Schweller, R. (1993) - Tripolarity and The Second World War

Tripolarity and the Second World War
Author(s): Randall L. Schweller

Source: International Studies Quarterly, Vol. 37, No. 1 (Mar., 1993), pp. 73-103
Published by: Wiley on behalf of The International Studies Association
Stable URL: http://www.jstor.org/stable/2600832 .
Accessed: 09/10/2014 22:23
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Wiley and The International Studies Association are collaborating with JSTOR to digitize, preserve and extend
access to International Studies Quarterly.
http://www.jstor.org
This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

All use subject to JSTOR Terms and Conditions
(1993) 37, 73-103
Intenational StudiesQuarterly
Tripolarity
and theSecond WorldWar
RANDALLL. SCHWELLER
ColumbiaUniversity
This essay seeks to offera new structuralaccount of the outbreak of

WorldWar II and a more determinatebalance-of-power theorybased on
two modificationsof Kenneth Waltz's theoryof internationalpolitics.
First,the distributionof power in the internationalsystemis more
preciselyspecified. Instead of simplycounting the number of Great
Powers to determinesystempolarity,units are divided into poles and
middle powersand weightedaccording to theirrelativepower capabili-
ties. Second, states are coded as either statusquo or revisionist.The
revised theorymore accuratelyreflectsthe twinfoci of classical realist
thought:the power and interestsof states.Several deductionsfromthe
model, however,contradictbasic tenetsof balance-of-power theory.At
the theoreticaland empiricallevels,the theoryis used to examine the
dynamicsof tripolarsystemsand to explain the alliance strategiesof the
seven major powers (three of which were poles) shortlybefore and
duringthe Second WorldWar.
Introduction
Althoughit is widelybelieved that systemicconstraintsand pressuresplayed an
important role in the origins of the First World War, scholars typicallydo not
associate structural factors with the outbreak of the Second World War. The
common image of World War I pictures statesmen hopelessly gored on the horns
of a structural dilemma, unable to prevent a war that none of them wanted,
whereas that of World War II evokes a tragic drama driven by a unique cast of
deplorable characters.
Reinhard Meyers suggests that the problem with the standard historical account
of World War II is that "the actors in the drama appear only as personified images,
no longer as real persons. . . . The drama has a villain (Hitler) and a sinner
(Chamberlain)-what more does one need to explain the outbreak of war in 1939,
especiallywhen the supportingroles are playedbylesservillainssuch as Mussolini
and Stalin, and lesser sinnerslike Beck and Daladier" (quoted in Richardson,
1988:305).
Most students of international relations also employ a villain/sinner image to
explain the origins of World War II. The father of neorealism himself, Kenneth
Waltz, at least partiallyendorses it:
Author's note:I would like to thankThomasJ. Christensen,Barbara Farnham,Ted Hopf, RobertJervis,

Stephen D.
Krasner,Edward D. Mansfield,James McAllister,Jon Mercer, Sue Peterson,Jack Snyder,and the reviewersand
editorsof ISQ for theirhelpfulcommentsand criticisms.Remainingerrorsare mine. The financialsupportof the
Pew Foundationis gratefully acknowledged.
? 1993 InternationalStudiesAssociation.
PublishedbyBlackwellPublishers,238 Main Street,Cambridge,MA 02142, USA, and 108 CowleyRoad, OxfordOX4
1JF,UK.

74 and theSecondWorldWar
Tripolarity
A small-numbersystemcan alwaysbe disruptedby the actionsof a Hitlerand the

reactions of a Chamberlain.... One may lament Churchill's failure to gain
controlof the Britishgovernmentin the 1930s, for he knewwhat actions were
requiredto maintaina balance of power. (1979:175-176)
Likewise,prominentgame theoristssuch as EmersonNiou and Peter Ordeshook

"sympathize . . . withthe analysesthatinterpretHitler'spersonality... as criticalto
the outbreakof WorldWar II ratherthan some breakdownin traditionalbalance-
of-powerforces"(1990:1231). And John Mueller argues thatHitler alone caused
World War II: "It almost seems that afterWorld War I the only person leftin
Europe who waswillingto riskanothertotalwarwasAdolfHitler"(1988:75).
Seeking to shed new lighton an importantcase which scholarshave come to
thinkof as familiar,in thisessayI attemptto showthatsystem-level factorsplayeda
crucial role in the outbreakof the Second World War. I argue thatin 1938 the
systemwas tripolar-not multipolar(5 or more poles) as claimed byWaltz (1979),
Posen (1984), and Christensenand Snyder(1990)-and thatthisexplainsmuch of
the alliance patternsand foreign-policy strategiesof the majorpowerspriorto and
duringthewar.
One reason nonsystemic explanationshave dominatedthe literatureon World
War II is thatthe predictionsof balance-of-power theory-themostwidelyaccepted
structuraltheoryof internationalrelationsand the cornerstoneof realisttheory-
are oftenindeterminateand thus not veryuseful (see Haas, 1953; Gulick,1955;
Claude, 1962; Liska, 1962:26-27; Wolfers,1962:ch. 8). In Theoryof International
Politics(1979), Waltzattemptsto solvethisproblembyrecastingclassicalrealismin a
more rigorousand deductivemold. Operatingat a highlevelof abstraction, Waltz's
theory addresses broad questions: why balances recurrentlyform after their
disruption;what degree of stabilityis to be expected of internationalsystemsof
varyingstructures; and how the constrainingeffectsof structurereduce the variety
of behaviorsand outcomesso thatbalancingbehaviorresultseven when no state
seeksbalance as an end.
Waltz'sideas have been both the intellectualspringboardforimportantresearch
within the structural-realist paradigm (Posen, 1984; Walt, 1987; Grieco, 1988;
Christensenand Snyder,1990; Mearsheimer,1990) and the main targetof neore-
alism's detractors(Kaplan, 1979:1-89; Rosecrance,1981; Keohane, 1986; Wendt,
1987; Dessler,1989; Milner,1991). Some criticsof Waltz'stheorychargethatin its
sacrificeof richnessfor rigor structuralrealismwinds up being a caricatureof
classicalrealismratherthana model ofit (Ashley,1984).1 Othersadmirethe power
and elegance ofWaltz'stheorybut complainthat,as a systemic-level theory,itis too
abstractto generate useful hypothesesabout specificforeign-policy behavior,as
Waltz readilyadmits (Keohane, 1983:512-527; Ruggie, 1983:267-268; Keohane,
1986; Nye,1988:245; Christensenand Snyder,1990). Bysacrificing some of Waltz's
parsimony,however,it is possible to turnhis theoryof internationalpoliticsinto
one offoreignpolicy.
The aim of thisessay,then,is twofold:(1) to offera new structuralinterpreta-
tion forthe originsof WorldWar II, and (2) to devisea systemstheorythatyields
determinatebalance-of-power predictions.The article opens with two modifica-
tionsof Waltz'stheory:(1) the distribution of capabilitiesis measurednot onlyby
the number of Great Powers but also by their relativesize, and (2) statesare
identifiedas eitherstatusquo or revisionist powers.The theoryis then applied to
analysisof tripolardynamics,to a discussionof alliance patterns,and to an expla-
nation of the alliance strategiesof the GreatPowersduringthe period 1938-1945.
IThough farmore sympathetic to Waltz'sworkthanAshley,Glenn Snyder(1991:123, 138) raisessimilarconcerns

in his critiqueofWalt (1987) and Niou, Ordeshook,and Rose (1989).

RANDALLL. SCHWELLER 75
The essay concludes by suggestingimplicationsof the analysisfor the post-Cold

Warworld,whichmaybe becomingtripolar.
Two Modificationsof Waltz'sSystemsTheory
TheDistribution
ofCapabilities
Waltzoffersa tripartite theoreticaldefinitionof systemstructure:(1) the ordering
principle,eitheranarchyor hierarchy;(2) the functionaldifferentiation of units;
and (3) the distributionof capabilities.He claims that because international
politicstake place withinan anarchicrealm and that"as long as anarchyendures,
states remain like units," internationalsystemsdifferonly along the third
dimension,the distributionof capabilities(Waltz,1979:93). Waltzoperationalizes
thisdimensionby simplycountingthe numberof GreatPowersin the system;the
totalnumberof GreatPowersdefinesthe polarityof the system.
Historically,however,the resourcesof the Great Powershave varied consider-
ably, and these imbalances often prove decisive in explaining their individual
foreign-policy strategies(Tucker, 1977). The keyquestion is, Does it matterthat
Waltzabstractsconsiderablyfromreality?
For Waltz, the answer is clearlyno; his theorypertainsto the propertiesof
systems,not individualstates. For those who would use his theoryto explain
foreignpolicy(e. g.,Posen, 1984; Walt,1987; Mearsheimer,1990), however,it does
pose a problem,as Christensenand Snydercorrectly pointout (1990:138). To turn
Waltz'sideas into a theoryof foreignpolicy,the descriptiveaccuracyof the theory
mustbe improvedto accountforpowerinequalitiesamong the majoractors.
Not surprisingly, recent attemptsto apply Waltz's theoryto analysisof the
post-ColdWarsystemhavefocusedon powerasymmetries among theGreatPowers.
For instance,Mearsheimerstates:"Both [bipolarand multipolar]systems are more
peaceful when equalityis greatestamong the poles" (1990:18). Many analysts,
however,strongly disagreewiththisproposition(Organski,1958:271-338;Wagner,
1986; Niou, Ordeshook,and Rose, 1989; Niou and Ordeshook, 1990). Niou and
Ordeshook conclude thatsystemstability does not require "eithera uniformor a
highlyasymmetricresource distribution"(1990:1230). And Wagner (1986) and
Niou, Ordeshook,and Rose (1989) maintainthatthemost"peacefuldistribution" is
one in whichone actorcontrolsexactlyhalfof thesystem'sresources.
To specifymorefullythe distribution of capabilities,I employa two-stepprocess.
First,each Great Power is weightedaccording to its relativeshare of the total
resourcesof the major-powersubsystem. This measurecapturesthe relativepower
disparitiesamong the GreatPowersand drivesthe analysis.By itself,however,it is
too unwieldyto be usefulas a wayto classify different typesof systems.To solvethis
problem,I furtherdividethe GreatPowersintotwotiers:poles and second-ranking
GreatPowers (hereaftercalled middle powers). To qualifyas a pole, a statemust
have greaterthan half the resourcesof the most powerfulstatein the system;all
other Great Powers are classed as middle powers. This conception of polarity
accordswiththe commonsensenotion thatpoles mustbe GreatPowersof the first
rank.Middle powerscan playimportantroles as kingmakersor stabilizers,but the
behaviorof the poles, due to theirsize,has the greatesteffecton the system.
TheCharacter
oftheUnits:StateInterest
Waltz assumes that statesare "unitaryrational actors,who, at a minimum,seek
theirown preservation and, at a maximum,driveforuniversaldomination"(Waltz,
1979:118). He furtherassertsthat,forhis theoryto work,it need not be assumed

Tripolarity
that "all of the competingstatesare strivingrelentlessly to increase theirpower,"

only that some statescontinue to maintainan interestin preservingthemselves
(1979:118). While acknowledgingthatunitsmay differaccordingto theirpower-
seekinginterests, Waltzdoes not incorporatethisunitvariationin his definitionof
internationalstructure-itsimplywashes out of the analysis.Statesare described
insteadas likeunits.This is importantfortworeasons.
First,the characteristicbalancing behavior of Waltz's self-helpinternational
systemis triggeredpreciselybystatesthatwishnot simplyto survivebut to weaken
and destroyotherstates,and, at a maximum,to "become so powerfulas to be able
to coerce all the restput together"(FriedrichGentz,quoted in Gulick,1955:34).
Since Waltz makes no attemptto determinethe extentof a state'sgoals fromits
structuralposition in the system,the catalystdrivinghis theorymust come from
outside the boundariesof his system(see Snyder,forthcoming; Wendt,1992:395).
Second, Waltz'stheoryassumes thatsystemsof the same polaritybehave similarly
despite differencesin the power interestsof the units.Walt (1987) convincingly
shows,however,thatstatesallyto balance againstthreatsratherthanagainstpower
alone. Thus, changes in unit interestsalone can drasticallyalter systemdynamics
and stability.
For these reasons,I treatthe powerpreferencesof the actorsas a model-based
featurethatdifferentiates both the unitsof the systemand, as a result,systemsof
the same structure.For the sake of parsimony,the continuousconcept of state
interestwill be reduced to a dichotomousvariable;unitsare identifiedas either
statusquo or revisionist.
Statusquo powersare usuallythosestatesthatwon the lastmajor-power warand
created a new world order in accordance with their interestsby redistributing
territoryand prestige (Wolfers, 1962:18, 84-86; Seabury, 1965; Buzan,
1983:175-186). As satiatedpowers,statusquo statesseek primarilyto keep, not
increase,theirresources.For thesestates,the costsofwarexceed thegains.
By contrast,revisioniststates-variouslycalled imperialist(Morgenthau,1948),
expansive (Weber, 1946), revolutionary(Kissinger,1957), have-not (Mattern,
1942), aggressors(Jervis, 1978), or unsatiatedpowers (Schuman, 1948:377-380)-
are thosestatesthatseek to increase,notjust keep, theirresources.For thesestates,
the gains fromwar exceed the costs.Revisioniststatesare oftenthose statesthat
have increased theirpower "afterthe existinginternationalorder was fullyestab-
lished and the benefitswere alreadyallocated" (Organskiand Kugler,1980:19).
Thus, theyoftenshare a common desire to overturnthe statusquo order-the
prestige,resources,and principlesof the system.
Revisioniststatesare not alwaysactivelyengaged in overturning the statusquo;
theymaybe temporarily passivebecause theylack the relativeeconomic,military,
and/or politicalcapabilitiesneeded to challengethe protector(s)of the statusquo
(e. g., the SovietUnion, 1919-39; Germany,1919-36; and Japan, 1919-31). Buzan
pointsout, "eventhe mostrabid revisionist statecannot pursueitslargerobjectives
if it cannot secure its home base" (1983:177). It should also be noted that
revisionist statesneed not be predatorypowers;theymayoppose the statusquo for
defensivereasons. As Schuman comments,revisioniststatestypically"feelhumil-
iated, hampered, and oppressed by the statusquo" and thus "demand changes,
rectifications of frontiers,a revisionof treaties,a redistributionof territoryand
power"to modifyit (1948:378).
Byelevatingthe concept of stateinterestto a positionequallyprominentas that
of the distributionof capabilities,the model more accuratelyreflectsthe twin-
pillaredaspect of traditionalrealisttheory-itsequal focuson both the powerand
interestsof states.Unlike Waltz's theory,which is all structureand no units,the
revised theorycontains complex unit-structure interactions:predictionsare co-

