Tom Brass Reply To Jairus Banaji

This article was downloaded by: [Jawaharlal Nehru University]
On: 20 June 2015, At: 10:30

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK
The Journal of Peasant

Studies
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/fjps20
Why Unfree Labour is Not

So-Called: The Fictions of
Jairus Banaji
Tom Brass
Published online: 05 Aug 2006.
To cite this article: Tom Brass (2003) Why Unfree Labour is Not So-Called: The
Fictions of Jairus Banaji, The Journal of Peasant Studies, 31:1, 101-136, DOI:
10.1080/0306615031000169143
To link to this article: http://dx.doi.org/10.1080/0306615031000169143
PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all
the information (the Content) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of
the authors, and are not the views of or endorsed by Taylor & Francis.
The accuracy of the Content should not be relied upon and should be
independently verified with primary sources of information. Taylor and
Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in
relation to or arising out of the use of the Content.
Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015
This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-andconditions
Debate
Why Unfree Labour is Not So-Called:

The Fictions of Jairus Banaji
TOM BRASS
In a recent critique of the deproletarianization thesis, which links the

reproduction of unfree labour mainly but not only in Third World
agriculture to class struggle prosecuted by capitalist producers,
Banaji maintains in effect that there is no such thing as unfree
labour. Equating the latter with nineteenth-century liberal ideas
about freedom as consent, he conceptualizes all historical working
arrangements simply as disguised wage-labour that is free, a
theoretically problematic claim first made during the Indian mode of
production debate. Such a view, it is argued here, ignores the fact
that unfree workers get paid and also appear in the labour market,
but not as sellers of their own commodity. Moreover, by abolishing
the free/unfree labour distinction, and adopting instead the view that
all rural workers are simply disguised hired labourers who are
contractually free, Banaji aligns himself with anti-Marxist theory
in general, and neoclassical economic historiography in particular.
The analysis of the commodity constitutes the starting point of the Marxist
system. Bhm-Bawerks criticism is primarily levelled against this
analysis. Rudolf Hilferding, Bhm-Bawerks Criticism of Marx.
A Spanish hacendado in Oaxaca locked his Yaqui serfs up at night in one
room so small they had to sleep on top of each other. With revolution, they
nailed him to the door of his beautiful mansion. The acceptability of debt
peonage to rural workers in late nineteenth-century Mexico, as recounted by
Beals [1932: 306].
Tom Brass formerly lectured in the Social and Political Sciences Faculty at the University of
Cambridge, UK. E-mail address: tom@tombrass.freeserve.co.uk
The Journal of Peasant Studies, Vol.31, No.1, October 2003, pp.101136
ISSN 0306-6150
DOI: 10.1080/0306615031000169143 2003 Taylor & Francis Ltd.
102
THE JOURNAL OF PEASANT STUDIES
INTRODUCTION
Of the many theoretical issues raised in the course of the 1960s/1970s

debate about the mode of production in Indian agriculture, the nature of the
capitalism/coercion link remains not only unresolved but perhaps also the
most contentious.1 The two following decades saw a welcome reengagement by development studies generally, and Marxist approaches in
particular, with what used to be one of their central theoretical concerns: the
issue of unfree labour.2 This re-engagement has, however, not been without
its problems: free labour continues to be equated unproblematically with
capitalism, and conversely, where this is not the case, there is frequently a
tendency to overlook prefiguring arguments concerning the fact of, and
reasons for, a capitalism/unfreedom link. In the 1980s, mine was virtually
the only voice arguing from a Marxist viewpoint that unfree labour was not
only unfree but was and is a relation of production that capitalists find
very acceptable.3 Suddenly, however, it seems as if everyone is not only
writing about the way in which unfree labour is still very much with us, how
contemporary forms of unfreedom are sought out and reproduced by capital,
but also claiming to have discovered this capitalism/unfreedom link in the
first place.4 Some even maintain, incorrectly, that this is what they have
argued all along.5
Others, by contrast, refuse to recognize the existence of a connection
between capitalism and unfree labour. According to this view, the presence of
unfree production relations stemmed from a failure of capitalism to develop,
the assumption being that once the latter process got under way, unfree
workers would everywhere and always be replaced with hired labour that was
free. Between the 1960s and 1990s this interpretation amounted to orthodoxy,
and its adherents included not only most bourgeois economists but also some
varieties of Marxism. Among them were exponents of the semi-feudal thesis
and the articulation of modes of production framework, who interpret unfree
labour as a non- or pre-capitalist relation of production. Both these viewpoints
perceived unfree relations as a sign of economic backwardness, irrational
obstacles to growth, and thus incompatible with capitalism. Such relational
forms were accordingly contrasted to free wage-labour which became
theoretically both a precondition for an economically dynamic commercial
agriculture, and thus evidence for a transition to capitalism. One of those who
subscribed to this epistemology was Jairus Banaji, who participated in the
original debate about the mode of production in Indian agriculture, and has
now tried to resuscitate this argument in the course of criticizing my theory
about deproletarianization, to which this is a reply.6
What follows, therefore, is in an important sense fallout from the mode
of production debate that took place three decades ago, in the form of an
WHY UNFREE LABOUR I S NOT SO CALLED
103
attempt by one of the participants to reassert the validity of his and its
dominant theoretical framework (wage-labour = hired labour = capitalist
relation, therefore paid labour = free wage-labour). Coercion is ubiquitous,
maintains Banaji, and has been applied historically to all known forms of
working arrangement. For this reason, he argues, no distinction has existed
or can exist between free and unfree labour. Because in his opinion I am
unsure as to whether hired workers have been, are or can be unfree, he
accuses me of failing to recognize that the debt bondage relation is nothing
other than free wage-labour . This together with myriad other issues such
as the fact that capital exploits kin workers in the peasant household, that
capitalist regulation operates at the level of the employing class and not the
individual employer, and the significance of the formal/real subsumption
argument escapes my notice, and is missing from the analysis contained
in my book. The latter, according to him, subscribes to the idea of
consent/contract embodied in the nineteenth-century liberal notion of free
labour. Needless to say, the reality is not merely different from the one he
describes, but much rather the opposite of what he claims to be the case.
The many assertions made by Banaji, about my work and his, are quite
simply wrong. Not only does he ascribe to me views that I do not hold, but
he then presents against me the very arguments that I myself have made
previously, the inference being that these are issues about which unlike
him I have nothing to say.7 Conversely, the positions he now criticizes are
found not in my work but his. This is not the first time Banaji has
misunderstood/misrepresented the arguments made by those he challenges,
a fact that emerges clearly both from the retractions contained in the
footnotes of his published articles and the comments by others on the way
he caricatures arguments made by those with whom he disagrees.8 To this
already extensive catalogue of errors are now added the inaccuracies,
distortions and misrepresentations set out below.
As interesting theoretically is the political identity of those Banaji
summons in defence of his argument. The real point of substance between
us his insistence that no difference exists between free and unfree labour,
and that this is a Marxist view is one that he defends by invoking the
authority of anti-Marxist neoclassical economic historiography and
postmodernism. This is not accidental, since both the latter have one thing
in common with Banaji: an attempt to banish the concept of unfree labour,
by claiming that it is nothing other than a form of free wage-labour. The
reason why he finds himself in the company of neoclassical economic
historians such as Steinfeld is not difficult to discern. Hence the focus of
Banaji is on the wage, from which he concludes that all relational forms
involving payment are nothing more than disguised wage-labour, to which
he frequently and contradictorily appends the term free. The result is an
104
essentialist and a-historical approach that in effect collapses both the

relational and systemic differences between ancient society and modern
capitalism.
By contrast, my focus is on the way the distinction between free and
unfree labour informs and shapes class formation and class struggle, a
dynamic whereby deproletarianization is one method of workforce
decomposition/recomposition used historically by capital to restructure its
labour process. Unlike Steinfeld, on whose interpretation of contract Banaji
bases his own view, for me the element of freedom lies not in the fact of the
contract itself, but rather on the ability of workers to exit from these same
relations subsequently that is, the retention/reproduction of the capacity
personally to recommodify their labour-power. Because he says little about
class struggle, Banaji fails to understand fully the central role in the latter
process of unfree labour, capitalist restructuring and deproletarianization. It
is equally clear, moreover, that the all-embracing notion of coercion utilized
by him is one that he does not problematize, nor is it applied by Marx to free
labour. Not only is the free/unfree distinction important, therefore, but
without it a Marxist theorization of a systemically and historically specific
transition from capitalism to socialism remains impossible.
MUCH ADO
To say, as Banaji does, that I am unsure whether wage labourers can be

unfree, and that consequently I ignore or am unaware of the connection
between wage-labour and unfreedom, is palpably nonsensical.9 The
absurdity of this claim is evident from even a cursory glance at what I have
written, since issues such as the wage form itself, the payment of wages to
and receipt of wages by unfree labour feature throughout my book.10 Thus
the focus of the first chapter, dealing with the definition of unfreedom, is
largely on the link between worker indebtedness and the wage form/level,
and how in particular the object of decommodifying labour-power is
(among other things) conceptually to separate the worker from the notion of
a wage and the value he/she produces.11 When considering gender and age
specific forms of unfree labour, a central issue for me is the fact that wages
are paid not to women and/or child labourers but to senior kin.12 Similarly,
my analysis of the methodological difficulties in identifying the existence of
unfree labour simply from official wage data is based precisely on the fact
that undisaggregated wage levels fail to indicate who appropriates (and
why) from the workers concerned [Brass, 1999: 247].
The same is true of the chapters presenting case-study materials from
Peru, Bihar, Punjab, and Haryana, where I document ad nauseam instances
of wages received by unfree workers [Brass, 1999: chapters 2, 3, 4, 6, and
105
7]. How, otherwise, would it be possible to characterize unfree labour in

these (and other) contexts as cheap, without addressing the issues of wages
and wage differentials between it and free labour?13 Moreover, how can I be
accused of failing to take into account unfree wage-labour when one of the
forms of unfreedom (= reverse bondage) considered by me is precisely the
withholding of wages, or the keeping back of payment owed by employers
to workers, a practice whose object is to prevent them from leaving and
simultaneously to compel them to borrow for subsistence purposes?14 Each
case study in my book makes it clear that wages are paid to all those whom
I classify as unfree labour: indeed, in extending the concept of unfreedom
the way I do, to cover non-permanent forms of agricultural worker (migrant,
seasonal, daily) who are not able personally to commodify their labourpower, it would be impossible not to address the issue of unfree wagelabour. So much for Banajis claim about my being unsure as to whether
wage-labourers can be unfree.
The fact that many of the arguments made now by Banaji have been
made before, not by him but by me, does not prevent him from attempting
to convey the opposite impression. Hence the claims that unfreedom is
compatible with capitalism, that exponents of the semi-feudal thesis are
wrong for not recognizing this, and that debt-servitude is a method of
controlling workers, all made by Banaji as though no one else had realized
them, are not only central to my own case about deproletarianization (see
below) but also ones that I have made repeatedly over the years in
everything that I have written on the subject.15 When observing that a
majority of Marxists are probably still reluctant to abandon the comforting
idea that slavery precludes capitalism because capitalism is founded on
free labour, therefore, Banaji manages to convey a doubly misleading
impression [Banaji, 2003: 81]. First, that this is a view with which I
disagree (I dont), and second, that it is his view (it isnt: see below). The
same applies to his assertion about the cash advances made by employers to
workers not being interest-bearing debt but rather a method of keeping
wages low, a view that is in fact no different from mine.16
Equally odd is his inference that kinship is a dimension missing from my
analysis of the way in which capital exploits rural workers. When Banaji
states that the liberal conception of capitalism which sees the sole basis of
accumulation in the individual wage-earner conceived as a free labourer
obliterates a great deal of capitalist history, erasing the contribution of
[family] units of labour-power, therefore, he creates a doubly inaccurate
impression: namely, that he himself addresses the class position of kin
labour within the peasant household whilst I by contrast fail to do this
[Banaji, 2003: 82]. Needless to say, the situation is not as Banaji describes:
it is he, not I, who has failed hitherto to problematize both the class position
106
of and the control exercised over kin labour within the peasant household.
Again, the connection between kinship and class (= kinship-as-class) has
been central to all I have written about unfree labour, my argument being
that one of the most effective ways for capital to enforce debt-servicing
labour obligations is through the formal/informal control embodied in
actual/fictive kin ideology exercised through kin networks. Not only do the
latter correspond to class relationships, therefore, but such pressure can be
sustained even when workers migrate and where unfree labour is illegal.
As strange is his opinion that I subscribe to a nineteenth-century liberal
notion of consent embodied in the free labour contract. About this it is
possible to make the following observations. To begin with, Banaji not only
misrepresents my views but does so by placing them in the wrong category.
Marx, Banaji accepts, uses the term in two specific ways: free labour is
both defined historically and contested ideologically [Banaji, 2003: 74].
Put slightly differently, for Marx free labour is a relation that for a variety
of non-free workers (slaves, serfs, etc.) is an achievement historically, but
an achievement which nevertheless does not signal the end of struggle,
merely a new phase: within capitalism, and this time for socialism. These
two distinct and contrasting aspects of freedom are free labour as the apogee
and thus the legitimation of capitalism, and free labour as both the outcome
of class struggle under capitalism and simultaneously a precursor and
necessary condition of a transition to socialism. Whereas the first meaning
corresponds to consent as understood by bourgeois economics, the second
by contrast informs Marxist theory about continuation of class struggle in
order to transcend capitalism. Whereas Marxism and liberalism both
privilege free labour, therefore, each draws a very different conclusion from
its presence. Rather than connecting my arguments to the second of these
two meanings where they belong he wrongly links them to the first.
The same is true of consent: here, too, Banaji is simply wrong to
maintain that I perceive free labour as a voluntary contract.17 Much rather
the opposite, since I equate contract with unfree relations (e.g., recruitment
by labour contractors) and make clear that it signals the presence of a
production relation which deprives the labourer of consent. Contrary to his
assertion, I do in fact problematize the liberal interpretation of consent:
indeed, how could I not, given that my critique is aimed specifically at
neoclassical economic historiography?18 Neither is it correct that I perceive
the worker simply as an individual (= nineteenth-century liberal view) as
distinct from class: once again, the exact opposite is the case, since like
Marx I maintain that the worker is an individual who is also and
simultaneously a member of a particular class.19 My argument is that one
important object of deproletarianization is precisely the attempt by
employers to shift the identity of the worker from one (= worker-as-part-of-
107
a-collective) to the other (= worker-as-individual). Nor is it true that I

