You are on page 1of 30

The Regional Subsystem: A Conceptual Explication and a Propositional Inventory

Author(s): William R. Thompson


Source: International Studies Quarterly, Vol. 17, No. 1 (Mar., 1973), pp. 89-117
Published by: Wiley on behalf of The International Studies Association
Stable URL: http://www.jstor.org/stable/3013464 .
Accessed: 12/06/2014 23:52

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 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
The Regional Subsystem
A Conceptual Explication and a Propositional Inventory

WILLIAMR. THOMPSON
Departmentof PoliticalScienice
University
of Minnesota

MichaelBanks(1969: 351-352)hasdelineated fivethemes in


the applicationof thevarioussystems perspectives
to regional
studies:(1) themacroapproachto theidentification ofregions
usingaggregate data; (2) the neofunctionalapproach;(3) the
socialcommunication approach;(4) thecomparison ofexisting
integration theorieswitlhothertlheories
of international
proc-
esses;and (5) theapplication of broadcategories
of systems
theoryto regionalstudies.WhileBanksbriefly discussedthe
finaltheme,hereafter labeledtheregionalsubsystem approach,
verylittleof his reviewexplicitlytouchedupontheproblems
encountered by thisparticular mode.' Thisresearch
analytical
note will selectivelydiscusssome of these problems.The
specificand modestobjectives aretwofold:(a) The definitions
of the basic analyticalunit, offeredby 22 published
(1958-1971) analysts,will be comparedand "dissected."A
1. The special issue of InternationalStudies Quarterlyin which Banks' article
appeared was devoted to internationalsubsystemsper se, which includesthe diverse
functionaland nonregionalsubsystemsstuchas the Commonwealth,the Communist
states, or coalitions of petroleum-producing
states. Since only some of the world's
subsystemsare regional,Banks need not be takento taskforhis brevity.

AUTHOR'S NOTE: This paper was originallypresentedat the AnnualMeetingof


I.S.A.-West,Portland, March 1972. I am indebted to ProfessorsGeorge Modelski,
FremontLyden, and Fouad Ajami,all of the Universityof Washington,and to David
Osterbergof the Universityof Lancasterfortheircriticismsof earlierversions.I am
equally gratefulfor the editorialguidanceof K. J.Holsti and his anonymousreaders.
International Studies Quarterly, Vol. 17 No. 1, March 1973, 1973 IS.A.

[891

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[90] INTERNATIONAL STUDIES QUARTERLY

standardizeddefinitionwill be constructedin order to aid


futuresubsystemcomparison.(b) An inventory of unprocessed
propositionswill be extractedand presentedin orderboth to
generateinterestin the theoreticalpotentialand aspirationsof
regionalsubsystemresearchand to underlinethe littleprogress
made, to date, toward the goal of scientificexplanationsof
regionalsubsystem behavior.
Many of the ideas concerningregionalsubsystemsspring
from the generalconceptionof the internationalsystem.In
fact,the firstpublisheduse of the conceptwas in reactionto
Morton Kaplan's (1957) pioneeranalysisof the international
system. Kaplan postulated that the internationalsystem's
patternof action in the 1950's could best be explained by a
bipolar model. An area specialist,Leonard Binder,disagreed
withthe universalextensionof thisgeneralizedmodel,particu-
larlyin referenceto the MiddleEast. Binderwent on to argue
that Kaplan's model contained a "great power" bias which
could not effectivelybe used to explain the behavior of
less-than-great
powers.This was said to be chieflydue to either
the smallerpowers'nonbipolarperspectiveor, alternatively, to
theirattemptsto manipulatethe bipolarsituationto theirown
advantage or to insulate themselvesfromit (Binder, 1958:
409-411).
In essence,Binder contendedthatworldpoliticswas not as
bipolaras Kaplan and othershad postulated.Justas studentsof
comparativepoliticshave struggledto shed theirethnocentric
preoccupations,analystsof world politicshave also begun to
realize the limitationsimposed by an exclusive great or
superpowerfocus. Scholars,of course,have neverreallydenied
the complexity of world politics. The problem remains,
however,that,if a model such as the bipolarone is foundto
be too restrictive,
studentsmustcontinueto searchformodels
capable of broader scope. While not all regionalsubsystem
analystshave attemptedto devisealternative models,someview
the world politicalarena as a networkof systemlevels (e.g.,
global,regional,national,and local). An importantadvantageof
such an approachis its flexibility.
As Oran Young (1964: 250)
has stressed,thereis no need to makea once-and-for-all
decision

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [911

concerningresearchfoci. Instead,the analystselectsthelevel(s)


which appears to offer him or her the best chance of
understanding and predictingthe behaviorchosen for study.
Thus, those analystsdealingwith regionalsubsystemsinvesti-
gate one of the territorially intermediatelevels of analysis
somewherebetweenthe global and the nationallayersand, in
doingso, theyofferthe followingrationalesfortheirapproach:
(1) The analysisof regionalsubsystems permitssomereduction
in the complexityof world politics. Boundariesare drawn
whichallow a morelimitedfocusand whichwould assistmore
comprehensive examinationsof relevantvariablesand processes
than mightotherwisebe possible. Essentially,the goal is to
increasethe subtletyof analysis.(2) Whereasanalyticalbound-
ariesare ultimatelyarbitrary,
some are mostjustifiedand more
usefulthanothers.If regionalsubsystems existin worldpolitics
as distinct"theatresof operation,"theydeservea shareof the
analyst'sattention.This is especially the case when national
elites considertheirregionalenvironments to be of primary
concern.Such an approach compensatesfor past and present
biases in favorof superpoweromnipotence.(3) The studyof
regionalsubsystemspresentsan opportunity fortheintegration
of the findingsof area specialistsand international relations
students.Both groupspresumably have something to offereach
other. A mutual exchange could only benefit our current
understanding of world politicalbehavior.(4) Finally,thereis
an equally excellent opportunityfor gainingfurtherinsight
throughcomparativeanalysis.For how else are we to learn
which formsof behaviorare "universally"regionaland which
are peculiar to specifictypes of regions?In this fashion,it
should be possible to avoid area-centricitiesor at least learn
where they are appropriate.It is this last justificationof
regional subsystemresearch,the rewardsof comparison,to
whichtheremainderof thisresearchnotewillbe addressed.

CONCEPTUAL DISPARITY

The accumulationof publishedworksdevotedprimarily to a


considerationof regionalsubsystems
has been slow and uneven.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[92] INTERNATIONAL STUDIES QUARTERLY

TABLE 1
REGIONAL SUBSYSTEM LABELS

Terms Analysts
Subordinateinternationalsystem Binder
Regional subsystemor subsystem Modelski,Aron,Kaiser,Hellmann,
Thompson,Yalem, and Young
Subordinatestatesystem Brecher,Zartman,and Bowman
Systemof nations Reinton
Partialinternationalsystem Hoffmann
Internationalsubsystem Haas, Dominguez,and Shepherd
Subordinatesystem-international
region Cantoriand Spiegel,Sigler,and Miller
State system Hodgkin

It is only quite recentlythatany criticalmass of literaturehas


begunto develop.Even so-or perhapsbecause thisis so-there
is not a greatdeal of definitionalagreementon what exactly
constitutes a regionalsubsystem.To beginwith,severaltermsor
labels are used to identifythe focusof concern.Of evengreater
importance,however,are the contentsof the various defini-
tions.Afterassemblingthe availabledefinitions, it was possible
to distinguish21 attributesused to identifythe regional
subsystem.Table 3 matchesthe variousattributes, cited by the
respectiveanalystsin theirdefinitions.Excludingthe 4 authors
who did not provideexplicitdefinitions, less than 22% of the
4
potentialcells are checked. Table representsthe resultsof a
dyadicindex of interanalyst agreement(or thelack thereof).It
was derivedsimplyby countingthe sharednumberof attributes
cited by each pair of analystsand dividingby thetotalnumber
of itemsgivenby both membersof the pair. For example,if
writerX mentionsattributes1, 2, 4, and 6, and writerY uses
attributes1, 4, 7, and 8, theyshareattributes1 and 4 and the
total numberof attributesgivenis 6 (1, 2. 4, 6, 7, and 8). The
X-Y indexof interagreement would thenbe 2/6or .333.
There are, no doubt, severalways to interpret Table 4, the
most obviousbeingthatthereis littleconsensusin theregional
subsystemliteratureon this basic definitionalquestion. Only
four pairs have reasonably high indices. The highestpair,

