “Supplement to
Gulliver”: The Medieval Volumes of The
435
9780826443595_Ref_Final_txt_print.indd 435
BIBLIOGRAPHY
History of England’, Eighteenth-Century
Life 25(2) (2001), pp. 32–46.
Siebert, D.T., ‘The Sentimental Sublime in
Hume’s History of England’, The Review
of English Studies, n.s. 40(159) (1989),
pp. 352–72.
Stockton, C.N., ‘David Hume: Historian
of the English Revolution’, Eighteenth-
Century Studies 4(3) (1971),
pp. 274–94.
Sullivan, M.G., ‘Rapin, Hume and the
Identity of the Historian in Eighteenth-
Century England’, History of European
Ideas 28 (2002), pp. 145–62.
Voigt, U., David Hume und das Problem der
Geschichte (Berlin: Duncker & Humblot,
1975).
Walton, C., ‘Hume and Jefferson on the
Uses of History’, in D.W. Livingston and
J.T. King (eds), Hume: A Re-evaluation
(New York: Fordham University Press,
1976), pp. 389–403.
Wertz, S.K., Between Hume’s Philosophy
and History: Historical Theory and
Practice (Lanham, MD: University Press
of America, 2000).
–––––, ‘Moral Judgements in History:
Hume’s Position’, Hume Studies 12(2)
(1996), pp. 339–67.
Wexler, V.G., David Hume and The
History of England (Philadelphia, PA:
American Philosophical Association,
1979).
–––––, ‘David Hume’s Discovery of a
New Science of Historical Thought’,
Eighteenth-Century Studies 10(2) (Winter
1976–7), pp. 185–202.
Wootton, D., ‘Hume, “the Historian”’,
in D.F. Norton, The Cambridge
Companion to Hume (Cambridge:
Cambridge University Press, 1993),
pp. 281–312.
2/9/2012 9:38:34 PM
9780826443595_Ref_Final_txt_print.indd 436 2/9/2012 9:38:35 PM
 INDEX OF NAMES

Adams, Marilyn 297 Boyle, Robert 246

Addison, Joseph 24, 350 Bramhall, Bishop Thomas 214
Ajax 328 Bruyère, Jean de La 22
Alexander the Great 328, 366–7, 372 Buckle, Stephen 241n. 4
Alfred the Great 289, 374 Bunyan, John 350
Allen, John 225 Butler, Joseph 80, 168, 251
Allestree, Richard 297
Aquinas, Thomas 246 Campbell, Archibald 332
Árdal, Pall 280, 285 Cantillon, Richard 334
Ariosto 353, 355 Carnap, Rudolf 243n. 29
Aristotle 1, 211, 333 Carr, Edward 201
Arnauld, Antoine 376n. 6 Carroll, Lewis 207, 354
Ayer, Alfred 381, 383, 389, 393n. 17 Cavendish, Margaret 329
Cervantes, Miguel de 356–9
Babbage, Charles 243n. 34 Child, Josiah 333
Bacon, Francis 56n. 47 Chisholm, Roderick 170
Baier, Annette 204, 206, 265, 277–8n. 16, Churchland, Patricia 53
286, 296, 299, 300n. 8, 319–20, Churchland, Paul 53
324, 326 Cicero 29, 246, 247, 300n. 10, 302n. 22
Bayes, Thomas 238–40 Clarke, Samuel 215, 385–6
Bayle, Pierre 24, 264n. 32 Coady, Anthony 236
Beattie, James 379 Cohon, Rachel 210, 286, 287n. 1
Beauchamp, Tom 73 Collins, Anthony 215, 218
Bentham, Jeremy 287n. 13, 300n. 7, 301n. Condillac, Étienne de 334
13, 316, 337 Coutts, John 332
Berkeley, George 18, 100n. 90, 334, 377, Coventry, Angela 141
379–83, 386 Craig, Edward 132, 136, 144, 197n. 20,
Blackburn, Simon 145n. 24, 291 242n. 15
Blair, Hugh 34, 240 Croce, Benedetto 201
Bodin, Jean 333 Cullen, William 20
Boisguilbert, Pierre de 333
Bolingbroke, Henry 33 D’Alembert, Jean-Baptiste 4, 33
Boufflers, Comtesse de 25, 27, 320–2 Dauer, Francis 46–50

