Some Systemic Effects of The Spread of Nuclear Weapons Technology A Study Through Simulation of A Multi-Nuclear Future

Some Systemic Effects of the Spread of Nuclear Weapons Technology: A Study through
Simulation of a Multi-Nuclear Future

Author(s): Richard A. Brody
Source: The Journal of Conflict Resolution, Vol. 7, No. 4 (Dec., 1963), pp. 663-753
Published by: Sage Publications, Inc.
Stable URL: http://www.jstor.org/stable/173121
Accessed: 28-11-2017 16:15 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to
The Journal of Conflict Resolution
This content downloaded from 148.210.106.123 on Tue, 28 Nov 2017 16:15:20 UTC
All use subject to http://about.jstor.org/terms
future
technology: a study through
SomeSome
systemic effectssystemic
of the effects of the
spread
spread
of nuclear weaponsof nuclear weapons
simulation of a multi-nuclear
Richard A. Brody
Department of Political Science, Stanford University
Chapter 1.41 Computer
Models in SiScimeNuclear
ulnceatiTABLE
.o.ns. . . .Weapons 43.126 IFDnaCanuatnpraidoabaSdnmiuolacietrnySicotena.sslt.eQ.mu.e.st..io..n..s.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ......... .. ........... .. 7619031692
. . . OF. . . . .CONTENTS
. . .Technology"
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 668673
for "Some Systemic Effects of the Spread of
ACKNOWLEDGMENTS ...................................................................................................................... 665

S~ Y~PREFXCIg .............................................................................................................. 666
I. VAmETmS OF SLUULATXONS XN INTmaNATmNAL RELATmNS RESEARC~ . . . . . . . . . . . . 668
1.2 Types of Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669

1.3 Political Simulations in the Study of International Relations . . . . . . . . . . . . . 672
1.5 All-Man Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675

1.6 Theory Playing--Simulation and the Exploration of Hypotheses . . . . . . . . . . 680
1.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
I . Tm~ NTIt-COUNTRY t)1tOBLI~t: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
B. 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 688
2.2 Some Key Assumptions in the Deterrence Literature . . . . . . . . . . . . . . . . . . . 688
2.3 Inventory of Propositions About the Nth-Country Problem Extant in the
Literature . . . . . . . . . ._ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
2.4 Propositions Which Predict Problems Occasionecl by the Spread of Nuclear
Weapons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
2.5 Propositions About National Motives for Attaining Independent Nuclear
2.6 Propositions About Effects of the Spread of Nuclear Weapons on the Inter-
2.7 The Feasibility or Likeliho d of Nations Achieving an Independent Nuclear
2.8 The Approach to this Study . . . . . . . . . . . . . . . . . . . . . . . . . . . 693

2.9 A Model of the Cold War System . . . . . . . . . . . . . . . . . . . . . . . . . 694
g. 10 The Cold War System and the Spread of Nuclear Weapons . . . . . . . . . 696
I I. Snv~tmAa~ON AND a~E Nra-COUNXaaY SITUXTXON: RESE~Ca D~mN . . . . . . . 09
3.2 The Inter-Nation Simulation: A Brief Description . . . . . . . . . . . . . . . 70

3.3 INS-8: Structuring the Basic Simulation to Admit the Exploration of the
Nth-Country Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
3.4 The Participants and Their Selection . . . . . . . . . . . . . . . . . . . . . . 704
3.5 Time "to": The Starting Situation for INS-8 . . . . . . . . . . . . . . . . . 705
3.7 Content Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707

3.8 Event Flow in INS-8 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
IV. SIMULATION AND THE NTIt-COUNTRY SITUATION: EXPLOBATION OF ~ MODEL .... 713
4.3 Units of Analysis and the Po ling of Data . . . . . . . . . . . . . . . . . . . 714

4.4 The Cold War System: Conditions and Linkages . . . . . . . . . . . . . . . 715
4.5 The Nth-Country Situation: Conditions and Linkages . . . . . . . . . . . . . 731
4.6 Summary: The Status of the Model . . . . . . . . . . . . . . . . . . . . . . 74
V. QUESTIONS OF GENERALIZINO: FltOM INS-8 TO ~E RmL WORLD . . . . . . . . 747
APPF~NDLX I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 749
REFERENCES ..................................................................................................................................... 750
da/saccompl
cepolalnfcoeiscitcdiuctseaehilnce.sdihed,
iAcknowledgments scoirenenMistapwereise"ctcWhael.t.haTgraduate
thDriefau.vter"hoqruandstudents
eistigoGeorge
rnataenfd,iunl wetaResearch.
tishmeBesitseetsraiatenndxetUnipesNOTSMiwhorvimecrhesawotoinettlysrthe,Drik(ePesdpvhicompl
etclrhi,iaprlsoeyBeuveghttiaorcaoProfe
hltn,hineDoofdilsonvtheaoidg,lrs-
665
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY
mental rorloeltehteyhepylaypledayinedthienftuhnedinfundig of ng of
this studuydy.. FuFundsnds forfothristwohirskwocamrke fcraome from
The research reported in this study was
CCoontractctAFAF494(6938(6)-3782)-a7n2d ReansdeaRerchsearch
team research and interdisciplinary research
ggrrant AFAF--AFAOFOSSR 9R5-6935,-6Ai3r, FAiorceFoOfrfcieceOffice
in the fullest sense of these new traditions.
of ScienntitfifciRec Researscehar, cOfh,fiOffice ofcAee orfspAeacerospace
The director of the project--Professor
Harold Guetzkow--holds professorships in
Dr. Thomas Milburn, Head, Behavioral
psychology and sociology, as well as polit-
Sciences Group, United States Naval Ord-
nance Test Station, aided material y in the
Wright, who figured so significantly in preparation of this report by generously
this project, are psychologists. The simula- allowing the author to devote part of his
tion staff, without whom the summer 1960
running of INS-8 never would have been study.
Thanks are due to the faculty and staff
political science, history, English, and of the Program of Graduate Research and
journalism. The author of this report is Training in International Relations, North-
The future referred to in this study may

Harold Guetzkow, Richard C. Snyder, and
never come to pas - nuclear capability may Chadwick F. Alger, and to Dr. Robert C.
or may not spread beyond the countries Noel and Mrs. Elaine Pancoast.
now in posses ion of this capability. It was
not the focus of this study to ascertain the
uals were instrumental in the execution of
likelihood of spread taking place- the study
Skinner, and Steven Star); I am grateful
thus, does not hold problematical the like- to them as I am to the fifte n other research
liho d of nuclear dif usion. To support
such a project, an agency of the govern- hot summer of 1960.
ment must have respect for basic research; Final y, I wish to thank the 357 students
the Air Force Office of Scientific Research, who participated in these experiments; it
by its support of this research, has evi- is to them that this study is dedicated in
the hope that the world of their future may
to the Air Force and Drs. Charles Hutchin- se les conflict than the futuristic worlds
son and Herman Sander for the instin- of their past.
666moaprediwhinnrmaTheedticxhepelmcitedarebnspnilmprioattcseleffects
oaeanpglbtcforeahslrea"eimtnpisonoertflinusualyinhotigrscci;aaemsladtlvngauybildraimyyntieuctof,clrahn"leiutastmodehsoacewioldnshNoanrprooefsprdiiotnbhltnmy-o. tandprompted
as"diynfhuoentbsdp.teoseexplrretnmvqetdtum,hiencoanetred.fyrlutcuyabymoeitnsrn,cethiand"'ethemurelwedcav(teencthairoinnucleemhcreasnppreea,rscloyibeld,aoeinfreimhoc)deo.trcaplesaietdnpisanldependence
RICHARD A. BRODY ofraihbutieyc,ldligcet,meotaynhrus.fndTheiertnafmtg--,
Summary--preface dependency,yb,obronronfonfecnescietys,itony,thoenntuh-e nu-
clear membebre.rI.nIpnarptarthetdhepednedpencdyernec-y re-
The effectiveness of future deterrence lationship isisdudeuteotohetheosthiolstyilotfythoef the
systems depends in part on. the future state
nuclear-armmededlealdeardeorf o.tfhe.tohpeospingosbilnocg bloc
of affairs. One uncertain aspect of the and in part totohtehneucnlueacrledairspdaristpyarwiityhiwin thin
future environment of a aleterrence system the al ilance.eI.tIwat was ins itnhetchoentceoxnt oefxthoifs this
is .the nature of international affairs in a
theoretical momodedl etlhathtahetehfe ecftseocftshoef the
world of many rather than few nuclear spread of nunculecalreawer weapoanpsoWensrWee esreayesdayed
powers: the so-called "n-country problem."
It is widely asserted that an increase in the
The expected change in the cold-war
number (n) of nations with nuclear and system is the fragmentation of the blocs
thermonuclear weapons capability will pro- which comprise it. This change wil be
duce adverse effects on the stability of the
present international political system. The of the formerly nonnuclear national actors
creased l'rkelihood of war; an increased dif- increased sense of independence and a low-
ficulty of obtaining arms control or dis- ering of the perceived threat from the op-
posing bloc wil reduce intrabloc cohesion
ability of upset ing the "balance ofterror";
and a shift in the Alliance structure. This
study represents an at empt to gain insight Minor changes in the basic inter-nation
into this pos ible future state of affairs by simulation were made in order to admit
means of the control ed manipulation of the study of thes hypothes . SeVente n
western Inter-Nation Simulation.

students as decision-makers. Each of the
17 runs began at the same starting point.
Scope and method
The experimental variable (that is, the
number of nuclear powers) was manipu-
the positing of a theoretical model of the
lated in the same control ed manner for
"cold war" interaction system, the predic-
al 17 replications. Attitude scales, and
tion of changes in this system which flow
communication content and flow were used
from the dif usion of nuclear weapons capa-
to operationalize the key concepts of the
bility, and the exploration of the theoretical
the study of these phenomena (Fig. 2.1).

The cold-war system is postulated as Two principal questions were asked of
being composed of two hierarchical y orga- the data: Did the starting situation and
nized bloc-al iances with the leading nation
fusion, conform to the model of the cold
of nuclear strike and counterstrike capa- war? And did changes in the cold-war
bil ty. Moreover, there is tension-generat-
ing hostil ty betwe n the blocs, which re- the predictions of the theoretical model?
inforces cohesion within the blocs and Data from the binuclear periods indicate
that hypotheses concerning the structural
The relationship within the blocs is one of and perceptual characteristics of this situa-
CONFLICT RESOLUTION VOLUME VII NUMBER 4
hesi
bers onperceived
produced(Fig.themselves
significance blo14.5).
cs whiForchaswere nationscatisinohnowethepatterns
muni
hierar-it change
dependent
example, d structural
no cletookar teplnanddeace.ncperceptual
yCommuni
to dire667c-t
tition
on werewereborneborne
out atouta hiatgh laevelhighof level of ofofnucnucllearecarapabcapabi
ility (thleit"yn-c(otunthery""n-country"
situ- situ-
aatition)on), da,tdata
a conficonfi
rmed botrmedh the both
hypoththe- hypoth-
was found that there was a high degree of eesisizezdedchachanges
nges and thande direthectiondiofrection of
intrabloc cohesion. Nonnuclear bloc mem-
variables. It was found, for example, that
upon nuclear bloc members. intrabloc tensions increased, and that as
The combination of dependence and co- intrabloc cohesion was reduced, perceptions
of external threat were also reduced. In
chically organized; authoritative decisions
conjunction with the relative fragmentation
tended to flow from the bloc leader rather
of the blocs, significant changes in com-
than from the other bloc members, and the
pattern of communication had the proper-
cation between members of different blocs
ties of a "wheel" communication net. In
increased; in the n-nuclear system, in con-
short, the characteristics of the situation in
the first part of the experiment were sim- trast to the binuclear system, less-developed
ilar to those of the cold war under condi-
tions of bipolarity. their communication to the economic-mili-
After the experimentally-induced spread tary leader of their al iance.
tresearch"
hseedsitsaudictutisteudeisotechni,nemisntpludithoqyieeresearch
carelations
voter- hnigdacpoqtednuecesto"rsdranugairveyovwenr woiwipthetnhagrleutlconcept
uessgnhargoviwteneincaorder tahreohry,pothe mosizerd kasnocwrulecdiagletweo whhavteiasboue-t
imodrbecuewoladconsireitlekpydh".edimntIreredmoodelermcpetoplynr"britaewiwi.sneeltflnhbedtyse.h.texplSiaeDeilnpvhceioserednlipolopmcilfiiunetrdigcsoealtnda.
CHAPTER I1
Varieties of simulations in international
In thIn
e past decade and a the
half, increasing past decade and a half, increasing
at enattention
tion has be n paid to the development has been paid to the development simulations are a type of "operating model,"
of meof
thods for the studymethods
of politics. The for the study of politics. The
1.1 Models in Science

further understanding of the process of A "model" is a collection of as ertions
voter decision-making. Decision-making about some reality--past, present, or pre-
dicted. It is a set of statements which
primarily from social psychology, have at- purports to describe patterns of relation-
tempted to add the weight of rigor to the ships holding between components- units
arguments of those political theorists who and variables- of that reality. But a model
postulate a "group theory of politics." Con- is les than a total representation of the
tent analyses have been employed by po- details of the phenomena of interest to the
litical scientists in examination of the theo- model builder. Othelwvise, the scientist
retical foundations of their own literature
as wel as to systematical y analyze govern- itself in its total richnes . Model building
mental documents and other political data. is an at empt to achieve parsimony in the
Simulation or gaming, as a method of ana- representation of a range of unique and
lyzing political interaction proces es, is
another of these post-war developments. A model is a scientific to l. If a model
The Northwestern Inter-Nation Simulation is to be useful for description, explanation,
is but one among a growing number that or prediction, there must be some manner
have been developed by students of inter- of cor espondence betwe n ~e model and
national relations. Our task now is to de-
scribe a number of these and to examine model involves abstracting from reality
the purposes for which they have been con- those components and relationships which
stmcted and operated. In order to ground
scientific method, the logic of simulation in the model to the exclusion of redundant
and distracting detail. "The bet er the
This chapter ap ears in H. Guetzkow, et
al., Simulation in International Relations: De- the conditions under which neglected vari-
velopments for Research and Teaching, ?
1963. Prentice-Hal , Inc., Englewo d Clif s, ables do or do not make a dif erence"
N.J. Reprinted by permis ion. (Brodbeck, 1959, p. 381). Do outcomes
CONFLICT iRESOLUTION VOLU}~E VII NUMBER 4
bywoprercompl
ibidelntfioflieodwhes rtehr"Poiucgthoritahl,e" Ho"Veursbae li,s"t"rMacatehd-, limitation trioenlas.tionships betwe n these units with the
sdenpit cretmentary
eonualreitallsy, pictorial
ihantverbalvoetshevmaterimodels;
eidreal asdaylva.relations
stnemta.g669eIns
in thin
e model so resemblethe
outcomes in the model so resemble outcomes in the anand
d flow, asflow,
Meadows pointsasout, "iMeadows
n the points out, "in the
realitreality
y it purports to represent as to war- it purports to represent as to war- lolong-run
ng-run it is not pos ible itto evadise thisnot possible to evade this
rant rant
making inferences about aspects of making inferences about aspects of
the rthe
eality by using the model? Treality
he con- by using the model? The con- cannot be readily pictorial" (1957, p. 7).
firmfirmarion
arion or disconfirmation of a model, or disconfirmation of a model, It is interesting to note that pictorial
via thvia
is pragmatic test, informthis
s the theorist pragmatic test, informs the theorist models usually need to be elucidated with
aboutabout
the adequacy of his abstraction, and, the adequacy of his abstraction, and, words. Often the social scientist uses a
thus,thus,
about the usefulnes of his ab revi- about the usefulness of his abbrevi- verbal model to construct a "meta-theory"
atedtatedtheory
heory to cope with reality. to cope with reality. within which the pictorial model explicates
limited and particular features of a broader
1.2 Ty1.2
pes of Models Types of Models schematization. Hence, the common use of
It se It
ms useful to distseems
inguish betwe n useful to distinguish between diagrams in texts, which exploit the leonit
clas eclasses
s of models. The four types of models of models. The four types of models features of a two-dimensional representa-
tion, but simultaneously remedy its defect
ematical," and "Simulational," are those
commonly found in use in the social sci-
1.2.b wR,AL MODr. LS
ences but they do not exhaust the universe
of the possible. Verbal models, ff they are more than just
1.2.a PICTOBIAL MODELS over pictorial models. It is, for example,

Pictorial (or leonit) models have en- pos ible to describe relationships betwe n
joyed wide currency among social scientists. units in ordinary language which could not
Consider, for example, the numerous or- be pictured. Furthermore, it is pos ible to
ganization charts found in a basic political imbue verbal models witha logical con-
science text. Here, by the use of a few sistency which admits deductive predic-
symbols (lines for channels and boxes for
units), pat erns of authority and subordi- Verbal models abound in polit cal sci-
nation are described, the "path" that a ence. The Republic is a verbal model of
Plato's conception of the ideal sta e. Be-
or certain aspects of the inter elations of yond this, it contains descriptions of an
the branches of government are il ustrated. actual sta e of af airs. It propose a series
"Since early times," Deutsch points out, of steps which it predicts, if fol owed,
"men have tended to order their thoughts
would ac omplish the transformation of the
in terms of pictorial models. The model it-
self. served more or les ef iciently, to
recent times, Easton, Dahl, Key, and other
order and cor elate men's acquired habits
authors have contributed to the store of
and experiences, and perhaps to suggest a
verbal models at the disposal of the polit-
selection of new gues es and behavior pat-
ical scientist- unquestionably this type of
terns for unfamiliar situations" (1951,
model has be n the principal to l of the
p. 232).
discipline. Bentley and Truman of er sim-
Pictorial models seem to be used pri-
marily for description. They are essentially ilar but separable models of "political pro-
static and do not yield easily to the deduc- ces ." Here the units are groups of individ-
tive logic necessary for prediction. While uals; the models purport to describe the
they can be very useful for depicting order
piSnohmvenoeltvmiednesad.onpersoedkucsetoc hnasveqtuhencvesrywhruilcehs matical model thereby becoming one type
de. .s,cr"before
mayexampl
670 haveibeed, stobyubsdraw
t.hanti.e.,vea.cumul
lr.baycorrel
ldimodefaferent
tivaletes. knowl oftioWhere
Benwhatever
'y"oWndhaappropri
with.endge.gs"). quirement,
meani hhaicalculus
s padpeavatensmathemati
twhenever thetwhe,nmathtechusedale'smimodelaszaecor-t'welli[coarsl
isagebeing
processes
RICHARD A. BRODY
enend
d of explainiofng and pexplaining
redicting the proc- and predicting the proc- of relatiof
onship which constitute the axirelationship
oms which constitute the axioms
ess.
Despite the advantages of verbal models respond to the social processes themselves.
over pictorial models, they also stiffer from Whenever the consequences fail to produce
significant disadvantages. It is difficult, for the reality, or under the more rigorous re-
from verbal models because of the ambigu- as the consequences fail to mirror the
ities of the symbol system. The same word, empirical social system one is attempting
in the systems of two different theorists, to represent--alternative calculi may be
sought, so that the model may become a
The terms of relationship (e.g., "more than more adequate representationof the system
of human behavior being considered.
are suggestive but without precision. More-
over, the development of crucial experi- can be found, they are powerful scientific
ments from verbal models generally in- tools. The abstractness of the symbol sys-
volves the translation of the verbal symbols tem facilitates the recognition of similarities
into a more explicit symbol system. One and congruences between models and,
result of this type of translation can be a therefore, between the realities they rep-
mathematical model of the system formerly resent. In this way the process of accumu-
lating knowledge may be expedited.
1.2.c MATHEMATICAL MODELS

models can be made to yield information
Mathematical models of political systems
about a variety of states of the system with-
are of relatively recent origin. They are, as
out changing the basic model. The ques-
Beschers puts it, "constructed by abs~act-
ing the properties of some data by measure- 'power,' or some other variable] of a unit
ment, and by expres ing these properties in
changes with respect to the system?" can
a set of symbolic statements that include be explored in the model. And this ques-
the logical relationship that holds for the
tion can be asked and examined for many
entire set of statements" (1957, p. 33).
combinations of units and variables.
Richardson's "Mathematical Psychology of
In order to derive ful benefit from this
War," although published in 1919, lay un-
feature of mathematical models, it is de-
noticed until quite recently. In the past
sirable to be able to "try" a large number
few years several authors, including Dahl,
of parametric values or to "try" one set of
Downs, Mat hews, Shubert, and Simon,
parameters over spans of time. To facili-
have contributed to our meager stock of
mathematical representations of political tate this kind of manipulation, mathemat-
ical models can be programmed on high-
To util ze mathematical models, the hu- spe d computing equipment- the mathe-
man behavior symbolical y represented in
its equations must be So structured that the of "operating" model- in this instance, a
"computer simulation."
relationships among the units and variables
The heavy reliance of mathematical
are isomorphie to the reality one portrays. models upon verbal supplements is note-
worthy. The variables posited in the equa-
CONFLICT RESOLUTION VOLUI~{E VII NUMBER 4
vSihduebsitkat,is"tiscalmyeunseafbuleStaommaplesn; irpeuliabtilotny control ed fashion of ers an op ortunity to
sanditryiasodetAstainesomcontipnsuec.rnm.aendepidnon.nTherespecially
elanguage.
equations, tNaitcupoatalulmeasurements
genci wsrnoeascltyiaareheslxwhenspebe"ynbeisratifemnuorudiscontinuities
t"dausedsre,ixnhopedienawrftod,ifiemtvcuarese-ernlet, expensi
geasnlhandl They coven,tandewhinddangerous
exampl cthhamit ghtffprovethey excessi
were invtro-ely
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY 671
tiontions
s are often explicated through are
verbal often explicated through verbal which would be imimpopsoisbliebloeroirmpimrapctriac tlical
to perform on the entity it~ portrays, The
cure precision in handling these variables operation of the model can be studied, and
are operationalized in verbal language, too. from it, properties concerning the actual
And sometimes the relationships among the system or organism can be inferred" (1958,
p. 1). The fact that manipulation takes
place in a model of reality can itself be
expounded by verbal means. Often the in- considered an advantage. Changes in a
vocation of "simplifying assumptions" when system can be tried in a "pilot plant," for
the mathematics become intractable is ex-
plained by verbal meta-theoretieal consid-
duced into the full-scale operation. Simu-
lation allows the study of induced variable
1.2.d SIMULATIONS
change in situations where it mightbe
Simulations are physical and/or biolog- otherwise difficult or undesirable to induce
ical representations of systems which at-
this change--this might be the case for
tempt to replicate sociopolitical proces es.
many, if not most, social systems. With
They are models' which yield information simulations, the problem of numbers can
about unit and variable changes over time. be solved by replication of runs. The num-
The theories they incorporate, therefore, in-
ber of runs is limited primarily by the re-
clude propositions abouttime-change in the sources--and the stamina--of the re-
seamher.
The manipulation of simulation models
to manipulate, the social .theorist perforce has been termed .........
pseucto-expenmentauon
by Helmet and Reseher--"'pseudo,' be-
or some substitute in car ying out his con- cause the experiments are carried out in
firmatory studies. The "natural" experi- the model, not in reality" (1959, p. 48).
ment has the advantage of providing data
from the empirical system itself. Here the
researcher can observe the operation not of . . pseudo-experimentation may ef ectively
a model but o? the actual system; validity annul the oft-regret ed infeasibility o? car ying
out experiments proper in the social sciences by
providi'ng an ac eptable substitute which,
have two serious drawbacks from the point moreover, has be n tried and proved in the
of view of the requirements of a scientific ap lied physical sciences [1959, p. 49].
methodology: (1) an experimental situa-
Simulation of ers advantage to the
theorist Working with models where time
which definitively confirms or disconfirms
hypotheses, and (2) when natural experi- is likely to be an important variable. The
ability to compress or expand: time in a
ments are found, it is unlikely that they
wil be found in suf icient number to pro-
study the ef ects of a given variation upon
becomes a problem (Campbel and Stanley; the system in a way which might be other-
1961, p . 81-98). wise impos ible. Thus, fit becomes feasible
"A simulation . model," ac ording to to study foreign policy decision,making in
a series of situations where the decision-
human participants. Although iconic,
brwovwhupoDecisions
evallhaoeytersrieaiboscnkumyabhtyatelslswima,aptetlsgihaveaprmoinelywhiththeva.dunder
tmakers
vacompressions iceensmnmaigtdifferentially
hdeveheotaulihdfsnraie."normal"
scbcanoutcomes trmHeidaeyoldnsimulated.
bebcvteres ithroutcyetalstpai,hiassocihorcNonetdmotimedshccircumstances
meo.evpcsnliHeeucaa;tprstedlaheotcasedeadlines.
nciaonrecwinvgaxstesnthvr.udy,iehbfAsripainyac-tl mephenomena,
usihsnieametvworki
thypothes
useful dhueso,lidabnwhitse.iniorder
ognunecdeveloping
ssnghausingtwiwibaotovoutlethmhelbepolphysical
ipnpnldternati
oiitheRICHARD
yastucicsimulation,
uesadiltheorismewisand/orduyonalltsbahtetemssA.lidotpoliwnfexplBRODY
biolog-,.ttheiriecMostroicoimcnhrealgte-
672
limitations are precisely the reason why one

may be different from those made in the moves to an operating representation of the
face of "ultimata" or "crises"; these time
ical means for the model construction.
The advantage of simulation techniques
in theory building and testing are summed 1.3 Political Simulations in the Study
up in the ideal of "control." Because the of International Relations
researcher is to a large extent "master" of Political simulations are developed and
his system, he can work with many of the
consequences of changes in particular sub- commonly, political games have been used
systems or within the total model. He can
systems. For ~e student of international
variety of alternative "policies" in relation relations the experimental manipulation of
the object of his study is out of the ques-
short, at empt to represent future states of
tion at present; he must rely on other
Work in simulation involves the utiliza-

Thre subtypes of simulations may be
tion of other modes of model-building. Pic-
identified: (1) machine or computer simu-
torial representation of computing proce-
lations; (2) mixed or man-machine simu-
dures is widespread. The use of ordinary
lations; and (3) al -man simulations or
language throughout man-computer simu-
games. What varies here is the degre of
involvement of human actors in the opera-
models. In developing programs, the com-
puter components in simulation models rely tion of the model (as "units" or "sub-units"
of the model, not as experimenters). The
Guetzkow and Noel (1963) il ustrate, to simulations of international political sys-
to l, it has be n neces ary to make recta- in al thre varieties. The thre subtypes
theoretical explication of the hypotheses show similar lines of development and sim-
ilar raisons d'~tre.
Now let us examine the contributions of
means of a verbal model the changes made a number of the researchers who already
in evolving the inter-nation simulation.
Guetzkow at empts to tighten his under- niques in their studies of various aspects
standing of the simulation by using mathe- of international relations. A variety of
matical models to explicate the programmed
hypotheses embodied in the simulation and emphases on the two basic objectives,
theory and training. But, because for either
the unprogrammed relationships which were activity the simulation must ap roximate
implicitly incorporated in the game by its reality, these researchers have shown con-
cern for ?urthering the state of knowledge
of international relations.
beposwistsoiifmmotAmong
nisuuheneilbryatamaleBetiiraooehpnustaeasocisnltrdoifuliencsimulations
TdMachine coatahalnebdlsianman.lesciesvgHopccesteswilontiwnciofpntiefamohsvlprevotits,iedecsrnt,aresideconomi
ecwinmoinricsgcttboperational
eaihrsntie"gimacoprentulsomlhtsac.iaepnttrhaveiuaidiceotltaon.eslygr"-- reUniLebagandrc0oThee.ts1n38eda0tnon,program
tStaaDactiSLetneimpovtHunga
slneseioioratmn,lendBriapagtvarirrwhich
WaireocDtady,iPsprui:nlacoKoreVi,tblm,theespecifies
easgtnatsraioc,SoviaresmGuaRen,leanitTatetheiUnimoninweasalaolrelation-
nn,,p,Egypt
possi rWesIonddo-bu673lctet-,
1.1.4 Com4puteComputer
r Simulations Simulations situatisituations
ons and the ef ects on situations of and the effects on situations of
a givean action-countgiven
eraction cycle. The action-counteraction cycle. The
models which have been programmed for modelmodel
consi ts of two sets of variables- consists of two sets of variables--
action variables and situation variables--
high-speed computing equipment. In
these simulations, humans are involved pri-
marily in the role of experimenters. A ship of variables :to outcomes.
model of decision-making behavior (so-

acting nations, nine possible target nations2
called "rules of action") is included as part
of the larger model to replace the human and nine possible intensity levels of action.
decision-makers found in "man-machine" The nine actor states used initially are the
simulations and "games." nine major powers of the actual scene: The
made the most use of this type of simula- ern Germany, France, Italy, India, China, and
tion. Milbrath (1958) has suggested the Japan. The nine target states are chosen from
Pool and Abelson (1961) have demon- nesia, and Iran. The nine intensity levels of
strated the feasibility of developing com- action . . . are defined in the program as rep-
puter simulations which do significantly resenting from ten to ninety per cent of effort
. . . roughly corresponding to a scale of in-
creasingly serious action [Benson, 1959, p. 2].a
tribution of voter at itudes toward salient
is ues. The "Simple Diplomatic Game," The situational variables describe the
developed by Oliver Benson (1958, 1959) "state of the system" at any given moment
at the University of Oklahoma, is perhaps when action is taking place. The relevant
the single instance of a primarily machine variables are: (1) national power expressed
in terms of war potential;4 (2) the distri-
2The
2Thenumber
number of acting
of acting
statesstates
and target
and target
Howard (1961) at General Motors has states
statesisisarbitrarily
arbitrarily
limited
limited
by thebystorage
the storage
developed an al -computer simulation of a capacity
capacityofof
thethe
particular
particular
computer
computer
being being
colonial socioeconomic development sys- programmed
programmed (IBM(IBM
650, 650,
in this
incase).
thisTheo-
case). Theo-
retically,
retically, anyany
number
numbero? states
o? states
for which
forthe
which the
tem, and his col eague Kaehler (1961) has
requisite
requisite data
data
are are
available
available
could be
could
employed.
be employed.
been preparing for the study of problems aa The
Thenine
nineintensity
intensity
levelslevels
of action
of are
action are
named
named"to"to
make
make
them
them
more more
vivid": vivid":
computer. Abt (1961), at Raytheon Com-
pany, has be n developing an extremely 0.20 United Nations Action
complex computer simulation of strategic
0.40 Propaganda-Subversion Campaign
decision-making, which of ers promise as a 0.50 Boycot and/or Reprisals
0.60 Tro p Movements
0.70 Ful Mobil zation
simulation program. designed to repro-
0.90 Al -out War (Benson, 1959, p. 2).
duce in simplified form a number of the 4 Eight determinants of war potential are
features of the international political sys- used: "Mil tary age manpower, transportation,
tem" (1959, p. 1). Fundamental y, this
per capita, energy production, ste l production,
literacy, and atomic capabil ty" (Benson, 1959,
counteractions to particular actions in given p. 4).
apsraogxirmesityv"en(eBse,nmisonli,ta1r9is5m9, fpl.ex5i)b.il ty, tension, talhctisiocnh,so.itcTheetiwpsorpogesraisbomnilsitsoyrsugtecivahmesthamitmigmohltdldoypifesritcaoa-e
ibis pbeiolanrmgv, elmeaplnyadyaccepti
progra
tentati
674 ted.ihgehtBens"rnbiulgpoeoncertai iepldifbyiedtMor-ioannsd foluniorani1959,uctntomalBenson
sl"aharsp).secnsifmproposi tavioews:riagettscyp.etio.crs5).al.rs.describes
tyeavtseasleactsdn.d aonesloigignscaction
altcountera
hem cycle
newctiwaoasnr
RICHARD A. BRODY
butbution
ion of this poofwer;this
5 (3) the power;5
interest (3) the interest withwith
[geographic] distance, and is[geographic]
height- distance, and is height-
lelevel,
vel, i.e., thei.e.,degreethe
of involdegree
vement of of involvement of enedened
by al iance and by pos es ionby
of mili- alliance and by possession of mili-
oone
ne statestate
with anothwith
er;6 and (4another;6
) the pro- and (4) the pro- tarytary
bases in the second state" (Bbases
enson, in the second state" (Benson,
pepensity
nsity to act (ortocountacteract) (orof eachcounteract) of each
actor state.7
Benson has developed his model by