determinedbythe powerand interestsof the unitsand the structures withinwhich

theyare embedded. Because neitherlevel is "ontologicallyprimitive," the theory
offersa partialsolutionto theagent-structure problemraisedbyWendt (1987).
Other factors,such as geographyand militarytechnology,mightbe added to
the theory(seeJervis,1978; Hopf, 1991; Christensenand Snyder,1990). I mention
these elements when necessary, but space limitations prevent their full
incorporationwithinthe model. It is mycontention,however,thatthesefactorsare
generallyless importantthan the powerand interestsof the units-withoutwhich
geographyand militarytechnologyprovide only partialanswers.Supportingthis
view,Posen testshypotheseson geographyand militarytechnologyagainst the
militarydoctrinesof interwarFrance, Britain,and Germanyand findsthat these
factorsalone are indeterminate(1984:236-239; also see Levy,1984).
The nextstepis to linkunit-structure interactionsto expectedoutcomes,thatis,
systemstability,alliance patterns,individualforeign-policystrategies.The theory's
twoindependentvariablescombine to produce manypermutations, each of which
mustbe analyzedaccordingto itsown systemicproperties.For thisreason,it is not
possible to make quick deductionsfor all typesof systems,let alone testsfor all
classes of cases. The discussionwill thereforebe limited to tripolarity and the
Second WorldWar,arguablythe twomostmisunderstoodcases.
TripolarSystems
Given the definitionalrequirementthatthe weakestpole musthave greaterthan
half the resourcesof the strongestpole, there are four possible tripolarpower
configurations:
(1) A=B=C
(2) A>B=C,A<B+C
(3) A<B=C
(4) A>B>C,A<B+C
Definitions
(1) Revisionist
powers:statesthatseek to increasetheirresources.
(2) Statusquo powers:statesthatseek onlyto keep theirresources.
(3) To qualifyas a pole, a statemusthave greaterthan halfthe resourcesof
the mostpowerfulstatein the system;all otherGreatPowersare classed as
middlepowers.
(4) Resources= militarypowerpotential.
(5) Systemstabilitymeans thatno actorin thesystemis eliminated.
Assumptions
(1) Wars are costly;for revisioniststates,the gains fromwar for nonsecurity
expansionexceed the costs;forstatusquo statestheydo not.
(2) The alternativesamong which the membersof the triadmaychoose are:
(a) do nothing; (b) align withanother member to block an attack; (c)
align withanother member to eliminatethe thirdmember; (d) wage a
lone attackto eliminateone or both membersof the triad.
(3) A strongermemberor coalitiondefeatsa weakermemberor coalition.
(4) The strengthof a coalition equals the combined total resources of its
members: IfA = 3 and B = 2, thenAB = 5.

78 Tripolarity
and theSecondWorldWar
(5) In a coalition attack,the resources of the victimare divided propor-

tionatelyamong the winningcoalitionmembers.In a lone attack,theyare
absorbedin totalbythevictor.
(6) Resourcesare increasedonlybyeliminatinga memberof the triad.States
do notvoluntarilycede resources.
Among otherthings,thislast assumptionand the inclusionof statusquo states

distinguishthe present analysis from that of Niou, Ordeshook, and Rose
(1989:93-97). Glenn Snydernotes in his critiqueof Niou, Ordeshook,and Rose
that their assumption of infinitelydivisible and freelytransferableresources
(1989:47) allowsthe authorsto assert
thatstateswillvoluntarily
cede resourcesto a stronger
challengeror attackerif
necessaryto avoideliminationand that,in turn,thechallengerwillacceptthe
resourcetransfersin preferenceto gainingthesameamountbywar.... Hence,
thedeviceof peacefulcedingof resourceswas introduced to permitthe third
partyto saveitself... bybuyingoffa memberofthethreatening coalitionat a
priceshortofsuicide.The logicalfitbetweenbargaining-set
theoryand a desired
theoreticaloutcome-stability-was improved,but at considerablecost to
empiricalrealism.(Snyder,1991:134,136)
Thus, Niou, Ordeshook, and Rose find,but I do not, that "the key featureof a
three-country systemis thatno nationwillbe eliminated"(1989:95).
Conversely, Wagner'sanalysisof three-actor systems(1986:554-559) assumes,as
I do, thatthe powerof statescan onlybe increasedbyconquest.But in contrastto
my assumptionthat the spoils are divided proportionatelyamong the victors,
Wagnerassumesthatstatesabsorb the victim's"resourcesat a maximumrate of r
unitsper 'day"' and "iftwostatesare at warwitha third,the ratesat whichtheycan
each absorb the third'sresourcesare determined. .. by how the victimtargetsits
resourcesat the twoopposing states"(1986:549-550). Giventhe rulesof Wagner's
game, it followsthatthe victimmusttargetitsresourcesunequally,therebyforcing
the more-opposedattackerto defectbefore the less- (or un-) opposed attacker
gainsmore thanhalftheresourcesof the system.Thus,Wagneralso concludesthat
no actor is ever eliminatedin a three-actorsystem.The usefulnessof Wagner's
assumptionis an empiricalquestion,but it is worthnotingthatNazi Germanyand
Soviet Russia, long rivals,eliminated Poland; neither attackerdefected to the
victim'sside.
Type1. TheEquilateralTripolar A =B =C
System:
This is the mostunstableof all powerconfigurations because it cannotbe balanced
by externalmeans (alliances). Whereas thisis true of any odd-numberedsystem
composed of equal units,the imbalance is proportionallygreatestin this triad,
whereA and B combinedpossess66 percentof the totalresources,as againstCs 33
percent share. (See Burns, 1957:494-499; Caplow, 1968; Dittmer,1981; Kaplan,
1957:34; Kaplan, Burns,and Quandt, 1960:244; Healy and Stein,1973; and Waltz,
1979:163. Conversely,Garnham, 1985:20; Niou, Ordeshook, and Rose, 1989;
Ostromand Aldrich,1978:766; and Wagner,1986:575 findthree-actor systemsthe
most stable.) To ensure its survival,each pole must preventthe formationof a
hostiletwo-against-one coalition.This, in turn,depends upon the specificmix of
revisionistand statusquo poles withinthe triad.
One RevisionistPole. Givenonlyone revisionist

pole, a stablesystemshould result.
Potentialaggressionwilllikelytake the formof the lone revisionistpole attacking
the neareststatusquo pole. But such an attackis improbablebecause individual

RANDALL L. SCHWELLER 79
aggressionin tripolaritytendsto be self-defeating. On the one hand, ifa stalemate

results(and thisis mostlikelyamong equallypowerfulpoles), both combatantsare
weakened in relation to the neutral, or enjoying,third (tertiusgaudens),who
obviouslyhas no incentiveto join the fray.On the other,if the revisionistpole
appears to be winning,the neutralthirdmustjoin the statusquo pole to ensureits
own survival.Consequently,war is unlikelybecause the revisionistattackercan
expectonlya warof attrition or certaindefeat.
A geographicallydistanced pole may assume one of four roles: (1) tertius
gaudens,(2) the abettor,(3) the eyewitness, or (4) the mediator.The firsttwoare
likelyto be playedbya revisionist pole, the lattertwobya statusquo pole. As tertius
gaudens,the remotepole turnsthe dissensionof the twogeographicallyproximate
poles to its own advantageby askingan exploitativeprice forits support.As the
abettor,it instigatesconflictbetweenthe other membersof the triadforits own
purposes. In the role of eyewitness, it does nothingand seeks nothingfromthe
conflict.In the mediatorrole, it declares neutrality towardthe conflictand works
to stabilizethe triad.2
Two Revisionist Poles.A tripolarsystemcomposed of two revisionistpoles is

typicallyunstablebecause both power-maximizers are highlymotivatedto augment
theirresourcesat the expense of the lone statusquo pole. As a result,the two
revisionistpoles can be expected to put aside theircompetitiontemporarily so as
to make substantialgains. And because the winning coalition consistsof two
equally powerfulpartners,each membercan expect to gain half the spoils,such
thatneitherpole willbe leftvulnerableto the otherafterthe war. Hence, thereis
littleto preventthe partitioningof the isolated statusquo state. Eliminatinga
memberof the triadtransforms the systeminto a relatively stablebut competitive
bipolar system,as the tworemainingpoles (both power-maximizers seekinggains
at the other'sexpense) are expectedto resumetheirrivalry.
A favorablegeographic position may temporarilyprotectthe lone statusquo
pole, but it will not appreciablystabilizethistypeof tripolarsystem.Suppose that
the two revisionistpoles are in close geographicalproximityand the statusquo
pole is distanced fromthem. The triad contains,in effect,two subsystems:one
bipolar,the other unipolar,whereinthe rivalrybetweenthe two revisionistpoles
protectsthe remotestatusquo pole bypreventingthemfromgangingup againstit.
The statusquo power may assume the role of either "the enjoyingthird"or the
"abettor."In eithercase, it remainson the sidelines,hoping thatthe tworevisionist
poles willbleed each otherto death.
But thisis a dangerousstrategy forthe statusquo power.One of the revisionist
poles maybe able to defeatthe othereitherbyforminga winningalliance withone
or more revisionistpowersor by devouringweakerstatesuntilenough additional
resourcesare acquired to defeatthe nearestpole. Once accomplished,the victo-
riousrevisionist pole would be in controlof at least twiceas manyresourcesas the
lone statusquo pole. (I willdiscussHitler'suse of thisstrategy and whyitfailed.)
ThreeRevisionist Poles.In this case, in which all three membersare of equal

strengthand are revisionistpowers,any possible coalition-AB, BC, and AC-is
equally likely,makingit an extremelyvolatilesituation.All three membersseek
coalition,since isolationmeans extinction.But the structureof thistypeof tripolar
systemprohibitsexternal"balancing"behavior(alliances forthe purpose of coun-
21 have borrowedthe terms"abettor"and "eyewitness" fromLiska (1962: 163-164); "tertius

gaudens"and "the
mediator"are takenfromCaplow (1968:20); who,in turn,borrowedthemfromSimmel (1950:148-149).

Tripolarity
terbalancinga strongeror more threateningpower or alliance) because any

the isolatedthirdmember.In addition,since all threemem-
coalitioneasilydefeats
powers,each pole mustalwaysbe suspiciousof the other two
bers are revisionist
and none can enjoy true, long-termsecurity.Consequently,this systemshould
exhibitthe tendencyto evolve into a stable bipolar system,as two poles will be
temptedto destroythe isolatedthirdto gain lastingsecurity.Eliminatingone pole
throughpartitioningshould not presentmajor difficulties because each member
of the winningcoalitionis of equal strengthand is thereforeentitledto an equal
shareof thespoils.
ThreeStatusQuo Poles.This, of course,is the moststabletripolarsystembecause

the interestsof the actorsare not strictly
opposed to each other.In thissystem,in
whichno pole seeks the eliminationof anyother,the integrity of all threeactorsis
virtuallyassured. Continued systemicstabilitysimplyrequires the poles to make
theirintentionsknown,avoid provocativeacts,and coordinateand consultwiththe
othersabout theirindividualforeignpolicies. Their tacitagreementnot to attack
other poles in the systemmay be formalizedby mutual non-aggressionpacts
between all dyads and/or a collective securityagreement,wherein each pole
promisesto come to the aid of anyattackedpole.
Type2. TheParadoxofPower:A > B = C butA < B + C

In thistypeof tripolarsystem, A is slightly
strongerthanB and C,who are of equal
strength.All threemembersstrivefora coalition,as any combinationdefeatsthe
isolatedmember.Paradoxically,when A is a revisionist pole, itsstrengthprovesto
be a handicap,since both B and C findA less attractive as a coalitionpartnerthan
each other.This is true because, in eitheran AB or AC alliance, A would be in
controlof its weakerpartner.Consequently,A, in a coalitionwitheitherB or C,
would be expected to gain at leastan equal share of the rewards,and probablythe
lion's share-furtherdisadvantagingtheweakerally,who mustreceivethe bulk of
the rewardto gain security.Alternatively,a BC coalition(in whichboth B and C are
of equal strength)secures an equal distributionof the reward and does not
threateneithermember.
Conversely,when A is a statusquo pole, eitherB or C mayalign withit against
the isolatedthird.Such an alignmentis especiallylikelyifeitherB or C is a status
quo state,in whichcase an overpoweringstatusquo coalitionwillformto oppose
the lone revisionistmemberof the triad.Finally,when all three poles are status
quo states,no coalitionis predictedbecause thereis no threat.(This "no-coalition"
predictionholds foranysystemcomposed ofall statusquo units.)
Thirdand theBalancer:A < B = C

Type3. ThePartitioned
ThePartitionedThird.When both B and C are revisionist, A cannot align with
either,because once the coalitionhas partitionedthe isolatedmemberof the triad,
A willbe destroyedby its strongerally.The onlyremainingalliance, therefore,is
BC,whichcan safelypartitionA because itsmembersare of equal strength. Indeed,
forsecurityreasonsalone, B and C should partitionA: B mustblock AC and C must
preventAB. Therefore,the mostlikelyscenariois the formationof a BC coalition
forthepurpose of partitioning A and achievingan equal balance betweenB and C.
TheBalancer.However,if B and C are mutuallyhostile,then it is clearlyin A's

by actingas a balancer,gainingat the expense of
interestto prolong theirrivalry
the other two poles. The role of balancer, however,is a dangerous one for the
weakestpole in the systemto play because it must guarantee throughskillful

RANDALL L. SCHWELLER 81
diplomacythatthe twowarringpoles do not temporarily set aside theirdifferences

to gang up againstit. This is no easy task,forby playingthe role of balancer and
continuallyfrustrating the desiresof B and C, A cultivates-throughthe years-a
certain amount of bitternessfromboth B and C. Thus, one would expect that
typicallythe balancer role would be played by a statethatwas strongerthan the
othertwocombined,not bytheweakesttriadicmember.
4. TheUnbalanced
Type Tripolar A > B > C butA < B + C
System:
Althoughthe logic is counterintuitive, a tripolarsystemis most likelyto exhibit
stabilitywhen at least some of the poles are of unequal weights.Consider, for
instance,a tripolarsystemin whichthe powerratioamong the threepoles is A = 5,
B = 4, C = 3. Let us furtherassume thatall membersof the triad are revisionist
powers.At firstglance, thissystemappears to be extremelyunstable (i. e., one of
the actorswillbe eliminated)and warprone, since anycontestbetweentwoactors
is decisiveand anycoalitionis a winningcoalition.Yetit provesto be a verysimple
and stableformof a balance-of-power system.
First,consider an isolated attackwithinthis triad. In such a situation,it is
immediatelyobvious to the third power that it must block the effortsof the
attackerby joining the weaker side, or else be dominated by the victor.The
attackerknowsnot onlythatthe thirdpower mustresistits effortsto destroythe
initialvictim,but also that it has no hope of prevailingagainstsuch a coalition.
Thus, a strongermemberof the triadwillnot be temptedto attacka weakerpower;
the dynamicsof the systemdiscourageoffensive actions.
Now let us consider an attackby a coalition against an isolated thirdpole.
Althougheverycoalition is a winningcoalition,all pairingsin a 5-4-3triad are
unbalanced-they consist of a strongerand a weaker member. Thus, if any
coalition formed,went to war, and won as expected, the systemwould be
transformed into a dyadwiththe weakerpole at the mercyof itsstrongerpartner.
The weakerallywillbe imperiledeven if the spoils of victoryare dividedequally
ratherthanaccordingto each member'sproportionateresources.3
In summary, giventhe dynamicsof thissystem,anypairingwillinevitably result
in onlyone remainingactor; since thisis obvious to all threepoles, no coalition
formsin the firstplace. Though all actors may be power-maximizers and every
pairingformsa winningcoalition,the workingsof the systemvirtually assure the
continuedintegrity of the actorsand the absence of war among them.This is true
forall mixesof revisionistand statusquo states,exceptthefollowingcase.
TwoStatusQuo Poles.Ironically,when thereare twostatusquo poles, the system

is potentiallyunstable.In thissituation,the statusquo poles maybe motivated,for
defensivepurposes,to wage a preventivewar to destroythe revisionistpole, which
poses a latentifnot immediatethreatto theirindividualsecurity.
Hypotheseson AlliancePatterns
So far,the discussionhas been limitedto the behaviorof onlypolar powers.This
analysisleads to broad expectationsabout the overallstabilityof the fourtypesof
tripolarsystems, but it does not explain the specificdynamicsof a givenhistorical
3Evensupposingthatthe strongermemberpromisesin advance to turnover the lion's share of the spoils to its
weaker partner,the latter must still reject the offer.This is because, afterthe initiallytargeted pole has been
vanquished,the strongerpole no longerhas anyincentive(and it cannotbe coerced) to complywiththe agreement.