437

9780826443595_Index of Names_Final_txt_print.indd 437 2/9/2012 9:39:50 PM

 INDEX OF NAMES

Davidson, Donald 201–2 Haakonsen, Knud 342

Dawkins, Richard 264n. 37 Haldane, John 305
Dees, Richard 299n. 1 Hardin, Russell 300n. 7, 301n. 13
Defoe, Daniel 334 Harman, Gilbert 56n. 28
Dennett, Daniel 54n. 6, 54n. 7, 54–5n. 8 Harris, James 300n. 8
Descartes, René 4, 6, 7, 42, 43–4, 46, 52–3, Hart, Herbert 316
58–9, 98n. 64, 100n. 90, 100n. 95, 132, Hartley, David 100n. 103, 220
138, 144, 171, 267, 276, 278n. 28, Harvey, William 249, 335
300n. 4, 386, 392n. 5 Helvetius, Claude 334
D’Holbach, Paul-Henri 20, 394n. 32 Herdt, Jennifer 300n. 4
Diderot, Denis 26 Herodotus 295, 366–7
Dodgson, Charles 213n. 36 Hertford, Lord 25, 27–9
Dretske, Fred 56n. 28 Highmore, Joseph 224
Du Tot, Charles de Ferrière 334, 346n. 22 Hobbes, Thomas 8, 214, 215, 218, 219,
Dubos, Jean-Baptiste 359 223, 225, 280, 303, 305, 306, 308, 312,
314, 316, 343, 381, 385
Earman, John 143 Holden, Thomas 257, 259–60, 263
Elliot, Gilbert 34, 244, 252 Home, Henry see Kames, Lord
Epicurus 154, 201, 246, 256, 347n. 48 Homer 323, 355–6
Euripides 323 Horace 307, 323, 352
Howson, Colin 143
Fairholm, Adam 332 Hunter, William 1
Ferguson, Adam 33 Hutcheson, Francis 19, 24, 34, 79–81, 221,
Fielding, Sarah 329 289–90, 293–4, 296, 382, 394n. 33
Flew, Antony 72, 242n. 13, 393n. 16
Fodor, Jerry 54n. 3, 394n. 27 Inwagen, Peter van 143
Fogelin, Robert 166n. 2, 166n. 5, 233–4,
242n. 13, 243n. 23 Joan of Arc 367
Forbannais, François de 334 Johnson, Samuel 75
Franklin, Benjamin 334 Jonson, Ben 33
Fricker, Elizabeth 236
Friedman, Milton 333 Kail, Peter 144, 145n. 15, 266, 269, 276
Kames, Lord 195–6n. 2, 394n. 31
Galen 246, 249, 261 Kant, Immanuel 1, 46, 52–3, 245, 253,
Galiano, Ferdinando 334 264n. 21, 281, 314, 379, 381, 387
Garrett, Don 73–9, 80, 101n. 110, 102n. Kemp Smith, Norman 247, 311, 382,
118, 102n. 121, 136, 166n. 7, 188–94, 394n. 33
196n. 8, 196n. 13, 197n. 17, 286 Kenyon, John 102n. 126
Gaskin, John 252–3, 261–2, 265, 266–7, Keynes, John Maynard 337
278n. 21 King, Bishop William 215
Gee, Joshua 334 Knowles, David 201
Genovesi, Antonio 334 Korsgaard, Christine 202, 208, 211n. 9
Gervaise, Isaac 334
Govier, Trudy 51–2 Laird, John 307, 394n. 31
Green, Thomas 392n. 5 Larrère, Catherine 346n. 18
Gregory, James 225 Law, Edmund 215
Grose, Thomas 392n. 5 Law, John 333–4