1. The action card [which names the state
about the relationship of action-inputs to acting, the form of the action, and the place
outcomes, given various situations. For ex- in which the action occurs] is entered in the
ample, the "nature of the universe," as de- machine as a 'play.' It informs us that a cer-
fined by the distribution of power, sets the tain large state has taken an action of a given
intensity level against a certain small state,
counteraction behavior pattern of states by
specifying the "rules" by which the "game"
2. Following the action card, the program
for each of the other eight large states in the

ton Kaplan (1957) for three "empirical"
systems of action (balance of power, loose 3. After the choice of counteractions, the pro-
gram recomputes the relative strength of the
A set of "elementary if rigid as umptions"
potential indices. By comparing these with the
links the interest index into the system: original set, gain or los from the cycle of
"The degree of interest of one state in
another varies directly with the ratio of 4. At the choice of the user of the program,
mutual trade to total trade and inversely the action cycle may substitute the modified
indices of power distribution resulting from the
previous play for the original set, and use
Based Based
on "power distribution. f two on "power distribution... ff two
them as the basis for the selection of counter-
coaliticoalitions
ons are dominant to the extent of con- are dominant to the extent of con-
trol introlling
g 75 per cent of the power, the universe 75 per cent of the power, the universe
chastic touch to the program, since the new
is cal eis
d lo se bipolar . . Ifcalled
two coalitions loose bipolar .... If two coalitions
ar ay may also change the rules for selection of
are doare
minant to a marked degre - condominant
trol ing to a marked degree--controlling
counteraction.
90 per90cent o? the strength- thper
e universe is cent o? the strength--the universe is
bipolar . . If power distribution is bipolar .... If power distribution is 5. Also at the choice o? the user, the program
tight tight
o a given degre [i.e , f les than 75 to a given degree [i.e., ff less than 75 wil hold the results of two suc es ive cycles
even teven
per ceper
nt of the total power is control edcentby of the total power is controlled by and compare them for relative advantage or
any twanyo coalitions] the universe of action two
is coalitions] the universe of action is disadvantage to the actor state or states. By
definedefined
d as one of bal nce of power" (Benson, as one of balance of power" (Benson,
1959, 1959,
p. 4). p. 4). the program as a competitive game, though
6"Inte6"Interest"
rest" is determined by an index of is determined by an index of this feature is intended to make pos ible the
four ifour
nterest indicators: "trade, coalition mem- interest indicators: "trade, coalition mem- exploration of alternative strategies rather than
bershibership,
p, presence or absence of military bases presence or absence of military bases simply to entertain [1959, p . 2-3].
of theof
actor in the target area,the
and geographic actor in the target area, and geographic Despite the se ming rigidity of the simu-
7 "Data on propensit ] to act (initial y based

on Q. Wright's field theory indices of social, tions can be introduced. The basic informa-
economic and political conditions: such labels tion which defines the situation can be
brought up to date or replaced with other
stabil ty, violence, defensivenes , frustration,
data deemed more theoretically significant.
and internationalism wil be arbitrarily at ached
to weights intended to measure the probabil ty It is also possible to modify the assump-
of overt action)" (Benson, 1959, p. 2). tions upon which the program is built to
lation
va. r. iiasticoonncwieilv,edhoaws eavnere,daulctaetriotnhaelcehxaerraccitseer. hcraivsesmu, ecltoinpolme meic omfbenrshivpes, pthroepiangtaenrdna-l
tional system. 675
adadmit
mit new hypothenew
ses about thehypotheses
interna- about the interna- regions of rerseosuorucresceasndancdapacbiplatbielsitaineds taond to
provide s'anunluatlaetdedxpexrpienrcienwicethwisotmhesomf tehoef the
strategies offstsateactreacfrta.fPte.rPheaprshatphse mothestmoims-t im-
In this exercise, as in all computer simu-
portant funcntciotnionis itso teonceonucroagueraimgeagimnatgivneative
lations, the computer itself is a drudge; any and manipulaltaivtievceoncsotnrsutcruiocntsionfspofs ipboles ible
outcomes derived from the model could also internationalaslysytesmtesm[1s9[519b5,9pb.,1]p. 1].
result from paper-and-pencil operations. The MeClel alnadndexexrceirsceisepilsaypeldaybeyd by
However, and this is the principal argu-
ment for using computing equipment, the
two op posingngtetameasm(osf(ofrofmromneotnento n
full exploration of the model would be im-
players each)h.)P.lPaylatyakteaskpelsacpelaocneaoman apmap
of the worldldwhwhichichashbase bnedinvideivdidnetdo into
practical without the capability of rapid
twenty regioinosn.sE.acEhactheatmeasmelescetsleact"snua-"nu-
repeated operation of the model which
cleus region"n"ororhohmoembeasbeasnedafnodurfoaduri-ad i-
computers can afford. Practically, the in-
tional as osciaitaetdedregrieogniso;ntsh;etsehetsweotgwrougpsroups
sights into the logical implications of the
of five regioinosncsocmopmripseritshe tihneitianliteiarl- ter-
model can hardly be gained in another
ritory and rerseosuorucersceosf otfhetohpe ospingosbinlogcsb, locs,
manner.
The remaininigngtenternegrieognisobnescobmecotmhe the
Benson's simulation, for all its v'n'tues,
"targets" forora saersiersieosf odfipldoimplaotmic-amticli-tmariylitary
has only slight provision for accmnmodating
"moves." A sesqeuqeunecncoefotfhrtehrmoe vmoes cvoems-com-
human decision-making. By design, once pletes a period or "decade."
an action-input has been selected, the out-
"The decision-makers . . . are free
come is in the lap of IBM. The "compet- carry on several kinds of strategies ac
itive" variation of the game holds the
prospect of providing data on human re- ing to their own preferences and hope
outcomes. A player can be aggressiv
actions to various states of the international
system. The further development of this passive; he can attempt to cooperat
engage in conflict; he can back a 'U
World movement.'... In general,
of this ~game- it then becomes a "man-
machine" simulation. player decides the kind of action [he
pursue" (1059b, p. 3).
1.5 All-Man Simulations There are four classes of moves: polit
1.5.a A SHOnT, SIMPLE EXEiRCISE
subversion moves, "mostly
The polar opposite of Bensoh's and "United World"
movements; these are roughly analogous to
machine, very little man" simulation is
Benson's "intensity" levels. The style of very
Charles McClelland's "mostly man,
little machine" World Politics plaGame.
y, the specific level of commitment, and The
purposes of these two operating thegames
reaction to a given move are determined are
also nearly opposite: Benson by tsays
he decision-makers. Where the teams that,
"the game as designed is a pattern for test-
ing general subjective statementsdebate over iniabout
tiatory and response strate- the
international behavior of states"
gies and the "fe dbacks" from the out(1958,
comes
p. 1). McCIelland indicates that his simu-
of moves can perhaps ad new dimensions
to the student's perceptions of foreign
policy~making.
It is intended to build interest in the geography
Two clas es of capabilities are employed
of world af airs, to increase sensitivity to the dis-
parities in the distribution among countries and in the gamem"Basic Capability" and "Mil-
outcome"
regi ons.moretoicapaci
attempt
even swrestdetermity,thethenied.niopposi
tiThere
ative.ngSiareteamnceprob-eachcan text,
gamewiRAND
witlhthese
between
basic
reported upon.
abilities as ociated with possible outcomes cnseinmvteulredawe.t)iroen)reweprese natbeadn,donebut odnlays t"huonpro-
bethistheme
diexerci
inother scussed se re-
haveasties,oinsnbeen
varivolume. thethebeifundamental
Variations RAND
g noted.oncon-theat
nemployed
676 PdCHARD A. BRODY
itary Capab~hty.
? '~8
sented bsented
A teamrs total y the RAND exercise has enjoyed capability
by the RAND exercise has enjo
?
is derived from the potential st cur ency of any ofheld

the widethe in the
the games widest currency of any of the ga
"nucleus region" and the five associated
During play a team makes a move--of MIT, Columbia University, West Point,
a given intensity level--toward a target re- University of Wisconsin, and the Air Force
gion, thereby establishing initiative in the Academy. Because of the basic similarity
area. By "bidding up," i.e., by committing
team defines its own interests in a given Beginning about 1954 there began to be
region, it wil also decide when it has made developed at the RAND Corporation acon-
its ultimate "bid'P--the team with the larg- ception of political gaming which may be
est commitment in a region is the "initiator" characterized as "Role-Playing--Crisis Play-
of action. When action has ceased, a "crisis ing." "RANDrs interest in political gaming
grew out of work in political analysis and
previous experimentation with the use of
and the determination of the outcome is gaming techniques for other purposes"
made stochastical y- an element of un- (Goldhamer and Speier, 1959, p. 72). At-
certainty is thus added to the game. For tempts at devising a quantifiable Cold War
example, in the political crisis the "odds on game (presumably, some sort of machine
the outcome cards are' i in 2--initiator
wins; i in 4--defender wins; i in 6--war; ductive" ac ording to Goldhamer and
i in 12--status quo ante" (1959b, p. 3). Speier. The present technique was set led
The provisions in the simulation for de-
cision-making in the face of uncertainty Goldhamer and Spei r of er the fol ow-
heighten the "realism" of the experience for ing description of the prototypie game:
the students. They also tend to make the
The government of each country was to be
game useful for checking certain hypotheses repres nted by a separate player or group of
about strategic behavior, proposed by the players. (In practice, of course, al countries
"Theory of Games," against the actual be-
havior of participants in the game. garded as most significant for the geographical
or problem area around which the game was
1.5.b 1ROLE-PLAYING--CRISIS-PLAYING
GAMES
In ad ition, 'Nature' was to be represented
by an individual or a team, and there was to repre-
The type of political simulation
be a team of refere s. The role of 'Nature'
s "Basic Capabilities. The real life reference
was to provide for events of the type that
is to the material resources, government and
hap en in the real world but are not under
social organization, industw, science, etc., that
the control of any government: certain tech-
support a society. BCs are units of funda-
nological developments, the death of impor-
mental wealth and strength of the region. Each
tant people, non-governmental political action,
region also has MCs or Militaw Capabilities.
famines, popular disturbances, etc.
The term militaw is defined broadly to include
armed strength but also the persuasive Participants in the game were to be ,-trea and or-
ganized abilities that can be used specialists who could dto
raw upon their knowl- change,
direct, or control the internationaledge and ac umulated area experience. Wibehavior
th
of the other [team]" (McClelland,the excepti1959b,on of the American Team, al p. 2).
CONFLICT RESOLUTION VOLUME VII NU~BER 4
ganyumeinesi.onadmitarlepsolnauiystetratiirons we,oinnrteorlpietogremnhavecite redpspeci,ohrtoswa,lesvpregard
erc,hteos run,government
Thebasedstrategyonteainpapers
mformati
s canarecontrol
majothernedinaforma-
onacontai minountthe677
government
government teams were to act as they judged teams were to act as they judged
andand
"Nature" team are use"Nature"
ful for inducing team are useful for inducing
'their'
'their'
governments wouldgovernments
in the circumstances would in the circumstances
thethe
sense of urgency essesense
ntial ff strategic of urgency essential ff strategic
prevailing
prevailing at any given time of the gameat any given time of the game decidecision-making
sion-making in crisis is to be simulated. in crisis is to be simulated.
('p('predicted
redicted strategy'). The American Teamstrategy'). The American Team
waswas
less restricted;less
it was permitted
restricted;
to pursue it was permitted to pursue TheThe
s t ra t egy pa p ers recordstrategy
the moves papers record the moves
anyany it judged to be optimal; in which it judged to be optimal; ofin
strategy whichstrategy theof
teams both projthe ected and actual. Ateams both projected and actual. A
particular,
particular,
the United States Team was notthe United States Team was notset set
of these papers is prepa
of red prior tothese
the papers is prepared prior to the
required
required
to follow the foreign policy
toline offollow the foreign policy line of
scenario; during the run, revisions in these
for the constraints placed upon American
foreign policy in reality by domestic consider- papers can be made and/or totally new
ations. The game was thus designed to permit strategy papers can be executed.
tests of a wide range of United States strate-
tion variable in the game. By "classifying"
The refere s had file task of ruling on the
or "declassifying" papers or by "leaking" or
feasibfi ty of each move; that is, they were
to disal ow any move that they did not regard
"withholding" their content, the "Nature"
as within the constitutional or physical power team and, to a lesser extent, the several
of the government proposing it. For thre
reasons the refere s also played the role of of information extant in the system.
'Nature.' This ar angement saved manpower;
According to Goldhamer and Speier,
it restricted the number of arbitrary moves
which might have be n made had ful -time The game was so designed as to meet six
players represented 'Nature'; and it permit ed main requirements:
the refere s to make certain non-governmental 1. Minimal forrealization: The government
moves which constituted indirect, partial evalu- teams were not limited to any prescribed
ations of the state of af airs that had be n set of moves . . . nor did the game contain
reached at any chosen point of the game. For any preestablished prescriptions automatical y
example, the refere s could introduce such entafi ng certain consequences from particular
evaluations in the form of pres roundups, trade types of moves . . .
2. Simulation of incomplete and incorrect
made in the United Nations, etc. (The govern- information: In foreign affairs, state secrets,
which al governments keep with varying de-
chal enge the plausibility of such moves.) . . . grees of success, are important obstacles in
Prior to the start of gaming, considerable the process of decision-making. In our game
time was spent on the preparation of a the introduction of "game-clas ified" moves
'scenario' and 'strategy papers' [1959, p . 73- and their unpredictable handling by the referees
4]. tried to take account of this factor.
The "scenario" and "strategy papers" are 3. Simulat'wn of contingent factors: In
important devices for setting the stage for political life many events are beyond the con-
the game. In the first few "runs" the his- trol of the most powerful actors, a fact desig-
torical present was used as the starting nated in political theories by such terms as
fortuna, "chance," "God's wil ," "changes in
point for the game; however, for several the environment," etc. We tried to simulate
reasons--e.g., the "overtaking" of the game this fact by moves of "Nature."
by real world events and the consequent 4. Plausibility of game events: We vested
"noise" introduced by the nongame press--
insurance against iraplausible game events not
only in the political iudgment of the referees
it became desirable to project the game
but also in that of the participants responsible
further into the future. The scenario for
for governmental moves . . .
the "Fourth Game" described a "future"
5. Clarification of is ues: Our aim was
some nine months hence which became the not to move on rapidly from point to point of
starting point for the exercise. The scenario
the game but to clarify by discussion the is ues
raofthgacehnmrgadmiitnthehganhatseimvxvariepeulberataeistinon.haopensvdeianbegogitheencatRAND
whiducti phatWadehipsreinensgrounded
lh,estpaudersys-tinctis- pedagogical.
ever,
ferenti
were enodeonprlainlcSaaotc"game tnaahc"etCoiprmlnaopywgleirnsehsnlread"tpiarobneysd.ecnOgartpneimtaig"meEntxiuhsgemcaAmeusUnigtiravmreti"cdgawinieSstTaehtiaemnds
ngihiareolcynlalptieinlncxireason
getireaeither
rt(cmCsiuoses.elnlt,usbarri"runsmbitheornThefogearnactualtmanUniguRANDenrtheaytvleibepresent
zlyrachoits,crunsiiohatmensycv)e"e,
678 RICHARD A. BRODY
rairaised
sed in the course of play. Sucin
h discus ionthe
s course of play. Such discussions ffoocusuosfonfonenoxpneretxrpeunsrits,runsby definsi,tibyon, definition,
to took
k place during the gamplace
e within each team during the game within each team
befbefore
ore a move was proposed or onaoc asionmove was proposed or on occasion
A second source of variation in the basic
betbetween
we n a government team and the refeare s, government team and the referees,
andand
after the game amafter
ong al participants . . the game among all participants ... game comes through the selection of dif-
6. Expl6.orationExploration
of novel strategies: We of novel strategies: We
tritried
ed to stimulate efforts toto
meet this restimulate
quire- efforts to meet this require-
mement
nt by prescribing "predictby
ed" and "optimprescribing
ar' "predicted" and "optimar' or in the very near future. There is, how-
strstrategies
ategies respectively to various teams in ad- respectively to various teams in ad-
vanvance
ce of the play [1959, pp. 75of
-7]. the play [1959, pp. 75-7].
of games further in the future, or, for that
TheThere
re is a series of points-of-eis
ntry a series of points-of-entry matter, in the historical past. Goldhamer
and Speier have speculated that "such a
tem. The particular variation used reflects game, removed from the immediate polit-
the goals and facilities of the group con- ical reality, could conceivably lead to the
discovery of entirely new problems and to
The most common variation has been in new insights and provide a greater em-
the "credentials" of the participants. Where
RAND used "area experts" to staff the (1959, p. 8a).
countries of their expertise, several similar If the results of "future-oriented" games
exercises have not. When the primary goals are to be more than merely suggestive, in-
neces ary. Thus far, the only example of

pants. Some of the MIT exercises have the multiple gaming of a single scenario
used graduate students; several of the se ms to be the runs Lucian Pye (MIT),
games there and at other institutions have
used undergraduates to staff the nations. and Major Abbot Greenleaf (West Point)
At MIT both the student participants and conducted in 1960 with undergraduates at
their instructors who conducted the exer- their respective institutions, The results of
cise, agreed on the value of political gam-
ing for instructional purposes (Blo mfield not be n reported as yet.
and Padelford, 1959, p. 1 14). Bernard Profes or Cohen's variation consists in in-
Cohen reports similar positive results from creasing the impact of "domestic" consider-
his Wisconsin simulation in Spring, 1960.o ations upon foreign policy-makers. This was
However, he qualified his endorsement of
the technique, noting that benefits se m to
the game, that is, within the over-al inter-
vary directly with the amount of back-
ground information one brings to the game.
Thus far, there have been no performance
internal-political game. By this technique
comparison data upon Which judgment can

in international af airs had to defend their
be based. Given the RAND conception,
policies not only to the refere s but also to
this is not real y a question. Since the values
of political gaming for RAND derive from
the confrontation of expert-with-expert, the
to be found which would satisfy the values
of "national elites" (who, in turn, had ref-
o Cohen, Bernard I., personal communication,
November, 1960. Se also (Cohen, 1962). erence groups of their own) as wel as the
CONFLICT RESOLUTION VOLUI~E VII NUI~IBER 4
aletnetnstitounatoin p.rHobwlemvseri,neavemon whre ientenhsie nsiamtiuonlatliornelsatihonats.have hypothesis "testing"
polgatertsOtailothoienwbehgsrliemore"ris.tso"predi
sfolcomes
between
Academy aSptivatlapparentl
cehrnamoitflleum"tacdireted"hosxinspeusybraoyloniapandenrtlmodihergeand,teah"optiqneuthadfabaperhaps,
iltfeeysmdnocichum"sanother
u"tnipsihredimqneugulmstrate-emorecatehated"hiaas-envriger paaSiulsidmnaequaetitulsmao.an,tJuselyioyasn.bet develahasIsrnetpthihrbeeneopedsBeewantsmadoneeyodn, ybyasnybesasefforthumateprograma delsnimipasbulmeratermaamedti679codine-
vavalues
lues of the policyof-makersthe themselvespolicy-makers
. themselves. kind of "prediction" has resulted, system-
ThThis
is is a more cisompleax polmore
itical environ- complex political environ- atic "explanation" has not. The focus 'of
mement
nt than is ordinthan
arily found inisthis ordinarily
type found in this type the BANNDD games has be n on the explora-
of game. tion of strategies, not theories.~?
The exercise conducted at the Air Force
1.5.e I~ANN--MAACCI~/WE SIMULATIONS
pect of the RAND game. In this simulation It is quite pos ible to mix men with
no distinction appears to have been made machines to develop complex forms of sim-
gies. With the relaxation of the require- strategy in the Northwestern Inter-Nation
ment that all nations save the United States
follow predicted strategies, the game be- to reap the benefits of both ways of con-
structing the operating model. Those parts
imaginative. It would probably be the case, of the theory in which knowledge is more
in games not grounded in the present, that
the tendency would be for dI teams to for the computer; those aspects of the
model which fal on the ambiguous, implicit
represents the "mostly machine, very lit le

suggested several but they are, as yet, un- man" end of a continuum, and the M?-
tried. One of the striking features of simu- Clel and-RAND-Bloomfield-Padelford crisis
game represents the other end, "mostly
to the imagination of the theorist and the man, very lit le machine," so the simulations
participant. New ideas come much faster developed by the Operations Research Of-
than the opportunities to try them out. fice (Harrison and Lee, 1960) and North-
The claims made for this style of polit- western (Guetzkow, 1959) may be located
ical gaming have been cautious, even with in the middle range of this continuum,
respect to its use in the clas ro m. As a "partly man, and partly machine."
device for car ying out research, its prime It se ms useful to describe the mixed
virtue, it is as erted, resides in bringing to- simulation in the context of model building,
contrasting these models with the al -man
games developed by Schel ing (1960) and
man er than would be the ease with other by Kaplan, et al., (1960). Experience with
methods. The bet er the experts, the more
thoroughly they become involved in the as their primary focus has, thus far, be n
role, the closer the ap roximation of reality limited. In part, this reflects the relative
se ms to be the basic as umption of those youth of simulation methods. It also re-
who would use the RAND gaming tech- flects the "state of the discipline" of inter-
nique for research.
It has hap ened that moves in the game,
which have ap eared to: be o? dubious ~o In many respects the RAND game ap-
legitimacy to the refere s, subsequently proach is epistemological y related to "single
case" studies o? empirical systems. The ap-
have come to pas in the real-world, equiva- proach exhibits the advantages and limitations
of the single case method (Paige, 1959).
isntysmsitculsoemplmcaot.niocMoenltuestridaoencthsa.pnieThecqiufeiscfolatlolyos,wiytsnhgemypreatrsieecneretna-- sttosentiyhesmo6.2.Ttincrease
680 reh.msaFiPermiel'I.sngwonatiebhteoqruodrathertcapabilities.
nalensdefeated
,triactor.aohtbnoelsthan,gdoatermhvorespasstlaoiconsusbptlehupanotdrraiseanwpenfopportuni
nrededsl,inesagcseunntiopmotany-tl
RICHARD A. BRODY
1.1.6
6 Theory PTheory
layingmSimulation andPlayingmSimulation and [whhiichhchcarhacrtearcitzertihze btehaeviborehofavtihoernao-f the na-
tional acatocrtso]:rs]:
ththe
e ExploratioExploration
n of Hypotheses of Hypotheses
1. AAccttotioncirnecasre caspeabcialptiaebsibluitinesgbotuiat enegotiate
InInthis secthis
tion, thre qusection,
asi-experimental three quasi-experimental ratherrthtahnafnighfti.ght.
simsimulations
ulations wil be examined: (1) Mortwill
on be examined: (1) Morton
KaKaplan,
plan, Arthm- Le BumsArthm-, and Richard Lee Bums, and Richard
3. Stop fighting rather than eliminate an es-
QuQuandt's
andt's use of gaming to stuusedy "balanceof gaming to study "balance
ofofpower" power"
systems (Kaplan, et al.,systems
1960); (Kaplan, et al., 1960); 4. Act to oppose any coalition or single actor
(2(2)
) ThomasThomas
C. Schel ing's study of bC.ar- Schelling's study of bar- which tends to assume a position of predomi-
gagaining
ining and limited way (Sand
chel ing, 196limited
0); way (Schelling, 1960); nance with respect to the rest of the system.
5. Act to constrain actors who subscribe to
anand
d (3) Joseph(3)O. Har isJoseph
on and Edward O. Harrison and Edward
supranational organizing principles.
M.M.Le 's maLee's
n-machine simuman-machine
lation of strate- simulation of strate-
gigic
c (military(military-political)
-political) problems (Har ison problems (Harrison national actors to reenter the system as accep-
anand
d Le , 1960).Lee,
A fourth stu1960).
dy, Harold A fourth study, Harold table role partners or act to bring some previ-
GuGuetzkow
etzkow and Richard A. Brody'sand explora- Richard A. Brody's explora- ously inessential actor within the essential actor
clas ification. Treat all essential actors as ac-
tition
on of the n-coof
untry prthe
oblem"~ withn-country
the problem"~ with the
ceptable role partners [Kaplan, 1957, pp. 22-3;
ininter-nation
ter-nation simulation, wfil constitute simulation,
the wfil constitute the cf. Brody, 1960b, pp. 14-8].
cocore
re of this repoofrt. The athis
p lication of report. The application of
Having thus described a "balance of
power" system in this earlier work, Kaplan
search in international relations is of such
and his co-authors proceed to seek answers
recent origin that no study (these three in-
to questions about the "stability" of the
cluded) has as yet formal y reported out
tion is based on the statements of the Is the system stable- that is, does it have the
property that it tends to be maintained as a
authors prior to the completion of their re-
system of independent nations, characterized
spective studies and is neces arily tentative by groupings into rival transient al iances; or
does it tend to be transformed into a dif erent
Kaplan and his co-authors have as theh' system, either by a reduction in the number
of states or by ag regation into a system of
major purpose to "examine the balance of
relatively permanent al iances? Or wil the
power international system in terms of a
restricted set of variables" (1960, p. 240), the conditions or as umptions built into the
in order to build toward a theory of this system? Is there some lower bound upon the
number of nations neces ary for stability? An
gaged in an at empt to "bring out and ex- up er bound? Since stability obviously de-
pends in part upon the behavior of the mem-
amine the stabilizing and other potential-
bers of the system what motivationsor mech-
ities at ributed to the 'balance of power' anisms trig er the behavior that makes for
system" (Kaplan, et al., 1960, p. 240). stability? [Kaplan, et al., 1960, p. 240, em-
Kaplan points out: phasis added.]
The term 'balance of power' has be n used

in the literature and makes intuitive sense f be answered in the natural set ing, i.e., in
it is ap lied to the description of the inter- the ongoing or historical international sys-
national system that persisted throughout the
eighte nth and ninete nth centuries and per-
petitive international political or balance of
haps the early part of the twentieth century.
. . . The 'balance of power' system is char- power game was constructed. In the au-
acterized by the fol owing set of es ential rules
CONFLICT RESOLUTION VOLUME VII NUI~-{BER 4
ta.hc.teoEseracsah. sSnuimatiupoltnioadniseolpiukrseolvyiu,dneucnaecrfeturaitnahtinyetr,ytrheafbtioniuest-,
assrhteroulTaproduced.
hinedtasbeu,tshvaaovrreisaebscploeen.coTheimfyicfrulucorentshstemusrainsttusa,ml oopwntitoforhnes neodposition
nreiptblhsoaeitydrelad'osneasnipdsliiotanefnyocitoendmasindependent
absorpti i adohnkelapdfrgerl,orwaytnhertd.organiiePvornilscanationytnooehrzaatistinkforntstate,nohnweoratriebsacriirsatthatehra681tevlr-
.. a.meansmeans
of displaying tofhe fordisplaying
mal proper- the formal proper- that would improve its position, but [it] would
tties
ies of anyofone oany
f a familyoneof modofelsmviaz., ofamily
f of modelsmviz., of also prefer a stil les srisky situation to its pres-s-
tthose
hose which aswhich
ume that theassume
'balance of powethat r' the 'balance of power' ent one . . Each nation desires to maintain its
anandd otherother
internationalinternational
systems are (i) es en- systems are (i) essen-
ttiany
iany competitcompetitive
ive and co perative at onceand, (i ) cooperative at once, (ii) is, no economies of scale or values dictate either
eessentially
s ential y responsive to responsive
military force or the to military force or the
tthreat
hreat of it, andof(i ) seit,nsitiveandin the lo(nigii)run sensitive in the long run fice for a different nation [ibid., p. 24~].
ttoo chanchange
ge in their consintituentheir
ts' economicconstituents'
po- economic po-
ttentiM,
entiM, from whicfrom
h alone theirwhichmilitary forcealone
s their military forces The operating model is described by
Arthur Lee Bums in an Appendix to the
The object of our game, therefore, must in- main body of the article. There is similar-
clude both competitive and cooperative ends, ity between this game and the M cClel and
though the proportion or mixture of these
game discussed above. Both are "board"
the possibility of making war . . And last, the games with moves characterized by bid-
equivalents for 'military force' to be displayed ding. In the Appendix, Bums offers the
in the game must flow in the last resort from following "brief description of the game":
something equivalent to the instruments of
productionrathe rate of flow again may be The players are Nations competing for pieces
varied [ibid., pp. 240-1]. (chips or dice). Pieces represent units of re-
sources; or they can be set aside as reserved
The components of this model are the forces which, when 'deployed' on 'frontiers'
nation states conaprising the international against other Nations, become 'sources of fire-
system. For this study the nations were not power.' Undeployed resources, i.e., those
differentiated internal y, the implicit hy-
of forces earn proportionate income, but in
pothesis being that the internal political
proportions that may vary by chance from
system does not constrain the national ac- round to round.
tors from behaving in the manner pre- Players take turns to move. A move can be
scribed by the rules of the international pas ed; or it can be used to deploy, reserve, or
withdraw forces, or to 'make war' with forces
system. The model further as umes that no
already deployed. If at war, op osed pieces on
political impediments to the decision-mak- a given frontier exchange at a rate determined
ing process exist, nor are there any con- by throwing the deployed dice and do so until
ultilization of resources. eliminated. The regular sequence of moves is

suspended by warfare, until the initial conflict
The decision-makers in the model have
complete information about the economic has taken some pieces from the defeated, and
and military resources of the other national each contestant has had the op ortunity to re-
the decisions made by the other nation- that the game wil be played to a finish or
states, except where this information is after a specific number of rounds it wil con-
elude with a payof which (in either ease) wil
voluntarily transmit ed, is also a feature of
depend on the number of pieces then pos es ed.
the model. 'Al iance' is sup osed to consist in two or
more Nations' reciprocal withdrawing of forces
about the nature of the national actors; from a common frontier(s). 'Pres ure' consists
in a Nation's increasing the forces deploying
against another. Degre of pres ure is increas-
ment of the "es ential rules" of the system:
ingly a mat er of predominance at specific
frontiers because of the Lanchester-type ex-
it prefers its present situation to a risky gamble change-rate inwar [1960, p. 247].
earch destheign TclargermaketoiholednThetwahiisersystemic
nbpoleatpietolyisc,alibdloien.steractinot proenseinntsuchthe raessystem,
682 oreversiirnsyyc,os.tmmoem.plbrl.eee.context.Thetomovesve.,erlba,egpraogawhilsSchelling
tisunalilnaetcgetheysisoeant"describes
sribargainatgmiocnoan-nlg",
PaCHARD A. BRob?
GeGenerated
nerated from the model arefrom
hypotheses the model are hypotheses limitetdewad rwa. Trh.isThiis onseipsrocnes prwitohcinesa within a
whwhich
ich presumably cpresumably
an be explored in the can be explored in the
gagame.
me. These hypothThese
eses concern thehypotheses
stabil- concern the stabil- his project as
itity
y of the bofalance ofthepower sysbalance
tem, the of power system, the
.... an experimental study of the bargaining
tetendency
ndency for wars to be an ofor utcome ofwars to be an outcome of process involved in limited war and other con-
flicts in which bargaining is by maneuver as
tendency for alliances to form and for the much as by words, communication is poor, legal
members of the alliances to be rewarded, enforcement is unavailable, and the participants
and finally, hypotheses about the relation-
are uncertain about each other's values, and
ship of the number of national actors to they have the power to inflict gratuitous dam-
the stability of the system. age on each other [1961, p. 50]?1
The following are incIuded as examples
participants are representative of
of these hypotheses:
"actors in a situation" where the "situation"
1. The stability of any balance of power, is a variant of the international scene. Thus,
and the motivation it affords for alliances and the group to whom Schelling would gen-
other such balancing alternatives to war, are
eralize is the group of decision-makers in the
acutely sensitive to the form and intensity of
the military exchange rate, and to the power
of the weapons to destroy economic resources. structed by Schel ing at empts to be struc-
2. There is a strong temptation to go to rural y isomorphic with the international
war in order to increase one's national strength system: Deeds, i.e., actions, are likely to
in relation to at least some other nations.
be se n as expres ive of intentions, com-
3. There is some tendency in the system for
alliances to form and for alliances to culminate munication is les than perfect and informa-
in at least limited wars.
4. The nation to whom the alliance is of- the units are relatively unenforceable, and
fered knows that it can get a better deal from self-restraint (or self-interest) is the princi-
the nation that is left out. Since that nation
pal check on violence.
does not want to be on the losing end of a
war, it wil accept less than a 'fair' offer [as In game theoretic12 terms the conflict
the price of alignment] [Kaplan, et al., 1960, situation in the game, as in the real world,
p. 2431. is a "non-zero-sum" situation. In inter-
The discussion in the article, unfortu-
l Se ealsoal(sSoch(elSicnhg,e1l9i6n0,g,p .19 6-10,8 pp.). 9 -108 ).
~o MoModeldseblasedbaosnetdheoTnhetorhyeoThef Gamoresy, of Games,
by means of which the hypotheses about
anddthtehoep roaptienrgamotindeglsmocal edeploslitciacl ed political
the balance of power are to be explored in ggaammeseasreaqrueitequdistiencdit esntintiecst. TehnetTihteioersy. The Theory
the game. However, the constituents of of GGaamems (e"Gsam("eGTahmeoeryThe") provridye"s)aprmeoanvsides a means
the model are represented in the game. Re- of dedescrsicbrinigbitnhegstrhategsictbreahtaevgioircofbeonheaovrior of one or
search on thehypotheses should, therefore, moreraectaocrstowhrso hwhoave tohamavke tcho imaces iknecocnh-oices in con-
ffllictctsitsuiatiuoantsi(oganmses()gwhamerest)hwhee payorfesthe payoffs
(potetnetniatl ioaultcomutesc)oamre easf)unacrteiona ofuncf the tion of the
Schel ing's gaming ap roach is in many choiciecsemas dmae bydealbypartailes tpao trhteiecosnftloict.he conflict.
ways similar to that of Kaplan, Bums, and TThehe GaGamemThe oThery moordyel imos nodrmelativse,noinrtmhat tive, in that
Quandt. The research focus of Schel ing is, it pprrescersicbersibtheschtohice corhcoimcebinoartiocnomf bination of
however, markedly dif erent. Where Kap- choiciecsewhs iwhich lecadhtoletahde bteosttphayeofbeusntdepar yoff under
the circiurmcsutmsancetsaonfcaegsivoefn caongiflivctensitucaotniofln.ict situation.
lan and his co-authors focus on the inter-
national system as a whole, Schel ing con- actor who wil always fol ow this best strategy.
centrates on the problems of bargaining and A political game (or simulation) is an operat-
tocdahpenrtmThesicnmeaoigcanbolefllnatinsectfoilrcbniatodoppone
tidence
on gaiarenconflict
national s thus
fromsianmsituations,
ulintternati
aneousmutual
ocharacteri
nal conflgainsisctit orcis lnssi.m.ocLnoSiitknbesc'icespewhtaantstniivsheibot,sehr)pc.eihavehmaWhemupyascotiicutreataaydll toabtstsioorurecogni
oueinstnsgnitodeun(saorasrtwia,niicoe'tparti foneasxtn,ledpbtgoalyhoirwetssiezh.:ee,st"OptlBhiIcoramnoersufiemakimnotlsrdhaiacmyltrapnsibnteotgiarsgrbo"aadherence
htuetnecosrsgaeueibst,o.stuei"hn(dtsrfhoteasyeorhussegcmhoniwhitoattfiaaelen-coinhyt-
683
borders s(e.(ge.,gt.h,ethYaeluYRalivuerRi) pvaert)akpeaorftake of

mutual losses can result from conflict inter- this qualaitlyit.yS.chSeclhinegl ipnogintpsoiunt,s out,
action. Or put another way, what one na- Limits hahvaevteo tboe qbuealqituatilvteatnivdediasncrdetde,iscrete,
rather thtahnaqnuaqnutaitnativteatainvdecaontdincuonust.inThuiosus. This
not necessarily equal to the losses of the is not jusutsatmaa mat ert oefrmaofkimang kvionlagtiovniosleatsiyons easy
nation it opposes. Conflict and interdepen-
force on one's own coreananders; it concerns
the need of any stable limit to have an evident
of the interaction system. It is worth noting
symbolic character, such that to breach it is an
that Scodel and his associates find that even overt and dramatic act that exposes both sides
when the conflict situation is objectively to the danger that alternative limits wil not
non-zero-sum, those involved in the situa- be easily found.
tion tend to perceive it and treat it as ff it The need for qualitatively distinguishable
limits that enjoy some kind of uniqueness is
were a zero-sum (pure conflict) situation enhanced by the fact that limits are generally
(Scodel, et al., 1959). found by a process of tacit maneuver and
negotiation. They are jockeyed for, rather than
their disposal the capacity to inflict severe negotiated explicitly. But ff the two sides
must strike a 'bargain' without explicit com-
munication, the particular limit has to have
is mutual, it is argued, mutual y bargained, some quality that distinguishes it from the
nondestructive strategies wil be optimal continuum of possible alternatives; otherwise
there is little basis for the confidence of each
for the participants. This is to say that
side that the other acknowledged the same
limit [Schelling, 1960, pp. 261-2].
self-limitation on the means of violence in
Schelling's argument can be summarized
It is Schel ing's contention that under

wil yieId the highest absolute payoff for
the participants) require the adherence to
limits on the capacity for mutual destruc-
there is a low level of communication be- tivenes . Given a low level of communica-
twe n the parties to the conflict, these lim-
its can be imposed through a proces of bargaining, f the means of conflict and
"tacit bargaining," ff conditions are right. the arenas of conflict exhibit discrete
The proper condition for the functioning
of this proces is the presence of funda- tions of the op onent wil come from per-
mental y dichotomous circumstances which ceptions of his activities; the more ambigu-
serve as symbolic boundaries. For example, ous this information source, the les likely
violent or nonviolent acts, nuclear or non- wil be the finding of mutual y optimal
nuclear weapons, or the participation of ad-
visers but not armed tro ps, are dichoto- means and arenas which contain no clear
mous conflict conditions with easily dis-
sources and therefore detract from the like-
liho d of the at ainment of optimal solu-
ing model which represents an at empt on the
part of the theorist, through the representation
of an empirical system, to provide himself with
"The game," which Schel ing developed
information about states of the real system.
strategic game which incorporates both
mucaboroauhtitocartiwhsy.nt"gomaneuver
plconfl
through
684 icgepahnrbtiearerastgeiasihyponratherposs csiatbnhlemos.thanes stthroethrough
ausghilyebxe- itselfpiwaathese(nan)ThedrtthOnsale.ismsservesebfunctigemodeyitscinanoetgnmiroltla.n,elthreeossnsttaiouatrrenuethedatdctefunctions:
osoinirmachi
aetbtrerhlheneptredcohno.tasmputnhe,Hee(1)enpritA.ine"comput-
cRICHARD esacdetieffect,
pecby,imetpcaptottm-ishnedgea
rBRODY
onone
e of 'bargaininofg by mane'bargaining
uver,' of signal- by maneuver,' of signal- militamilitary
ry and political factors. The machine and political factors. The machine
linling
g intentions, printentions,
oposals, threats, refusals, proposals, threats, refusals,
anand
d informationinformation
about one's preferences about one's preferences ing," i.e., the converting of decisions into
the "language" of the model and calculat-
words" (1961, p. 52). Here, as in the ing the results of these decisions; (2)
Kaplan, Burus, and Quandt game, there is"bookkeeping," i.e., the storage of the re-
structural similarity to the exercise devel- sults and the combining of them with previ-
oped by McClelland. The game employs aous decision-outcomes and the results stem-
map and includes players who vie for ter- ming from the opponent's decisions; and
ritorial control. It is basically a two-person (3) "display," i.e., the feeding back to the
(or two-team) game. However, accordingdecision-makers the outcomes of their own
to Schelling, expansion of the number ofand their opponent's moves. By performing
Schelling describes this research as "ex- serves as the interaction linkages between
the humans who serve as decision-makers
A good deal of trial and error wil be in-

volved in working out the analytic framework
? . . the object [at this stage of development] prises the decisional set ing for the partici-
is not to test a set of available hypotheses, so
ploratory experimentation, to manipulate the