conceive capitalism entirely from standpoint of individual capital, thereby
ignoring the fact that the logic that regulates capitalist economy is,
necessarily, that of total social capital. Indeed it is, and I make just this
point: hence my argument concerns the need of capitalists as a class to
control workers as a class. This is exactly the reason why I take issue with
neoclassical economic historians who question the feasibility of interemployer collusion i.e., an agreement among employers not to poach one
anothers workers where clearly this would be contrary to the interests of
individual capitalists. Against this, I have argued that they refrain from
doing so precisely in order to maintain the unity of capitalists as a class, and
stop wages rising.20
Significantly, Banaji remains silent about the fact that many of the
positions he now criticizes are found not in my work, as he maintains, but
rather in his. One would not guess, for example, that a free labour contract
is a concept to which he himself subscribes (see below). Nor is there
mention of the fact that this nineteenth-century liberal/individualist notion
of wage labour is one to which he also adheres. First, when describing the
specifically capitalist characteristic of what he terms disguised wage
labourers as being that they were not free to dispose of their crop as they
chose [Banaji, 1978: 387]. And second, when observing that J B Smith,
orthodox Free Trader and president of the Manchester Chamber of
Commerce argued that the Indian peasant was not a free agent but in thrall
to moneylenders, a view that according to Banaji was fundamentally
correct [Banaji, 1978: 414]. He also fails to point out that the case about the
epistemological link between liberalism and unfree labour has been made
earlier, and better, by Kerr [1997] and Rao [1999].
The attempt to present arguments I have made as though I have not made
them remains wholly bizarre and until linked to both the analogous
exclusion by Banaji of my contributions from the debate about formal/real
subsumption, and his correspondingly partial account of the latter
inexplicable.21 Hence the assertion that between 1979 and 1990 [o]ver a
decade passes without any substantial theoretical discussion of these issues
is not just incorrect, but one that he must know to be incorrect.22 In making
this claim, therefore, Banaji manages to overlook the numerous
contributions by me to this discussion that appeared throughout the 1980s.23
All the latter used a Marxist analytical framework to explain the
acceptability to capitalism of unfree labour, arguing that far from attempting
to eliminate unfree relations that were invariably characterized in the
literature about economic development as pre-capitalist archaic remnants,
capitalist producers in many Third World countries actually preferred an
unfree workforce, because it was easier to control and thus cheaper.24 For
108
this reason, I argued, the incidence of bonded labour was increasing, and
thus much greater than generally supposed.
That Banajis account involves not just omission but also commission is
evident from his observations about the work of Mundle, which manages
simultaneously to combine silence about my earlier defence of the latter
with claims about his own non-existent and/or problematic support for the
same work. In defending Mundle against the criticisms made by
Ramachandran in 1990, therefore, Banaji conveys the impression that until
his own principled championing of Mundles work, no one apart from him
had noticed it, let alone defended it against critics. Such a claim fails to
mention that I, too, have defended the views of Mundle, both earlier and
against exponents of the semi-feudal thesis.25 Also omitted is the fact that,
until now, Banaji has made no mention of Mundle, let alone defended the
latter against the criticisms made of his work. It is clear, moreover, that his
defence of Mundle is nothing more than self-endorsement, since he claims
that Mundles work is based on his own.26 For Banaji, the real missing hero
of this debate is not Mundle whose work on bonded labour is reduced to
a subsidiary role, its author depicted as a ventriloquist dummy merely
relaying the words of another but Banaji himself.27 The inaccurate, not to
say condescending, nature of this claim is evident from a number of things.
First, the fact that the text which Banaji identifies as the source of Mundles
arguments about bonded labour has nothing to do with unfree labour, a point
its author has himself conceded.28 Second, and linked to the foregoing, the
analysis by Mundle of the connection between unfree labour and the process
of formal/real subsumption is obviously based on Marx, and makes no
mention of Banaji.29 Third, when Mundle cites Banaji he also cites the
unpublished thesis by Arvind Das as an equally influential source, a costarring role about which Banaji remains curiously silent.30 And fourth,
unlike Banaji, Mundle does not regard bonded labour in relational terms
simply as another form of hired labour, nor does he subscribe to Banajis
problematic view that no difference exists between free and unfree labour.31
Why Deproletarianization Matters
On the face of it, it is difficult to understand why Banaji should attempt both
to misrepresent my views about unfree labour and to exclude them from the
debate about this subject. Over the years I have argued at length and in
many different publications that in the course of class struggle capital
restructures its workforce by replacing free labour with unfree equivalents,
a process of workforce decomposition/recomposition I term
deproletarianization. The latter entails both economic and ideological
decommodification, since preventing workers from selling their labourpower also conceptually devalorizes the wage form.32 For this reason, the
109
free/unfree distinction is not only important, but its significance extends

beyond the usual restriction of this transformation to a particular moment
of historical transition from feudalism to capitalism and locates it not as
a moment but as a process within capitalism itself. Theorized thus,
unfreedom becomes a crucial and twofold part of the class struggle waged
by capital to ensure the reproduction of the capitalist system. On the one
hand, therefore, workforce restructuring involving the replacement of free
workers by unfree equivalents, or the conversion of the former into the
latter, enables capital to combat economic crisis by cheapening the cost of
labour-power. On the other, this same process of workforce
decomposition/recomposition permits capital to restructure class relations
in its labour process, the object of deproletarianization being either to
undermine or to pre-empt working class consciousness/solidarity and
agency based on this.33 To both these ends, capital decomposes the
proletariat and similtaneously recomposes its workforce as unfree labour
with or without wages and contract.
The historical and contemporary significance of deproletarianization,
and its implications for the direction taken by class struggle, as capitalism
becomes global, is not difficult to discern. As Marx himself pointed out, it
is precisely in the existence of a space between unfree and free labour
dismissed by Banaji as formalism and the desirability not just of a
transition from one to the other but also the reproduction of proletarian
selfhood once achieved, that workers are afforded the opportunity and
possibility of a shift in and retention of a specifically working class
consciousness, through the realization of the fact of the appropriation from
them of value they produce.34 Hence the more perceptive observers note that
in the UK and Germany during the 1930s capitalist crisis the replacement of
free wage-labour with unfree workers was designed both to lower wage
costs by restructuring the capitalist labour process and simultaneously to
weaken trade union organization and break down working class
consciousness/solidarity.35 The freedom secured by foreign migrants in
Germany at the end of the First World War was short-lived, and unfree
labour which re-emerged at the beginning of the 1930s increased from
virtually nothing to around 43 per cent of the workforce during the mid1940s.36 Fascism merely represented the culmination of and continuity with
the more general process of capitalist response to crisis, and an important
reason for the deployment of unfree labour (low-paid foreign workers,
concentration camp prisoners) was to undermine working class resistance to
restrictions on mass consumption, thereby permitting capital investment in
rearmament [Roth, 1997]. In a context where full employment might
otherwise generate pay increases, deproletarianization enabled capital to
maintain wages at the level of the 1930s Depression.
110
Much the same is true of late twentieth-century capitalism in the socalled Third World, where a consequence of the rising incidence of
landlessness is that the labour force is increasingly composed of workers
who sell not the product of their labour but rather labour-power itself.
Accordingly, unfree labour of one sort or another can be introduced or
reintroduced by industrial capitalists, agribusiness enterprises and/or rich
peasants, all of whom employ it in preference to existing free wage-labour.
This has occurred not only in the United States (prison labour and peonage
in the American South, migrant farm workers from Mexico and Central
America, H-2 contract labour from Jamaica employed in the Florida
sugarcane industry, workfare schemes, sweatshops) and South Africa (the
apartheid system), but also in India (debt bondage) and Latin America (the
enganche system).37 Where such workers, who earlier have personally
commodified their own labour-power, subsequently become unfree (their
labour-power either ceasing to be a commodity, or being commodified by
someone other than themselves), what has occurred is a relational
transformation that corresponds to deproletarianization. These workers are
still landless, they still work for someone else on a permanent, seasonal or
casual basis, they can still be employed in conjunction with advanced
productive forces, they still receive cash wages, they may be migrants or
work locally, and (under the control of a contractor, or in the form of
changing masters) their labour-power can still circulate in the labour
market, but and this is the crux of the issue they are no longer able
personally to sell their own labour-power. This is a theoretical position that
has remained constant for the past two decades, and has been defended by
me against numerous critiques by neoclassical economic historians,
postmodernists, and exponents of the semi-feudal thesis.38
ABOUT NOTHING?
Rather than attempting an in-depth examination, both of my arguments

about unfree labour and the data presented in support of them, Banaji
merely restates his own (much criticized) position developed some three
decades ago.39 There he attempted to differentiate what he claimed was the
authentic meaning of wage-labour, the one that Marx used, from those
definitions he dismissed as inauthentic and vulgar marxist. Criticizing
Dobb for arguing that the mode of production is deducible from relations of
production, Banaji accused him of equating coercion simply with feudalism
and serfdom, the inference being that consequently Dobb was unable to
make the epistemological connection between unfreedom and accumulation
[Banaji, 1977: 6]. The definition to which Banaji objected then, and objects
still, is the view of wage-labour as a commodity that results from separation
111
from the means of production (= dispossession), a conceptualization he