438

9780826443595_Index of Names_Final_txt_print.indd 438 2/9/2012 9:39:50 PM

 INDEX OF NAMES

Leechman, William 24 Norton, David 208, 226n. 9, 300n. 8

Leibniz, Gottfried 100n. 90, 168, 215, 386 Norton, Mary 208, 226n. 9, 300n. 8
Libet, Benjamin 203
Livingston, Donald 364–5 Ogilby, John 350, 353
Locke, John 4, 8, 18, 39, 62, 72, 73, 79, 83, Orde, Nancy 320
86, 87, 89, 100n. 90, 100n. 103, 167–8, Oswald, James 333, 348n. 54
177, 214, 216, 219, 223, 225, 314, 333, Ovid 323, 352
377, 379–83, 386 Owen, David 72, 74–8, 97n. 50
Loeb, Louis 98n. 67, 103n. 127, 130n. 37
Lucian 22, 24, 28–34 Paul, Saint 255
Lucretius 201, 256, 394n. 32 Pinto, Isaac de 332
Plato 1
Mackie, John 133, 313 Pluche, Noël-Antoine 334
MacLaurin, Colin 246, 250–1 Plutarch 372
Magri, Tito 207 Pope, Alexander 224, 361
Mandeville, Bernard 25, 221, 290, 294–6, Popkin, Richard 376n. 8
303, 305, 312, 317, 334 Popper, Karl 49
Mappes, Thomas 73 Price, Henry 44–5
Marx, Karl 1 Price, Richard 76–7, 80–1, 101n. 105,
Massie, Joseph 334 238
Medea 328 Priestley, Joseph 220, 225, 262
Melon, Jean-François 334 Pritchard, Duncan 166n. 6
Mill, John Stuart 56n. 47, 313, 337, Putnam, Hilary 53
345n. 9, 381, 383
Millar, Andrew 276 Quesnay, Francis 333–5
Millican, Peter 225, 239 Quine, Willard 53, 317
Milton, John 350, 353 Quintus Curtius 366–7
Mirabeau, Victor 334
Montaigne, Michel de 33 Radcliffe, Elizabeth 205
Montesquieu, Charles-Louis de 25, 334, Rawls, John 306, 310, 312
339, 343 Reichenbach, Hans 239
Moore, George 241n. 4 Reid, Thomas 80–1, 100n. 92, 100n. 96,
Morellet, André 29, 332, 334 168, 225, 311, 379, 381, 383,
Moses, Gregory 365 392n. 5
Mossner, Ernest 364–5 Ricardo, David 333
Motte, Antoine de la 342 Robinson, John 136, 144n. 8
Mure, William 348n. 54 Rochefoucauld, François de La 35n. 4,
301n. 16
Nagel, Thomas 47 Rousseau, Jean-Jacques 4, 20, 22, 27, 31–3,
Newton, Isaac 19, 42, 66, 114, 152, 153, 320, 322, 334, 335
159, 191, 192, 197n. 22, 241n. 2, Russell, Bertrand 53, 377, 381, 383, 389,
245–6, 248, 304, 322, 382–3, 392n. 10
394n. 33 Russell, Paul 225
Nicole, Pierre 376n. 6 Rymer, Thomas 359
Nietzsche, Friedrich 394n. 32
Noë, Alva 48–9 St Clair, General James 23, 24–5, 27, 332,
North, Dudley 333 343, 344n. 2

439

9780826443595_Index of Names_Final_txt_print.indd 439 2/9/2012 9:39:51 PM

 INDEX OF NAMES

Say, Jean-Baptiste 333 Swift, Jonathan 22, 334

Sayre-McCord, Geoffrey 284, 286 Tacitus 343, 352, 366
Schafer, Karl 213n. 35 Tawney, Richard 315
Schelling, Thomas 303, 309 Thucydides 367
Scruton, Roger 392n. 10, 394n. 29 Tucker, Josiah 332, 334
Sellars, Wilfrid 202 Turgot, Anne-Robert-Jacques 332–4, 339
Seth, James 380 Tweyman, Stanley 251
Shafer-Landau, Russ 282
Shaftesbury, Anthony Ashley-Cooper, Earl Vanderlint, Jacob 334
of 80, 337 Vincent de Gournay, J.-C.-M. 334
Shakespeare, William 356 Vitellius 328
Sidgwick, Henry 337 Voltaire 263n. 12
Siebert, Donald 376n. 11
Sieyès, Emmanuel-Joseph 346n. 18 Wallace, Reverend Robert 346n. 24
Smith, Adam 4, 15, 30–4, 247, 303, Walpole, Horace 28
311–12, 317, 332–3, 335, 337–9, Walpole, Robert 343
340–1, 342, 346n. 22 Walsh, William 377
Smith, Michael 141 Warburton, William 21–2, 276
Snare, Francis 285 Waxman, Wayne 49–51
Sobel, Jordan 243n. 26, 243n. 32 Williams, Bernard 395n. 36
Spinoza, Baruch 100n. 90, 172–3, 386 Windelband, Wilhelm 392n. 7
Sterne, Laurence 4, 28, 354 Wittgenstein, Ludwig 53, 192, 199, 359
Steuart, James 332, 339 Wootton, David 376n. 6
Stewart, Dugald 247, 376n. 12 Wordsworth, William 268
Stove, David 72, 94–5n. 19 Wright, John 138–40, 141–2, 144, 225
Strahan, William 30–4
Strawson, Galen 139–40 Xenophon 333
Stroud, Barry 44, 72, 96–7n. 39, 102n. 126,
138, 144n. 4, 313 Yandell, Keith 266