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [93]

TABLE 2
REGIONAL SUBSYSTEM ATTRIBUTE LIST
(1) Proximityor primarystresson a geographicregion.
(2) Actors' pattern of relations or interactionsexhibit a particulardegree of
regularityand intensity.
(3) Intrarelatedness-a condition wherein a change at one point in the system
affectsotherpoints.
(4) Internalrecognitionas distinctivearea.
(5) Externalrecognitionas distinctivearea.
(6) One or moreactors.
(7) At leasttwo actors.
(8) At leastthreeactors.
(9) Small powersonly.
(10) Unitsof power are relativelyinferiorto unitsin the dominantsystem.
(11) Subordinationin the sense that a change in the dominantsystemwill have a
greatereffecton the subsystemthanthe reverseand thereis more intensiveand
influentialpenetrationof the subsystemby the dominant system than the
reverse.
(12) Geographical-historicalzone.
(13) Some degree of sharedethnic,linguistic,cultural,social, and historicalbonds.
(14) A relativelyintegratedand unifiedarea.
(15) Some evidence of integrationor a professed policy of achieving further
economic,political,and social integration.
(16) Functionallydiffuse.
(17) Explicit institutionalrelationsor subsystemorganization.
(18) Autonomy-intrasystemactions and responses predominate over external
influences.
(19) A distinctiveconfiguration of militaryforces.
(20) A regionalequilibriumof local forces.
(21) Common developmentalstatus.

Cantoriand Spiegel-Miller, is possiblyartificial.Miller'sarticle


was evidentlywrittenforCantoriand Spiegel's(1 970a) book of
readings,and while Millerdoes not actuallyprovidea regional
subsystemdefinition,she gives no indicationof disagreement
with her editors. Of the other threepairs: Brecher-Bowman,
Kaiser-Hellman,and Modelski-Thompson, Bowmani Hellman,
and Thompsonovertlyrecognizethe meritof theirpair-mem-
ber's earlierdefinitionso that at least theiragreemenit indices
are not coincidental.It is, of course,entirelypossiblethatthe
measurementmethod exaggeratesinteranalyst differences. The
attributesare not weighted(nor are theanalysts),and the more
criteriainvolvedin a definition,the less likelytherewill be

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[94] INTERNATIONAL STUDIES QUARTERLY

TABLE 3
ANALYST-ATTRIBUTE MATCH
Nunibers
Attrtbute
I 2 3 4 5 6 7 8 9 10 /I 12 13 14 15 /6 1 7 18 19 20 21
Bander(1958) NOT I XPLICIT
HodgkinI196 1) NOT I XPLI( IT
Modelski(1961) X X X X X X
Aron( 1962) X X X X X X X
Bredher ( 1963) X X X X X X X
FktItniann(1963) X X X
Si'% art/inin( 1966) X X X X
Reinton 1967) X X X X
Zartman 1967) X X X X X
Kaiser(1968) X X X
Young(1968) NOT I XPLI( IT
Bowman1968) X X X X X X X
Hellinann 1969) X X X X
Sigler(1969) X X X
J 969)
( antoriand Spiegel X X X X X
Haa 1 970) X X X
Thompson(1970) X X X X X X
Yaledm1970) NOT I XPLI( IT
Mliler 1970) X X X X X
Shepherd)1970W X X X
lDomingue7 11971) X X

TABLE 4
INTERANALYST AGREEMENT

> - t
- - -~
Modelki 18 33 00 00 25 22 29 30 25 13 22 29 71 22 29 14 22
Aron .30 38 10 22 20 43 27 38 11 33 25 30 33 11 13 24
Bre her 25 11 1I 22 29 86 25 13 22 13 50 22 11 00 24
Hottmann 14 00 00 17 22 14 00 13 00 11 13 00 .00 10
S.lhwart/inan 14 13 0() 10 00 17 29 00 00 29 00 00 09
Reinton 29 40 10 60 40 29 40 25 29 40 .20 26
Zartinan 14 20 29 14 11 14 22 11 00 17 15
Kaj%cr 13 75 20 33 50 50 33 20 25 .29
Bowmilan 22 11 20 11 44 20 22 00 22
Hellmnann 17 29 40 43 29 17 20 28
Sigler 33 20 13 33 20 00 16
( antoriand Spiegel 33 22 1 00 14 17 23
Haas 29 33 20 25 22
Thompson 22 29 14 .28
Miler 14 17 27
Shepherd 25 .16
DIomingue/ 12
(averagemean) 21

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [95]

substantialagreement.In addition,not all analystsofferhighly


precise definitions, nor is it always clear whetherthe analysts
are referringto regional subsystemsin general or to the
particular ones they happen to be studyingat the time.
Nevertheless,only 4 of 136 nonredundantpairs showingany
substantialconnectivitycannot fail to appear as anythingbut
dismal.And since Tables 3 and 4 listtheanalystsin chronolog-
ical order, it is clear that no trend toward rectifyingthe
situation is evident. While some readers may perceive this
problemto reflectdesirableacademicindividuality, therecan be
littledoubtthatthosewho strivetowardan empiricalscienceof
world politics must bear the burdenof this basic conceptual
disparity.For unless studentssharea commonanalyticalunit,
theirinvestigations cannotbe compared.And withoutcompar-
ison, it is impossibleto movefromthe descriptionof,say,Arab
or Latin Americanpeculiaritiesto regionalsubsystemgenerali-
zationsand theories.
It is a basic rule of researchthat conceptsmustbe precisely
definedif theyare to serveanyusefulpurpose.Yet thepractice
of each investigator proposinga new definition regardlessof its
claim to precisionwill no longersuffice.That is not to imply
that the problem is one of false definitions.The problemis
rathera lack of uniformity in that the concept of regional
subsystemobviously does not mean the same thing to all
analystsat all times.The likely consequenceshave been and
may continueto be misunderstanding and noncomparability. In
orderto attemptto rectifythepresentstateof affairs, it willbe
necessaryto engagein whatCarlHempel(1952: 12) has termed
explication.Using the 21 attributesas the basis of discussion,
the goal of explicationis to reduce the "limitations,ambigui-
ties, and inconsistencies."A reinterpretation, settingforththe
necessaryand sufficient conditionsforutilizingthe concept of
regional subsystem,will be offered. As with definitions,
explicationscannotbe judged eithertrueor false.The standard
must be the degree of adequacy obtained in reachingthe
previously statedgoals.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[961 INTERNATIONAL STUDIES QUARTERLY

CONCEPTUAL EXPL1C I TION

The multiplecharacteristics attributedto the regionalsub-


systemand similaranalyticalunits have already been listed.
Beforediscussingeach featuire separately,it would be beneficial
to considerthe implicationsof the most prevalentlabel, the
regionalsubsystem.The term"subsystem"is fairlyunambig-
uous. A subsystemis simplya componentof a largersystem
with systemicpropertiesof its own-i.e., "a structurethat is
perceivedby its observersto have elementsin interactionor
relationshipsand some identifiableboundariesthat separateit
from its environment"(McClelland, 1966: 20). The term
"regional"is less clear.In context,it identifiesthenatureof the
"identifiableboundaries";thatis, theinteraction boundariesare
regional.This bit of circularitynaturallyevadesthekeyissueof
what qualifiesas a region.However,thisneed not be a critical
stumblingblock. Strictlyspeaking,regionalsubsystemsneed
not be geographicalregions per se. Rather,the subsystems
consistof the interactionsof nationalelites,not the physical
entitiesof politicalunits,of whichtheinteractions are observed
to have moreor lessregionalboundaries.In thissense,it should
only be necessaryto employ the minimalregionalcriterion-
namely,generalproximity.2
If one examinesthe literatureattributelist,it will be noted
that the first2 attribuItes are alreadyimpliedby the regional
suibsysteinterm.The first2 attributesalso happen to be the
most consistentlycited by the 18 analystsand, of all the 21
attributes,the first2-proximityand regularinteraction-come
closestto supplyingthe necessaryand suIfficientconditionsfor
applyingthe coinceptof regionalsuibsystem. If theremaining19