dismisses as a simple category and thus un-Marxist.40 In its place, he
proposes a concept which in his view not only operates at a deeper level of
abstraction, where wage-labour [is] capital-positing, capital-creating
labour, but is one to which Marx himself adhered.
About these definitions, and their authenticity/inauthenticity in terms
of Marxist theory, the following comments can be made. In support both of
his own conceptualization, and also of his criticisms of what he deems the
incorrect view, Banaji invokes the authority of what Marx himself wrote on
page 463 of the Grundrisse.41 The latter page, however, contains things not
mentioned by Banaji. There Marx states quite clearly that free labour is in
terms of a consciousness of labour as the property of the self an
improvement (= advance) over the way this relation is perceived by unfree
workers.42 On the same (and the following) page, moreover, it is stated
unequivocally that in addition to the separation of the worker from the
means of existence the means of self-preservation of living labour
capacity, a capacity of capital to reproduce both the existing and surplus
living labour capacity, and value creation a workforce composed of
wage-labour is premised on the existence of a free exchange relation and
not on the master-servant relation between it and capital [Marx, 1973:
4634]. Neither is it true as Banaji asserts that Dobb equates compulsion
with pre-capitalist relations of production, since the latter recognized that,
even in metropolitan capitalist contexts (the US, Europe) during the midtwentieth century, capital has imposed unfreedom on labour where market
conditions favoured workers and not employers.43 Similarly, in invoking the
authority of Lenin, Banaji fails to mention that he, too, subscribed to the
free/unfree labour distinction.44
This in turn highlights the problematic notion of class that Banaji
inherits from his earlier theoretical flirtation with the semi-feudal thesis and
the colonial mode of production.45 The difficulty lies not with the
perception of the moneylending rich peasant as capitalist but rather with the
way he categorizes those at the other end of the agrarian class structure, the
colonial workforce.46 The latter, according to Banaji, were what he termed
disguised wage-labour, a working arrangement whereby smallholders in
the colonies continued to operate their own labour process, but no longer as
independent units of production. The reproduction of such units was
governed by capital, and when the simple commodity enterprise operated in
this fashion, it no longer had its own laws of motion but was instead a
quasi-enterprise with the specific social function of wage labour [Banaji,
1977: 34]. The price which the producer received for the enforced sale of
his/her crop was now a concealed wage, since disguised wage-labour
worked mainly on its own land and not that of or for others. For Banaji,
112
therefore, rural households are not differentiated internally along class lines
(= peasants-as-workers), composed of family members who in the main
actually sell their personal labour-power to others. Rather, he perceives the
rural household mainly as a unitary form (= peasant-as-cultivator),
composed of family members all of whom receive a hidden wage for
growing crops on their own land.47 In nineteenth-century colonized peasant
nations, therefore, capital integrated peasant economy into commodity
production by a process called forced commercialization, which did not
require dispossession of the direct producer, since s/he continued to work on
his/her own land.48
Why Every Relation is Not Disguised Wage-Labour
The resulting problems Banaji has with the existence of historically specific
components of a working class, and where capitalism is present his
reduction of them all to disguised wage-labour, are fourfold.49 First,
because he perceives rural households as cultivators who are not in the main
dispossessed, and as such cultivate produce for capital, Banaji dismisses the
need for capital either to dispossess producers or to form a labour market.50
Second, he fails to recognize that forced commercialization of peasant
cultivation is not the same as bonded labour.51 Whereas debt bondage entails
the appropriation by capital from indebted workers of surplus-value, forced
commercialization can involve the appropriation of surplus product by
agribusiness from what are in fact other small agrarian capitalists.52 Third,
notwithstanding that the disguised wage-labour is that of smallholding
peasants, such cultivators are termed by him a proletariat, despite the fact
that they are separated neither from the control of a particular employer
they are unfree as a consequence of being indebted to monied capitalists
nor from their means of labour (land).53 And fourth, because he is unsure as
to whether or not debt bondage is actually a capitalist relation, Banajis
theory is unable to account for the fact that rich peasant households can and
do become accumulators of capital using intra-kin coercion.54 Rather than
originating externally and affecting all household members uniformly,
therefore, such compulsion operates from within the peasant household
itself, and affects only the landless members of these same units.
The difficulties to which his kind of theorization (= a proletariat
composed of peasants) give rise are evident from Banajis confusion about
the historical development of capitalism in India and its connection with the
expansion of the working class, a lack of clarity shared by Breman.55 Having
proclaimed confidently that a rapid growth of [Indias] working class
during the First World war and its aftermath was co-terminous with an
expansion of national capital, Banaji subsequently admitted that a
substantial modern proletariat would emerge only after Independence, with
113
the new period of capital expansion that started in the [nineteen] fifties.56
Significantly, the concept of peasant-as-cultivator is also a legacy of the
semi-feudal thesis: for exponents of the latter, therefore, although capital
reproduces the peasant household economically, it nevertheless remains
external to it. Another legacy, shared by semi-feudalism and the colonial
mode of production alike, is the view of the colonial state as the main
exploiter of the Indian peasant, albeit mediated by monied capitalists
[Banaji, 1978: 364, 379].
What all this suggests is the following. Disclaimers notwithstanding,
therefore, that Banaji has still not broken completely with the main
assumption of the semi-feudal thesis to which he originally adhered, in that
his disguised wage-labourer remains a cultivator working on his/her own
land, albeit at the behest of capital. During the course of the earlier mode of
production debate, Banaji changed his mind not once but twice. Having
been an enthusiastic supporter of Utsa Patnaik (then as now an exponent of
the semi-feudal thesis), he then changed his mind and became an opponent
of her views.57 The same is true of his initial endorsement of the concept
colonial mode of production, which he also jettisoned subsequently, but
this time rather more surreptitiously.58 In reality, the theoretical gap between
on the one hand Banajis smallholder as a disguised wage-labourer, and on
the other Patnaiks pauperized peasant is vanishingly small. An apparent
distinction does, however, exist. Exponents of the semi-feudal thesis define
systemic differences in terms of production relations: bonded labour
therefore feudalism, wage-labour therefore capitalism. Banaji maintains the
opposite, and defines production relations by means of the economic
system: if capitalism, then all relational forms necessarily correspond to
wage-labour, which itself designates the workforce as a proletariat. In the
end, both these approaches are not only the same, in that the wage form
signals the presence of capitalism, but break with Marx, who maintained
that capitalism could and would utilize a variety of relational forms for the
purpose of accumulation, a position consistent with my own view about
deproletarianization, or accumulation on the basis of unfree labour.59
In many ways, Banajis epistemology is the mirror image of mine:
whereas I argue that in the course of class struggle capital attempts to
replace free labour with workers that are unfree, he by contrast maintains
that the free/unfree distinction is non-existent, and that all workers in receipt
of a wage are simply hired labour. What Banaji has done, in short, is to
fetishize the wage, the presence or absence of which for him determines the
nature of every and any historical and contemporary relation. Not
surprisingly, this unambiguously procrustean theoretical procedure results
in a certain amount of confusion: having accepted that capital can and does
use a variety of relational forms (my argument), he immediately qualifies
114
this, then contradicts himself in order to reassert the validity of his claim
that all relational forms are nothing more than variants of (disguised)
wage-labour.60 Both the epistemological tentativeness of this procedure, and
the recognition that it breaks with Marxist theory, are evident from the raft
of qualifications accompanying the claim (it would surely represent an
advance in Marxist theory to think of capitalism working through a
multiplicity of forms of exploitation based on wage labour original
emphasis).
Having essentialized hired labour in this manner, Banaji no longer has to
probe beneath the surface existence of any other relational form.61 Rather
than problematize unfree labour, he undertakes what might be termed a
semantic readjustment, asserting that wherever it occurs coercion is
simply evidence for the presence of free wage-labour. The kinds of
contradiction to which this gives rise are evident from the following
juxtaposition encountered in his recent book. At the outset, he proclaims
confidently both that no evidence survives as to the kind of contracts
through which these rural labourers were attached to their employers,
that binding peasants to the soil was unnecessary and unconnected with
agricultural production, and that workers employed in this were free
[Banaji, 2001b: 5, 90, 93]. Subsequently, however, it transpires not only that
big landlords wanted lifelong control over [labour with] skills which may
have been in short supply, that the reason why landowners felt it was
necessary to bind workers through a [contractual] clause of this type may
well have been theirmobility between several jobs (different estates) in
the course of the year, that [i]t seems likely that these were bonds which
employers used to tie labour down or strengthen their own control over its
exertion a common feature of rural labour markets, but also that [d]ebt
was the essential means by which employers enforced control over the
supply of labour, fragmenting the solidarity of workers and personalizing
relations between owners and employees.62 All the latter statements might
have been written by me, since they make exactly my point: the object of
debt bondage, in other words, is precisely to curtail the freedom of the
worker to change jobs that is, to sell his/her labour-power elsewhere. Try
as he might, Banaji cannot in the end deny that what he essentializes as
free wage-labourers were unfree workers.63
Not the least of the many difficulties that result from this failure to
problematize unfreedom is his mistaken assumption that any/every kind of
labour-power that receives wages must as a consequence be free, thereby
failing to realize that unfree labourers also get paid and appear in the
market, but not as sellers of their own commodity. In doing this, Banaji
ignores a number of crucial issues. First, Marx did indeed say that capital
used previous relational forms in the accumulation process, but he did not
115
say that these forms became free wage-labour as a result. Plantations in the
southern US and the Caribbean were indeed capitalist enterprises, but the
workforce deployed on them was not composed of wage labour,
disguised or otherwise. Historically and contemporaneously, therefore,
many different kinds of worker slaves, indentured labourers,
sharecroppers, labour-service tenants, convicts, smallholders have been
and are paid a wage, which of itself indicates neither that the workers
concerned are the actual recipients of the payment made, nor that they are
free, and become thereby a proletariat. Second, the fact that Marxism
defines class not in terms of the presence/absence of a wage but rather in
terms of relations of production, an important aspect of which is whether a
worker is able personally to commodify his/her own labour-power. And
third, where his argument differs from mine, it is in many respects no
different from that advanced not by Marxists but rather by anti-Marxists,
such as neoclassical economic historians and postmodernists. This
contention can be demonstrated easily, by examining the politically and
theoretically contentious sources he invokes in support of his case, a
problem which can in turn be attributed to an understandable (and
forgiveable) lack of familiarity on Banajis part with the debate about unfree
labour.
W ITH FRIENDS LIKE THESE
Not the least significant difficulty faced by Banaji is a failure to question the
either the theory or the methodology of the sources he invokes in support of
his claims about wage-labour. Consequently, he accepts the arguments of
those such as Steinfeld, Breman, McCreery and Bauer at their face value,
without asking about the way in which these are constructed and what
theoretical assumptions inform and what political discourse accompanies
their epistemology.64 Although like Banaji they misinterpret as free what is
actually unfree labour, it is their reasons for doing so that are of particular
interest. Theorizing unfreedom as a form of equal exchange, or a benign
(and thus tension-free) arrangement to the benefit of all parties to the
relation, non-Marxists like Breman, anti-Marxists like Bauer, and
postmodernists like McCreery, maintain that others wrongly interpret as
unfreedom what in their opinion is the voluntary entry by rural labour into
reciprocal and essentially desirable work relationships which provide the
labourer concerned with economic security in the form of a subsistence
guarantee or employment insurance.65 The debt component of indenture,
peonage, contract or bonded labour relations is accordingly recast by them
as evidence not of the coercive power exercised by employers but much
rather of the enhanced bargaining power exercised by workers.66
116
Much the same views are expressed by neoclassical economic

historians, for whom silence on the part of the worker is interpreted as
assent: having banished coercion, neoclassical economics is left with a form
of market essentialism, whereby all worker/employer relationships
necessarily involve harmonious and voluntary exchanges between choicemaking subjects.67 For exponents of this view, slaves were attracted to and
willingly remained on the antebellum plantation, because in the latter
context they enjoyed cultural autonomy, a high standard of living, usufruct
rights to land, good incomes, and were not maltreated or overworked, nor
were their families split up and sold.68 The antebellum plantation can now
be doubly redefined: as a non-coercive and thus a capitalist unit, employing
a workforce that corresponded in everything but name to free labour.69 In
this way, slavery is recast as empowering, and both the plantation system
and capitalism itself are rescued from the taint of compulsion and
simultaneously declared economically efficient, the subtext being that
systemically it cannot be transcended hence the need and desire for
socialism are eliminated. Not only is class and class struggle abolished by
the neoclassical concept of an employer/worker equilibrium, but wagelabour becomes like capital (= saving) itself an ever-present (and thus
natural) category, a pan-historical form that underlies all known working
relations/arrangements, and now features in the economic transactions of
the trinity formula merely as one factor of production.70 All these points
have been made before, by those opposed to Marxism: what is unusual,
therefore, is not that they are made again, but that they are made now by
someone who claims to be not just a Marxist but a Trotskyist, and moreover
one who cites neoclassical economic historians in support of his case.
That bourgeois economists seek to defend the capital/labour relation by
dehistoricizing it is unsurprising: what does require explanation, however,
is why Banaji uncritically replicates so many of their views. Not only did
his earlier interpretation of feudalism lean rather heavily on that of Kula,
whose Chayanovian and marginalist assumptions were conveniently
overlooked, and like him privilege consumption and demographic variables
over production and class, but Banajis a-historical concept of (disguised)
wage-labour is not unlike that deployed by neoclassical economic theory.71
Hence the enthusiastic citation by Banaji of Steinfeld, another anti-Marxist
exponent of neoclassical economic historiography.72 The extent of overlap
between their arguments about wage-labour and unfreedom is both
remarkable and significant. Like Banaji, Steinfeld maintains that, as during
the nineteenth century neither wage labour nor contract labour had a fixed
content, wage-labour was essentially no different from contract labour, and
consequently no real difference existed between wage (= free) labour and
contract (= unfree) labour.73 Again like Banaji, Steinfeld both complains that
117
wage-labour is not viewed as unfree, and erases the difference between free
and unfree labour on the grounds that all employment is to some degree
regulated.74 Steinfeld then goes on to argue that the free/unfree labour
polarity is a false (= non-existent) dichotomy, since both free and slave
labour possess the same ability to choose to work.75 Not only does this
difference-dissolving claim overlook abundant evidence to the contrary, but
the categorization of both free and unfree labour as choice-making is
redolent of the concept subjective preferences at the centre of neoclassical
economic theory. And yet again like Banaji, Steinfeld argues that whether
or not coercion exists is ultimately a question of law (economic coercion is
governed by law).76
Steinfelds argument about labour being free despite the presence of
coercion only works if one accepts his legalistic definition of contract, as
Banaji clearly does.77 Freedom of contract for Steinfeld means quite simply
the ability of workers to enter working arrangements: that they are
subsequently unable to withdraw from them without the consent of their
employer does not affect his meaning.78 In this he is no different from other
neoclassical economists, such as Bardhan, who similarly classify such
relations as free solely because workers may enter them voluntarily.79
Together with Bardhan and Banaji, therefore, Steinfelds assumption is that
workers against whom contractually binding stipulations were enforced
once employed somehow remained free (and retained their identity as a
proletariat), when the latter term covers only their entry into and not the
possibility of exit from the relation in question.80 It is from this
epistemologically inconsistent approach restricting the conceptualization
of the working arrangement to the manner of its inception (= entry into), and
consequently excluding the conditions governing the reproduction of the
whole relation (entry into + exit from) that all their difficulties and
resultant confusions stem.81 Adopting their partial definition, for example, it
would be possible to explain neither the act of self-sale into slavery during
periods of famine nor the process of transformation occurring over time
within the same production relation, whereby an initially free working
arrangement becomes unfree (and vice-versa).82
By contrast, it is this aspect the incapacity of a worker personally to
commodify or recommodify his/her labour-power that for Marxism defines
the relation as unfree, and distinguishes it from those relations where
workers retain this capacity of commodification/recommodification.83 The
assertion by Banaji that Marx did not conceptualize wage-labour as free or
labour-power as a commodity at the level of the individual worker, fails to
survive even the most cursory examination. Marx himself insisted that the
two defining and interrelated characteristics of wage-labour under capitalism
are that it must be free, for which its owner must have and keep the ability
118
personally to commodify his/her labour-power.84 Not only did Engels hold