440

9780826443595_Index of Names_Final_txt_print.indd 440 2/9/2012 9:39:51 PM

 INDEX OF TOPICS
Academic scepticism 150–1
action 8, 199–213, 214, 374–5
regularity of human 8–9, 202, 220,
222–3, 333, 373–4, 385
aesthetics 327, 330, 349–63, 366, 375, 382
delicacy of taste 326, 351, 356–9
judges and critics 16, 352, 360–2
normativity 16, 349–63
projection 6, 141–2, 350
rules of composition and art 16, 353–62
standard of taste 15, 101n. 109, 145n.
33, 145n. 34, 324, 326, 349–63
afterlife 117, 271
agnosticism 2, 9, 262
altruism 13, 295, 305
analogy, argument from 246, 248–9, 251,
253–8
see also argument to design
animals 14, 50, 82, 90, 95n. 28, 99n. 84,
128n. 17, 211, 320, 382
anthropomorphism 11, 84, 250, 252–3,
256, 268–78
argument to design 2, 10, 11, 245–64, 265,
266, 273, 275, 276, 277n. 11
associationism (association of ideas) 5, 14,
104
see also imagination
atheism 2, 19, 172, 257–8, 262–3, 276,
311, 386, 388
accusations of 24, 33
moral 263
Bayes, theorem 10, 238–40, 333
beauty 292–3, 297, 350–1
441
9780826443595_Index of Topics_Final_txt_print.indd 441
moral 293, 301n. 12, 375
belief
as distinguished from imagination 41, 43,
46, 52
Hume’s dissatisfaction with his account
of 54n. 4, 206–7
intentionality of 4, 43–9, 50, 52
and judgement 8, 142–3, 207–9
and motivation 204–9
natural 191, 227–8, 251–3, 261, 272,
278n. 21
scepticism and suspension of 158–60,
163
as vivid idea 4–5, 8, 38, 42, 63, 83, 113,
115, 118–19, 119–20, 122–4,
206–7, 371
beneficence 288, 305, 323, 325–6
benevolence 3, 12–13, 14, 280, 288, 290,
291, 292, 296, 301n. 17, 328, 330, 341,
342
business 31, 297, 375
Cartesian scepticism 59
causa secundum esse 254–5
causa secundum fieri 254–5
causal maxim 60, 76, 77, 106, 218
causal powers 5–6, 71, 94n. 12, 222
observable 133
relative idea of 139, 140–1
secret 5, 128n. 15, 128n. 22, 139, 143,
153–4, 189, 200, 203, 369, 375
causal reasoning 4–5, 7, 9–10, 57–103,
104–30, 150, 161–2, 166, 199–200,
208, 233
2/9/2012 9:37:48 PM
 INDEX OF TOPICS

rules by which to judge of causes and
effects 6, 78, 88, 120–1, 125, 129n. 24,
138, 142
causation 5, 7, 38, 131–45, 274, 320, 379,
381, 385, 392n. 4, 393n. 13
causal maxim 60, 76, 77, 106, 218
causal powers 5–6, 71, 94n. 12, 128n. 15,
128n. 22, 133, 139, 153–4, 189, 200,
203, 222, 369, 375
New Hume debate 5, 95n. 21, 225
principle of association of 5, 176–7,
372
two definitions of cause 131, 135–7, 138,
202–3, 211n. 1
see also causal reasoning; cause, two
definitions of; necessary connection
cause, two definitions of 131, 135–7, 138,
202–3, 211n. 1
chance 218–19
character 12–13, 15–16, 17, 18, 199–204,
220, 279, 286, 289, 290–1, 293, 295–6,
301n. 12, 307, 320–2, 329, 352,
373, 374
national 343, 352
cognitive science 54n. 3, 86–9, 198n. 29,
394n. 27
colligation 377, 384
common point of view 5, 12, 281–5, 286–7,
290–2, 294–5, 297, 328, 367
compatibilism, with respect to free will 8,
203, 214, 222–3
Conceivability Principle 61, 62, 64, 66, 68,
70, 97n. 45, 112, 127n. 4, 171–3
contiguity
spatial 5, 105–6, 111–12, 114–17, 122,
126, 128n. 10, 128n. 22
temporal (succession) 105–6, 111–12,
128n. 10
contracts 295, 312, 314, 317
conventions 3, 294, 296, 306, 308–10, 312,
313, 315, 316
copy principle 38–56, 368–9
cosmological argument and the origin of the
universe 11, 245, 265, 273,
277n. 11, 386
custom 4–5, 52, 59, 63, 65, 69, 70, 72, 73,
84, 87, 89, 107, 113, 117, 121, 122,
125–6, 129n. 26, 129n. 27, 132–3, 165,
199, 205, 228, 272, 278n. 21
death
afterlife 27
Hume’s 4, 25, 30–4, 335
suicide 4, 23
deism 9, 11
accusation of 24, 28
attenuated 261–2, 265
English 215
delicacy of taste 326, 351, 356–9
demonstrations 62, 64, 66, 68, 70, 75, 76,
83, 85, 213n. 37, 228
design, argument to 2, 10, 11, 245–64, 265,
266, 273, 275, 276, 277n. 11
desire 8–9, 79–81, 203–7, 210,
279–81, 286
determinism 8, 218
dialogue form 247
Dialogues concerning Natural Religion,
publication of 30–4
dogmatism 326
duty 279
artificial 312, 315–16
natural 312
economics 1, 15–16, 332–48
physiocrats 334–5, 348
specie-flow mechanism 339
Edinburgh, University of 24
education 11, 15, 116, 118–19, 125–6
egoism 279
empathy and sympathy 12, 13, 14, 17,
18, 25, 200, 211, 279, 280, 285, 286,
290–6, 298, 303, 307–8, 325–7, 328,
372, 375
empiricism, 6–9, 38–56, 245, 249
British 18–19, 44–5, 377, 379–95
meaning 39, 141, 381
and personal identity 181–98
Enlightenment, The 54, 395n. 35
Enquiry and Treatise, relation between 9,
63, 65, 68, 69, 70, 76, 85, 87, 89, 95n.
21, 103n. 129, 139, 143, 153–4, 223–4,
379, 385, 388
enthusiasm 374

