A Social Psychological Approach to Enduring Rivalries
Author(s): Cameron G. Thies
Source: Political Psychology, Vol. 22, No. 4 (Dec., 2001), pp. 693-725
Published by: International Society of Political Psychology
Stable URL: http://www.jstor.org/stable/3792483
Accessed: 09/09/2009 13:33
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=ispp.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].
International Society of Political Psychology is collaborating with JSTOR to digitize, preserve and extend
access to Political Psychology.
http://www.jstor.org
Political Psychology, Vol. 22, No. 4, 2001
A Social PsychologicalApproach
to EnduringRivalries
Cameron G. Thies
Departmentof Political Science, LouisianaState University
Therecentscholarlyworkon theconceptof enduringrivalriesoffersa promisingwayto
examinestrategicinteraction
periodsof time.Afocus
amongdyadsof statesoverextended
on rivalry,and on the mechanisms
thatprovidefor suchinteraction,mayoffera wayto
bridgeexistingtheoriesof internationalrelationsthat rely exclusivelyon structureor
thepotentialfortheory-building
hasnotbeenfullyrealizedbecause
process.Unfortunately,
researchintorivalryhastendedto be inductive.Thispaperseeksto rectifythatproblemby
situatingthe rivalryconceptwithina social psychologicalapproachto international
relations.Therivalryconceptis appropriately
locatedin a theoreticalapproachthatviews
theinternational
systemas a socialsystemwhereactorsare conditioned
bymechanisms
of
andsocialization.
competition
KEY WORDS: enduringrivalry;socialization;dissonance;social proof; mechanisms;international
system.
The recentscholarlyworksurroundingthe concept of enduringrivalriesoffers
a promising way to examine strategic interaction among dyads of states over
extendedperiodsof time. The notionof an enduringrivalryrequiresus to examine
interstatebehavioras a time-dependentprocess, ratherthanas a series of discrete
events. A focus on rivalry,andon the mechanismsthatprovidefor suchinteraction,
may offer a way to unite existing theories of internationalrelations that rely
exclusively on structureor process. The rivalryconcept thus offers considerable
potentialfor theory-buildingin internationalrelations.
Unfortunately,that potential has not been fully realized because of the way
that this literaturehas developed. The majorhindranceto theory developmentin
the rivalryliteraturehas been the inductive natureof its birth.Enduringrivalries
were first noticed as empiricalphenomenaassociated with the Correlatesof War
(COW) Militarized InterstateDispute (MID) data set. Goertz and Diehl (1992)
launchedthis literaturewhen they emphasizedthatcertainstate dyads were more
693
0162-895X ( 2001 InternationalSociety of Political Psychology
Publishedby Blackwell Publishers,350 Main Street, Maiden, MA 02148, USA, and 108 Cowley Road, Oxford, OX4 IJF, UK.
694
Thies
war prone than others. Indeed, such rivalrous dyads appearto have engendered
more than half of all interstate wars. Goertz and Diehl were correct to note that
this phenomenon deserved greaterscholarly attention. However, until recently,
most of this researchinvolved refinements in the data-drivenoperationalization
of rivalries, with scant attention paid to theoretical explanations for this
phenomenon.
This papersituatesthe rivalryconcept within a social psychological approach
to internationalrelations.I arguethatthe currentconceptualfocus on competition
among states is but one structuralfeature conditioning the emergence, maintenance, and terminationof rivalries. Socialization is another,equally important,
structuralfeature of the internationalsystem that adds a vertical dimension of
politics to the conventionalhorizontaldimension accountedfor by competition.
Incorporatingsocializationinto the rivalryliteratureoffers new hypothesesto test
againstdyads thathad been inductivelyidentifiedas rivalries.The rivalryconcept
is appropriatelylocated in a theoretical approach that views the international
system as a social system where actors are conditionedby both competitionand
socialization.
The Conceptualization and Operationalization of Enduring Rivalries
Goertz and Diehl (1992, p. 153; 1993, pp. 154-155) suggested that enduring
rivalries have three conceptual components: competitiveness, time, and spatial
consistency.Competitionover an intangiblegood (e.g., prestige,influence)or over
a tangible good (e.g., territory,resources) underlies the conflict in a rivalry. A
rivalry cannot be "enduring"unless it persists for an extended period of time.
Finally, rivalries must include a consistent set of states in their spatial domain,
normally defined as dyadic in nature,althoughGoertz and Diehl (1993, p. 155;
1997) and Diehl (2000, pp. 241-262) suggested that rivalries might involve
linkages amongmultiplestates.A prominentexample of a multi-staterivalryis the
triangleformedby the United States, U.S.S.R., and China(Goldstein& Freeman,
1990).
Goertz and Diehl (1992, pp. 155-156) operationalized enduring rivalries
according to the following components: a minimum number of militarized
interstate disputes or MIDs (Gochman & Maoz, 1984), a minimum time frame
for dispute occurrence, and a rule that determines termination of the rivalry
after a minimum passage of time in the absence of a dispute. Enduringrivalries
are conflicts between the same two states that involve at least five MIDs within
20 years. The rivalry terminates if 10 years pass without a MID between the
two states. This operationalizationproduced a list of 59 enduringrivalries. The
authorsalso adopted a thresholdcriterionto indicate isolated conflict (one MID
with no dispute in the following 10 years) and proto-rivalries(two to four MIDs
with no more than 10 years between disputes). Their empirical results indicate
that the frequency of MIDs is twice as great in enduring rivalries as it is in
695
Enduring Rivalries
isolation.1Additionally,morethanhalf of all interstatewarstakeplace in enduring
rivalries.Finally, a dispute in an enduringrivalryis almost twice as likely to lead
to waras one in isolation.These results,despitethe lack of theoreticalspecification,
suggest thatrivalriesare indeed an importantsubjectof scholarlyattention.
Unfortunately,because of the lack of a priori theoreticalspecificationof the
rivalry concept, much of the literaturehas yielded different lists of rivalries
generatedby differentoperationaldefinitions.2This battle of operationaldefinitions has led to the tendencyfor rivalrylists to be used as case selection devices to
study a numberof other phenomenain the conflict literature(e.g., Geller, 1993;
Huth, Gelpi, & Bennett, 1993; Huth & Russett, 1993; Sorokin, 1994). The result
is thatlittle theoreticalwork has been done on the rivalryconcept itself. As Diehl
(1998) admitted,"Because little work has been done on the beginningand end of
enduringrivalries,thereis a notablegap on whathappensin rivalriesbetweenthese
endpoints"(p. 15).
This statementsuggests three general aspects of rivalries that remain unexplained by the inductive approach.First, how exactly are these series of conflicts
linked together?What intangibleor tangiblegoods providethe basis for rivalrous
interaction?Could something other than competition account for the linkage
between conflicts? The inductive approachto rivalryis able to incorporatethese
linkages into its definitiononly by assumption.Second, how are rivalriesinitiated
and terminated?The inductive approachcan only make such determinationsex
post facto. The inductive approachadopts a wait-and-see approachto determine
whethera rivalryhas commencedor expired.Third,why does the level of intensity
varyduringthe courseof a rivalry's life cycle? Accordingto the inductiveapproach
with its adoptionof the MID framework,the occurrenceof varyinglevels of conflict
(includingwar) throughouta rivalryis again assumed,not explained.
Toward Theory: Competition
The first problem,that of accountingfor the linkage in rivalrousinteraction,
has received some attentionin the rivalry literature.Vasquez (1996) defined a
rivalry as "a relationshipcharacterizedby extreme competition,and usually psychological hostility, in which the issue positions of contenders are governed
primarilyby their attitudetoward each other ratherthan by the stakes at hand"
(p. 532). Rivalriesarethusmarkedby the adoptionof an actor dimensionto issues,
ratherthan a stake dimension. An actor dimension is adopted when issues are
defined in terms of what they mean for the other actor, as opposed to a stake
dimension, which defines issues in terms of their intrinsic value (Mansbach&
1 These results are somewhat temperedby the fact that dispute frequencyis also a componentof the
definitionof an enduringrivalry.
2 See Goertz and Diehl (1993) for a
comparisonof a numberof operationaldefinitionsof rivalry.Also
see Bennett (1996, 1997, 1998) and Thompson(1995, 1998) for alternativeoperationaldefinitions.
696
Thies
Vasquez, 1981, p. 60). Vasquez's (1996) definitionthusencompassesthreedimensions:issues, competition(overthe issues), andpsychologicalhostility.His general
argumentis that one of the main factors leading to war among equals is the
occurrenceof territorialdisputes,especially where contiguousterritoryis at issue.
In his analysis, territorialcontiguity is used as an indirect measure of territorial
disputes,which is itself the operationalizationof a "contentiousissue."Essentially,
what promised to be a conception of rivalry that incorporatedcompetition over
issues and psychological hostility is reducedto territorialcontiguity.
Bennett (1996, 1997) similarly incorporatedissues into his definition of
rivalry.Bennett (1996) defined an interstaterivalryas
a dyad in which two states disagree over the resolutionof some issue(s)
between them for an extended period of time, leading them to commit
substantialresources(military,economic, or diplomatic)towardopposing
each other, and in which relatively frequentdiplomaticor militarychallenges to the disputed status quo are made by one or both of the states.
(p. 160)
The issues at stake in rivalries could include territory,externalpolitical policies
(e.g., promotion of a religion or ideology), or internal political policies with
internationaleffects (e.g., the treatmentof ethnic minorities). Bennett's (1996)
operationalizationof interstaterivalriesrequiresthatthe issues at stakein the dyad
must be connected over the life of the rivalry.The first MID between two states
marksthe onset of the rivalry,the issue is identifiedby the demandor claim made
by one state against the other, and the rivalrycontinuesuntil the issue(s) at stake
are settled. In 25 of the 34 interstaterivalrieshis operationalizationproduces,both
states were concerned with the issue of control over contiguous or homeland
territory.Bennett(1997, p. 231) furtherarguedthatterritorialissues are the key to
understandinginternationalconflict generally, and because conflict is the key to
identifyingrivalries,we shouldexpect territorialissues to be crucialto identifying
interstaterivalries.
The focus on competition over issues in the work of Vasquez (1996) and
Bennett (1996, 1997) is a good startat conceptualizingrivalries.But as we have
seen, most of the "issues"collapse into the issue of territory,or more specifically
territorialcontiguity. It would seem that the study of rivalrieshas a built-in bias
toward territorialcontiguity, as it assumes a central position in these and other
scholars' researchon rivalry(e.g., Diehl, 1985; Hensel, 1994, 1996; Huth, 1996).
This may be an artifactof the MID data:At least half of the 14 types of military
acts that are drawnupon to code a MID include an explicit referenceto the threat
or actualuse of force on or againstthe territoryof anotherstate(Gochman& Maoz,
1984, pp. 587-589).