game and its environment in an effort to bring pilot model is the fol owing. One contestant
the suggested hypotheses into clear relief, and or the other in a limited war recognizes that he
to rationalize the results in terms of a theo- is in an unfavorable position so far as objectives
retical model that can be identified within the
therefore either to extend the scope or to re-
structure of the game [1961, p. 63].
lax the limitations, or both, so as to improve
The participants in the Schel ing game his position. But if he unilateral y takes such
have available to them a full "spectrum of action the situation is reversed. The opponent
force," i.e., they can bring to bear in a con- or second contestant now finds himself in the
flict situation forces to the limit of their same situation that the first contestant was in
initial y, and must himself decide whether to
resources. There is no differentiation of the
extend further the scope or to relax the limita-
"weapons systems" in the game. With the
tions. This sequence of events repeats itself
means of destruction a continuous rather for a number of cycles each leading to a war
than a discrete variable, fol owing the logic of broader scope and/or higher intensity than
of Schel ing's theory, we would expect that the last. The situation is finaUy resolved in
territorial boundaries wil provide the points one of several ways:
limited objective and/or a pos ible los of the

struck. By comparison, in Har ison and
Le 's research (1960) determining the ef- war on the next higher level.
fects of two contrasting force spectra- (b) A compromise is reached in which both
continuous and discretemis the main ex- contestants ac ept limited objectives and lim-
perimental task. ited forces as being les undesirable than con-
tinuing the spiral of relaxation of limitations.
The "Stratspiel Pilot Model," developed
(c) Both contestants suc es ively relax the lim-
by Har ison and Le , is a man-machine itations and extend objectives tmtil al -out war
ensues.
CONFLICT RESOLUTION VOLUI~E VII NUMBER 4
tcearnmibnedc. oerced. Force ap lied by either barThisgainsituation
are (1)(2to) tomaximinmimizieze"poltheitdamage
--and ical control ed solutiowasns. simulated by depriving
to one'"--s deterrent capability versus a graduated de-685
WhWhich
ich of these alterofnatives athese
ctual y oc urs alternatives actually occurs
dedepends
pends on each contestaonnt's evaleach
uation of thecontestant's evaluation of the
terrent capability" (ibid., p. 16).
imimportance
portance of the expected gainsofrelativetheto expected gains relative to
ththe
e risks ofrisks
al -out war andofon his all-out
estimate of war and on his estimate of
hihiss op onenopponent'
t's views regarding this esvaluviews
ation regarding this evaluation one player of the "intermediate" range of
[H[Harrison
ar ison and Le , 1960, p. 6and
]. Lee, 1960, p. 6]. force and allowing the other player the full
spectrum. It was found that while "re-
ThThe
e objectiveobjectives
s in this conflict interactionin this conflict interaction
stricted force availability on one side tends
to moderate the competitive behavior . . .
operationalized initial y in terms of popula-
a player who was [so deprived] was at a
tion size and, ultimately, in terms of both
disadvantage relative to his opponent"
the population and geographic power bases
(ibid., p. 2).
The principal limitation of this model,
"condition"--which can be conceived o? as
which the authors forthrightly present, is
the economic power base. The nature of
the lack of any alternative to violence as a
these two variables is such that the deci-
mode of interaction. The model assumes
sion-maker is faced with the problem of
at the outset the existence of limited war.
considering, "whether it is in his best in-
This situation plus the fact that force is a
terest: (a) to increase his force allocation
continuous variable--"different types of
so as to increase his probability of winning
?orces and weapons systems are not spe-
the incident and thus increase his political
cifically considered, the player being con-
control, or (b) to limit his force allocation
cemed only with how much force he should
so as not to provoke his opponent into
bring to bear on his opponent, not what
using more force and degrading his con-
kind" (ibid., p. 16)--may account for the
dition" (Harrison and Lee, 1960, p. 7). finding that even though there are non-
The decision-makers have at their dis- zero-sum aspects to the conflict situation,
posal force by means of which the opponent cooperative behavior was minimal.
Schelling's theory can help us understand
player is measured in terms of potential why, when the ?orce spectrum of one party
fractional damage to the op onent's "con- was broken into two clearly separable parts
dition." Thus, for example, f one con- --the gap providing a prominent tacit bar-
testant ap lies force at a level of 0.50, he gaining point--competitive behavior was
has the potential of reducing the op onent's "moderated." Nonviolent interaction-alter-
condition by 50 per cent; whether this po- natives, ff included in the game, would pro-
tential is realized or not is randomly de- vide a still richer spectrum of mutually-
The initial research with the pilot model

was designed to "test the ef ect of limiting 1.7 Conclusio~
the spectrum of force available to one . . These thre examples are il ustrative of
of the players. If one player is constrained the international relations research which
to back up his bids for political control might be undertaken with simulations in
with either a ]arge measure of force or none any of its forms, al -computer, mixed, or
at al , and the other player is permit ed to al ,man. There is no fundamental reason
util ze a complete spectrum of force, then why Benson could not make replicative
one ap roximates a situation of mas ive runs, using an experimental design to test
lnacpioqmtsuieodtnawhtnletdreacnleaxrtiheoatinryseapinrsdoetbisplmoerecsocmuoirnmpelspi,crmabetsueldijaoutnxiohtnia-t, tbaiduotneatldhre"isrlaitmdieouvnlesa.toipoenrsars'hmoaredaelcso.mSmiimtumlaetnitontos
pevantages
686 short- qandcruci
offered.als examiA typinedcalpoidinstatodvantage
rThets. materi uenmathemati
t"quality
alycomponents
odological
data becontrols.
en tcontror'
chale theory
cas.eMathemati
tRICHARD
handaitn ctheadequate
ertaconstructi
icnA.alpltheory
ameth-usibolne
BRODY
somsome
e of Quincy Wrighoft's theorieQuincy
s about the Wright's theories about the disciplines (Festinger and Katz, 1953)-
dedeterminants
terminants of war. Nor would it be im- of war. Nor would it be im-
popossible
s ible for those interested infor
role-playing those interested in role-playing
crcrisis
isis games to set up mogames
re than case stud- to set up more than case stud- On the plus side, simulation offers the
iesies
of contempoforary realcontemporary
ity. In fact, in the reality. In fact, in the opportunity to do exploratory experimenta-
latlatter
er instance, it wouldinstance,
se m neces ary that it would seem necessary that tion in many areas of interest which other-
susuch
ch ef orts be undeefforts
rtaken in order to avoid be undertaken in order to avoid wise would not be amenable to systematic
bebeing
ing trap ed by the iditrapped
osyncrasies of the by the idiosyncrasies of the study. Beyond this, the technique forces
paparticular
rticular scenario used and the specialscenario
per- used and the special per- the scholar to come to terms with his theory
sonsonalities
alities of those who participated aofs ex- those who participated as ex- building. The complementary role of verbal
of operating models cannot be overempha-

comings of simulation as well as to ad- sized. Verbal theory guides the selection of
of the approach is its potential for misrep- aids in specifying the relationship between
resentation due to the lack of empirical components. As Schelling points out,
grounding of the model. Because little
To build a game of this sort and especial y
quantitative data has been gathered in the to build into the game particular features that
area of international relations (cf. Deutsch, one wishes to represent, requires that one de-
1960), the parametric loadings of the varia- fine his concepts operationally. A game . .
bles themselves may be quite inadequate. imposes discipline on theoretical model build-
By providing pressure for more or les de- ing; it can be a test of whether concepts and
propositions are meaningful, and a means of
finitive conceptualization, simulation may demonstrating so when they are. In the actual
force premature modeling in areas in which construction of the game, and in discussion of
empirical research has been sparse. In the the game's features with persons who have
use of all-man simulations, one may actually played it or observed it played, it has fre-
be overrepresenting the phenomena, in
concepts had to be abandoned when an ef ort
one's effort to obtain surface richness, to identify them (or to incorporate them) in
thereby seducing the researcher away from the game revealed that they were meaningles ,
his goal of parsimony. On the other hand, or innocuous, or that they rested on ines ential
it may turn out that the realities of inter- distinctions [Schel ing, 1961, p. 5.7].
A science arises out of scholarly dialectic.

simulation cannot but oversimplify and un- In its exploration with simulation, social
der epresent the crucial variables. For ex- science may have another method for use in
ample, the distortions of time within a the development of the study of interna-
mixed simulation present serious problems
which are but lit le understo d. How can
the weeks of decision burden, with its con- 1.7.a SUMMARY
To"pictorial,
'' ' '""verbal," and "mathemat-
ical" models of social processes have been
the economic factors? Just because one has
dif er in structure and mode of operation
one stil confronts in common those found
by experimenters within al social science the proposition that social realities can be
CONFLICT RESOLUTION V O L U l~il E VII NUMBER 4
iwisnigtotControlled
nihfinthiicternati
anstsictouatinotexperimentation
nalriobn,utpoliaointdozen
.ical processes.
or soinresearch-
interna- vehi
Thecle chapters
tempt
national toofbehavior,
exploratiwhich
explore ousing
n. follow
hypotheses about
are ainter-
simulation aspres-
the687
undeunderstood
rsto d via the stviaudy ofthescaled-study
down of scaled-down
veversions
rsions of thoseofrealthose
ities. realities. entation of the details of the author's at-
tional relations is out of the question. Faced
Chapter I is a presentation of the "prob-

ers have put their theories to work in simu-
lation. By means of a variety of techniques, lem" to be studied the so-called "nth-
country problem." Chapter I I is a discus-
these scholars have sought to gain insights
sion of the experimental design and the
adaptation of the inter-nation simulation
Whether programmed for high-speed for the execution of this design. Chapter
computers or set into motion by human
IV contains the execution of the design--
decision-makers enacting roles, a simula,
the exploration of a hypothetico-deductive
tional model requires an attention to opera- model which predicts effects of the spread
tionalizing theory that is unusual in social of nuclear weapons capability. This is fol-
research. In the end this development of lowed by Chapter V which contains a dis-
operational theory may be simulation's most cussion of the question of generalizing from
the simulation to the real world.
tdeigsor,hbnfyaMaalnrtrdhe.ixlposofPoa,tialnuosociittdhiescrosaeLterltysxtani(Bgcsinnaarn.alonytheldsbytaes, 2.eatsrou2sbcjmioSomeenvcetdmar.cbatonKeyekbrsesigwhinAsaitnhsceiuhdsmptiscbuoyrnvcseetoyrrns.noAltinihnegimtheprtphrimeosuairgoyhn
nelasorutsarThecFonhatupsoirahectngsihnriioebeterestosfrnatisenwvttHoaeiodlrIIofynaebzahiewacstudents
tCHAPTER
The "nth-country problem"
2.1 Introduction so-calso-called

ed "nth-country problem" alone is "nth-country problem" alone is
now now
nearly as extensive as that on al asnearly
- as extensive as that on all as-
military aspects of international politics is pectspects
of deter ence a few years ago. If of deterrence a few years ago. If
not a new phenomenon. Historians at least the mathe
gnitude of writing is an inmagnitude
dex of of writing is an index of
as far back as Thucydides have chronicled interinterest,
est, there is high interest in this topic. there is high interest in this topic.
WhatWhat
can be said of the content? can be said of the content?
the military-political transactions of states;
and they have raised the reporting of these
Deterfence Literature
at least as far back as Machiavel i have
been wil ing to acknowledge proces ual Writers on deterrence have been pri-
contiguities betwe n war and politics. Von marily concerned with behavior in the
Clauzwitz's dictum--"War is to be re- dyad. The explicit units of analysis--i.e.,
garded not as an independent thing, but as the constituents of this dyad--are the
a political instrument"--is essentially paral- United States and the Soviet Union. The
leled in the writings of such political phi- bloc-allies of these two nations are gen-
erally seen as acting through the bloc lead-
The acknowledgment of an intimate con-
international interaction system. Underlying
recognition that military preparation, strat- this point of view are two implicit postu-
egy, and level of capability (one kind of lates: (1) "If the Soviet Union is deten'ed,
power) are data of consequence in the so are its allies" and (2) "If the United
analysis of international relations. This im- States deters, its allies do not provoke."
pres ion is reinforced by the priority given Thus, cohesion of the two bloc-alliances is
these items as "bases of power" by the
There is, of course, a literature which
holds this "cohesion" problematical (e.g.,
drawn from the burgeoning literature on Wolfers, 1959; Knorr, 1959; Brzezinski,
many facets of contemporary military-po-
1960; Modelski, 1960; Osgood, 1962) but,
litical problems, e.g., deter ence, disarma-
ment, arms control, or the "cold war." The articlarticles
es wil have to be examined for his survey will have to be examined for his survey
thirty-od bo ks and articles on deter ence of thof
e psychological as umptthe
ions in the strate- psychological assumptions in the strate-
gic litgic
erature being done for theliterature
Institute of being done for the Institute of
1960a) now represent a smal fi'action of InternInternational
ational Order. Osgo d is not at empting Order. Osgood is not attempting
the total literature.1 The literature on the
of the magnitude of the bo k literature on this
1 Charles Osgo d (personal communication,
April, 1962) estimates that over 3,0 0 journal Brentano's Pentagon branch- it is awesome!
CONFLICT RESOLUTION VOL Ul~ E VII NUMBER 4
ttLhritoeWielrrmonuc presentati
by and large, there has been little overlap fusion n, place.
parallelsasbetween
direct proposi
quotatitoions.ns examcinleaiatriowannswir WhiolbvbeiocuihnsdilPreycaintded.circetases [Kis-
oatvtheurceenrultecaliernardyanotwedaab.plTheoenesx."ctepntdeionnsc,ywriamtongers der2.4 Propos
recases,searchtaking
are. opresented
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY 68o
tern; and (4()4t)htehfeafsiebaisl tbyiliotfynoucflenarucdliefa-r dif-

between it and the deterrence literature.
The concentration of focus on the United The propositions will be rank-ordered,
States-U.S.S.R. dyad is not without rationale, within categories, from "least specific" to
given a desire to understand the contem- "most specific." The propositions, in all
poraneous international system. In terms
of a statistical metaphor, the behavior of This is done in order to avoid possible in-
these two national actors is a major source accuracy of paraphrase and to communicate
of variance of the behavior of the entire the flavor of the literature. The hypotheses
system. To acknowledge this state of af- to be explored in this study wil be stated
fairs, however, it is not necessary to agree later on in this chapter; at the time of that
with Morgenstern's contention that "the
fundamental tendency toward bipolarity from the literature and the hypotheses un-
will remain a characteristic of the power
game . . ." (1961, p. 208). The presentation of these propositions
without comment should not be construed
on deterrene have not looked beyond the to mean that the author is satisfied about
their "truth status." Some of the proposi-
those who have examined the "2 + N" nu- tions are mutual y contradictory; a fact
clear-power world has been to multiply the which emphasizes the need for empirical
sources of danger in the present conflict
system by "N," i.e., to es ay the "nth-
country problem" qua problem. They rea- Problems Occasioned by the Spread
son es ential y in the following manner:
"There are serious problems facing the two of Nuclear Weapons
2.4.1 The spread of nuclear weapons wil
major powers, e.g., the stability of the de-
create problems of which our current
terrent relationship, the potential for ac i-
ones are pale shadows [Morgenstern,
dental war, and the difficulty of achieve-
1961, p. 136].
ment of a disarmament rapprochement; 2.4.2 The nth-nation problem promises
these problems wil be exacerbated with an to be practically an infinite series
increase in the number of nations with con- leading to a catastrophic divergence
[Inglis, 1959, p. 25].
2.4.3 . . . a further increase [in the number
of nuclear powers] would make the
2.3 Inventory of Propositions About the international situation and the achieve-
Nth-Country Problem Extant in the
ment of a general disarmament more
difficult [U.N. Review, 1960, p. 16].
In order to sup ort the above character- 2.4.4 The ?ewer fingers on nuclear triggers,
the fewer chances of their being
ization of the reasoning on this topic, an
pulled [Dory, 1960, p. 824].
inventory of propositions drawn from the 2.4.5 With many countries possessing nu-
literature wil be presented. These pro- clear weapons, the possibility of nu-
positions cluster in four general categories:
(1) problems oc asioned by the spread of singer, 1961, p. 242].
2.4.6 The spread of nuclear mfi tary capa-
nuclear weapons capability; (2) national
bil ty wfil almost certainly decrease
motives for se king nuclear capability; (3) the stabil ty of aleter ence systems.
ef ects of spread on the international sys- The greater the number of nations
690 ositions aweapons
rsmalce oafsleedrr.dependence
greater eiDnatidsi:fmoreounssiothann.,.ont.hwoulegreatertheserefodreweaponshave,power.plactoes
RICHARD A. BRODY
that
that possesspossess
the capabilitythe
of launch-
capability of launch- secursecurity
ity problems nations face. For problems nations face. For
ing
ing a nuclear
a nuclear
strike, the greater
strike,
the the greater the the spthe
read of nuclear weapons wil spread of nuclear weapons will
probability
probability that there will
thatbe a strike
there will be a strike almosalmost
t inevitably result in an ever inevitably result in an ever
[Brown
[Brown& Real, 1960,
& Real,
p. 25]. 1960, p. 25].
2..4.7
2..4.7To upsetTotheupset
existing [deterfence]
the existing [deterfence] and a decreasing dependence and
balance
balance . . . all [a. nation
. . all
not aligned
[a nation not aligned emphasis on conventional weapons
totothe the
U.S. or U.S.
U.S.S.R.] or
has toU.S.S.R.]
do is has to do is [Tucker, 1962, p. 6].
produce
producea sufficient
a sufficient
quantity of con- quantity of con- 2.4.12 The danger in the spread of nuclear
ventional
ventional nuclear bombs
nuclear
[Morgen- bombs [Morgen-
stem,
stem, 1961, p.1961,
137]. p. 137]. The leaders of the irresponsible
2.4.8
2.4.8[With[With
the spread ofthe
nuclearspread
weapons] of nuclear weapons]
the
theprobabfiity
probabfiity
of war by 'mistake'
ofas war by 'mistake' as do more effectively what they can
well
wellas by as
deliberation
by deliberation
is likely to in- is likely to in- already do in part. These leaders
crease.
crease.Vulnerability
Vulnerability
to attack and to attack and already have the power to start wars
the
theincidents
incidents
of 'mistakes' can
of. . .'mistakes'
be can . . . be that might involve the big two and
affected
affected by the wayby we shape
theourway we shape our so start a spiral into a world struggle.
posture,
posture,includingincluding
our decision-mak- our decision-mak- With atomic weapons at their dis-
ing
ingprocesses.
processes.
However, the prolifera-
However, the prolifera- posal, they might be more than ever
tion
tionof nuclear
of forces
nuclear
has an essential
forces has an essential tempted to adventure, and the wars
connection
connectionwith the difficulty
withofthesolv- difficulty of solv- that they fight wil be incomparably
ing
ingthesethese
two problemstwo simultane-
problems simultane- more destructive. But the greatest
ously.
ously.It placesItan places
increased burden
an increased burden difference would seem to lie in the
onon
a system
a system
of positive control,
of positive
and control, and fact that the line between conven-
therefore
therefore increases theincreases
problem of its the problem of its tional and nuclear weapons would
protection.
protection. The dispersionTheof nuclear
dispersion of nuclear have already been passed before the
weapons
weapons complicatescomplicates
the problem of the problem of great powers got drawn into the fight
responsible
responsible deterrenee bydeterrenee
increasing by increasing and the difficulties this would
the
the ambiguity
ambiguity
as to source of as
attackto source of attack create in confiningit [Hilsman, 1959,
[Wohlstetter,
[Wohlstetter, 1961, p. 371]. 1961, p. 371]. p. 160].
2.4.9
2.4.9ThereThere
is no questionisthatnoincreasing
question that increasing 2.4.13 With the multiplication of natdonal
the
the numbernumber
of nations armed
of with
nations armed with strike forces, the control problem be-
nuclear
nuclear weapons weapons
will add to the risks
will add to the risks comes especially acute. If many na-
that
that some some
local outbreak
local
of violence
outbreak of violence tions have the power of decision, and
will
willspiralspiral
into all-warinto
[Hilsman,all-war
1959, [Hilsman, 1959, if, in addition, each nation decentral-
p. 168]. izes its control to a multiplicity of
2.4.10 According to the 'statistical theory,' subordinates . . . it is evident that
the probability of a global thermo- the situation can get out of hand very
nuclear war increases as the number easily. The difficulty of distinguish-
of nuclear powers increases, because ing accidents from attack or, if the
attack is actual, in identifying its
(a) the larger the number of these
powers, the greater the probability source, would be enormously in-
that nuclear weapons will be used in
some conflict (both because of more increased burdens on control [Wohl-
opportunities and a greater chance of stet er, 1961, p. 363].
2.4.14 The problem of achieving interna-
irresponsibility); and (b) if nuclear
weapons are used in the conflict, the tional arms control wfil become vastly
risk of its expanding into a global more dif icult when the thre powers
war is greater than ff the conflict now having nuclear weapons are
remained nonnuclear [Ik16, 1960, joined by a fourth and then a fifth
and pos ibly more [National Plan ing
p. 391].
2.4.11 The increasingly widespread posses- Association, 1958, p. 39].
sion of nuclear weapons will have the As counterhypotheses the fol owing prop-
general effect of aggravating the
NucleartfesaCahrusaelmenfpiUnicaltebmisad,lelt.inetbdyescboNa.aemuUndecsteaimoconefusrin]athadrohimesvpeichnnocdasiveartbcebelbuoem-ptno- itcai[nmIonedndpressigletspihwiines,uptdse1959,h,onngrsuhtnthatoecaup.thldecer26]l'edkchaeterrent
oer.ndpt,groetowhlwieortvshetirinrcnudepen-
hfcatunhtuugerriese
691
2.2.4.15
4.15 Of itself, theOf
possessionitself,
of nuclear the possession of nuclear . . f.or increa.sed secur.ity . . for
for increased security . . . for
weapons
weaponsshould not increaseshould
the will- not increase the will- ag raggressive
es ive purposes. [Morgenstern, purposes... [Morgenstern,
ingness
ingness
of a minor nuclear
ofpowera tominor nuclear power to 1960,1960,
p. 137]. p. 137].
gogo
to wartoagainstwar
a major opponent
against a major opponent 2.5.62.5.6
The pos es ion of nuclear weaponThe
s possession of nuclear weapons
[Kissinger,
[Kissinger,
1961, p. 242]. 1961, p. 242]. has bhas
ecome a symbol of beingbecome
a great a symbol of being a great
2.2.4.16
4.16 The counterargument
The [tocounterargument
the 'statisti- [to the 'statisti- nationation,
n, influential at the council table influential at the council table
cal'cal'
argument]argument]
is that the diffusion of is that the diffusion of
nuclear
nuclear capability mightcapability
make the in- might make the in- 2.5.7 A motivation [for obtaining indepen-
volvement
volvement of major powers in global
of major powers in global dent nuclear capability] has been the
conflicts
conflicts appear to be moreappear
risky and to be more risky and
hence
hencerender it lessrender
likely. In other it less likely. In other dence' wil also allow Britain a free-
words,
words,nth country capabilities
nth might country capabilities might dom of action in world affairs that
either
either
help to deterhelp
local aggressiontoal- deter local aggression al- she could obtain by no other means
together
togetheror they might helporto isolate
they might help to isolate [Bums, 1961, p. 190].
locallocal
conflicts [Ik16,
conflicts
1960, p. 391; [Ik16, 1960, p. 391; 2.5.8 A nation may be attracted toward a
of.of.
prop. 2.prop.
4.10]. 2.4.10]. nuclear development of its own..
not only because nuclear weapons
2.2.5
5 Propositions
Propositions
About National About National mean power, honor, and a fuller
Motives
Motives[or Attaining Independent
[or Attaining Independent sovereignty in general in our time,
but also because in particular a na-
tion so equipped can now itself make
2.5.1 The spread of the fis ion weapons is the key decisions, to defend itself
itself a kind of 'fis ion proces '; each with these weapons and pull its allies
nation that acquires weapons induces into the fight [Mil er, 1959, p. 37].
more nations to get them too [Morgen- 2.5.9 The chief motivation for the acquisi-
stem, 1961, p. 137]. tion of nuclear capability within a
2.5'.2 [Our hopes for a strengthened Euro- coalition, and for neutrals who might
pean force and disarmament through join a coalition, is the payoff associ-
decisions [Snyder, 1961, p. 71].

al owed to set up gigantic nuclear
2.6 Propositions About Ef ects of the
Spread of Nuclear Weapons on the
provide the French forces with a
International System
nuclear armament [Ambas ade de
France, 1960, p. 3]. 2.6.1 The fortunes of nation-states, and
2.5'3 . . . f France wanted to maintain its likewise the structure of their rela-
military position as a major power at
the precise moment when inevitable in the instruments of military art
changes in the concept of war became [Bums, 1961, p. 187].
2.6.2 If the dif usion of nuclear weapons
ments [then she would have to de-
velop nuclear weapons] [ibid., 1960, of international relations wil be pro-
v. 8]. foundly altered [Kis inger, 1961, p.
2.5.4 .. '. for Britain and France, the de- V,40].
cision turned not only on strategic g.0.a . .. we might foresee a world in
but also on political and economic which gradually more powers became
considerations, prestige, influence in
Washington, commercial interests, and own right. . Assuming that the
scientific ambitions, as well as other control exercised by a bloc over its
reasons only remotely related to the constituent members would be very
Russian threat [Dory, 1960, p. 819]. tight, it would se m likely that the
2.5,5 Nations want [nuclear weapons] for proces would come to an end in a
a variety of reasons . .. for prestige very rigid world of separate, au-
692 Wagumresnatw, nPaucletar. .di.f(2u)sitohnemaopposy theirte- osNaNuitticoilnsoenarsarwilltmaCAceriaoffered:
woulcand tcyo'smetandoff'
necessi
you to beneaconnected
r tevery
o dictaothertwiintghpolnation
icy:
some- ejphsoarbeiwobenucivlumorelntdylgecaoarnnstable
tpoiInnwuedeertpto,oecbenotheudA.delednpBRODY
RICHARD atelnwdeaynst
extent
tartarchic,
chic, nuclear powers having nuclear
very powers having very experiences. Given a adidmimniuntuiotnionfof
litlittle
le to do with each toother . . Ido
f with each other .... If those pres sures, the ededpepnednednecncoefof
the nth power is alslosobobuonudntdotdoe-de-
in the world, there is not a pressing crease [ibid., p. 19]. .
2.6.8 OOtther things bebienigngeqeuqaul,al,cao cloitailontion
one else who can do so, too [Burns,
1958, p. 38]. that decision-making is centralized,
2.6.4 In that situation [when not only the i.e., the coalition is hierarchically or-
Soviet Union, but also China has nu- ganized [Snyder, 1961, p. 70].
clear weapons] the condition of nu- As counterhypotheses the {oilowing prop-
clear deterrence . . . would also in
effect come to obtain within the
2.6.9 A measure of independent [nuclear]
Communist system [Modelski, 1960,
strength for each of the NATO al ies
r. 30]. is likely to contribute to the cohesive-
2.6.5 Two limiting states of military art can
hess of the al iance . . . [Hilsman,
be imagined which in the circum-
stance of the late twentieth century 1959, p. 168].
2.6.10 A nuclear arsenal under national con-
trol [is not] a means for becoming
(1) the invention of a very cheap
independent of big power tutelage.
make-it-yourself, al -purpose deterrent
would invite every nation to be isola- A major nuclear power, confronted by
an nth country not backed by another
tionists in military affairs, fragment-
ing such al iances as NATO and the
strike preemptively. Thus nth coun-
situation, in which an adequate de-
on the support of a major nuclear
terrent weapon-system had become
so expensive that only the superpower [Kis inger, 1961, p. 242].
2.6.1 . within one bloc, the wider dis-
powers or very large blocs could af-
tribution of nuclear weapons poses
ford it. This second state of the art
great problems to the bloc: if the
would almost prescribe a world align-
control over the weapons is uncon-
ment much like that which obtained, ditional y given to some of its mem-
say, about 1956-58 [Bums, 1961,
bers, the bloc is les firm as an en-
r. 188]. tirety . . . nevertheles , the members
2.6.6 It has been argued that nuclear dif-
wfil stil be held together by com-
fusion must prove destabilizing for
mon {ears, ideals, interests, and
the reason that nations acquiring
treaties, the real force binding them
nuclear weapons either would actually
is the existence of the superpower
be in a more independent position
than before or would at least feel around which the bloc is formed.
that they enjoyed a greater degree of the fundamental tendency toward
independence. According to this ar- bipolarity wil remain a characteristic
of the power game [Morgenstern,
fore be expected to have a disruptive 1960, pp. 206-8].
ef ect on the cohesivenes and unity
of the major al iance systems- partic- 2.7 The Feasibility or Likeliho d of
ularly the Western al iance system
[Tucker, 1962, p. 13].
2.6.7 The nth country would [not] con-
tinue to be as dependent on the sup- There is relatively lit le in the unclas i-
port of its major nuclear al y, for the fied literature on the question o? the feasi-
nth power's dependence on its major bil ty and likeliho d of nuclear dif usion
al y is obviously no more than a func-
coming about. The National Plan ing As-
tion of the pres ures brought against
it and the consequent insecurity it sociation (NPA), in two pamphlets on the
p.1
httc1965.hoalausttb'h"e]"snwieExperirtelidmuceoesmd,enceaebdywimiawitsbhotihmuknsparce[htseongttpredirlhyeapt"uenbcructilifslohrenseadr- isntieymwadulysiftaestaidrotyenn.stwoiatlrewiledmtshewiirnglteh?bee blaopcpros whpirichatceomto-
progres
potenti asperhaps
l iyn capabl
It is nine
tween thiands noteworthy
areasixteen
e ofwereindependentl
appearithatnwilltheyg wiattaihave
nations thinn- neresources,
NPA wi"inalrvulprobabl
future . ye";alother
nerablincluding sophistictheatedhardlycapabi
so begeography,
beyond weapons li693ty
avail-
subjebctj(e1c9t58;(191609),5es8tim; a1te9s6th0at)b,e- stimates that be-

able to other powers for years to come [1962,
nuclear weapons by 1970 if the present p. 4361.
pattern of spread continues. General Gfffin's prediction hinges on the
qualffiers "expensive,' ......... and
complicated,
did not include Israel among those nations
system characteristics, e.g., missile delivery,
ing nuclear weapons; yet stories of Israel's high warhead cost-effectiveness ratio, mo-
bility, and positive command and control,
a year2 of the publication of The Nth-
Country Problem (NPA, 1960). Estimates of all but the two "superpowers" in the
of the status of China's nuclear develop-
ment have also proven inaccurate. A few But do these differences in the "state of
years ago "1961-1962" were the dates pre- the art" of weaponry make a difference?
dicted for China to test a nuclear "device." How crucial is level of development-sophis-
The current estimates seem to favor "1964- tication to the predictions contained in the
propositions in sections 2.4, 2.5, and 2.6?
argues for caution 'm accepting estimates of It is hoped that the tools of analysis and
who wil gain nuclear weapons, when.a measurement developed in this study of
Even if we were to feel certain about
these estimates, questions about the level examining these questions in the real world.
of capability attained would stil remain.
Thus, it is possible to agree with Patti Dory 2.8 The Approachto this Study
that "without international agreement and The approach taken in this study stems
a display of national self-control uncommon from the as umption that to conceive of
the "cold war" as a pat ern of interaction
is a useful heuristic. The "cold war" can
quency" (1960, p. 818). However, ac- be thought of as a system of action with
ceptance of the plausibility of this prophecy,certain describable regularities- -interaction
which goes to emergence of nuclear nationsproces esmwhich serve to define it as a
rather than level of development, does not
preclude agreement with General Giftin Proce ding from this heuristic, we may
ask the fol owing question: I~ one charac-
[In]themiltarymatrix. postulatedfor teristic oi the "cold war" system- -the num-
thenearfut re. ther areonlytwogreat ber of nations with nuclear capabil ty- -is
nuclearpowerspos e singveryexpensiveand changed, wil the system survive or wil a
verycomplicated,butinvulnerable,strike
forces.Membership nthiscompanycould
Another way to ask the question is to
onlybeachiev dbythecontrol fenormous
query the cruciality of the nuclear capabil-
~E.g,gN.e,wYNorekwTimYes,oDreckmTberi,m19e60s, December, 1960,
prise the "cold war" system. If the nuclear
aSomeofouruncertaintyonquestionsof dif erential is crucial to the interaction pat-
feasiblityandlikeliho dofspreadtakingplace
tern we cal the "cold war," then we would
on the topic (Beaton & Mad ox, 1962). expect that the spread of nuclear weapons
beiatlhaenrcealeiacnheostyhsetrm, ,thine magrenayter stphectsprthe- pfcooalrnacre.beeacdeh fotihneerd. as that behavioral state
res694theareHereiultes.xenpet,forth. taThil instetheoreti
rtheimenformal vention-ceal-stmodel
rproperti whihe inchducewasde-d Communist alternative to membership in
RICHARD A. BRODY
(t(the
he nth-counnth-country
try situation) would alter the situation) would alter the HHoowever, King indicates that, "thehenono-n-
papattern
t ern of interaction.of interaction.
TThe
he nul hypnull
othesis (H0)hypothesis
is that this (H0) is that this the United States collective defense net-
chchange
ange wil not matewill
rial y changnot
e the statematerially change the state work is not membership in a 'third force,'
ofofthe sythe
stem. Thesystem.
experimental hypotheThe
sis experimental hypothesis but neutralism, not the creation of another
(H(H1)
1) is that aismarkethat
d change in athe marked
num- change in the num- power center but the abnegation of power"
beber
r of nucleofar nationuclear
ns wil be reflectednations
in will be reflected in (1959, p. 113).
aamarmarked
ked (i.e., "step-lever')(i.chanege.i,n th"step-lever'
e ) change in the Taken together, these propositions indi-
,'c,'oldcwaoldr" systewar"
m- a new steasystem--a
dy state wil new steady state will cate that while nations may avoid alliance
with either bloc, they do so at the price
Figure 2.1 is a formal presentation of H~. of reduced international influence, at least
in the military sense. Given the absence
scribe the "cold war" system are stated and of a third force, in many significant respects
linked and the predicted consequences of the future of the bipolarity becomes the
future of the entire system. It should,
increase in the number of nuclear nations- - therefore, not do violence to reality to con-
sider the "cold war" system as tightly bi-
posited in advance of the experiments and
forms the basis for the exploration reported Betwe n the blocs there has developed
in Chapter IV. a state of high tension and, at least con-
ceptual y, antecedent to this tension are
2.9 A Model o[ the Cold War System mutual perceptions of hostility. "Tension"
In gross terms the "cold war" system can
be described as "bipolar," i.e., there are which is characterized by- (1) a focus of
two major centers of military and economic at ention on an object, (2) where the ob-
power toward which other nations are at-
ject is charged with affect, and (3) where
tracted. The reasons for attraction in a stereotypy and other proces es of cognitive
given direction would include historic and simplification characterize the as ignment
economic ties, fear of suzerainty being of meanings by members of one bloc vis-~t-
established by one superpower or the other, vis members of the other bloc (Milburn,
the actual presence of one or the other
1961).
major power's troops at the end of World Where the object of focus is charged
War I , and a feeling of being threatened
with negative affect---as is the case here,
by one or the other nuclear power.
with hostility as the tension antecedent--
Moreover, while nations exist outside of
behavior tends to take the forms of aggres-
sion and/or avoidance (Frank, 1958).
bipolarity may be considered "tight." "At
Tension and hostility, thus, tend to rein-
al times," ac ording to Wolfers, "weak and
vulnerable countries have sought refuge in
neutrality and found it a source of protec- This "tension-hostil ty" cycle fe ds into
tion, provided their stronger neighbors haveinter-bloc relations. The proces exhibits
held each other in check... the more some of the characteristics of Newcomb's
"autistic hostfi ty cycle."
American and Soviet power have come to
Autistic hostil ty is . . based upon a more
miumon neutrality" (1959, p. 10). or les fixed perception of self-other relation-
munication decreases; as communication NUCLEAR DIFFERENTIAL / 695
(N.B.: C1 indicates the ith condition in a chain of conditions which

is hypothesized as being relevant in the transformation of a steady
state A into another steady state B.)
WITHIN THE BLOC (GIVEN)