the same view, but he made it clear that he did so in the very same review
of Capital I that Banaji cites as evidence for the contrary.85 The centrality to
the wage-labour relation both of the freedom on the part of the worker
concerned to sell labour-power, to sell it personally, and to retain this ability
(= the absolute right of commodity owners to dispose of their
commodities), is also evident from what other Marxists have written.86 In
cases where labourers possess this capacity, and indeed exercise it by
entering a relation from which they are unable to exit, such workers have in
Marxist terms been deproletarianized: they were free, but are unfree. Like
Steinfeld, however, Banaji disguises the existence of this problem by
doubly redesignating workers: an initial category wage labour then
undergoes a surreptitious metamorphosis, and becomes free wage labour.
Symptomatically, both Steinfeld and Banaji avoid this difficulty in exactly
the same manner, simply by changing the way in which labour is
characterized, and shifting their description from a worker/employer
relation to the fact (undeniable in itself) that the work done is paid for.
Although at the outset both Steinfeld and Banaji like Fogel and Engerman
dismiss the free/unfree distinction as unimportant, therefore, each
nevetheless subsequently relabels wage/hired labour as free.87
That the wage contract once entered into must by definition be free
(entry into = choice) is accordingly central not to Marxism but to the
anti-Marxist framework of bourgeois economic theory. Hence the view of
Steinfeld internally consistent for a neoclassical economist but from a
Marxist perspective contradictory is that it was by means of contract
enforcement that nineteenth century employers made freer labor markets
work [Steinfeld, 2001: 318]. For exponents of bourgeois economics, such as
Steinfeld, contract is not just voluntary, but a reciprocal exchange agreed
between capital and labour, each of which is perceived as equal.88 It is in this
conceptualization, not that of Marx (or me), that free labour is imbued with
the additional notion of consent.89 Marxists, by contrast, see no such thing,
and argue that the wage relation is not given (= contract-as-agreement)
let alone eternal, reciprocal or equal but rather the object of struggle
(= disagreement), and under capitalism the object of class struggle. This
neoclassical economic historiography denies, since it does not recognize the
existence of class, let alone class struggle, each of which is politically
incompatible with discourse about contract based on reciprocity,
equality and consent.90 Moreover, formal conditions governing contract
say nothing about the reality on the ground, which again focuses on class
and class struggle, or the differential capacity of each party either to
disregard or to enforce contractually binding legal conditions.91
119
Why All Roads Do Not Lead to Rome

Most importantly, the pan-historical category of wage-labour (= contract)
which Banaji shares with neoclassical economic historians such as Steinfeld
together with its attendant confusions and theoretical implications can
now be found in late antiquity, which is precisely where Banaji encounters
it.92 It comes as no surprise that his claims about the prevalence of free
labour in the eastern Mediterranean countryside in the period between the
fourth and seventh century project backwards and simultaneously
replicate the mistaken assumptions of neoclassical economic
historiography. When observing that slaves worked as hired labourers, on
nineteenth-century cotton plantations no less than in ancient society,
therefore, Banaji makes a common methodological error, and assumes that
the sale of labour-power by slaves was no different from that by other wage
workers.93 Like Shlomowitz and other neoclassical economic historians,
however, Banaji fails to distinguish between a free market in labour, and a
free labour market. Whereas the latter involves labour-power that has been
commodified personally by its owner (= hired labour that is free), the
former by contrast incorporates labour-power (= indentured workers,
slaves, bonded labourers) that has been commodified by a different subject,
someone in other words who is not its owner.
Because he denies that the concept of economic freedom is premised on
the concept of labour-power as the property of the worker, Banaji fails to
differentiate the latter from property in the labour-power of the worker,
which by contrast is at the root of economic unfreedom. It was the sale of
labour-power by other than its subject that lay at the basis of the slave trade,
and slaves in the antebellum American South were no less such for being
leased out by their owners to other employers. Long ago, for example,
Joseph Clark Robert made the same kind of mistake, and argued that the
employment of slaves in the antebellum factories of Virginia constituted
evidence both for the growth of free labour and (thus) for the pre-abolition
presence of emancipatory/empowering trends in the south.94 It transpires,
however, that what was categorized by Robert as free Negro labor is
actually composed of plantation slaves leased by their owners to tobacco
manufacturers, the wages paid by the latter going straight back to the
slaveholder.95 In other words, and as de Ste. Croix pointed out with regard
to ancient society, in this situation what appears as free labour masks what
is actually unfree labour.
Not the least of the many ironies is, firstly, that Banajis argument about
the preponderance and economic role of free labour in late antiquity is
vulnerable to precisely the criticism that he himself made against Dobb
some three decades ago. Namely, that because of a failure to understand
120
wage labour Dobb would be compelled to argue that when some of the
most deeply entrenched feudal estates of thirteenth century England often
based their production mainly or entirely on paid labour (wage-labour in
Dobbs sense), specifically capitalist relations of production were
established [Banaji, 1977: 7]. Secondly, having relied on legal sources for
an understanding of modern unfree labour, Banaji then proceeds to criticize
precisely these same sources as a method for accessing unfree labour in late
antiquity.96 And thirdly, having adopted an historically ubiquitous concept of
wage-labour, Banaji criticizes Bhaduri for categorizing peasant debt as
precapitalist, on the grounds that the monetization of such relations is
indicative of their capitalist nature [Banaji, 1978: 410]. On the basis of this
logic one could argue that wherever peasant debt takes a specifically
money form, there too one finds capitalism which is in fact the position
Banaji now appears to hold.
What, finally, are the political implications of locating in late antiquity a
pan-historical category of contract that is free wage-labour? How can
workers fight to become free labour, a proletariat in the full sense of the
term, if that is what they already are and always have been? Conversely, any
struggle undertaken by the relationally distinct components of a rural
workforce (tenants, sharecroppers, artisans, lumpenproletarians) subsumed
by Banaji under the historically ever-present category of free wage labour
is not conducted by a proletariat (class-in-itself) to realize itself as such
(class-for-itself) but rather by diverse elements of a socio-economically
heterogeneous workforce unconnected with working class political
objectives (e.g. tenants wanting to convert usufruct rights into
landownership, rich peasants wanting lower input and labour costs, reduced
taxes, and/or higher crop prices). If there is no longer a concept of class,
other than one that is so all-encompassing as to be meaningless, then the
concept of struggle to realize class identity so as to transcend capitalism is
correspondingly impossible.
CONCLUSION
I am accused by Banaji of that most heinous of intellectual crimes: namely,

of unknowingly adhering to the argument I criticize, in my case nineteenthcentury liberal notions of consent embodied in free labour. Along with the
other issues supposedly missing from my book that I am unsure if wagelabour is unfree, that I overlook the way in which capital exploits kinship
and regulates at the level of class, and that I regard free labour as
voluntary/contract all this suggests is that prior to criticizing a text it is
always a good idea to acquaint oneself with its contents. If I regard free
labour as much more than unfree equivalents, then not only did Marx
121
himself, but so does Banaji (and indeed Steinfeld) when he agrees with me
that capital does indeed utilize unfreedom in order to exercise better control
over labour-power. More importantly, and contrary to what Banaji infers,
my argument is not and never has been that free wage-labour is an end in
itself. What I have said is that conditions where wage-labour is free are
better and perceived by workers themselves as such than where it is not
(a point which both Banaji and Steinfeld accept). This in turn enables
workers to organize in furtherance of their class interests as a proletariat,
in other words which for Marxists is a necessary first step on any road to
socialism. Were this not the case, it would obviously be possible to have
socialism in ancient society where unfree relations of production were
widespread, since the latter situation would pose no obstacle to the
realization of a socialist transition! Given his views about the ubiquitous
and non-specific nature of free wage-labour, one cannot be entirely sure
that this is not what Banaji actually believes.
Not only is his accusation that I adhere to nineteenth-century liberal
notions of freedom and consent untrue, but on closer inspection it turns out
to be a more accurate description of the views held by the accuser himself.
Since he conceptualizes wage-labour as contract, Banaji unwittingly
imbibes from bourgeois legal and economic theory the view that a work
arrangement is defined simply by the act of recruitment (entry into). This
contrasts with Marxist theory, for which a work arrangement is defined by
a process: the reproduction of the relational form (entry into + exit from),
as embodied in the ability of a worker personally to commodify and
recommodify his/her own labour-power. Because it has nothing to do with
Marxism, bourgeois contract theory licenses the claim made about the
free/unfree distinction by two kinds of anti-Marxist theory: a-historical
postmodernism and neoclassical economic historiography. The latter, like
Banaji, abolishes the free/unfree distinction, in order to be able to argue
wrongly that every form of paid labour is and always has been the same,
is based on consent and is therefore free, and thus empowers the subject
concerned. As no work relation is disempowering, it follows that
consequently there is no need (and no desire on the part of the workforce)
for socialism. In effect, Banaji ends up by standing Marx on his head:
whereas Marx argues that the wage relation under capitalism must be free,
Banaji by contrast maintains that wherever the wage relation occurs it must
also be free, and thus historically there too is found capitalism. Whether or
not he recognizes this, Banaji is now in the camp of those who assume the
permanence and immutability of capitalism, its eternal and nontranscendental character.
122

NOTES
1.
2.
3.
4.
5.
6.
7.
8.
The most useful accounts of the mode of production debate remain Rudra et al. [1978] and
Thorner [1982].
Among the more recent monographs by Marxists, or edited collections containing Marxist
analyses, about historical and contemporary forms of unfree labour are de Ste Croix [1981],
Miles [1987], Lichtenstein [1996], Blackburn [1988; 1997], Brass and van der Linden
[1997], and Brass [1999].
This is not to say that other theoretical analyses written from a Marxist viewpoint did not
appear in that decade, those by Miles [1987] and Cohen [1987] being of particular interest.
Such analyses, however, continued to regard the presence of unfree relations in the capitalist
labour process as in some sense anomalous (see the critique by Brass [1988]). Hence the
theoretical framework deployed by Miles was the articulation of modes of production, in
which free labour is equated with capitalism and unfreedom with non-capitalist modes. His
argument was that capitalist producers resorted to coercion only reluctantly, and then in
contexts where labour was scarce.
Where the capitalism/unfreedom link is concerned, therefore, the list of those who have been
unable to distinguish between a personal voyage of discovery and the existence of a welltravelled route includes not just popular journalism (for example, Bales [1999], and Klein
[1999]) but also texts by those with more serious intellectual pretensions (for example, Bush
[2000], and Harriss-White and Gooptu [2000]).
This was the subject of an exchange between Amartya Sen and me in July 2001. Objecting
to my observation [Brass, 2001: 174 note 48] that he had suddenly discovered rather late
in the day that bonded labour had not vanished, Sen insisted that he had written about
bonded labour for a great many years. In my reply to him I pointed out that the latter
justification avoided the issue in question, which was not the fact of having mentioned
bonded labour but what he actually said about it, adding that, until very recently, Sen like
many others had dismissed bonded labour as a residual form, and thus peripheral to the
development debate. I continued by noting that in his initial work on famine and entitlements
[Sen, 1980; 1981], there was no mention of bonded labour as the outcome of famine in India
(a well-documented phenomenon), and it was only much later that the existence of this link
was even acknowledged, and then only in passing. Now, however, unfree production
relations were regarded by Sen [1999] no longer as a residual category but much rather as
common in many developing countries, and as such loomed large in his analysis of
development problems. My final comment was that this was a welcome development indeed,
but nevertheless (as I had said in the review article) a veritable discovery on his part.
The texts in question are Brass [1999] and Banaji [2003]. This reply was submitted to
Historical Materialism, the journal in which the critique by Banaji appeared, but the editors
in effect refused to publish it (thereby replicating a previous episode, outlined in Brass [2000:
13940 note 4], when the right of reply to criticism a usual and unproblematic procedure
in most academic journals was denied). The reason for this was, perhaps, two things in
particular, each of which occasioned some embarrassment to the editors of what claims to be
a Marxist journal. First, their failure to spot that what they had regarded as a Marxist
analysis of wage-labour turned out to be nothing of the sort, being much rather an antiMarxist analysis, premised on neoclassical economic theory. And second, that this fact and
consequently the problematic nature of their original judgement was evident from the case
made in the reply.
Indeed, I am presented with the wholly bizarre situation whereby arguments I have made in
the past, and texts I have cited in support of these arguments, are now quoted back at me by
Banaji purportedly as evidence of my being unaware of these very same arguments/texts!
Enforced retractions scattered throughout the footnotes and/or postscripts of Banajis own
published writings, coupled with serious criticisms about misrepresentation levelled at him
in the writings of others, are difficult to avoid. Having confidently proclaimed that [m]any
of Chattopadhyays criticisms of U Patnaik spring from a failure to understand this position
[= the conceptual relevance of the colonial mode of production] or indeed its theoretical
significance, therefore, Banaji [1972: 122; 1977: 38 note 15; 1978: 418 note 24]
9.
10.
11.
12.
13.
14.
123
subsequently confessed that I would now repudiate completely my critique of