Thompson (1995) clearly distinguishedrivalries that concern territoryfrom
those that concern position in the internationalsystem. This distinctionbetween
spatialand positional rivalriesis useful for several reasons.First, it highlightsthe
Enduring Rivalries
697
differentmotives and goals behindrivalrousinteraction.Second, it helps sort out
the relationshipsof minor and majorpowers in rivalries.Third,it helps us understandthe differentroles playedby competitionandsocializationin the international
system. According to Thompson (1995, p. 204), spatial rivalriesusually involve
minorpowers thatlack the resourcesand/orthe independencefrom majorpowers
to fully resolve their territorialdisputes. Such disputes tend to flare up for brief
periods of time, as a resultof resourceconstraints,and then die down until one of
the rivals has the resourcesto resolve the issue to its favor. Spatialrivalriestend to
be morecommonandless deadlythanpositionalrivalries.They aregenerallyfound
where thereis power symmetrybetween the two states.They also should occur in
the states' immediategeographicarea (Diehl, 1985, pp. 1204-1205).
Positionalrivalriesalso requiresome roughsymmetryof capability,especially
at the major power level. Positional rivalries involve conflicts about relative
position at or nearthe apex of the internationalsystem.3These rivalriestend to be
less frequentand much more deadly than spatialrivalries,as states near the apex
of the system generallydo not preferto sharethe lead position (Thompson,1995,
p. 205). Vasquez (1996, p. 553) found empiricalsupportfor Thompson'sdistinction between spatialandpositionalrivalries.Because Thompsonhypothesizedthat
spatial rivalries will be the primary domain of minor powers, it is likely that
previousresearchon the "issue"of territorywill need to be refined. If Thompson
is correctin his categorizationof some rivalriesas positional, territorymay often
be the proximatecause of conflict between states, but is not necessarily what is
being foughtfor in a largersense. But if minorpowersarecompetingover territory,
and major powers are competing over position, then what explains rivalries
betweenmajorand minorpowers?
Vasquez (1996, p. 533) arguedthatrelativeequalityis a prerequisiteof rivalry,
because the behavior of major power-minor power dyads is fundamentally
differentfrom that of majorpower-major power or minorpower-minor power
dyads. Vasquez pointedto questionabledyads in Goertzand Diehl's (1993) list of
rivalries,such as U.S.-Haiti andIndia-Nepal. Thompson(1995) also questioned
these and 14 additionalrivalriesfrom GoertzandDiehl (1993) as dubiousin terms
of face validity. Goertz and Diehl (1993) and Diehl (2000, p. 25) suggested that
the inclusion of majorpower-minor power dyads as rivals is simply an empirical
question.However, it is also an importanttheoreticalquestionbecause this behavior must be explained.
In the searchfor a conceptualizationof rivalrythatseeks to avoid the inductive
trap of dispute frequency, I concur with Thompson (1995) in recognizing that
rivalriesrequirea certainlevel of identificationand recognitionby both members
of a rivaldyadto distinguishthemfrommerecompetition.This type of recognition
takes us beyond a mechanisticsystem of impersonalcompetitiontowarda social
3
Thompson (1995, p. 203) stated that positional rivalries can be dyadic, regional, global, or regional-global in nature.
698
Thies
system wherethe identitiesof actorsareimportant.Such an understandingis found
in Keunne (1989):
Rivalry is distinguished from competition among agents by its nonanonymity. In rivalrous interaction,each agent knows that his or her
autonomous actions will affect the welfare of every other rival nonnegligibly and that those rivals will react self-protectively.... In competitive environments,the externalwelfare consequencesof actors'decisions are dispersed so widely over other interested parties that their
initiatives cannot be identified and reactionsneed not be contemplated.
(p. 555)
The non-anonymity condition moves the rivalry concept beyond Goertz and
Diehl's (1992, 1993) inductive approachthat emphasizes perfect competition.
Indeed,theirapproachmust emphasizeperfectcompetition,as all possible combinationsof statesareconsideredpotentialrival dyads.The non-negligiblecondition
requiresthat states in a dyad are awarethattheirbehaviorwill affect each other.
Non-anonymous, non-negligible competition is analogous to oligopolistic
competitionamong firms in a market.Oligopolistic competitioninvolves a small
numberof firms dominatinga market.Such firms are aware of their competitors
and the impact that their decisions have upon them. This is quite differentfrom
perfectcompetition,in which firms' decisions affect othersin the abstract,because
no single firm has enough marketpower to affect the livelihood of any single
competitor. The great powers are clearly states with "marketpower" in the
internationalsystem. The type of oligopolistic competitiondescribedby Keunne
is characteristicof great-powerrivalries.However, oligopolistic competitionmay
also be foundin regionalsubsystems.Neighborswithina particularregionarevery
aware of who their competitors are, and how their decisions affect them. If
neighborshave enough capabilitiesto affect each other,then competitionmay also
explain rivalriesbetween minorpowers within particularregions.
Competitioncannot explain rivalries that are unbalancedin terms of power
capabilities. Thompson's and Vasquez's concerns about major power-minor
power rivalriesarevalid. The long-lived conflictualinteractionbetweenthe United
States and Haiti, or that between India and Nepal, is not properlyexplained by
competition. Major differences in capabilities among states introduce a status
dimensionto internationalpolitics. It may be competitionthat spursthe development of those capabilities,but once a statusis attained,its maintenanceentersthe
realm of socialization. Waltz's (1979) structuraltheory of internationalpolitics
views competitionas one of the two mechanismsthroughwhich structureworks
its effects on units. For Waltz, competitionis a situation within which units find
themselves. The structuralcondition of anarchymeans that no higher authority
exists to regulate the actions of states other than the states themselves. Thus, all
units conditioned by anarchy must potentially interact with others for power,
prestige, and security. The situation of competition pushes all units to adopt a
Enduring Rivalries
699
similar form and internalorganization-whatever appearsto be most efficient at
securingpower,prestige,and security.Competitionas a situationdoes not say how
the unitswill go aboutsecuringthese thingsin the system. How they go aboutthese
tasks will largely be the result of socialization.
I expect that two mechanisms can account for competition among statesorganizationalcompetencyandrationalimitation.4Waltz is clear thatcompetition
shouldencouragethe samenesseffect amongthe units.Stinchcombe(1998) argued
thatstateshave developed certainorganizationalcompetenciesthatenablethem to
appropriatebenefits from certainlegitimateactivities, such as trade,and to be free
of liabilityfor whateverdamageis causedin the legitimatepursuitof those benefits.
The flow of benefitsthatthe statemonopolizes is used to maintainits statusor rank
in the system. Such rankor status is precariousbecause one's own appropriation
may be inhibitedby the appropriationactivities of others who are also not liable
for damages. This argumentis quite similar to that of Spruyt (1994), who also
arguedthatthe sovereignterritorialstatewas betterable at organizingitself to take
advantageof trade and commerce than its competitors,including city-states and
city-leagues. Thus, organizationaldevelopmentsthat might reap competitive advantage are quickly adoptedby other states in orderto maintaintheir rankin the
system.
Samenessin function,or functionalundifferentiation,refersto the fact thatall
statesface similartasks, includingraisingrevenue,establishinginternalorder,and
defending themselves from other states. Hedstrom(1998) proposeda mechanism
of rationalimitationthatcan accountfor this type of sameness.Rationalimitation
does not arise from any notion of the need to adhereto social conformity.Rather,
imitationis seen as a useful strategyfor arrivingat betterdecisions in the pursuit
of resources or position. Imitationis also seen as a strategyfor organizationsto
achieve legitimacy. When organizationsimitate already existing and accepted
models, they reduce the risk of being called into question by individuals and
institutionalactorsand therebyincreasetheirchance of survival.This mechanism
of rationalimitationseems to be whatWaltzbelieved to be operatingin the system.
But who will they imitate?It seems reasonablethat states with similar statuswill
imitateeach other-some avenuesof statebehavior,suchas theprojectionof global
force, are simply unavailableto states without similarcapabilities.
Rivalriesconditionedby competitionare likely to occur among greatpowers,
or between great powers and majorpowers. These two groups of states are more
likely to have the organizationalcompetencyto allow them to make the fullest use
of theircapabilities.This kind of competitionis exactly what the power transition
literaturedescribes(Organski& Kugler, 1980). To the extentthatgreatpowersand
majorpowersimitate,they will imitateeach other.Competitionamongsmall states
and emerging states is likely to take the form of rationalimitation.These states
simply do not have the capabilities (or the organizationalcompetency to exploit
4 For more on mechanisms,see Hedstromand Swedberg(1998).
700
Thies
the capabilitiesthey possess) to compete with majoror great powers;rather,they
will imitate the internalorganizationand externalform of their peers. In general,
rivalrous competition should occur between states with similarly matched
capabilities.
Socialization and Enduring Rivalries
Waltz (1979) consideredsocializationthe otherprimarymechanismthrough
which structureconditions the units of the system. Again, he was ratherelusive
abouthow he expected socializationto operate,but it is clear thathe expected it to
also producethe "sameness"effect. Undoubtedly,Waltz expected that units will
conformto the dictatesof the internationalsystem, or they will fail to survive.The
inclusion of socialization into Waltz's otherwise static theory of international
politics offers a way to think aboutboth structureand process in the international
system. Competition is an environment or situation within which actors find
themselves. Socialization,on the otherhand, is an ongoing process. Rivalriesare
ongoing dyadic processes that occur within a competitiveenvironment.Rivalries
are thus an ideal way to bringout the latentdynamicfeaturesof neorealismand to
bridgethe theoreticalgap between structureand process.5
Previousdiscussions of socializationprovidelittle guidancefor incorporating
this conceptinto the studyof internationalrelations.IkenberryandKupchan(1990)
discussed socialization within the confines of a hegemonic system, whereas my
intenthere is to demonstrateits importanceto the anarchicsystem thatis generally
assumed by rivalry researchersto form the backdropof internationalrelations.
Schimmelfennig (2000) offered an account of socialization that emphasizes the
self-interestedactionsof post-communiststatesoperatingwithinthe context of the
highly institutionalizedenvironmentof the EuropeanUnion. Again, however,most
states are generally not operating in such an environment.Both Ikenberryand
Kupchan(1990) and Schimmelfennig(2000) have run into the problemof incorporatingwhat is normallyconsidereda social psychological concept into international relationstheory driven primarilyby economic analogy. Their solution was
to select cases in which the environmentis highly institutionalized,such as under
hegemony,or in an exceptionalcase of regionalinstitutionalization.Neitherchoice
is particularlyhelpful to the task at hand-incorporating socialization into the
generallyacceptedanarchicview of the internationalsystem.
Resende-Santos's(1996) attemptto incorporatesocializationinto an explanation of the emulationof militarysystemsobscuredthe conceptof socializationeven
further.He arguedthat"emulationis more directlya productof socialization"than
5 Of course, by drawingout the logical implicationsof socialization,neorealismas understood its
by
practitionersmay be undermined.Socializationoffers a way to think aboutreplicationand transformation of the internationalsystem, or, as Wendt (1992) so succinctly put it, "anarchyis what states
make of it."
Enduring Rivalries
701
it is of competition (p. 208), even though he discussed both emulation and
innovationas featuresof competition.Waltz (1979, p. 127) gave an exampleof the
emulationof the Prussianmilitarystaff system to describethe effects of competition, thusdirectlycontradictingResende-Santos'sanalysis,which tendsto conflate
competitionand socialization.Theoretically,we shouldexpect thatemulationand
innovationare featuresof both socializationand competitionas expressedthrough
the operationof differentmechanisms.Wendt(1999) also arguedthatsocialization,
in addition to competition, should play an importantpart in the explanationof
behaviorin the internationalsocial system, but he failed to producea framework
that could offer general predictionsor explanationsof the operationof either one
of these structuralfeatures.The frameworkthat I propose for understandingthe
impact of these mechanismsshould illuminatenot only the study of rivalries,but
the study of internationalrelationsmore generally.