82 Tripolarity
case. For thiswe need to considerthe behaviorof middleand smallpowersas well

as poles and develop a general theoryof alliance dynamics,based on the two
variablesof the model-differencesin unit size and interests.Unlike the prior
analysisof tripolarsystems,thesehypothesesapplyto all classesofbalance-of-power
systems,thatis,bipolar,tripolar,and multipolar.
TheSizePrinciple
WilliamRiker's "size-principle" hypothesispositsthatthe coalition most likelyto
formis one that containsjust enough strengthto defeat the opposing players
(1962:32-33). This so-called minimum winning coalition is attributedto the
commonsense desire among the winners not to spread the spoils among
superfluouspartners.The greaterthe number of losers,the greaterthe sum of
theirlossesand thegreaterthegainsof thewinners;and the fewerthewinners,the
greaterthe share of each winner.Thus, giventhreeplayersof the followingsizes,
A = 4, B = 3, and C = 2, the size-principle
hypothesispredictsa BC coalition.
The added complexityof more-than-three actor games, however,reduces the
determinatenessof Riker'stheory.If A, B, C, D equal 4, 2, 2, 1, respectively, the
minimumwinning coalition or size principle predicts either an AD or BCD
coalition.4Experimentalevidencehas shown,however,thatAD is more likelythan
BCD because AD is a one-stepcoalitionwhereas BCD requires two steps. Hence,
bargainingcosts are cheaper and less complex for AD than BCD (Shears, 1967;
Russett,1968:292).
In contrastto Riker,Theodore Caplow (1968) assumesthateach playerdesires
controloverall others,includingthe membersof itsown coalition.Consequently,
each actorprefers,ceteris panibus,to alignwithweakercoalitionpartners.
Returningto the example where A, B, C, D equal 4, 2, 2, and 1, respectively,
Caplow's theorypredictsa BCD coalition instead of AD because (given the large
power disparity betweenA and D) D would be extremelyvulnerableto A afterthe
defeatof B and C. Instead,D feelsmore securewithB and C,whichare onlytwice
as strong as D, than with A, which is four times as powerful(see Rothstein,
1968:61). Broadly speaking, the anarchic internationalenvironmentforces all
states to consider seriouslythe possibilitythat today's ally will be tomorrow's
enemy.Analysisof coalitionformationmustconsiderthat "theweakestplayer,by
joining a nearlypredominantstrongplayer,onlycreatesa conditionin whichhe
willbe the nextvictim"(Kaplan, 1979:70). Caplow's theorytakesthisintoaccount;
Riker'stheorydoes not.
TheSizePrinciple:
StatusQuo and Revisionist
Coalitions
Riker's size-principlehypothesisclearly applies to revisionistpowers. This is
because the raison d'e^treof offensivealliances is to maximizeone's share of the
spoils of victory.Any additional member beyond what is needed for victory
diminisheseach member'sshare of the winnings.Hence, revisionist statesdesirea
coalition,as Rikerargues,just strongenough to defeatthe target,and no stronger.
Moreover,the demands of revisionist statesare not alwayscomplementary; thus,a
revisioniststate will want to put togethera coalition with as few members as
possibleand withstatesthatdo not hold conflicting territorial
interests.
Conversely,statusquo powers,whose primaryinterestis self-preservation and
systemstability,formalliancesto deteror defeatrevisionist statesor coalitions.It is
4Moreover,under real-worldconditions,even rationalplayersmaybe uncertainof the distributionof power in

the system.Riker's assumption of perfect informationmay thereforelimit the empirical validityof coalition
predictionsbased on thesize principle(see Riker,1962:48).

reasonable to assume, therefore,that they do not seek minimum winning

coalitionsbut ratherverylarge coalitions,whichbetterservedefensiveor deterrent
purposes. Indeed, the larger the coalition, the less cost to each member of
balancing againstthe threat.This is not to implythatfree-rider problemsdo not
exist in large coalitions (see Olson and Zeckhauser,1966; Sandler and Cauley,
1975; Murdoch and Sandler, 1982; Thies, 1987; Oneal, 1990). It is precisely
because alliance memberscan be expected to do less than theirfairshare thatit
makes sense to forma coalition that is clearlyoverpowering,not slightlymore
powerfulthan the aggressor(s).This is WoodrowWilson'sidea thatstatesare only
deterredbythe threatof a community of power,not a balance ofpower.
Moreover,when the expansionistthreatis large,thereis less of a temptationto
free-ride.This is because, asJervisargues:
Internationalcoalitionsare more readilyheld togetherbyfearthan hope of gain.
... It is no accident thatmost of the major campaignsof expansion have been
waged by one dominant nation (for example, Napoleon's France and Hitler's
Germany),and thatcoalitionsamong relativeequals are usuallyfound defending
the statusquo. Most gains fromconquest are too uncertainand raise too many
questions of futuresquabbles among the victorsto hold an alliance togetherfor
long. Althoughdefensivecoalitions are by no means easy to maintain . . . the
common interestof seeing thatno statedominatesprovidesa strongincentivefor
solidarity.(1978:204-205)
For these reasons, the concept of the minimumwinning coalition-which

derivesitslogic fromexpectationsof dividingthe spoilsofvictory-isnot operative
for status quo alliances, whose raisond'etreis defense and/or deterrence,not
conquest. Note that this hypothesis-that status quo states are attractedto
coalitionslargerthan necessaryto defeator deterthe opposing stateor alliance-
contradictsa centraltenetof balance-of-power theory,namely,thatstatesreact to
powerimbalancesbyjoiningtheweakerside.
Indeed, all stateswerewelcomed to pool theirresourcesto fightNapoleon, the
Kaiser, Hitler, and most recentlySaddam Hussein. Thus, by war's end the
resources of the status quo coalitions far exceeded those of the aggressors
(Kennedy, 1987). Because Riker assumes that all actors are power-maximizers
(none are statusquo), he cannot explain whycoalitionslargerthan necessaryfor
victoryhave repeatedlyformedthroughouthistory.Criticsof Riker'sanalysispoint
out that his assumptionof only maximizingunits also leads him to conclude
(1962:160) wrongly that balance-of-powersystems should be in continual
disequilibrium(Zinnes,1970:356-362;Wagner,1986:569).
Statusquo statesare not affordedthe luxuryof the size principle;theymust
endure the annoyances of large alliances, that is, increased transactioncosts,
disagreementsover rules and decisionmakingproceduresthathave distributional
consequences regardingthe burdensand benefitswithinthe alliance (Stein,1983;
Snyder,1984; Krasner,1991).
RevisionistStatesFlockTogether:
Bandwagoning Alliancesare rarelya mix
Alliances.
of revisionistand statusquo states.This is because revisionist
stateswillonlyjoin a
defensive,statusquo coalitioniftheirsurvivalabsolutelydemandsit;otherwisethey
will flock togetherto overturnthe statusquo and therebyimprovetheirpower
positions.Such bandwagoningbehavior,however,createstensionforthe revision-
istleader,who, seekinga minimumwinningcoalition,mustguard against"preda-
tory"buckpassing,that is, allowing smaller revisioniststates to gain unearned
spoils.Thus, thoughrevisioniststatestend to flocktogether,theydo so onlyin a
limitedsense: theywillnot balance againstthe dominantrevisionist powerbut will
supportit in the hope of attainingtheirown irredentistaims. Such behavioris
capturedbythe ItalianForeignMinister'sremarkof September25, 1938,that"the

84 and theSecondWorld
TriPolarity War
Duce and I, though we did not incite Germanyto war, have done nothingto
restrainher" (Ciano'sDiary,1937-1938,1952:162).
For the same reason that revisioniststates flock together,status quo states
cannotreadilyembracea revisionist state:to do so would be to riskunravelingthe
statusquo to whichtheyare committed.Further,statusquo coalitionspromisea
smallerpayoffto dissatisfiedstatesthan do revisionistcoalitions,since the former
cannot,in principleand fordomesticpoliticalreasons,offerterritorial incentives
to wean therevisioniststateawayfroma revisionist coalition.
Distancing.In addition to the formationof alliances throughbalancing and

bandwagoningbehavior,the proposed theoryidentifiesa thirdtypeof responseto
threats:distancing,or no coalition.This hypothesisposits that threatenedstatus
quo stateswill notjoin the weaker side if the potentialcoalition cannot achieve
enough deterrentor defensivestrengthto dissuadeadversariesfromattacking.
Suppose revisioniststateA threatensstatusquo statesB and C, and A > B + C: no
coalitionis predicted,since A' strengthexceeds B and C combined (whichmeans
thateitherB or C is not a pole, or bothare notpoles). In thesetypesof situations,
in
whichA is the dictator,a BC coalitionwillnot formbecause not onlydoes joining
the weakerside failto make the statesafer,it is also dangerous:the statecommits
itselfto being draggedintoa warit cannotwin.Associatingwiththeweakeralliance
also increasesthe likelihoodthatthe statewillbe seen as a potentialtarget;at the
same timeit depletesthe state'salreadyinadequate resourcesforthe defenseof its
allies-resources thatwould be betterspenton home defense.Consequently,itwill
seek to distanceitselffromother,more immediately threatenedstates,choosingto
remainisolatedeven thoughpotentialalliesare available.
Summaryof Hypotheseson AlliancePatterns

(1) If a statusquo stateor coalitionhas enough strengthto deter or defend
againstthe revisionist
state(s), itwillbalance againstit.
(2) If the combined strengthof twoor more statusquo statesis less than the
threateningrevisionist power,no coalitionwillform.Instead,less directly
threatenedstatusquo stateswill distancethemselvesfromavailable allies
thatare more directlythreatenedbythe aggressor(s).
(2a) Corollary:The strongertwoor more statusquo statesbecome in relation
to a morepowerfulrevisionist state(s), the more theywilldrawtogetherto
oppose it;theweakertheyget,themore theywilldrawapart.
(3) Revisioniststatesnot immediatelythreatenedbya strongerrevisionist state
or coalitionwillseek to bandwagonwithit and willnot balance againstit.
Revisioniststatesflocktogether.
(4) If an alliance formsfor offensivepurposes, it will operate according to
Riker'ssize-principlehypothesis.Revisioniststatesseek minimumwinning
coalitions.
(5) If an alliance formsfordefensiveor deterrencepurposes,itsmemberswill
seek additionalpartnersbeyondtheminimumrequiredto defeattheexter-
nal threat.Statusquo statesseek large,not minimumwinning,coalitions.
and theSecond WorldWar

The Case Study:Tripolarity
World War II was chosen as the case studyforthe followingreason: applyingmy
definitionof a pole to theCorrelates-of-War
computationof major-power capability
sharesyieldsonlyone instanceof tripolarityin the post-1815period: the 1936-39
system(only partialdata existsfor the years 1940 to 1945). Given the preceding

the case was essentiallychosen for me. Luckily,the late-

discussionof tripolarity,
interwarperiod is intrinsically
interestingand importantbecause it engendered
one of only two worldwars this century.It should also be emphasized that the
analysisexamines the alliance strategiesof all seven major powers (i. e., Britain,
France,Germany,Italy,Japan,the SovietUnion, and the U.S.). Hence, the case of
WorldWar II is actuallysevenseparatecases.
theVariables
Operationalizing
Capabilities.Data fromthe Correlatesof War project is used to measure the
relativecapabilitiesofthe majorpowers.5Though theCOW measuresstressmilitary
forces-in-being, I believe thatthisis appropriateforthe presentanalysis.It willbe
shownthatstatesmen,in choosingtheircourse of action,respondedto the current
balance of power:forexample,by 1937,Hider saw Germany'smilitary advantageas
a wastingassetand so decided to wage a seriesof offensive warsbeforehis window
of opportunity closed; and thoughit is truethatthe U.S., beforeenteringthe war,
could have extractedfarmore resourcesfromitseconomythanitdid, thisprovided
littlecomfortto Britainand Francein 1940 or to theSovietsin 1941.
COW capabilityscores reflectthree distinctmeasures of national power: (1)
military(forces-in-being), (2) industrial(war potential), and (3) demographic
(stayingpower and war-augmenting capability).Each component is divided into
two subcomponents.The militarydimension consistsof the number of military
personnel and militaryexpenditures;the industrialcomponent is measured by
productionof pig iron (pre-1900)or ingotsteel (post-1900)and fuelconsumption;
and the demographiccomponentis dividedinto urban and totalpopulation.The
compositepowerindex is the sum of each state'smean score forthe six measures
as a percentageof all scoreswithinthe Great-Power subset.
As Table 1 indicates,by1938 theinternational system was tripolar,
withGermany,
the United States,and the SovietUnion comprisingthe threepoles. Britainand
Francehad fallenfromthefirsttier,joiningJapanand Italyas middlepowers.
A statewillbe coded "revisionist"

Interests. if (a) it expressedbitterdissatisfaction
withthe territorialchanges,treatyrevisions,or reparationsthatresultedfromthe
lastmajor-power war (whetheritwas on the winningor losingside), or (b) it grew
to fullpowerafterthe new orderwas establishedand complainedthatitsincreased
power entitledit to greaterbenefits(territorial or prestige).All statesthatdo not
fallintoat leastone of thesecategorieswillbe labeled "statusquo."
TABLE 1. COW percentageshare distribution

of Great-Power
capabilities,1938.
USSR U.S. Germany UK Japan France Italy
25.0 22.7 20.2 10.4 9.4 6.9 4.9
Source-Compiled using the "Correlatesof War" capabilitydata-setprintout(December 1987) made available through
ConsortiumforPoliticaland Social Researchat the University
the Inter-University of Michigan.
5The COW capabilityindex providesa reasonablyaccurate pictureof the power bases held by the major actors
with respect to their relativefightingcapabilities. Originallya skeptic myself,I arrivedat this view afterhaving
constructedmyown "capability-index formula"(data available upon request) to testthe validityand reliability
of the
COW numbersfor the period 1938-40. My formulaconsistsof ten separate power indices-eight of whichare not
used in the COW capabilityindex-which attemptto measure the criticalmass (population and territory), economic
war potential,and militarycapabilitiesof the Great Powers.The resultof thismini-experiment was that the Great-
Powercapabilitysharesyieldedbymyindex and the COW index werevirtually identical.