[Chattopadhyay] as badly misdirected (original emphasis), and that [t]he characterization
of the mode of production as colonial [is one] which I myself accepted earlier, much too
hastily, etc., etc. Elsewhere, and in a similar vein, he [Banaji, 1980: 264] rescinds an earlier
opinion, accepting that the latter was subjectivist and a-historical. Hence the notion that
Permanent Revolution formed a historically viable alternative programme [in preindependence India] would have to be drawn out and re-examined much more critically, and
a properly historical Marxist analysis would have to probe much deeper to draw out
phenomena connected with the historical constitution of the [Indian working] class and
with the specific features of this process that were determined by the general evolution of
capitalism in India and by the nature of the sort of capitalist democracy that evolved post1947. (original emphasis). Replying to an accusation by Omvedt [1991: 138] of careless
use of data about actual rural struggles and agrarian relations, exemplified by his
misattribution of a quotation, Banaji is once again compelled to retract his original
attribution, and admits that I was clearly wrong to attribute the phrase lamb in the stomach
of the tiger to [Sharad Patil], while failing to identify his critic [Banaji, 1990: 297; 1995:
240 note 8]. Banaji has also been accused of a caricatured representation of Dobb, Laclau
and others by Wolpe [1980: 31].
This oft-repeated claim [Banaji, 2003: 74, 78] takes on an almost mantric significance.
That Banaji may not actually have read my book is further suggested by the fact that his
references to it see Banaji [2003: notes 28, 44, 94, 95 and 96] encompass all of eight
pages (out of a book of nearly 400 pages), mostly those corresponding to the Index entries
on wages. Clearly, Banaji thinks that this category is by itself sufficient to provide him with
all the information he requires in order to possess a thorough understanding of what the book
is about. This is a position that does not depart from the conclusion he reached
immediately after having received a copy of the book itself, when he complained to me that
I notice that in the Index to the book there is scarcely any reference to wage labour.
Communication from Banaji, 24 June 2001 (emphasis in the original), a point repeated in
Banaji [2003: 74]. Since the Index contains no entry for unfree labour either, by this odd
logic I could be accused of ignoring unfree labour in a book specifically about unfree labour!
See Brass [1999: 12ff., 30ff.]. The absurdity of the claim that I fail to consider the
unfreedom/wage-labour link is evident from the following exerpt [Brass, 1999: 16]: And
just as high wages may not indicate worker advantage or freedom, so low wages do not
necessarily signal the presence of bonded labour. Although low wages for all forms of labour
may be both the effect and the intention behind the operation of the debt bondage
mechanism, it is nevertheless the case that not all low-paid workers are themselves unfree to
sell their own labour-power to the highest bidder. A low wage is an effect of economic
unfreedom, and hence not of itself constitutive of the debt bondage relation.
Brass [1999: 22, 37 notes 31 and 32]. Indeed, one of the sub-headings in the first chapter
(Work Intensity, Wage Levels and Unfreedom) actually specifies the existence of an
unfreedom/wage-labour connection [Brass, 1999: 24]. Clearly, Banaji overlooked this.
Wage differentials between and thus the waged nature of both free and unfree labour
are mentioned throughout the book [Brass, 1999: 62, 789, 856, 8990, 92, 96, 103, 119,
176ff., 199, 2034, 214, 215, 220, 225, 228, 250]. Just why Banaji picks up so many
references equating wage labour with unfreedom in books/articles other than mine, yet
manages to miss the abundant references in my book making precisely this same connection,
is a matter for conjecture.
The connection between withholding wages and unfree labour is outlined clearly, and
unmissably, in Brass [1999: 5, 12, 27, 40, 41ff., 60, 845, 1034, 1245, 126, 129, 140ff.,
159ff., 176, 190, 207, 254, 296]. Exactly the same mechanism is considered by the
neoclassical economic historian Steinfeld [2001: 16], except that for him withholding wages
from a worker does not make the latter unfree. Significantly, this view is shared by Banaji
[2001b: 198], who evades the obvious difficulties it poses for the categorization of such
workers as free by observing merely that these types of coercion are true of labour markets
throughout the world and are scarcely a distinguishing characteristic either of the Roman
world or the late Roman period.
124
15. It is impossible for Banaji to claim that he is unfamiliar with my views about semi-feudalism,
since he described an article by me criticizing Utsa Patnaik, an influential exponent of the
semi-feudal thesis, as not only devastating but enjoyable as well, adding I wonder what she
thought of it (original emphasis). In fact, Banaji liked that critique so much he asked to
include it in a festschrift then being prepared by him for the eminent Indian Trotskyist, A.R.
Desai. These observations were contained in a communication from him dated 24 October
1995, and the article in question was my reply to comments by Patnaik on an earlier text by
me (for this exchange, see Brass [1994], Patnaik [1995], and Brass [1995]).
16. Banaji [2003: 856]. Compare Banaji [2003: 87], the depiction of wages as loans is simply
a device to control labour in conditions where the competition for labour is likely to drive up
the bargaining power and wages of workers, and Banaji [2001b: 199], [the object of] a
consumption loan, [or] the advance payment of wages isto strengthen the employers
control over the employee (the employer becomes a creditor but his essential interest is in
the exaction of labour), with Brass [1999: 12]: The object [of an advance payment or loan]
is to ensure the availability to the creditor-employer of the worker thus indebted, the
ideological decommodification of the wage, and ultimately to reduce the price of labourpower. In ideological terms, therefore, a bonded labourer works to pay off a debt rather
than for a wage.
17. Nowhere do I endorse an interpretation of free wage labour as being based either on contract
or on consent. My references to free wage labour as consent are made precisely in order to
criticize such a view, held by neoclassical economic historians.
18. This is evident from, for example, Brass [1999: 14950]. The issue of consent, as projected
in bourgeois economic discourse about the voluntary nature of work relations, is considered
in more detail below.
19. For the incorrect accusation that I conceive capitalism entirely from the standpoint of
individual capital, see Banaji [2003: 80, original emphasis]. That this assertion is wrong is
evident from the following [Brass, 1997a: 223]: [W]hat is so wonderful about being a free
wage labourer in a neo-liberal capitalist context? What if any benefit accrues to the
working class (as distinct from individual labourers or categories of labour) from freedom
achieved under capitalism? [My view] is that freedom has to be considered on a collective
(= workers-as-a-class) basis under socialism and not simply in individualistic terms (=
workers-as-individuals) under varieties of capitalism. Hence the understandable
scepticismabout the theoretical and political closure implied in both the fact and the
desirability of a transition from unfree to free labour; as long as the latter is confined to (and
defined theoretically in terms of) a capitalist mode of production, the true achievement and
extent of worker emancipation must remain problematic. The expression of a similar view
in Brass [1999: 2645] merely serves to underline the suspicion that Banaji may not actually
have read my book.
20. On employer collusion, see Brass [1999: 20, 34, 93, 121, 138, 21819, 248, 252].
21. This exclusionary approach contrasts with my supportive references to the endorsement by
Banaji of the real/formal subsumption argument, which specifies how in particular
contexts/conjunctures capital utilizes relational forms that are not free wage labour.
However, the erroneous conclusion Banaji draws from this that all relational forms
deployed by capital are by definition wage labour, and thus evidence for proletarianization
is not one shared by me.
22. Banaji [2003: 72]. This period is the one between the publication of Mundle [1979] and
Ramachandran [1990]. Banajis ire at the failure of the latter to mention him is
understandable. Unlike Banaji, however, Ramachandran was aware of my arguments: not
only does he cite my work [Ramachandran, 1990: 238] but he also clearly took account of
criticisms I made of his theoretical framework when this was presented to a seminar in
Cambridge during late 1986.
23. Throughout the 1980s my arguments about the fact of, as well as the reasons for, the
capitalism/unfreedom link appeared in a wide variety of academic journals, connecting the
spread of unfree labour to capitalist development, class struggle and workforce restructuring
in Latin America (Peru), India (Punjab, Haryana) and the Caribbean. These publications are
listed in Brass [1999: xii, 308].
125
24. Hence the process of workforce restructuring, whereby employers either substituted unfree
labour for free equivalents, or converted the latter into the former. That the
capitalism/unfreedom link was presented by me specifically in terms of formal/real
subsumption [Brass, 1986: 567] would have been difficult for Banaji to miss, since he was
then on the advisory board of the journal this one in which this discussion appeared.
25. A defence of Mundle is found in articles by me published during the 1980s (see, for example,
the ones which appeared in vol. 11, no. 3, and vol. 14, no. 1 of this journal). Since these same
points in defence of Mundle are contained in Brass [1999: 111, 130, 169, 170, 172, 247], it
is something of a mystery as to how Banaji could possibly be unaware of them.
26. According to Banaji [2003: 72 note 24], therefore, Mundle [1979: 92ff.] is simply taking up
Banaji [1978].
27. Ironically, Banaji [2003: 72, 78] exhibits the very same condescension towards Mundle of
which he accuses Ramachandran.
28. Banaji accepts that my Deccan peasantry piece [i.e., Banaji, 1978] actually has nothing to
do with the issue of free and unfree labour (personal communication, 24 June 2001).
29. This analysis appears in Mundle [1979: 8296], the source for which is identified rightly
as the Appendix: Results of the Immediate Process of Production [Marx, 1976:
9481084], and not Banaji, whose article is cited at the beginning of the following chapter,
on p. 97.
30. Mundle [1979: 97 note 1], where the reference to Banaji [1978] is also found.
31. See, for example, Mundle [1979: 94], where it is plain that in contrast with Banajis view
the capacity of an agricultural labourer freely to commodify his/her labour-power on a
personal basis is regarded as important.
32. This point was not lost on rural capitalists and their representatives. In the case of late
nineteenth-century Mexico [Beals, 1932: 307], for example, [t]he Governor of Coahuila
reported the farm-workers, who received three to five dollars a month, as being too
proletarian.
33. Although he objects to my applying the term restructuring to the process of replacing more
expensive forms of labour-power with cheaper equivalents, Banaji fails to point out that he
is a recent convert to this very view (see, for example, Banaji [2001a]).
34. Hence the view [Marx, 1973: 4645]: What the free worker sells is always nothing more
than a specific particular measure of force-expenditure [Kraftusserung]; He sells the
particular expenditure of force to a particular capitalist, whom he confronts as an
independent individual. It is clear that this is not his relation to the existence of capital as
capitalNevertheless, everything touching on the individual, real person leaves him a wide
field of choice, of arbitrary will, and hence of formal freedom. In the slave relation, he
belongs to the individual, particular owner, and is his labouring machine. As a totality of
force-expenditure, as labour capacity, he is a thing [Sache] belonging to another, and hence
does not relate as subject to his particular expenditure of force, nor to the act of living labour.
In the serf relation he appears as a moment of property in land itself, is an appendage of the
soil, exactly like draught-cattle. In the slave relation the worker is nothing but a living labourmachine, which therefore has value for others, or rather is a value. The totality of the free
workers labour capacity appears to him as his property, as one of his moments, over which
he, as subject, exercises domination, and which he maintains by expending it (original
emphasis).
35. The freedom of mobility has been taken away from labour to a degree which makes one
question whether one can really still speak of a proletariat such as we have known since the
Industrial Revolution, notes Kuczynski [1939: 42], adding that [o]ne can speak of the
German worker only in a very limited sense as a free-wage worker, free to sell his labour
where he gets the least lowly price for it. The same point was made by Wal Hannington, of
the Unemployed Workers Movement, who rightly identified the role of residential training
camps in the Distressed Areas of 1930s Britain as an attempt by the state to unmake the
working class (= outclassed, de-classing), using the cheap unfree labour of the
unemployed to undermine trade union opposition to wage cuts. This strategy, he noted
[Hannington, 1937: 11213, 114], is the sort of thing that tends to make the men feel that
they are being outclassed; that they no longer belong to the wage-earning class The mental
126
36.
37.
38.
39.
40.
41.
42.
43.
effect of such a system is likely to produce a slave psychology the tendency will be for
them to regard themselves as fortunate if they can get an ordinary job at wages far below
trade union rates We must resist any scheme which has the effect of de-classing the
unemployed. When work is available, whether it be for a private employer or for the
Government, the unemployed should be entitled to receive proper payment for that work and
so be restored to their normal position of wage-earners.
See the important analysis of unfree labour in Germany by Roth [1997].
For these and many, many other instances of capitalist restructuring involving unfree labour,
see Brass [1988:18397; 1994: 26974, notes 33, 34, 35, 38], and also Brass [1999: Chapters
2, 3, 4, 6 and 7]. Confident pronouncements about what the enganche relation is or is not
see, for example, Banaji [2003: 86 note 95] mask the fact that their author has not actually
undertaken fieldwork in rural Latin America. Indeed, it is not at all clear that he has ever
conducted research on unfree labour anywhere, even in India. This absence contrasts with the
approach of Marx himself, who in 1880 published A Workers Enquiry see Marx [circa
1930] setting out guidelines for investigation of labours experience of capital in the
capitalist labour process, plus the kinds of consciousness and agency generated by existing
working conditions in short, a questionnaire designed to elicit the concrete situation of
labour, and how this gives rise to class formation and class struggle.
This theoretical defence against these critiques is set out clearly and unmistakably in Brass
[1999].
Hence the reference by Banaji [2003: 80 note 56] to his earlier argument contained in Banaji
[1977], and the redeployment of the latter in Banaji [2003: 80 passim]. Among those who
have criticized the views to which Banaji adhered during the 1970s are Chattopadhyay
[1978a; 1978b], Wolpe [1980], and Rudra [1990].
Banaji [1977: 67]. Those dismissed as adherents of the inauthentic view include not just
Dobb but also Sweezy.
Not the least of the many oddities characterizing this attempt by Banaji [1977: 67] both to
reconceptualize wage-labour and to attribute the result to Marx is the scattered nature of the
textual evidence invoked. Whereas references to capital-positing, capital-producing labour
are found on page 463 of the Grundrisse, evidence for the concreteness of the resulting
category is sought from a section of the text some 363 pages earlier! Using this methodology,
it would be possible to concoct almost any kind of concept and then to ascribe it to Marx.
Given what else is found on page 463, Banajis resort to such procrustean methods is in a
sense necessary and understandable.
The relevant text [Marx, 1973: 463] is as follows: The recognition [Erkennung] of the
products as its own, and the judgement that its separation from the conditions of its
realization is improper forcibly imposed is an enormous [advance in] awareness
[Bewusstsein], itself the product of the mode of production resting on capital, and as much
the knell to its doom as, with the slaves awareness that he cannot be the property of another,
the existence of slavery becomes a merely artificial, vegetative existence, and ceases to be
able to prevail as the basis of production.
In American mining towns, notes Dobb [1946: 14], it is common for the company to own
practically the whole town, sometimes not excluding the magistrates and police. Where this
occurs, the employee may be prevented from transferring to alternative employment or
refusing to accept the employers contract for fear of losing his home, or in the other case he
will be prevented from spending his wages in the cheapest market by the manner in which
he is paid. Many of the schemes to lessen the labour turnover and tie the worker to a
particular firm, of which much is heard in America, tend to have a similar effect (emphasis
added). Much the same is true of Britain in the period 193945, about which Dobb [1940:
332] observes: The crucial limit seems to befull employment in the labour market to raise
wages to such an extent as to precipitate a sharp shrinkage of surplus-value, and
consequently to change the value both of existing capital and of new investment. So
abhorrent and unnatural does such a situation appear as to cause exceptional measures to be
taken to clip the wings of labour evento curtail the normal working of competitive forces
whenever labour scarcity shows signs of becoming an enduring condition in the labour
market (emphasis added). It is perhaps examples like this that Wolpe had in mind when he
127
accused Banaji of caricaturing Dobbs arguments (see note 8 above).