442

9780826443595_Index of Topics_Final_txt_print.indd 442 2/9/2012 9:37:48 PM


Epicureanism 154, 201, 246, 256, 347n. 48
evil, problem of 9, 253–5, 258–60, 266,
268–9
inferential 258–60, 277n. 11
logical 258
experimental method 2, 152, 165–6,
392n. 11
external world, belief in 7, 42–3, 45, 51,
52–3, 82, 86, 88, 127, 169, 174–5, 176,
185, 186, 191, 194, 251, 252, 264n. 20,
272, 274, 379, 381, 385
externalism, epistemological 53
faculties 86–7, 101n. 109
feminism 14, 319–31
fideism 276
force and vivacity 4–5, 38–56, 63, 83,
108–10, 114–17, 123, 126, 206, 280,
299n. 4, 323, 365, 370, 371, 372
as functional role 51–2
and intentional content 4, 43–9, 50, 52
as qualitative feel (qualia) 4, 43–6, 50
as verisimilitude 4, 49–51
France 25–8, 33, 295, 320–5, 327, 333–4, 336
free will 8, 203, 214–26, 341, 379
compatibilism 8, 203, 214, 222–3
determinism 8, 218
and God’s prescience 9, 220, 223–4
libertarianism 221
friendship 3–4, 320–2, 327, 329–30, 337,
342, 373
gender 14, 319–31
general point of view 5, 12, 281–5, 286–7,
290–2, 294–5, 297, 328, 367
Glasgow, University of 335
greatness of mind 328
habit 4–5, 52, 59, 63, 65, 69, 70, 72, 73,
84, 87, 89, 107, 113, 117, 121, 122,
125–6, 129n. 26, 129n. 27, 132–3, 165,
199, 205, 228, 272, 278n. 21
happiness 3–4, 9, 12, 14–15, 81, 299,
332–48
history 4–5, 364–76
and character 18, 199–200, 289, 296
historiography 200, 205
443
9780826443595_Index of Topics_Final_txt_print.indd 443
INDEX OF TOPICS
Hume as a historian 391n. 3
Hume’s interest in 1, 14, 17–18, 24–5,
223, 324, 344
monkish historians 366
and normativity 201, 375
and philosophy 8, 17, 373, 394–5n. 34
Hume, David
death 4, 25, 30–4, 335
wealth 27
Hume’s fork 65, 205, 241n. 1, 381, 393n. 16
hypocrisy 24
ideas (and impressions) 38–56, 209, 281
and double existence 175
of existence 110
relative 139–41
theory of 4, 379, 392n. 5
Image of God hypothesis 5, 132, 138, 144
imagination
contiguity and succession 5, 105–6,
111–12, 114–17, 122, 126, 128n. 10,
128n. 22
copy principle 38–56, 368–9
and history 17, 365–6, 370–3
and personal identity 183–95
principles of association 40–1, 45–6, 63,
168–9, 174–7, 211, 216, 297, 299n. 4
and reason 83–7
and resemblance 5, 41, 105–6, 114–17,
118, 126, 176–7
indian prince 229–30, 241n. 10
induction 114–15, 239, 379, 385
inductive reasoning 4–5, 57–103,
199–200, 218, 251, 272, 336
problem of 4, 9, 107, 112, 124, 132,
261, 381
regularity theory of 227, 240
inference to the best explanation 114, 262
innateness 289–90, 296
intentionality, of beliefs 4, 43–9, 50, 52
introspection 170
Ireland 29–30
irony 29, 247
irreligion 2, 9, 19, 384–9
judgement