Tripolarity
Applyingthesecriteriato the interwarperiod,Britain,France,and the U.S. won

the lastmajor-power war,establishedthe new order,and thusweresated statusquo
powers.By contrast,Germanyand Russia6were both defeatedin the FirstWorld
War and saw the map of Europe redrawn largelyat their expense. (Austria-
Hungarywas the biggestloser, of course.) Germanyalso sufferedthe loss of its
colonies, huge reparation payments,and severe limitationson the size and
equipment of its armed forces.Italyand Japan, though technicallyvictorsin the
FirstWorldWar,feltso betrayedbythe Versaillespeace settlementthattheycould
not be relied upon to defend the new order. By the 1930s, both states had
substantiallyincreased their militarypower and sought to expand beyond their
present territorialborders. As expected, Rome and Tokyo pursued naked
revisionistaims: Mussolinitriedto create a second Roman Empire,whileImperial
Japan strovefor hegemonyover East Asia (the so-called Co-Prosperity Sphere,
which it announced in 1938). In fact,these two middle powers unleashed the
initialblowsagainstthe statusquo order.
Table 2 convertsthe COW numbersinto ratiosand specifiesthe interestsof the
states:
TABLE 2. The capabilitiesand interestsof the majorpowers,1938.
State Type Power Interest
USSR pole 5.0 revisionist

U.S. pole 4.5 statusquo
Germany pole 4.0 revisionist
Britain middle 2.1 statusquo
Japan middle 1.9 revisionist
France middle 1.4 statusquo
Italy middle 1.0 revisionist
A.W. DePorte's discussionof thisperiod nicelysummarizesthe essence of what

Table 2 shows:
. . . two thingsare sure: first,that both Germanyand Russia were profoundly
anti-statusquo, and second, that the sum of their potential power was much
greater than the sum of the power of those European countries that were
pro-statusquo (assumingthe continuedpoliticalabstentionof the United States).
... We mayconclude, then,thatin the long run the statusquo was bound to be
changed,butwhen,how,and to whoseadvantagewereless certain.(1979:30-31)
Table 2 also suggeststhatwhethera stateis revisionistor statusquo is not an

endogenous functionof the distributionof capabilities,as some realist,Marxist,
and geopoliticaltheoriesposit: twoof threepoles were revisionistand halfof the
second-rankingGreat Powers supported the status quo. Perhaps it is fair to
conclude that for some reason the historyof nations supportswhat the French
characterized by the bon mot "l'appetit vient en mangeant," translated by
Shakespeareas "theappetitegrowsbywhatitfeedson" (Mattern,1942:59).
AllianceDynamicsof theMajor Powers

This section attemptsto explain the alliance behavior of all seven major actors
throughthe use of the twoelementsin Table 2: the capabilitiesand interestsof the
GreatPowers.
5For the Soviet Union as a "revisionist"power, see Mandelbaum (1988:104); Hochman (1984); Henderson
(1940:258); Fischer(1969:349); DePorte (1979:31-32, 40); Carr (1951:123).

Britain
Historiansand politicalscientistshave wonderedwhyBritaindid so littleto help
France against Germanyand allowed Hitler to pursue piecemeal aggression.
Recent interpretationsof the case maintain that Britain perceived defensive
advantagein military technologyand thusattemptedto ride freeon the balancing
effortsof France. In short,Britainpassed the balancing buck to France (Posen,
1984; Christensenand Snyder,1990).
I argue insteadthatunlike 1914,when the statusquo alliance of Britain,France,
and Russia enjoyed a huge resource advantage over the revisionistcoalition of
Germanyand Austria-Hungary, 1938 saw German power alone (4.0) exceed the
combined strengthof the European statusquo states,Britain (2.1) and France
(1.4).7 In thistriadic-although not tripolar-situation(whereA > B + C; A is a
pole, B and C are middle powers),the theorypredictsno coalition,since B and C
cannot combine to defeatA and A does not require a coalitionto defeateitherB
or C. I will attemptto show that,on both theoreticaland empiricalgrounds,this
interpretation is superiorto the morewidelyacceptedview.
The buckpassingargumentis part of a largertheorythatpredictsthe effectof
perceived offensive/defensiveadvantage on multipolar alliance patterns
(Christensenand Snyder,1990). When offenseis perceivedto have the advantage,
the theoryposits,statesbalance aggressively and unconditionally:once one state
goes to war,its allies immediatelyfollow,as if on a chain gang. Conversely,when
defense is perceived to have the advantage,statesattemptto ride free on the
balancing effortsof others: allies pass the buck. The perception of offensive
advantagein 1914 and defensiveadvantagein 1939 is said to explain the difference
betweenthe alliance dynamicspriorto the twoworldwars.
The logic is consistentand convincing,but does the theorysuccessfully explain
theBritishcase? For severalreasons,I thinknot.
First,how does the theoryaccount for the reversalof Britishpolicy after
Munich?Accordingto Christensenand Snyder:
One of the reasons that Chamberlain appeased Hitler at Munich was his
exaggeratedestimateof German strategicbombing capabilitiesand his fear that
Britain'sown retaliatory
capabilitywould not deter attackson Britishcities.After
the Munich crisis,Chamberlainpushed for a reorientationof Britishair power
expendituresfrombombersto fighters. Believingtheseeffortsto be successful,he
concluded bymid-1939thata Germanattackon Britainwould probablyfail.This
allowed him to guarantee Poland withless fear of the immediatecasualtiesthat
thismightproduce. (1990:165)
In Christensenand Snyder'sown words,Britain'sbehaviordirectlycontradicts

theirhypothesis.Perceivedoffensiveadvantage(a Germanknockoutblow) causes
Britainto buckpass,not chaingang,whereasdefensiveadvantage(faithin theirair
defenses) causes it to balance more aggressively,
that is, guarantee Poland and
pledge to stationtroopson Frenchsoil.
My distancinghypothesis,by contrast,predictsBritain'spartialpolicyreversal.
AfterMunich, Anglo-Frenchrearmamentbegan closing the gap in Germany's
militaryadvantage,and Britainsaw itselfand not France as the more immediate
targetof Germanaggression(see Watt,1989:ch.6). As a result,Britaindrewcloser
to France and actuallypledged troopsto a continentaldefense.In directsupport
of thisargument,the Chiefsof Staff(COS) paper "The StrategicPositionof France
in a European War" (February1, 1939) indicatedthat"Britishdefenseswerein the
processof being strengthened," but stressedthat"improvements would be shown
during the next two years" (Wark, 1985:217). Taking pains to contrast the
7IfItalyis added, the equation becomes even more lopsided in favorof the revisionist
coalition.

88 Tripolarity
"expectedbrighterfuturewiththe present,"the COS concluded: "We shallbe in a

positionto defendourselvesat home and at the same timeto affordconsiderably
more assistanceat sea, in the air and on land, than we could today" (Wark,
1985:217). When war came, however,the strengthof Britainand France did not
yetequal thatof Germany.Thus, Britain,havingaccepted a limitedcommitmentto
the Continent,did not entirelyreverseits distancingpolicy regardingFrance,
thoughitmighthave in 1941.
Second, a central theme of the "buckpassing"argumentis that Britishelites
fearedthathighermilitary spending,particularlyfora largerand better-equipped
army,would tempt the French to reduce their own militaryexpenditures.By
passing the costs on to France, Britaincould confidentlybelieve that Germany
wouldbe balanced at minimalcostto itself.
The actual militaryexpendituresof the Great Powers tell a differentstory,
however.In the crucialperiod 1934-38, Germanyincreasedmilitary expenditures
by470 percent,Japan by 455 percent,the Russiansby 370 percent,and Italyby 56
percent. By contrast,the Britishincreased spending by 250 percent, and the
Frenchbya pitiful41 percent(Milward,1977:47). Hence, whileone can argue that
Britain should have spent more on the militaryto keep up with its potential
enemies, it is implausiblethat Britainrefusedto do so because it worriedthat
France would spend less. As TimothyMcKeownpointsout, "increasesin [French]
militaryspendingfrom1934 to 1938 were so anemic thatit is difficult to imagine
them doing less, even had a strongerBritish effortofferedthem a greater
opportunity to freeride" (1991:270).
Third,the buckpassingargumentrelies on the assumptionthatthe defenseof
Britainand France compriseda collectivegood. None of the authors,however,
explicitlymakes this case, which is particularlytroublingin lightof the obvious
differencesbetween British and French securityrequirements. Britain was
threatenedbyan air and sea assault;France bya land invasion.Thus, ifone of the
allieswas able to secureitsbordersagainsta Germanattack,it does not followthat
the other'sfrontiers weretherebydefended.
In fact,both Britishand Frenchleaders believed thatGermanycould launch a
successfulamphibious assault against Britain'shome territory independent of a
GermanoccupationofFrance.DouglasJohnsonobserves:
[Beginning in 1939], the idea began to be put about that Germany was
contemplatinga move againstGreatBritain,perhaps preceded bya move against
Holland and Belgium,ratherthana moveagainstFrance. If thatwas the case then
France mightwell be able to stand by and allow thatto happen. She could then
do a deal withHitler,or Mussolini,or both. (1983:58)
This viewis supported,forinstance,by Harold Nicolson's diaryentryon October

31, 1939, of a discussionhe and otherBritishpoliticianshad withFrenchpolitical
leaders,includingEdouard Daladier and Paul Reynaud:"[The French] thinkthat
we shallsuffer verymuchfromthe air and thatthistimeitwillbeweand notthey who
areinvaded"(Nicolson,1967:45,emphasisadded). Finally,on May9, 1940,one day
before the German attackon the Low Countriesand France, the Chamberlain
cabinetdiscusseda reviewof the strategicsituationpreparedbythe Chiefsof Staff.
The reportpredictedthatGermany's"mostlikelycourse . . . would be to launch a
majoroffensive againstBritain,and the main threatto the UnitedKingdomwas an
intensiveair offensivewhich,ifsuccessful,mightculminatein an attemptat actual
invasion.An enemyoccupationof the Low Countrieswould seriouslyaggravatethis
menace" (Butler,1957:267).
The veryfact that BritainsurvivedafterFrance's defeat suggeststhat Anglo-
French securitylacked the propertyofjointnessrequired of a collectivegood. Of
course, this is not to implythat there was no relationshipbetween Britishand

French securityor thatBritishleaders expected thatthe fallof France would not

affectthe riskof a successfulGermaninvasionof the BritishIsles. It is simplyto say
thatBritain'sfrontiers would have been even less secure had Britaindone more to
bolsterFrance'sdefensesand so less to strengthen itsown.
The non-rivalry propertyof a collectivegood is also suspectin thiscase. Anglo-
French forces assigned to defend one of the allies' frontierscould not
simultaneously be used forthe defenseof theother's.As Wallace Thies observes:
Because defensiveforces generallycould not be transferredquicklyand easily
fromone allyto another,alliance membersdid not viewthe forcesof theirallies
as substitutesfortheirown.The Frenchmighttake comfortin the knowledgethat
the Britishwould almost certainlytake their side in the event of war with
Germany,but thisdid not relieve them of the requirementto build an armyas
large or nearlyas large as Germany's;the Britishnavymightrule the seas but it
would be of littlehelp in shieldingFrance againsta Germanattack.(1987:323)
And untilthe Britishhad secured theirown borders,theyhad no intentionof

shieldingFrance againsta German attack.To do so would be to make Britaina
more invitingtargetthan France. Thus, the BritishChiefs of Staff,convinced
preciselyof this,predictedin theirlastprewarstrategicappreciation(February20,
1939) that a German attackwould be concentratedagainst Britainratherthan
France because "the vulnerability of the islands to combined air and sea attack
would make [Britain]a temptingtarget"(Wark,1985:215). The COS gave four
reasonsforexpectingthe initialGermanassaultto be directedagainstBritain:
France could not resistalone afterthe defeatof Great Britain;a full-scaleattack
on France would involve heavy losses which might have a serious effecton
German morale; Britishsupportto France under attackwas likelyto prove more
effectivethan French supportto Great Britain;and lastly,by use of her air force
alone againstthiscountryGermanywould exploither relatively strongestweapon
withthe leastexpenditureof lifeand economic resources.(Butler,1957:172)
Believingthatitwouldbe attackedfirst,Britainmade everyeffort duringthefinal

weeks of the twilightwar to hasten the productionof anti-aircraft equipment,
especiallyBoforsguns, bomber and fighteraircraft,and fullytrainedcrews (see
Butler,1957:172). The outcomeoftheBattleofBritainconfirms theChiefsof Staffs
beliefthat,forBritain,the "cruxof thematteris air superiority"
(Barnett,1972:583).
Even ifthe Britishhad understoodhow quicklyFrance would fall,it is farfrom
clear thattheywould have,could have,or should have acted otherwise.According
toJ.L. Richardson,"evenifmore adequate resourceshad been available,theArmy
would probablynot have adopted thekindof strategy and equipmentlikelyto have
made a crucialdifferenceto the campaignin France in 1940" (1988:303). Indeed,
manyBritishelitesexpressedreliefoverFrance'sdefeat,as Eleanor Gatesrecords:
Others quite openlyviewedFrance as a heavyburden,whichit was necessaryto
slough off, and gave full vent to their feelings after the French defeat.
Chamberlain. . . spoke withreliefof being "at anyrate freeof our obligationsto
the French,who have been nothingbut a liabilityto us," even going so faras to
claim: "It would have been farbetterif they[the French] had been neutralfrom
the beginning.... Lord Hankey... also feltthatFrance had been "a debitrather
than an asset in the presentwar"and found it "almosta reliefto be thrownback
on the resourcesof the Empire and America.". . . AirMarshalDowding,who had
jealously guarded his fightersquadrons throughoutthe Battle of France, went
farthestof all. At the end of thisagony,he actuallysaid to Halifax: "I don't mind
tellingyou thatwhenI heard of the Frenchcollapse I wentdown on mykneesand
thankedGod." (Gates, 1981:566)
To manyBritishelites,the riseof air powerhad made France and the Continent

less importantforBritishsecurity-justas, yearslater,ICBMs and mutualassured

Tripolarity
destructionmade the balance of power in Europe largelyirrelevantfor U.S.