44. As with Marx (see above), this information is found on those pages of Lenin cited by Banaji.
Thus the resort by capitalist producers to bonded labour is described by Lenin not only as
preventing him from changing his master that is from personally commodifying his/her
own labour power, but also as a situation whereby the producer istied to a definite place
and to a definite exploiter. These references are encountered in Lenin [1960a: 216; 1960b:
484], the same two pages cited in Banaji [1977: 8].
45. References to what is termed the colonial mode of production are found in Banaji [1972;
1973; 1977; and 1978].
46. Banaji [1977: 33, 36]. Class and class struggle were categories absent from the very start.
Symptomatically, when considering the rival approaches of Beidelman and Dumont to the
jajmani relation, widely recognized as the locus classicus of bonded labour in rural India,
Banaji [1970a] does this in terms of hierarchy and status, and fails even to mention class and
class struggle. Nor is the absence of the latter from British social anthropology mentioned by
him [Banaji, 1970b] as being among the causes of its crisis.
47. See, for example, Banaji [1973: 394], where it is stated that [i]n the colonies which
capitalism subjugated [it] did not eradicate tribal modes of production [sic][t]he
populations it encountered consisted largely of peasants and far from uprooting their existing
forms of production through their expropriation and conversion into wage labourers, so as to
lay the foundations for an internal expansion of its own mode of production, capitalism
imparted a certain solidity to those forms and even extended them (original emphasis).
48. See Banaji [1972: 2499; 1977: 35], and Banaji [1978: 3812, 385] where it is stated clearly
that this represented the necessary forms of appearance of capitalist relations in the
conditions of a small production economy where the process of labour remained the process
of the small producer (emphasis added).
49. Having criticized Patnaik for what he dismisses as verbal contortions (proto-feudal rich
peasantry, proto-bourgeois rich peasantry), Banaji [1978: 397, 399, 400, 410]
unknowingly deploys some verbal contortions of his own: namely, semi-wage labour
peasantry and the unintentionally humorous disintegration of an independent middle
peasantry into the ranks of a depressed, capitalistically exploited lower middle peasantry.
50. Banaji [1977: 36]. As always, it is necessary to introduce a caveat, since in the case of
nineteenth century Deccan he maintains [Banaji, 1978: 377, 393, 395, 396, 4012, 404] that
peasant dispossession, land transfers and labour market formation all occurred.
51. Further details about this are contained in Brass [1999: 16, 312 note 10].
52. Hence the forced commercialization of tenant produce by landlords in the Peruvian province
of La Convencin during the late 1950s resulted in the unfreedom not of the tenants
themselves, who were rich peasants, but rather of poor peasant sub-tenants and/or
agricultural labourers employed by these same rich peasants to meet their own labour-service
obligations. This case study is examined in Brass [1999: Chapter 2].
53. For the equation of peasant smallholders and artisans with a proletariat, see Banaji [1978:
3656, 368, 385, 401, 4067]. The kinds of theoretical contortions to which this gives rise
are evident from his palpable confusion in categorizing the georgoi of late antiquity as both
landless workers and substantial peasants. [Banaji, 2001b: 192]. That a proletariat is
composed of the disguised wage labour of peasants, and the resulting idea that all
production relations are capitalist, is firmly anchored in his original conceptualization of the
mode of production as colonial. According to Banaji [1972: 2500], the difference between
the capitalist mode of production in metropolitan and colonial contexts is that, unlike the
latter, in the former accumulation was based on advanced productive forces. In the colonies,
therefore, accumulation took place on the basis of an immense super-exploitation of variable
capital i.e. not machinery but labour-power. There are enormous difficulties with this view.
To begin with, it implies that in metropolitan capitalist context no equivalent process of
super-exploitation of labour-power is to be found, which is wrong. As the examples of
agriculture in the southern US and Germany during the 1930s, and the current economic
importance in Europe and America of the informal sector (= sweatshop industries), all attest,
accumulation in metropolitan contexts also involves super-exploitation. Equally problematic
is the fact that the metropolitan/colonial dichotomy, overlaid as it is with an additional
128
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
mechanized/labour-intensive polarity, creates a space for the argument made by exponents of

the semi-feudal thesis that unfree labour occurs in colonial contexts because it is
incompatible with advanced productive forces, which of course is also wrong.
On the question of whether or not bonded labour is a capitalist relation, Banaji [1978: 3878,
392, 410, original emphasis] offers diametrically opposed views, both in the same text. At
one point, therefore, he states: The peasants whom Norman saw [in 1874] were not bonded
labourers of any varietythe compulsion or element of unfreedom that this assistant
collector saw pertained specifically to the fact that the small peasantswere not free to
dispose of their crop as they chose, so long as they were bound by a capitalist who year after
year paid their subsistence-costs (their wages) and to one extent or another controlled their
means of production. That is to say, in these relationships there was no fixed political and
social relationship of supremacy and subordination of the sort that characterizes the feudal
economies of Europe. Later on, however, we are informed both that rich peasants regularly
utilized permanent farm-labourers on the labour-mortgage system [= bonded labour], and
that the tying of labour to this or that individual capital (landowner) does not in the least
alter the content of the social relation as one of capitalist dominationsuch forms of
bondage are precisely a characteristic of the formal subordination of labour to capital, that is,
of a system in which capital retains its individual capital.
Not the least of the many ironies informing Banajis current argument is his invocation of
Breman in support of his own ideas. Not only is the latter an exponent of the semi-feudal
thesis, but he is also clearly confused on issues of theory. For the different historical
conjunctures at which the pre-capitalist hali system in Gujarat is said by Breman to have
ended, and thus accumulation based on the presence of capitalist production relations to have
begun, see Brass [1997b: 342, 351 note 15]. Moreover, from around 1990 onwards Breman
has himself changed his mind about the nature of the rural workforce, and relabelled as neobonded labour what he previously categorized as a landless proletariat. On this volte face,
see Brass [1997b and 2000].
Cf. Banaji [1980: 214, 264].
As one noted commentator [Thorner, 1982: 1998] on the mode of production debate
observes, [w]here Banaji had earlier intervened as a supporter of Utsa Patnaik, he now
classes heras a practitioner of extreme formalism.
Having noted his earlier enthusiasm for the colonial mode of production, the same
commentator then observes of a subsequent contribution by Banaji to the debate that this
time he also eschews any reference to a colonial mode of production. On this particular volte
face, see Thorner [1982: 1965, 1998].
That Banaji [2003: 80] still shares with exponents of the semi-feudal thesis the view that the
wage form necessarily signals capitalism is evident from his assertion that the real issue of
theory here is whether we can sensibly visualise the accumulation of capital being founded
on unfree labour And the obvious response is, no. The same is clear, for example, from
his claim [Banaji, 1978: 3878] that in mid-nineteenth century Deccan the domination by
monied capital over smallholding peasants did not correspond to bonded labour, because
for him the latter relation is a characteristic of European feudalism.
See Banaji [2003: 823].
The conflation of all relational forms, and the consequent idea that essentially no difference
existed between wage labour and slavery, was one that Engels [1907: 21112] himself
mocked in the following fashion, when it was deployed by Eugen Dhring: And when he
[Dhring] explains that our modern wage slavery is only a somewhat transformed and
ameliorated inheritance of chattel slavery and not to be explained from itself (that is from the
economic laws of modern society) it only signifies that wage slavery, like chattel slavery, is
a form of class domination and class subjection as every child knows, or it is false. So we
might with the same right maintain that slavery is only a milder form of cannibalism, the
established original method of disposing of conquered enemies.
Banaji [2001b: 186 note 108, 191, 204, 205]. The latter are just a few of the many examples
found in Banaji [2001b, Ch. 8] that confirm labour-power was unfree.
The reasons why Banaji subsumes unfree workers under the rubric of free labour, why the
result is not merely not Marxist but anti-Marxist, and in whose political company he finds
129
himself as a consequence, are considered below.

64. For endorsing references to texts by Bauer, Breman and McCreery, see Banaji [1977: 26] and
Banaji [2003: 79 note 46, 87 notes 103 and 105]. Paradoxically, one of the more appropriate
supporters invoked by Banaji [2003: 87ff.] in furtherance of his argument is Sartre, that wellknown commentator on the relational intricacies of the agrarian labour process in Latin
America and India. Sartre, of course, was not well informed either about the empirical data
or the theory concerning the substance of and transformations in the agrarian structure of
Third World countries, points which emerge clearly both from his exchanges with the French
Trotskyist Pierre Naville on the subject of existentialism and Marxism [Sartre, 1947: 65ff.],
and from his introduction to the populist Third Worldist text by Fanon [1963].
65. For details see Brass [1999: 223ff., 260ff]. Even an uncritical analysis of Bremans views
such as that by van der Linden [2003: 249ff.] accepts that these are the arguments he
makes. Significantly, perhaps, Banajis objections to my views first surfaced at a 2001
conference paying homage to Jan Breman, whose confused and mistaken views about unfree
labour I have criticized extensively [Brass, 1994; 1997a; 1997b; 1999; and 2000]. The claim
by Banaji that at the conference his ideas went unchallenged (personal communication, 6
June 2002) is contradicted by the conference report, which states [Kannan and Rutten, 2002:
1982] it was argued during the discussion that Banaji seems partly to ignore the present-day
reality of unfreedom of labour imposed by the capital of multinational corporations; there is
a tendency to control labour through the informalization of the formal sector within the
context of globalization and liberalization. Not only does this raise once again the
problematic issue mentioned above (see note 8), but it also suggests that the critique made at
the conference of Banaji follows along the same lines as that made by me here.
66. See Brass [1999] and also Brass [1997b and 2000]. Like Breman, McCreery has changed his
mind about the nature of bonded labour, but in the opposite direction. Whereas Breman
reintroduced the element of coercion into his analysis of the rural work force in Gujarat (see
note 55 above), McCreerys conversion to postmodernism has resulted in his decoupling
peonage and coercion, and recasting the debt bondage relation as an empowering form of
cultural otherness.
67. This is the much criticized revisionist argument made by Fogel and Engerman [1974], and
Fogel [1989], about antebellum plantation slavery.
68. The assertion by cliometricians that slaves were attracted to the antebellum plantation is in
fact replicated by Banaji [2001b: 188], who maintains that permanent (= unfree) labour was
attracted to the estate [in late antiquity] by the prospect of stable long-term employment.
Like cliometricians, moreover, Banaji provides no evidence for this view, in terms of how
the workers themselves perceived their situation. That claims made by Fogel and Engerman
about the empowering situation of slaves under a benign plantation regime have
subsequently been shown to be false does not prevent other neoclassical economic historians
from continuing to defend them. One of these is Steinfeld [2001: 67, 24 note 50, 317],
whose arguments Banaji finds so appealing.
69. Hence the claims made by Steinfeld [2001: 7, 8] that [t]he dynamics of slavery have begun
to look a little more like the dynamics of wage labor [t]his changed picture of slavery
reveals that slavery shared certain basic dynamics with wage labour[w]e have to give up
the idea that so-called free and coerced labor inhabited completely separate universes and try
to understand both in terms of a common framework.
70. It is impossible to improve on Sweezys [1946: 67] description of the a-historicism
structuring marginalist economics: [T]he particular concept wages, which plays a role in
all modern [bourgeois] economic theoriesis taken from everyday language in which it
signifies the sums of money paid at short intervals by an employer to hired workmen.
Economic theory, however, has emptied out this social content and has redefined the word to
mean product, whether expressed in value or in physical terms, which is imputable to human
activity engaged in a productive process in general. Thus Robinson Crusoe, the selfemployed artisan, and the small peasant proprietor as well as the factory labourer all earn
wages in this sense, though in common parlance, of course, only the last-named is properly
to be regarded as the recipient of wages. In other words, wages becomes a universal
category of economic life (the struggle to overcome scarcity) instead of a category relevant
130
71.
72.
73.
74.
75.
76.
to a particular historical form of society (emphasis added). A similar critique of marginalist