and belief 8, 142–3, 207–9
2/9/2012 9:37:48 PM
 INDEX OF TOPICS
moral 8, 12–13, 17, 204, 209, 279–87
justice 14, 220, 294–6, 304–7, 310,
311–17, 323
legal theory 315–16
libertarianism (metaphysical) 221
libertarianism (political) 340
liberty (and necessity) 8, 214–26
of indifference 203, 219, 221
libertarianism (metaphysical) 221
of spontaneity 203, 219
logical positivism 4, 53, 381–2, 390,
393n. 17
love 14, 293, 320–2, 324–5, 327, 329–30
luxury 15–16, 341–2
Manicheanism 259
materialism 257
mathematics 5, 62, 65, 66, 88, 96n. 33,
129n. 28, 381
meaning empiricism 39, 141, 381
memory 41, 49, 52, 58, 83–5, 87,
101n. 111, 104, 108, 111, 116, 117–18,
120, 128n. 23, 131, 209, 323, 365–6,
367, 368–9, 373
metaphysics 2, 6, 8–9, 18, 151–2, 155, 162,
214, 218, 222, 224, 225, 275, 381, 382,
388, 389
Conceivability Principle 61, 62, 64, 66,
68, 70, 97n. 45, 112, 127n. 4
Separability Principle 61, 173
middling station (middle class) 15, 340–2,
344
miracles 2, 9–10, 12, 28, 72, 94n. 17, 201,
222, 227–44, 266, 277n. 11, 376n. 8
and marvels 242n. 10, 367
and passions and emotions 234–5
see also testimony
mirroring 13, 303, 306, 307–8
mitigated scepticism 4, 60, 96n. 33, 150–1
monkish virtues 13, 284, 296–9
moralists 29, 221, 304
morality 16, 73, 127, 153, 209, 374,
379, 382
anti-realism and realism 279, 281–2,
286–7, 300n. 12
duty 279, 312, 315–16
444
9780826443595_Index of Topics_Final_txt_print.indd 444
egoism 279
emotivism 284–5
moral beauty 293, 301n. 12, 375
moral intuitions 306–7
moral judgement 8, 12–13, 17, 204, 209,
279–87
moral rules 291, 293–4, 298
moral sense theory 221, 301n. 12,
308, 311
moral sentiments (sentimentalism) 12–13,
141, 209–10, 279–87, 291–3, 298–9,
301n. 13, 326, 327, 328–9
and motivation 82, 204–9, 279, 280,
313, 317, 332
non-cognitivism and cognitivism 209,
284–7
and normativity 12, 13, 300n. 7, 310–11,
317
partiality 17, 305–6, 314, 315–16, 342
projection 141–2
and reason 13, 279, 293–4, 311
religious 11, 12, 13, 221, 280, 284,
296–9, 308, 311, 385
responsibility 203, 214, 220
rights 314, 317
subjectivism 12, 281, 285
utilitarianism 286, 301n. 12, 310, 311,
312–13, 316, 317, 337
see also virtue
motivation
and belief 204–9
Humean theory of 8, 204–9
moral 82, 204–9, 279, 280, 313, 317, 332
and reason 204–9, 382
national character 343, 352
natural abilities 13, 221, 292, 300n. 8,
300n. 10
natural relations 105, 107, 114–17, 136–7
naturalism 4–5, 53–4, 73, 199, 290, 292,
306–7, 311, 382–3, 392n. 11
materialism 257
religious 11, 240, 257–8, 262–3, 265–78,
298
and scepticism 4–5, 18–19, 384–5, 387
see also experimental method; science
of man
2/9/2012 9:37:49 PM