What then is meant by socialization?According to Wentworth(1980), "socializationis the activitythatconfrontsandlends structureto the entryof nonmembers into an alreadyexisting world or a sector of that world"(p. 85). Socializing
activityestablishesa tensionbetweenthe "member"andthe "novice."This tension
has several implications.First, there are relative differences in power, status,and
prestige between the member and the novice. Second, the novice is within the
sphere of influence of the member.Third,a variety of "others"may intervenein
the socialization of the novice. Fourth,there is a varying degree of asymmetry
between the member'sandthe novice's view of reality.Fifth,an ongoing historical
institution,consistingof structureandroles, precedesa new generationof potential
members.
The tensionbetweenmembersandnovices in the internationalsystemrequires
additionalexplanation.Who are the members?The membersare those states that
are establishedin theirroles in a given system. Who are the novices? The novices
are statesthatemerge duringa periodin which the system is stable.Whathappens
when the system changes? In a new system, all states are novices in a sense.
However, because of the peculiarities of the international system, some
states-namely great powers-are always "members."Greatpowers, because of
their greater capabilities, retain their status and primaryrole(s) from system to
system. The great powers will be the dominant socializers in any international
system. However, small states should be considered "novices"when the system
changes,just as emergingstatesare also considered"novices"at whateverpoint in
historythey enterthe system.
The relative differencesin power, status,and prestigebetween membersand
novices supportsmy contentionthatsocializationmay provideexplanatorypower
for rivalriesconsisting of states with differentialcapabilities.Greatpowers have
traditionallyheld sway over stateswithintheirspheresof interest.The greatpower
that dominatesa particularsphere of influence will socialize the novice states of
that region; if a great power no longer holds sway, then a regional power will
assume the role of membervis-a-vis the novice. The regionalpower itself may be
702
Thies
a novice vis-a-vis a great-powermemberof the internationalsystem. Thus,a novice
may be subject to socialization pressurefrom both a great power and a regional
power. However, "others"such as local peer states may also intervene in the
socializationprocess.
The asymmetrybetween a member'sand a novice's view of realityis crucial
to the socializationprocess.The memberlargelystructuresthe realitywithinwhich
the novice must operate.The memberdeterminesthe roles, norms,and principles
that are appropriateguides for behavioralinteractionin the system. The socialization processoperateson behalfof the member,with the purposeof bringingnovices
in line with the member'snormativeexpectations.However, entryinto the system
is a process of mutual accommodationand negotiation.Disagreementsbetween
novices and membersover appropriateroles and norms may persist, thus giving
rise to rivalries.Further,novice states,andparticularlyemergingnovice states,are
in the processof constructingtheiridentity.Identityformationis partiallythe result
of internalprocesses that may reject outside impositionof roles and norms,but is
also conditionedby interactionwith the "other,"accordingto social identitytheory
and identitytheory.6Thus, socializationis crucialto the constructionof identityin
the internationalsystem.
Finally, the ongoing historical institution that we discuss in the context of
international relations is the existence of an international system. The deep
structure(anarchy) that forms the base of all systems is regardedas constant by
Waltz (1979). The surface-level component of structure-the distribution of
capabilities-changes to give rise to a new system. Certain roles are also part
of the ongoing historicalinstitution.Waltz (1979) acknowledgedthe special role
that great powers play in the internationalsystem. This should not preclude the
existence of other roles in the system. Members (largely great powers) must
socialize emerging states and resocialize existing states to the new reality. Even
Waltz (1979, pp. 194-210) emphasized the system maintenancefunctions performedby the greatpowers. Thus, the main elements of continuityin the international system are the deep structureand the socializing role of the greatpowers.
During the socialization process the novice undergoes assimilation, which
makes the novice more similar to the members.However, the novice also exerts
influence on the members, thus necessitating accommodation on the part of
members (Moreland, 1985, p. 1174). This process fits well with the neorealist
conceptionof the internationalsystem as an environmentof mutualadaptationand
adjustment.The speed with which novices are socialized depends on the level of
commitmentof the novice to the membergroupandparticularrelationshipswithin
the group,the extent of the differencesbetween the novice and members,and the
numberof novices thatjoin the membergroup at any one time (Moreland,1985,
p. 1174). These factorsmay explain the genesis of enduringrivalriesin the Middle
6
See Mercer (1995) for an applicationof social identity theory to internationalrelations,and Wendt
(1999) for an applicationof identitytheoryto internationalrelations.
EnduringRivalries
703
Eastern subsystem. Israel, Jordan,Lebanon, and Syria all joined this subsystem
within a few years of each other-a large numberof emerging novices that all
formed enduringrivalries in the region. Israel, for obvious reasons, is also quite
different from the rest of the Middle East and only recently has expressed any
commitment to other members in the region. Thus, it has been the target of
numeroussocializationattemptsthatchallenge its very existence.
Stryker and Statham (1985), Sarbin and Allen (1968), and Biddle (1986)
discussed severalprocesses identifiedin the literatureas responsiblefor socialization: direct instruction,imitationor modeling, and altercasting.7However, these
processes can be collapsed into two general socialization processes: those that
involve the direct interalization of communicatednormativeexpectations, and
those that involve the indirectassimilationof normsthrougha process of identification with socialization agents who exemplify the norms. Socialization should
occurindirectlythroughimitationor modelinganddirectlythroughinstructionand
altercasting.
Socialization mechanisms should primarilyaffect the external behavior of
states, with some requisite adjustmentsin internalorganization.I postulate two
mechanismsfor the operationof socializationin the internationalsystem:the social
proof heuristicand dissonancereduction.
The Social Proof Heuristic
The social proofheuristic(Cialdini,1993;Hedstrom,1998) is the socialization
mechanismmost likely to reinforce existing norms and roles in the system. The
social proof heuristic is essentially this: When you are not sure what to do, look
aroundto the actions of others for possible clues as to what your own behavior
should be (Hedstrom, 1998, p. 314). As Cialdini (1993) suggested, "we view a
behavior as more correct in a given situation to the degree that we see others
performingit"(p. 116). This mechanismis essentiallyone of imitationormodeling,
and is a pervasive featureof social as well as political life.
Cialdini (1993) argued that the principle of social proof is more frequently
observedundercertainconditions.First,the operationof the social proof heuristic
is generallyobservedunderconditionsof uncertainty,when the situationis unclear
or ambiguous (p. 129). However, the use of social proof can be problematicin
uncertainsituationswhen everyone is looking to see what everyone else is doing,
but no one in the situationhas a clear understandingof what is actuallygoing on.
This can lead to the phenomenon of "pluralisticignorance," which has been
invoked to explain why groups of bystandersfail to aid victims in their time of
need. Pluralisticignorancehas gained notorietyin cases such as the 1964 murder
of Catherine"Kitty"Genovese in Queens, New York, because it seems to operate
7
Altercastingrefersto a memberstate casting a novice in a role, which entails specific normativeand
behavioralexpectations.
704
Thies
best amongstrangers(p. 129). However,pluralisticignoranceneed not alwayshave
negative consequences.In many cases, social proof is a heuristicthatworksto the
benefit of both the individualand the group (p. 157).
A second conditionthatpromotesthe use of the social proof heuristicis when
we are observingindividualswho are similarto ourselves (Cialdini, 1993, p. 140).
Similarity heightens the confidence that we attach to the appropriatenessof
behavior,especially given an uncertainsituation.Finally,the social proofheuristic
is most probablyobservedin relativelycontainedgeographicareas.This aspect of
the social proof heuristicis generallynot emphasizedbut is nonethelessimportant.
Copycat suicides and murders,ratherunfortunateexamples of social proof, occur
only within those geographicalareaswhere newspaperor television stories about
them were publishedor broadcast(p. 147). It makes sense thatimitationaccording
to social proofwill have some geographicboundaries,otherwisewe would observe
behaviorsandattitudesspreadingquicklyanduniformlyacrosscountriesandeven
throughoutthe globe. In some cases this may be observed, but most situationsin
which individuals resort to use of the social proof heuristic are probablyfairly
localized.
These threefactors-uncertainty, similarity,and geographicconcentrationcombineto provideadditionalunderstandingof situationsin which the social proof
heuristicoperatesto guide statebehavior.The social proofheuristicshouldbe most
evident in emergingstatesas they seek to understandthe role of the sovereignstate
itself. An emerging state finds itself in an uncertainsituationas a new memberof
a system. Any confidenceaboutappropriatebehavioror actionwill probablycome
from imitation of similar states (i.e., states with similar capabilities) in its geographicalregion. Other possible occasions for the operationof this mechanism
include learningsymmetricalroles such as thatof an ally, or even thatof an enemy.
In general, the social proof mechanism should operate when there is little
question about the appropriatenessof the state's own role conception. The main
questionthatthe social proof mechanismhelps to answeris how to properlyenact
a role in conformitywith others' expectations(Biddle, 1986, p. 78). To the extent
that the social proof heuristicpredominates,we should see a fairly stable system
in which the units engage in its reproduction.The rangeof acceptablebehaviorsin
the system should be fairly constrained.However, work by Schelling (1978) on
"tippingpoints"suggests thatif a rogue statewere able to engage in novel behavior
with impunity,andotherstatesadoptedthatbehaviorvia the social proofheuristic,
then, aftera certainnumberof statesadoptedthe behavior,the entiresystem would
accept thatbehavioras normal.
However, innovationsin roles, norms, or their behavioralmanifestationsare
usually met with skepticismby the relevantothersin any social system (Stryker&
Statham,1985, p. 353). Deviance fromexpectationsis permissiblein the short-run,
as actorsengage in "aligningactions"to bringtheirbehaviorin line with standards,
but over the long-runsuch behaviorwould be punished(Stokes & Hewitt, 1976).
The only exception to this rule is that social deviance could persist in a situation
Enduring Rivalries
705
of structuralfailure (Stryker& Statham, 1985, p. 365). In the case of the state
system, structuralfailure would characterizecertain regions where interaction
capacity is low, such that it is difficult to even think of a system whose members
could constrainagentbehavior.The social proofheuristiccould transmitinnovative
roles or normsfrom stateto stateby diffusionin such a low-interactionsystemonly
if states were actively seeking models for their own behavior and this could
eventuallylead to theiracceptanceas standard.We areunlikelyto observerivalries
based on competitionor socializationin areas of structuralfailure. For example,
there are only three examples of African rivalries in a region populated by
"quasi-states"(Jackson, 1990).