2. Misinterpretation of the political nature of regionalsubsystemshas led one


scliolar to state essentiallythat there is no region that can be identifiedas a
subsystemof the internationalsystem(Russett, 1967: 168). The argumentfor
objectivemultiplecriteriaas a means of delineatingsubsystemsappearsto be in part
a legacy of the geographers'search for spatial homogeneity.By all means, these
dimensionsshould be explored in orderto assess theirimpacton subsystempolitics,
but they do not necessarilydeterminethe boundaries of a regional subsystem.
Classification rigor is equally desirable where possible, but it is national elite
interactionand orientationwhichmustbe measured.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [97]

attributescan be shownto be neithernecessarynor sufficient,


thequestionof conceptualdefinition
can be resolved.

so frequently
is a characteristic
Attribute3: Intrarelatedness
includedin systemdefinitionsthatits adoptionherewould be
almost redundant.Since intrarelatedness is virtuallysynony-
mous with system, its definitionalutility lies mainly in
the meaningof systemor subsystem,and thisshould
clarifying
properly be considered as both a qualificationand as a
consequenceof attribute2.

Attributes4 and 5. Internaland externalrecognition appear


to be the productof commonsense. The regionalsubsystem
concept would have less than a heuristicvalue if eitherits
nationalelites or its penetrators wereblissfullyignorantof its
boundary-erecting existence.The problemarises,however,that
if attributes1 through3 pertain,politicalactorsare bound to
be aware of the subsystemic eventhoughit may
distinctiveness
not always be politic to pay homage publiclyto the idea. In
addition,the foreignpoliciesof greatpowershave on occasion
given the impressionthat regionalsubsystemseitherdo not
existor at least need not be takenseriously.Dulles' attemptsin
the early 1950s to gain several Arab adherentsto what
eventuallybecame a NorthernTier anti-Sovietalliance comes
immediatelyto mind. A more timelyexample mightbe the
belated recognitionof the Vietnamconflictas an Indo-China
war. In any event,attributes4 and 5 are significant and would
be necessaryand sufficientin their own rightor as natural
derivativesof the first3 attributes.Their presencein a basic
definitionis beneficial,but they may need to be appliedless
rigorouslythanwouldotherwisebe desirable.

Attributes6, 7, and 8: Analystshave invoked minimal


membershipthresholdsof from 1 to 12. This is an empirical
questionwhichhas not reallybeen studied.On theotherhand,
it maynot be a questionof greatsignificance.Since thiswriter
is unawareof any singleactorwho satisfiesattributes1 through

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[98] INTERNATIONALSTUDIES QUARTERLY

5 (in termsof attribute1, such a situationwould seem highly


illogical!), attribute 6 (one or more) can be dismissed as
inconsistent.As for choosing between attributes7 (two or
more) and 8 (three or more), until contraryevidence is
provided,7 (at least two actors)is preferable because it is more
inclusivethan its rival. While all contemporary regionalsub-
systemshave at least threeactors,thereis no pressingreasonto
limit the concept to the currentscene. For that matter,the
concern for membershipthresholdsclearly betraysa nation-
state-centricbias, the perceivedutilityof which,despite the
obvious problemsforquantification, is likelyto diminishin the
near future. Multinational business corporations,guerrilla/
terroristgroups,and certainindividualsdeserverecognitionas
regionalsubsystemactorswhereappropriateand irrespective of
their national affiliationor lack thereof.Even the smallest
subsystemis thenlikelyto have manyactors.

Attribute 9: There equally does not seem to be any


overwhelmingrationale for restricting regionalsubsystemsto
small powers only. Middle-ranking powers need not be leftin
limbo. For that matter,some observersfeel that the Western
European subsystemhas potentialfordevelopinginto a greator
superpowerin its own right.If it did, it would mostlikelyretain
of a regionalsubsystemforsome time.
the characteristics

Attributes 10 and 11: Whetherthe dominant system is


thoughtof as bipolar,tripolar,quadrilateral,or evenpentagonal
in nature,all state actors' unitsof power are relativelyinferior
to the unitsin thedominantsystem.But whatif the "dominant
system"happened to be a regionalsubsystemas well?One does
not have to searchveryfarin worldhistoryto findexamples.
Again, there is no need to restrictthe regional subsystem
concept to the immediatepast and present.This last criticism is
applicable to the idea of subordinationas well. Severalwriters
have placed particularstresson thisaspect.Again,thisseemsto
be a fairlyuniversalyet contemporary featureof worldpolitics.
It may be an overly absolute judgmentas well. Subordinationis

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [99]

an importantbut variableprocess. Its natureand degreeneed


not be uniformfor all subsystemsnor constantovertimefor
any singlesubsystem.For thatmatter,certaincrisesoriginating
in the "subordinate"arenashave a way of thrusting into the
interactionof the dominantsystemin sucha mannerthatthese
subsystems'subordinatestatusbecomesfarless obvious,at least
in the shortrun. In a different
direction,it may be questioned
whethersubsystemsare subordinatedto the dominantsystem
or to certainactorsof the centralsystem(M. Haas, 1970: 101).
Nevertheless,the centralpointto be madeis thatsubordination
is not a necessarycondition,and perhaps it is too easily
assumed. The precisenatureand degreeneed to be carefully
examinedforeach subsystem and fortheappropriateperiodsof
time. The question should not be whethera subsystemis
subordinatedor not, but rathersubordinatedto what extent,
underwhatconditions,and to whom.

Attributes 12, 13, 14, and 15: Attributes12 and 13


(geo-historicalzone and shared bonds) are the least objection-
able of this group,but withan emphasison interaction, these
characteristics
are renderedunnecessaryeven thoughtheymay
be frequentlypresentand of some significance. Attributes14
and 15 are anothermatter.These criteriaare bothlegaciesor at
least "spillovers" of the older interestsin the process of
regionalism.Theiremphasisis a potentialliabilityto the study
of regionalsubsystemsin two ways.First,as Miller(1970: 364)
has noted,thetermssystemor subsystemcan be instrumentally
neutralvis-a-visthe growthof cohesion. Second, in empirical
systemsanalysis,integration is best interpretedas a specialcase
of systemtransformation or change(Kaiser, 1967: 410). Justas
some studentsof regionalismapparentlyhave little use for
regional subsystemresearch(E. Haas, 1970: 612), regional
subsystemanalystsneed not immediatelyburden themselves
withtheproblemsof regionalintegration studies.If theregional
subsystemmode of analysisprovessuccessful,the studiesof
regionalintegrationwill eventuallyrequirereconciliationand
mergeras specialcasesof subsystemic change.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[100] INTERNATIONAL STUDIES QUARTERLY

Attributes 16 and 17: Both these attributesare again


commonand probablyapplicableto most regionalsubsystems.
But, once again, they are not necessary conditions. Their
applicabilitymay depend at what point in time one studiesa
specific subsystem.Organizationalfunctions3 may not be
diffuseearlyin the developmentof the subsystemnor will the
institutionalboundaries always coincide with those of the
regional subsystem.The Organizationof AfricanUnity is in
some respectsan example.

Attribute 18: The autonomy characteristicis at once an


easily exaggeratedderivativeof the conceptof systemas well as
an implicitdenial of both the subsystemand subordination
themes.In the absolutesense,an autonomoussubsystemis no
longera subsystem,nor can it be consideredsubordinated.This
problem of varyingdegree also involves what Cantori and
Spiegel (1970a: 297) have labeled the "intrusivecondition."
Intrusiveor penetratingactorscan extricatethemselvesfroman
arena,whileregionalsubsystemmemberscannot.In an age of at
least threeglobal powers,completeautonomyforany regional
subsystemis problematic.At the same time,thisdoes not rule
obt some degree of autonomy on certainissues or types of
issues. One might also find that, for example, sub-Saharan
Africanelites,membersof a less-penetrated subsystem,possess
more autonomy than their counterpartsin Eastern Europe,
Latin America, or Southern Africa. The autonomyvariable
must remain a subject of empiricalinvestigation,but it is
neither a necessarynor sufficientcriterionfor the regional
subsystem.