SENSITIVITY
SENSITIVITYTOTOREALREALORORPERCEP-
PERCEP-
TION
TIONOFOFOUT-GROUP
OUT-GROUP(OTHER (OTHERBLOC)-<-
BLOC)-<-
HOSTILITY
HOSTILITY(C1(C1ANDANDC2)C2)
DEPENDENCY
DEPENDENCYRELATIONSHIP OFRELATIONSHIP OF
1
INCREASE
INCREASEIN INTI~RBLOCIN
TENSION
INTI~RBLOC TENSION
NONNUC
NONNUCLEAR BLOC MEMBERS
LEARTO'<BLOC
-'fr..~, MEMBERS TO'<-'fr..~,
NUCLEAR
NUCLEARBLOC MEMBERSBLOC
(C.4) '"-' MEMBERS (C.4) '"-'
--.... INCREASED INTRABLOC COHESION (C5) ............ ---
AD HOC HIERARCHY WITHIN BLOC (C6)
EXPERIMENTAL INTERVENTION
NUCLEAR PARITY
PARITY (GIVEN)
(GIVEN)
INCREASED DISSATISFACTION REDUCTION OF PERCEIVED OUT-

.......-- --.....
WITH AD HOC HIERARCHY (C8) GROUP THREAT (C7)
INCREASED INTRABLOC TEN-
SION (C9) !A
REDUCED INT BLOC COHESION
(C10)
DISSOLUTION OF AD HOC
HIERARCHIES (C11)
RELATIVE FRAGMENTATION OF BLOCS;

A "STEP-LEVEL" CHANGE TO A NEW
"STEADY STATE" (C12)
Fxc. 2.1. Hypotheses linkings for INS-8.
shipships,
s, the persistence of which is maintaithe
ned by persistence of which is maintained by into intra-bloc relations. The presence of
bar barriers
iers to communication (overt and/or covert) to communication (overt and/or covert)
perceived out-group hostil ty tends to re-
withwith
the object of hostil ty [Newcomthe
b, 1947, object of hostility [Newcomb, 1947,
inforce the dependency relationship within
p. 73].
the blocs and to increase intrabloc cohe-
In other words, as hostility increases, com- sivenes . The hostil ty becomes one aspect
of a threat situation with which the non-
decreases, hostil ty increases. nuclear bloc member fe ls inadequate to
The "tension-hostil ty" cycle also fe ds deal. This inadequacy leads to increased
at racted toward membership in an al iance. hierarchical blocs would have more or les
(1959,Thistother
theUni
from
leader.edthep."dependency"
9)member
. Miandl erdisparity
States
nuclear itnsalsodiEuropean
nations. catesstems,
that,introduc-
"ThebetweeninAl"thei
iethes]r" indicates
lpart, woul
propertid expect
ethats associa"authoritative"
hiaetedrarchiwitchalabl"wheel
oc to exhi" com-(i.bei.t,
decisions
dependence
dependence on the
on nuclear
the nuclear
armed bloc-
armed bloc- Hierarchical
Hierarchical organization,
organization,in
inthis
thiscontext,
context,
those decisions which commit the bloc to a

given international policy) tend to flow
"superpower" who heads the alliance and from the bloc-leader rather than from the
other bloc members; in addition to yield-
tion of nuclear weapons," Wolfers asserts, ing information about the locus of decision-
"helped to tighten the bonds between [the making in the bloc, the hierarchy of or-
ganization yields predictions about the pat-
tern of communication within the bloc. We
existence greatly increased . . . the inter-
dependence of nations in the alliance"
(1959, p. 37). munication net; we would expect that non-
Morgenthau asserts that, "a nation will
shun alliances if it believes it is strong "al -channel" communication nets.
enough to hold its own unaided" (1959, In summary, the "cold war" system is
p. 185). By in~plication, ff a nation be- composed of two hierarchical y-organized
lieves that it cannot hold its own, it will be bloc-al iances with the leading nations in
the bloc posses ing a virtual monopoly of
Thus, the dependency born of perceived nuclear strike and counter-strike capability.
out-group hostility is increased by percep- Moreover, there is tension-generating hostil-
tions of military. inadequacy vis-~t-vis the ity between the blocs which reinforces co-
hostile threatener. The attractiveness of hesion within the blocs and which itself is
bloc membership from this source also reinforced by this cohesion. The relation-
heightens intra-bloc cohesion--that is to ship within the blocs is one of dependency,
say, cohesion varies positively with the at- born of necessity. In part, the dependency
tractivehess of group membership. relationship is clue to the hostility of the
This within-bloc cohesiveness is itself a nuclear armed leader of the opposing bloc
factor in exacerbating between-bloc hostil- and in part to the nuclear disparity within
ity. The process operating here is described the al iance. It is in this context that the
by Pepitone and Reichling as follows: effects of the spread of nuclear weapons
wil be essayed and explored.
Members of cohesive groups provide each
other with strength and support, which enables 2.10 The Cold War System and the
them to overcome internal and external re-
straints against retaliation in the event of attack Spread of Nuclear Weapons
. . . highly cohesive groups . .. will be less It has already been asserted---H~---that
restrained when under attack than will rela- the result of an increase in the number of
tively less cohesive groups. The greater ability
nuclear nations, in the above-described sys-
of highly cohesive groups to reduce restraints
wfil be reflected . . . in a greater volume of tem, will be a marked change in the "cold
expressed hostility and a more direct expres- war" steady state. In this section, an at-
sion of hostility with respect to its source
tempt will be made to explicate the nature
[1960, pp. 141-2]. and direction of this change.
The combination of dependence and In two respects the relationship of each
cohesion has produced blocs which are nation vis-~t-vis its own alliance and toward
more or less hierarchically organized. the other alliance can be expected to be
CONFLICT RESOLUTION V O LU1Vf E VII NUMBER 4
hocredithoughts
temstiblfole i]nl,toewthey
ntfrom
] migtheseht havechanges. mactaptahniraktibei,nnyoggtewihrxtheoepdamageltrhctwiiahneldilnitagdhnoremains
followed the follow- inflicting sece.cbebreliaThusonsfceg.teintact.hcq,teutheadhil,byepMoreover,
"ropawnuarrecrahbliecsar"l
chachanged
nged by the achievement of indepen- by the achievement of indepen- lingg.. ThThee potpoenteianltaigalreasggror'sescapoarb'sil ctyapoafbility of
dendent
t nuclear capability: (1) tnuclear
he reduction capability: (1) the reduction
of of
the dependencthe
y on the bloc-leaderdependency
, and on the bloc-leader, and the former weakling's perception of the
(2) (2)
the reduction ithe
n the level of percereduction
ived in the level of perceived "bully's" hostile intentions also are un-
outout-group
-group threat. Other changes in the sys- threat. Other changes in the sys- changed. But the credibility of the "bully's"
aggressiveness is reduced as a function of
To il ustrate the process of this change the former weakling's felt capability to re-
the following is offered. taliate in kind. To this extent, the percep-
"The Parable of the 97-1b. Weakling" tion of threat is reduced.
Doubtless we all remember the "Charles Reduced also, it is hypothesized, is the
Atlas" ads which pictured the "97-1b. weak-
willingness of the former weakling to re-
ling" confronted by the bully on the beach.
If we could have recorded the weakling's main part of an alliance which he feels no
longer functions for his protection, at least
ing pattern: to the extent of the costs involved in main-
"You are able to hurt me"; [the threat is
leads us back to Morgenthau's proposition:
"You appear to want to hurt me"; [he perceives "A nation wfil shun al iances if it believes
that it is strong enough to hold its own
"You wil hurt me if I can't stop you and I am
unable to hurt you back"; [he fe ls that he unaided or that the burden of the com-
hasn't the capability to deter] mitments resulting from the al iance is likely
"I can't stop you or hurt you back without to outweigh the advantages to be expected"
help. Therefore, I'd bet er se k help- I'm
(1959, p. 185; Caplow, 1956; Gamson,
wil ing to pay the cost of being al ied."
19~9).
At this moment, who should come along to Another aspect of the "cold war" system
save our weakling from his dif iculty but
Charles Arias, himself. Arias, in addition to
forestailing at ack by the bully, helps, either by
example or actual instruction, the weakling organization of the blocs. If, as has be n
build himself into a g10-1b. non-weakling, the as erted, a raison d'~tre for the ac eptance
equal of anyone around. of a hierarchical structure is the nuclear
Now, when our hero comes upon the same
disparity, it should fol ow that the achieve-
bully, his thoughts might fol ow this pat ern:
ment of nuclear parity wil raise questions
"You can hurt me";
"You appear to want to hurt me"; about the equitability of that hierarchy
"But, I can now stop you because I can hurt (Prop. 2.6.7, above).
you as much as you can hurt me; you don't Moreover, if we ac ept with Las wel
like being hurt anymore than I do";
"I can now stop you without help. Therefore,
I no longer need to pay the cost of being al ied go d index of participation in decision-mak-
with Atlas." ing, then we would expect that as this
This fantasy is designed to il ustrate the (nuclear) base Of power increases for the
hypothesis that only some aspects of a former non uclear bloc members, their wil -
threatening interaction situation wil be ingnes to remain in a subordinate decision-
modified by increased capability. It wil be
noted that not al elements of the threat sit- Mil er was quoted above (Prop. 2.5.8)
uation were changed by the at ainment of as as erting that "a nation may be at racted
independent capability by our former weak- toward a nuclear development of its own
nuciiPetennreenoCsawhiisplssfooieiwithnuoaed.nrpnscchonor,
cdeciparticular
tpower,
self di'lhtsdee(Chihhsdcaoowever
vtheseauwietnewmnlination
dsoaetfpandhlaatmeeconuclaweapons
tsrdsiioenadbgthasounifuller
oeawicarsifenequipped
itdnucmachtnati
wiant)phandtrsovereignty
elhkeo"caeiwoulCnrdnhitoiovwerwoultemcanerhpullddamuniearupprob-cCo"cdnowohotniuiitsmcbeoinstan]--l changed.
tiacnbeiioKaplntvvheealliances
bership
The dd),.ptatrWhehnhunder
erdeeihypothesi
acstytehmaysfdterarfomrmhierarchical
tasurvive
hgwitmehlRICHARD
ezsntesprooetapatdthatithisbyoansbis.nltstructure.
ygatraumatic
trebelBRODY
A."the sowhiecmlwieblcsshle
. ...not onlynot
because nuconly
lear weapons meabecause
n nuclear weapons mean bership wil lbebel slesnedn,edpa,rptiacrutliacrulyarmelymme- m-
general in our time, but also because in

input but they will not survive it un-
itself make the key decision, to defend it-
allies into the fight" (1959, p. 37). This hierarchical in structure the bloc actors are,
the more possible it becomes for national
ably be only one of many crucial decisions actors to dissociate themselves from the
bloc and the more impetus there is in the
capable of exercising influence. direction of some sort of 'balance of power'
The capability of decisional indepen- system" (1957, p. 40).
dence and the increased unwil ingness to In 'this model, the phrase "in the dh-ection
remain subordinate in the bloc (i.e., inof some sort of balance of power system"
has been translated into the word "frag-
structure of the bloc) wil , it is hypothe- mentation." The expected change is the
sized, lead to an increase in intra-bloc fragmentation of the blocs (Prop. 2.6.5,
This hypothesis is consistent with Model- emerges wil lo k like Kaplan's "balance
ski's prediction about the effect of China's of power system" (1957, p. 30if) or "unit
veto system" (1957, p. 50if) wil depend
international system." He as erts (Prop. on decisions which are oc asioned by the
2.6.4, above), "in that situation [when spread. If national strike forces are main-
the condition of nuclear deter ene , familiar the trait veto model. If, however, spread is
from Soviet-American relations, would also fol owed by disarmament, the balance of
power model wil probably describe the
system. In either case where there were
munist system" (1960, p. 30). When we
two power centers there wil now be multi-
add to China, Czechoslovakia, The German
ple power centers (Prop. 2.6.3, above).
Democratic Republic, and Poland-- the
In summary, the expected change in the
other Communist nations estimated by the
"cold war" system is the fragmentation of
NPA tobe capable of having nuclear weap-
the blocs which comprise it. This (step-
ons by 1970- tension within the bloc level) change wil be prompted by the in-
should reach substantial levels. creased independence of the formerly non-
nuclear national actors after they achieve
group" threat, the increase in intra-bloc nuclear capability; the increased sense of
independence and a lowering of the per-
in the cohesion which previously charac-
terized the bloc-al iance system (Prop. les en the intra-bloe cohesion and eventuate
2.6.6, above). The at ractivenes of mem-
CONFLICT RESOLUTION VOLU/~{E VII NUI~IBER 4
"onCHAPTER
nraatturuyraailnle"gxeptxeoperaimrtiIIIemfnietcnimuatlsyresearch
situation:
Simulation swetrbeand
aprseodluegfchtetwi.the
heth"ctohldesign
de nth-country
Frecannbechproducti
nuclearvdee ofvelhypotheses
opments arande undeofferr-
33..11InItnrotdrucoduction tion ststatistical

atistical model in mind.1 Butmodel
these cases in mind.1 But these cases
TThehe devdeelovpemleonpmet of whnta ofit iswhahopetd ist is hoped is
potential reality checks for ideas and mea-
a cochoeherenrt,eantal,ytainc moalydetliocfmodethe ef lecofts ofthe effects of
suring tools developed in the simulation.
theesprsepadreoafdnuofclearnucwealpeonasrdweoes anpotons does not
elimminaitneathe ptrhobel mproblof altemrnatofive faolrtmeru-native formu-
Currently, studies of the British and
latioinosnsandamundtuamutl y counatrladyictcorynthrypaodtihc-tory hypoth-
esese.sT.hThee probprobl
lem of feimndinofg mefiansdifonrg means for way. Harvey deWe rde, of BAND Corpo-
evalaulautiantgitnhge ptlahuseibipll tayuosfibalitleirtnyativofe alternative ration, has been engaged in intensive his-
fformul
ormulatioantsiroensmainres wimthaiuns. with us. torical research on Britain's nuclear pro-
DDeeprpivriedv,eads ,weasrewe, of aonr-eg,oinofg, con-ruciagl,oing, crucial, gram. Ciro Zoppo, also of RAND Corpora-
tion, has produced an interesting paper on
choice betwe n waiting for the nth-country French efforts to develop~ the force de
situation to evolve in actuality or work'mg frappe (1962). The UCLA National Se-
with the two extant, approximate cases- curry Studies Program, under the direction
of Richard Bosecrance, has been conduct-
i.e., the British and French developments-
ing a series of seminars on nuclear diffusion
war" system and then experimental y in- with emphasis on the British and French
ducing the spread of nuclear capability.
programs. None of these research efforts
is completed; at their present stage of de-
In the final analysis, the confirmation or
velopment, emphasis has been on the do-
disconfirmation of the model wil depend
upon the ef ects of nuclear dif usion in the
mestic political and economic proces es
real world; in this sense we have no choice which have shaped the emergence of nu-
clear capability in these two countries and
but to wait for the situation to evolve. But
on the military implications for the Western
if our interest is in evaluating the plausibil-
Alliance- the implications of these develop-
ity of the model prior to the actual oc ur-
ments for the cohesivenes of the Western
rene , substitutes for a ful -blown, crucial,
Alliance and its effects on the "cold war"
have received les :emphasis. Studies on
The two alternative ap roaches which
these points are precisely what is needed if
are available- -case studies of extant devel-
the model presented in Chapter I is to
opments and simulation- are by no means be confirmed or disconfirmed'
mutual y exclusive enterprises. The paucity
of case material augurs il for the explora- On the problems and benefits of using case
tion of the nth-country situation with any materials, se Paige (1959).
The particular gaming approach em- economic advisor to the CDM and performs prsuersefnoertxmedrmotseydrmbebtyohltainhcaeoln"yoep(roaltes,ihteio.rgn.t,"hCDbaenMcobmaynedas
7OO RICHARD A. BRODY
With
Withonly two real worldonly
cases to draw two
participants. Theparticipants.
"Central Decision-Maker" real world The
case"
upon--and
upon--andboth of these being in one bloc both (CDM) represents(CDM)
the chief-of-state and of these being
represent
--at
--at
this time only suggestive
this evidencetime
can perforonly
ms the experforms
ecutive function of govern- the exec
suggestive ev
be be
adduced foradduced
or against the model. Gam- forment. He maintament.
ins his posor
ition by satis- He mainta
against the mo
ing,
ing,
therefore, remains
therefore,
as a potentially use- fying those whofying
validate hisremains
of ice-holding. those
as a who v
poten
fulful
approach. approach. The "Internal DecThe
ision-Maker" (IDM) rep- "Internal Dec
resents the direcresents
tor of the budget; he is the direc
ployed-the Northwestern Inter-Nation
Simulation (hereafter, INS )--was chosen such other tasks as are assigned to him by
because of the author's experience with this the CDM. The "External Decision-Makers"
operating model and because it was felt that (EDMx and EDMy) represent the foreign
the INS was flexible enough to offer a rea- relations structure of the nation and per-
sonable prospect of providing a model form the corresponding function. They are,
world in which the spread of nuclear weap- as are the IDM and the "Decision-Maker
ons could be examined. with respect to Force" (DMF), dependent
upon the CDM for their continuance in of-
3.2 The Inter-Nation Simulation: flee. The DMF is the military advisor to
A Brie[ Description the CDM; it is his responsibility to advise
Since detailed descriptions of the /NS on matters of strategy, allocation of re-
exist elsewhere, it would be redundant to sources to force capability and defense. The
present more than a synopsis here. For full- division of labor among these decision-
blown discussions and descriptions of the makers is a matter for the CDM to decide;
INS, the reader should see: Guetzkow the "cabinet" holds office at the pleasure
(1959); Brody and Noel (1960); and of the CDM. Outside of the "government"
Guetzkow, et al. (1963). is an aspiring Central Decision-Maker
The /NS is comprised of a group of (CDMa) who represents the leadership of
groups, i.e., a system of interacting units, competing elites in the nation who aspire
where the units are teams of individuals. to become the group in power.
The unit-actors are styled "nations." The The number of participants who fil these
individuals who act for the nation are its roles wil vary with the number of subjects
,,'l ? ? I 7,
ctemslon-maKers.available and the problem being studied.
As a convenience, before entering a
The six key roles are always represented
cussion of the interaction between na
but one participant may be called upon to
the "domestic" structure of a single n
will be explicated.
The nation inIDM. The CDMa rolethe
has often be n rep- INS is a syste
roles. Some of these roles (the "dec
participant); on these oc asions the pres-
making" roles) are occupied by live
?? 7> ?, 1 ? . 7,
viduals--the participants or
part of the over-al pres ure exerted by the
the other roles (validation roles)
insofar as they"validators" on the decision-makers. On provide a sour
straints on the only one oc asion has the number of partici- freedom of act
pants per nation be n les than thre ; in
live decision-makers.
The internal that instanorganization
ce, two decision-makers per na- of t
is determined, tion Were used with a specific research pur- in large measu
CONFLICT I~ESOLUTION VOLUME VII N U l~f B E 12 4
pose in mind (Hermann and Hermann, been
termedon the
"forcedomesti uitnuByder,macaonndsiipcdoulenrasatutiimnogner.thdesme apandrsa.metric values
c settin" gTheof onedecision(any
capability.
the lthe
evel of the nation's militarlevel
y strength- of the nation's military strength--
1962).
The "validators" are coneeptually the form also informs the CDM about the min-
group or groups, within a nation, to whom imum amount of basic capability which
decision-makers (especially the CDM) are must be allocated to generating consumer
responsive and responsible. The office- goods and services. Thus, the decision
holding of the CDM (and, consequently, form gives the CDM basic information
the office-holding of his "cabinet") is func- about the nature o? his domestic environ-
tionally related to the level of "satisfaction" of ment at time' . . . . which is ~,--rotormatron
the validators. Since not all decision-makers vital to his decision-making at time "t.;+l."
(in real nations) are equally responsive and Emphasis, thus far in this synopsis, has
responsible to their validators, we have
allowed the degree of relationship to vary one) nation in the INS; this emphasis is
in the simulation. This degree of relation- not inordinate. In the INS, in contradis-
ship (termed "decision latitude") is ex- tinction to the other operating models pre-
pressed in terms of a ten-point scale--the sented in Chapter I, an effort has been made
higher the deeision latitude, the less sensi- to represent the domestic constraints on the
tive is the CDM's office-holding to the level [oreign policy-making process. These con-
of validator satisfaction. straints are presented to the decision-maker
Validator satisfaction stems from two as parametric values for a set of (hope-
basic sources: (1) from "consumption," fully) prototypic variables--public opinion,
i.e., from expenditure (by the CDM) for economic level, military level, decision lat-
consumer goods and services beyond the
minimum needed to sustain the population;
and (2) h'om felt "national security," i.e., at the outset of the simulation (i.e., time
from the level of capability (military and "to"), paral els to real world nations can be
economic) of the nation and its allies in drawn. In this manner, role-playing of one
comparison with its strongest nonally and sort (i.e., where the participant is charged
that nation's allies. ? with the responsibility of reproducing a
Time, in the simulation, is expressed in "predicted" strategy for a nation which is
terms of decision periods. A period (h'om not his own, but about which he knows a
50-70 minutes of real time) is a unit great deal) can be avoided. The partici-
of domestic, fiscal decision-making---other pants in the INS are not instructed to "re-
types of decisions are made more or less act the way the Soviet Union would react
continuously and are unprogrammed with to the strategy of the United States"; rather,
respect to game time, they are instructed to "react to the situation
At the beginning of a period, the CDM in which you find yourself." To further
receives a "decision form." This form con- reduce pres ure for this type of role-play-
stitutes the CDM's primary source of ining, nations in the INS are not given real
formation about his domestic, political-eco- world names. A set of imaginary nations
nomic setting; on it are found the results (viz., Algo, Erga, Ingo, Omne, and Utro)
of his previous period's decision-making. have been created, With eharaeteristies
These "results" include data on validator which vary with the research problem
satisfaction, the level of the nation's econ-
omy--expressed as "basic capability"---and Despite the emphasis on the domestic
moreover,
nati
702 In the "exploratory," "feasibility," and lvbaignerhypotlitgeonspg-mrtcwaenumbeonhtdse"isdybipseyrraoebjsmadeof,out.ctwaksnaiwasbentgaishocannsivoutimcorwoullropedlaiuirnscdhectdahgi.pePvabldyaeyiusnldtdyea,,-
tmayseektSioimn,leuoaornlvremain
anetmaydigitohanvewars"etwpatradeoforeiorlatrAdtdmayipeacecmigprewipinatntsbeaithh.sord"eotherl;aismalEengage
xfliptraedlencesonatiraorstiinopomayototalnsnswar;andib;belae
RICHARD A. BRODY
constrconstraints
aints on decision-making, the INS is on decision-making, the INS eieighth
ghis
th run ing of therunning
INS. INS-8 was it, of the INS. INS-8 was it,
an opean
rating model of the interoperating
national sys- seself
model of the international sys-lf run sevenrun
te n timesMiseventeen
.e., sevente n timesMi.e., seventeen
tem; ptem; ititerations
erations of the same startingofsituationthe
rototypie clas es of international be- prototypie classes of international be- with same starting situation with
haviorhavior
exist in the model. Thus, a nation exist in the model. Thus, a nationdidiffering
f ering groups of decision-mgroups
akers. These of decision-makers. These
ruruns
ns were designwere
ated INS-8-I thrdesignated
ough INS-8-I through
INS-8-XVII.
A detailed description of the emendation
of the simulation, for this study, exists else-
formed for a variety of reasons and with where (Guetzkow, et al., 1960); a brief
differing degrees of cohesion; international presentation here should be sufficient to
organizations may be founded or foundered. give the reader a feel for INS-8.
Several media of international communica- For INS-8, it was decided that the there-
tion are available: written messages, face- tofore usual number of nations (five) would
to-face conferences, intemational organiza- not be sufficient; two new nations were
added Yora and Zena. We felt that the
In general, in the INS, international in-
teraction is les highly programmed than closer approximation to the real world sys-
the domestic set ing--in this arena partici- tem we were modeling and provide a more
pants are dealing with each other rather stable statistical base for the testing of
than with a conceptual reference group. hypotheses- for example, if we are testing
Since the pat ern of interaction is the pri-
mary locus of analysis, we have endeavored in a five-nation system twenty such (non-
symmetric) dyads exist; by adding two
to respond to and shape the international nations we more than double (N = 42) the
situation in which they find themselves. number of nonsymmetric dyads.
In three ways we increased the kinds of
3.3 INS-& Structuring the Basic
goods and services that could be generated
by the al ocation of basic capability: (1)
of the Nth-Country Situation force capability, which had been of one
type only, was now differentiated into
"clas ro m laboratory" runs of the simula- "nuclear" and "conventional"; (2) the na-
tion (runs INS-1 through INS-7) we ad- tions were permit ed to al ocate basic capa-
hered to our policy of nonintervention once bility to the "hardening"- -i.e., pas ive de-
the simulation was underway. These runs fense against at ack- of both nuclear force
were invaluable in the development and capability and basic capability; and (3) a
'shaking down' of the model (Guetzkow, procedure for conducting "research and de-
eta/., 1963).
As a result of this experience, we felt these changes were dictated by the problem
that certain changes and restructuring of
the basic system were required if the theo- The dif erentiation of force capabil ty,
retical model of the "cold war" and the
predicted ef ects of the spread of nuclear veloping two sets of "destruction ratios"-
capability (presented in Chapter I ) were nuclear weapons are more ef ective than
to be given a fair test. The basic system conventional weapons against certain tar-
so amended was dubbed "INS-8"mthe
CONFLICT lIESOLUTION V O L UI~E Vii NUMBEll 4
sctaipuatbe.il ty, we are grateful to Dr. Frank Both- f"otl o."wsT:his decision has be n explained as
grasibasdteozrasemirmtccethreshol lpncatiyraomolgocrnoanmetoosen.exqnucl(peRuen&nseciD)arvees wipadecisnteoaynltgoffiseohaidnoenvabebreleaigsottoiernwignsiwa.,naditticheitatedzreanfromtnaaiotniononetwoasthelretoacrbetogilDeciviasenfourhsizes14,aditoin0aoperi00tnFormstBC'iwaomdsses
emedhgsiatepasin.cbdexperiailnditedyvbeforeieminlsotentapdeemrcneildanaltienidlooterventi
mormore
e expensive to produce; i.e., itexpensive
takes to produce; i.e., it takes ers' nnee ed for a plausuisbilbele"c"ocvoevr"erf"orfotrhethe
mormore
e basic capability to produce onbasic
e unit capability to produce one unit experimental intervrevnetniotino.nR. R& D& waD was psrep-re-
of nuofclear force capnuclear
ability than it does to force capability than it does to sented to the particicpiapnatnstas afsofl olwosw: s:
proproduce
duce one unit of conventional force one unit of conventional force . . one of the decisisoinosnswhwhiichchaidasidescoe-co-
capabfiity.2 nnoomic growth is thehededceisciosniontotaol oaclaotceabteasibcasic
For the sake of realism, it was ruled resources to researchchanadndedvevloeplmopemnte.nTth. iTshis
that a nation's economy (i.e., absolute level
of basic capability) would have to pass a (BCrd decision, Decision Form, line 59) at
the beginning of each period. A nation must
have accumulated 2,000 BC's [units of basic
capability would be permitted; this thresh- capability] allocated to research and develop-
old was set high enough to preclude at- ment before it can expect to receive any return.
tainment of this eapabfiity prior to the pro- Both the amount of return and the time of
return are to some extent uncertain. A nation
The possibility of acquiring pas ive de- before a research and development project wil
fense was introduced because of the em- begin to pay off; once payoff begins it wil last
phasis it has received in the literature on for three periods. There is one chance in six
deterrence strategy (e.g., Morgenstern, that a project wil fail; i.e., never pay off. Once
1959; and Kahn, 1958). We made it pos-
in the first period of payoff, 10,000 BC's in
sible for a CDM to achieve invulnerability the second period, and 6,000 BC's in the third
for a portion of his nuclear force capability period; however, there is one chance in six that
and to "dig in" or "harden" part of his in a given payoff period these amounts wil be
basic capability (a feature analogous to reduced by half or lost entirely for that one
period. Once a nation has received al the
civil defense). Both of these forms of pas- return from one 2,000 BC 'project' it may
sive defense are expensive-- the defense of begin another research and development project
which would operate in the same manner as
than the defense of nuclear force capabfiity. the first project. There is also a smal prob-
ability that research and development wil
The cost of defense notwithstanding, this
yield return in the form of nuclear force capa-
provision made pos ible the adoption of bility rather than BC's [Brody and Noel, 1960,
strategies which reduce the pres ures for p. 14, emphasis added].
pre mption- we sought to provide deci- The italicized sentence was our cover for
sion-makers with a wide range of potential
the programmed spread of nuclear eapabil'
ity. All nations had a R & D project under-
The third change- the ad ition of re-
way at time "t"o; this project yielded nu-
clear capability to the nonnuclear nations
- -was prompted by two considerations: (1)
ac ording to a schedule which wil be pre-
the experimenters' desire to acid a category
of decision-making which had important
We decided, for INS-8, to have an inter-
(Rumre l, 1961); and (2) the experiment-
2 FoFr ohirs ashi tsancaesinidsevtalonpicgethiens cdoste/veloping these cost/

ef ecftievcntesivreatnioseosn urclaetriaonsd conventiouncal ear and conventional In earlier work, the simulation's tendency
wel , Center of Naval Analyses, Franklin In- great enough to obviate any ne d to start the
system-run with a structured organization al-
gani
runszatiwereon ialnirteady
704 iated iwin texih santence.internati
The ocharter csmales).
nal or- chosen able.
permanent delegate. With seven nations me"avnertsa)gein"doirca"taebtohvaetaovuerapgaer"tiocniptanhtis vwearie-
oat"Thecelgaosryifromfcrucial
weed,rewa709grsvariable
iThose oastudents
whouspefoldparticipated
ltowsginetested
:thboth122er i"ncbyselecting
oancnaDriver
rteioen,".
were
PdCHARD A. BRODY
readyinexiready
stenceamongthenations.How. in existence among the nations. How.
ev r,considever,
erable f ortoftenwasdevotedby considerable effort often was devoted by
thenationsthe
tothedev lopmentofaninter- nations to the development of an inter-
nationalorgnational
anization,beitapolitcal,miltary,
during June and July, 1960.
organization, be it a political, military,
and/oreconand/or
mic nsti ution.Ther fore,inan economic institution. Therefore, in an
at empt oattempt
alowconcentrationuponthen- subjects and assigning them to nations in
to allow concentration upon the n-
countriesprcountries
oblemperse,thes vente nsy tem- the INS," according to Driver, "was gen-
problem per se, the seventeen system-
eralized abstractness. The degree of ab-
stractness was measured by a form of the
of this 'I0' provided for universal membership
with regular sessions of permanent representa- Situational Interpretation Test [SIT] de-
tives.. [Guetzkow, et al., 1960, p. 5]. veloped by Schroder and Hunt" (Driver,
The EDMy was the permanent delegate 1962, p. 36). The SIT score was used to
identify the subjects' locations on a "con-
to the IO.
crete-abstract" dimension. From these re-
As was indicated above, roles for up to
sults, the subjects were classified into three
six participants are provided for in the
simulation. To have filled all of these roles groups, viz., "abstract" (Type IV), "con-
---one person per role---for multiple runs
crete" (Type I), and "neutral" (Types II
would have taken us far beyond our po- and III). The distribution of participants,
tential supply of subjects. Faced with a
122 "neutral," and 133 "abstract." When-
choice between more runs with fewer par-
ticipants per nation and fewer runs with ever possible, subjects from the same SIT
more participants per nation, we chose the
former. Our experience has been that com- Care was taken to distribute nations, so
composed, randomly over the seventeen
binhag roles works--i.e., the functions em-
bodied in these roles get performed. runs (Guetzkow, et al., 1960, pp. 16-7).
In addition to the SIT, Form 40-45 of
We used three participants per nation--
combining the CDM and IDM, and EDMx the California F Scale was used as a selec-
and DMF, and giving the EDMy the twin tion instrument. The F Scale correlates
responsibilities of aiding the EDMx with with the SIT and was used by Driver to
foreign affairs and attending the IO as increase the polarization of "Type r' and
"Type IV" subjects. The distribution of
per run and sevente n runs ac omplished, participants' scores on the F Scale is
?~ , I I I ,,
357 subjects participated in INS-8. smgm-peaKea, X = 5
s -- 8.6g--the distributi
3.4 The Participants and Their The subjects were almo
Selectio&
juniors and seniors; the
The participants in INS-8 were 357 few participants who
North Shore and Northern Chicago high their high school traini
school students (172 males and 185 re- to the beginning of the
a Dr. Michael Driver, Educational Testing Data (from the schools) on !.Q' (un-
Service, assumed primary responsibility for fortunately, not complete and not alw
the selection and assignment of participants--
drawn from comparable measuring inst
for this I am deeply grateful. For a detailed
description of the selection and assignment
process, see: Driver (1962), and Guetzkow,
et al., (1960).
CONFLICT RESOLUTION V'OLUME VII NUMBER 4
schoolparticipants,
orsnations
characteri
with
possible,
SYSTEMIC ticolcEFFECTS
s. legeandstudent
soruns asreduce
operati
OFas SPREAD
towere g a TheWEAPONS
nNUCLEAR
OFerror
populated startingTECHNOLOGY
situation was programmed
705
TheThe
ages o{ the subjages
ects ranged o{
from 15-the subjects ranged from 15- furtfurther
her oriorientetnthe stheubjecsubjts toetctshe opetortheationoperation
18.18.
The averageThe
age was 16. average
As far as age was 16. As far as of ththee simsiulmatulioan.tion.
In tthihis tsratraiininng,inthg,e stheeriousserinesousness
of the of the
eentel
ntel~r~iriseswae swasemphaemphasi
sized. Thezed.subjTheects'subjects'
variance from these subject psycho-social mamastery
stery ofofthethemechmechani
anics, involcs,veminevolnt, vanement,
d and
ssober
ober apapproach
proach to thtoe ethexperiexperi
ment wams ental- was al-
Despite the degree of simplification, ttogether
ogether remremarkabl
arkable. As e.viAsdenteviofdtenthis inof- this in-
operating a nation in the INS is stil a com- volvvemement
ent andandsobrisobri
ety, tehty,e facthet is offefactredis offered
plex task--it has been suggested that oper- tthathat outoutofofoverover9,0009,in0t00er-ninatter-nati
ion mes aognes messages
ating an INS nation is a task o{ about the ccontent
ontent codecoded,d, les ltehssan than
one peronecentperwercent
e were
same order of complexity for the high ccoded
oded "e"extra
xtra systsystem,
em," i.e., "coint.ea.,incontai
ed maten-ed mate-
riallnotnotperperti
tinennt entto thtoe simtheulatsiiomn.ulation.
real nation is for the more experienced
adult decision-maker (Driver, 1962). It is 3.5 TiTimme "eto"to":
": The TheStartiStarti
ng SitunatgioSin tuation
impossible to assess the merit o? this sug- for IN$-8
gestion but it is an interesting idea--this
complexity is a potential source o{ bewiI- to yield two bloc-al iances (one o{ :three
dement to the subjects. To reduce potential nations, the other of {our) which were very
bewilderment o? the participants (and, con- nearly alike in their military and economic
sequently, the error variance deriving there- capacities. Only one nation in each bloc
from) a two-phase program of training in had nuclear force capability at the outset-
the mechanics of the INS was instituted. In these nations were also the major economic
the first phase, the basic operating docu- powers. Table 3.1 summarizes the starting
ments of the INS (.the Participants Manual, parameters- al sevente n runs began in
a handbook covering the details of one of the same manner.
the roles, and information about the start- Table 3.2 presents the "generation rates"
ing situation, i.e., time "to") were mailed o{ the several nations for the production o{
to the subject in advance of his participa- goods and services. The generation rate is
tion. In addition, a one-hour, on-the-job, an expres ion of t~e ratio o{ units o{ a given
training period (phase two) was used to commodity produced to a unit of basic
TABLE 3.1
BLOC-ALLIANCE STHUCTURE WITH INITIAL BASIC AND FORCE CAPABILITIESa
T~m O~N~.-Yoat-EmoA BLOC T~m UTRO-ZmNA-INGo-ALcO BLOC