We should expect to see rivalriesformingin partbecause of the social proof
heuristicunderseveral conditions.First,as mentionedabove, the speed of socialization is affected by the level of commitmentof the novice to the membergroup
and particularrelationshipswithin the group.If the level of commitmentbetween
the novice and membersis low, then the socializationprocess may be lengthy and
potentiallyconflictual.The level of commitmentrelatesdirectlyto the uncertainty
of the situation faced by an emerging state entering an established system. If
memberstates act in a conflictualmanner,then novices may adoptthe same types
of behaviorin return.Second, the extent of differencesbetween the novice andthe
members also affects the speed of socialization.If similaritybetween the novice
and membersis low, then an emerging state is unlikely to look towardmembers
for guidance regardingthe norms of interactionin the system, thus socialization
will proceedslowly andwill likely proveconflictual.Finally,the numberof novices
thatjoin the membergroup at any one time affects the pace of socialization.The
more novices enteringthe system in a particulargeographicalregion, the greater
the uncertaintyabout interactionin the system, and the less ability that member
stateshave to controlthe behaviorof any particularnovice. Again, the resultcould
be a slow and painful socialization process for the novice state. Any of these
conditions could set the stage for the formationof an enduringrivalrybetween a
novice and a member.On the other hand, if an emerging state enters the system
underthe protectionof a member,or is perceivedto be similarto the members,or
emerges by itself, it would seem unlikely that an enduringrivalryconditionedby
socialization would form, because the pace of socialization should proceed relatively quickly underthese conditions.
Dissonance Reduction
The second socializationmechanismis dissonancereduction(Festinger,1957)
as discussed by Jervis (1976, chapter 11), Elster (1998), and Kuran (1998).
Accordingto dissonancetheory,inconsistencyamongcognitionscauses a motivational state called dissonance (or "cognitive strain"in Sarbin & Allen, 1968,
p. 541). Dissonance leads to an aversive state of arousal.The aversive state of
arousalleads to attemptsto reducethe arousal/discomfortandachieve consonance.
706
Thies
This often leads to attemptsto avoid informationor situationsthat would lead to
increasedlevels of dissonance. Jervis (1976) discussed a numberof implications
for internationalrelationsderivedfrom dissonancetheory.
First,by attemptingto reducedissonance,policymakersaltertheirbeliefs and
perceptions, thus altering the premises of later deliberationsthat affect future
perceptionsand decisions (Jervis, 1976, p. 382). In combinationwith the normally
high costs of policy change in domestic society, and the reputationalcosts in
internationalsociety for breaking commitments, initial decisions may become
"lockedin," with futuredissonantinformationhaving little negative impacton the
continuanceof the original policy. Second, we should also observe a "spreading
apartof the alternatives,"in which thereis an increasingperceptionthatthe policy
chosen was clearly better than the rejected alternatives(p. 388). However, some
decisions may actuallydestroy the availabilityof alternativepolicies by radically
altering the state's environment, such as the decision to go to war (p. 389).
Dissonantinformationaboutpolicy choices may be irrelevantonce such a course
of action is taken.
Third,expendingresourcesincreasesdissonanceand the correspondingpressure to believe that the policy is succeeding (p. 393). Therefore,states that suffer
high casualties in wars are likely to believe that their sacrifice is noble and
worthwhile,despite any informationto the contrary.Further,the sacrificeentailed
in warmay lead to a perceptionthat"winning"the fight has become morevaluable,
thatthe probabilityof success is higher,andthatwhatmightbe consideredfailures
by disinterested observers are actually viewed as successes by the leaders
(pp. 394-396). Finally, and most troublingto Jervis(p. 399), is the "psychologyof
insufficient reward,"in which there exists an inverse relationshipbetween the
incentives given for an alternativecourse of behaviorand attitudechange. For our
purposes,this is importantbecause it suggests thatthe more otherstatesattemptto
negatively sanction a given state's behavior, the less likely its leadershipis to
change theirminds and adoptalternativebehaviors.
Dissonancereductionmay also be thoughtof as a social processratherthanas
a purely cognitive mechanism (Kuran, 1998). Agreement upon the operationof
certainroles and normsis essential for any social system to remainstable.If a state
challenges the roles assignedto it in a stable system, the resultmay be a balancing
of other states against it to prevent the enactment of the nonsanctionedrole.
Balancing can thus be conceived of as a dissonance reductionmechanismin an
interstatesocial system. States that are content with the currentdistributionof
great-powerroles are expected to thwartattemptsby minorpowers to breakinto
the ranksof the greatpowers, and preventa currentgreatpower from establishing
a hegemonic role for itself. Any moves in these directionsto destabilize a system
will produce "dissonance"and attempts at dissonance reductionthrough direct
instructionor altercasting.Thus, balancingand positional rivalriesare not purely
a function of competition, as neorealists argue, but are also conditioned by
socialization.
EnduringRivalries
707
Most rivalries conditioned by socialization will probably exhibit characteristics of dissonance reduction.Obviously, novice states or other membersthat
enact roles or normsthat are inconsistentwith the expectationsof othermembers
will receive cues from those membersto altertheirbehavior.These cues may take
a benign form of socialization,such as diplomaticprotestsor condemnationin the
United Nations. However, the offending statemay choose to ignorethese cues and
enact its role or norm regardlessof memberopinion. This may lead to increased
negative sanctions(e.g., economic sanctionsor militaryaction) againstthe stateto
attemptto socialize it into an appropriaterole or norm. If the offending state is
resolute and/orhas the capabilitiesto resist socializationefforts, then this process
could endurefor an extended period of time, hence forming an enduringrivalry.
In fact, dissonance theory strongly suggests according to the "psychology of
insufficient reward"that rogue states will be unlikely to change their behavior,
particularlyas the negative sanctions increase in strength.Policy inertia and the
"spreadingapartof the alternatives"also decrease the chances that the offending
behavior will be changed. The choice to go to war may even alter the state's
environmentin such a way thatno alternativepolicy choices remainavailable.The
result for many socialization attemptsin the internationalsystem is that rivalries
become "locked in" early on, and dissonantinformationis ignored,or alternative
courses of action are eliminated,thus producingan ongoing series of militarized
disputesbetween socializer and socializee.8
There are two variantsof enduringrivalriesthatcould stem from a socialization perspectiveemphasizingdissonance.First,rivalriesconditionedby socialization may be situationsin which efforts to socialize a state to a particularrole or
norm are incomplete, or fail altogetheracross time.9Even in dyads in which the
capabilityimbalanceis enormous,it may be difficultto reducedissonancebetween
actual and expected behavior by enforcing conformity to norms or roles, as
explained above. Waltz (1979) suggested that states that fail to adapt to the
pressuresof competitionand socializationwill be selected out of the system. Does
this mean that states are eliminated?Judgingfrom the historicalrecord,states are
not frequentlyeliminated."Selectionout"of the system may take on a more social
tone as states arejudged "pariahs"or "rogues"by other membersof the system.
Therefore,the list of enduringrivalriesis quite likely to containmost of the states
that have been consideredrogues or revisionists throughouthistory.The ultimate
failure of socializationattemptsoccurs when a rivalryescalates to war, especially
if thatrivalrycontinuesbeyond the war to a series of conflicts or wars.
The second possibility is that the norms or roles under contention in an
enduringrivalrydo not remainthe same throughtime. Instead,the glue thatholds
Goertz and Diehl (1998) and Diehl (2000) also arguedthat rivalries become "locked in" relatively
quickly.
9 This situation would be analogous to Bennett's (1996) requirementthat the issues at stake in a
competitionremainthe same over the course of a rivalry.
8
708
Thies
a rivalrytogetheris the relationshipbetweenthe socializerandsocializee, no matter
what the norm or role undercontentionis at any particulartime. It seems highly
likely thatas statesendurein the system,they will adoptdifferentroles andcomply
or fail to comply with differentnorms.This is especially the case as certainstates
increasetheircapabilitiesand attaina differentstatusthroughtime. This explanation of enduringrivalrieswould arguethatthey shouldpersistuntil such a time that
the differencein capabilitiesbetweenthe socializerand socializee erodes,suchthat
the socializer is no longer able to structurethe socializee's reality.
The Interaction of Socialization and Competition
Socialization and competition have independentand interactive effects on
state behavior. Competitionis largely conditionedby capabilities.Greatercapabilities make a bettercompetitorin an anarchicworld,all otherthingsbeing equal.
Socialization is conditioned by the instruction/learningof the appropriateroles,
rules, and normsof interaction.I proposea simple taxonomyof statesconditioned
by the interactionof the mechanismsof competitionand socialization:emerging
states, minormemberstates, majormemberstates, and greatpowers.
A stateis considereda greatpower if the statematchesWaltz's (1979, p. 162)
great-power categorization and the major-power categorization of Small and
Singer (1982). Majormembersare those statesconsideredby Small and Singer as
majorpowers, yet not consideredgreatpowersby Waltz.Minormembersarethose
states with lesser capabilitiesthathave enduredwithin a system for more than 10
years.Emergingstates arethose statesthathave endured10 yearsor less upontheir
first MID with anotherstate. These four categoriesby no means exhaustthe types
of roles that states may adopt. However, each of these categories is an ideal type
thatmay serve as a "masterstatus,"or a rolethatis salientin every situation(Stryker
& Statham,1985, p. 357).
Socialization also depends on capabilities. Usually, the actor with greater
capabilitiessocializes the actorwith lesser capabilitiesinto a normativeorderthat
favors the continuityof the former's status and position. However, as mentioned
above, socialization is also a process of negotiationand mutualaccommodation.
In general, we can expect the aforementionedstates to be conditionedby socialization and competitionin the following ways:
Emergingstates shouldbe most circumscribedin theirbehavior.These states
will probablyengage in imitativebehaviorfor theirinternalorganizationandtheir
external policies. They will have a small number of well-defined roles in the
system, and the roles they have are likely to be ascribedto them by memberstates.
They may also be unfamiliarwith the norms and rules of interstateinteraction.
Emerging states will be subject to the most intense socializationpressureof any
kindof statein the system,andwill be the least likely to resistsuchpressurebecause
of theirlow or uncertaincapabilities.
Enduring Rivalries
709
Minor memberswill have a largernumberof roles, and more well-developed
roles than emerging states. Minor members, having a greater sense of their
capabilitiesand identity,will attemptto achieve roles in the system in additionto
their ascribedroles. These states may still be imitatingothers,but theirexpanded
role location efforts may subjectthem to conflict if othermembersdisagree with
their role conceptions. Minor members may be both membersof an established
system and novices when the system changes.
Majormembers,because of theirgreatercapabilities,will have a greaterarray
of well-developed roles, and more of these roles will be achieved than ascribed.
The greater capabilities that they have are partly a function of their natural
endowments,but also a function of their ability to innovate their internalorganization to make the best use of those capabilities.Majormembersare responsible
for socializing emerging states and minor member states in their geographic
subsystem. However, majormemberscannot act with impunity, as they are still
subjectto socializing activitiesby the greatpowers at the level of the international
system and in the regional subsystemif it is a greatpower's traditionalsphereof
influence.
Greatpowers, because of theiroverwhelmingcapabilitiesrelativeto the other
types of states, will have the largestnumberof well-developed roles. These roles
are likely to largely reflect the achievementsof greatpowers on the basis of their
capabilities.Greatpowers are still subjectto socializing influence as membersof
a system, but they can often force the adoptionof a role for themselvesor disregard
norms in the face of externalpressure.Greatpowers usually maintaintheir status
as the system changes, unless theircapabilitiesand statusare severely depreciated
duringthe war thatleads to transition.Greatpowers areresponsiblefor socializing
all othertypes of states in the system.