security.Changes in the Air Ministry'sprograms that obviated offensiveand
defensivedeploymentof the RAF in the Low Countriesbolsteredthisbeliefin the
of the Continent,as Michael Howardobserves:
expendability
The Royal Air Force was [in autumn 1937] developing weapons-systems which
were beginningto make a footholdon the continentof Europe, fromtheirpoint
of view, expendable. Consciouslyor unconsciouslythis may have affectedthe
attitudeof theAirStaffto the continentalcommitment.(1972:117)
Resourcesdevotedto the RAF were "believedto increaseBritishcapabilityboth

autonomouslyto deter a German attackon Britainand to deal withthat attack
should deterrencefail" (Posen, 1984:171). Even Churchillcalled Britain'sfighter
squadrons"in effectour MaginotLine" (Butler,1957:184)-a statementthatsheds
considerabledoubt on Baldwin'sclaim in 1934 thatBritain'sfrontierslay on the
Rhine.
Finally,accordingto the buckpassinginterpretation, Britainattemptedto ride
free on the strengthof France's balancing efforts.By contrast,the "distancing"
hypothesispositsthatalliances will not formamong statusquo stateswhen their
combined strengthis lessthan the opposing revisioniststate(s). Accordingto this
view,the main reason Hitlerwas able to pursue piecemeal aggressionunopposed
was the power asymmetryin his favor,not Britain's perception of France's
defensivestrength.
The empirical record supports the distancinghypothesis.Contraryto the
buckpassingview,Chamberlainnever considered relyingon the effortsof other
SidneyAstercomments:
statesto provideforBritain'ssecurity.
[Chamberlain]held France and itsstatesmenin near contempt.... Of particular
concern were French economic and industrial troubles which hindered its
rearmamentprogrammes.His attitudeto the United States . . . was that "it is
alwaysbest and safestto count on nothing
fromthe Americansexcept words.". . .
[He] had no faith in Soviet militarycapabilities and considered the Russians
untrustworthyas a potentialally.... Consequently,Britaincould onlylook to its
own resourcesforprotection,havingno reliableallies. (1989:242-243)
Upon succeeding Baldwin as Prime Ministeron May 28, 1937, Chamberlain

formulateda double policyof deterrenceand appeasement,whichwas designedto
provideforBritain'ssecurity
withoutrelyingon allies:
His over-ridingprinciple... was thatBritain'sbest defence policy"wouldbe the
existence of a deterrentforce so powerfulas to render success in attack too
doubtfulto be worthwhile." As he wroteon 9 February1936, in practicalterms
thismeant "our resourceswillbe more profitably employedin the air,and on the
sea, than in building up great armies."Nothing,includingthe outbreakof war
withGermanyin 1939, everchanged his viewon thissubject. (Aster,1989:247)
Seeing onlyweak and unreliableallies,Chamberlainreacted to the Czech crisis

by "dismiss[ing]as impracticalthe Churchillianalternativeof a 'grand alliance'
againstthe dictators."SupportforCzechoslovakia,Chamberlaindeclared, "would
simplybe a pretextforgoing to war withGermany. .. [and] thatwe could not
thinkof unlesswe had a reasonableprospectof being able to beat her to her knees
in a reasonable time,and of thatI see no sign" (Aster,1989:246). In otherwords,
Britaincould notjoin or put togethera winningcoalition,and so, as the theory
predicts, no coalition formed. Rejecting the militaryoption, Chamberlain
redoubled his effortsto mediate possible areas of conflictwithGermany-at least
untiltheBritishrearmamenteffort had begun to payoff.
Even afterGerman troops occupied Prague on March 15, 1939, Chamberlain
persistedin his attemptto conciliateHitler.On March 19 he wrote"I neveraccept
theviewthatwaris inevitable"(Aster,1989:252-253). A week laterhe expanded on
hisview:

I see nothingforus to do unless we are prepared ourselvesto hand Germanyan

ultimatum.We are not strong enough ourselves and we cannot command
sufficientstrengthelsewhereto present Germanywith an overwhelmingforce.
Our ultimatumwould thereforemean war and I would never be responsiblefor
presentingit. (Aster,1989:253)
As noted, the German move did provokea limitedBritishcommitmentto the

Continent. Given the woeful state of the TerritorialArmy,however,Britain's
pledge was merelya symbolicgesturedesigned to encourage France to fightand
not to remainneutralor, worse,side withthe enemyif GermanyattackedBritain
directly(Watt, 1989:94, 102). Barely equipped for survival,Britain was in no
positionto help itsallies.
In sum,duringthe 1930s the combined strengthof Britainand France was less
than thatof Germany,and so no coalitionformed.Instead,Britaindistanceditself
fromFrance and adopted a dual policyof deterrenceand appeasement.As the
rearmamentprogramsof Britainand France began closingthe gap in Germany's
militarylead, Britaindrewcloser to France and began to standup to Hitler.That
the gap was neverentirelyclosed explainsBritain'snot doing more to aid France.
France
Beforethe FirstWorldWar and the Bolshevikrevolution,Francewas able to enlist
Russia (then a statusquo power) in an anti-Germanalliance that,given the size
and strengthof the alliance, allowed both countries to retain their offensive
militarydoctrines. After 1919, France's inferiordemographic and industrial
potentialrelativeto Germany,coupled withthe loss of Russia (now a revisionist
power) as an ally,resultedin French dependence on allied supportand a strictly
defensivemilitary posture(Hughes, 1971; Challener,1955:ch.3).
Thus, throughoutthe interwar period,Francetriedto constructa massivedefen-
sive alliance systemcenteringon the LittleEntente,the League of Nations,and a
troubledpartnership withBritain.As the theorypredicts,a statusquo middlepower
directlyconfrontedby a much strongerrevisionistpole will seek a large alliance
(not a minimumwinningcoalition)to deteror defendagainstthethreatening state.
But one mightask why,givenFrance's desirefora large coalitionand Britain's
reluctanceto join, did Paris not push harder to secure a formalmilitaryalliance
withRussia,itstraditionalallyagainstGermany?There are threeprincipalreasons.
First,French and Britishmilitaryexpertsagreed that the Red Armywould be a
formidableopponent to anyinvaderbut,decapitatedbyStalin'spurges,would not
be capable of mountingan offensivecampaign (Herndon, 1983). Said Neville
Chamberlainof the SovietUnion in March 1939: "I have no beliefwhateverin her
abilityto maintainan effective even ifshe wantedto" (Feiling,1947:403).
offensive,
The West'sassessmentof Sovietcapabilitiesprovedremarkably accurate.
Second, Anglo-FrenchelitesdeeplydistrustedStalinin particular,who theycon-
sidered bloodthirsty and opportunistic,and the Soviet Union in general,which
theyregarded as a power-seeking, revisioniststate.The French premierEdouard
Daladier voiced theseconcerns-the defensivenatureof the Red Armyand distrust
of the SovietUnion-in a March9, 1939,conversationwiththe PolishAmbassador
in France,JuljuszLukasiewicz.The U.S. Ambassadorin France,WilliamC. Bullitt,
noted:
Both Daladier and the Polish Ambassadorwere of the opinion that the Soviet
Union was to be counted on for nothing.They both feltthatit was certainthat
internalconditionsin the SovietUnion would preventthe Red Armyfromtaking
any activepart in any war anywhereand both agreed that no reliance could be
based on any promisesof Soviet support in the formof supplies to Poland or
Rumania. Both agreed also thatifHitlershould be willingit would not take a half
hour to form an alliance between Germanyand the Soviet Union. Stalin was
pantingforsuch an agreement.(Foreign RelationsoftheUnitedStates1939, 1956:30)

Tripolarity
In short, Daladier believed that the USSR would not honor its alliance
commitmentsand, even if it did, the Red Armywas in no condition to open a
second frontshouldwarbreakout in theWest.
Finally,Britain,upon whom France depended, pressuredthe French not to
alignwiththe SovietUnion because "itwas farmore likelythatthe Germanswould
move againstthe Soviet Union, and thereforemuch more probable thatFrance
and consequentlyBritainwould be drawn into conflictto aid the Soviet Union
than thatshe would have to call upon SovietassistanceagainstGermanaggression
in the West" (Wolfers,1940:308). Most French observersagreed, viewing an
alliance withRussia as more of a liabilitythan an asset,and adding thatit would
alienate Poland, Rumania, and other membersof the Little Entente. For these
reasons, Paris and London applied only weak pressureon Colonel Beck to get
Poland to grant permission for the passage of Soviet troops-Moscow's
precondition for continuing militaryconversationswith France and Britain
(August1939).
Despite these reasons not to seek an alliance with Soviet Russia, however,
Daladier believed war could not be avertedwithoutthe threatof the Red Army
enteringon the side of the statusquo powers.In a May 16, 1939, telegramto the
U.S. Secretaryof State,Bullittwrote,"[Daladier] believed that it was essentialto
have Russia in the combination.Only thuscould a sufficient combinationof force
be built up to deter Hitler fromriskingwar" (ForeignRelations of the UnitedStates
1939, 1956:255). Daladier's thoughtsare consistentwithmyhypothesisthatstatus
quo statesseek verylarge alliances (not minimumwinningcoalitions) for the
purposesof deterringaggression.
In sum,the Frenchwere tornbetweentwocontradictory impulses.On the one
hand, theyfeared that aligningwithrevisionistRussia would jeopardize French
securitybyriskingthe unravelingof the statusquo order and alienatingthe Little
Ententeand Britainin the process. (Similarfactorsexplain theweak Frencheffort
to gain an alliance withItaly;see Parker,1974.) On the other hand, the French,
wantingto constructas largea coalitionas possible,fearedthatHitlercould not be
deterredwithoutRussian membershipin the countervailingstatusquo alliance.
Both French concerns,thoughpullingin different directions,are consistentwith
the predictionsof the model. In the end, these conflictingimpulses paralyzed
French policy vis-a-visthe Soviet Union. This resulted in the Nazi-Sovietnon-
aggressionpact and France'sswiftdefeatbyGermanyin 1940.
Gemany
and discussionsof the balance of powersuggestthathe viewedthe
Hitler'sstrategy
worldas a tripolarsystemcomposed of Germany,the United States,and the Soviet
witheringunder the shadowscastby
Union. The tripolarimage of Europe's vitality
the growthin Americanand RussianpowersurfacesthroughoutHitler'swritings.
Of the United States,Hitlerwrote:"Withthe AmericanUnion, a new powerof
such dimensionshas come into being as threatensto upset the whole former
powerand ordersofrankof the states"(1928:83). Continuingon the theme,Hitler
remarked,"SincetodayGermany'seconomic fatevis-a-vis Americais in factalso the
fate of other nations in Europe, there is again a movement . . . to oppose a
European union to theAmericanUnion in ordertherebyto preventa threatening
worldhegemonyof theNorthAmericancontinent"(Hitler,1928:103).
Hitler's beliefs about the American threat were supported by his leading
geopoliticalstrategist,Major General Dr. Karl Haushofer,who declared in 1938:
"Potentially,the United States is the world's foremostpolitical and economic
power,predestinedto dominatetheworldonce itputsitsheartintopowerpolitics"
(Strausz-Hupe,1942:67).

In fact,of all the GreatPowers,Hitlermostfearedthepotentialmilitary strength

of the UnitedStates.As earlyas 1937,Hitlerand G6ring,the head of Germany'sair
force,had authorizedthedevelopmentof bombersto strikeat New Yorkand other
East Coast cities;and, as GerhardWeinbergputs it, "ifnothingmuch eventually
came of these projects,it was not forlack of trying"(1981:xiii). ByJanuary1939,
Hitler'sstatementsindicatean obsessionwithAmerica'sshadowon the European
scene. "Fromthattimeon,"John Lukacs asserts,"he began to considerRoosevelt
as his principalenemy-a convictionthatHitlerheld to the end" (1989:172). Of
the Americandanger to Germany,Hitlercommented,"confrontedwithAmerica,
thebestwe can do is to hold out againsther to theend" (Trevor-Roper, 1953:199).
Prior to engagingin the inevitablebattle againstthe United States,Germany
would firsthave to defeatthe SovietUnion and gain continentalhegemony.This
would not be an easytask,as Hitler'scommentsin 1941 illustrate:
The more we see of conditionsin Russia,the more thankfulwe mustbe thatwe
struckin time.In anotherten yearstherewould have sprungup in Russia a mass
of industrial centres, inaccessible to attack, which would have produced
armamentson an inexhaustible scale, while the rest of Europe would have
degenerated into a defenceless plaything of Soviet policy. (Trevor-Roper,
1953:586-587)
Hitler'sstrategyforaccomplishingGermanworld dominationwas a Schlieffen

in reverse.Afterthe Anschluss, German forceswould wage a lightningstrikeat
Czechoslovakia to neutralize that power and acquire the additional strategic
resourcesand manpower(35 Czech divisions)beforeswitching the armywestward.
Germanywould thenpacifyitswesternflankbyconqueringBelgium,Holland, and
France. In the process,Germanywould keep its war economy afloatby gaining
directcontrolover the rawmaterials,food supplies,and labor reservesof Western
Europe (Carr, 1978:87). Seeing no hope for victory,Britainwould then sue for
peace and join in the titanicwar againstthe Soviet Union, the success of which
would determinethe fateof Hitler'sultimategoal of Lebensraum-living space in
the SovietUkraine.Under Hitler'sgranddesign,"WhatIndia
the East,particularly
wasforEngland,the territories of Russiawillbe forus" (Trevor-Roper,
1953:24).
Withthe resourcesof the entireEurasianlandmass,Germanywould be able to
defeattheUnitedStates(the thirdpole). Weinbergexplains:
[T]he Americanswere the real threatto German predominance in the world.
Hitler's deduction fromthis analysiswas simple: only a Eurasian empire under
German domination could successfullycope withthis menace. A thirdwar was
now added to the originaltwo.Afterthe firsttwowarshad enabled it to construct
a continentalempire fromthe Atlanticto the Urals,Germanywould take on the
United States.One of the major tasksto be performedby the National Socialist
movement,therefore,must be the preparation of Germanyfor this conflict.
(1964:1009)
Emphasizingdemographics,HitlerviewedGerman continentalhegemonyas a
prerequisitefortheultimatewarwiththe U.S.:
It is ridiculous to thinkof a world policy as long as one does not control the
Continent.... A hundred and thirtymillionpeople in the Reich, ninetyin the
Ukraine. Add to these the other States of the New Europe, and we'll be four
hundred millions,compared with the hundred and thirtymillion Americans.
(Trevor-Roper, 1953:93)
Accordingto the Fiuhrer,the BritishEmpirewas also endangeredbyAmerica's

inexorable economic, military, and naval growth:"[Roosevelt]wants to run the
worldand rob us all of a place in the sun. He sayshe wantsto save England but he
means he wantsto be ruler and heir of the BritishEmpire" (Toland, 1976:693).