CAPA]BILITI~$ CAPABILITIES
Nation Basic Force Nation Basic Force
FCe Fan FCc Fan
Omne 34,000 1,600 28 .... Utro ' 30,00'0 1,500 9.,5
Yora 12,000 700 Zena 8,000 200 --
Erga 8,500 750 -- Ingo 9,000 800 --
Algo 7,500 100 --
Total 54,000 3,050 28 ' ' 54,500 2,600 25
a The "basic capability" of the nation (subsuming, as it does, resources, both physical and human) represents
the nation's ability to produce all goods and sende. es. The "conventional" (FCe) and "nuclear" (FCn) force
capability of the nation represents its ability to mount military operations against other nations. All units in the
simulation are arbitrary, reflecting only the relativeposition of one nation vis-~-vis another.
aePeprs,rsntpuhec3.let2apairvrfeto"ircc(itephacniasptasdobweilctury-e. wellttheneef3.tdc(eox2uira6tr)pesemigcdlenexperiDaasryiedntodsuymetmilitary
vasplotnawtrteiangtneigasl."opaWr,oaprmelrdehtTABLE a)ite.saolmfinozSources emfbetanmaeentagxslioytcinhsoegalliances;
trnlpbeinsamlitronsnovuterventi
,wpucse,grn.daeotmencibothtedoasrn.a,Thetofgemenraredos,(mbateycshesecuredoumetnrihmfeffects
easerdu(e"inre.tbyecoi.se",-
The bloThe
c-al iancesmthe nations of Omne, bloc-alliancesmthe nations of Omne,
GENERATION RATESa
Yora, aYora,
nd Erga making up one bloc and Utro, and Erga making up one bloc and Utro,
Zena, InZena,
go, and Algo the other- are trade as Ingo, and Algo the other--are trade as
B.
B. ATE
ATE
Nation BC/BC CS/BC FCn/BC FCc/BC
formal treaty [Quoted in. Guetzkow, et al.,
Algo 0.8 1.0 0.06 0.5 1960, pp. 24-5].
Erga 0.8 1.5 0.02 0.5
Ingo 0.9 1.0 0.02 1.5 In this manner, the participants learned
Omne 1.0 1.5 0.08 2.0 of the situation in which they would begin
Utro 1.0 1.5 0.08 2.0 to operate their nations--the situation
Yora 0.9 1.0 0.06 1.5
Zena 0.8 1.5 0.02 0.5 which emerged from their decision-making
a BC -- basic capability unit. beyond time "to" was of their own making.
CS- consumption satisfaction unit. The "worlds" beyond time "to" were un-
FCn -- nuclear force capability unit.
FCc = conventional force capability unit. structured by the experimenters except for
capability allocated--it is thus a measureof which the study was designed to de-
of the efficiency of production.
In the absence of a medium of exchange,
the differential generation rates, between
nations on the same commodity, permit
In order to determine the effects of the
barter-trade according to comparative ad-
operational definitions of the variables)

In addition to this information about the
were drawn from two clas es of data from

given information on the initial military-
INS-8' (1) system endogenous data; and
economic-diplomatic situation v/a the fol-
"System endogenous data" are those data

ment was offered in lieu of a scenario):
which result from the operation of the sim-
In the simulation world two nations are the ulation by the participants: decision forms,
dominant centers of power. Omne and Utro, in
addition to being the only nuclear powers at
this time, are also in pos es ion of large and IO transcripts, pres releases, trade and
amounts of BC's, that is, economic potential. aid transactions, treaties--in short, the
The other five nations cluster about these
power centers in two bloc-al iances; the two sion-makers, which lend themselves to the
bloc-al iances are relatively equal in compari-
son with each other- neither enjoying a clear
superiority in al areas of present position or to time "t,." These data have be n used to
growth potential. All nations have research aid in the production of the "history" of
and development projects underway which if each of the sevente n runs of INS-8 (Brody,
they bear fruit wil yield increased economic
et el., 1961). The use of endogenous data
The present positions of the nations represent
in the exploration of the model wil be dis-
the product of both internal growth and inter-
national trade. The principal trading partners The "system exogenous data" were gen-
have be n Omne, Yora, and Erga, on the one erated from the administration of a bat ery
hand, and Utro, Algo, Ingo, and Zena on the of "pencil and paper" instruments. At thre
other; trade has not be n exclusively among
these groups but it has tended to fol ow these points during the simulation- just prior to
lines. the experimental intervention, just after the
used(nDdtoris.v"thier,bysinstrument,
Aiiscaling Driv. erThionlisyninhistrument
1962)
study s workdeveloped
sofar
was as Driwionvler'INenter personaffl fabehavi
sandS-8re- itated
personal wectospend
rs ocr.oulIadnmoment sions'
sftTheruoabstiwenapargqdulrfientoeirhtcsmeiap,noatnprlhytuesratewewsnaoartseiohApaplniskapenisnend,dtheiohx. el"Ip.rsfayutsletnetmshe,. ttoohrwieTgamhrindseaslceosuteihxtgeerhresst.n;Diptvhrsateoricausebostlneftsuli.nufbeohmramviisoarpscriesrsecenumtleaedrd
beorderexpretosdiscussing
eaccompl
d as combiisith-
experimentexperimental
alintervention,and uringthe intervention, and during the
finalperiodfinal
- theparticpantswer asked period--the participants were asked and its application to the data of INS-8.
tocomplettoeachofthre "PolicyPlan ing complete each of three "Policy Planning Leary's research was designed to illumi-
nate the effects of personality on inter-
"Planning Aid #1," a multi-dimensional
this a system of classification of message
content was developed (1957, pp. 63ff.).
Leary reports that,
[i]n surveying the list of more or less generic

sults are drawn upon for confirmatory or interpersonal trends, it became clear that they
disconfirmatory data. all had some reference to a power or affiliation
"Planning Aid 4P2," a form of the factor. When dominance-submission was taken
"Semantic-Differential" (Osgood, et al., as the vertical axis and hostility-affection as
the horizontal, all of the other generic inter-
1957), was used to assess decision-makers'
perceptions of the nations in the system. nations of these four nodal points. The various
Data from the "Semantic-Differential" do types of nurturant behavior appeared to be
not enter into the analyses which follow; blends of strong and affectionate orientations
to blend hostility and weaknes . . . [a] cir-

"Planning Aid 4P3," a rating she t con- cular two-dimensional continuum of sixte n
sisting of five "Lickert Scales," was used generic variables represented the optimal de-
to as es the at itudes of decision-makers gree of refinement of interpersonal themes
[1957, p. 64].
While Leary identifies "skte n generic

variables," his analyses are made on the
and co perativenes of the other nations"
basis of eight interpersonal proces es: (1)
and to expres their perception of the like-
managerial-autocratic; (2) competitive-nar-
liho d of the other nation "becoming in-
cis istic; (3) aggres ive-sadistic; (4) rebel-
volved in a war" with their nation. This
lious-distmsfful; ( 5 ) self-effacing-maso-
chistic; (6) docile-dependent; (7) co pera-
3.7 Content Analyses4 tive-overconventional; and (8) responsible-
hypernormal ( 1957, p. 65).
Before essaying the laboratory explora-
tion of the model, a further note
continu m whose octants represeon
nt content- meth-
odology is in order. Throughout the balance
of this chapter, reference will abe s Figure 3.1 (after Leafy, 19made
57, p. 65). to
"Leary Content Processes"; since the
The octants are labeled with a two-let er con-
tent analysis approach of Timothy code which indicates the placement of the Leary
(1957) has not been widely used in polit-
ical science, communication should be facil- Leary defines two scale-scores, "DOM"
and "AFL," which are composites of pri-
4 In the work reported in this section, the mary and (weighted) secondary dimen-
assistance, aid, and comfort provided by Dr.
George Wright (University of Cincinnati) were
invaluable. The original impetus for this mode DOM=AP-HIq-0.7(NO
of analysis carne from Wright. q-BC-FG-JK)(3.1)
fsaCimnOfepoNdleFaostwLrfsIi:goClTnowmRs:eEtSriOcLfUorNaTmrciIuOsliasNeti:c VODistruLstUfulME VII NUMB
708 s"cHryuMapibsoemcsHOSh(HiisatEsiIcVIRit+Ior-. nCoopei4LMc,tDE"lyraa-tnpivadAPeLMth"ocl-omgDE)0ic.p7ael;tiantiedornp"restcatoiroens.
VII . Responsible- RICHARD A. BRODY
II.II.CompeH~ive.
CompeH~ive. '~ AP
NO ,~
Overconventional
'~'
Bc -~ LM
IV. IV.
Rebel ious- Rebellious- V.Self-Effacing-
Fro. 3.1 Leary octant proces es.
AFL = LM - DE + 0.7 (NO ThusThus,

, DEP - JK- BC + 0.7 DEP -- JK- BC + 0.7
(3.2) - BC-FG + JK) (3.5)
Following
Following Leary's
Leary's
logic, "dependency"
logic, "dependency"
(DEP)
(DEP)andand
"hostility"
"hostility"
(HAS) scale-scores
(HAS) scale-scores (NO- BC-FG + JK) (3.6)
can
canbebederived
derived
from the
from
manipulation
the manipulation
of of
Were they to be used, "trust," "suspicion,"
The scale-score of the ith axis is defined could be derived in the same manner.
These labels (i.e., the Leafy proces es)
ith axis score = i- (1 + 180?) it should be understo d, are merely a short-
70.7 [(i+45?) + (i-45?) hand for .the complex of characteristic be-
-(i- 135?) - (i + 135?)] (3.3)
havior (including communication behavior)
The scale-score of the i-F 90? axis in ais
given octant. Leary's identification of
de-
both "pathological" and "normal" (i.e.,
adaptive) content in each category pro-
i -F 90 o axis score = (i + 90 o ) _ (i - 90 o )
+ 0.7 [(~ + 4~o) - (~-4~o) The scale-scores profile the content of an
-(i- 135?) + (i + 135?)] (3.4)
individual's interaction with others and
(In both formulae, 0.7 -- sine theta [the
sug est the dominant loadings.
angle made from moving counter-clockwise
To ap ly this system of categorization to
from the nearest boundary of the i + th90?
e analysis of ap roximately nine thousand
octant to the ith octanti - sine 45 o. ) inter-nation mes ages, the fol owing steps
form contaIDniscoded.
sages were ritnecedut riemoreoqnu.est.than one theme, an BApFSaGrertisquaaagognecmliutchconference.
CrUlRPower agoexiilstasdfizmermtsoauunyttn.lmiruggl,je,.,ccsdea..tipgseaudperi.,letcoyiian,ttuimvpese.nt. 709
wewere
re taken: (1) ttaken:
he unit of analys(1)i was the unit of analysis was Of er a deal such as as tartadreadweitwihth
dedefined;
fined; (2) the operation(2)al speciftheicatior operational specificatior a specific transactiononsusgugestesdt,ed,
e.g, such as psecificicamaomuonutsntso to
ofofthe conthetent categcontent
ories was ac omplishecategories
d: was accomplished:
be traded or a specififcitcmtiemfeorfor
(3(3)) the relthe
iabil ty o?reliability
coding was deter- o? coding was deter-
mined; and (4) a decision was made on
how to handle different levels of intensity, II. REJECTION OR MANIPULATION
= BC
It should be noted that sampling problems
Self-praise in a manner derogatory
did not emerge--all nine thousand mes- or disparaging of others.
Withholding, e.g., "I won't give in
To extract maximum richness from this until you do," with a challenging
rich source of data, the transmitted mes-
Aggres ive confidence theme, e.g.,
sage-form was chosen as the unit of anal'
"I don't need you to go ahead."
ysis. However, if a transmitted message-
Selfish motivation--self-centered.
appropriate number of codes was as igned "Don't cal us. . we'l cal you."
to it. If, for example, both hostile and I I . . HOSTILE-AGGRESSIVE- DE
friendly themes were contained h a single Readiness to use power in aggressive
message, both were coded (thereby in- manner-- "either . . or."
dicating the ambivalence of the message). Themes of punishment, ? threat, co-
We, of course, have no direct measure ercion, anger.
o? the effect of the message---i.e., the de-

gree to which nuance was picked up by Hard-hearted.
the receiver--but the thrice given attitude Brutal y frank.
scales seem to confirm the content analysis. Impatient, in a hostfie sense.
For example, hostile communicators tend
to be perce{ved as hostile; in correlat'mg IV. SUSPICIOUS - FG
the receipt of hostile themes from a given To be hurt, e.g "You are doing
nation with the rating of the hostility level bad' things to me."
of that nation by the nation receivingthe Fe lings of distrust, iealousy, scep-
communication, we find the following:
rs = 0.53; t- 3.95; d[- 40; p ~ 0.0005. Complain'rag.
We established the following scoring
guide for each octant5 (the principal proc- ANXIOUS-SUBMIS IVE = HI
ess appears in all caps): Fe lings of personal discomfort, e.g.,
I. DOMINATING- AP
The giving of orders without hostile Inadequate, weak,
overtones; Ambivalence, unable to commit self.
Setting the terms of interaction.
Compliant ('overly).
Giving suggestions or opinions which
would induce the recipient to fol-
low the sender's lead. DEPENDENT = JK
Provoking some kind of help
Asking for "aid" as op osed to

5The specification of these categories is pri- "trade" among equals.
marily George Wright's work.
_ErgaLegend:DF ? 1 2X Xx 3xX 4xX5xX 6x X x7Xx8 Xx9
Algo Period
FCn's no I~ yes '

DF i x x x x x x X x
FCn's I ~'no
yes-- .... ~
Ingo DF i x x x x x x x x
I ............. ....~ .....
FCn's .... I no ,yes- _
Omne DF ~0 x x x x x x ' x x X .
FCn's yes
? ~ - , .......~ ,, -- ~
Utro
DF ~ x x x x x x x x
FCn's
~ yes ...........- ............
DF
i
~ i .......
-'-----
Yora I ...............
FCn's ! no
i r"'" ~ yes
......................... -
DF I x x x x x x x x
Zena i
FCn's ! no ~ -yes
iO meeting I ooo ooo ooo ooo ooo ooo ooo ooo
Planning Aid , I xxx x
DF - Decision Form.
FCn's - Presence of nuclear force capabil ty .(Note: Phase transition from bi-nuclear to n-nuclear.)
X'-Time of ar ival of DF to CDM (Note: The delivery of the DF's was stag ered through the period but each CDM had his form in his pos es-
sion the same length of time.)
o o ~ Span of 30-minute IO meeting.
x x-' Time of administration of exogenous measures- 30 minutes for each administration.
Fro. 3.9,. Event flow in INS-8.
c(benytWr; Nig=hPGiAsksSProposesrut3ovhepmutas8tbeipoe6nnttilacer)tgaford.ountwafavor.eaatdgnslh.asobesusemprtradel(necionaaevan.enouofecntiethcs..camongwiowetermsrtn)ahksaeinid.r.equaletertrheedsstc)hwi.oidgteh-hs patciliationnhlraeoetlcayesmdceausleetrlwheacoding)visrceewinshnotlDafteobiclseyieldedtoqiweucox.andclisuFptsorainvagreement
Eagerstronger.
Apple to please.
polishing. mcetl,ysdtuh(tresheasdrPwheineaoverrtigeohrde-
messages' were double coded and then

Demonstrates trust in someone
randomly assigned to one or the othe
for reconciliation of differences, Ch
"You set the terms of our interac- this reconciliation coding (v/a doubl
tion."
90 per cent and this procedure, i.e., double
coding and reconciliation coding by one or
VII. FRIENDLY-COOPERATIVE ---- LM the other coder, was followed for system-
runs INS-8-II through INS-8-XVII.
out the dominance implied in Leary observed that differing levels of
intensity of expression affected the inter-
Proposes nonaggression pact for action of communicators; our experience
Going out of way to be nice, agree- in coding INS-8-I confirmed this observa-
tion. To account for intensity, we weighted
Anticipates hap y ending. coded themes--the expression of a given
Sharing, sharing information or con- Leary Process was either coded as a single
fidence, e.g., about a third party. or double frequency depending on the in-
VI . SUP ORTIVE _- NO tensity. Thus, for example, a rejectlye
theme would be coded "II"; a very rejectire
Be very kind and gentle.
theme would be coded "II+" and counted
Think of other's welfare primarily.
as double the frequency of occurrence.
Sympathy, understanding, soft-
Persistent doubts about the "interval"
nature of the circular continuum and
Leary's statement that, "while units around
In ad ition to the eight Leary categories, ment is cor ectly ordered" (1957, p. 6 ),
we felt a ninth (neutral) category was have militated for the selection of non-
needed to ac urately reflect the mes age
content; 9 per cent of the coded themes Leary Proces Scores are involved. These
was as igned to this category. This ninth
category does not enter into subsequent
analyses. However, its presence reduced 3.8 Event Flow in INS-8
the tendency to read one or another proces
Before turning to the actual experiment
into a/ themes during coding.
and the testing of hypotheses, it is felt that
To check the reliability of coding, the
"feel" for INS-8 can be enhanced by the
mes ages from INS-8-I were double coded presentation of a model of the flow of
events in the simulation.
compared (.the mes ages from this system- Each of the seventeen system-runs cov-
run were also used to refine the eight cate- ered four half-days of real time--twelve
gories)- the level of agre ment (60 per periods of simulation time. In each case,
the first period was given over to the on-
enough to justify the procedure of single the-job training discussed above. The deci-
coding the remaining sixte n system-runs. sion-makers, therefore, received and com-
As a result of this reliabil ty check, al
RICHARD A. BRODY
12 Decision Form was completed but not in during Periods 5 and 6--our feeling was
returned to the decision-maker after cal- that a two period "transition" from binu-
culation). The international organization clear to n-nuclear would aid in maintaining
met once each period--eleven times in all. our cover story, i.e., seem less unnatural
The "pencil and paper" instruments were to the participants.
administered during Periods 4, 7, and 12. Figure 3.2 summarizes the flow of ex-
The experimental intervention was phased- perimenter controlled events in INS-8.
icwasetsmaettegtrsiiginovwhinsgenngawhinctfromhod taichnrhee- miul4.atsha2kcThietr,oiFundaho.iyssnpchapter
presmeditieocrstnieoacnttnhbiposoensualitabzstloeitdcurthaelofliasbertytihlsdoyitannpievmsmkedihcoesairs.platylcondi
hypot taohntehdme"tsceehonsledticseaowaxlndevotedpltQueraoinrasyetdsitoietonnmtsbe"thehegsitnpresenta-
hasbeolwiirsehttheicdaa?l
CHAPTER IV
Simulation and the nth-country

situation: exploration of the model
4.1 In4.1
troduction Introduction exception being the expepreirmimenetnaltayl-yin-tirnot-ro-
In ChaIn
pter I , the at empt waChapter
s made to duced spread of nuclearrweweapaopnosn. s.
II, the attempt was made to
serve to define the "cold war system" tion of the results of the exploration of the
theoretical model in the inter-nation sim-
(Figure 2.1, above), These "conditions"
and "linkages" are not rules of behavior for
decision-makers, such as serve as defining basic question set.
characteristics of Kaplan's eight "systems"
(1957); rather, they are a series of pre-
Drawing upon the hypotheses which ap-
pear in Chapter I , it se ms reasonable to
the interactions of l national.actors differ- ask questions of the data generated tin
ential y structured with respect to their INS-8 at two levels of analysis- macro and
economies, military capabilities, and do-
The fundamental, macro-level question
Chapter I I dealt with the development is: Is the international system dif erent be-
of a laboratory set ing in which this "island fore and after the spread of nuclear weap-
of theory" could be explored. Subjects were ons?
asked to make domestic and international At a more micro-level, the following
decisions- i.e., to commit their nations to questions may be asked: If there is :a
courses of action- in a bipolar, binuclear, 'change in the system, what is its relatiOn-
seven-nation, simulation "world." Informa- ship to the island of theory? That is to
Do we get two blocs which relate :to each

participants prior to their participation. In- other and have an intra-bloe structure as
formation about the structure of the inter, predicted? Is the pre-spread system tightly
national system was kept to a minimum- bipolar?If the "cold war" does emerge;
the entire scenario was containedin a thre - does it fragment under the impact of the
paragraph "World Perspective" (Section spread of nuclear weapons?
3.5, above). The development of~ the na- It is neces ary to answer the micro,
tions and the international system beyond level questions first, v/a the examination of
time "to" was almost entirely the product
of the participants' decisions- the principal model. The answer to the macro-level ques-
714 YAUE0 -+
Throughout the analyses which follow, Natifsdeobecrevlooanunr-lpebtew"melsetr(awihifcr+a.ehosl.ur,d1U-creme7b"kscsieogtdyryihn+seeoYtdanlerltd--mueeoanddt-srZanoe"l++utset).neaisrHotmy.dseTwiontphneemv.nedsamerect,noicsrtai.ecsnulTcrhewesitstehrrioess
RICHARD A. BRODY
Nation A E I O
- q-
- q-
I - q- - q-
- q-
q- - -q-
_ q-
Z + - q- - +
q-: intra-bloc dyad; N -- 18
- -- inter-bloc dyacl; N: 24
Fig. 4.1.~ Behavioral dyads in
tion follows from this examination--the reduced- becaureduced--because

se of Driver's design for of
answer wili serve as a summary evaluation subject placemsubject
ent, for any given nation, placement, f
of the model. personality ispersonality
randomly distributed over .the is randoml
sevente n :sysseventeen
tem-runs; (2) ag Tegation :system-r
4.3 Units of Analysis and the Pooling over the rimsover
provided more data in each the rims provid
of Data ofthe dyads anofthe
d increased the stabil ty of dyads and incr
the measures.the
To insure: that ag regation measures. To ins
was legitimatewas
the basic unit of analysis is the behavioral , various measures were legitimate, var
che kedfor "pcheekedfor
o labil ty. . Runs tesq' "poolab
?
dyad aggregated over the system-runs. Two

(SiegeI, 1956, p(SiegeI,
types of dyads---intra-bloc and inter-bloc-- . 52-8') performedon sev- 1956, pp. 52-
were defined (Figure 4,1). This definition
within dyads; betwe n ',worlds." It was
is in terms of the alliances indicated to the
participants in the "World Perspective."decided that an analysis of variance should
Since the measurement of behavior for
any pair of dyads composed of the same
nations (e.g., AE and EA) is not expectedThe cmciality of the measure of hostility
to be Symmetrical, our total number (Sof ystems Questionnaire, Scale #1, Ap-
dyads is forty-two.s pendix I ) guided its selection. The scores
For several reasons it was were east into a two-way table having k
.. considered de-
sirable to aggregate over system-runs: (1) columns and N rows (k = 17; N- 42) re-
the effects of personality differences are fleeting the 42 dyad scores, on this measure,
This treatment is based on the assurnption

that the behavioral dyads are statistically inde-
pendent of one another. "Statistical indepen- range of values for measures used to describe
dence," in this context, means that a score behavior or perception in one dyad is not
(.message frequency, rating, etc.) found in one af ected by the value in another dyad, the
dyad is unaffected by the score found in
another dyad--in the sameway, for example, not to argue that the behavior in the dyads
that the appearance of a head or 'taft: in one is psychological y independent. The degre
coin-flip is unaffected by the findings in previ- of psychological nonindependence resulting
ous flips. If the range of possible values for from dif erent "treatment" (e.g., intra-bloc vs.
an event "e~." is affected by the value of event inter-bloc) is preciseIy what the analyses which
"ca," the events are said to be "not indepen-
CO~,{FLICT RESOLUTION 'VOLUME VI NUMBER 4
tnEranked 12
aCohvutaHOScsni,eosdntioTastnlIiLb,oRIwithinaTInelsYiandtaoinsrd.se,dyadsLaNivnoarkailtahgbwacrossels atertnhrunsUenP2irv,2oe6andrg8siatmya, blTthecuhorecvspedyadscrawiolbet.ahbTheiwilzietlrieexperiosrankmeas aonmchiiaentagtnhered ulwinonhypothes
ithpercepti
voacriuarnicse,on.nsicne
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY :715
should
shouldaid
aidinin
binding
binding
aggregates
aggregates
of nations
of nations
"Friedman Two-Way Analysis of Variance" into
intobloc-alliances.
bloc-alliances.
(ibid., 1956, pp. 166-72) was performed, Null
NullHypothesis--Ho:
Hypothesis--Ho:perception
perception
of~ in-of~ in-
with the following results.2' ter-bloc
ter-blochostility
hostilityis no
is greater
no greater
than the
thanper-
the per-
Null Hypothesis--Ho: There is no differ- ception
ceptionofof
intra-bloc
intra-bloc
hostility.
hostility.
Hi: the
Hi:
inter'
the inter'
ence between replications on hostility-per-
ception levels. Hi: There is a difference of hostility than wil the intra-bloc dyads.
between replications. Statistical Test--Systems Questionnaire
(Appendix I), Scale # 1 is the measure of
perceived hostility. On it, each of the
x2 = ~ (Rj)2-3N(k + 1);
Nk(k+l)~= seven nations rated the other six nations as
to theft "friendliness-unfriendliness"mhigh
where Rj is the sum of the ranks for the
scores are toward the "unfriendly" end of
jth column (i.e., the jth replication).
effect, predicts the relative rankings of per-

N -- 42; k - 17; ~ (Rj) 2 -- 2,458,489.75;
j=l ceptions across bloc lines in contrast with
df = 16; perceptions within the bloc. The "Mann-
Whitney U Test" (Siegel, 1956, pp. 116 if.)
9 _ 0. 0 93 (2,458,489.75)
Xr
is appropriate for analyz'mg these data.
= 18.40; p -- 0.30. Significance LevelsLet a = 0.05 (a --
probability
Decision.--Ho not rejected; the of a Type I error). nl = 18 =
data are
poolabIe. the number of intra-bloc dyads; n2 = 24 =
the number of inter-bloc dyads.
4.4~ The Cold War System: Sampling Distribution--For n2 > 20 the
following formula yields values of z (ibid.,
1956, p. 123):
4.4.a PERCEPTION Or OUT-GROUP
z -- U - (nln2/2)
The first "condition" (C1, Figure 2.1, X/(ni) (n2)(nl q- n2 +1)/12 (4.1) .
above) to be established in our exploration

of the "cold war system" is the perception under H0 of values as extreme as an ob-
of the external bloc (i.e., the nations with served z may be determined by reference to
whom a nation is not allied) as being a table of probabilities under the normal
hostile. The tendency of groups to dis-
tinguish "ins" and "outs" is pervasive and tleiection Region- Since H1 predicts the
serves to bind aggregates of actors into a direction of the dif erences, the region of
group (Campbell and LeVine, 1961). Byrejection is one-tailed. It consists of al
extension, the presence of such distinctions values of z which are so extreme that their
as ociated probabil ty under Ho is equal to
hich this analysi of vari- data upon which this analysis of vari- or les than a = 0.05.
' The d'ata upon wThe
Decision--The hostility perception scores
of Graduate Training and Research in Inter-
of each of the 42 dyads are shown in Table
4.1, together with the rank of each dyad.
24. ZY 2302 27.5
Foos(ribce1.iqdtauh.t,ieevs1dael9ewi5nAE
6t,lyhp,.Va12l0u)e:230
s as Extrem27.5
e as Ob- 1.HOSTIAILITY 162 11
TABLE 4.1
PERCEPTIONS OF INTER-BLOCANDINTI1A-BLOC HOSTILITY
INTER-BLOC DYADS INTBA-BLOC DYADS

Dyad Score Bank Dyad Score Rank
2. AO 229 26 2. AU 169 13
3. AY 234 29 3. AZ 178 18
4. EA 225 25 4. EO 134 2
5. E1 224 23.5 5. EY 149 7
6. EU 252 37 6. IA 160 10
7. EZ 239 33 7. IU 171 15
8. IE 224 23.5 8. IZ 165 12
9. IO 268 40 9. OE 151 8
10. IY 238 31.5 10. OY 142 5
11. OA 186 19 11. UA 177 17
12. OI 238 31.5 12. UI 139 3
13. OU 285 42 13. UZ 141 4
14. OZ 251 36 14. YE 144 6
15. UE 237 30 15. YO 124 1
16. UO 277 41 16. ZA 170 14
17. UY 259 39 17. ZI 173 16
18. YA 195 20 18. ZU 152 9
19. YI 99~2 22
20. YU 246 35 ~ -- 171 -- R~
21. YZ 242 34
22. ZE 218 21
23: ZO 254 38
~--732-- tG
data, R1 - 171 and /t2- 732. p ~ 0.00 030. 3Si.nSceintche pthiseleps isthlaensa-than a-
The value of U is found by substituting the 0.05, ouorudrecdeisicoinsiosntoisretieoctrHoeieicntfHoavorin favor
observed values in the fol owing formulae of H~. Na. Natiotnisoni stihne etxhterenxaltbelroncalreblpoerc~are per~
ceiveddasabseibenginmogrmoe horsetilhoe tshtainlenathioans nations
withinnthtehbeloblc. oc.
n2 ( n, -H 1)
U = nln9 + -/12 (4.2a) 4.4.b Wm~Wm~SMSIrMEIrD EXDPREXPRES
ES ED SED
nl(nl + 1) The second "condition" (C2, Figure 2.1,

U -- nln2 + - R1. (4.2b)
above) is the transmis ion of hostile mes-
FromFormula4.2awefindthat: sages betwe n the blocs (i.e., "real" in con-
tradistinction to "perceived" hostil ty).
U- 732-R2=0;
Nul Hypothesis- Ho: the transmis ion of
usingFormula4.1,wefind: hostil ty betwe n the blocs is no greater
z -- U- 216/39.34 = 5.40. than the transmis ion of hostil ty within the
Reference to the "Table of Probabilities As- blocs. H~: there is more hostil ty trans-
mit ed in inter-bloc dyads than in intra-
served Values of z in the Normal Distribu- bloc dyads.
tion" (ibid., p. 247) reveals that z = 5.40 Statistical Test- The HOS Leary Proces
has a one-tailed probabil ty under Ho of Score (Formula 3.6, above) is the measure
CONFLICT IRESOLUTION VOLUME VII NUMBEIR 4
2341O879 25689
2wi1teaF2.04.h3.1.7n5.f33o..lerosAEYOAEAAOAYEZEIOUmebcZUOEstedu,alrctpovarne.4dt.1azincwitesdltihn--10.-23.--b-25aScoree8.88601057me67t...d3349573u248moesteadrgemtshooinsdentdilbe4321Ra235.3try8685onm..55khrmeienfmeTABLErzs- 4.isn21f2.4.859FoDya.uIYUAAUEOOErZndOdmubyla 4s.u2bsa'titut--22.--65.iScore246234n6.781g.58783342the observedRa2370nvaklues
RECEIVED EXPRESSED HOSTILITY IN INTER- AND INTBA-BLOC COMMIJNICATION
INTER-BLOC DYADS INTRA-BLOC DYAD$

Dyad
1. AI
-17.2
3. AZ -15.8
5. EY -43.4
7'
6. EU q- 0.4 6. IA 16
-17.8 7. IU
8. IE -17.1 -23.1 19
9: IO 3
10. IY 10. OY -18.9 fl.5
-19.0 15
12. OI 12. UI -40.3
13. UZ 1
14. OZ 14. YE
15. UE 10
16. UO 16. ZA
17. UY -32.5 17. ZI -31.9 12
18. YA 18. ZU -43.2
19. YI -21.9
-17.1 2&5 ~l- 237'- R1
21. YZ 41
24. ZY 37
-- 666 -- Ra
of tranof
smit ed hostil ty. The hypothtransmitted
esis, in hostility. The hypothesis, in
U - 732-/t2 - 6 .
bers of the op osing bloc than wil be con-
ta'med in mes ages sent within the bloc. Using Formula 4.1, we find:
The U Test wil be used in the analysis of z- U-216/39.34- 3.81.
Significance Level- Same as 4.4.a, above. Reference to the table of normal probabil-
ities (ibid., p. 247) reveals that z- 3.84
Sampling Distributiorv- Since n2 > 20,
has a one-tailed probability under Ho of
and the probability as ociated with the oc- p < 0.00007. Since this p is less than a-
cur ence under Ho of values as extreme as 0.05, our decision is to reject H0 in favor of
Hx. More hostility is transmitted external
ence to a table of normal probabil ties. to the bloc than internal to it.
/1eiection Region- Same as 4.4.a, above.