All states will begin their life as emerging states, as imitatorsand adaptorsto
the internationalsystem.As they successfullyendurein the system,emergingstates
will at some point enter one of the other categories. Most states will move to the
minormembercategoryandundergoresocializationas the systemchangesthrough
time. Othersmall statesmay become "members"of the system,not becauseof their
own capabilitiesor intentions,but because other states toleratethem as members
despite their imperfect imitation of the form and function of states (i.e., quasistates).Finally,stateswith greatercapabilitiesmaybecome majormembersor great
powers, with the abilityto exert theirinfluenceregionallyor globally, respectively.
Propositions Derived from Socialization and Competition
Concerning Rivalries
Theoretically,both competitionand socializationshould be importantfactors
conditioningstate interaction.We can derive a numberof propositionsabout the
operationof both factorsfrom the previoustheoreticaldiscussion.Capabilitiesare
importantdeterminantsof both competitionand socialization.Capabilitiesallow
Thies
710
states a certain statusor position in the internationalsystem. The maintenanceof
this status,once attainedthroughcompetition,is a primaryconcernfor socialization
efforts. Thus, socializationefforts emanatefrom states with greatercapabilitiesto
those with lesser capabilities.This should not imply that states with lesser capabilities will not try to resist socialization efforts. Such resistance can lead to
enduringrivalries.
I deduce the following propositionsregardingrivalrousinteractionfrom the
theoreticaldiscussion of competition and socialization in an internationalsocial
system. These propositionsare a first cut at examiningthe effects of socialization
and competitionwithin rival dyads.
P-1.
a.
b.
c.
d.
Greatpowers compete with othergreatpowers.
Greatpowers socialize majormemberstates.
Greatpowers socialize minormemberstates.
Greatpowers socialize emerging states.
P-2.
a.
Majormemberstates compete with othermajormemberstates.
b.
Majormemberstates socialize minormemberstates.
c.
Majormemberstates socialize emerging states.
P-3.
Minormemberstates compete with otherminormemberstates.
a.
Minormemberstates socialize emerging states.
b.
P-4. Emergingstates compete with and socialize otheremergingstates.
P-5. The majorityof rivalrousinteractionwill be within a region.
P-6. Greatpowers will socialize outside of theirregion, as theirinterest
is in maintaininga global orderfavorableto theircurrentstatus.
P-7. System changes will requireresocializationof existing majorand
minormembers.This should resultin the beginning of new rivalriesand
ending of old rivalriessince 1945.
What evidence is there to support these propositions? Although in-depth
analysisof each case will be necessaryto documentthe effects of socializationand
competition,it is possible to gather some preliminaryindirectevidence from the
most recentlisting of rivalriesfrom 1816 to 1992 (Bennett, 1998, pp. 1215-1216).
This list is reproducedwith my coding for state statusin Table I. The state status
listed in Table I correspondsto the ideal statetypes or masterroles conditionedby
the interaction of the mechanisms of competition and socialization found in
Figure 1. The ideal types give us a rough idea of how these states will be
conditionedby competition and socialization, and should help us understandthe
characteristicsthat led to their participationin the rivalriespresentedin Table I.
711
Enduring Rivalries
Table I. Bennett's (1998) UpdatedList of EnduringRivalries, 1816-1992
State 1
U.S.
U.S.
U.S.
U.S.
U.S.
U.S.
U.S.
Ecuador
Brazil
Bolivia
Chile
U.K.
U.K.
U.K.
U.K.
Belgium
France
France
France
Aust.-Hung.
Italy
Greece
Greece
Russia
Rus./U.S.S.R.
Rus./U.S.S.R.
Somalia
Jordan
Iraq
Egypt
Syria
Lebanon
Afghanistan
China
China
India
Thailand
Iran
N. Korea
Honduras
Germany
Italy
Greece
Serbia
U.S.S.R.
Spain
Morocco
Israel
Status
State 2
Status
Major
Major
GP
Major
Major
GP
GP
Minor
Emerging
Minor
Minor
GP
GP
GP
GP
Minor
GP
GP
GP
GP
GP
Minor
Minor
GP
GP
GP
Emerging
Emerging
Minor
Minor
Emerging
Minor
Minor
Minor
Major
Emerging
Minor
Minor
Emerging
Emerging
GP
GP
Minor
Minor
GP
Minor
Emerging
Emerging
Haiti
Mexico
Ecuador
U.K.
Spain
U.S.S.R.
China
Peru
U.K.
Paraguay
Argentina
Germany
Rus./U.S.S.R.
U.S.S.R.
Ott./Turkey
Germany
Germany
Ott./Turkey
China
Italy
Ott./Turkey
Ott./Turkey
Turkey
Ott./Turkey
China
Japan
Ethiopia
Israel
Israel
Israel
Israel
Israel
Pakistan
Japan
India
Pakistan
Kampuchea
Iraq
S. Korea
Nicaragua
Italy
Yugoslavia
Bulgaria
Bulgaria
Norway
Morocco
Algeria
S. Arabia
Emerging
Emerging
Minor
GP
Minor
GP
Major
Minor
GP
Minor
Minor
GP
GP
GP
Minor
GP
GP
Minor
Minor
GP
Minor
Minor
Minor
Minor
Minor
GP
Minor
Emerging
Minor
Emerging
Emerging
Minor
Emerging
Minor
Emerging
Emerging
Emerging
Minor
Emerging
Emerging
GP
Emerging
Emerging
Emerging
Minor
Emerging
Emerging
Minor
FirstMID StartYear* End Year
1869
1836
1952
1837
1850
1946
1949
1891
1826
1906
1873
1887
1833
1940
1876
1914
1830
1897
1860
1904
1880
1854
1958
1876
1862
1895
1960
1948
1967
1948
1948
1965
1949
1873
1950
1947
1953
1953
1949
1907
1914
1923
1913
1913
1956
1957
1962
1957
1891
1859
1972
1858
1873
1966
1969
1911
1849
1927
1897
1919
1876
1961
1905
1938
1866
1920
1898
1926
1908
1878
1978
1898
1898
1917
1980
1968
1991
1968
1968
1985
1974
1894
1971
1967
1975
1973
1970
1929
1939
1953
1940
1940
1978
1979
1984
1981
1915
1927
1992
1903
1898
1992
1972
1992
1865
1938
1984
1955
1907
1992
1926
1954
1955
1939
1929
1930
1928
1923
1992
1923
1992
1992
1992
1992
1992
1979
1992
1992
1992
1951
1992
1992
1992
1992
1992
1962
1956
1956
1954
1956
1992
1992
1992
1992
712
Thies
Table I. (cont.)
State 1
Syria
Iraq
U.K.
U.S.
U.S.
U.S.
China
S. Korea
Cyprus
Rus./U.S.S.R.
Congo
Uganda
Ethiopia
Thailand
Thailand
Status
State 2
Status
Emerging
Minor
Major
GP
GP
GP
Major
Emerging
Emerging
GP
Emerging
Emerging
Minor
Minor
Minor
Jordan
Kuwait
Iraq
Cuba
Peru
N. Korea
S. Korea
Japan
Turkey
Iran
Zaire
Kenya
Sudan
N. Vietnam
Laos
Emerging
Emerging
Minor
Minor
Minor
Emerging
Emerging
Emerging
Minor
Minor
Emerging
Emerging
Minor
Emerging
Emerging
FirstMID StartYear* End Year
1971
1990
1984
1979
1992
1975
1976
1977
1988
1933
1987
1989
1987
1980
1980
1949
1961
1958
1959
1955
1950
1950
1953
1965
1908
1963
1965
1967
1960
1960
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
1992
*Bennett's (1998) operationalizationmarks the startof a rivalry 20 years after the first MID. This
contradictsGoertzandDiehl' s (1992) operationalizationof rivalryinitiationat the firstMID. Rivalries
shown as ending in 1992 may still be continuing.
CompetitionMechanisms
SocialProof
Heuristic
RationalImitation
OrganizationalCompetency
EmergingStates
MajorMembers
Low/Uncertain
Capabilities
Few Roles
AscribedRoles
HighSocializationPressure
HigherCapabilities
MoreRoles
Roles
Achieved/Ascribed
MediumSocializationPressure
Socialization
Mechanisms
MinorMembers
Dissonance
Reduction
ModerateCapabilities
MoreRoles
Roles
Ascribed/Achieved
MediumSocializationPressure
GreatPowers
GreatestCapabilities
MostRoles
AchievedRoles
Low SocializationPressure
Figure 1. Interactionof competitionand socializationmechanisms.
EnduringRivalries
713
My approachto the inclusion of socializationas a conditioningstructuralfeature
is similarto Goertzand Diehl's (1992) originalassumptionthatall rivalrousstates
are engaged in competition.As discussed previously, competitioncannot necessarilyaccountfor the rivalrousinteractionin unbalanceddyads. I assume,pending
furtherinvestigation,thatsocializationcan accountforrivalrousinteractionin these
dyads.The examples of rivalriesconditionedby socializationshouldserve primarily as illustrationsof the potentialof this approachto explainingrivalries.Future
researchshould focus on testing specific hypothesis about socializationand competition in the cases identified throughthis initial plausibilityprobe of previously
identifiedrivalries.
P-1.
a.
There are eight cases of rivalry that begin between two great
U.K.-Russia/
powers: U.S.-U.S.S.R.,
U.K.-Germany,
U.S.S.R., U.K.-U.S.S.R., France-Germany, Austria-HungaryItaly, Russia/U.S.S.R.-Japan, and Germany-Italy. These are
clearly cases of competition.
b.
Therearetwo cases in which greatpowersmay be socializingmajor
members:U.K.-U.S. and U.S.-China.
c.
There are 12 cases in which great powers may be socializing
minor members:U.S.-Ecuador, U.K.-Ottoman Empire/Turkey,
Germany-Belgium, France-Ottoman Empire/Turkey,FranceChina,Italy-Ottoman Empire/Turkey,Russia-Ottoman Empire/
Turkey,Russia/U.S.S.R.-China, U.S.S.R.-Norway, U.S.-Cuba,
U.S.-Peru, andRussia/U.S.S.R.-Iran.
d.
There are three cases in which great powers may be socializing
emergingstates:U.K.-Brazil, Italy-Yugoslavia, and U.S.-North
Korea.
These findings about great-powerinteractionin rivalryare quite interesting.
Almost one-thirdof the cases are competitionsamong the greatpowers. Thereare
fewer cases of greatpowers potentiallysocializing emerging states than might be
expected theoretically.However, it seems possible that great-powersocialization
of stateswith lesser capabilitiesmightbe accomplishedratherquickly.These cases
might fit into Goertz and Diehl's (1992) isolated conflict or proto-rivalrycategories, in which there are not enough disputesto be classified as an enduringrivalry.
Some of these cases of socializationby a greatpowerthatdo attainenduringrivalry
status are probablyfailures or incomplete attemptsat socialization in which the
same roles or normsare undercontention.The aforementionedcases do include a
numberof rogue or pariahstates, such as NorthKoreaand Cuba,which may resist
behavioralchangeaccordingto ourexpectationsfromdissonancetheory.Although
the Cold War between the U.S. and Soviets is often seen solely in terms of
competition,it actually involved socializationattemptsby both poles to maintain
the integrityandcohesion of theirrespectiveblocs. Further,both superpowerstried
714
Thies
to capturestates from the opposing bloc. The U.S. certainlydesired to socialize
North Korea,Cuba,and Chinaout of theirroles as revolutionarycommuniststates
in the post-1945 era.Its failureto achieve thatgoal spawnedenduringrivalrieswith
these states.