Attributes 19 and 20: Attribute 19 (distinctivemilitary


forces configuration)is not very informative.Attribute20
(regionalequilibriumof local forces)is perhapstoo informative.
3. A distinctionshould be made, however,between what Young (1968: 69-71)
has called "special purpose" and "generalpurpose" subsystems.The formerperform
a specialized purpose in the operation of a largersystem,while the latterdo not. A
regional subsystemis more likely to be "a separatelyoperatingsubsystemwithina
largersystem"and thereforequalifiesas a generalpurpose system.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
/REGIONALSUBSYSTEM[101]
Thompson

At the same time,both are vague in terminology as well as in


conceptualrationale.In general,the formerseems to add little
to thedefinition, and the latteris unnecessarilyrestrictive.
What
is a nondistinctive configuration? Would a regionalsubsystem
cease to exist if it lacked a local equilibriumof militaryforces?

Attribute21: Common developmentalstatus is similarto


attribute9 concerning smallpowers.Such a restriction
has little
merit.Proximateand interacting actorsmaybe richor poor. In
fact, if all the actors of a regionwere, say, verypoor, it is
theoreticallyconceivablethat they would not interactat all
(Reinton, 1967). But, if we were to add a middle-incomeactor
to a groupof proximateand verypoor actors,we are likelyto
observe a regional subsystemsomewhat on the order of
SouthernAfrica.
In concludingthis exercisein explication,I have proposed
that the necessaryand sufficientconditionsfor the regional
subsystemare as follows:
(1) The actors' patternof relationsor interactions exhibita particular
degreeof regularityand intensity to theextentthata changeat one
pointin thesubsystem affectsotherpoints.
(2) The actorsaregenerally proximate.
(3) Internaland externalobservers and actorsrecognizethesubsystem as
a distinctive
areaor "theatreofoperation."
(4) The subsystem logicallyconsistsof at least two and quiteprobably
moreactors.
As was suggested before, the success of this enterpriseis
dependenton the degreeto whichregionalsubsystemanalysts
findthisproposeddefinitionuseful.

HYPOTHESIZED GENERALIZATIONS OF
REGIONAL SUBSYSTEM BEHA VIOR

The conceptualexplicationwhich precededthis sectionhas


not been promotedwithouta purpose.A majorobjectiveof an
empiricalpoliticalscienceis the establishment
of generalizations
which are capable of explainingand predictingpoliticalbehav-

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[ 102] INTERNATIONAL STUDIES QUARTERLY

ior at all levels of social and political interaction.These


generalizations can only be establishedby repeatedtestingwith
similarresults.An initialrequirementof the testingprocessis
thattheconceptualvocabularymustbe reasonablyuniform.An
importanttermshouldmean the samethingto all investigators.
Regional subsystemresearch,althoughcertainlynot alone in
politicalscience studies,has been lackingin thisrequirement.
The potential consequence is noncumulativeresearch,which
shouldbe viewedas a wasteof scarceresources.This factoris all
the more irritating because, scatteredthroughoutthe regional
subsystemliterature,are a numberof hypothesizedgeneraliza-
tions which could ultimatelybe establishedas principlesof
regionalsubsystembehavior.In orderto begin the processof
overcomingthisproblem,thissectionwill presentan inventory
of propositionsconcerningpoliticalbehaviorwithinand about
regionalsubsystems.4
The propositionshavebeen collectedsolelyfromthatpartof
the publishedliteraturethatmoreor less consciouslymakesuse
of a conceptsimilarto the regionalsubsystem. Clearly,suchan
inventorycould be greatlyexpanded by tappingthe "tradi-
tional" literatureconcernedwithregionalismand the processes
of integrationand world order. This, however,mustremaina
task forfutureresearch.Since thepropositionsbeingpresented
in this collectionhave been takenfromtheiroriginalcontext,
some have had their terminologymodifiedso as to improve
theircommunicability, whileothersmayhavelost some of their
full meaning.In all cases,interestedreadersare encouragedto
refer to the native source. Finally, and in agreementwith
Przeworskiand Teune (1970: 8-9) that a goal of comparative
researchis to substitutethe namesof variablesforthenamesof
social systems,only those generalizationswhichappearto have
been intendedforapplicationto morethanone subsystemhave
been selected.
4. This writeris aware that such an undertakingmightbe more valuable if it
advanced a theoreticallylinked inventoryof hypothesesratherthan an unprocessed
collection.WhileI hope to be able to contributeto regionalsubsystemtheoryin the
future,such an attemptat the presentappears premature.Hopefully,otheranalysts
will be assisted by the present inventoryin preparingand presentingtheir own
theoreticalapproaches.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/REGIONALSUBSYSTEM [103]

SUBSYSTEMDEVELOPMENTAND TRAIVSFORMATION

(1.1) Regional subsystemsare the result of changinggreat power


configurations and variableregionalintegration thereis
pressures;
nothingnecessarily natural,inevitable, or permanent about them
(Modelski,1961: 150).
(1.2) The followingconditionsare conduciveto regionalsubsystems
comingintotheirown:
(a) the absenceof large-scale internationalwarand accompanying
polarization,
(b) thegradualdiffusion of effectivepower,
(c) the riseor resurgence of smallnumbersof new powercenters
ofmajorsignificance butlessthansuperpowers,
(d) the demiseof colonialism and therapidgrowthin thenumber
ofindependent states,
(e) the generalrise in levels of politicalconsciousnessand the
spreadofactivenationalism, and
(f) the emergence of new linesof conflictas effectiveinfluence
combinedwiththe superpowers
diffuses, becomingaware of
theirowncommoninterests (Young,1969: 341).
(1.3) Thereis a long-term structural trendtowardinternationalfragmen-
tation-abreakdown intosubsystems. thedomainof
Consequently,
the centerof the international systemhas been reducedto the
extentthattheperipheries havebrokenaway.To theextentthat
empireshave crumbled,the domainsof the subsystems in the
peripheries havebeenreduced(Dominguez,1971: 178).
(1.4) Subsystemtransformation occurswheneitherI or II takesplace:
(I) A new subsystem beginsif thereis a changein thenumberof
poles,suchas thefollowing:
(a) a majorpoweris demotedthrough military
defeat,
(b) one or moremiddlepowersriseto majorpowerstatus,
(c) a majorpowerentersthesystemanewor exitsfromtheregion,
(d) newmilitary alliancesinvolving majorpowersarise,and
(e) existingmilitary alliancesinvolving majorpowerscollapseor
membership is reshuffled.
(II) A new systembeginswhen a dramaticeventprecipitates a
trendtowardtightening or looseningthe powerdistribution,
suchas thefollowing:
(a) a previouslynonalignedpower becomes an ally of another
majorpower,
(b) an allianceamongmajorpowerscausesa schismto develop,

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[104] INTERNATIONAL STUDIES QUARTERLY

(c) manynonmajorpowersbecome attachedto or breakaway


frombloc arrangements, or
(d) pluralisticsecurity communities wax or wane(M. Haas, 1970:
102).
(1.5) Intergovernmental regionalsubsystems stimulatetheformation of
otherregionalsubsystems of thesametypeiftheformer areactive
in the realmof highpolitics-i.e.,mattersof nationalprestige,
survivaland defense, or powerstatus(Kaiser,1968: 94).
(1.6) Comprehensive regionalsubsystems contribute to theformation of
othersubsystems ofthesamekindif
(a) therehas been priormutualinvolvement on the level of
societies(as in a transnational societysubsystem) and ifthe
comprehensive subsystem has hurtor can be expectedto hurt
theinterests ofotherregions (thethreateffect), orif
(b) the comprehensive subsystems arerelevantas modelsto other
regions(thedemonstration effect;Kaiser,1968: 95).
(1.7) Withinoverlapping regionalsubsystems thatalso sharea transna-
tional society subsystem,the pressureto formor strengthen
comprehensive regionalsubsystems in theareawillgrowwiththe
increasingintensityof interactionin the transnational society
(Kaiser,1968 02).
(1.8) Subsystemic boundariesbreakdownwhensubsystems are linked
throughfusion,international regionalorganizations, initiatives
fromcentercountries, or a combination of these.Althoughthe
foreignpoliciesof the countries of theperipheries continueto be
local, the area withwhichthey are concernedhas broadened.
Local subsystems of limiteddomainand scope arebreaking down
whilecontinentalsubsystems are experiencing a growthprocess
(Dominguez,1971: 186).
(1.9) Only the simultaneous coexistenceof a numberof states-say,
arounda dozen-createsfavorableconditions fortheemergence of
certainsystemproperties-for instance,regionalstructuresof
cultureandsolidarity (Modelski,1961: 148).
(1.10) A pan ideologywhichpervadesa geocultural areamaycharacterize
the earlystagesof systemgrowthin interstate relations(Fleming,
1969: 101).