Decision- The HOS Score for each of the 4.4.c THE PERCEPTION OF THREAT
EXTERNAL TO THE BLOC
42 dyads is shown in Table 4.2, together
with the rank of each dyad. For these data, The third "condition" (C3, in Figure 2.1,
Rx = 236 and /19 = 6 7. The value of U above) is specified as "inter-bloc tenston
atseiom14.
718 monp"trsaestohanmereaOZ
intetnierngvfeutnlhiya197
ngopnvaeartariotinaobsnwlaei25. above.
tzprhein"otveTABLEhnde-5 4.lhypna14.
ik3teiloytnhtseosribaste,YE
ciodnmtehfeineovctoh116
,lveprdesdixncnataswatihor2antwisnaotnhtitohnes
RICHARD A. BRODY
PERCEPTIONS OF INTER-BLOC AND INTRA-BLOC THREAT
INTER-BLOC DYADS INTRA-BLOC DYADS

Dyad S core Rank Dyad S core Rank
1. AE 206 29 1. AI 173 18
2. AO 243 36 2. AU 189. 21
3.AY 197 25.5 3.AZ 156 15
4.EA 207 30.5 4.EO 149 12
5.E1 211 32 5.EY 144 9
6.EU 257 39 6.IA 146 10
7.EZ 207 30.5 7.IU 172 17
8.IE 198 27 8.IZ t52 14
9.IO 272 40 9.OE 136 6
10.IY 224 34 !0. OY 141 8
i1. OA 186 23 11. UA 174 19.5
12. OI 220 33 12. UI 148 11
13. OU 286 42 13. UZ 130 3.5
15. UE 185 22 15. YO 136 6
16. UO 279 41 16. ZA 136 6
17. UY 236 35 17.ZI 114 1
18. YA 150 13 18. ZU 130 3.5
19.YI 189 24
20. YU 246 37 ~l -- 182.5 = R~
21. YZ 169 16
22. ZE 174 19.5
23. ZO 252 38
24. ZY 203 28
~: -- 720.5 -- R2
Statistical Test---Systems Questionnaire

ItsIts
inclusion
inclusion
was guided by the was
ubiquity guided by the ubiquity
ofof
"tension"
"tension"
as a concept in theas
analysisa ofconcept in the analysis of (Appendix I) Scale 4P2 is the measure of
international
internationalrelationships. However,
relationships.
at- However, at- perceived threat. On it, each of the seven
"likeliho d of becoming involved in a war"

futile- it is perhaps bet er left as a prim- with the perceiving nation- high scores in-
itive used to describe a total interaction dicate a higher perceived likeliho d. The
situation. Moreover, the presence of the
concept "threat" in the post-spread model outside of the bloc are se n as being more
(C7, Figure 2.1, above) argued for its in-
clusion in the pre-spread analysis. Accord- rating nation than are the nations of the
ingly, "threat" wil be substituted for "ten- rating nation's own bloc. The U Test wil
sion" in this analysis. be used in the analysis of the data.
Nul Hypothesis- Ho: the perception of Significance Level- Same as 4.4.a, above.
threat acros bloc lines is no greater than Sampling Distribution- Same as 4.4.a,
the perception of threat within the bloc.
Hi: nations outside of the bloc wil be se n Rej ction Region- Same as 4. :a, bove.
Decison- The threat perception for each
bloc. of the 42 dyads i shown in Table 4.3, to-
hrsReference rbitesisva.enTtcioharle--, theTaThe212)bl:sampleproba4.4. ibniglitdiysatris 7ob4ciuti,0a4to6edn iwis dithstrithebutedoccur-as
yeolpauotldiho,enhosisbw,toeientvwtheerfe, whetablnct,nhepreoftshdeeinormalcymetsaraepssouusprobabfi . (4. ) 719
ggeethtehr ewirthwithtehrantkhoef eraachndkyadof. Foerach dyad. For SpeSpearman

arman Rank OrdeRankr CorreOrder
lation CoeCorrelation
f- Coef-
thhesesdeatada, Rt1a=,1R182.5 a=nd182.R2 =572a0.n5.dThR2e = 720.5. The fificient
cient (rs) (ib(rs)id, pp.(ibid,,
202 if.) wipp.l be202used if.) will be used
vvaaluleuoef Uofis fUounisd bfound
y substitubytionsoufbsthetitution of the ttoo testest
t this hypot
thisheshypothesis.
is.
oobsbserevrdvevadluesvainltuoeFsorminultao4.2Formul
a: a 4.2a: SiSignificance
gnificance Level- LeLevel--Let
t a = 0.05. N = a = 0.05. N =
4242= t=he the
numbenumber
r of dyads ranofked dyads
on these ranked on these
U - -732732-- Ra - Ra1 .5;- 11.5;
ttwo
wo meameasures.
sures.
uussinignFgormFormul
ula 4.1, wea f4.in1d,: we find:
SaSampling
mpling DistributDistribution--Since
ion--Since N> 10, N> 10,
z -- -U- 2U-16/3216/39.
9.34 - 5.203.4 - 5.20.
Student's t with dr= N- 2. The following
formula is used to determine the value of
ities (ibid., p. 247) reveals that z = 5.20
has a one-tailed probability under Ho of t for given observed values of rs (ibid., p.
p <0.00003. Since this p is less than
a----0.05, our decision is to reject H0 in
favor of H~. There is a greater perception VN-2
-- r8 1 - (rs)2
(4.3)
of threat [rom nations outside o[ the bloc
than [rom nations within the bloc.
rence under H0 of values as extreme as an
4.4.d THE LINKAGE OF PERCEPTIONS observed t wil be determined by reference
OF HOSTILITY TO PERCEPTIONS
to the "Table of Critical Values of t" (ibid.,
OF THREAT p. 248).
Reiection Region--Since H~ predicts the
Figure 2.1 indicates a hypothesized link
direction of the correlation, the region of re-
between hostility and tension (C1 --> C3);
jection is one-tailed. It consists of all values
the decision to trade "tension" for "threat"
of t which are so extreme that their prob-
carries over to this aspect of the exploration
of the model. Here, as will be the caseability
with of occurrence under H0 is equal to
or less
all the linkages examined, the use of dis-than a- 0.05.
Decision--The
crete measures precludes establishment of rankings of each of the
dyads on the two variables are shown in
antecedent-consequent (let alone, cause and
effect) relationships. Measures of correla-
tion are the best we can hope for;For these dthese
ata, ~ d2 -1,501.5. The value
of rs is found by substituting the observed
strengthen our confidence in the description value into the fol owing formula (ibid., p.
of the "cold war" interaction system. 2o6).
Nul Hypothesis- Ho: hostility and threat 6 ~d2
rs-- 1 ---------.
Na- N
are unrelated. Hi: there is a positive rela-
tionship betwe n perceptions of threat and
Thus, for thes dat ,
perceptions of hostility; i.e., as perceptions
of hostility increase, perceptions of threat 6 ( ~,50 ~.5)
rs- 1- = 0.88;
wil also increase.
Statistical Test- The data in Tables 4.1 using Formula 4.3, we find-
and 4.3 are the ranked measures of per-
ceived hostility and perceived threat. The t - 0.8 (13. 4) - 1 .7
Ref rence to the Table of Critcal Values
of t rev als that - 1 .7, with dr-40,
232102.3.375806.938... YAYIZUUYAOEZYOUEUEAZAOUOE 4232123O7953659180...55 422313079595086...5 ---1178422050432813....555 1321640015902266.....2255555
HostiRa27.nlkiOt5yN
CORRELATI Rank
Threat 325.0.5 - 3.5 72.20.12.12.25
TABLE 4.4
OF P~cnv~v HosTtarrY iND PERC~ TH:REAT
Dyad d d2
1. AE
26 36
4. EA
5. EI
8. IE
9. IO
10. IY
11. OA
12. OI
14. OZ -10.5
15. UE 2
16. UO 41 41
19. YI
4
21. YZ
21 19.5
24. ZY 0.5
25. AI 49
26. AU 21
27. AZ --3
28. EO
10
31. IU 17
33. OE
34. OY 9:
19.5 2.25
36. UI
37. UZ
16
6
40. ZA 14
41. ZI 16 -15
42. ZU
da- 1,501.5
has one-tailed proba ilty under Ho f 4. e T~ LIN4.4.e

KAGE OF REC IVED T~ LINKAGE OF RECEIVED
p<0. 0 5. Since this p is les than EXPBESSED HOSTILITY AND PERCEPTIONS
a = 0. 5, our decison is to rej ct Ho in OF THREAT
favor of H1. Ther is a positve relatmhip The linkage of hostility to threat in

betwe n perceptions of hostilty and percep- Figure 2.1 (C2 --> C3) implies a relation-
tions o~ threat. ship of the receipt of hostile messages (or
anbaotvieo.nal actor is considered "dependent" to tmeranmsbemirt) ewid ldecpoentdeaincymo. There dehypotpendehenscisy,
dependency
hostile messages is unrelated to perceptions notthese
gated. here themes
beisgoalsatisfied
withneeds.
whom ithen communi
Thecaticoncern
dependent
unilaterally. oactor
n di721-
ththemes
emes in mes ages) ainnd ,themessages)
perception of and ,the perception of
ththreat;
reat; this implica~othis
n was also iimplica~on
nvesti- was also investi-
communicates his feelings of dependence,
Null Hypothesis--Ho: the receipt of i.e., to whom he looks for aid in satisfying
of threat. Hi: there is a positive relation- NuU Hypothesis---Ho: the expression of

ship between the receipt of expressed hostil-
ity and perceptions of threat. rected toward the nuclear bloc member is
Statistical Test--The data in Tables 4.2 no greater than the expression of these
and 4.3 are the ranked measures of received themes in communication directed toward
expressed hostility and perceptions of threat.
the nonnuclear members of the bloc. Hi:
The Spearman r8 wil be used to test H0. a greater level of dependency wil appear
Signii cance Level--Same as 4.4.d, above. in messages directed to the nuclear bloc
Sampling Distribution---Same as 4.4.d,
members than to nonnuclear bloc members.
Statistical Test---The DEP Leary Process
Reiection RegionwSame as 4.4.d, above.
Score (Formula 3:5) is the measure of
Decision- -The rankings of each of the 42
dyads on the two variables are shown in
in effect, predicts that communication di-
Table 4.5. For these data, :~ d2 = 7,02 .5.
rected to the bloe4eader (i.e., the nuclear
The value of r8 is found by substitution
into~ Formula 4.4:
themes, i.e., have a higher DEP score, than
0 (7,022.5)
r8 -- 1 = 0.43; communication directed toward othernon-
74,046 nuclear bloc members. The U Test Wil be
using Formula 4.3, we find: used in the analysis of these data.
t- 0.43 (7.0) - 3.01. Significance Level- -Let a - 0.05, n2 =

13- the number of intra-bloc dyads where
Reference to the Table of Critical Values of
the nuclear member is not the receiver;
t (ibid., p. 248) reveals that t= 3.01 with
nl- 5 =the number of intra-bloc dyads
d[ = 40 has a one-tailed probability under
where the nuclear nation is the receiver.
Ho of p < 0.005. Since this p is less than
Sampling Distribution- -The probabilities
a- 0.05, our decision is to reject H0 in
favor of H~. There is a positive relationship as ociated with the oc ur ence under H0
between the receipt of expressed hostility o? values as smal as an observed U for n2
and perceptions of threat.
betwe n 9 and 20 are found by using
Siegers "Table K" (1956, pp. 274 fl.)
4.4.{ EXPRESSION OF DEPENDENCY which provides the probability of given
o~ BLoc LEADER critical values of U ......
The ?ourth "condition" (C4, Figure 2.1, Reiection Region---Since H~ states the
direction of the predicted difference, i.e.,
above) is specified as the dependency of
nonnuclear bloc members (i.e., Algo, Erga, DEP Scores will be higher where the nu-
Ingo, Yora, and Zena) on the nuclear mem- clear nation is involved in the intra-bloc
bers of the blocs (i.e., Omne and Utro). A communication, the region of rejection is
one-tailed. It consists of all values of U
the: degre thatit has goal ne ds which can- which are so small that the probability as-
s43211oO259048903.2.6c....DyadiZUYZIOYIaAYAOUOAYAtAIEed with their oc u2314322rO89675088..55ence under Ho is 21343213Ra25.6870955O920.n...55k 4.2: ----12131456d91.0098846.1..55
722 Hostility TABLE
Threat4.5 2156387946.25
1,34211850720.14622..2255
PdCHARD A. BRODY
CORRELATION OF RECEIVED EXPRESSED HOSTILITY AND PERCErVED THREAT
Rank d2
1. AE 35.5 42.25
18 36 324
35.5
4. EA
5. EI 31
6. EU 42
7. EZ
8. IE
9. IO
11. OA
12. OI 17
13. OU 49,
14. OZ 12.5 156.25
15. UE 121
41
17. UY
38 13
21
20. YU
21. YZ
22. ZE
23. ZO 40
37
25. AI
26. AU
27. AZ
28. EO 12 100
29. EY 4
--6
31. IU 12 144
32. IZ 25
-17
14 30.25
36. U!
37. UZ
38. YE
2 -16
41. Z!
20.25
d2 -- 7,022.5
ThevealuevoaflUuiesdoetfrmUinedisbydsuebstei ur-mined by substitu-

equal to or les than a- 0.05. tiononfthoefbsterhvedvoalbuesinetrovFoermdulavalue into Formula
Decision- The DEP Score of each of the
intra-bloc dyads is shown in Table 4.6, to- U- 80-R1- 13.
gether with the rank of the dyad, Ref rencetoSieg l's"TableK"rev alsthat
Forthese data, R2- 104 and R1- 67. U=13(nl=5;n2=13)has one-tailed
CONFLICT RESOLUTION VOLUI,IE VII NUMBER 4
de3.pendeIUncy dexirp6.4ercetsdiotnosw14.5
wiarld tbhe dnirueclteadr test H7.8.4.o. UAOYIA -14.3
-26.5 712.
-34.3
TABLE 4.6
DEPENDENCY THE~ES IN INTRA-BLOC DYADIC COMMUNICATION
DYADS INWHICHNUCLEAR NATIONISI:~.ECEIVER DYADS IN WHICH NONNUCLEAR NATION IS RECEIVER

Dyad Score Rank Dyad Score Rank
1. AU 6.4 14.5 1. AI - 7.2 10

2. EO 45.2 18 2. AZ - 2.1 12
3. EY 1.0 13
4. YO -25.9 3
5. ZU 32.5 17 5. IZ - 8.2 9
6. OE -22.9 4
9. UI -16.5 6
10. UZ -17.9 5
11. YE 6.8 16
12. ZA - 8.4 8
13. ZI - 3.4 11
--67 -- R1 ~: - 104- /h
probaprobability
bil ty under H0 of p < 0.05. Since under H0 of p < 0.05. Since towtoward
ard the nuclear member of a natithe
on's nuclear member of a nation's
this pthis
is les than a- 0.05, our decisiop
n is is less than a--0.05, our decision is owown
n bloc-al iance. bloc-alliance.
to reto
ject H0 in favor ofreject
H1. There~ is H0 in favor of H1. There~ is StaStatistical
tistical Test- The DEP Leary Proces Test--The DEP Leary Process
greatgreater
er tendency for dependency themes to tendency for dependency themes to Scores (Table 4.6) for the nonnuclear na-
ap eaappear
r in mes ages directed to nuclear bloc in messages directed to nuclear bloc tions directed toward the nuclear nation in
memmembers
bers than to non uclear bloc members. than to nonnuclear bloc members. their bloc (viz., the following dyads: AU:
EO, IY, YO, and ZU) is the measure of de-
4.4.g4.4.g
THE LINKAGE OF DEPENDENCY THE LINKAGE OF DEPENDENCY pendency. The nonnuclear nation's percep-
TO PERCEPTIONS OF THREAT tion of threat (Table 4.3) from the external
bloc leader (viz., the following dyads' AO,
Figure 2.1 'indicates a .hypothesized link
EU, IO, and
between the level of perceived threat YU, and ZO) is the measure of
threat. The
the level of dependency on the bloc-leader
Spearman rs will be used to
(C3--> C4). Theargument here is that
Significance LevelsLet a- 0. 5. N-
threat reduction is a goal-need experienced
5- the number of nations ranked on these
by the nonnuclear nations which they can-
two measures.
not unilaterally satisfy; aid and assurances Sampling DistributionsSince N is les
from the nuclear nation in their thown
an !0, the sampling distribution is i/N! bloc
will help satisfy this need, therefore
and the "probabil ty of oc ur ence under the
dependency on the bloc leader. H0 of any particular value of r8 is propor-
Null Hypothesis--Ho: perceptionstional to the number of permutations..giviof
ng the
external nuclear armed bloc leader as rise to that value" (ibm., p.. 210). Siegers
threatening are unrelated to expressions"Table
of P" (1956, p. 284) gives critical
values for r~ for N less than 10.
bloc leader of the nation's own bl0c-al iance. Rejection Region--Since H~ predicts the
Hi: the more threatening the nuclear leader direction of the correlation, the region of
of the external bloc is seen, the more rejection is one-tailed. It consists of all
critical values of r8 for N- 5 which are so
_ReiCnOfeTaNr2.1.eFbnLclIeCt4.To7SRie.EgAE.SerOs1L"2UT0aTbIleO41P"N rVOeveLa5.0
lU2.5
s MEthat VII NUMB
tFihnoavrems1.0 sinlgaEep4tlRu.i4nse:iga4vreintqoutiahnleictrye1.00
teau1.5
Rank on Perception Rank
TABLEon4.7 2.25
aslbeowowiutrilhtd;.idnDecfrpienaistediseonits RICHARD A. BRODY
724
CORRELATION OF DEP~NDF. NCY AND THEEAT
Nonnuclear cff Threat from External Dependency on d d2

Nation Nuclear Power Bloc Leader
3. I 5 2.5 -2.5 6.25

4. Y 2 1.0 -1.0 1.00
5. Z 3 4.0 1.0 1'.00
Z d" = 11.50
extrememe tehathtahte ptrhoebaprbilotbyaobfiloictyuroefncoec ur ence StatiStatistical

stical Treatment- The HOS Leary Treatment--The HOS Leary
uundender rH0H0is eiqsuaelqtuoaolrtleos otrhalnesa- 0t.h05a.n a- 0.05. ProcProcess
es Scores (Table 4.2) for the nuclear Scores (Table 4.2) for the nuclear
DDeecisisoino-nT-heTrheankrinagnkiofneagchonfoneacu-h nonnu- nationations
ns directed toward their non uclear, directed toward their nonnuclear,
clearrnanatiotni,on,tohne twhoevatrwiabolevas, risiashbolwesn, is shown nonanonallies
l ies (viz., the fol owing dyads: OA, (viz., the following dyads: OA,
UE, UE,
OI, UY, and OZ) is the meOI,
asure of UY, and OZ) is the measure of
For these data, Ii d2 - 11.5. The value of hosthostfifty.
fifty. The DEP Leary Proces Scores The DEP Leary Process Scores
rs is found by substituting the observed (Tab(Table
le 4.6) for the non uclear nations di- 4.6) for the nonnuclear nations di-
value into Formula 4.4: rectrected
ed toward the nuclear nation :in their toward the nuclear nation :in their
6 (11.5) blocbloc
(viz., the fol owing dyads; AU,(viz.,
EO, the following dyads; AU, EO,
rs-- 1- =0.43. IU, YIU,
O, and ZU) is the meaYO,
sure of depen- and ZU) is the measure of depen-
dencdency.
y. The Spearman rs is the measure of The Spearman rs is the measure of
correlation.
rs - 0.43 (N- 5) has a one-tailecl prob- Results--The rankings of each of the
ability under H0 of p > 0.05. Since this p nonnuclear nations on the two variables are
is greater than a' 0.05, our decision is to shown in Table 4.8. For these da. ta,
not reject Ho. Dependent themes appear to
d2 -3.5. The value of rs is determined
increase with perceptions o[ threat (the cor- by substitution of the observed value into
relation is in the right direction) but the
relationship may very wel be random. o (a.5)
rs = 1 .....= 0.825.
4.4.h HOSTILITY AND DEPENDENCY 120
The search for cor elates of dependency For N--5, this rs enters the "region of
led to the examination of a linkage not doubt" (i.e., 0.10 > p > 0.05) andis worth
specified inFigure 2.1- the relationship of
the receipt of HOS (hostile) themes from
the external nuclear bloc leader and the the receipt o[ hostile themes [rom the ex-
transmis ion of DEP (dependency) themes ternal nuclear power but the relationship
to the nuclear power in the nonnuclear na- may be random.
tion's own bloc. The hypothesized linkage
4'4.i Tm~ COHESION OF THE BLOCS
predicts this relationship unmediated by
perceptions of threat. However, since this The concept of "cohesion" (C5, Figure
hypothesis was not stated in advance of the 2.1) has enjoyed a somewhat checkered
running of the simulation, the formal pre- care r in group psychology. It tends to
sentation of the hypothesis wil be eschewed.
hf1951,reseTwoiqoune2.E ncoheizatsiovnea-l hypodaftbyhroaev0.0
.np.cyo9).nsor,istemorent fingedinn5egsrali,n"oTrhega5.0
Nation Nuclear Power Power dme.ot.hbFesoseirsm.veudlav4al.1u,ewein0.00
tfoinFdo:rmula 4.2a: 725
TABLE 4.8
COILRELATION OF DEPENDENCY AND HOSTILITY
Rank on DEP
Rank on Receipt
Nonnuclear
of Directed
I-IOS fromtoward
External Internal Nuclear d d2
1.A 1 2.5 1.5 2.25

3.I 3 2.5 -0.5 0.25
4.Y 2 1.0 -1.0 1.00
5.Z 4 4.0 0.0 0.00
:e da = 3.50
ofof"cohesion""cohesion"
leave something to be desired leave something to be desired com uniccommunication
ation frequency betwe n the frequency between the
frfrom
om the point of view othe
f operationalizapoint
tion: of view of operationalization: blocs. Hx:blocs.
the interaction betwe n bloc Hx: the interaction between bloc
"C"Cohesiveness
ohesiven s is the at raction of member- is the attraction of member- membersmembers
(i.e, com unication frequency (i.e., communication frequency
shship
ip in a group for iints memberas" (Bacgroup
k, for its members" (Back, in intra-bin
loc dyads) wil be greater than intra-bloc dyads) will be greater than
the interathe
ction acros bloc lines. interaction across bloc lines.
nes of a group is the resultant of al forces Sta isticalStatistical
Test- The frequency of mes- Test--The frequency of mes-
acting on al members to remain in the sages betwsages
e n dyadic om unication part- between dyadic communication part-
group" (Cartwright and Zander, 1960, ners is thners
e measure of interaction. The U is the measure of interaction. The U
p. 74). Test wil Test
be used to test he xperimental will be used to test the experimental
These caveats notwithstanding, two
themes continually recur in the literature Significance Level- Same as 4.4.a, above.
on cohesion: cohesion is related to uni- Sampling Distribution- -Same as 4.4.a,
formity of opinion and to communication
Rej ction Region- Same as 4.4.a, above.
Decis on- The frequency of communica-
and group research can aid in specifying tion for each of the 42 dyads is shown in
measures of cohesion. These findings are Table 4.9, together with the rank of each
reported by March and Simon: (1) "The
greater the cohesivenes of a group, the For thes dat , R1 - 5 1 and R2 - 352.
The value of U is found by substi ution of
greater the unifonuity of opinion . . (2)
The interaction of the group increases as
group cohesivenes increases" (1958, p. U - 732- R2 = 380;
60).
It has already been shown that there is a z = U- 216/39.34 = 4.17.
high degree of uniformity of opinion within Reference to the Table of Normal Probabil-
the bloc with respect to the hostility and ities (Siegel, 1956, p. 274) reveals that
threatening nature of the external bloc z--4.17 has a one-tailed probability under
(Sections 4.4.a and 4.4.c, above); the level H0 of p ,5 0.00003. Since this p is less than
of interaction is our second measure of co- a--0.05, our decision is to reject Ho in
favor of H~. There is more interaction
Nul Hypothesi - Ho: com unication fre- within the blocs than between the blocs--
quency within the bloc is no greater than the blocs are highly cohesive units.
action (a negative relationship). observed value iOnt7(o14,6F0,4o76r96m)ula 4.4: RICHARD A. BRODY
726
TABLE 4.9
MESSAGE FREQUENCY IN INTER- AND INTRA-BLOC DYADS
INTER-BLOC DYAI)S INTI1A-BLOC DYADS

Dyad Frequency Rank Dyad Frequency Rank
1.AE 33 3 1.AI 68 17.5
2.AO 99 30 2.AU 130 37
3.AY 34 4 3.AZ 69 19
4.EA 43 8.5 4.EO 161 41
5:EI 58 14.5 5.EY 94 29
6.EU 71 23 6.IA 80 26
7.EZ 41 7 7.IU 122 36
8.IE 50 19. 8.IZ 73 2,4
9.IO 70 21 9.OE 136 38.5
10.IY 5.9 16 10. OY 136 38.5
11. OA 85 27 11.UA 107 32
12. OI 58 14.5 12. UI 103 31
13. OU 121 35 13. UZ 111 33
14. OZ 37 5 14.YE 91 28
15. UE 46 10.5 15.YO 163 4.2
16. UO 117 34 16. ZA 68 17.5
17. UY 53 13 17.ZI 70 21
18. YA 43 8.5 18. ZU 154 40
19.Y! 79 25
20. YU 70 21 ~; -- 551 -- R~
21. YZ 31 2
22. ZE 39 6
23. ZO 46 10'.5
24. ZY 24 1
-- 352- Rs
4. ,j TH4.4,j
E LINKAGE OF COHESION THE LINKAGE OF COHESION Rei ctReiection
ion RegionmSame as 4. d, above. RegionmSame as 4.4.d, above.
AND TAND
HREAT THREAT Decis oDecisiorv--The
rv- The rankings of each of the rankings of each of the
dyads,dyads,
on the two variables, are shown in on the two variables, are shown in
FigureFigure
2.1 indicates a link betwe n per- 2.1 indicates a link between per-
Table 4.10.
ceivedceived
external threat and intra-bloc o- external threat and intra-bloc co-
For these data, ~ d" - 16,796. The value
hesionhesion
(C3- ~ C5); this is one aspect of (C3--~ C5); this is one aspect of
of rs is determined by substitution of the
the famthe
il ar "out-group" hypothesi . familiar "out-group" hypothesis.
Nul HNull
ypothesi - Ho: there is no rela- Hypothesis--Ho: there is no rela-
tionshtionship
ip betwe n perceptions of threat and between perceptions of threat and r8- 1- =-0.36;
levels of interaction (the measure of co-
hesion). Hi: the higher the level of per- using Formula 4.3, we find:
ceived threat, the lower the level of inter- t - 0.36 (6.78) - 2.4 .
Reference to the "Table of Crit cal Values
Statistical Test- Tables 4.3 and 4.9 pro- of t" (ib d., p. 248) reveals that - 2.4
vide ranked measures of perceived threat with dr- 40 has a one-tailed probabil ty
and level of interaction. The Spearman rs under Ho f p ~ 0. 1. Since this p is les
wil be used to test this hypothesis. than a- 0. 5, our decis on is to rej ct H0
Significance Level- Same as 4.4.d, above. in favor of H1. There is a negative relation-
Sampling Distribution- -Same as 4.4.d, ship betwe n perceptions o[ threat and level
above. of interaction.
CONFLICT RESOLUTION VOLUAIE VII NUMBER 4
23410569274.83.2.631....TZIUZEAEOAAEAAYUOAOOUOAYOCOHESION 43213325.O68O76950..5 23431117598004...55 and cohesion----21.a22113127r40O8656987O4e...555positively rela11t,,e428420.426342.306.462.1083d0581597373O46.200..22225555
SYSTEMIC EFFECTSThreatOFRank TABLE
SPREAD OFCohesi on Rank4.10 WEAPONS TECHNOLOGY
NUCLEAR
727
CORRELATION OF THREAT AND COHESION
Dyad d d2
29 -26 676
m 6:
4
8.5
5. EI -17.5
6. EU 39
7. EZ 3O.5 -23.5 552.25
8. IE
9. IO 361
10. IY 34 16 324
12. OI 14.5
14. OZ 25.5
15. UE 10.5 -11.5
49
17. UY 35
18. YA 8.5
19. YI
20. YU 21
21. YZ 16 -14 196
19.5 6 -13.5
756.25
24. ZY 729
25. AI
21 37
27. AZ 19 16
41 29
4OO
30. IA 26 16
31. IU 17 36 19 361
14 24 10 100
33. OE
34. OY 38.5 930.25
12.5 156.25
36. UI
3.5 29.5
38. YE
39. YO
17.5
41. ZI 21 40O
36.5
d2 -- 16,796.0
4.4.k
4.4.kxx-m
xx-mItELATIONSHIP
ItELATIONSHIP
OF DEPENDENCY
OF DEPENDENCY
Statistical Test- Since the DEP Leary
Figure 2.1 indicated a hypothesized link Proces Score is to some extent relative to
betwe n expres ions of dependency and co- the fi'equency of communication, it can ot
hesion (C4--> C5). meaningful y be cor elated with our mea-
Null Hypothesis--Ho: cohesion and de- store of cohesion (communication frequency).
penUeney are unrelated. Hi: dependency As a substitute measure of cohesion, the
3.owAFLn inProcessTabAlIe 4.Scoreil. 2.5
aLearybroevseh.1. (Formula 3.2)4 L1.5
oema.dership (Sub12o2.25
1ran(d-1.5 0rdinate 0 level)
7ss RICHARD A. BRODY
TABLE 4.11
CORRELATION OF DEPENDENCY AND AFFILIATION
Hank on DEP Hank on AFL

Nonnuclear Directed to Directed to d dS
Nation Nuclear Member Nuclear Member
2. E 5.0 5 0.0 0.00
4. Y 1.0 2 1.0 1.00

5. Z 4.0 3 -1.0 1.00
Z d2 = 6.50
6 (6.5)
was chosenrathe AFL score is a measm'e rs- 1- ~ = 0.675.
of the affiliative themes contained in mes- Ref rence to the "Table of Crit cal Values
sages. The Spearman q wil be used to of rs" (ib d., p. 284) rev als that rs - 0.675,
determine the relationship between these N- 5, has a one-tailed proba il ty under
two variables.
Ho f p > 0. 5. Since this p is greater than
Significance Level--Let a: 0.05. N --
a = 0. 5, our decis on is to not rej ct Ho.
5 = the number of nonnuclear--dependent Dep ndent hemes ap ear to be related to
--nations ranked on these two measures. af il ation themes (the cor elation is in the
Sampling Distribution--Same as 4.4.g,
right direction) but he relationship may be
Reiection t egionmSame as 4.4.g, above.

Decision- The rankings of each of the 4.4.1 mERARCHICAL ORDERING
non uclear nations, on the two variables, OF THE BLOCS
The sixth "condition" (C6) is specified in

For these data, ~ d2 - 6.5, the value of rs Figure 2.1 as the hierarchical ordering of
is found by substitution of the observed the blocs. A sytsem is "hierarchical" to the
value into Fo~xnula 4.4: degree that operationally differentiable
level)
0 0 0
FIG. 4.2. Commu
paidentifies
ollariaitnyc. es, wi"leaders"l be usaseddistributors
as our operinatitheironal above.
"lea"leaders"
ders" and "subordinates" can be dis- and "subordinates" can be dis- ceive anadnidnitniaiteiamoteremocomremuconmicamunition thcantion than
tintinguished.
guished. The smal group literature The small group literature the nnoonnunucclearlenartinaonst.ions.
Statistsitcaicl aTlesTet ~stCo~mComumnicmuniation cfartei-on fre-
information networks, i.e., leaders tend to quenciceiseisnitnhetihnetrian-btlroca-dbylaodcs dya(Tabldes (Table
receive and initiate more communication 44..9) is sthtehne cneescaersy ameryasumere;atshuereco;mtphaeri-compari-
than do subordinates (Bavelas, 1950; Bales, son wiwil lbebemadmae bdetwbee ntwinetrna-bilnotcrday-abdlsoc dyads
et al., 1951). whwheererethtehneucnuclearlenatrionanstairoenosnearoef tohne of the
The communication pattern associated ppaartiesestotohethcoemcmuomnimunicationcantdiointran-dbloicntra-bloc
with this leader/subordinate system is the whwheererethtehyeayreanroet noinvtolivnedvoinlvtehde cionmt-he com-
"wheel" (Figure 4.2, after Guetzkow and mumuninicactaiotnio. nTh. eTheU TeUst Tewilstbewiusledbein used in
Simon, 1955, p. 237). Guetzkow and the exepxlploroatrioantiofntohfis thhypisothypohesis. thesis.
Simon call this wheel pattern a "two-level Signnii[i[ciacnacne cLevLeel-v-eLle-t -aL-et0.a05-. nl0.'-05. nl '-
hierarchy." Arnold Wolfers has described 88--theenunumbe
mber ofrionftrai-nbtlorca-dbylaodcs idyan ds in
the communication pattern in the Soviet whwhiichhthtehneucnuclearlenartionan itsionnot iaspnoarty atoparty to
and American alliances as "wheel-like" the cocmommumuninicaticoant;ino2n-;1n2-0- th10-e numthbernumber
(1959, p. 7). The flow of communication of inntrtar-ba-lobclodcyadyads indswhinchwhithecnhuctlehaer nuclear
along the dyadic channels, within the bloc- nnaationonisiasparpatyrtoythteoctomhemuconmicamunition. cation.
SamppllinignDgisDitrsibturtibonut- iSoanm-e aSsam4.4e.f,as 4.4.f,
measure of hierarchy.a
Null Hypothesis- Ho: there is no differ- Rejection Region- Same as 4.4.f, above.
ence betwe n the frequency of communica- DecisionraThe frequency of communica-
tion directed toward and initiated by the tion for each dyad is shown in Table 4.12,
nuclear nations in the blocs and the fre- together with the rank of the dyad.
quency of communication directed toward For these data, Rx- 36 and R2- 135.
and initiated by the nonnuclear nations in The value of U is found by substitution of
the blocs. Hi: the nuclear nations wil re- the observed value into Formula 4.2.a:
U-- 1U--
35- R2 = 0. 135- R2 = 0.
a It shouald be noted that iIt
n several of the should be noted that in several of the
analyseanalyses
which fol ow, com unication flow is which follow, communication flow is ReferReference
ence to Siegel's "Table K" (1956, to Siegel's "Table K" (1956,
used as used
the operational measure of what ap ear as the operational measure of what appear
p. 27p.
4) reveals that U =274)
0, nl = 8; n2 = 10 reveals that U = 0, nl = 8; n2 = 10
to be thto
re dif erent variables (viz., "be
cohesion," three different variables (viz., "cohesion,"
has ahas
one-tailed probability undear Ho of one-tailed probability under Ho of
"hiera c"hierarchy,"
hy," and "tightnes of bipolarity"). My and "tightness of bipolarity"). My
contenticontention
on is that hes are but dif erent as- is that these are but different as- p < 0p.0 1. Since thi<s p is les th0.001.
an a = 0.05, Since this p is less than a = 0.05,
pects ofpects
the same factor; namely, modal oci of of the same factor; namely, modal loci of our dour ecision is to reject H0 in favdecision
or of H1. is to reject H0 in favor of H1.
interactinteraction.
ion. When the level of analysi is at he When the level of analysis is at the The nThe
uclear members act as distribunuclear
tors in members act as distributors in
to al sytotal
stem level, i.e, inter-bloc rs. intra-bloc, system level, i.e., inter-bloc rs. intra-bloc,
theirtheir
communication nets; the nets exhibit communication nets; the nets exhibit
we speakwe of interaction as a measure of thespeak of interaction as a measure of the
cohesivecohesiveness
nes of the blocs. When the level of of the blocs. When the level of the pthe roperties o[ the whe l strucproperties
ture- a o[ the wheel structure--a
analysi analysis
is at he bloc-al iance level, the modal is at the bloc-alliance level, the modal two-ltwo-level
evel hierarchy. hierarchy.
interactinteraction
ion pat ern can be used to identify a pattern can be used to identify a
bloc leadbloc
er (if one merges), in terms of his leader (if one emerges), in terms of his 4.4.m ~ TmHT~ESS OF
locationlocation
in a com unication net. When a in a communication net. When a BIPOLABITY
nation hnation
as be n identif ed as a "leader" in one has been identified as a "leader" in one
bloc andbloc
we investigate the pos ib lity of its and we investigate the possibility of its One question remains concerning the
being unbeing
dif erentiated from the "leader" of undifferentiated from the "leader" of communication structure and, consequently,
the othethe
r bloc, we are using modal interaction other bloc, we are using modal interaction
the international system prior to the spread
pat ernpattern
as a measure of the tightnes of the bi- as a measure of the tightness of the bi-
of nuclear capability: How marked is the
730dichoifmefermuNation
arrecnhticiaeltssinaaglrseoexcbhalausresaidvcNuclear
teolynrizwimoe threetdihtehstanainntcietocinoon'Nation
s- Test wil be used to tesRICHARD
t this hypA.othBRODYesis.
TABLE 4.12
COMMUNICATION FREQUENCY IN INTai-BLOC DYaa)S
Dyads which
Involve Communication Dyads whieh Communication
Nuclear f Bank Do not Involve f Rank
1. AU 130 13 1. AI 68 1.5
2. VA 107 l0 2. AZ 69 a
a. go 161 17 3. mY 94 S
4. OE 136 14.5 4. IA 80 6
5. IU 122 12 5. IZ 73 5
6.: UI 103 9 6. YE 91 7
7. YO 163 18 7. ZA 68 1.5
8. OY 136 14.5 8. ZI 70 4
9. ZU 154 16
10. UZ 111 11 ~: ----. 36 = R~
prpreference
eference for communicating to the leafor
der communicating to the leader ency to cocmommumunicnatiecawitethwithe lteahderleoafder of
ofofthenationthenation's
's own bloc? own bloc? their own blocs (i.e., AU, EO, IU, YO, and
ItItcan be acan
rgued that sinbece economargued
ic that since economic zu).
Statistical Test---Communication fre-
betwe n nuclear and nonnuclear nations quency (Table 4.9), of nonnuclear nations
(Table 3.1, above), the preference for com- directed toward the nuclear nations, is the
municating with Omne and Utro may needed measure. Comparison will be made
only reflect the search for economicaid and of communication within the bloc with
not for military as istance. If this is the communication across bloc lines. The U
case, we would expect that tendencies for
Significance Level- Let a- 0.05. nl =
own bloc leader would not emerge. From 5- the number of inter-bloc dyads in
the point of view of their ability to tender which the nuclear~ nation is receiver; n2 =
economic aid, there is no distinction be- 5-the number of intra-bloc dyads in
tween Omne and Utro; both have the which the nuclear nation is receiver.
Same economic potential and started at Sampling Distribution- The probabilities
the same economic level (Tables 3.2 and as ociated with the oc ur ence under H0 of
3.1, respectively). values as smal as an observed U for nl, n2
If a clear preference for communicating les than or equal to 8, are given in Siegers
with the internal bloc leader does emerge, "Table J" (1956, p. 271).
it Would seem justified to argue that the Reiection Region Since H1 states the di-
rection of the predicted dif erence, the re-
nomic dif erential and that a "tight". bi- gion of rejection is one-tailed. It consists
polar system emerged in the binuclear of al values of U which are so smal that
periods of INS-8. the probability of their oc ur ence under
Nul Hypothesis- Ho: non uclear nations H0 is equal tO or les than et- 0.05.
have no 'clear preference for the nuclear Decision- The frequency of communica-
nation with which they commtlnicate. H~. t i o n f o r e a c h d yad is shown in Table 4.13,
the non uclear nations wil exhibit a tend- t o g e t h e r wi t h t he rank of the dyad.
war system."
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY rathiibodvs.es comparison. . . . . . .
731
TABLE 4.13
COMMUNICATION FREQUENCY OF NONNUCLEAR NATIONS WITH NUCLEAR NATIONS
Inter-bloc Intra-bloc
Dyads f Rank Dyads f 'Hank
1 AO 99 5 1.AU 130 7
2.EU 71 4 2.EO 161 9
3.IO 70 2.5 3.IU 122 6
4.YU 70 2.5 4.YO 163 10
5.ZO 46 ! 5.ZU 154 8
~: -- 15- R1 ~; -- 40- R,
For
For these
these
data, R1-
data,
15 andR1-
Rg--40. Null Hypothesis--Ho:
15 and Rg--40. there is no differ-
The
The value
value
of U isof
determined
U is determined
by substitu- by substitu-
ence between the level of perceived threat,
tion
tion of the
ofobserved
the observed
value into Formula
value into Formula
from the external bloc, before and after the
4.2a: spread of.nuclear capability. Hi: the level
U -- 40- R2 -- 0. of perceived threat will be lower after the
Reference to Siegel's "Table J" (1956, spread of nuclear weapons.
p. 271)reveals that U, 0 (nl - 5; n2 - 5) Statistical Test--Our measure of threat
has a one-tailed probability under Hd of perception is, as before, Scale #2 on the
p < 0.004. Since this p is less than a 0.05, Systems Questionnaire (Appendix I). The
our decision is to reject H0 in favor of H1. comparison will be made on the 24 inter-
There is a tendency for nonnuclear bloc bloc ratings before and after spread took
members to communicate with the leader place (i.e., Periods 4 and 12, see Figure
of their bloc rather than with the external 3.2, above). The U Test will be used in
nuclear power--the system prior to the
spread of nuclear capability is not only bi- Significance LevelsLet a - 0.05. nl =
polar, it is tightly bipolar. 24- the number of inter-bloc dyads in the
pre-spread periods; n2 - 24- the number
4.5 The Nth-Country Situation:
Conditions and Linkages of inter-bloc dyads in the post-spread pe-
The reader will, we believe, grant, on
Sampling Distribution- Same~ as 4. b,
the evidence contained in theforegoing
section, that the"cold war system" as we
Region o/ t e] ction- Same as 4. a,
have defined it, emerged in the binuclear
periods of INS-8. In thissection we will Decis on- The level of perceived threat
attempt to assess the effects of the diffusion
for each of the 48 dyads is hown in Table
of nuclear weapons capability on this "cold
4.14, together with the rank of the dyed.
For thes dat , R1 - 81 .5 and R2- 364.5
4.5.a PERCEPTIONS OF THi~AT FROM The value of U is det rmined by substi u-
NATIONS EXTERNAL TO THE BLOC IN THE
tion of the observed value into Formula
IgTH-COUNTItY SITUATION
4.2b'
Figure 2.1 indicates a predicted reduc-
U -- 876- R1 -- 64.5.
tion in the level of perceived threat from
the externai bloc (i.e., C7) as a result Usingof Formula 4.1, we find:
the spread of nuclear capability. z - U- 288/47.94 -- 4.66.
73221it303.2.5.21..i.eYZAEAYAOOIEIsOU(Siegel, 19562222111054007498976,1063369p. 247) rev34212e876096.a5ls that bil211T4t3.7645y.Dya.hZOAUYAYOZUOUEEU"eYdeighth "cond11i86546678t17492ion" (C8, Fi2321g8947565u5...5re 2.1,
Score TABLE 4.14 Score
RICHARD A. BRODY
PERCEPTIONS OF INTER-BLOC THREAT
PI?~-$PtkEAD POST-SPtPOST-SPttEAD
EAD
Dyad
Dyad ~Rank Ranknk
35 1. AE
42 2. AO B
197 29.5 156
4. EA 36.5 4. EA
5. El 176 19
6. EU 257 45 10
7. EZ 36.5 7. EZ
8. IE 8. IE 201
9. IO 9. IO 192
10. IY 224 10. IY 165 11
11. OA 186 24 130
12. OI 154
13. OU 13. OU 175
14. OZ 29.5 172 15.5
15. UE 185 22
16. UO 47 2OO 32
17. UY 236 41 21
18. YA 18. YA
19. YI 189 19. YI 13
20. YU 172 15.5
21. YZ 13 21. YZ 178 2O
22. ZE 174 22. ZE 159
23. ZO 189
24. ZY 156
~: = 81 .5 = R1 ~:- 364.5- R~
RReef renfceteorhetanblecofenormtaolprotbahil-e table of normal probabil- 4.15, totgoegthetrhewirthwithe rtahnke orfanthke odfyatdh. e dyad.
For thehsesdeatda,tRa1, =R1301=a3n0d1/1a2n-d60/12. 2- 602.
z = 4.6 has a one-tailed probability under TThehe valauleuoefoUfisUfoisunfdobuyndsubsytitsutibnsgtithueting the
Ho of p < 0.0 0 3. Since this p is les than observevdedvalvuaelu(/e12()/i1n2to) iFnortmouFlaor4.m2au: la 4.2a:
a = 0.05, our decision is to reject Ho in U - -73723-2R-2R2= 13=0.130.
favor of H1. After the spread of nuclear From FoFromrumlau4l.a1,4we.1,fweind:find:
weapons, the perception of inter-bloc threat
z = U-216/39.34 = 2.19
is markedly reduced.
p = 0.0143.
4.5.b INTEn- VS. rNTRA-BLOC THREAT After the spread of nuclear capability
IN THE POST-SPREAD PERIODS there is still a significant difference between
the levels of inter- and intra-bloc threat.
Figure 2.1 does not predict the nature
There is more to threat than just "capa-
of the inter-bloc, intra-bloc comparison, on
threat perception after the spread of nuclear
4.5.C IMAGES OF THE BLOC LEADER
capability, but data on this comparison wil
AFTER THE SPREAD OF NUCLEAR WEAPONS
be presented to aid in the description of
the nth-country situation.
The level of perceived threat for each of above) relates to the level of dis atisfaction
the post-spread dyads is shown in Table with the hierarchy after the spread of nu-
1879 231854790 15 218956
S2c1Dyaa0l3,72.986521.et..aiYdOAAEEIAOAYUOOIsUErOfUZac tpioabniloirtyd.isHoScore1a5tw7isefvaecrti,onnowidti3118.1hr6722e..55tchemebTABLEloac- 4.hnMo1eu1l5Dyacp63.52.1.rlf.euAZZadAIYAUolrA.Ovenra,titohnes ubslScoreeo211cf07465324829elwaedrerdoims isnea43n1204c.e5asthabemouets733
SYSTEMlC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY
PERCEPTIONS OF INTER-BLOC AND INTRA-BLOC THREAT
INTEI~-BLOC DYADS INTl.-BLOC DYADS
Rank Rank
161 174
148 189 37.5
156
4. EA 187 4. EO
5. EI 176 5. EY
6. EU 164 6. IA 169 25
14.5 7. IU 151
201 41 8. IZ 149
192 9. OE
10. IY 165 10. OY 16.5
130 11. UA
154 12. UI 3
13. OU 175 13. UZ 164
14. OZ 172 14. YE 21
18~.
2OO 4O 11
17. UY 17. ZI 137 4
18. YA 169 18. ZU 161 18.5
19. YI 169 25
172 -- 301 -- R1
21. YZ 178
22. ZE 159
23. ZO
24. ZY 156 12.5
:e = 602 = Ra
##44, S,ysSteymstQeumstsionQuaire)s.tPiroe-npostaciorme-). Pre-post com-