94 Trapolarity
Given his misguidednotion of an underlyingAnglo-American rivalry,the Reich

Chancellorconfidently expected Britainto join a German-ledEuropean coalition
againsttheNorthAmericancontinent.
Butjust in case the "thick-headed"Britishfailedto see wheretheirtrueinterests
lay, Hitler went to great lengthsto convince Britainthat Germanysought only
freedomof action on the Continentand had no intentionof threateningBritain's
naval supremacyor Empire. In the Anglo-GermanNaval Treaty(1935), whereby
Hitler conceded a proportionalnaval strengthof 35:100, "Germanytook the
initiativein the negotiationsin the spiritof makinga gesturein Great Britain's
interest;the German proposals themselveswere clearly aimed at reassuring
[Britain]on the questionof Germannavalrivalry" (Hinsley,1951:6-7).
Havingmade severalgenerousoffersto gain an Anglo-Germanalliance without
success,the Fuihrergrewdisillusionedwiththe British.He could not understand
whythe Britishrefusedto see thatGermany,unlikeFrance and the United States,
soughtonlyEuropean hegemonyand would gladlyallow Britainto maintainher
worldambitions.How could the BritishresistGermanfriendship, he wondered,in
the face of the obviouscommon threatsposed byAmericaand the Bolshevikmen-
ace? Was it not obviousthatGermany'ssecurityhad been severelycompromisedby
the Franco-Russianalliance and that Russia was, in Hitler's words, "now the
greatestpower in the whole of Europe"? (Robertson,1963:54). Surely,thismade
an Anglo-Germancombinationthe mostnaturalof alliances (Hitler,1925:181-185,
892-965; Toland, 1976:536-537,614-616, 692-694; Calleo, 1978:95-115).
Unable to gain an Anglo-Germanalliance,Hitlerturnedto the SovietUnion. In
accordance with the prior discussionof tripolarity with two revisionistpoles, to
overturnthe tripolarsystemHitler needed to augment German resources by
waginga seriesof offensive warsagainstsmall and middle powers.This depended
on cooperationwiththe revisionist pole, the USSR. At the veryleast,Hitlerhad to
preventthe formationof a hostileAnglo-Franco-Russian coalition,as had occurred
in 1914. Withoutthe Nazi-Sovietpact, Germanycould not have attackedPoland
and France.
Aftertakingcare of France and Britain,Germanywould turnaround and attack
itspenultimatetarget,the SovietUnion. That Hitlerplanned thisstrategy priorto
signing the Nazi-Soviet pact is made clear in his statement to the High
Commissionerof the United Nations,Burckhardt,on August 11, 1939, in which
the Fuihrerdecried what he believed was the West's stupidityin not coming to
termswithGermany:"Everything thatI undertakeis directedagainstRussia;ifthe
Westis too stupidand too blind to understandthis,then I willbe forcedto reach
an understandingwiththe Russians,smashtheWest,and thenturnall myconcen-
trated strengthagainst the Soviet Union" (Hillgruber,1981:69). The prophesy
about Russiashowsthatthe Sovietpact did not signala change in Hitler'sultimate
goal to destroythe SovietUnion; rather,itwas nothingmore thanan expedientto
avoid a two-front war-a necessaryshort-term act ofpure powerpolitics.
Afterthe defeat of France, Hitler again felt optimisticabout his chances of
negotiatingan end to the warwithBritain.Together,Britainand Germanywould
firsteliminatethe SovietUnion and then turntheirenergiesagainsttheAmerican
continent.Of the lattertask,Hitlersaid: "I rejoice on behalfof the Germanpeople
at the idea that one day we will see England and Germanymarchingtogether
againstAmerica"(Trevor-Roper, 1953:26).
Backed by U.S. materialassistance,however,Britainchose to fightratherthan
jump on the Nazi bandwagon.London's decisionforwarwas eased byprogressin
its civil and air defense systemswhich greatlymoderated the public's fear of a
German"knockoutblow."
Hitler'searliernaval concessionsprovedcostlyin the Battleof Britain;as late as
September1939, the German Navywas woefullyunpreparedforwar withBritain

(Hinsley,1951:6-7). Then, in early1940, the Germanfleetsufferedheavylosses in

the Norwegian campaign, which seriouslyimpaired its abilityto support an
invasion of Britainfour monthslater. Confrontedby the prospect of a war of
attritionwithBritainthatriskedprematurelybringingthe U.S. into the conflict,
Hitler decided that the bettergamble was to turn Germany'sattentionto the
defeatof his primarytarget,the SovietUnion. If Russia could be knocked out of
the war, Britainwould lose its last potential continentalally and would, Hitler
believed, sue for peace. More important,in controlof the European continent,
Germanycould thenwage thefinalhegemonicwaragainstNorthAmerica.
But unlike in the Battle of France, Germany'sblitzkriegstrategy was foiled by
the vastexpanse of the SovietUnion, whichallowed the Red Armyto trade space
for time. He mighthave won his gamble and defeated Russia in 1941, Hitler
thought,ifnot forMussolini's"idioticintervention" in Greece (Calleo, 1978:108).
RescuingItalycost Germanyprecious timeand resourcesneeded forthe Russian
campaign.
The failureof the Wehrmacht's great offensivein 1941 gave the U.S. time to
mobilize its enormouswar machine and join the SovietUnion in an overwhelm-
inglypowerfultwo-against-one winningcoalition. "As Hitler's observationsimply,
once the United States and Russia gathered and combined theirvast resources,
Germany'sfatewould be sealed" (Calleo, 1978:108). Had Hitler'svictorycome in a
hurry,the tripolarsystemwould have been transformed into a bipolar one, pitting
the Eurasian continentagainst the weakerAmerican continent(Hitler believed
thatthe U.S. would annex Canada). The finalwarwould then have convertedthe
internationalsystemfrombipolarityto hegemonyunder Germanrule.
Luckily,the Wehrmacht stalledin 1941 and the U.S. and the SovietUnion went
on to defeatand partitionGermany.This transformed the volatiletripolarsystem
into a stable bipolar one and finallyeliminatedthe "GermanProblem"thathad
caused twoworldwarsin twenty-five years.
Italy
Dissatisfiedwithitspowerpositionand temptedby the prospectof gains in North
Africaand CentralEurope, Italyreadilyjumpedon the Germanbandwagon,as the
theorypredicts.Britainand France,who wishedto defend,not destroy,the status
quo, were reluctantto align themselveswitha revisionist
state.For thisreason,the
Westernpowers could not match Germany'sconcessions to Italy. Says Arnold
Wolfersof Franco-Italianrelationsbetweenthe twowars:
. .. the decisiveobstacle to co-operationbetweenFrance and Italylay in the fact
thatItaly,unlikePoland or the LittleEntente,was a dissatisfiedcountryand could
not be attractedto France by mere guaranteesof the establishedorder. She was
out forchange,not enforcementof the statusquo,and manyof the changeswhich
she desired could be effectedonlyby far-reaching French concessions.. . . [But]
more was involvedfor France than this or that concession, this or that naval
agreement, this or that cession of colonial rights or territory.The whole
conception of the preservationof the statusquo could not be harmonizedwith
Italian or Fascist"dynamism" drivingforgreaterpower.... (1940:143-144)
The Spanish CivilWar and the Anglo-French-inspired League sanctionsagainst

Italy in response to the Ethiopian War drove Mussolini firmlyinto Germany's
embrace. The Duce believed that,althoughbandwagoningwithGermanymeant
satellitestatusforItaly,aligningwithFrance and Britainofferedstillless room for
Italian maneuvering.On this point, Denis Mack Smith writes,"As one of his
Ambassadorssaid, betterbe number two with Germanythan a bad thirdafter
Franceand Britain;onlywithGermanycould he challengethe dominantpowersin
the Mediterraneanand break out of what he called Italy's'imprisonment'in the

96 Tripolarity
inland sea" (1983:260). Moreover,HitlerrecognizedthatGermany(weight= 4.0)

could not defeata combinationof Britain(2.1), France (1.4), and Italy(1.0). Italy
thusenjoyeda strongbargainingposition,affordedby-its kingmakerrole.
Seeking to maximize Germany'sshare of the spoils, however,Hitler did not
desire activeItalianparticipationagainstBritainor France (Germanycould defeat
themwithouthelp). Germanysignedthe Pact of Steel mainlyto preventan Anglo-
Franco-Italianalliance. Hitler also reckoned that the threat alone of Italian
interventionwould tie down Western forces in southern France. Perhaps
unwittingly, Hitlermade it easyforItalyto assume itstraditionalrole as thejackal,
trailingthe lion (Germany,in thiscase) to scavengethe scrapsitleavesbehind.
Japan
The hypothesesthatrevisioniststatesflocktogetherand seek minimumwinning
coalitions are consistent with Japan's alliance policy. By 1940, Japan had
major powers(Italy,Germany,and the USSR).
establishedtiesto all the revisionist
But, as predictedby our hypothesisthatrevisioniststatesformminimumwinning
coalitions,Tokyo did not seek activesupportfromany of themuntilwar withthe
Anglo-Americancoalition appeared inevitable.The Japanese then enlisted only
Germanparticipation,believingtheywere forminga minimumwinningcoalition.
In so doing,however,theyoverestimatedGerman-Japanese strength(accordingto
myfigures,4.0 + 1.9) relativeto thatof the Anglo-American coalition (2.1 + 4.5).
But the principalfailureofJapan'sstrategy was itsfalsepresumptionthatGermany
would soon sign a negotiatedpeace withthe SovietUnion, freeingNazi forcesfor
thewaragainstAmericaand Britain.
TheSovietUnion
In his speech at the plenarysession of the Central Committeein January1925,
thatwould guide Sovietforeignpolicyin
Stalinrevealedthe tertiusgaudens strategy
the nextworldwar:
... ifwarbegins,we shallhardlyhaveto sitwithfoldedarms.We shallhaveto
comeout,butwe oughtto be thelastto comeout.Andwe shouldcomeoutin
orderto throwthedecisiveweighton thescales,theweightthatshouldtiltthe
scales.(Deutscher,
1949:411)
In orderforthe SovietUnion to gain the enviablepositionof the enjoyingthird,
Stalin had firstto preventthe formationof two potential hostile coalitions:an
Anglo-Franco-Germancombination and a German-Japanesealliance. Since
Germanywas the onlycommon memberof the twofearedcoalitions,the solution
to the Soviet securityproblem was a rapprochementwith Berlin. And because
Stalin did not believe thatBritainand France would come to Russia's aid in the
eventof a Germanattackin the East, an alliance withthe dreaded Nazis in effect
offeredMoscowitsonlyand thereforeminimumwinningcoalition.
The "power"side of the model thereforepredictsa Nazi-Sovietalliance,but only
one in which the Sovietsactivelyparticipatein the fightingand are assured of
receivingalong withGermanyan equal shareof the spoils. (See discussionabove of
equilateraltripolarsystemwithtworevisionist poles.) Byjoining Germanyto divide
Poland, Stalinbehaved accordingto the "PartitionedThird"versionof tripolarity.
True, Poland was not a pole; but the powerconfiguration of Poland, Germany,and
the SovietUnion was A < B = C, and so the logic of whatI call a "Type3" tripolar
systemstill applies. Conversely,by encouragingGermanyto attackwestwardby
itself,Stalin acted contraryto the logic of the "power"side of the model, and
Russiapaid the priceforhis mistake.

The "interests"side of the model also predictsa Nazi-Sovietalliance.Specifically,

the pact supportsthe hypothesisthatrevisionist statesflocktogether.For Russia,a
rapprochementwithGermanyofferedmore than security:Soviet expansion was
possibleonlyin collusionwithHitler.In the end, the twodictators'sharedinterest
in revisionismbridgedtheirideologicaldivide.As Louis Fischercomments,ideology
"provedno barrierwhenHitlerwantedwarand Stalincovetedterritory" (1969:322).
Stalin's own explanationfor the Nazi-Sovietpact (given in 1940 to the British
Ambassador,Sir StaffordCripps) testifiesto the strengthof shared interests,
whetherrevisionistor status quo, in deciding who aligns with whom: "[The]
U.S.S.R. had wantedto change the old equilibrium... England and France had
wanted to preserve it. Germanyhad also wanted to make a change in the
equilibrium,and thiscommon desireto get rid of the old equilibriumhad created
the basis forthe rapprochementwithGermany"(Weinberg,1981:7). Similarly, Sir
Nevile Henderson viewed the Nazi-Sovietpact as the inevitableoutcome of a
concession-making competitionbetween statusquo and revisionistsuitorsof an
unsatiatedpower:
[ByAugust11, 1939] Moscow... was askingfora freehand in the Baltic States.
Russia'sreal objectivewas thusbecomingapparent;and, withGermanysecretlyin
the market,the scales were being heavilyweightedagainst the WesternPowers.
They could not barteraway the honor and freedom of small but independent
countries,but Germanycould. (Henderson, 1940:259)
In a limitedsense,Stalin,like Hitler,conceivedof the systemin tripolarterms.