SUBSYSTEM STABILITY

(2.1) or lossof autonomy


The disintegration ofa keysubsystemmember
will have subsystemwide repercussions; the domestic
therefore,

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [105]

stabilityof most subsystemunits is necessaryfor subsystem


maintenance, which,in turn,is conduciveto stability in theglobal
system (Brecher,1963: 234).
(2.2) Containableinstability of nondominant actorsdoes not affectthe
subsystem (Bowman,1968: 260).
(2.3) The subsystem can handleitsfunctions onlywhenproblem-solving
and autonomyare the subjectof consensusand are highervalues
thanpoliticaladvantage, ideologicalcommitment, or a particular
outcomeofa dispute(Zartman,1967: 549).
(2.4) The mostlikelychangein a component of thesystem,capableof
causingits breakdown, is the destructionof its seemingly fragile
autonomy (Zartman,1967: 561).
(2.5) The relationship betweenthe numberof military powercenters
and subsystemstabilityis negative.A subsystem containing one
pole with a pluralityin militarypower(unipolar)is the most
pacific.Mostmembers preferto engagein waroutsidethesystem,
if at all. A subsystem withfouror moremilitary powercenters,
usuallyunaligned majorpower(multipolarity), is theleastpacific.
Bipolarity is associatedwithlongerwars,and countries seekingto
upsetan existingdistribution of resourcesare mostsuccessful in
thesesubsystems (M. Haas, 1970: 121).
(2.6) Social cohesionseemsto havelittleeffectuponsubsystem stability
(CantoriandSpiegel,1970a: 364).
(2.7) Conflictis greatestin the subsystemcore with the greatest
disparityof powerlevels(withthe exceptionof U.S.-dominated
LatinAmerica;CantoriandSpiegel,1970a: 368).
(2.8) A subsystem, someofwhosecoremembers belongto organizations
that cross systemicboundaries,while theirrelationswithother
coremembers are by no meanscohesive,shouldno doubtproveto
be rather unstableas a system (Miller,1970: 376).
(2.9) The riseof regionalorganizations theworkof countries
facilitates
which act as intermediaries in disputesin a linkedcontinental
subsystem. Intermediaries matterbecausetheirprimaryconcerns
are continental milieugoals (goals whichseek not to defendor
increasethepossessions heldby a stateto theexclusionof others
but to shapeconditionsbeyondnaturalboundaries).Thesegoals
include conflictresolution,the effectiveness of regionalproce-
dures,and theprojection oftheirowninternational positionin the
continent (Dominguez,1971: 191).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[106 ] INTERNATIONAL STUDIES QUARTERLY

INTRASUBSYSTEMIC INTERACTION

(3.1) A subsystem in due coursecreatesandmaintains itsownsolidarity


(Modelski,1961: 151).
(3.2) Subsystems replicatethe hierarchical patternof theinternational
systemso that most countries(exceptforthe weakestand the
strongest) are engagedin superordinate and subordinate relation-
shipswhichstrongly shapethe scope and domainof theirforeign
policies(Dominguez,1971: 208).
(3.3) In a feudalsubsystem, a systemof unitswithtwoproperties, rank
concordanceand interaction in a pairstrongly dependenton total
rank of the pair,the interactions will be determined more by
geography, the lower the positionof the unitin the systemof
stratification(Schwartzman, 1966: 74).
(3.4) Lower-ranking actorsmay identifyin normsand behaviorwitha
higher-ranking actorwhenthereis a possibility of mobility upward
in thesubsystem structure.It is onlyunderconditions ofmobility
thatthe demonstration effectfromhighto low exists(Reinton,
1967: 357).
(3.5) An actor'sstatusin a particular systemdetermines its intensityof
participation, the characterof its behavior,and its capabilityof
influencein the system.The lower the averagestatusof units
withina pair,thelowertheintensity ofinteraction; thegreater the
discrepancy in statusbetweenunitswithina pair,the stronger is
thedomination ofone unitin interaction.
(a) Pairs with unitsof highrankhave eitherhighor medium
interaction.
(b) Pairs with units of mixed (high and medium)rank have
mediumorlow interaction.
(c) Pairswithunitsofmediumrankhavelow interaction.
(d) Pairswithunitsof mixed(mediumandlow) rankhavemostly
low or minimal interaction.
(e) Pairs with unitsof mixed (high and low) rankhave either
minimal orno interaction orthelow-rank unitis dominated by
thehigh-rank unit.
(f) Pairs with units of low rank have eitherminimalor no
interaction (Reinton,1967: 332-348).
(3.6) Egalitarian weaknessin thedifferent components ofa subsystem's
power configuration leads to a tendencyto rejectpoweras the
basisforinternational relations (Zartman,1967: 553).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson
/REGIONALSUBSYSTEM[107]

(3.7) The conceptof systemdeterminancy (statepolicydetermined by


thecharacter anddistribution ofpowerwithinthesystem ofwhich
it is a member)may not have a highdegreeof applicability in
emergingregionalsubsystems (Thompson,1970: 159).
(3.8) An emergingregionalsubsystemwith a combinationof the
followingattributes:antagonistic actor heterogeneity, unstable
interactionpatterns, highlydiffused and generallylow-level power
distribution, inconsistentnormative boundaries,and a substantial
subsystemic ideologydeemphasizing actordifferentiation is likely
to possessa weak and fluctuating statusstructure (Thompson,
1970: 162).
(3.9) The difficulties of integration or cooperationamongthe subsys-
tem'sactorswillbe greaterthelowerthepositionof theactorin
thesystem (Schwartzman, 1966: 58).
(3.10) The volumeof exchangeof politicsin the regionalenvironment
will not necessarily be determined by the volumeof attitudinal
orientationoftheinternal elite(Fleming,1969: 101).
(3.11) The higherthe aspirationforregionalleadership, the higherthe
orientationto environmentsbeyond contiguousboundaries
(Fleming,1969: 101).
(3.12) The stronger the pan ideologyof the regionalenvironment, the
morepoliticswillbe oriented towardrelationships beyondimmedi-
ateneighbors (Fleming,1969: 101).
(3.13) As an actorbecomesmorehighlydifferentiated (an increaseinthe
totalnumberof rolesorlinesof communication amongstates)and
articulated(an increasein the numberof different types of
international communication or interacting roles),it willbecome
more relativelycentralto its regionalsubsystem (Bernsteinand
Weldon,1968: 174).
(3.14) Actorswill be diplomatically salientto each other-thatis, they
willmaintainhighlevelsofpersonnel exchange-tothedegreethey
share other measurablebonds of commoninterest.Diplomatic
contacts would thus be heaviestwithinregionalsubsystems
(RussettandLamb,1969: 39).