sure of satisfaction-dis atisfaction was em- parirsoinsowners wemaderoenmaal ofdetheosnmaelsuroesf. these measures.
ployed in INS-8 (mea culpa). In alafloref' toerste,'thesatmse,ftivhe dysadsmwer five dyads were
By the use of indirect measures, we can usede(dviz(.,vAiUz,.EOAU,, IU, YEO,Oan,dIZU,). TYhOe , and ZU). The
say a great deal about the "image" of the fol owloinwg tianblge intdaibcaltes tihnedniuclatnedsext-he nul and ex-
bloc leader, held by the former nonnuclear perirmiemntaelnhytpaolthesy,ptohethvaeluseosf,Ut, hsige- value of U, sig-
nations after dif usion to k place. If this niffcaincaenlevcel olfetvhelobosefrvetdhUe, aondbstherved U, and the
image is positive, we can infer satisfaction; deciscoinsiuonderuHn0.dTehrdH0at u.pTonhwehidchata upon which
if it is negative, we can infer dis atisfaction. Table 4.16 is based are found in Table
4.17.
leader wil be taken as an indication of a Table 4.16 indicates no uniform change
similar at itude toward the hierarchy. in the image of the bloc leader, after nu-
Four measures were used to investigate clear diffusion took place. In communicat-
the image of the bloc leader: (1) HOS ex- ing with the bloc leader, the former non-
pres ed toward the leader; (2) DOM ex-
pres ed toward the leader; (3) perceptions but expres about the same level of hostil ty.
of the leader's "fi' endlines " (Scale #1,
Systems Question aire) and (4) percep- as friendly as before dif usion but les
tions of the leader's "helpfulnes " (Scale
734 tahlte ercodnobymictdheveilnoducpmendt swhpreichadmuofstnucprelcTABLEeadre 4.ct1ae6psTHEFitabtgihlureiBLOCitsyh.y2.p1otihnedisisc.atesDecia hypothes
sion under ized in-
PdCHARD A. BRODY
IMAGES OF THE BLOC LEADEBSa
Ho I~ al n~ U pb Ho . - .
HOSp,e- HOSpo~t HOSp,e < HOSpo~t 5 5 7 ns accept

DOMpro- DOMpo,t DOMpro < DOMpost 5 5 4 0.028 reject-wrong direction
Fr. pro- Fr.post Fr.p,o > Fr.post 5 5 13 ns accept
He.pro -- He.post He.pro ~ He.post 5 5 3 0.02.8 reject-right direction
aaDOMDOM
~-desire~-desire
to dominate to dominate b ppbabasedsoendSioengelS'si"eTgabele'sI" (1T9a5b6,lep. I2"71(1).956, p. 271 ).
HOSHOSm exprems eexpressed
d hostility hostility
Fr.Fr.~- pe~--perceived
rceived as friendly as friendly
HeHe.. - - pe----rceivedperceived
as helpful as helpful
ThiThis
s situatsituation
ion must have beenmust
frustratinhave
g been frustrating ppeercreciveidvhedlpfheulnpesful(thneeosther(tfhinediontghs er findings
foforr the dethecision-decision-makers
makers in Algo, Erga, Ingo, in Algo, Erga, Ingo, nnootwtwithisthansdtinagn)diisnag)signisofadsi sgatnisfoacftiodin.s atisfaction.
YoYora,
ra, and Zenanda. TheyZena.
~communicaThey~communicated
ted more more
ssubmissiveness
ubmis ivenes (i.e., les domina(i.ncee) t.o,Omneless dominance) to Omne 4.5.d.dTHTHEE CONCONDI
DITION OTFITOanNEAOFT W~TaTmnNEAT W~TmN
aand
nd UtroUtro
but felt lebut
s rewarfelt
ded for iless t (i.e., rewarded for it (i.e.,
OmneOmneand Utrand
o were sUtro
e n as les were
helpful seen as less helpful
to their bloc allies). This ambivalence re-
crease in 'mtra-bloc "tension" (now, "threat;"
flects the manner in which INS-8 was i.e., C9)after the nth-country situation
emerges. :As fellow bloc members gain ilin
weapons. Only the nuclear power differ- military capability, they should appeai rela-
ential was manipulated; the economic dif- tively more threatening to each other.
ferentials ~ere not,4 Even after the spread Null Hypothesismilo: there is nochange
of nuclear weapons the former nonnuclear in the level of perceived intra-bloc threat
nations had economic needs which they after the spread of nuclear weapons. Hi'
could not satisfy unilateral y- in other there wil be an increase in perceived intra-
words; while increased nuclear capability bloc threat after the diffusion of nuclear
changed their perception of threat from the
external bloc, it did not reduce the eco- Statistical Test- Our measure of threat
nomic dependence felt by the former non- perception is Scale ~2 on the~:Systems
nuclear nations. Questionnaire (Ap endix I). The compar-
If economic dependency continued into ison wil be made betwe n the ratings of
the post-spread situation and the leaders the 18 intra'bloc dyads before the spread
of their al iance were judged les helpful to k place (Period 4) and after experience
than before, we would expect a turning to with the n-nuclear world (Period 12; Figure
the opposing bloc leader for aid. If we 2.2, above). The U Test wil be used to
find this to be the ease (se 4.5.j, below),
it may be legitimate to argue that reduced
Significance Level- Let a - 0.05. nl -
18- the number of intra-bloc dyads prior
4 This is highly unrealistic if one thinks of
to spread; ns- 18- the number of intra-
bloc dyads after spread.
independent nuclear development in the real
world- it is, however, highly defensible in Sampling Distribution- The probabil ties
terms of experimental controls. as ociated with the oc ur ence under H0
CONFLICT P~ESOLUTION VOLUME VII NUMBER 4
178436 162276598wiit-107.7
4521.YZ2.AE1.UO EOAU-106.8 enobtlo,c. ont163r89ary to Figure 2.1RRR,58341726aaannnkkkp ear
tdhoetsh175 TABLE 4.17 Rank
IMAC~ OF THE BLOC LEADERS: DATA ON HOSTILITY, DOMINANCE, FRIENDLINESS, AND HELPFULNESS
A.A.
HOS Themes
HOS Expressed
Themes to Bloc Leader
Expressed
Pre and Post (R~--, 33)to Bloc Leader Pre and Post (R
Dyad Pre HOS Score Rank Post HOS Score
1. AU - 38.0 9 - 48.9
3. IU - 54.8 7 -63.8
4. YO - 21.2 10 - 67.5
5. ZU - 62.8 5 -62.7
B. DOM Themes Expressed to Bloc Le
Dyad Pre DOM Score Rank Post DOM Score
1. AU 18.7 6 - 1.7
2. EO 0.4 4 17.1
3. IU 31.7 8 25.0
4. YO 55.9 10 - 3.3
5. ZU 32.3 9 32.7
C. Perceived Friendliness of Bloc
Dyad Pre Fr. Score Rank Post Fr. Score
10
2. EO 134 3 128
3. IU 171 9 142
4. YO 124 i 143
5. ZU 152 6 161
D. Perceived Helpfulness of Blo
Dyad Pre He. ~core Rank Post He. Score
10
152 1
3.IU
143 129.
ofvaluesa smal sanobservedUfor ThischangeThis

ap earstostemdirectlyfrom change appearsto stem directly from
9<n2< 0ared terminedbyusing thespreadthe
ofnuclearweaponscapbilty- spread of nuclear weapons capability---
Sieg l's"TableK"(1956,p .274- ).
Rei ctionRegion-Sameas4. f,above. to be mediated by the level ofdis atisfaction
Decison-Thel velofperc ivedthreat
foreachdyadis howni Table4.18
Forthes dat,R1-386.5and/t2= 4.5.e COHESION AFTEa T~E SPREA~
279.5 ThevalueOfUisdet rminedby OF NUCLEAR WEAPONS
substi utionoftheobservedvalue(R1) The tenth "condition" (C10, Figure 2.1,

intoF rmula4.2b: above) is a predicted decrease in'the co-
U495-R1- 108.5 hesivenes of the blocs. In the pre-spread
Ref rencetoSieg l's"TableK"(1956,p. system, it was found that there was a
27 )rev alsthatU-108.5(nl-18;n2 marked tendency for nations to interact
=18)has one-tailedproba iltyunderH0 within the bloc rather than betwe n the
ofp<0. 5 Sincethispisles than blocs. The prediction is that this tendency
-0. 5,ourdecisonistorej ctH0infavor wil not persist after the dif usion of nuclear
ofH1Afterthespreado[nu61earweapons, capability.
ther isani creas inthep rception:of Null Hypothesis--He: there is no differ-
threatfrom embersoftheinternalboc. ence between the level of interaction within
1 4 814.5 2536
tThiis110.nep652.7439s.tDyadreZUAIYOYEUZsOYAUOEAUrAdaiscitnepisoetnrh.nieoThetdiaslt.aycTheoftmrhopemhypasraPi1111mse347784o1683262r09enitohdwiestl7istbhwieartmaoluwaTABLEgbdhesRa232n29k.54.at1bh8oavte.the bloc pe1114658r72499sisted unchanged23128659i.nto the
INTRA-BLOC THREAT PERCEPTION
Pre-s?ore Post-score Rank
1. AI 31 174 32.5
34
3. AZ 156 2O8
4. EO 17.5
5. EY 144 14 140
15 169
151 19
8. IZ 152 2O 149 17.5
136
12 24
157
12. UI 16 132
4.5 164
116 2
136 172
155 21
17. ZI 137 10
18. ZU 130 161
~l- 279.5- R~ -- 386.5- Rx
and and
betwe n the blocs. Hi: thebetween
tendency the blocs. Hi: the tendency
to into
teract along blocinteract
lines (found in the along bloc lines (found in the n-nuclear world.)
pre-pre-spread
spread system) persists after the spread system) persists after the spread Sampling Distribution- Same as 4.4.b,
of nof
uclear capability. nuclear capability.
Statistical Test--Message frequency in
Reiection Region- Same as 4.4.a, above.
each of the 42 dyads is the measure of Decision- The dyadic communication
frequencies are shown in Table 4.19, to-
betwe n inter-bloc and intra-bloc dyads. gether with the ranks of the dyads. For
these data, R1 = 418 and R2 = 485. The
used to investigate cohesion in the pre- value of U is determined by substitution of
the observed value (R2) into Formula 4,2a:
U- 732- R2 = 247.
12. The over-al increase in communication
From Formula 4.1, we find:
rate (4.5.f, below) in the post-spread pe-
z - U- 216/39.34 - 0.79.
riods precludes meaningful, direct, pre-post
comparison. The U Test wil be used in the Reference to the table of normal probabil-
examination of this hypothesis. ities (Siegel, 1956, p. 247) reveals that
Significance Level- Let a = 0.05. nl - z- 0.79 has a one-tailed probabil ty under
18- the number of intra-bloc dyads; n2- H0 of p- 0.2148. Since this p is greater
24 = the number of inter-bloc dyads. (The than a- 0.05, our decision is to not reject
terms of reference "inter-bloc" and "intra- Ho. There may be as much interaction
bloc" refer to the dyads in their initial bloc (cohesion) between the blocs as within the
configuration; i.e., at time "to," they are blocs--the marked preference [or communi-
used to make comparison with the pre- cating within the bloc, which existed prior
spread system pos ible and not to indicate to nuclear diffusion, is no longer evident.
CONFLICT RESOLUTION VOLUME VII NUMBE~ 4
22. ZE 75 9.5 737
SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR W
TABLE
TABLE4.19 4.19
MESSAGe.
MESSAGe. FRSQUENCYFRSQUENCY
rN INTER- AND INTRA-BLOC
rN DYADS
INTER- AND INTRA-BLO
INTER-BLOC DYADS INTBA-BLOC DYADS

Dyad f Rank Dyad :[ Rank
1. AE 64 3 1. AI 68 5
2.AO 168 38 2.AU 127 27
3.AY 60 I 3.AZ 72 6.5
4.EA 79 11.5 4.EO 254 42
5.EI 75 9.5 5.EY 111 21
6.EU 148 33 6.IA 67 4
7.EZ 80 13 7.IU 160 36
8.IE 74 8 8.IZ 89 15
9.IO 157 34.5 9.OE 211 41
10.IY 91 16 10.OY 157 34.5
11. OA 128 28 11. UA 118 24
12. OI 139 31 12. UI 137 30
13. OU 185 40 13. UZ 120 25
14. OZ 116 22 14. YE 106 20
15. UE 121 26 15. YO 163 37
16. UO 181 39 16. ZA 72 6.5
17. UY 93 17 17.ZI 79 11.5
18. YA 63 2 18.ZU 143 32
19.YI 96 19
20. YU 117 23 2: -- 418 -- R~
21. YZ 95 18
23. ZO 136 29
24. ZY 86 14
-- 485 -- R2
TABLE 4.20
MESSACE FREQUENCY IN SYSTEM-BUNS
Run Pre f Rank Run Post f Rank
1. 305 26 1. 472 33
2. 212 11 2. 237 14
3. 180 8 3. 212 11
4. 303 25 4. 321 27
5. 262 18 5. 292 22.5
6. 238 15 6. 459 31
7. 176 7 7. 276 20
8. 166 5 8. 226 13
9. 157 4 9. 422 30
10. 297 24 10. 620 34
11. 212 11 11. 469 32
12. 184 9 12. 325 28
13. 174 6 13. 246 16
14. 254 17 14. 330 29
15. 263 19 15. 292 22.5
16. 133 2 16. 154 3
17. 98 i 17. 286 21
-- 208 ~ R~ ~ -- 387 -- P~
rnieuxnfp.ie1c1r23796584a..inmteintcarleiansterinvetnot1i1ao7763425839ln.5.coThemmunire iscatsiogn-1263578.5 otpfievrVapcley>rivis0aetnd.a1cb0telh.errmeee--v1a56te0326a.a5slusrteh.at:
Rank Pre TABLE
Rank Post4.21 - 5.4.5 360909..
25.330.20.360..02255
738 PdCHARD A. BRODY
CORRELATION OF MESSACE FREQUENCY Prm AND POST BY RUN

System-run d
o 16 -1 1.0
4
16 10
14
-1 1.0
13 100.0
10. 4.0
11. 9.5
11
13.
12
8.5
16. 1 - 1.0 1.0
1
d2 -- 333.00
4.5.f RELIABILITY OF COMMUNICATION po lapoolable.

ble. A Friedman Two-Way Analysis A Friedman Two-Way Analysis
FREQUENCY
N = 7 = number of nations
A rough estimate of the stability of com-
K = 17 = number of replications
munication frequencies (a ubiquitous meas-
ure in this study) can be gained by an
df=16
2-- (0.005) (80,262.5) 378-- 23.3
examination of message frequencies in the Xr
seventeen system-runs before and after the

If we take into ac ount within-world
variability, the relationship of pre-post com-
after the spread of nuclear weapons (tiaough munication frequency is higher than f we
not neces arily because of it). Table 4.20 ag regate within worlds. Table 4.2 shows
contains the pre-post totals for each system- the rankings of the 42 dyads pre and post.
For these data, r8 = 0.71; t = 6.43; p <
Forthes dat,U=5 (nl- 17;n2= 0.0 05. Communication frequency is a rela-
17);fromSieg rs"TableK"(1956,p.274)
wefind:thatp<0. 1.Ther isasignif-
4.5.g THREAT AND COHESION AFTER
icantincreas inthel velofinteractioni
THE SPREAD OF NUCLEAR CAPABILITY
Periods7through12.
This ncreas notwithstanding,ther la- Figure 2.1 indicates a linkage betwe n
tivepositonofthesytem-runsi fairly threat perception and cohesion (C7- +
constant.Table4.21indicatestherankings C10); this linkage was found in the pre-
ofthes vente n"worlds"preandpost. spread system (Section 4.4.j, above) but it
Forthes dat,r8=0.59;t=2.85;p< does not hold up after the shift in level of
0.01.
The message-frequency data are also Nul Hypothesis- Ho: there is no rela-
23456792. UYOZAAE
1SYSTEMIC 231475698. 32416879O05. above. --1235468950378..5 436196840..025
214.2.6.71.9844356Dya.. YAEEAEZOAZAOIOZUYEUOUUOUEOAEd EFFECTS OF311O408..55SPREAD OF NUCLEAR
349.5 useDWEAPONS
dectiostoens-t t-Th11i23TECHNOLOGY
6s5e..5hryapnoktihnegsiso.f each2272.100.2864.53141d966020y..,0022a55d, on
739
TABLE 4.22
COILRELATION OF MESSAGE FREQUENCY Ptm AND POST BY DYAD
Rank PrRank
e Pre Rank Post d dfl . dfl ..
3
38
3. AY
11.5
5. EI 25.0
23
8. IE
182.2.5
10. IY 16
1.0
12. OI 14.5
240.25
34
- 6.5
19. YI
20. YU 21 23
21. YZ
23. ZO 10.5 342.25

169.0
25. A! 156.25
-10 100.0
27. AZ -12.5 156.25
28. EO 41 1.0
29. EY 29
30. IA
31. IU 36 0
32. IZ 81.0
34.5
36. UI 31
-8 64.0
38. YE
42 25.O
40. ZA 121.0
41. ZI 11.5
4O
ds -- 3,530.5
tionship between perceptions of threat and

level of interaction (our measure of co- Significance Level- Same as 4.4.d, above.
hesion). Hi: the higher the perceived Sampling Distribution- Same as 4.4.d,
threat, the lower the interaction.
Statistical Test--Tables 4.15 and 4.19 Rej ction Region- Same as 4. d, above.
provide ranked measures of perceived threat
and interaction. The Spearman rs will be the two variables, are shown in Table 4.23.
432C110.0984367.46526.O...ZZYIOZAOAEZOIUOUENEUYAYOOEUFLICT R(6E)S(1O4L,7U4234311T85609772..5IO)N VOLUME324133O76059884VI9..5I NUMBER 4---4213317659185O0...5 1,7342624516805874825934720.961160622...O000222555
7402.4.3. EAAYAO -34.-22.-15.5 506.961.132.02255
240.
RICHARD A. BRODY
TABLE 4.9.3
COIlRELATION OF THREAT AND COHESION
Dyad
Dyad Threat Rank Cohesion Rank d d2
1. AE 18.5
12.5 -11.5
36 11.5 1,190.25
5. EI
21 12 144.0
14.5
8. IE 41 1,089.0
9. IO - 4.5
16
26
10 21
13. OU 81.0
81.0
17. UY 17
18. YA
19. YI 19
20. YU
21. YZ 18 -15
22. ZE 16.5 9.5
29
25. AI
110.25
27. AZ 1,260.25
28. EO 1,296.0
29. EY 21 16
30. IA -21
31. IU 72.9.0
32. IZ
33. OE 1 41
34. OY 16.5 18
35. UA
36. UI 27
37. UZ 21 16.0
21 1.0
11
41. ZI 11.5
42. ZU 18.5 32
0!' -- 14,744.5
For
Forthese
these data,
data,
~d2 ~d2
- 13,689.5.
- 13,689.5.
The value
The value t- -0.195 (6.45) ---1.26.
of
ofrsrsisis
determined
determinedby substitution of the of the Reference to the "Table of Critical Values
by substitution
observed
observed value
value
intointo
Formula
Formula
4.4: 4.4: of t" (Siegel, 1956, p. 248) reveals that
t--1.26 (dr- 40) has a one-tailed prob-
r8- 1- ---0.195. ability under Ho of p > 0.05. Since this p
74,046
is greater than a---0.05, our decision is to
Using Formula 4.3, we find: not reject Ho. The predicted link between
t15.
on13.
4.EO s. YO
h1.ese variablUZ
eAI 5161815210abov-14
12.5 16 i6.251
-2.5169
e.13 196 741
SYSTEMIC EFFECTS O
TABLE 4.24
CO~TION OF THREAT AND COI-IESION: INTllA-BLOC
myad Threat Rank Cohesion Rank d d~
2.AU 17 11 - 6 36
3.AZ 18 3.5 -14.5 210.25
5.EY 4 8 4 16
6.IA 14 I -13 169
7.IU 7 15 8 64
8.IZ 6 6 0 0
9.OE I 17 16 256
10. OY 10 14 4 16
11. UA 9 9 0 0
12. UI 2 19. 10 100
14. YE 12.5 7 - 5.5 30.25

16. ZA 8 3.5 -4.5 20.25
17.ZI 3 5 2 4
18. ZU 11 13 2 4
2: d~ --
ththreat
reat and cohesiandon does nocohesi
t emerge. Thois n does not emerge. This SamplinSampling
g Distribution- Same as 4.4.d, Distribution--Same as 4.4.d,
nnegati
egative findinvg semfis tonaridise fnromgtheseems to arise from the
vvari
ariabil tyaobif thlreiatpyercepoftion inthreat
the ex- perception in the ex- Rejection Region- Same as 4.4.d, above.
teternal
rnal bloc- -thisbljudogmc---thi
ent is based onsthejudgment is based on the Dec,ion- The rankings of each dyad on
ffiindinng diin thne ngextisenctionthe. next section. the two variables are shown in Table 4.24.
For these data, ~ d2 - 1,298. The value of
44..5.h5rN.Tha~-BrNTa~-BLOC
LOC T~mEAT XNV COaESXONT~mEAT XNV COaESXON r8 is determined by substitution of the ob-
served value into Formula 4.4:
FFiigurge 2ure.1 predict2.s a 1negapredi
tive relationc- ts a negative relation-
6(1,298)
shshiip betpwe nbetween
perceived intra-bloperceic threat ved intra-bloc threat r8 -- 1- ---0.34.
5,814
anandd level ofleinvelteractionof(C9 - i>nC10teracti
). on (C9 --> C10).
NNulul HyplothHypothesi
esi - Ho: there is no resla---Ho: there is no rela- Using Formula 4.3, we find:
titionsohipnshibetwe npercbetween
eption of threat witpercepti
hin on of threat within t---0.34 (4.25) ---1.445.
ththee bloc anbld inotercactiandon withinitnheteracti
bloc. on within the bloc.
Reference to the "Table of Critical Values
HH~:~: the grtheeater thegreater
perception of thrtheeat, perception of threat,
of t" (ibid.) reveals that t- 1.445 (dr--
ththee lower lthoe wer
level of intheteractionl(ceohvelesion). of interaction (cohesion).
16) has a one-tailed probability under H0
SStati
tatistical Tse tit- Tcaballes 4.1Test--Tabl
5 and 4.19 es 4.15 and 4.19
of 0.05 < p < 0.10. Since this p is greater
cocontai
ntain data onnthedata levels of thonreat pethercep- levels of threat percep- than a- 0.05, but lies in the "region of
t-t-iion anodni terandaction initnhe teracti
intra-bloc dyados. n in the intra-bloc dyads. doubt" in the predicted direction, we do not
TThehe SpearmSpearman
an r8 wil be used to ter8st thiswil be used to test this reject Ho but consider the finding worth
hhypothesi
ypothesi . ? s. ? investigating in the real world. Threat per-
SSiignifgicniancefLeviecl-ance
Let a = 0.05Level
. N = --Let a = 0.05. N = ception and level of interaction (in intra-
1188 = the=numtheber of innumber
tra-bloe dyads rankofed intra-bloe dyads ranked bloc dyads) appear to be negatively related
but the relationship may be random,
above.
mental intervention. CAPABILITY
742 RICHARD A. BHODY
TABLE 4.25
COMa~tONmATION Fr~-Q~"~NC? IN INTaA-BLOC DYAnS
Dyads which Dyads which

Involve Nuclear f Hank Do not Involve [ Hank
Nationa Nuclear Nationm
1.AU 127 11 1.AI 68 2

2.UA 118 9 2.AZ 70. 3.5
3.EO 254 18 3.EY 111 8
4.OE 211 17 4.IA 67 1
5.IU 160 15 5.IZ 89 6
6.UI 137 12 6.YE 106 7
7.YO 163 16 7.ZA 72 3.5
8.OY 157 14 8.ZI 79 5
9.ZU 143 13
10. UZ 120 10 Z ---- 36 = B~
Z = 135-- Rs
a The label "Nuclear" refers to the nation prior to the experimental intervention.
4.5i HIE4.5.i
RARCHY AFTEH THE SPREAD HIERARCHY AFTEH THE SPREAD Rej cjteiocntRieogin-RSaemgiaosn4.-f,Sabomve. as 4.4.f, above.
OF NUCOF
LEAR WEAPONS NUCLEAR WEAPONS Deciscoins-iTohne-freTquhenecyforceomquenica-y of communica-
The levThe
nth "conditon" (Cl, Figure eleventh "condition" (Cll, Figure tion fnorfeaochrdyeadcish odwynaindTaibsle 4s.2h5,own in Table 4.25,
2.1, abov2.1,
e) is described as dis olution of above) is described as a dissolution of togegthertwhiethrthwie ratnkhofthe dyrad.nk of the dyad.
the hierthe
a chical structure of the blocs. In hierarchical structure of the blocs. In For trhesthdaets,eR1d-a36tan,d R2-1 353. 6 and R2- 135.
this ectthis
ion, we wil examine the com uni- section, we will examine the communi- The vealuveaofluUeis fofundUbyisubfstoi utniodofby substitution of
cation ocation
f the blocs to det rmine whether of the blocs to determine whether the oebseorbvesdevraluve indtovFoarlmuela 4i.n2a:to Formula 4.2a:
the twothe
"whe l" structures found above two "wheel" structures found above U = 1=35-1R325- -R02. - - - 0.
(Section(Section
4. 1) continue after the xperi- 4.4.1) continue after the experi- Ref rfencre tonScieg lt'so"TaSbile Kg"e(ilb'sd.,"Table K" (ibid.,
p. 2742)7re4v)alrs tehvateUal=s0t(nhl -at8;Un2=10) (nl - - 8; n2 = 10)
Nul Hypothesis- Ho: there is no dif er- has sonae-toainled-prtoabileidty upnrdeobHaobfil ty under Ho of
ence betwe n the frequency of intra-bloc p ( 0. 01.0Sinc1e.thSisnlocisel sthiasn alo- 0i.s5,les than a - 0.05,
communication directed toward and initi- our drecidsoencis tio rnej citsHotion fraveojreocftH xH. o in favor of H x.
ated by Omne and Utro and that directed The hei rha icehrcalrstcuhtiucraelofstherbuloctduoerse of the bloc does
toward and initiated by the former non- not btreabkrdeowankundeorwthne imupnactdoefrthethe impact of the
nuclear nations. Hi: Omne and Utro wil spreradeoafdnucolefarnweuapcolnes.aTrhiswea cpleaorlnys. This is a clearly
continue to receive and send more com- negatgivaetfinvdeingf!iPnerhdaipns tghe!aPnelyrsihanps the analysis in
munication than the former nonnuclear na- the sectsioenctoifolnowtofilfsoheldoswme lwfightil shed some light
tions within the blocs. on potpenotialernatsoinaslforethaescontsinufanocerofthe continuance of
Statistical Test- Communication fre- the hei rhaicehyr?archy?
quency in the intra-bloc dyads (Table 4.19)
wil provide the data base. The comparison 4.5j T.HjETIHOETNTESI$OHF THNE BEIPSO$LABOITFY THE BIPOLABITY
wil be made betwe n intra-bloc dyads in AFTEH THE $PHEAD OF NUCLEAH
which Omne and Utro are parties to the
interaction and dyads in which they are not. The hierarchies have persisted in the
The U Test wil be used in this analysis. nth-country system but the perception of
Significance Level- -Same as 4.4.1, above. external bloc threat was significantly re-
Sampling Distribution- Same as 4.4.f, duced. If the hierarchies are based solely
on military neces ity, these are inconsistent
CONFLICT HESOLUTION VOLUME VII NUMBER 4
htelepfeucloneosm. ic dif erentials which were part above. 743
Y E Z A
FIo. 4.3. Pre-spread international communication.
findinfindings.
gs. However, the findings can be con- However, the findings can be con- WWee founfdoinutnhedprei-nspretadhsey tpemrae-spread system a
sideresidered
d consistent f some other motivation cleareparef prerncefoer inetnracteionfwoitrh thienteraction with the
consistent ff some other motivation
for sufor
bmis ion to the hierarchical strsubmission
ucture to the hierarchical structureleadaerdoefra naotifn'saownabtloic.oTnhe'sworldwn bloc. The world
can becan
found. be found. communmuication siyctaemtilonkedslyikesthem lo ked like the
CluesClues
to this motivation are present in to this motivation are present in idealeizadlimzoed l inmoFigurde 4.l3;in thFis mgoduelre 4.3; in this model
the "ithe
mages of the bloc leader" held b"images
y the of the bloc leader" held by the
the let elrstare nratisonsarned thnealitneisorneprse-and the lines repre-
former nonnuclear nations (Section 4.5.c, sentnhte mtohdael comom udnicaltiocnolinmks.munication links.
above). The leaders were seen as being Thihs ciosm cuonimcatimuon synteimcawtsiloabnel dsystem was labeled
less "helpful" and just as "friendly" after a "ti"gthig(.eh,thie(ria.ceh.ic,al)hbipeolra itry.c"hItical) bipolarity." It
spread took place. If threat is reduced, the remaminsatoinbesdetormibnedwdhetherthmiscionme-d whether this com-
locus of this decreased helpfulness might mmuunicantioncsaytieomnpersiytsatloenmgwitphehresists along with h e
reasonably be expected to be in the military persirtsenicsetofetnhechei roafchitesh. e hierarchies.
realm. Unfortunately, the participants were NNuul HlypoHythesip-oHt:hesfoirsm-erHno -: the former non-
not asked to differentiate between kinds of nuculeacrlneatironsnhavetniocnlesarphreafvrencneforclear preference for
the belocbleadoercwlitehawdheomrthwiey ctohm whuni- om they communi-
It was argued above (Section 4.4.m) that cate.tHei:.thHie fo:rmtehr neo fnuoclreamneatrionsnon uclear nations
if a clear preference for interaction with the wwiil clonticnouento eixnhiubteateondencxyhtoicbomit- a tendency to com-
leader of a nation's own bloc emerged, the mmuunicanteipcraimtaerilypwrithmthaerl aidleyr ofwithetirh the leader of their
hierarchy must have some basis beyond the own bloc.
economic dif erentials. By the same reason- Statistical Test--Communication fre-
ing, if no clear preferences are exhibited, quency (Table 4.19), of former nonnuclear
the hierarchy might be considered to reflect nations directed toward Omne and Utro,
is the needed measure. Comparison will be
of the starting situation and persisted made between frequency within the bloc
throughout INS-8.5 and frequency across bloc lines. The U
5 Stilman and Pfa f (1960) argue that he
Test will be used.
cohesion of the real world bloc-aliances has Significance Level--Same as 4.4.m, above.
already decreased under the impact of in- Sampling Distribution--Same as 4.4.m,
creased econ mic ap bilty and that along
with this decreased cohesion has gone a dis-
solution of the post-W I hiera chical decison
Rej ction Region- Same as 4. m, above.
structures (p. 147 if. ) DecisonJThe com unication frequency
ep-co-n0om.5i0.c dSiifnecrenthialis. p is greater than i4.dea3;liazenewd modelsystemappeahasrs asemerged.
744 Figure 4.4Thi. s
PalCHARD A. BRODY
TABLE 4.26
COaXCATION Fm~QUF~CY or NONN~CL~R NATroNS wrrH NUCLF~R NATIONS~
Inter-bloc Intra-bloc
Dyads f Rank Dyads f Rank
1. AO 168 8 1. AU 127 1
2. EU 148 4 2. EO 254 10
3. IO 157 5 3. IU 160 6
4. YU 185 9 4. YO 163 7
5. ZO 136 2 5. ZU 143 3
25 -- 28 -- R, 25 = 27 = R1
a The labels "Nuclear" and "Nonnuclear" refer to the nations prior to
ooff each each