necessary connection, idea of 5, 60–1, 106,
111, 128n. 22, 131–45, 185, 186, 188–9,
191, 194–5, 203, 216–17, 222, 326
non-cognitivism 5–6, 131, 134, 137, 141–3
projectivism 5–6, 131, 134, 137, 141–3
regularity theory of 5–6, 131, 134,
137–8, 143, 202, 225
sceptical realism 5–6, 131, 134–5, 137,
138–41, 143–4, 225
New Hume debate 5, 95n. 21, 225
normativity
and aesthetics 16, 349–63
epistemological 4, 6, 74, 78, 79, 88,
120–1, 125, 129n. 24, 138, 142
and history 201, 375
and justice 314
and morality 12, 13, 300n. 7, 310–11,
317
psychological 304, 307, 309
see also general point of view
omnipotence 11, 253, 260
omniscience 253
ontological argument 98n. 64, 245
other minds, existence of 252
problem of 307
partiality 17, 305–6, 314, 315–16, 342
passions 14, 73, 100n. 100, 114, 115, 120,
126, 153, 220, 221, 291, 319, 326,
366, 375
calm 141
direct 209–11, 215–16
indirect 209–11, 212n. 29, 215–16
love 14, 293, 320–2, 324–5, 327,
329–30
and reason 8, 14, 82, 117, 204–9, 284,
285, 330
and religion 11, 268–72
personal identity
scepticism with respect to 8, 182,
299–300n. 4
self 7–8, 127, 167–98, 201, 260, 299n. 4,
379, 381
Philo’s U-turn 252, 261–2
philosophical relations 105
physiocrats 334–5, 338
445
9780826443595_Index of Topics_Final_txt_print.indd 445
INDEX OF TOPICS
poetry 17, 201, 323, 353, 365, 372, 373,
374
politics 202, 323, 374, 379
contracts 295, 312, 314, 317
conventions 3, 294, 296, 306, 308–10,
312, 313, 315, 316
polytheism 11–12, 254, 267–75, 299
population 336, 376n. 10
Port-Royal logic 376n. 6
positivism, logical 4, 53, 381–2, 390, 393n. 17
prayer 272
pride 13, 297, 300n. 4, 326
priority, temporal
probability 62–3, 65, 68, 76, 94n. 9, 115,
119–20, 122–5, 128n. 13, 217, 228–9,
238–40, 246, 254, 258–61, 365
Bayesian 10, 238–40, 333
see also causal reasoning
probable reasoning see causal reasoning
problem of evil 9, 224, 253–5, 258–60,
266, 268–9
inferential 258–60, 277n. 11
logical 258
problem of induction 4, 9, 107, 112, 124,
132, 261, 381
projection
with respect to aesthetics 6, 141–2, 350
with respect to causation 5–6, 131, 134,
137, 141–3, 227, 269
with respect to morality 141–2
with respect to religious belief 269–72
promises 294–6, 304–5, 313–15
proof 228, 242n. 12, 242n. 13
Pyrrhonian scepticism 58, 72, 96n. 33,
150–1, 154, 160, 274
qualia 4, 43–6, 50
rationalism
and belief 41
innatism 46, 50, 289–90, 296
reason
faculty of 62, 64, 65, 70, 71, 79–83, 215,
329–30
nature of 4, 60, 72–87, 208, 212n. 27,
326, 332
and the passions 8, 14, 82, 117, 204–9, 330
2/9/2012 9:37:49 PM
 INDEX OF TOPICS
and motivation 204–9, 382
regularity theory of causation 5–6, 131,
134, 137–8, 143, 202, 225
relations
natural 105, 107, 114–17, 136–7
philosophical 105
theory of 94n. 9
relative ideas 139, 140–1
religion 9–12, 128n. 23, 227–78, 379
afterlife 117, 271
agnosticism 2, 9, 262
anthropomorphism 11, 84, 250, 252–3,
256, 268–78
argument to design 2, 10, 11, 245–64,
265, 266, 273, 275, 276, 277n. 11
atheism 2, 19, 24, 33, 172, 257–8, 262–3,
276, 311, 386, 388
concealment of scepticism 9, 153, 247,
394n. 31
controversy 21–2, 28
cosmological argument 11, 245, 265,
273, 277n. 11, 386
deism 9, 11, 24, 28, 215, 261–2, 265
Dialogues concerning Natural Religion,
publication of 30–4
and emotions (passions) 11, 268–72
fanaticism and enthusiasm 374
fideism 276
Hume’s own attitude to 1–2, 9, 20, 34–5,
304, 382, 389
hypocrisy of clergy 24
Image of God hypothesis 5, 132,
138, 144
irreligion 2, 9, 19, 384–9
Manicheanism 259
miracles 2, 9–10, 12, 28, 72, 94n. 17,
201, 222, 227–44, 266, 277n. 11, 367,
376n. 8
monkish historians 366
natural belief 191, 227–8, 251–3, 261,
272, 278n. 21
naturalism 11, 265–78
omnipotence 11, 253, 260
omniscience 253
ontological argument 98n. 64, 245
Philo’s U-turn 252, 261–2
polytheism 11–12, 254, 267–75, 299
446
9780826443595_Index of Topics_Final_txt_print.indd 446
prayer 272
problem of evil 9, 224, 253–5, 258–60,
266, 268–9, 277n. 11
and projection 269–72
religious experience 275
superstition 24, 29, 34, 73
resemblance
as a principle of association 5, 41, 105–6,
114–17, 118, 119, 126, 176–7
with respect to sympathy 280
rights, natural 314, 317
rules of art 141
sceptical realism 5–6, 131, 134–5, 137,
138–41, 143–4, 225
scepticism 1, 2, 5, 6, 18, 24, 126–7, 146–54,
209, 222, 311, 322
Academic 150–1
Cartesian 59
mitigated 4, 60, 96n. 33, 150–1
and naturalism 4–5, 18–19, 384–5, 387
Pyrrhonian 58, 72, 96n. 33, 150–1, 154,
160, 274
with respect to personal identity 8, 182,
299–300n. 4
and suspension of belief 158–60, 163
Scholasticism 254–5
causa secundum esse 254–5
causa secundum fieri 254–5
science of man 4–5, 6, 8, 17, 18–19, 73,
127, 132, 151–2, 154, 157, 163, 200,
204, 241n. 2, 322–3, 325–7, 328, 332,
364–5, 373–4, 382–3, 385
Scottishness 1, 3, 30, 217, 303, 311, 332,
344n. 2, 392n. 5, 392n. 6
secret powers (secret springs) 5, 128n. 15,
128n. 22, 139, 143, 200, 203, 369, 375
self, the 7–8, 127, 167–98, 201, 379, 381
bundle theory 7, 173, 182, 189, 191
immateriality of 7, 260
and the passions 300n. 4
role of causation 7, 177, 185
role of memory 176–7
role of resemblance 7, 176–7
scepticism with respect to 8, 182,
299–300n. 4
and sympathy 299n. 4
2/9/2012 9:37:49 PM