INTR USI VE-PENETRATI VE BEHA VIOR

(4.1) The greatpowers,especiallythe superpowers, are increasingly


to modifytheirconflicts
constrained inanygivensubsystemby the
fact that they are apt to have common interestsin other

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[108] INTERNATIONAL STUDIES QUARTERLY

subsystems whichtheydo not wantto jeopardize.At the same


time,these powersare oftenhamperedin the exploitationof
commoninterests bothwithinsubsystems and on thegloballevel
by the factthttheyfindthemselves engagedin sharpconflicts in
othersubsystems (Young,1969: 338).
(4.2) Localism,fragmentation, linkages,and resourcescarcities lead to a
varietyof countryrankings and layersof international activity-
that is, to a new international hierarchy.It is based on the
international centerand local subsystem centerseventhoughthe
moretraditional aspectsof the hierarchy havebeeneroded.This
hierarchy is recognized by thecountries of theperipheries through
restrictionson theirforeign policydomainsandby thecountries of
the centerthrough theirdifferingbehavioral patternsin thevarious
subsystems of the peripheries than on the center(Dominguez,
1971: 208).
(4.3) The restrictions (largelyself-imposed obstaclesto effective inter-
vention)on thecountries ofthecenteroftendo no morethanlead
to differentialcenterbehaviorin thevariousperipheral subsystems
(Dominguez,1971: 207).
(4.4) Generally, super-and greatpowersare allowingextantsubsystems
increasinglatitude in managingsubsystemicaffairs(Bowman,
1968: 256).
(4.5) Whilethevarioussubsystems are significantlydiscontinuous,there
are nevertheless importantinterconnections betweenthem.For
this reason,it is sometimespossibleto achieveadvantagesby
utilizingpoliticalcreditand reputation acrosssubsystems (Young,
1969: 338).
(4.6) Wherethe capabilitiesof the systemare overburdened and the
systemcan no longerperform its functions,the dooris openedto
external influences (Zartman,1967: 549).
(4.7) If powerwereto be likenedto raysof light,we mightsay that
extra-areapoweris refractedwhen projectedinto a subsystem.
Policyinitiative mustbe leftto subsystem member initiative.
Great
power involvement in subsystemsmust be as unobtrusiveas
possibleif it is to be successful.In subsystems withlow levelsof
power,actionto maintainthe statusquo willbe moresuccessful
thanrevisionism (Binder,1958: 414-415).
(4.8) Intrusive powersare relatively weakin theirabilityto altersocial
cohesionand only slightlystrongerwith respectto economic
cohesion.Theirmainefforts are usuallyconcentrated on political
andorganizational cohesion(CantoriandSpiegel,1970a: 293).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson
/REGIONALSUBSYSTEM[109]

(4.9) Whenthe actorsin a subsystem distinctively lack internalcom-


munications, intrusivepowerstendto concentrate on thedomestic
arena and supportfew regionalprojects(Cantoriand Spiegel,
1970a: 378).
(4.10) Militaryaid,intervention, and subversion,whichaltermilitary and
motivationalpower,have the most directand dramaticeffects
upon the regionaldistribution of power.Economicassistance,
althoughpotentially moresignificant in thelongrun,ingeneralhas
less effecton immediateshort-run gainsin power.Accordingly,
where intrusivepowers are in competition,militaryaid and
motivational influencetend to have morepoliticalrelevance;in
areaswhereone intrusive poweris dominant or wheretheintrusive
powers collaborate,the prospectof raisingmaterialpower is
greater(Cantoriand Spiegel,1970a: 293-294).
(4.11) Intrusivepowers,as theyaffectthestructure of relations,
oftenare
as important in limiting conflictand even,on occasion,encourag-
ingcooperation as theyarein intensifying strife.Wherea dominant
intrusivepower operates,or at leastwhereintrusive powersare
allied with each other, local conflicttends to be aggravated
(CantoriandSpiegel,1970a: 294).
(4.12) Dependentupon timing,local conditions,and the degreeof
primacyof the intrusive power,the creationof socialcohesionis
extremely difficult and is consequently rare.Wherea singlepower
is dominant,cohesioncan be encouraged; wherethesuperpowers
are competing,their effectupon cohesionis largelynegative
(CantoriandSpiegel,1970a: 376-377).
(4.13) The levelof instusive competition, thelevelof local conflict, and
the degreeof subversion seemto be directly related(Cantoriand
Spiegel,1970a: 269).
(4.14) In subsystems wherea regionalpoweris biddingforprimacy, the
superpowerconfrontation recedesinto the background(Liska,
1968: 20).
(4.15) Colonieseitherserveas a focusof tensionand disputeor simply
allowthe colonialpoweran opportunity to playa surrogateroleas
a member of thesubsystem (Cantoriand Spiegel,1970a: 259).
(4.16) Small-state alliancesare at once a means for keepingthe great
powersand theirconflicts"out" of a regionand a means for
attenuating or concealing thefactofthegreatpowersbeing"in" as
the ultimatelyindispensableor unextrudablemakeweights and
safeguards (Liska,1968 7).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[11 0 ] INTERNATIONAL STUDIES QUARTERLY

(4.17) In subsystems wheretheprimacy issuehasapparently beensettled,


the trendhas been away fromrigid alignment and integration
towardless alignment withthe leadingsuperpower, leavingopen
thepossibilityfora politicallyfreerhand in conditionsshortof an
acutemilitarythreat(Liska,1968: 21).
(4.18) In the followingcontexts:(1) when international conferences
take place,(2) in determining participationin externalwars,and
(3) when internationalagreements are negotiated, when an
intrusivegreatpowertakes the the
initiative, smallactorsfollow;
the smallertheyare, the most eagerly they (Reinton,
participate
1967: 354).
(4.19) An allianceorganization thatis createdas theinstrumentality for
intrusioninto subsystems may prompt the strengthenedresistance
of rivalstate actors,therebyincreasing ratherthanameliorating
tensions(Miller,1970: 377).
(4.20) Overlapping regionalsubsystems, in whicha superpower partici-
pates,interactin sucha wayas to fostertheformation ofadvanced
formsof comprehensive regionalsubsystems excludingthesuper-
power. A transformation of overlappingregionalsubsystems
towarda typeof comprehensive regionalsubsystem thatinvolves
advancedintegration is improbableif a superpower participates.
The greaterthe integration, the slighterthechanceforparticipa-
tionofthesuperpower (Kaiser,1968: 105-106).

SUBSYSTEM-PERIPHER Y BEHA VIOR

(5.1) Relationsbetweencoreand periphery frequently takeprecedence


overand are moreimportant thanrelationswithintheperiphery.
The relationswiththe international systemtakeprecedence over
relationswiththesubsystem (CantoriandSpiegel,1969: 375).
(5.2) A subsystem's periphery centerby
is linkedto the international
actorsin thesubsystem's interaction
center.Directand significant
withinand betweena subsystem's peripheralactorsis unlikely
(Dominguez,1971: 204).
(5.3) A peripheralsector'schief characteristics-noncohesiveness and
relations-are
lack of internal compensated for,in somedegree,by
stronger bonds with the core and relationswithactorsin other
partsoftheworld(CantoriandSpiegel,1970a: 171).
(5.4) The antagonistic relationsbetweensubsystem and periphery are
one of theelementsthathelpkeep the system together(Zartman,
1967: 547).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [ 1111

(5.5) Dependentterritories playinga bufferrole suggestthat some


peripheralareasmayhavesomeof the characteristics of member-
shipin thesystem (Zartman,1967: 547).
(5.6) In theirexternal theperiphery
relations, members(especiallythose
which are isolated) often attemptto manipulatethe security
objectivesof theintrudingpowerstowardtheirownends(Cantori
andSpiegel,1970a: 171).
(5.7) Whenperipheral actorsinteractto a higherdegreethanin other
peripheries,theyare likelyto be united,at least in part,by a
commonintrusive power(CantoriandSpiegel,1970a: 171).