dyad is shodyad
wn in Tableis4.26,shown
to- in Table 4.26, to- longelonger
r described by the model in Figure described by the model in Figure
ggerber
erber with the rwith
ank of eachthedyad. rank of each dyad.
FFor
or these dthese
ata, R1- 27 adata,
nd R2- 28. R1--27 and R2--28.
TThe
he value ovalue
f U is givenofby ForUmulais4.2a:given by Formula 4.2a:
UU- 40--- R2 40-- 12. R2 -- 12. 4.6 Summary: The Status of the Model
ReReference
ference to Siegel's "Tabtole J" Siegel'
(ibid., s "Table J" (ibid., We are now in posses ion of sufficient
pp.. 271)271)
reveals threveals
at U - 12 (nl - 5;that
n2 - 5) U - 12 (nl - 5; n2 - 5) data to permit an answer to the "macro-
hahass a one-at ileone-tailed
d probability under Hoprobability
of under Ho of question" posed above: "Is the interna-
tional system different before and after the
a - 0.05, our decision is to not reject Ho and spread of nuclear weapons?" The answer
reject H1. After the spread of nuclear is an unqualified "Yes."
weapons, the former nonnuclear nations In the binuclear system, al of the pre-
may be as likely to communicate with the dicted "conditions" were established, but
external bloc leader as with their own bloc two of the predicted linkages (C3--> C4
leader. The bipolarity appears to fragment; and C4---) C5) were not.
the hierarchies appear to be based on the The system was comprised of two hier-
archical, cohesive bloc-alliances. The na-
The world communication system is no tions of the external bloc were perceived as
Fro. 4.4. Post-spread international communication.
tnieointhbertwceorn ethlaetibolncsowaefs ichiaernatEXPERIc waerTIisGzeHThMdiENTALgbBIhyPOLARINnshTERVENTI
yuSascYtTHREATENI I
lSeebmTaeEMr."nsyOrsNteepmeNaGtisedt.i)ghtly bipolar.
NUCLEAR
NUCLEAR
DIFFERENTIAL TRANSMIS ION OF PERCEPTI
DIFFERENTIAL
ON OF TRANSMISSION OF PERCEPTION OF
WIWITHIN
THIN THE BLOC (GIVEN) HOSTILITY THE
BETWEEN THE EXTERNAL
BLOC (GIVEN) HOSTILITY BETWEEN THE EXTERNAL
THETHE
BLOCS BLOC AS HOSBLOCS
TILE BLOC AS HOSTILE
DEPENDENCY
DEPENDENCY
ON NUCLEAR PERCEPTION OF THE ON NUCLEAR PERCEPTION OF THE
BLOC MEMBER EXTERNAL BLOC AS
INTRABLOC COHESION
HIERA CHIALBOCS
I
~ARITY (GIVE~
BLOC LEADER SEEN
'~ EXTERNAL BLOC SEEN
AS LE
AS LESS
INTERNAL
/
AS MORE TH
/
INTRABLOC CO
I
BIPOLARITY FRAGMENTED
FIC. 4.5? A model of empirical findings INS-8.
momore
re hostile andhostile
threatening than thand
e na- threatening than the na- the post-spread situautaiotnio.nI. Isnumsu,mth,ethbie-bi-
titions
ons of the inteofrnal blothe
c. The cominternal
munica- bloc. The communica-
Four key elements of the pre-spread
a greater level of expres ed hostil ty than system are dif erent after the spread of
the communication within the blocs. The nuclear capability. (1) threat external to
non uclear nations expres ed dependency the bloc is reduced; (2) threat internal to
upon the nuclear bloc leader- a depen- the bloc is increased; (3) the cohesivenes
dency which ap ears to be related to out- of the blocs is reduced; and (4) the bi-
group hostil ty and threat. (However, polarity is fragmented. All of these ad up
to a step-level change in the "cold war
enough to rule out random oc ur ence.)
The dependency on the bloc leader and the Figure 4.5 is a reformulation of Figure
hierarchical ordering of the blocs stem from 2.1, in light of the tid ngs in INS-8. (To
both military and economic insuf iciencies- aid comparison with Figure 4.5, Figure 2.1
the economic insuf iciency continued into
DIS OLUTION OF AD HOEXPERIC MENTAL INTERVENTION
NUCLEAR DIFFERENTIAL
(N.B.'
(N.B.'C1C1indicates
indicates
thethe
ith ith
condition
condition
in a chain
in a of
chain
conditions
of conditions
which which
is
is hypothesized
hypothesized as as
being
being
relevant
relevant
in theintransformation
the transformation
of a steady
of a steady
state
stateAAinto
intoanother
anothersteady
steady
statestate
B.) B.)
WITHIN THE BLOC (GIVEN)

SENSITIVITY TO REAL OR PERCEP-
TION OF OUT-GROUP (OTHER BLOC)~-
HOSTILITY (C1 AND C2)
INCREASED INTRABLOC COHESION (C?) . . . -
AD HOC HIERARCHY WITHIN BLOC (C6)
NUCLEAR PARITY (GIVEN)
INCREASED DISSATISFACTION REDUCTION OF PERCEIVED OUT-

--....
WITH AD HOC HIERARCHY (C8) GROUP THREAT (C7)
INCREASED INTRABLOC TEN-

SiON (C9)
REDUCED INTRABLOC COHESION

(C!0)
HIERARCHIES (C1 )
-.,,.
RELATIVE FRAGMENTATION OF BLOCS;
A "STEP-LEVEL' CHANGE TO A NEW
"STEADY STATE" (C12)
Fig. 2.1. Hypotheses linkings for INS-8.
itoanaeyuIpredi,NtlhichawoS-8onsthe
sofrcCHAPTER ,renclregi"gestidwabeopafynsiostnteredrwasVnethereal
recnxeaccompl siwhirmaimulyceahtntworld
spesuccesses. ohisdethatntgivagheatsltliiomespovelnaetonu-d"l TheconfisOrhaycisg,tehsmopmedst?entcsrhttueodiethalearsstetgucacontent
ioedsthahed;inivAgneterna tdrlmeeyn,pletnswhaatninnetahandaidts.leyweftshiinsehaidrsevmealaebtiomesrnuaritsrhoenrilpyg
Questions of generalizing: from ins-8
performed before we wil lfe le netirerleyly
the same qualifies as the standard measur-

system, after the spread of nuclear capabil- ing instrument, e.g., semantic-differential.
it-y, conformed in most respects to the theo-
retical model posited in advance of the ex- But these "suc es es" by no means elimi-
periments. The pre-spread system was a nate al problems and reservations concern-
plausible model of reality, i.e., the pattern ing the meaning of the findings. To ques-
of interaction, which developed prior to tion the "meaning" of the findings is to
diffusion of nuclear capability, resembled make public the author's concern about the
generality of knowledge derived from the
since the end of World War I . laboratory. What is the relationship of
This "pat ern matching" is highly desk- knowledge about the behavior of the simu-
able. With respect to the validity of ex- late to knowledge about the real system?
perimental findings, Zelditch and Evans
of the simulate to the real international
late the investigator ought to run a suf-
ficiently long series of uniformity trials to We do not argue that the inter-nation
show that, left to itself, the simulate be- simulation is a complete representation (i.e ,
haves like the proces simulated at least is total y isomorphic) of al aspects of the
international system. However, we do not
(1962, p. 60). The real world pat ern
matching in the pre-spread periods of INS- gained knowledge about the behavior of
8 would se m to meet this requirement.
Of les direct theoretical import but none- situation. But where betwe n these ex-
theles significant was the development of tremes, on this continu m of generalizabil-
sensitive measuring instrt ments- partic- ity, do the findings from INS-8 fal ?
ularly the technique of content analysis. Any answer to this question must derive
The instruments wil permit the study of initial y from the as umptions of the simula-
cold war phenomena outside of the lab- tion designers about the cor espondence of
oratory. Further, purely methodological in- "concrete" and "analytic" structures (Levy,
vestigations of the inter elations among 1958) in the simulation to the same struc-
measuring instruments wil have to be tures in the real international system, and
interorine.lTheory
atloiadencesscr.ibe actions found in both research situation
748 model
erate
tory aconditioning
tlife.isionusuallysMoreover,
neienddidesigned
tcfactors
oatRICHARD
besetherwhiesassumed
toeexperimental
tc.hachieveTA.hvaisBRODY
rtoiiasbnottop-lhese
ulultimately
timately must derive frommust
the pragmaticderive from the pragmatic
tetest--prediction
st- prediction of real world behavior. of real world behavior.
Among
Among the "concretethe
" structure"concrete"
s (i.e., structures (i.e.,
obsobservable
ervable entities or types entities
of behavior) or types of behavior) actual effects observed outside of the lab-
rerepresented
presented in the simulation arine "natithe
on," simulation are "nation," oratory but equivalent kinds of effects using
"d"decision-maker,"
ecision-maker," "al iance," "trade," "war," "alliance," "trade," "war," the most economical representations of
"m"message,"
es age," and "internationaland
organizat"international
ion." organization." causal variables" (1962, p. 109).
SoSome
me of the "anofalytic" the
structures"analytic"
(i.e., non- structures (i.e., non- The ability to set key variables and their
obsobservable
ervable relations and procrelations
es es) repre- and processes) repre- interrelations in motion and to achieve
sesented
nted in the INS ainre "publthe
ic opinion,INS
" are "public opinion," equivalent kinds of effects or to achieve a
~, ? . ? ,, ,, ? o ,,
communlcauon, degree of correspondence between labora- commu
"validation toryof decisions,"
and real world structures is a function a
of the adequacy of the theory which in-
From these two lists, it can be seen that forms the construction of the operating
the "concrete" structures appear most dif-
ferent after the transposition from the nat- and interrelations are central to the reality
ural setting to the laboratory. The "analytic" being modeled; the correspondence of the
structures are linguistic conventions used model to reality after operation (i.e., its
predictiveness) is a test of the salience of
set ings and are, by definition, isomorphic the variables chosen for inclusion in the
in both loci. simulate, and of the fidelity of the repro-
In the context of these concepts the prob- duction of the interrelations among these
lem of generalizing fi'om the artificial set ing variables.
to the political world is the problem of Failure to predict is a signal that adjust-
determining whether the change in "con- ments in the theory need to be made--new
crete" structure also alters the nature of variables need to be included and/or old
the performance of the unit in its relation-
ships with other units. For example, the es ence of the theory-bufiding aspect of
proces by which a decision is reached in laboratory experimentation: a cycle of
the simulate is more apt to be the same THEORY-BUILDING, EXPERIMENTA-
after transposition to the laboratory than is TION, PREDICTION, OBSERVATION,
the content of decisions under the same THEORY-BUILDING ... until the model
transformation. This belief informed the is fine-tuned.
decision to concentrate on interaction proc- INS-8 is the culmination of three phases
es in this study. of this cycle: theory-building and model
Snyder contends that, "the primary aim construction in INS's I through 7; experi-
of laboratory representations is to set key mentation and prediction in INS-8. Obser-
variables and their interrelationships in vation of the predicted phenomena, in the
motion and to create, explicitly, imitations real world, must await the arrival of the
of and substitutes for the hidden explana- nth-country situation.
?yourn_ation. 5 : : 6 -:
courtofetheous worlId situation.
nature
APPENDIX
Policy Planning Aid #3

This document is designed to aid you in your planning of future international decisions. The
questions are a simplified form of the kinds of questions foreign policy decision-makers ask of the
world situation. The fact that your answers may differ from time to time reflects the changeable
In addition to helping you clarify your thinking about tile other nations in the world, this
document wil serve as a permanent record of your views of the world situation and, thereby,
help us to understand your picture of the world for experimental purposes.
In the questions below you wil be asked to rate the other nations on various qualities; two or
more nations may be rated as equal but a rating must be given to each of the six nations.
Example: Indicate the courteousness to your nation of the other sLx nations.
: extremely
1 : :
very
g.3quite
:
slightly
: '
: 4
slightly
:
quite ' - very extremely discourteous

(Note, that in the example nations 2 and 3 are rated equally, while nations 1, 4, 5, and 6 are
rated non-equally. In the example, numbers were used for the nations; in your answers, you will
use the first letter of the nation's name: A, E, I, O, U, Y, and Z.
1. Indicate the friendliness to your nation of the other six nations.
extremely ?riendiy
very~ quite slightly
I ' : I
l
I : I I 1
slightly quite very extremely unfriendly
g. Indicate the likelihood of each of the other six nations becoming involved in
extremely likely very quite slightly

slightly quite vet37 extremely unlikely
3. Indicate the trustworthiness of the other six nations in their dealings with your
: I i I i 1
extremely trustworthy very quite slightly

slightly quite very- ' extremely- untmstworthy
4. Indicate the helpfulness to your nation of the other six nations.
: I : :
extremely helpful yew ' quite slightly

i : : I l
slightly quite very extremely harmful
5. Indicate the degree of cooperation or conflict which the other six nation
extremely co perative : very ' quil~e " slightly

I : : i l
' slightly quite very extremely conflicting
Expesgraphed.
earcrhimiEn)eLD,nIntsteiL.rnnPoP.atilo"itPnioaclaliRtGaieclaamltiiGaonng,sm,"i1Amen9g.61".rU.icaSn. versiD2tEReDemocratic
BLOOMFI New 81.
(1i3dro0956)uyvn-cdi5ate2York:ltywn.,iPo,Press,489-n5rlai4etsRandom
T(c91,es3.1960.91t"6i0n5Ame)3g,.S3House,e4r-RICHARD
iInstitutions, iv5c7a1960..ne,n.P1do9l.6i2t.iA.cal BRODY
Science
750
t[EFERENCES BrBrtONOWSKI,
tONOWSKI, J. The Common Sense of Science. J. The Common Sense of Science.
ART, C. Design for a Strategic Model. Wal- BROWN, H. and REXL, J. Community o[ Fear.
tham, Mas .: Raytheon Co., 1961. Santa Barbara: Center for the Study of
AMBAS AnE DE FamNCE. "France's First Atomic
Explosion." Service de Pres e et d'Informa- BRZEZINSKI, Z. The Soviet Bloc: Unity and
tion, February 13, 1960, p. 34 (Mimeo- Diversity. Cambridge, Mass.: Harvard Uni-
BAcx, K. "Influence through Social Communi- BtmNs, A. L. "Military Technology and Inter-
cation," Journal o[ Abnormal and Social national Politics," Yearbook of World Af-
Psychology, 46 ( 1951 ), 9-23. fairs, 1961, London: Stevens and Sons, Ltd.
B~.ATON, L. and MADnOX, J. The Spread of ( 1961 ), 185-203.
Nuclear Weapons. New York: Praeger, 1962. CAMPBELL, D. T. and LEVINE, R. "A Proposal
BENSON, O. "A Simple Diplomatic Game--Or for Cooperative Cross-Cultural Research on
Putting One and One Together." University Ethnocentrism," The Journal o~ Conflict
of Oklahoma, August, 1958. (Mimeographed.) Resolution, 5 ( 1961 ), 82,-108.
~. "A Simple Diplomatic Game--Or CAMPBELL, D. T. and STANLEY, J. C. "Experi-
Putting One and One Together." University mental Designs for Research on Teaching."
of Oklahoma, September, 1959. Paper read In N. L. GAGI~ (ed.), Handbook of Research
at the American Political Science Association, on Teaching. Chicago: Rand McNally Co.,
September, 1959. (Mimeographed.) 1963.
BF.S~IERS, J. M. "Models and Theory Construc-CAPLOW, T. "A Theory of Coalitions in the
tion," American Sociological Review, 22 Triad," American Sociological Review, 21
(1957), 32-8.
CARTwmGRT, D. and ZANnEa, A. (eds.).
Army War Col ege, Carlisle Barracks, Penn., Group Dynamics, 2nd ed. Evanston, I 1.:
November, 1959. (Mimeographed.) Row, Peterson, 1960.
~. "Political Gaming," U.S. Naval In- COrIEN, B. J. "Political Gaming in the Clas -
stitute Proce dings, 86 (1960), 57-64. room," Journal of Politics, 24 (1962), 367-
BLOOMFIELD, L. P. and PADELFOILD, N. "Three
DAvis,R.P olitcalGaming.CarliseBar-
Political Science Review, 53 (1959), 1 05- racks,Pen .:U.S ArmyWarColeg,
15. March, 1959.
BROnBrCK, M. "Models, Meaning, and Theories." DEUTSCH, K. W. "Mechanism, Organism, and
In L. GRoss (ed.), Symposium on Sociolog- Society: Some Models in Natural and Social
ical Theory, Evanston, Ill.: Row, Peterson Science," Philosophy of Science, 18 (1951),
& Co., 1959.
BRODY, R. A. "Deterfence Strategies: An An- ~. "Toward an Inventory of Basic Trends
notated Bibliography," The Journal of Con- and Pat erns in Comparative and Interna-
flict Resolution, 4 (1960a), 443-57.
"Three Conceptual Schemes for the
Study of International Relations," Journal of "Dis emination of Nuclear Weapons," United
Student Research, 2 (1960b), 8-25. Re- Nations Review, 6 (1960), 15-6.
printed in R. SNYrER, et al. (eds.), Foreign DOTY, P. "The Role of the Smal er Powers,"
Policy Decision Making. New York: Free Daedalus, 89 (1960), 818-30.
Press of Glencoe (1962), 250-74. DroVER, M. J. Conceptual Structure and Group
BRODY, R. A., BEACH, P., and SAMUELS, E. Proces es in an Inter-Nation Simulation.
Chronologies of Runs in INS-8. Evanston, Princeton, N.J.: Princeton University and
Ill.: Program of Graduate Training and Re-
FESTINC~, L. and KATZ, D. Research Methods
BRODY, R. A. and NOEL, R. Inter-Nation Simu- in the Behavioral Sciences. New York:
lation Participants' Manual. Evanston, Il .:
Program of Graduate Training and Research FRANIC, I. "The Great Antagonism," Atlantic
in International Relations, 1960. Monthly, 202 (195.8), 58-62.
CONFLICT RESOLUTION VOLUME VII NUMBE~ 4
2ssearch
Moc3rip-tt5o.0r).s andCorpTeachiorationn,g.DeEnglfenseewoodSystemClsiDiffs,vi- 5lPower,
vBi2se.ahedrssiictymanuscri 3-9.
C'oP"urBehaviersse,:1pGa9t.5o8mral). inScig oernce,Sim5ul(a1t960)ion,E240-xer-
GAMSON,
GAMSON,W.W.AATheory
Theory of of
Coalition
Coalition
Formation.
Formation. flict,"Geflict,"
n ralMot rsCorporation,Def nse General Motors Corporation, Defense
Ann
Ann Arbor:
Arbor:University
UniversityMicrofilms,
Microfilms,
Inc.,Inc.,
1959.1959. SystemsDSystems
ivsion,March29,1961.(Unpub- Division, March 29, 1961. (Unpub-
GIFFrN,
GIFFrN,S.S.F.F."Tomorrow's
"Tomorrow's Military
Military
Matrix,':
Matrix,':
World
World Politics,
Politics,1414(1962),
(1962),
433-8.
433-8. KAa-rN, H. Report on a Study of Non-Military
GOLDmaMEa,
GOLDmaMEa,H.H.and andSPEmR,
SPEmR, H. "Some
H. "Some
Ob- Ob- Defense. Santa Monica, Calif.: The RAND
servations
servationsononPolitical
PoliticalGaming,"
Gaming,"
World
World Corp. ( 4P R-322-RC), 1958.
Politics,
Politics,1212(1959),
(1959),71-83.
71-83. KAPr~N, M. System and Process in International
GUETZKOW,
GUETZKOW,H.H."The "The
UseUse
of of
Simulation
Simulation
in the
in the Politics. New York: Wiley & Sons, 1957.
Study
Study of
ofInter-Nation
Inter-NationRelations,"
Relations,"
Behavioral
Behavioral KAPLAN, M., BtmNS, A. L., and QUANDT, R. M.
Science,
Science,44(1959),
(1959),183-91.
183-91. "Theoretical Analysis of the 'Balance of
GUETZKOW,
GUETZKOW,H.,
H.,ALGER,
ALGER,C.,C.,
BRODY,
BRODY,
R., NOEL,
R., NOEL,
R.,
R., and
andSNYDER,
SNYDER,R.R.C. C.
Simulation
Simulation
in Inter-
in Inter-
national
nationalRelations:
Relations:Developments
Developments[or [or
Re- Re- KING, J. E., Jrt. "Col ective Defense: The
Mil tary Com itment." In A. WOLFERS
N.J.: Prentice-Hall, 1963. (ed.), Al iance Policy in the Cold War.
GUETZKOW, H., BRODY, R., DRnTF_aa, M., and Baltimore: The Johns Hopkins University
BEACH, P. An Experiment on the N-Country Pres (1959), 103- 45.
Problem through Simulation. St. Louis, Mo.: KIS INGER, H. A. The Nec s ity [or Choice.
The Social Science Institute, Washington New York: Harper, 1961.
University, 1960. KNoma, K. (ed.) NATO and American Secur-
GUETZKO~V, H. and Sr~ON, H. "The Impact of ity. Princeton, N.J: Princeton University
Certain Communication Nets upon Organiza- Press, 1959.
t-ion and Performance in Task-Oriented LF~Y, T. Interpersonal Diagnosis of Personal-
Groups," Management Science, i (19'55), ity. New York: Ronald Press, 1957.
LEVY, M. "Some Aspect of 'Structural-Func-
HARRISON, J. O. and LE , E. M. The Stratspiel tional' Analysis in Political Science." In R.
Pitot Model. Bethesda: Operations Research YOUNC (ed.), Approaches to the Study of
Office, The Johns Hopkins University, 1960. Politics. Evanston, Ill.: Northwestern Uni-
HELMER, O. and RESCram, N. "On the Epis-
temology of the Inexact Sciences," Manage- MALCOLM, D.G. (ed.) Report of System Sim-
ment Science, 6 (1959), 25-52. ulation Symposium. Baltimore: Wavefly
HEa~ANN, C. F. and HERMANN, M. G. On the Pres , 1958.
Pos ible Use of Historical Data for Valida- MARCH, J. and SnwoN, H., with GrmTZKOW, H.
tion Study of the Inter-Nation Simulation. Organizations. New York: Wiley & Sons,
China Lake, Calif.: U.S. Naval Ordnance 1958.
Test Station, 1962. MARCENAU, H. "The Competence and Limita-
HILSMAN, R. "On NATO Strategy." In A. tions of Scientific Method," Operations Re-
WOLFERS (ed.), Alliance Policy in the Cold search, 3 (1955), 135-46.
War. Baltimore: The Johns Hopkins Uni- McCLELLAND, C. "A World Politics Game."
versity Press, 1959, 149-83. San Francisco State College, 1959b, 1-5.
HOWARD, W. D. "The Computer Simulation of ( Mimeographed. )
a Colonial Socio-Economic System." General ~. "Improvement in Technique in the
sion, May 1 , 1961. (Unpublished manu- cises." San Francisco State Col ege, 1959a,
1-1 . (Mimeographed.)
IKLk, F. C. "Nth Countries and Disarmament," MEADOWS, P. "Models, Systems and Science,"
Bul etin of the Atomic Scientist , 16 (1960), American Sociological Review, 2 (1957),
391-4.
INGLIS, D. "The Fourth-Century Problem: MEY R,H."OntheHeuristcValueofScien-
Let's Stop at Three," Bulletin of the Atomic tifcMocles,"Philos phyofScience,18
Scientists, 15 (19'59), 22-6. (1951),1 -23.
KAEHLER, R. C. "A Systems Engineering Ap- MmBRATH,L."PredispositonsTowardVoting
proach to the Problems of International Con- Decisons,"PROD,I(1958),26-9.
We184-Sysat2pe12.monss,"TeAmechnolricoagy.n EcWaonosmhincgtReovn,iewD.,C5.0: 83.relgtUCrio.aSapnti.hc.seNaoednal"Wargaming
3-11. rRe.nv)asAnalloevaOrecdyhnBsi,ae"snh.OpcaM.evtheiToA.reastThesilColdioApSntasptsRe,irWar,"oUninas,ceha1vre9ersics6U..h1",.tSIyn.6
MILBBURN,
URN, T. W.T. "W.The"CTohencepCot onfcDepteroefncDee: ter ence: PublPublic
ic Opinion Quarterly, 25 (1961), 167- Opinion Quarterly, 25 (1961), 167-
SomeeLoLogicgalicanldaPnsydchPsoloygcihaol CogniscdaelrCoa- nsidera-
tions,s",J"oJournuarl nofalSocfialSIoscuieasl, I1s7 u(1e9s6,1),17 (1961), PatcE, M. T. "Applying Wargaming to the
Cold War," PROD, 3 (1959), 3-6.
MmLF_a% W. L. "The American Ethos and the
Alliance System." In A. WOLFmS (ed.), Naval Institute Proceedings, 85 (1959), 44-
Alliance Polic~t in the Cold War. Baltimore: 7.
The Johns Hopkins University Press, 1959, RIc~, R. P. "Simulation as an Aid in Model
31-48. Building," Operations Research, 3 (1955),
MODELSKI, G. The Communist International 15-9.
System. Princeton, N.J.: Center of Inter- PtUMMEL, Pt. J. Technology and War: A Cor-
national Studies, 1960.
MORgENSTEaN, O. "The N-Country Problem," of Hawaii, 1961.
Fortune (March, 19'61 ), 136 if. SCHF, LLrNC, T. C. "An Experimental Game for
? The Question of National Defense. the Study of Bargaining," World Politics, 14
New York: Random House, 1959. ( 19.61 ), 47-68.
MOUgENTm~U, H. J. "Alliances in Theory and -------. The Strategy of Conflict. Cambridge,
Practice." In A. WOLFEnS (ed.), Alliance Mass.: Harvard University Press, 1960.
Policy in the Cold War. Baltimore: The SCODEL, A. J., et al. "Some Descriptive Aspects
Johns Hopkins University Press (1959), of Two-Person Non-Zero-Sum Games," The
Journal of Conflict Resolution, 3 (1959),
NATIONAL PLANNING ASSOCIATION. 19'70 With- 114-9.
out Arms Control: Implications of Modern SmCEL, S. Nonparametric Statistics. New York:
McGraw-Hill, 1956.
N.P.A., 1958. Smmnz, M. "Simulation, Its Uses and Potential,
. The Nth Country Problem. Washing- Part I," Anticipation Project Expository and
ton, D.C.: N.P.A., 1960. Development Paper, No. 2, General Electric
Ngwco~, T. M. "Autistic Hostility and Social Corp., Operations Research and Synthesis
Reality," Human Relations, i (1947), 69-86. Consulting Service, June 25, 1958. (Mimeo-
ORCUTT, G. H. "Simulation of Economic
o "Simulation, Its Uses and Potential,
(1960), 893-907. Part I ," Anticipation Project Expository and
OSGOOD, C., SucI, G., and TANNENBAIYlVI, P. Development Paper, No. 3, General Electric
The Measurement of Meaning. Urbana, I 1.: Corp., Operations Research and Synthesis
University of Il inois Pres , 1957. Consuiting Service, May 4, 1959. (Mimeo-
Osgo D, Pt. E. NATO: The Entangling Alliance.
Chicago: University of Chicago Pres , 1962. Snv~oN, H. and NEWEL , A. "Heuristic Prob-
PADELFORD, N.J., SESVELL, G. H., and H~RI- lem Solving: The Next Advance in Opera-
SON, T. "Report on Political Exercise on Sim-
ulated Far Eastern Situation." Mas achuset s (1958), 1-10.
Institute of Technology, January, 1961, 1-14. SNYDF~, R. C. Deterfence, Weapons Systems,
( Mimeographed. ) and Decision-Making. China Lake, Calif.:
PAINE, G. "Problems and Uses o? the Single
Case in Political Research." In G. PAINE, ? "Experimental Techniques and Political
The Korean Decision (June 24-30, 1950): A Analysis: Some Reflections in the Context of
Reconstruction of Decision-Making Events,
Doctoral Dis ertation, Northwestern Uni- D.C. CaAaL~WORTa (ed.), The Limits of
versity, 1959, Chapter 10. Behavioralism in Political Science. Philadel-
PEruTONE, A. and PtEICHLINg, G. "Group Co- phia: American Academy of Political and
hesion and the Expres ion of Hostility." In Social Science, 196g.
D. CXRTWRmHT, and A. ZANDFaa (eds.), SOCIAL SCIENCE DMSION. Experimental Re-
Group Dynamics, 2nd ed. Evanston, Il .: search on Political Gaming, RAND Report
Row; Peterson, 1960, 141-51. P-1540-RC. Santa Monica, Calif.: BAND
POOL, I. and ABELSON, R. "Simulmatics Project," Corporation, November 10, 1958.
8St2.anford, Calif.: Stanford University Pres , lCorpora tion, 1962.es: A Methodological Anal-
ated Bureaucraci
STn.
STn.LaVIAN,
LaVIAN,E. E.
andand
PrAtt,
PrAtt,
W. The
W. New
The Poli-
New Poli- WOHLSTr.
WOHLSTr.TWr_~, A. "Nuclear Sharing: NATO TWr_~, A. "Nuclear Sharing: NATO
tics:
tics:America
Americaandand
thethe
EndEnd
of Postwar
of Postwar andandthe N + I Country,the" Foreign AffairsN
, + I Country," Foreign Affairs,
Wor/d.
Wor/d.New
NewYork:
York:
Coward-McCann,
Coward-McCann,
1960. 1960. 39 39
( 1961 ), 355-8(7. 1961 ), 355-87.
To~,
To~,R.R.W.
W.Stability
Stability
andand
the the
Nth Nth
Country
Country WoLrERs
WoLrERs,
, A. (ed.). Alliance Policy in the Cold A. (ed.). Alliance Policy in the Cold
Problem.
Problem.Washington,
Washington, D.C.:D.C.:
Institute
Institute
for for WaWar.
r. Baltimore: The Johns Baltimore:
Hopkins Uni- The Johns Hopkins Uni-
Defense
DefenseAnalyses,
Analyses, 1962.
1962. verversity
sity Pres , 1959. Press, 1959.
UHR,
UHR,L.L."Intelligence
"Intelligence
in Computers:
in Computers: The Psy-
The Psy- ZELDrrcH,
ZELDrrcH,M., JR. and EVANS, W. M. "Simu- M., JR. and EVANS, W. M. "Simu-
chology
chologyofofPerception
Perception in People
in People
and Ma-
and Ma-
chines,"
chines,"Behavioral
BehavioralScience,
Science,
5 (1960),
5 (1960),
177- 177- ysis." In H. GOETZ~OW (ed.), Simulation
in Social Science: Readings. Englewood
VAN DYXE, V. "Models." In V. ~LAN DYICE, Cliffs, N.J.: Prentice-Hall, 1962.
Polit cal Science: A Philosophical Analysia. Zorro, C. France as a Nuclear Power, RAND
Report P-2485. Santa Monica, Calif.: RAND
1960, 104-7.

Some Systemic Effects of The Spread of Nuclear Weapons Technology A Study Through Simulation of A Multi-Nuclear Future

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

Some Systemic Effects of The Spread of Nuclear Weapons Technology A Study Through Simulation of A Multi-Nuclear Future

Uploaded by

Copyright:

Available Formats

Some Systemic Effects of the Spread of Nuclear Weapons Technology: A Study through

Simulation of a Multi-Nuclear Future

for "Some Systemic Effects of the Spread of

ACKNOWLEDGMENTS ...................................................................................................................... 665

I. VAmETmS OF SLUULATXONS XN INTmaNATmNAL RELATmNS RESEARC~ . . . . . . . . . . . . 668

1.2 Types of Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 669

1.5 All-Man Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675

2.7 The Feasibility or Likeliho d of Nations Achieving an Independent Nuclear

2.8 The Approach to this Study . . . . . . . . . . . . . . . . . . . . . . . . . . . 693

3.2 The Inter-Nation Simulation: A Brief Description . . . . . . . . . . . . . . . 70

3.7 Content Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 707

4.3 Units of Analysis and the Po ling of Data . . . . . . . . . . . . . . . . . . . 714

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY

The future referred to in this study may

western Inter-Nation Simulation.

the study of these phenomena (Fig. 2.1).

CONFLICT RESOLUTION VOLUME VII NUMBER 4

Varieties of simulations in international

1.1 Models in Science

CONFLICT iRESOLUTION VOLU}~E VII NUMBER 4

1.2.a PICTOBIAL MODELS over pictorial models. It is, for example,

1.2.c MATHEMATICAL MODELS

CONFLICT RESOLUTION VOLUI~{E VII NUMBER 4

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY 671

limitations are precisely the reason why one

Work in simulation involves the utiliza-

CONFLICT RESOLUTION VOLUME VII NUMBER 4

model of decision-making behavior (so-

Benson has developed his model by

for each of the other eight large states in the

7 "Data on propensit ] to act (initial y based

CONFLICT RESOLUTION VOLUME VII NUMBER 4

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY

is derived from the potential st cur ency of any ofheld

CONFLICT RESOLUTION VOLUME VII NU~BER 4

neces ary. Thus far, the only example of

comparison data upon Which judgment can

CONFLICT RESOLUTION VOLUI~E VII NUI~IBER 4

represents the "mostly machine, very lit le

The term 'balance of power' has be n used

CONFLICT RESOLUTION VOLUME VII NUI~-{BER 4

ultilization of resources. eliminated. The regular sequence of moves is

CONFLICT RESOLUTION VOLUME VII NUMBER 4

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY

borders s(e.(ge.,gt.h,ethYaeluYRalivuerRi) pvaert)akpeaorftake of

It is Schel ing's contention that under

A good deal of trial and error wil be in-

ploratory experimentation, to manipulate the

limited objective and/or a pos ible los of the

CONFLICT RESOLUTION VOLUI~E VII NUMBER 4

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY

The initial research with the pilot model

of operating models cannot be overempha-

A science arises out of scholarly dialectic.

CONFLICT RESOLUTION V O L U l~il E VII NUMBER 4

SYSTEMIC EFFECTS OF SPREAD OF NUCLEAR WEAPONS TECHNOLOGY

tional relations is out of the question. Faced

Chapter I is a presentation of the "prob-

The "nth-country problem"

2.1 Introduction so-calso-called

CONFLICT RESOLUTION VOL Ul~ E VII NUMBER 4

tern; and (4()4t)htehfeafsiebaisl tbyiliotfynoucflenarucdliefa-r dif-

CONFLICT RESOLUTION VOLUME VII NUMBER 4