The U.S.-U.K. rivalryis a case in which the roles undercontention,and the
statusof the formerparticipant,changedthroughoutthe course of the rivalry.The
U.K. remainedthe U.S.'s dominantsocializer from its emergence to the point at
which it became a greatpower. The U.S. as an emerging state attemptedto adopt
a neutralrole for itself on several occasions by emulatingmembersof the Armed
Neutralityof 1780, which comprisedsmalltradingstatessuch as Denmark-Norway
and Sweden, and eventually the Holy Roman Empire,Prussia,Portugal,and the
two Sicilies as well (Bukovansky,1997). The emulationof similarstatesis exactly
what would be expected of an emerging state accordingto the principleof social
proof. Unfortunatelyfor the U.S., this role was rejectedby the U.K. severaltimes,
ultimatelyleading to the War of 1812 as a method of dissonance reductionusing
a very directform of instruction(war).
As a minor member of the system, the U.S. attemptedto adopt the role of
regionalleaderandprotectorfor LatinAmericathroughthe issuanceof the Monroe
Doctrine in 1823. The U.K. had previously attemptedto altercastthe U.S. in the
role of regionalcollaboratorin LatinAmericaas partof a partnershipbetween the
two states. However, as a minormemberconditionedby the dissonancereduction
mechanism,the U.S. viewed the U.K. proposalas subservientand dissonantwith
its goals of expansion and preeminencein the region. The U.S. held onto its own
conception of a regional leadershiprole despite British pressureto abandonthe
role. Ultimately, the U.K. grudgingly accepted the role by providing the tacit
protection of the British navy in order to prevent its European enemies from
reattaininga foothold in Latin America. By the time of the 1895 boundarycrisis
in Venezuela, the U.K. fully accepted the U.S. regional leader and protector
roles for Latin America. This crisis was seen by many as the emergence of the
U.S. as a great power both in terms of capabilities and social status. The U.S.
then went on to socialize Spain out of its ratherhollow great-powerrole with the
Spanish-AmericanWar in 1898. The U.S.-U.K. rivalry drew to a close soon
thereafter,reflecting the changed status of the U.S. vis-a-vis the U.K. As a great
power enacting its roles of regional leader and protectorin Latin America, in
additionto its role as bloc leader duringthe Cold War,the U.S. intervenedin the
affairsof manyLatinAmericanstates,includingminormembersPeruandEcuador,
to attemptto socialize themto the U.S. vision of the globalpolitico-economicorder.
The interventionby a series of greatpowers (includingthe U.K., France,Italy,
and Russia) in the affairsof the OttomanEmpireoffers an interestingwindow into
the interactionof competition and socialization. Nationalists, supportedby the
various great powers, had begun to carve independentstates out of the declining
OttomanEmpirein the 19thcentury.Serbia,Greece,andRomaniawere established
by 1870, and were joined by Bulgaria,Montenegro,and Albania before the First
Enduring Rivalries
715
WorldWar.Competitionamong greatpowers explains much of the machinations
of this time period. However, the competitionbetween great powers provided a
link to their separate rivalries conditioned by socialization with the Ottoman
Empire.The fact that the OttomanEmpire was often formally treatedas a great
power, despite the lack of capabilitiesto back up thatrole, suggests thatsocialization may be an importantfactorin explainingthe behaviorof statesinvolved in the
Ottomanquestion. The OttomanEmpirehad declined in power to such a precipitous state by the 19thcentury that it was maintainedas an independentpolitical
entity primarilybecause the great powers of Europe could not decide how to
properlydivide it up withoutengaging in a full-scale war.The Russianrivalrywith
Iranalso datesto this period,as the instabilityproducedby a weak OttomanEmpire
threatenedthe entire Middle Easternregion.
The Ottomanquestion that had plagued the Europeangreatpowers since the
18th centurywas eventually settled by the First World War. Until the conclusion
of the war, the Ottoman Empire emulated the behavior of a great power and
certainlyavoided any dissonance-producinginformationto the contrary,particularly the informationprovided by its inability to prevent the formationof independent states in its Europeanterritoryduringa series of Balkan wars supported
by the greatpowers.Itjoined the CentralPowersin the war,andwas soon socialized
out of its pseudo-role as a great power. War in this case, as in the War of 1812
between the U.S. and U.K., is an example of the operationof dissonancereduction
as a social mechanism.The newly independentstates formedbefore the empire's
dissolutionalso proveinterestingto a socializationperspectiveon rivalriesbecause
they all emerged in the same geographicregion in a relatively shorttime. Indeed,
as illustratedin Proposition3, Greece formedrivalrieswith Bulgariaand Turkey,
and Serbiaand Bulgariaformeda rivalryas well. These emergingstates probably
imitated the conflict-pronebehavior exhibited by each other in their subsystem
accordingto the social proof heuristic.Yugoslavia, formedin the aftermathof the
war among the ruins of Austria-Hungary'sBalkan possessions, also formed a
rivalry with Italy based on the latter's continuedinterestin maintaininga sphere
of influence along the AdriaticSea.
P-2.
a.
There are no cases of majormemberscompetingwith each other.
b.
Thereare two cases in which a majormembermay be socializing a
minormember:U.K.-Iraq and the U.S.-Spain.
c.
There are four cases in which majormembersmay be socializing
emerging states: U.S.-Haiti, U.S.-Mexico, China-India, and
China-South Korea.
The interestingthing about Proposition2 is that there are so few cases of rivalry
that meet its standards.The absence of interactionsbetween major members is
probablybecause theircapabilitiesarelimitedto theirown regions.The case of the
U.K. socializing Iraq can probablybe construedas the U.K. acting as though it
716
Thies
were still a greatpower,becausethe MiddleEastwas a traditionalsphereof interest
for the U.K. when it truly was a greatpower. Many observersarguethatthe 1956
Suez Canalcrisis markedthe final socializationof the U.K. out of its role as a great
powerby the U.S. (Waltz, 1993). The U.S.-Spain rivalrystartedas the U.S. began
to assume the role of regional leader and protectorin Latin America, leading to
conflicts with Spain over Cuba.The U.S.-Spain rivalrycame to a conclusion as
the U.S. assumed the mantle of a great power and socialized Spain out of any
remainingdelusionsof great-powerstatusin 1898-a situationin which warclosed
alternativepolicy choices for Spain, and the social dissonance caused by Spain's
continuedpretenseat great-powerstatuswas eliminated.The four cases of major
memberssocializing emerging states all involve regionalinteraction,as expected.
The U.S., in particular,continuedto exercise its role as regionalleaderandprotector
in its interactions(and military interventions)with Haiti and Mexico. Similarly,
afterChinaconsolidatedundercommunistrule,it attemptedto socialize the newly
independentstatesof SouthKoreaandIndiainto its vision of regional,if not global,
communism.
P-3.
a.
Thereare 10 cases in which minormembersmay be competingwith
other minor members: Ecuador-Peru, Bolivia-Paraguay,
Chile-Argentina, Greece-Ottoman Empire/Turkey,GreeceTurkey,Iraq-Israel, Lebanon-Israel, China-Japan, Iran-Iraq,
and Ethiopia-Sudan.
b.
There are 12 cases in which minor membersmay be socializing
emerging states: Somalia-Ethiopia, Egypt-Israel, AfghanistanPakistan,Thailand-Kampuchea,
Greece-Bulgaria,Serbia-Bulgaria,
Spain-Morocco, Saudi Arabia-Israel, Iraq-Kuwait, TurkeyCyprus,Thailand-North Vietnam,and Thailand-Laos.
The remarkablething aboutProposition3 is that one-thirdof all rivalriesinvolve
minormemberscompetingwith and/orsocializing otherminormembersor emerging states. Most of the minor member-minor member dyads appearto involve
competitionover territory,such as Chile-Argentina, Ecuador-Peru, BoliviaParaguay,or Greece-Turkey after the Second World War. However, some of
these cases (such as Iraq-Israel and Lebanon-Israel) do contain elements of
socialization.Both Iraqand Lebanon,in additionto Egypt and SaudiArabia,have
long attemptedto socialize Israel out of the first role adopted by any emerging
state-that of the sovereign state.10Without interventionby the U.S. and other
majormembers,such as FranceandBritain,they may have accomplishedthis goal.
However,even in the absenceof great-powerintervention,ourpreviousdiscussion
of the social proof mechanism and the pace of socialization would expect that
10See Barnett(1993) on the role of the sovereign state in the Middle Easternsubsystem.
Enduring Rivalries
717
Israel's introductioninto the Middle Easternregional subsystemwould producea
long and troubledsocializationprocess. The Iraq-Kuwait and Turkey-Cyprus
rivalriesare also examples of minormemberschallengingthe adoptionby emerging states of the role of the sovereign state.
It is possible that"peer"socializationmay be moreimportantthanmost of the
social psychological literaturewould expect.1I This may make sense if minor
membersthemselveshave previouslybeen socializedinto the appropriateroles and
normsof the internationalor regional system. The greatpowers or majormember
stateshave alreadyinstilledtheirsense of the appropriatenormativeorderintothese
minor members,which then pass that informationon to emerging states through
socializationattemptsin theirlocalities. This may be evidence of a division of labor
wherebygreatpowers socialize majormembersandminormembers,which in turn
socialize emergingstates.This may not be a very efficient methodof socialization,
as many of these emerging states seem able to resist the socialization efforts of
minormembers.
P-4. Thereare 10 cases in which emergingstatesmay be competingwith
and/or socializing other emerging states: Jordan-Israel, Syria-Israel,
India-Pakistan, North Korea-South Korea, Honduras-Nicaragua,
Morocco-Algeria, Syria-Jordan, South Korea-Japan, Congo-Zaire,
andUganda-Kenya.
Rivalrous interactionamong emerging states is also fairly commonplace. I
would expect that these dyads have the potential for the most conflictual and
longest-lived rivalriesin the absence of outside intervention.If two statesemerge
into the system at the same time, compete over some good, and look to each other
for guidanceas to appropriatebehavior,then they essentially socialize themselves
into an ongoing cycle of conflict throughthe social proof heuristic. The actual
source of contentionno longer mattersat some point, as the actorsinvolved adopt
an actor dimension (Mansbach& Vasquez, 1981, p. 60) and reduce all issues to
"us versus them."The group of rivalries in the Middle East, including JordanIsrael, Syria-Israel, and Syria-Jordan, may be examples of the operationof the
social proof heuristic.Because many of these rivalrieshave emerged since 1945
and are ongoing, we may have to wait for some time to see how they play out.
Recent efforts to end the conflict between Israeland her Arabneighborshave met
with continuedhesitancyon the partof both sides to reducethe dissonancebetween
the experienceof the otheras an enemy andthe proposalsthatcast them as partners
in peace.
P-5. As you can tell from a visual inspection of the list of rivalries,the
vast majorityof them occur within a geographicregion. The exceptions
usually involve at least one great power, as expected. Socialization and
l1 See Harris(1998) for an importantexception.
718
Thies
competition are more likely to take place in arenas that offer greater
possibilities for interaction.Regional subsystemsoffer such an arena.
P-6. As noted in P-5, great powers are found to potentially engage in
socialization outside of their region when it comes to major members,
minormembers,and emerging states.The greatpowers have the greatest
stake in maintaininga normativeglobal orderfavorableto their current
status.