SUBSYSTEMIC ORIENTA TIONS AND ISSUES

(6.1) With the exceptionof a few borderlinestateswho may be


consideredto belongto two,everynation-state (no matterhow
strongor weak) is a memberof only one regionalsubsystem
(CantoriandSpiegel,1970a: 5).
(6.2) The presentinternational systemis the firsttrulyglobalinterna-
tional system:the regionalsubsystemshave only a reduced
autonomy;the"relationships of majortension"blanketthewhole
planet,the domesticpoliticsare dominatednot so muchby the
region'sproblemsas by purelylocal andpurelyglobalones,which
conspireto diverttheregion'smembers fromtheinternal affairs
of
theirarea,and,indeed,wouldmakean isolatedtreatment ofthose
affairsimpossible. As a result,each nation,neworold,findsitself
placed in an orbitof its own,fromwhichit is quitedifficult to
moveaway;fortheattraction oftheregional forcesis offset
by the
pullof all theotherforces(Hoffmann, 1966: 864-865).
(6.3) Whilethe specificissuesand the mixturebetweenregionaland
globalconcernswillvaryfromone subsystem to thenext(Young,
1969: 339), the contemporary internationalsystemis regional
subsystem-dominant because,to most actors,regionalrolesand
problemsare of considerably greaterimportance thansystemwide
issues(Holsti,1970: 291).
(6.4) Nation-states act differently as membersof a regionalsubsystem
thantheydo as members of theglobalsystem. The differencesare
greaterandof a different natureforemerging subsystems thanthey
areforestablished subsystems (Lamb andRussett,1969: 28).
(6.5) Apart fromthe most local settings,subsystems in the world's
peripheries are likelyto be linkedsubsystems, not single,unified,
undifferentiated, regional,continental, or imperialsubsystems.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[1121 INTERNATIONAL STUDIES QUARTERLY

There are linkagesfor some goals and not for others.Only the
international actorswho are engagedin the linkageprocessraise
local subsystem actorsto higherlevelsof localism.Whenthese
actorsare inactive,worldperipheral actorsrevert to themostlocal
levelsof theirforeign policies,fortheylackresourcesto do more
and theirinternational problems, opportunities, and goalsare less
thanregional (Dominguez,1971: 178, 198-199).
(6.6) Decision-makers of developing states,but not of developedstates,
perceivethe settingfor decisionsin political-diplomatic issues
primarily in termsof theirregionalsubsystem characteristicsand
pressures(Brecheret al., 1969a: 91).
(6.7) The higherthe exchangeswith the regionalenvironment in a
developing polity,the lowerthe involvement in the greatpower
cleavage(Fleming,1969: 115).
(6.8) There are discontinuities between the conceptuallyanalogous
patternsof international politicswithinthesubsystems on theone
hand and the global formatfocusingon the directrelationship
betweenthesuperpowers on theotherhand.Underthesecircum-
stances,a substantiveissueinfluenced simultaneously by severalof
thesedivergent patternsof international politicsis virtually
bound
to becomeextremely complexandambiguous (Young,1969: 344).

SUBSYSTEMIC GOALS AND ROLES

(7.1) A "benign"regionalenvironment is an objectiveof everystate.


Strategiesconduciveto the creationandmaintenance of a regional
environment congenialto a state's developmentinclude: (1)
conquestand thecreationof buffer areason its own borders;(2)
regionalalliancesso as to maintaina stableregionalbalanceof
power;(3) "splendid"isolation(holderof thebalanceof power);
(4) federative attachments; (5) allianceswithactorsoutsidethe
regionalsubsystem so as to strengthen theactor'sregional position.
(a) Allor mostactorsattempt strategy 1,butinsularstatestendto
be lesssuccessfulin theiruse ofit.
(b) Insularstatesare lesslikelythancontinental actorsto pursue
strategy2.
(c) Insularstatesare morelikelyto followstrategy 3, provided
thattheregionalenvironment conforms to the classicbalance
of powersituation.
(d) Insularstatesarelesslikelyto pursuestrategy 4.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
/REGIONALSUBSYSTEM[113]
Thompson

(e) Insularstatesare morelikelyto pursuestrategy 5 (Holt and


Turner,1969: 218).
(7.2) Local subsystem actorsremainrelatively able to findsupportfor
theirlocal possessiongoals(goalswhichaimat theenhancement or
thepreservation of one or morethingsto whichthestateattaches
value),buttheyare also relatively unableto gainsupportforsuch
goalsbeyondtheirlocalsubsystems (Dominguez,1971: 198).
(7.3) The originsof the role of regionalleader (a national role
conceptionreferring to dutiesor specialresponsibilities that a
government perceivesfor itselfin its relationto statesin a
particularregionwithwhichit identifies) includesuperior capabil-
itiesandtraditionalnationalrole(Holsti,1970: 261, 296).
(7.4) The originsof the role of regionalprotector(a nationalrole
conception whichmayimplyspecialleadership on a
responsibilities
regionalbasisandemphasizes thefunctions ofproviding protection
for adjacentregions)includeperceptionof threat,geographic
location,traditionalpolicies,and needsof the threatened states
(Holsti,1970: 261-262,296).
(7.5) The originsof the role of regionalsubsystemcollaborator(a
nationalroleconceptionreferring to far-reaching commitments to
cooperative effortswithotherstatesto buildwidercommunities)
include economic needs, sense of "belonging"to the region,
commonpolitical-ideological traditions,and geographiclocation
(Holsti,1970: 265, 296).
(7.6) Negativemembersof a regionalsubsystem (e.g.,Israeland Union
of South Africa)are followersof the regional"lowerway"-i.e.,
theirdecisionmakersaccept wholeheartedly what constitutes a
"negativeidentity"in formalized reactionto thestandards of the
dominantnormof theirregion.These membersperform at least
two importantfunctions:(1) theyprovideone of the unifying
themesin regionalsubsystem politics;and (2) theytakesomeof
thepressure frommembers of thesubsystem who deviatein some
matters butwho remainbasicallysimilarto theirneighbors in the
rest.Pressuresforsolutionofregionaldisputeswithinthe"family"
are increasedby a desireto reaffirm solidarity againstnegative
members (LambandRussett,1969: 14).

INTERSUBSYSTEMIC INTERACTION

(8.1) Relationsappear to be closestwithinregions,while relations


among regionsappear to be relayedthrough"representative"

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[114] INTERNATIONAL STUDIES QUARTERLY

nationsforeach region.That is, a singlenation,usuallythemost


highlydifferentiated withinthe region,acts as an intermediary
betweenthe nationsof thatregionand otherregions(Bernstein
andWeldon,1968: 173).
(8.2) As a nation becomes more differentiated and more relatively
centralto its geographic region,it will becomemorearticulated
with nations in other regions(Bernsteinand Weldon, 1968:
174-195).
(8.3) The level of politicaland economic transactions remainshigh
withinlocal subsystems and withthe international centerbut low
across local subsystemic boundariesin the world's peripheries
(Dominguez,1971: 198-199).
(8.4) The international actors(centerpowersor regionalorganizations)
whichlinksubsystems in theworld'speripheries aretheonlyactors
which are common to the linked subsystems. They are full
membersof the subsystemswhich they have penetratedor
encompassed (Dominguez,1971: 199).
(8.5) Comprehensive regionalsubsystemsweaken otherregionalsub-
systemsof all threevarieties(a) if theyhurtthe interestsof
membersof otherregionalsubsystems, (b) if theyare reasonably
strongand offerrewardsforparticipation, and (c) if participation
in or accessto the benefitsof the comprehensive subsystem is an
optionthatis reasonablyopen to indiVidual members oftheother
subsystem (Kaiser,1968: 101).
(8.6) In contrastto cooperative organizations(organizations designedto
aid in settlingmembership disputesor to presenta common
militaryand perhapsdiplomaticfrontagainstan outsideactoror
actors),alliancegroupings are morelikelyto increasetheinterrela-
tionshipsthat statesof one subsystemmay have withthoseof
anotherwithoutpromoting greatercohesiveness in the qualityof
those relationsin the process.The long-range effectof such a
situationmaybe to undermine theeffectiveness ofthecoalitionas
a single actor in the world arena even while intersubsystem
relationsincreasein scopeandnumber(Miller,1970: 362, 377).
(8.7) Comprehensive regionalsubsystems tendto reinforce each other
(a) if thereexistsa mutualinvolvement betweenthe societiesof
the two subsystems, (b) ifregionalprocesseswithinone subsystem
hurtthe interests theother,and(c) ifthereis no realpossibility
of
ofparticipating in thesubsystem (Kaiser,1968: 100).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson/ REGIONAL SUBSYSTEM [ 115]

CONCL USION

As initiallynoted, this paper had two fairlymodest objec-


tives: (1) to constructa synthesized and standardizeddefinition
of the regional subsystem, and (2) to make available an
unprocessed inventoryof propositionsconcerningregional
subsystembehavior.The firstgoal was formallyachieved,but
its fullrealizationis dependenton itsreceptionand consequent
acceptance or rejection.Success in termsof the second goal is
moredifficultto evaluate.The inventory's contentsare certain-
ly a mixed bag. Due to the conceptual disparitydiscussed
earlier,not all of the generalizations may belong in the same
collection.A few are data-based, while manyothersneed to be.
Most of the propositionsderivefromthe study of a single
subsystemand will requirereplicativeand comparativestudies
in other subsystems.What the inventory,by no means an
exhaustiveone, revealsmost clearlyis the immaturity of the
study of regionalsubsystemsand, in causal consequence,the
great lack of theorycapable of explainingand relatingthe
existing generalizations.In the regrettableabsence of the
essentialtheoretical"glue," regionalsubsystem analystsmustbe
preparedto show a greaterconcernformakingtheirinvestiga-
tionsmoreof a joint ventureby speakingthe same dialectand
by exchangingtheirfindings. Thispaperwas writtenin orderto
prod analyststoward this state,foronlythenwillit be possible
to accumulatea genuinebody of verifiedknowledge.And only
thenwillregionalsubsystemanalystsbe preparedto contribute
somethingto the assemblingof the puzzle box knownas world
politics.