But unlikeHitler,who saw the European systemas bipolarwiththe U.S. out of the
picture, Stalin believed that Britain and France together constituteda third
European pole-one that could effectively balance Germany. Stalin's "major
blunder"-as Isaac Deutschercorrectlypointsout-was that "he expected Britain
and Franceto hold theirgroundagainstGermanyfora long time;... he overrated
France's militarystrength; and he underrated Germany's strikingpower"
(1949:441; also see Ulam, 1974:227-229). Consistentwiththese observations,in
1939 Stalinstatedthat"peaceful,democraticstates... are withoutdoubt stronger
thanthe fascistones both militarilyand economically"(Ulam, 1974:264).
But givenhis misperceptionof the powerdistribution, it is easyto see whyStalin
made a deal withHitler.A non-aggressionpact withGermanywould destroythe
statusquo, affordeasyspoilsin EasternEurope and Finland,and instigatea war of
attritionamong the capitalistpowers.Betterstill,the SovietUnion enjoyedthe role
of kingmaker,as both Germanyand the democraticpowersneeded it to forma
winning coalition. Thus, Germanywould be made to pay heavilyfor Soviet
assistance in a war from which Russia could safelyabstain. Stalin used his
bargaining power to prolong the Anglo-Franco-Soviet negotiationsjust long
enough to extractadditionalconcessionsfromHitler,in exchange forwhich the
Soviet leader put his signatureto the pact with Germanythat he desperately
wanted anyway.It appeared that Stalin had succeeded in the role of abettor:
Germanywas deflectedto the west,while the Sovietscomfortably looked on from
the sidelines,gainingat the others'expense.
But contraryto appearances, Hitler duped Stalin. With the September 1
deadline forthe Polish attackonlya week away,Hitlerwaswillingto givethe Soviet
dictatorwhateverhe wanted to secure Sovietneutrality and therebydeter Britain
from honoring its pledge to Poland. Indeed, the Ffihrerseemed to be in an
especiallygivingmood, confidentthat he would attackand crush Russia in the
near future,takingback all he had given and more (see Rich, 1973:chs. 14-18;
Leach, 1973).
In retrospect,we knowthatitwould have been farbetterforthe Sovietsto have
balanced against,ratherthan bandwagoned with,Germany.In that case, Stalin

Tripolarity
would have presented Hitler with the prospect of a two-front war, seriously
underminingthe Fiuhrer'sstrategyand perhaps causing its abandonment. But
because he mistakenlybelieved that Europe was structurally a tripolar,not a
bipolar,systemwithFrance and Britainas the thirdpole, Stalinexpected a war of
attritionin the West. The fall of France abruptlyended Stalin's dream of easy
conquestsin a postwarperiod when the restof Europe would be exhausted.But
when the long war failed to occur, whydid not Stalin immediately join Britain
againstGermany?The answerto thisquestionrevealswhythe model of equilateral
withtwo revisionistpoles could not have predictedthe Sovietcase. As
tripolarity
the historianJames McSherrypoints out, the intensityof Bolshevik ideology
overrodetripolarsystemiclogic:
Withthreeor moreapproximately equal states,a balanceof poweroperates
almostautomatically. Should stateA appear to be growingtoo powerful and
dangerous, statesB and C combineagainstit.IfstateB becomestoo powerful, A
and C forman alliance.Tsarsand foreign ministers in St.Petersburgreactedin
thisclassicalpatternalmostinstinctively....But theBolsheviks sawthemselves
winning everythingor nothing. Theyperceived onlytwopowers:theSovietUnion
and an implacably world.Aslongas thecapitalists
hostilecapitalist didn'tunitein
a crusadeagainsttheU.S.S.R.,whatmatterifone capitalist statebecamemore
powerful thanthe othersor evenbroughtsome of the othersunderitssway?
Once Hitlerhad conqueredFrance,theSovietUnionwas in mortalperil.But
StalinrealizedthefullextentofthedangeronlyonJune22,1941.(1970:254)
TheUnitedStates
As the lone statusquo pole in an equilateraltripolarsystem,the United Stateshad
to guard against the formationof a hostile two-against-one coalition. Distanced
fromthe European fray,the United Stateshad timeto watcheventsunfoldand to
change roles-the eyewitness,the mediator,tertiusgaudens, and the abettor-to
suit the situation.At a high level of abstraction,the theoryaccuratelypredicts
Americanforeignpolicy.The U.S. enteredthewarwhen one of the tworevisionist
poles appeared to be triumphing overthe other.Priorto thattime,Americastayed
on the sidelinesand providedarms and economic assistanceto Britainand later
the SovietUnion.
Though America'sentryinto the war is usuallyattributedto theJapaneseattack
on Pearl Harbor, the AtlanticCharter-arrangedprior to theJapanese attackon
the U.S. but afterthe Germanattackon the USSR-and the "Europe-first" strategy
indicatethattheRooseveltadministration was preparingforwarand was principally
concernedwithdefeatingGermany.Essentially, Roosevelt's"shoot-on-sight" speech
on September11, 1941,whichput intopracticethepolicyof convoyingBritishships
halfwayacross the ocean, committedthe U.S. to the Battleof the Atlanticagainst
Germany.Roosevelt'smid-Octoberdecision to revisethe Neutrality Act, enabling
Americanmerchantmento carrysuppliesacrosstheAtlanticto Britishports,would
have resultedin warwithGermanyin a matterof months.True, Hitlerdesperately
wantedto avoid warwiththe United Stateswhilethe Wehrmacht was sluggingit out
againsttheRed Army,but theReich Chancellorcould not standbyidlyas American
shipscarrieda majorportionoflend-leasesuppliesto Britain.This would havebeen
tantamountto Germany'sgivingup the Battleof theAtlantic-a concessionbeyond
the limitof Hitler's forbearancetowardthe United States. "And once Germany
began sinkingAmericanships regularly," RobertDivine opines, "Rooseveltwould
have had to ask Congressfora declarationofwar"(1969:46).
The overwhelmingly powerfulU.S.-Soviet coalition ultimatelydefeated and
divided Germany,transforming the unstable tripolarsysteminto a stable bipolar
one. Moreover,the UnitedNationscoalitionwas a defensivealliance,and so, as the

theorypredicts,no effortwas made to limitparticipationto a minimumwinning

coalition.
ImplicationsforthePost-Cold WarWorld
Over the course of the last twoyearsthe unprecedentedrate of global change has
as tryingto paint a movingtrain.
made the taskof politicalforecastingas difficult
Yet thereare clear signsthattheemergingpost-ColdWarworldis again becoming
tripolar,withthe United States,Germany,and Japan as the poles-each in control
of a sizableregionalbloc.
Supportingthis view,WalterMead envisionsa world made up of three rival
blocs-Europe, East Asia, and the Americas-with the U.S. heading the "weakest
and mosttroubled"of the three (1992:335). Similarly,Leonard Silk positsthatthe
post-ColdWarworld
has become "tripolar"economically,withthe United States,Japan and Germany
(or, in regional terms,North America, the Pacific Rim and the European
Community) bound together in a complex relationship,both rivalrous and
interdependentlike a tempestuousmarriage.
Depending on the waythe menage a troisbehaves, the relationshipmaysplit
apart or strengthenand mature.Threesomes are inherentlyunstable, however;
the immediatedanger,theJapanesebelieve,is thatthe Americansand Europeans
willgang up on them.(Silk, 1991:2)
The significance of economic tripolarityhas heightened "now that

geoeconomics is turning geopolitics and all warfare into a provincial
phenomenon" (Luttwak,1992:13). Today more thanever,scholars,statesmen,and
citizensalike appreciatethe linksbetweensecurityand economic issuesand assess
states' relative power according to their economic, not military,resources.
Typifying the new awarenessof economic-security links,Leslie Gelb reasons that
"in the absence of the Sovietmilitarythreat,the Americans,West Europeans and
Japanese have lost incentivesto set aside economic differences.As a result,
economic conflictshave become the mostpronounced source of tensionbetween
nations,and disputesare becomingmore difficult to resolve"(1991:54).
The political balance is also becoming tripolar.Witness Germany'srecent
decision to recognizeSlovenia and Croatia againstthe objectionsof the U.S., the
E.C., and the U.N., demonstratingthat it is no longer an economic giant and a
politicaldwarf(Tagliabue, 1991; TheEconomist, 1991-92). Some even argue that
the "newGermany,like its predecessors,has proved thatit is a revisionistpower,
intenton reshapingEurope" (Lind, 1991:A33). AnthonyLewis agrees,and warns:
"At a time when angry nationalismis flaringup in so many places, it is in
everyone'surgentinterestto knitGermanyintoa largerEurope" (1991:A19).
Like Germany, Japan is beginningto assertpoliticalpowermore in accordance
with its status as an economic superpower.In the firstU.S.-Japanesebilateral
negotiationssince the collapse of the SovietUnion, the two countriespledged to
share responsibilityfor maintainingpeace and prosperityin the region. Yet, the
competitiveatmosphereat the Tokyo summitpromptedTheEconomist to viewit as
a turningpoint in U.S.-Japaneserelations:"Fromvalued ally,Japan is being cast
increasinglyas a dangerous competitor,more of a threatthan an opportunity"
(1992:52). FiftyyearsafterPearl Harbor, some see Japan realizingits dream of a
"GreaterEast AsianCo-Prosperity Sphere" (Sanger,1991:Dl, D22).
Focusingsolelyon the tripolarstructureof the emergingpost-Cold War system,
one is temptedto predictincreasingsystemicinstability, possiblyfulminatingin
general major-powerwar.As I have writtenelsewhere,however,the huge increase

100 Tripolarity
in the numberof democraticstates,and the factthat,unlike the interwarsystem,

all three poles of the post-Cold War systemwill be democracies,should greatly
mitigatethe destabilizingeffectsassociated with its volatile tripolar structure
(Schweller,1992:268).
References
ASHLEY, R. (1984) The PovertyofNeorealism.International Organization38:225-286.
ASTER, S. (1989) "'GuiltyMen': The Case of Neville Chamberlain."In Pathsto War:NewEssayson the
OriginsoftheSecondWorldWar,edited by R. Boyce and E. M. Robertson,pp. 233-268. London:
Macmillan.
BARNETT, C. (1972) TheCollapseofBritish Power.New York:WilliamMorrow.
BURNS, A. L. (1957) FromBalance to Deterrence:A TheoreticalAnalysis.World Politics9:494-529.
BUTLER, J. R. M. (1957) Grand Strategy, Vol. 2: September 1939-June1941. London: Her Majesty's
StationeryOffice.
BuzAN,B. (1983) People,States,and Fear: TheNationalSecurity Problem in International
Relations.Chapel
Hill: UniversityofNorthCarolina Press.
CALLEO, D. (1978) The German Problem Reconsidered: Germany and theWorldOrder,1870 to thePresent.
London: CambridgeUniversity Press.
CAPLOW, T. (1968) TwoagainstOne: Coalitions in Triads.EnglewoodCliffs, NJ:Prentice-Hall.
CARR,E. H. (1951) German-Soviet Relationsbetween theTwo WorldWars,1919-1939. New York:Harper &
Row.
CARR,W. (1978) Hitler:A Studyin Personality and Politics.London: EdwardArnold.
CHALLENER, R. D. (1955) The FrenchTheory of theNation in Arms,1866-1939. New York: Russell &
Russell.
CHRISTENSEN, T. J.,ANDJ. SNYDER (1990) Chain Gangs and Passed Bucks:Predicting Alliance Patternsin
Multipolarity. International
Organization44:137-169.
Ciano'sDiary,1937-1938 (1952) TranslatedbyA. Mayor.London: Methuen.
CLAUDE, I. (1962) Powerand International Relations.New York:Random House.
DEPORTE, A. W. (1979) Europebetween theSuperpowers:TheEnduringBalance,2d ed. New Haven, CT: Yale
University Press.
DESSLER, D. (1989) What's at Stake in the Agent-Structure Debate? InternationalOrganization
43:441-473.
DEUTSCHER, I. (1949) Stalin:A PoliticalBiography,2d ed. New York:OxfordUniversity Press.
DITrMER, L. (1981) The StrategicTriangle: An ElementaryGame-TheoreticalAnalysis.WorldPolitics
33:485-515.
DIVINE, R. A. (1969) Rooseveltand WorldWarII. Baltimore:TheJohns Hopkins Press.
TheEconomist (1991-92) EC and Yugoslavia:Countdownto Recognition.December 21-January3:57.
TheEconomist (1992) A Poor Matchin Tokyo.January11-17:51-52.
FEILING, K (1947) TheLifeofNevilleChamberlain. London: Macmillan.
FISCHER, L. (1969) Russia'sRoadfrom PeacetoWar:SovietForeign Relations,1917-1941. New York:Harper
& Row.
ForeignRelationsof theUnitedStates1939 (1956) Vol. I. Washington,DC: United States Government
PrintingOffice.
GARNHAM, D. (1985) 'The Causes of War: SystemicFindings." In Polarityand War: The Changing
StructureofInternational edited byA. N. Sabrosky,pp. 7-23. Boulder,CO: WestviewPress.
Conflict,
GATES, E. M. (1981) End oftheAffair: TheCollapseoftheAnglo-French Alliance,1939-40. London: George
Allen & Unwin.
GELB, L. H. (1991) FreshFace. NewYorkTimes, December 8, Sec. 6:50, 54.
GRIECO, J. (1988) Anarchyand the Limitsof Cooperation: A Realist Critique of the NewestLiberal
Institutionalism. International
Organization 42:485-507.
GULICK, E. V. (1955) Europe'sClassicalBalanceofPower. New York:W.W.Norton.
HAAs,E. B. (1953) The Balance ofPower:Prescription, Conceptor Propaganda?World Politics
5:442-477.
HEALY, B., AND A. STEIN (1973) The Balance of Power in InternationalHistory.Journalof Conflict
Resolution 17:33-61.
HENDERSON, N. (1940) Failureofa Mission:Berlin,1937-1939. New York:G. P. Putnam'sSons.
HERNDON, J. S. (1983) "BritishPerceptions of Soviet MilitaryCapability,1935-9." In The Fascist
Challenge and thePolicyofAppeasement, edited byW.J. Mommsenand L. Kettenacker,pp. 297-319.
London: Allen & Unwin.