self-interest 13–14, 298, 301n. 17, 303–6,
310, 311, 313, 317
self-love see self-interest
sense data (the given) 44–5, 48–9
sentimentalism
aesthetic 141, 350
moral 12–13, 141, 209–10, 279–87, 291–3,
298–9, 301n. 13, 326, 327, 328–9
Separability Principle 61, 171–3
sociability 215, 342, 344
soul see self
specie-flow mechanism 339
state of nature 305, 343
Stoics 15, 154, 200, 201, 211, 211n. 5,
336–7, 347n. 48
substance (substrata) 7–8, 127, 167–80, 187
suicide 4, 23
superstition 24, 29, 34, 73
sympathy and empathy 12, 13, 14, 17,
18, 25, 200, 211, 279, 280, 285, 286,
290–2, 292–6, 298, 303, 307–8, 325–7,
328, 372, 375
see also mirroring
taste, standard of 15, 101n. 109, 145n. 33,
145n. 34, 324, 326, 349–63
tenderness 14, 319, 352
testimony 2, 9–10, 75, 118, 125–6, 128n.
23, 199
epistemology of 228–44, 275, 277n. 11
and history 365, 367–73
reductionism and non-reductionism de-
bate 10, 236–8
tragedy 323, 372
transparency of experience 56n. 28
Treatise and Enquiry, relation between 9,
63, 65, 68, 69, 70, 76, 85, 87, 89, 95n.
21, 103n. 129, 139, 143, 153–4, 223–4,
379, 385, 388
trust 320, 329, 340
447
9780826443595_Index of Topics_Final_txt_print.indd 447
INDEX OF TOPICS
uniformity of nature see Uniformity
Principle
Uniformity Principle 62, 64–5, 67–9, 71,
75–6, 77, 79, 103n. 129, 251,
252
universe, origin of 10, 222
utilitarianism 286, 301n. 12, 310, 311,
312–13, 316, 317, 337
verification principle 381
virtue (and vice) 12, 14–15, 18, 101n. 105,
200–1, 203, 221, 282, 283–5, 286,
288–302, 312, 323–4, 325–6, 328–30,
342, 344, 374–5, 381–2
artificial 12, 290, 294–6, 298, 312, 313
intellectual 292
monkish 13, 284, 296–9
natural 294–6, 312, 313
natural abilities 13, 221, 292, 300n. 8,
300n. 10
and utility and agreeableness 12, 284,
286–7, 288–9, 292–3, 294–5, 297
and voluntariness 221
see also beneficence; benevolence; justice;
pride; tenderness
virtue epistemology 200
vivacity and force 4–5, 38–56, 63, 83,
108–10, 114–17, 123, 126, 206,
280, 299n. 4, 323, 365, 370, 371,
372
as functional role 51–2
and intentional content 4, 43–9, 50, 52
as qualitative feel (qualia) 4, 43–6, 50
as verisimilitude 4, 49–51
well-being 3–4, 9, 12, 14–15, 81, 299,
332–48
will, the 8, 79–82, 209–11
wisdom 72, 121, 373–4
women 3, 4, 25, 319–31
2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 448 2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 449 2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 450 2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 451 2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 452 2/9/2012 9:37:49 PM
9780826443595_Index of Topics_Final_txt_print.indd 453 2/9/2012 9:37:50 PM
9780826443595_Index of Topics_Final_txt_print.indd 454 2/9/2012 9:37:50 PM