REFERENCES

ARON, R. (1966) Peace and War: A Theoryof InternationalRelations (R. Howard


and A. B. Fox, translators).New York: Praeger.
BANKS, M. (1969) "Systems analysis and the study of regions." International
Studies Q. 13 (December).
BERNSTEIN, R. A. and P. D. WELDON (1968) "A structuralapproach to the
analysisof internationalrelations."J. of ConflictResolution12 (June).
BERTON, P. (1969) "Intemational subsystems:a submacro approach to interna-
tional studies."InternationalStudiesQ. 13 (December).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
[116] INTERNATIONAL STUDIES QUARTERLY

BINDER, L. (1958) "The Middle East as a subordinateinternationalsystem."World


Politics 10 (April).
BOWMAN, L. W. (1968) "The subordinate state system of Southem Africa."
InternationalStudies Q. 12 (September).
BRECHER, M. (1963) "Internationalrelations and Asian studies: the subordinate
state systemof SouthernAsia." WorldPolitics 15 (January).
B. STEINBERG, and J. STEIN (1969a) "A frameworkfor researchon foreign
policy behavior."J. of ConflictResolution 13 (March).
---(1969b) "The Middle East subordinatesystemand its impacton Israel's foreign
policy." InternationalStudies Q. 13 (June).
CANTORI, L. J. and S. L. SPIEGEL (1970a) The InternationalPolitics of Regions: A
ComparativeApproach. Englewood Cliffs,N.J.: Prentice-HalL
--- (1970b) "The internationalrelationsof regions."Polity 2 (Summer).
---(1969) "International regions: a comparativeapproach to five subordinate
systems."InternationalStudies Q. 13 (December).
DOMINGUEZ, J. E. (1971) "Mice that do not roar: some aspects of international
politicsin the world's peripheries."InternationalOrganization25 (Spring).
FLEMING, W. G. (1969) "Sub-SaharanAfrica: case studiesof intemationalattitudes
and transactionsof Ghana and Uganda," in J. N. Rosenau (ed.) LinkagePolitics:
Essays on the Convergenceof National and InternationalSystems. New York:
Free Press.
HAAS, E. B. (1970) "The study of regionalintegration:reflectionson the joy and
anguishof pretheorizing."InternationalOrganization24 (Autumn).
HAAS, M. (1970) "Internationalsubsystems:stabilityand polarity."Amer.PoL Sci.
Rev. 64 (March).
HANSEN, R. D. (1969) "Regional integration:reflectionson a decade of theoretical
efforts."WorldPolitics 21 (January).
HELLMANN, D. C. (1969) "The emergence of an East Asian intemational
subsystem."InternationalStudies Q. 13 (December).
HEMPEL, C. G. (1952) "Fundamentals of concept formationin empiricalscience,"
in InternationalEncyclopedia of UnifiedScience 2, 7. Chicago: Univ.of Chicago.
HODGKIN, T. (1961) "The new WestAfricanstatesystem."Univ. of Toronto Q. 31
(October).
HOFFMANN, S. (1966) "Obstinate or obsolete? The fate of the nation-stateand the
case of WesternEurope." Daedalus (Summer).
--- (1963) "Discord in community: the North Atlantic area as a partial
internationalsystem."InternationalOrganization17 (Summer).
HOLSTI, K. J. (1970) "National role conceptions in the studyof foreignpolicy."
InternationalStudies Q. 14 (September).
HOLT, R. T. and J. E. TURNER (1969) "Insular polities," in J. N. Rosenau (ed.)
Linkage Politics: Essays on the Convergence of National and International
Systems.New York: Free Press.
KAISER, K. (1968) "The interactionof regional subsystems."World Politics 21
(October). Politics 21 (October).
(1967) "The United States and the EEC in the Atlanticsystem:the problemof
theory."J. of Common MarketStudies 5 (June).
KALLEBERG, A. L. (1966) "The logic of comparison:a methodologicalnote on the
comparativestudyof political systems."WorldPolitics 19 (October).
KAPLAN, M. A. (1957) System and Process in InternationalPolitics. New York:
JohnWiley.
LAMB, W. C. and B. M. RUSSETT (1969) "Politics in the emergingregions,"in Wm
Isard and J. Wolpert (eds.) Peace Research Society Papers 12, Cambridge
Conference,1968.

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions
Thompson / REGIONAL SUBSYSTEM [ 1 17]

LISKA, G. (1968) Alliances and the ThirdWorld. Baltimore: JohnsHopkins Press.


McCLELLAND, C. A. (1966) Theory and the InternationalSystem. New York:
Macmillan.
MILLER, L. H. (1970) "Regional organizationsand subordinatesystems,"in L. J.
Cantori and S. L. Spiegel (eds.) The Intemational Politics of Regions: A
ComparativeApproach.Englewood Cliffs,N.J.: Prentice-Hall.
MODELSKI, G. (1961) "International relations and area studies: the case of
South-EastAsia" InternationalRelations 2 (April).
PRZEWORSKI, A. and H. TEUNE (1970) The Logic of ComparativeSocial Inquiry.
New York: Wiley-Interscience.
REINTON, P. 0. (1967) "Internationalstructureand intemationalintegration:the
case of Latin America."J. of Peace Research4.
RUSSETT, B. M. (1967) InternationalRegions and the InternationalSystem: A
Study in PoliticalEcology. Chicago: Rand McNally.
--- and W. C. LAMB (1969) "Global patternsof diplomaticexchange,1963-64." J.
of Peace Research 1.
SCHWARTZMAN, S. (1966) "International development and internationalfeu-
dalism: the Latin American case," in Proceedings of the InternationalPeace
ResearchAssociation: InauguralConference.Assen,Netherlands:Van Gorcum.
SHEPHERD, G. W. (1970) Nonaligned Black Africa: An InternationalSubsystem.
Lexington,Mass.: D. C. Heath.
SIGLER, J. H. (1969) "News flow in the North Africaninternationalsubsystem."
InternationalStudies Q. 13 (December).
SINGER, J. D. (1969) "The global system and its subsystems:a developmental
view," in J. N. Rosenau (ed.) Linkage Politics: Essays on the Convergenceof
National and InternationalSystems.New York: Free Press.
THOMPSON, W. R. (1970) "The Arab sub-systemand the feudal pattern of
interaction:1965." J. of Peace Research 2.
YALEM, R. J. (1970) Regional Subsystemsand WorldPolitics.Universityof Arizona
Instituteof GovernmentResearch4.
YOUNG, 0. R. (1969) "Political discontinuitiesin the internationalsystem,"in J. N.
Rosenau (ed.) InternationalPoliticsand ForeignPolicy. New York: Free Press.
--- (1968) "A systematicapproach to internationalpolitics." PrincetonUniversity
Centerof InternationalStudies ResearchMonograph33.
--- (1964) "The impact of generalsystemstheoryon political science," in L. Von
Bertalanffy and A. Rapoport (eds.) GeneralSystemsYearbook 9.
ZARTMAN, I. W. (1967) "Africa as a subordinate state systemin international
relations."InternationalOrganization21 (Summer).

This content downloaded from 91.229.229.212 on Thu, 12 Jun 2014 23:52:22 PM


All use subject to JSTOR Terms and Conditions

You might also like