HILLGRUBER, A. (1981) Germany and theTwo WorldWars,translatedby W. C. Kirby.Cambridge,MA:

HarvardUniversity Press.
HINSLEY, Cambridge:CambridgeUniversity
F. H. (1951) Hitler'sStrategy. Press.
HITLER,A. ([1925] 1941) MeinKampf:New York:Reynaland Hitchcock.
HITLER,A. ([1928] 1961) Hitler'sSecret Book,edited by G. L. Weinberg,translatedbyS. Attanasio.New
York:GrovePress.
HOCHMAN, J. (1984) TheSovietUnionand theFailureofCollective 1934-1938. Ithaca, NY: Cornell
Security,
University Press.
HOPF, T. (1991) Polarity,the Offense-DefenseBalance, and War. AmericanPoliticalScienceReview
85:475-493.
HOWARD, M. (1972) TheContinental Commitment. London: Temple Smith.
HUGHES,J. M. (1971) To theMaginotLine: The Politicsof FrenchMilitaryPreparationin the 1920's.
Cambridge,MA: HarvardUniversity Press.
R. (1978) CooperationUnder the SecurityDilemma. World
JERVIS, Politics30:167-214.
JOHNSON, D. (1983) "The FrenchView."In 1939: A Retrospective FortyYearsAfter,edited by R. Douglas,
pp. 54-63. London: Macmillan.
KAPLAN, M. A. (1979) TowardsProfessionalism in InternationalTheory:Macrosystem Analysis.New York:Free
Press.
KAPLAN, M. A. (1957) System and Processin InternationalPolitics.New York:JohnWiley& Sons.
KAPLAN,M. A., L. A. BURNS,AND R. E. QUANDT(1960) Theoretical Analysis of the "Balance of Power."
Behavioral Science5:240-252.
KENNEDY, P. (1987) TheRiseand Fall oftheGreatPowers:EconomicChangeand Military from1500 to
Conflict
2000. New York:Random House.
KEOHANE, R. 0. (1983) "Theoryof World Politics:StructuralRealism and Beyond."In PoliticalScience:
TheStateoftheDiscipline,edited byA. Finifter,pp. 503-540. Washington,DC: AmericanPolitical
Science Association.
KEOHANE, R. 0. (ED.) (1986) Neorealism New York:Columbia University
and Its Critics. Press.
KISSINGER, H. A. (1957) A WorldRestored:Castlereagh, Metternich, and theProblemof Peace, 1812-22.
Boston:Houghton Mifflin.
KRASNER, S. D. (1991) Global Communicationsand National Power:Life on the Pareto Frontier.World
Politics43:336-366.
LEACH,B. A. (1973) German StrategyAgainstRussia,1939-1941. Oxford:Clarendon Press.
LEvy,J. S. (1984) The Offensive/Defensive Balance of MilitaryTechnology. IntenationalStudies
Quarterly 28:219-238.
LEwIS,A. (1991) StillLittleEngland?NewYorkTimes,December 16:A19.
LIND,M. (1991) Recognize the Powerof the New Germany.NewYorkTimes,December 27:A33.
LISKA,G. (1962) Nationsin Alliance:TheLimitsofInterdependence. Baltimore:JohnsHopkins University
Press.
LuKAcs,J.(1989) The Coming of the Second WorldWar.Foreign 68:165-174.
Affairs
LuTTwAK, E. (1992) Obsession. TheNewRepublic, February24:12-14.
MANDELBAUM, M. (1988) TheFateofNations.New York:CambridgeUniversity Press.
MATTERN, J. (1942) Geopolitics:DoctrineofNationalSelf-Sufficiency and Empire.Baltimore:JohnsHopkins
University Press.
McKEowN,T.J. (1991) The ForeignPolicyof a DecliningPower.International Organization45:257-279.
MCSHERRY, J. E. (1970) Stalin,Hitler,and Europe,Vol. II: TheImbalanceofPower,1939-1941. Arlington,
VA: The Open-Door Press.
MEAD,W. R. (1992) "On the Road to Ruin." In TheFutureofAmerican ForeignPolicy,edited by C. W.
Kegley,Jr., and E. R. Wittkopf,pp. 332-339. New York:St. Martin'sPress.
MEARISHEIMER, J.J. (1990) Back to the Future: Instabilityin Europe afterthe Cold War. International
Security15:5-56.
MILNER, H. (1991) The Assumptionof Anarchyin InternationalRelationsTheory:A Critique.Reviewof
IntenationalStudies17:67-85.
MILwARD, A. S. (1977) War,Economy and Society, 1939-1945. Berkeley:University of CaliforniaPress.
MORGENTHAU, H. J. (1948) PoliticsAmongNations:The Struggle forPowerand Peace, 1st ed. New York:
AlfredA. Knopf.
MUELLER, J. (1988) The Essential Irrelevanceof Nuclear Weapons: Stabilityin the PostwarWorld.
InternationalSecurity13:55-79.
MURDOCH,J., ANDT. SANDLER(1982) A Theoretical and Empirical Analysis of NATO. Journal of Conflict
Resolution
26:237-265.
ofHarold Nicolson:The War Years,1939-1945. New
NIcoLsoN, N. (ED.) (1967) TheDiariesand Letters
York:Atheneum.

102 Tripolarity
Niou, E. M. S., AND P. C. ORDESHOOK (1990) Stabilityin Anarchic InternationalSystems.American

PoliticalScienceReview84:1207-1234.
Niou, E. M. S., P. C. ORDESHOOK, AND G. F. ROSE (1989) TheBalance ofPower:Stability in International
Systems.New York:CambridgeUniversity Press.
NYE,J.S. (1988) Neorealismand Neoliberalism.World Politics40:235-251.
OLSON, M., AND R. ZECKHAUSER (1966) An EconomicTheoryof Alliances. Review ofEconomics and
Statistics
48:266-279.
ONEAL, J. R. (1990) The Theory of Collective Action and Burden Sharing in NATO. International
Organization 44:379-402.
ORGANSKI, A. F. K (1958) World Politics.New York:AlfredA. Knopf.
ORGANSKI, A. F. K, AND J.KUGLER (1980) The WarLedger. Chicago and London: University of Chicago
Press.
OSTROM, JR., C. W., ANDJ. H. ALDRICH (1978)The Relationship betweenSizeand Stability in theMajor
PowerInternationalSystem.American JournalofPoliticalScience22:743-771.
PARKER, R. A. C. (1974) Great Britain,France, and the Ethiopian Crisis, 1935-1936. The English
HistoricalReview89:293-332.
POSEN, B. (1984) The SourcesofMilitary Doctrine:France,Britain,and Germany between theWorldWars.
Ithaca,NY: CornellUniversity Press.
RICH, N. (1973) Hitler'sWarAims:Ideology, theNazi State,and theCourseofExpansion.New York:W. W.
Norton.
RICHARDSON, J. L. (1988) New Perspectiveson Appeasement: Some Implicationsfor International
Relations.World Politics40:289-316.
RIKER, W. H. (1962) TheTheary ofPoliticalCoalitions.New Haven, CT: Yale University Press.
ROBERTSON, E. M. (1963) Hitler's Pre-War Policyand MilitaryPlans, 1933-1939. London: Longmans.
ROSECRANCE, R. (1981) InternationalTheoryRevisited.International Organization 35:691-713.
ROTHSTEIN, R. L. (1968) Alliances and SmallPowers. New York:Columbia University Press.
RUGGIE, J. G. (1983) Continuityand Transformationin the World Polity: Toward a Neorealist
Synthesis.World Politics35:261-287.
RUSSETT, B. M. (1968) Components of an Operational Theory of InternationalAlliance Formation.
JournalofConflict Resolution 12:285-301.
SANDLER, T., AND J.CAULEY (1975) On the Economic Theoryof Alliances.JournalofConflict Resolution
19:330-348.
SANGER, D. E. (1991) Powerof the Yen WinningAsia: New "Co-Prosperity" Is DisplacingU.S. NewYork
Times,December 5:D1, D22.
SCHUMAN, F. L. (1948) International Politics:TheDestinyof theWestern StateSystem, 4th ed. New York:
McGraw-Hill.
SCHWELLER, R. L. (1992) DomesticStructureand Preventive War:Are DemocraciesMore Pacific?World
Politics
44:235-269.
SEABURY, P. (1965) 'The Idea of the StatusQuo." In Balance ofPower,edited by P. Seabury,chap. 22.
San Francisco:Chandler.
SHEARS, L. M. (1967) Patternsof CoalitionFormationin Two Games PlayedbyMale Tetrads.Behavioral
Science12:130-137.
SILK, L. (1991) Some ThingsAre More VitalThan MoneyWhen It Comes to CreatingtheWorldAnew.
NewYorkTimes,September22, Sec. 4:2.
SIMMEL, G. (1950) TheSociology ofGeorgSimmel, translatedand edited by K H. Wolff.New York:Free
Press.
SMITH, D. M. (1983) "Appeasementas a Factor in Mussolini'sForeign Policy."In TheFascistChallenge
and thePolicyofAppeasement, edited byW.J. Mommsenand L. Kettenacker,pp. 258-266. London:
Allen & Unwin.
SNYDER, G. H. (1984) The SecurityDilemma in Alliance Politics.World Politics36:461-496.
SNYDER, G. H. (1991) Alliances,Balance, and Stability. International Organization 45:121-142.
SNYDER, J. (forthcoming)Introductionto CopingwithComplexity in theInternational System, edited byJ.
Snyderand R.Jervis.Boulder,CO: WestviewPress.
STEIN, A. A. (1983) "Coordinationand Collaboration:Regimesin an AnarchicWorld."In International
Regimes, edited byS. D. Krasner,pp. 115-140. Ithaca,NY: CornellUniversity Press.
STRAUSz-HuPE,R. (1942) Geopolitics: TheStruggle forSpaceand Power.New York:G. P. Putnam'sSons.
TAGLIABUE, J. (1991) Bold New Germany:No Longer a Political 'Dwarf. New YorkTimes,December
16:A12.

THIES, W. J. (1987) Alliances and Collective Goods: A Reappraisal. Journalof ConflictResolution

31:298-332.
TOLAND,J. (1976) AdolfHitler.NewYork:Doubleday.
TREvoR-RoPER,H. R. (ED.) (1953) Hitler'sTableTalk, 1941-1944, translatedby N. Cameron and R. H.
Stevens.London: Weidenfeldand Nicolson.
TUCKER,R. W. (1977) The InequalityofNations. New York: Basic Books.
ULAM,A. B. (1974) Expansionand Coexistence: SovietForeignPolicy,1917-73, 2d ed. New York: Holt,
Rinehartand Winston.
WAGNER,H. R. (1986) The Theoryof Games and the Balance of Power. World Politics38:546-576.
WALT,S. (1987) TheOriginsofAlliances.Ithaca,NY: CornellUniversity Press.
WALTZ,K N. (1979) ThearyofInternational Politics.Reading,MA: Addison-Wesley.
WARK,W. K (1985) The Ultimate Enemy:BritishIntelligence and Nazi Germany, 1933-1939. Ithaca and
London: CornellUniversity Press.
WATT,D. C. (1989) How WarCame:TheImmediate OriginsoftheSecondWorldWar,1938-1939. London:
Heinemann.
WEBER,M. (1946) FromMax Weber: Essaysin Sociology, translated and edited by H. H. Gerth and C. W.
Mills.New York:OxfordUniversity Press.
WEINBERG, G. L. (1964) Hitler's Image of the United States. The AmericanHistoricalReview
69:1006-1021.
WEINBERG,G. L. (1981) Worldin theBalance:BehindtheScenesofWorldWarII. Hanover,NH: University
Press of New England.
Problem in InternationalRelations Theory. International
WENDT, A. (1987) The Agent-Structure
46:335-370.
Organization
WENDT, A. (1992) Anarchy Is What States Make of It: The Social Construction of Power Politics.
International 46:391-425.
Organization
WOLFERS,A. (1962) Discordand Collaboration:Essayson International
Politics.Baltimore:JohnsHopkins
UniversityPress.
WOLFERS,A. (1940) Britainand Francebetween theTwo Wars:Conflicting Strategies
ofPeacesinceVersailles.
New York: Harcourt, Brace.
ZINNES,D. A. (1970) "CoalitionTheories and the Balance of Power."In TheStudyofCoalition Behavior:
Theoretical and CasesfromFour Continents,
Perspectives edited by S. Groennings,E. W. Kelly,and M.
Leiserson,pp. 351-369. New York:Holt, Rinehart& Winston.


Schweller, R. (1993) - Tripolarity and The Second World War

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Schweller, R. (1993) - Tripolarity and The Second World War

Hochgeladen von

Copyright:

Verfügbare Formate

Tripolarity and the Second World War

Author(s): Randall L. Schweller

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

This essay seeks to offera new structuralaccount of the outbreak of

Author's note:I would like to thankThomasJ. Christensen,Barbara Farnham,Ted Hopf, RobertJervis,

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

A small-numbersystemcan alwaysbe disruptedby the actionsof a Hitlerand the

Likewise,prominentgame theoristssuch as EmersonNiou and Peter Ordeshook

IThough farmore sympathetic to Waltz'sworkthanAshley,Glenn Snyder(1991:123, 138) raisessimilarconcerns

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

The essay concludes by suggestingimplicationsof the analysisfor the post-Cold

Two Modificationsof Waltz'sSystemsTheory

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

that "all of the competingstatesare strivingrelentlessly to increase theirpower,"

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

determinedbythe powerand interestsof the unitsand the structures withinwhich

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

(5) In a coalition attack,the resources of the victimare divided propor-

Among otherthings,thislast assumptionand the inclusionof statusquo states

One RevisionistPole. Givenonlyone revisionist

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

aggressionin tripolaritytendsto be self-defeating. On the one hand, ifa stalemate

Two Revisionist Poles.A tripolarsystemcomposed of two revisionistpoles is

ThreeRevisionist Poles.In this case, in which all three membersare of equal

21 have borrowedthe terms"abettor"and "eyewitness" fromLiska (1962: 163-164); "tertius

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

terbalancinga strongeror more threateningpower or alliance) because any

ThreeStatusQuo Poles.This, of course,is the moststabletripolarsystembecause

Type2. TheParadoxofPower:A > B = C butA < B + C

Thirdand theBalancer:A < B = C

TheBalancer.However,if B and C are mutuallyhostile,then it is clearlyin A's

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

diplomacythatthe twowarringpoles do not temporarily set aside theirdifferences

TwoStatusQuo Poles.Ironically,when thereare twostatusquo poles, the system

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

case. For thiswe need to considerthe behaviorof middleand smallpowersas well

4Moreover,under real-worldconditions,even rationalplayersmaybe uncertainof the distributionof power in

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

reasonable to assume, therefore,that they do not seek minimum winning

For these reasons, the concept of the minimumwinning coalition-which

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

Distancing.In addition to the formationof alliances throughbalancing and

Summaryof Hypotheseson AlliancePatterns

and theSecond WorldWar

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

the case was essentiallychosen for me. Luckily,the late-

A statewillbe coded "revisionist"

TABLE 1. COW percentageshare distribution

USSR U.S. Germany UK Japan France Italy

25.0 22.7 20.2 10.4 9.4 6.9 4.9

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

Applyingthesecriteriato the interwarperiod,Britain,France,and the U.S. won

State Type Power Interest

USSR pole 5.0 revisionist

A.W. DePorte's discussionof thisperiod nicelysummarizesthe essence of what

Table 2 also suggeststhatwhethera stateis revisionistor statusquo is not an

AllianceDynamicsof theMajor Powers

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

In Christensenand Snyder'sown words,Britain'sbehaviordirectlycontradicts

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM

"expectedbrighterfuturewiththe present,"the COS concluded: "We shallbe in a

This viewis supported,forinstance,by Harold Nicolson's diaryentryon October

This content downloaded from 190.3.14.67 on Thu, 9 Oct 2014 22:23:09 PM