economic theory is found in Bukharin [1927].
The influence of Kula is evident in Banaji [1977]. As a review of Kula [1976] in Capital and
Class 2 (1977, pp. 1478) long ago pointed out, the book is based on microeconomic theory
which is predominantly marginalist (and sometimes explicitly Chayanovian) an analysis
more concerned with the study of markets and circulation than with surplus appropriation
and disposal. The neoclassical economic concept of disutility of labour is echoed in
Chayanovian theory of peasant economy as the drudgery of labour structuring the
labour/consumer balance on the peasant family farm. Even exponents of the semi-feudal
thesis such as Patnaik not only recognize Kulas Chayanovian epistemology but also
criticize Banaji for his procrustean attempt to synthesize Leninist views about the peasantry
with those of Chayanov [Banaji, 1976; Patnaik, 1999: 2 note 1, 63ff.].
For endorsing citations of Steinfeld [1991; 2001], see not only Banaji [2003: 70 notes 2, 5
and 6, 71 note 20, 79 note 52, 88 notes 107 and 111], but also Banaji [2001b: 208 notes 108
and 113]. Ironically, Steinfeld [2001: 13 notes 226] not only attacks Marx for adhering to
the free/unfree distinction, but also for using the same definition as do I (and other Marxists).
It is equally clear that Banajis views are acceptable to those who use a marginalist analysis
to examine employment relations. Thus the influence of Banaji is evident in a confused and
derivative account by Harriss-White and Gooptu [2000: 94ff.] of coercion and debt-bondage
as a form of contract. Not only is the usual theoretical framework of one of the two authors
a neoclassical economic one, but the input to the analysis of Banaji himself is acknowledged
[Harriss-White and Gooptu, 2000: 111]. Most significantly, however, his view about
disguised wage labour is reproduced verbatim: hence the claim [Harriss-White and Gooptu,
2000: 96, emphasis added] that small peasants and landless agricultural labourers [who] are
obliged to work by debt [and] by the coercive power of dominant landowners [are]
effectively reduced to being wage-workers in thin disguise [sic].
Steinfeld and Engerman [1997]. Steinfelds views are no different from those of his coauthor Engerman, whose marginalist economic framework informed the conservative and
highly influential revisionist account of antebellum plantation slavery as noncoercive/benign/mild. See Fogel and Engerman [1974], and the critique of their
neoclassical economic historiography in Brass [1999: 6, 7, 20, 34, 74, 145ff., 163ff., 255ff.,
299ff.]. That Banaji is unaware of this connection and common epistemology, as also of its
political implications for his arguments, suggests once again a lack of familiarity both with
the debate about unfree labour and the contents of my book.
Steinfeld [2001: 12]. In order to arrive at this view, Banaji occasionally engages in what can
only be described as a linguistic sleight of hand. Having noted that a Revenue Department
report of 1879 described the relation between moneylenders and smallholders as one
between masters andtheir slave, he glosses this relation as wage slavery [Banaji, 1978:
3889, 424 note 84].
That the epistemology informing neoclassical economic historiographic view about unfree
labour it is voluntary, and therefore in a sense free finds echoes in the justification of
slave labour advanced by defendants at the 1945 Nuremberg Trial [HM Attorney-General,
1946: 189] may or may not be significant. Hence the view that [w]hilst admitting the
deportation to Germany and the utilization for the war industriesof millions of workers
from the occupied territories, [Plenipotentiary-General for the employment of labour] Saukel
denied the criminal character of this action, affirming that the recruitment of labour was
allegedly carried out on a voluntary basis.
Steinfeld and Engerman [1997]. Significantly, Banaji [2003: 70 note 8, 72 notes 21 and 22]
eschews specifically Marxist analyses of legal theory, such as Renner [1949] and Pashukanis
[1978], preferring instead to base his case about contract on impeccably bourgeois sources
such as the Yale Law Journal and Harvard Law Review. The extent of this influence is
evident from the fact that many of the legal sources and case studies that appear in Banaji
[2003] are found originally in Steinfeld [2001]: thus, for example, Duncan Kennedy, and
Robertson v. Baldwin cited endorsingly in Banaji [2003: 70 note 5, 71 note 19, 72 note 21]
are also encountered in Steinfeld [2001: 22 note 44, 25 note 52, 72 note 113, 189 note 65,
and 270ff.].
131
77. For earlier references to the existence in nineteenth-century India of what are termed
variously wage contracts and labour contracts, see Banaji [1978: 393, 409]. The same is
true of the eastern Mediterranean during late antiquity, where according to Banaji [2001b:
198]: [o]ne can speak of a labour market wherever labour was recuited through contracts,
regardless of the nature of those contracts or the terms and conditions of employment
embodied in or entailed by them, and thus of the coercion and domination of labour which
such agreements may have represented (emphasis added).
78. All these claims are found in the symptomatic series of observations by Steinfeld [2001: 9,
13] that: the criminal enforcement of labor agreements was an integral aspect of the first
blossoming of free contract in labor markets. Freedom of contract implied that workers
should not be constrained to enter only revocable agreements but should be free to bind their
labor irrevocably as well In the case of wage workers under short contracts, criminal
punishment wasnothing more than a contract remedy to enforce certain kinds of voluntary
agreement. This is a view that Banaji [2003: 77] wholly implausibly ascribes to Marx.
79. In making the bizarre claim that I criticize Bardhan a neoclassical economist whose
concept of unfree labour is shared not only by Steinfeld but also by Fogel and Engerman
for regarding labour as choice-making, a view which in Banajis opinion is no different
from mine, the latter typically misrepresents (or misunderstands) my argument. What Banaji
fails to realize, therefore, is that my criticism of Bardhan is much rather that he holds the
opposite view to mine. Hence my objection is that Bardhan regards attached workers as free
because of his belief wrong, in my view that they choose to enter bonded labour relations.
Cf. Banaji [2003: 78 note 44] and Brass [1999: 224].
80. Among the neoclassical economists adhering to the same view of contract are Drze and
Mukherjee [1987], two of the many bourgeois economists whose views Banaji [2001b: 200,
203, 208 notes 63, 88, 108, 113] cites with approval. That Banaji [2001b: 197, 198] separates
the entry into work relationships from a capacity to exit from these subsequently is embodied
in his distinction between the forms in which employers recruited labour and the methods
of control which they used to regulate [workers once recruited] (original emphasis). The
labour market, he then adds, is defined simply by the act of recruitment, or entry into work
relations. Given his insistence on the all-embracing nature of coercion, informing free and
unfree relations alike, it is indeed ironic that Banaji misses the significance for his argument
of the inability of a labourer to exit from a work arrangement s/he entered voluntarily.
81. In claiming that Polish workers in Germany were merely free workers subjectedto
repressive forms of control Banaji [2003: 79] forgets one crucial point. As the sources he
himself cites make clear, although foreign migrants were free when recruited in Poland, once
in Germany they were controlled by means of work permits. The latter remained in the
possession of employers to whom such workers had been allocated, and without them it was
impossible for them to seek alternative and better-paid jobs. In short, Polish workers were
free but became unfree. This relational transformation is one that escapes not only Banaji but
as is outlined in Brass [2002] also Byres, an exponent of the semi-feudal thesis.
82. The historical link between famine conditions and self-sale into slavery, a voluntary act that
does not negate the unfreedom of the subsequent relation, is found not just in ancient Rome
but also in Africa, Russia, Latin America, and India, where specific categories of famine
slave such as ankla bhritta (= a person maintained in famine) exist. For
details/sources, see Brass [1999: 11, 30 note 2] and de Ste. Croix [1981: 169].
83. That this is the case is evident not just from Chapter 6 (The Sale and Purchase of Labourpower) in Capital I [Marx, 1976: 272] but also from the manuscript for A Contribution to
the Critique of Political Economy [Marx; 1988: 378], written in the period 186163. The
latter contains the unambiguous observation [Marx, 1988: 111, 135, emphasis added] that, so
as to confront capital as wage labour, [t]he worker must be free, in order to be able to
dispose of his labour capacity as his property, he must therefore be neither slave, nor serf,
nor bondsman. Equally, he must on the other hand have forfeited the conditions for the
realization of his labour capacity. He must therefore be neither a peasant farming for his own
needs nor a craftsman: he must have altogether ceased to be an owner of property By wage
labour we understand exclusively free labour which is exchanged for capital, is converted
into capital and valorizes capital.
132
84. It is clear from Marx [1976: 271] that to be free, a worker did not just have to have but had
to retain i.e. to be able to recommodify the capacity to sell his/her labour-power: He [=
the free worker] must constantly treat his labour-power as his own property, his own
commodity[in] this way he manages to alienate (verussern) his labour-power and to
avoid renouncing his rights of ownership over it (emphasis added). It is precisely these
aspects the constant ability to recommodify labour-power, and the non-renounciation of
personal ownership rights to the latter that break with bourgeois legal and neoclassical
economic notions of contract.
85. Banaji [2003: 71 note 18] cites Engels [1985a: 255] in support of the contention that wage
labour is not and cannot be free. However, in that same review Engels [1985a: 2445]
citing Marx [1976: 27071] writes as follows: But in order to enable the owner of money
to meet the labour-power as a commodity in the market, several conditions have to be
fulfilledlabour-power can appear as a commodity, in the market, so far only as it is offered
for sale, or sold, by its own owner, the person whose labour-power it is. In order to enable
its owner to sell it as a commodity, he must be the free proprietor of his labour-power, of his
person. He and the owner of money meet in the market, and transact business, as each others
peers, as free and independent owners of commodities, so far different only that the one is
the buyer and the other the seller (emphasis added). For more along the same lines, see
Engels [1985b: 289].
86. Thus, for example, Kautsky [1936: 5960] notes: Labour-power has to appear in the market
as a commodity. What does this mean? the exchange of commodities is based on the
absolute right of commodity owners to dispose of their commodities. The owner of labourpower, the worker, must therefore be a free man, if his labour-power is to become a
commodity. His labour-power must remain a commodity; consequently, he must not sell it
outright, but only for definite periods, else he would become a slave, and be transformed
from a commodity owner into a commodity (emphasis added).
87. See Banaji [2003: 7980] for evidence that what starts out as the relationally non-specific
category of wage labour is subsequently relabelled by him as free. The same procedure is
found in Banaji [2001b].
88. The bourgeois legal theory structuring neoclassical economic claims about the wage relation
as contract being free despite the inability of an indebted worker to exit from it is outlined
in Pawate [1953: 6]: Thus in a contract the assent of the promisor is at the root of the matter
and the apparent fetter or bondage that has sprung from the root of free assent cannot be
anything but an aspect of the promisors own freedom. The promisor would be lacking in
freedom were he not free to bind himself by his own will.
89. That Banaji criticizes others for espousing a discourse about consent when it is in fact he
who subscribes to a theory about contract in which this very same notion of consent is
implicit, requires no further comment.
90. At times Steinfeld [2001: 317, 320, 321] comes close to recognizing this. Without
mentioning class struggle, therefore, he accepts that employers were using forms of legal
compulsion to tie workers to jobs for longer or shorter periods because they hoped to reduce
labor costs by taking these steps Only when American labor became better organized after
the Civil Warwas legislation passed that cut back contract remedies even further In both
England and the United States a broad suffrage, together with a well-organized labor
movement and a set of arguments that resonated widely because they drew on fundamental
values, seem to have been crucial for establishing a contracts policy that imposed strict
limitations on [employers].
91. For an example from nineteenth-century Brazil of the gap between what the law states and
what happens at the grassroots, see da Cunha [1985].
92. The questionable claim made in Banaji [2001b] is that late antiquity not only saw the
emergence of a new aristocratic elite that invested in agricultural production, but and
more problematically that such landlords utilized a workforce composed of hired workers
(free wage labour, in other words).
93. See Banaji [2001b: 217 and note 21].
94. See Robert [1938: 1989]. The teleological link between slave-as-hired labourer and claims
about emancipatory/empowering trends can be gauged from the assertion [Robert, 1938: viii]
133
that [t]he broad privileges granted the factory bondsmen broke with tradition and indicated
an evolutionary movement away from slavery, a movement which, had it not been for the
cataclysmic changes of the 1860s, might have resulted in the peaceful granting of Negro
freedom.
95. Such slaveowners were, unsurprisingly, much in favour of higher wages paid to slaves
leased to factories, since these increases went to them and not the unfree worker employed.
96. Compare Banaji [2001b: viii], where he criticizes what he terms a traditional picture of late
antiquity derived from legal sources, with his obvious enthusiasm for the latter in Banaji [2003].
REFERENCES
Bales, Kevin, 1999, Disposable People: New Slavery in the Global Economy, Berkeley, CA:
University of California Press.
Banaji, Jairus, 1970a, Hierarchy and Power: Reflections of a Blind Materialist, Journal of the
Anthropological Society of Oxford, Vol.I, No.1.
Banaji, Jairus, 1970b, The Crisis of British Social Anthropology, New Left Review, No.64.
Banaji, Jairus, 1972, For a Theory of Colonial Modes of Production, Economic and Political
Weekly, Vol.VII, No.52.
Banaji, Jairus, 1973, Backward Capitalism, Primitive Accumulation and Modes of Production,
Journal of Contemporary Asia, Vol.3, No.3.
Banaji, Jairus, 1976, Chayanov, Kautsky, Lenin: Considerations towards a synthesis, Economic
and Political Weekly, Vol.XI, No.40.
Banaji, Jairus, 1977, Modes of Production in a Materialist Conception of History, Capital &
Class, No.3.
Banaji, Jairus, 1978, Capitalist Domination and the Small Peasantry: Deccan Districts in the Late
Nineteenth Century, in Ashok Rudra et al. [1978].
Banaji, Jairus, 1980, The Comintern and Indian Nationalism, in K.N. Panikkar (ed.) [1980].
Banaji, Jairus, 1990, Illusions About the Peasantry: Karl Kautsky and the Agrarian Question,
The Journal of Peasant Studies, Vol.17, No.2.
Banaji, Jairus, 1995, The Farmers Movements: A Critique of Conservative Rural Coalitions, in
Tom Brass (ed.) [1995].
Banaji, Jairus, 2001a, Workers rights in a new economic order, http://biblio-india.com/articles.
Banaji, Jairus, 2001b, Agrarian Change in Late Antiquity: Gold, Labour, and Aristocratic
Dominance, Oxford: Oxford University Press.
Banaji, Jairus, 2003, The Fictions of Free Labour, Contract, Coercion, and So-called Unfree
Labour, Historical Materialism, Vol.11, No.3.
Beals, Carleton, 1932, Porfirio Daz, Philadelphia and London: J.B. Lippincott and Company.
Blackburn, Robin, 1988, The Overthrow of Colonial Slavery 17761848, London: Verso.
Blackburn, Robin, 1997, The Making of New World Slavery: From the Baroque to the Modern,
14921800, London: Verso.
Brass, Tom, 1986, Unfree Labour and Capitalist Restructuring in the Agrarian Sector: Peru and
India, The Journal of Peasant Studies, Vol.14, No.1.
Brass, Tom, 1988, Slavery Now: Unfree Labour and Modern Capitalism, Slavery & Abolition,
Vol.9, No.2.
Brass, Tom, 1994, Some Observations on Unfree Labour, Capitalist Restructuring, and
Deproletarianization, International Review of Social History, Vol.39, Part 2.
Brass, Tom, 1995, Reply to Utsa Patnaik: If the Cap Fits, International Review of Social
History, Vol.40, No.1.
Brass, Tom, 1997a, Introduction: Free and Unfree Labour The Debate Continues, in Tom
Brass and Marcel van der Linden (eds.) [1997].
Brass, Tom, 1997b, Immobilized Workers, Footloose Theory, The Journal of Peasant Studies,
Vol.24, No.4.
Brass, Tom, 1999, Towards a Comparative Political Economy of Unfree Labour: Case Studies
and Debates, London and Portland, OR: Frank Cass Publishers.
Brass, Tom, 2000, Labour in Post-Colonial India: A Response to Jan Breman, The Journal of
Peasant Studies, Vol.28, No.1.
134
Brass, Tom, 2001, At Their Perfect Command? The Struggle of/(over) Post-Emancipation
Rural Labour, The Journal of Peasant Studies, Vol.28, No.3.
Brass, Tom, 2002, Rural Labour in Agrarian Transitions: The Semi-Feudal Thesis Revisited,
The Journal of Contemporary Asia, Vol.32, No.4.
Brass, Tom (ed.), 1995, New Farmers Movements in India, London and Portland, OR: Frank
Cass Publishers.
Brass, Tom, and Marcel van der Linden (eds.), 1997, Free and Unfree Labour: The Debate
Continues, Bern: Peter Lang AG.
Bukharin, Nicolai, 1927, The Economic Theory of the Leisure Class, New York: International
Publishers.
Bush, M.L., 2000, Servitude in Modern Times, Cambridge: Polity.
Chattopadhyay, Paresh, 1978a, On the Question of the Mode of Production in Indian
Agriculture, in Ashok Rudra et al. [1978].
Chattopadhyay, Paresh, 1978b, Mode of Production in Indian Agriculture: An Anti-Kritik, in
Ashok Rudra et al. [1978].
da Cunha, Manuela Carneiro, 1985, Silences of the Law: Customary Law and Positive Law on
the Manumission of Slaves in 19th century Brazil, History and Anthropology, Vol.1.
de Ste Croix, G.E.M., 1981, The Class Struggle in the Ancient Greek World: From the Archaic
Age to the Arab Conquests, London: Duckworth.
Cohen, Robin, 1987, The New Helots: Migrants in the International Division of Labour,
Aldershot: Gower Publishing Co.
Dobb, Maurice, 1940, Political Economy and Capitalism, London: Routledge & Kegan Paul Ltd.
Dobb, Maurice, 1946, Wages, Cambridge: Cambridge University Press.
Drze, Jean, and Anindita Mukherjee, 1987, Labour Contracts in Rural India: Theories and
Evidence, London: LSE Development Research Programme.
Engels, Frederick, 1907, Landmarks of Scientific Socialism: Anti-Duehring, Chicago: Charles
H. Kerr & Co.
Engels, Frederick, 1985a, Review of Volume One of Capital for The Fortnightly Review,
Marx/Engels Collected Works, Vol.20, Moscow: Progress Publishers.
Engels, Frederick, 1985b, Synopsis of Volume One of Capital by Karl Marx, Marx/Engels
Collected Works, Vol.20, Moscow: Progress Publishers.
Fanon, Franz, 1963, The Wretched of the Earth, New York: Grove Press.
Fogel, Robert William, 1989, Without Consent or Contract: The Rise and Fall of American
Slavery, New York: W.W. Norton.
Fogel, Robert William, and Stanley L. Engerman, 1974, Time on the Cross: Volume I The
Economics of American Negro Slavery, London: Wildwood House.
Hannington, Wal, 1937, The Problem of the Distressed Areas, London: Left Book Club/Victor
Gollancz, Ltd.
Harriss-White, Barbara, and Nandini Gooptu, 2000, Mapping Indias World of Unorganized
Labour, in Leo Panitch and Colin Leys (eds.) [2000].
Hobsbawm, E.J., W. Kula, Ashok Mitra, K.N. Raj, and Ignacy Sachs (eds.), 1980, Peasants in
History: Essays in Honour of Daniel Thorner, Bombay: Oxford University Press.
HM Attorney-General, 1946, The Trial of German Major War Criminals by the International
Military Tribunal at Nuremburg, Germany: Speeches of the Chief Prosecutors, London:
HMSO.
Kannan, K.P., and Mario Rutten, 2002, Labour and Capitalist Transformation in Asia: Report on
a Conference, Economic and Political Weekly, Vol.XXXVII, No.21.
Kautsky, Karl, 1936, The Economic Doctrines of Karl Marx, Hampstead: NCLC Publishing
Society Ltd.
Kerr, Ian, 1997, Free or Unfree? Railway Construction Labour in Nineteenth Century India, in
Tom Brass and Marcel van der Linden (eds.) [1997].
Klein, Naomi, 1999, No Logo: Taking Aim at the Brand Bullies, New York: St. Martins Press.
Kuczynski, Jrgen, 1939, The Condition of the Workers in Great Britain, Germany and the Soviet
Union, 19321938, London: Left Book Club/Victor Gollancz, Ltd.
Kula, Witold, 1976, An Economic Theory of the Feudal System: Towards a Model of the Polish
Economy 15001800, London: NLB.
135
Lenin, V.I., 1960a, What the Friends of the People Are, and How They Fight the SocialDemocrats, Collected Works, Vol.1, Moscow: Foreign Languages Publishing House.
Lenin, V.I., 1960b, The Economic Content of Narodnism, Collected Works, Vol.1, Moscow:
Foreign Languages Publishing House.
Lichtenstein, Alex, 1996, Twice The Work of Free Labor: The Political Economy of Convict
Labor in the New South, London: Verso.
Marx, Karl, c. 1930 [1880], A Workers Enquiry, London: Communist Party of Great Britain.
Marx, Karl, 1973, Grundrisse, Harmondsworth: Penguin Books.
Marx, Karl, 1976, Capital Volume 1, Harmondsworth: Penguin Books.
Marx, Karl, 1981, Capital Volume 3, Harmondsworth: Penguin Books.
Marx, Karl, 1988, Economic Manuscript of 186163, Marx/Engels Collected Works, Vol.30,
Moscow: Progress Publishers.
Miles, Robert, 1987, Capitalism and Unfree Labour: Anomaly or Necessity?, London: Tavistock
Publications.
Mundle, Sudipto, 1979, Backwardness and Bondage: Agrarian Relations in a South Bihar
District, New Delhi: Indian Institute of Public Administration.
Omvedt, Gail, 1991, Theorists of the Peasantry Should at Least Have Their Feet on the Ground:
A Response to Banaji, The Journal of Peasant Studies, Vol.19, No.1.
Panikkar, K.N. (ed.), 1980, National and Left Movements in India, New Delhi: Vikas Publishing
House.
Panitch, Leo, and Colin Leys (eds.), 2000, The Socialist Register 2001, London: The Merlin
Press.
Pashukanis, Evgeny B., 1978, Law and Marxism: A General Theory, London: Ink Links.
Patnaik, Utsa, 1995, On Capitalism and Agrestic Unfreedom, International Review of Social
History, Vol.40, Part1.
Patnaik, Utsa 1999, The Long Transition, New Delhi: Tulika.
Pawate, I.S., 1953, Contract and the Freedom of the Debtor in the Common Law, Bombay: N.M.
Tripathi Ltd.
Ramachandran, V.K., 1990, Wage Labour and Unfreedom in Agriculture: An Indian Case Study,
Oxford: Clarendon.
Rao, J. Mohan, 1999, Freedom, Equality, Property and Bentham: The Debate Over Unfree
Labour, The Journal of Peasant Studies, Vol.27, No.1.
Renner, Karl, 1949, The Institutions of Private Law and Their Social Functions, London:
Routledge & Kegan Paul Ltd.
Robert, Joseph Clarke, 1938, The Tobacco Kingdom: Plantation, Market, and Factory in Virginia
and North Carolina, 18001860, Durham, NC: Duke University Press.
Roth, Karl Heinz, 1997, Unfree Labour in the Area under German Hegemony, 19301945: Some
Historical and Methodological Questions, in Tom Brass and Marcel van der Linden (eds.)
[1997].
Rudra, Ashok, et al., 1978, Studies in the Development of Capitalism in India, Lahore: Vanguard
Books.
Rudra, Ashok, 1990, Indian Capitalism and Marxist Theory, in Gyansham Shah (ed.) [1990].
Sartre, Jean-Paul, 1947, Existentialism, New York: Philosophical Library.
Sen, Amartya, 1980, Famine Mortality: A Study of the Bengal Famine of 1943, in Hobsbawm
et al. (eds.) [1980].
Sen, Amartya, 1981, Poverty and Famines: An Essay on Entitlement and Deprivation, Oxford:
Clarendon Press.
Sen, Amartya, 1999, Development as Freedom, Oxford: Clarendon Press.
Shah, Gyansham (ed.), 1990, Capitalist Development: Critical Essays, Bombay: Sangam Books.
Steinfeld, Robert J. 2001, Coercion, Contract, and Free Labor in the Nineteenth Century,
Cambridge: Cambridge University Press.
Steinfeld, Robert J., and Stanley L. Engerman, 1997, Labor Free or Coerced? A Historical
Reassessment of Differences and Similarities, in Tom Brass and Marcel van der Linden
(eds.) [1997].
Sweezy, Paul M., 1946, The Theory of Capitalist Development: Principles of Marxian Political
Economy, London: Dennis Dobson Limited.
136
Thorner, Alice, 1982, Semi-Feudalism or Capitalism? Contemporary Debate on Classes and