P-7. The shift from multipolarityto bipolarityresultedin five rivalries
ending in the period 1945-1955.12 However, 18 new rivalries began
duringthis period. These new rivalriesreflect the division of the world
into two competing social spheres,with the U.S. taking an active role in
attemptingto bringback statesfromunderSoviet influence.The rivalries
spawnedby the Cold War will have elements of socialization and competition, because the U.S.-Soviet rivalry based on competition was
linked to a numberof otherrivalriesthroughsocialization.13
In general,these findings suggest that socializationmay indeed play an important
role in rivalry formation.More than half (N=34) of the rivalrousdyads contain
stateswith differentialpower capabilities.As I arguedabove, competitioncan only
account for interactionamong states with similar capabilities. Socialization is a
worthwhilecandidateto explain these unbalanceddyads. Further,46% (N=29) of
rivalrousdyads include at least one emergingstate.Thus,just as we would expect,
memberstatesof the system(of whatevercapability)maybe involvedin socializing
novices into the system. Furtherinvestigationis needed to explain whetheremerging states in these dyads are able to resist socializationefforts on particularnorms
or roles for extended periods of time, or whether the norms and roles under
contentionchange throughtime, thus requiringongoing socializationefforts.
The findings in this section are highly tentative. The propositionsdeduced
from the theoreticaldiscussion of competitionand socializationdemonstratethat
an approachto rivalry that considers both structuralfeatureshas the potentialto
explain some of the lingering questions about rivalries. This exploratorystudy
should open the door to the generationof testablehypothesesand the creationof a
databasewith informationon the contents of socialization(e.g., norms and roles)
as well as the contents of competition(e.g., issues). Even in its currenttheoretical
stage, this researchcan illuminateadditionalempiricalfindings that puzzle those
operatingin the inductive approach.
121 use Goertz and Diehl's (1995) method of allowing a 10-yearimpact window when assessing the
effects of political shocks (such as a system change) on rivalryformationand termination.I also use
GoertzandDiehl's specificationof the startof a rivalryas the incidenceof its firstMID.This approach
differs from thatof Bennett (1998).
13See Goertz and Diehl (1997) and Diehl (2000, pp. 241-262) on linkages between rivalries.
Enduring Rivalries
719
Additional Findings Explained by Socialization
The second majorproblemwith previousresearch,as notedabove,is its failure
to explain how rivalries are initiated and terminated.Goertz and Diehl (1995)
arguedthatenduringrivalriesrepresenta peculiarkind of stabilityin international
relations.To breakthe stabilityof such conflictualpatterns,it may be necessaryto
introducea large political shock. Indeed, Goertz and Diehl arguedthat political
shocks-defined as world wars, territorialchanges, alterationsin the international
distributionof power, civil wars, andnationalindependence-may accountfor the
initiationandterminationof rivalries.Further,ratherthanpredictingthe beginning
or ending of a rivalry at a fixed point in time, they expected shocks to have an
impactover a rangeof 10 years aftertheiroccurrence.They found that 87%of the
45 enduringrivalriesexaminedin theirstudybegan duringor within 10 years after
any of the five politicalshocks.Further,morethan53%of the rivalriesendedwithin
10 years of a political shock.
Goertz and Diehl (1995) did not flesh out the causal story that links political
shocks to the initiationor terminationof rivalries.A socializationperspectivecan
providea theoreticalexplanationfor the findingthatsystem shockslike worldwars,
large changes in the distributionof territory,and changes in the distributionof
poweraccompanythe onset andterminationof rivalries.Whenthe systemchanges,
all states except the remaininggreatpowers are novices and must be resocialized.
This means that old rivalriesbased on socializationto the norms and roles of the
old system should end, and new rivalriesbased on socializationto new normsand
roles may form. Some preliminaryevidence for this was found in Proposition7.
Emerging states are particularlysubject to socialization attempts. This is
consistentwith Goertzand Diehl's (1995) finding thatnationalindependencesare
a political shockthataccompaniesthe onset of rivalry.Some of these rivalriesform
as emerging states become locked in competition over their insecure territorial
claims; however, othersform as emergingstates take time to learnthe appropriate
norms and roles requiredof them to join the "club of nations" (Maoz, 1989).
Emerging states that are quickly socialized into the system (by great powers,
regionalpowers, or theirpeers) may never become rivals. They may accountfor a
numberof the isolated or proto-rivalriesidentified by Goertz and Diehl (1992).
However, othersmay find themselves in an ongoing relationshipwith a socializer
that will attemptto mold and shape their behaviorfor an extendedperiod of time.
The third general problem of the inductive approach,an explanationof the
varying intensityof rivalries,has received some attentionin the literature.Hensel
(1998) developed an evolutionaryapproachto rivalry that demonstratesthat the
conflict level of a rivalrygraduallyincreases throughtime. However, Goertz and
Diehl (1998) found no supportfor such a "volcano model" of rivalry in which
hostilities escalate over time into war. The little evidence on the varyingintensity
of rivalriesis thus mixed so far. A socializationperspectivewould arguethat the
intensity level of a rivalry will vary accordingto how offensive each episode of
720
Thies
role or norm location was to the normativeorderthe socializer was attemptingto
support.This might include an escalation of hostility if a state repeatedlyviolates
a norm or attemptsto enact a non-sanctionedrole. This may account for ups and
downs in the level of intensityin a rivalryover time, if the state violates different
norms or enacts different non-sanctionedroles. Presumably,repeated offenses
would elicit more intensely conflictualresponsesfrom the socializer. This turned
out to be the case for the U.S., which attemptedto enact a non-sanctionedneutral
role twice, ultimately leading to the War of 1812 with the U.K. A socialization
perspective also accounts for periods of relative calm, in which the socializee is
enacting its role appropriatelyor conformingto establishednorms.The inductive
approach,with its reliance on competition,has so far been unable to explain the
lapses between conflictual episodes.
The inductive approachhas also had some difficulty explaining Bennett's
(1998) and Cioffi-Revilla's (1998) findings, from analyses using survivalmodels,
that rivalriesexhibitpositive durationdependence.Positive durationdependence
means that a rivalry'shazardrateincreasesover time. In otherwords, the longer a
rivalry persists, the more likely it is to terminate.Essentially, rivalries do not
become entrenchedover time. Thus, truth-in-advertising
requiresthatwe dropthe
"enduring"appellationto rivalries. Bennett (1998, p. 1224) was puzzled by the
positive durationdependenceand asked what mechanismmight be responsible.I
suggest that socialization may account for positive durationdependenceoverall,
and for the variancethatCioffi-Revilla found amongrivalrousdyads accordingto
theirpower configurations.Socializationis an ongoing process within the system.
The membersof the system need to be ever ready to attemptto instructor coerce
other states to supportthe system. Those states thatresist socializationefforts and
their would-be socializers are likely candidatesfor rivalries. States with greater
capabilities,or thatare not fully involved in the internationalsystem, are likely to
resist socialization attemptsfor a greater period of time than those with lesser
capabilitiesor a greaterstake in the system.
Majorpower-major powerdyadsarethe least stable(shortest-lived)rivalries,
accordingto Cioffi-Revilla (1998, p. 86).14He suggested that because the maximum amount of power is available to deal with these competitions, they are
unlikely to persist over long periods of time. Major powers are not likely to be
socializing each otherunless a new greatpower enterstheirranks.Thus, competition is the dominantforce amongthe greatpowers, andresultsin the shortest-lived
rivalries.
Majorpower-minor power dyads are more stable (longer-lived).Thompson
(1995) and Vasquez (1996) have questioned the validity of including majorminordyads in the categoryof rivalries.It does not necessarilymake sense to think
of major-minor dyads as competitors,except in the general sense that all states
14Cioffi-Revilla(1998) referredto only two types of states:majorpowers and minorpowers.
721
Enduring Rivalries
are theoretically competitorsin an anarchic system according to Waltz (1979).
However, this dyadic combination is exactly what a socialization perspective
would expect. Major powers should be socializing minor powers within their
sphere of influence. The fact that these dyads persist into rivalriesindicates that
socialization attemptsmay often be incomplete or fail at times, or that they may
involve changes in the roles and normsundercontentionover time.
Minor power-minor power dyads are the most stable (longest-lived). This
may reflect Thompson's(1995) notionthatthese statesdon't have the resourcesto
settle their disputes once-and-for-all.Minor-minor dyads are competitorsand
usuallyneighbors.A theoreticalfocus on competitionwouldpredictthatneighbors
would be most likely to engage in competitive behaviors over time because
neighborsarein the immediatefield of competition(Vasquez, 1993, pp. 134-135).
A socializationperspectivewould also expect thatpeer statesengage in socializationefforts.This could take threedyadicforms:emergingstate-emerging state,
minormember-emerging state,andminormember-minor member.Dyads of two
emerging states might be particularlyconflictual in the absence of a dominant
regional or great-power socializer. A conflict among competitiveemergingstates
could easily become entrenchedif eitherof the new statesbelieve that its survival
is at stake.Once conflict is initiated,these statessocialize each otherinto acceptable
behavioraccordingto the logic of the social proofheuristic-essentially, reciprocity based on conflict. Rajmaira's (1997) work on the India-Pakistan rivalry
appearsto confirmthishypothesis.In the absenceof a great-powersocializer,these
rivalriesmay become entrenched.Bennett(1998, p. 1224) similarlysuggestedthat
institutionalizationmay occur within a rival dyad if it is not disruptedby some
outsidefactor,such as a greatpower.Minormember-emerging staterivalriesmay
be long-lived if the minor member does not have the capabilities to force the
emerging state into conformitywith the appropriateroles or norms.Minor member-minor memberdyadsengagedin competitionandsocializationmay similarly
face a challenge of balanced capabilitiesthat fail to enable either side to prevail.
These situations also may require the intervention of a great power or major
member, or the disruptionof the rivalry by a political shock such as a system
change.
Conclusions
Goertzand Diehl (1996) expressedan understandingof the implicit use of the
rivalryconceptin academictheoryandarguedthatthe time has come to incorporate
rivalryinto theory-building.The dominanttheoreticalexplanationfor rivalriesis
that they are competitions over territorialissues. However, as Waltz (1979) has
argued,competitionis but one of two structuralfactorsconditioningstatebehavior
in the internationalsystem. Socializationalso appearsto play an importantrole in
conditioning enduring rivalries. Further,socialization, along with competition,
may make explicit the role of the rivalry concept in existing researchprograms
Thies
722
such as Modelski's (1987) long-cycle approachandOrganskiandKugler's(1980)
power transitiontheory, among others.
I have sought to expand on Waltz's spare description of competition and
socialization by proposing several mechanismsthat may operateon their behalf,
therebyproducingfour ideal types of states. When these types of states interact,
we can make some general assumptionsabout their behaviors. The competitive
mechanisms may account for rivalries between states with similarly matched
capabilities.The socialization mechanismsmay account for rivalrousinteraction
between states with different levels of capability.Because these mechanismsof
socialization and competition are not peculiar to rivalrousinteraction,they may
accountfor all types of behaviorin the internationalsystem. This researchshould
therefore contributeto a renewal of Waltz's structuraltheory by incorporating
mechanismsassociatedwith the process of socialization.