Modes of Production in India, Economic and Political Weekly, Vol.XVII, Nos.4951.
van der Linden, Marcel 2003, Refuting Historys Occidentalism, in A. Das and M. van der
Linden (eds.) [2003].
Das, A., and M. van der Linden (eds.), 2003, Work and Social Change in Asia: Essays in Honour
of Jan Breman, New Delhi: Manohar.
Wolpe, Harold 1980, Introduction to The Articulation of Modes of Production, edited by Harold
Wolpe, London: Routledge & Kegan Paul.

Tom Brass Reply To Jairus Banaji

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tom Brass Reply To Jairus Banaji

Uploaded by

Copyright:

Available Formats

This article was downloaded by: [Jawaharlal Nehru University]

On: 20 June 2015, At: 10:30

The Journal of Peasant

Why Unfree Labour is Not

PLEASE SCROLL DOWN FOR ARTICLE

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

Why Unfree Labour is Not So-Called:

In a recent critique of the deproletarianization thesis, which links the

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

Of the many theoretical issues raised in the course of the 1960s/1970s

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

essentialist and a-historical approach that in effect collapses both the

To say, as Banaji does, that I am unsure whether wage labourers can be

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

7]. How, otherwise, would it be possible to characterize unfree labour in

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

a-collective) to the other (= worker-as-individual). Nor is it true that I

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

free/unfree distinction is not only important, but its significance extends

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Rather than attempting an in-depth examination, both of my arguments

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

from the means of production (= dispossession), a conceptualization he

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

Much the same views are expressed by neoclassical economic

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

personally to commodify his/her labour-power.84 Not only did Engels hold

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

Why All Roads Do Not Lead to Rome

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

THE JOURNAL OF PEASANT STUDIES

I am accused by Banaji of that most heinous of intellectual crimes: namely,

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015

WHY UNFREE LABOUR I S NOT SO CALLED

THE JOURNAL OF PEASANT STUDIES

Downloaded by [Jawaharlal Nehru University] at 10:30 20 June 2015