I have arguedin theoreticaltermsfor the necessity of consideringsocialization
as a conditioningfactorin rivalriesbetween stateswith differentlevels of capability. The propositionsabout socializationput forwardabove have shed some light
on a subset of rivalrycases that should be examinedin greaterdepth.These cases
may not all turnout to involve socialization,yet many of them will. The key point
is that all rivalries are not equivalent. The dynamics of rivalries between major
powers may indeed be similarto rivalriesbetween minorpowers, in that they are
both conditionedby competition.However, the mixed dyads are a differentsort
and are the likely candidatesfor socialization effects. Because the mixed dyads
make up more than half of the rivalry database, it seems logical to pursue a
socializationperspectivethatmay offer insightsinto theirdynamics.In conjunction
with work thathas alreadybeen done on competition,we may yet develop a more
comprehensive explanation of the peculiar feature of the internationalsystem
known as enduringrivalries.
ACKNOWLEDGMENTS
An earlierversion of this paperwas presentedat the 1999 annualmeeting of
the SouthwesternPolitical Science Association, San Antonio, TX. I thank Mark
Schafer, Steve Walker,and the reviewers and editors of Political Psychology for
their helpful comments. Correspondenceconcerningthis article should be sent to
Cameron Thies, Department of Political Science, Louisiana State University,
Baton Rouge, LA 70803-5433. Email:cthiesl @lsu.edu
REFERENCES
Barnett,M. (1993). Institutions,roles, and disorder:The case of the Arabstates system. International
Studies Quarterly,37, 271-296.
Bennett, D. S. (1996). Security, bargaining,and the end of interstaterivalry. InternationalStudies
Quarterly,40, 157-184.
Enduring Rivalries
723
Bennett, D. S. (1997). Measuringrivalrytermination,1816-1992. Journal of ConflictResolution,41,
227-254.
Bennett,D. S. (1998). Integratingand testing models of rivalryduration.AmericanJournalof Political
Science, 42, 1200-1232.
Biddle, B. J. (1986). Recent developmentsin role theory.AmericanReview of Sociology, 12, 67-92.
Bukovansky,M. (1997). Americanidentity and neutralrights from independenceto the war of 1812.
InternationalOrganization,51, 209-243.
Cialdini,R. (1993). Influence: Thepsychology of persuasion. New York:Quill.
Cioffi-Revilla, C. (1998). The political uncertaintyof interstaterivalries:A punctuatedequilibrium
model. In P. F. Diehl (Ed.), Thedynamicsof enduringrivalries(pp.64-97). Urbana,IL:University
of Illinois Press.
Diehl, P. F. (1985). Contiguityand militaryescalationin majorpower rivalries,1816-1980. Journalof
Politics, 47, 1203-1211.
Diehl, P. F. (1998). Introduction:An overview and some theoreticalguidelines. In P. F. Diehl (Ed.),
Thedynamicsof enduringrivalries (pp. 1-25). Urbana,IL: Universityof Illinois Press.
Diehl, P. F. (2000). Warand peace in internationalrivalry. Ann Arbor,MI: University of Michigan
Press.
Elster, J. (1998). A plea for mechanisms.In P. Hedstrom& R. Swedberg (Eds.), Social mechanisms:
An analytical approachto social theory(pp. 45-73). Cambridge:CambridgeUniversityPress.
Festinger,L. (1957). A theoryof cognitive dissonance. Stanford,CA: StanfordUniversityPress.
Geller, D. S. (1993). Power differentialsand war in rival dyads. InternationalStudies Quarterly,37,
173-193.
Gochman,C. S., & Maoz, Z. (1984). Militarizedinterstatedisputes, 1816-1976: Procedures,patterns,
and insights.Journal of ConflictResolution,28, 585-615.
Goertz, G., & Diehl, P. F. (1992). The empirical importanceof enduring rivalries. International
Interactions, 18, 151-163.
Goertz, G., & Diehl, P. F. (1993). Enduringrivalries:Theoreticalconstructsand empiricalpatterns.
InternationalStudies Quarterly,37, 147-171.
Goertz, G., & Diehl, P. F. (1995). The initiationand terminationof enduringrivalries:The impactof
political shocks.AmericanJournal of Political Science, 39, 30-52.
Goertz, G., & Diehl, P. F. (1996). Taking "enduring"out of enduringrivalry:The rivalryapproachto
war and peace. InternationalInteractions,21, 291-308.
Goertz,G., & Diehl, P. F. (1997). Linkingrisky dyads:An evaluationof the relationbetweenenduring
rivalries. In G. Schneider & P. A. Weitsman (Eds.), Enforcing cooperation: Risky states and
intergovernmentalmanagementof conflict (pp. 132-160). New York:St. Martin's.
Goertz,G., & Diehl, P. F. (1998). The "volcano model" andotherpatternsin the evolutionof enduring
rivalries. In P. F. Diehl (Ed.), The dynamics of enduring rivalries (pp. 98-125). Urbana,IL:
Universityof Illinois Press.
Goldstein,J., & Freeman,J. (1990). Three-waystreet: Strategicreciprocityin worldpolitics. Chicago:
Universityof Chicago Press.
Harris,J. R. (1998). Thenurtureassumption:Whychildrenturnout the way they do. New York:Free
Press.
Hedstrom,P. (1998). Rationalimitation.In P. Hedstrom& R. Swedberg(Eds.), Social mechanisms:An
analytical approachto social theory(pp. 306-327). Cambridge:CambridgeUniversityPress.
Hedstrom,P., & Swedberg, R. (Eds.). (1998). Social mechanisms:An analytical approach to social
theory. Cambridge:CambridgeUniversityPress.
Hensel, P. R. (1994). One thing leads to another:Recurrentmilitarizeddisputes in Latin America,
1816-1986. Journal of Peace Research, 31, 281-297.
724
Thies
Hensel, P. R. (1996). Chartinga courseto conflict:Territorialissues andinterstateconflict, 1816-1992.
ConflictManagementand Peace Science, 15, 43-73.
Hensel, P. R. (1998). The evolution of the Franco-Germanrivalry.In W. R. Thompson(Ed.), Great
power rivalries (pp. 86-121). Columbia,S.C.: Universityof South CarolinaPress.
Huth, P. K. (1996). Enduringrivalriesand territorialdisputes, 1950-1990. ConflictManagementand
Peace Science, 15, 7-41.
Huth,P., Gelpi, C., & Bennett,D. S. (1993). The escalationof greatpower militarizeddisputes:Testing
rationaldeterrencetheoryandstructuralrealism.AmericanPolitical ScienceReview,87,609-623.
Huth, P., & Russett, B. (1993). Generaldeterrencebetween enduringrivals: Testing three competing
models. AmericanPolitical Science Review, 87, 61-72.
Ikenberry, G. J., & Kupchan, C. A. (1990). Socialization and hegemonic power. International
Organization,44, 283-315.
Jackson, R. H. (1990). Quasi-states: Sovereignty, international relations, and the third world.
Cambridge:CambridgeUniversityPress.
Jervis, R. (1976). Perception and misperceptionin internationalpolitics. Princeton,NJ: Princeton
UniversityPress.
Keunne, R. E. (1989). Conflict managementin mature rivalry. Journal of Conflict Resolution, 33,
554-566.
Kuran,T. (1998). Social mechanismsof dissonancereduction.In P. Hedstrom& R. Swedberg(Eds.),
Social mechanisms: An analytical approach to social theory (pp. 147-171). Cambridge:
CambridgeUniversityPress.
Mansbach,R. W., & Vasquez, J. A. (1981). In search of theory:A new paradigmfor global politics.
New York:ColumbiaUniversityPress.
Maoz, Z. (1989). Joining the club of nations: Political development and internationalconflict,
1816-1976. InternationalStudies Quarterly,33, 199-231.
Mercer,J. (1995). Anarchyand identity.InternationalOrganization,49, 229-252.
Modelski, G. (1987). Long cycles in worldpolitics. Seattle, WA: Universityof WashingtonPress.
Moreland,R. L. (1985). Social categorizationand the assimilationof "new" groupmembers.Journal
of Personalityand Social Psychology, 48, 1173-1190.
Organski,A. F. K., & Kugler,J. (1980). The war ledger. Chicago:Universityof Chicago Press.
Rajmaira,S. (1997). Indo-Pakistani relations: Reciprocity in long-term perspective. International
Studies Quarterly,41, 547-560.
Resende-Santos, J. (1996). Anarchyand the emulationof militarysystems: Militaryorganizationand
technology in South America, 1870-1930. SecurityStudies,5, 193-260.
Sarbin,T. R., & Allen, V. L. (1968). Role theory. In G. Lindzey & E. Aronson (Eds.), Handbookof
social psychology (2nded., pp. 488-567). Reading,MA: Addison-Wesley.
Schelling, T. C. (1978). Micromotivesand macrobehavior.New York:Norton.
Schimmelfennig, F. (2000). Internationalsocialization in the new Europe: Rational action in an
institutionalenvironment.EuropeanJournal of Interational Relations,6, 109-139.
Small, M., & Singer, J. D. (1982). Resort to arms: Internationaland civil wars, 1816-1980. Beverly
Hills, CA: Sage.
Sorokin,G. L. (1994). Arms,alliances,andsecuritytradeoffsin enduringrivalries.InternationalStudies
Quarterly,38, 421-446.
Spruyt,H. (1994). The sovereign state and its competitors.Princeton,NJ: PrincetonUniversityPress.
Stinchcombe, A. (1998). Monopolistic competition as a mechanism:Corporations,universities and
nation-statesin competitivefields. In P. Hedstrom& R. Swedberg(Eds.),Social mechanisms:An
analytical approachto social theory(pp. 267-305). Cambridge:CambridgeUniversityPress.
Stokes, R., & Hewitt, J. P. (1976). Aligning actions.AmericanSociological Review,41, 838-849.
Enduring Rivalries
725
Stryker,S., & Statham,A. (1985). Symbolic interactionand role theory. In G. Lindzey & E. Aronson
(Eds.), Handbookof social psychology (3rded., pp. 311-378). New York:RandomHouse.
Thompson,W. R. (1995). Principalrivalries.Journalof ConflictResolution,39, 195-223.
Thompson,W. R. (1998). The evolution of a greatpower rivalry:The Anglo-American case. In W. R.
Thompson (Ed.), Great power rivalries (pp. 201-221). Columbia, SC: University of South
CarolinaPress.
Vasquez, J. A. (1993). The warpuzzle. Cambridge:CambridgeUniversityPress.
Vasquez, J. A. (1996). Distinguishing rivals that go to war from those that do not: A quantitative
comparativecase studyof the two pathsto war. InternationalStudies Quarterly,40, 531-558.
Waltz, K. N. (1979). Theoryof internationalpolitics. New York:McGraw-Hill.
Waltz, K. N. (1993). The emergingstructureof internationalpolitics. InternationalSecurity,18, 44-79.
Wendt, A. (1992). Anarchy is what states make of it: The social constructionof power politics.
InternationalOrganization,41, 335-370.
Wendt,A. (1999). Social theoryof internationalpolitics. Cambridge:CambridgeUniversityPress.
Wentworth,W. M. (1980). Contextand understanding:An inquiryintosocializationtheory.New York:
Elsevier.