Review: What is Nationalism and Why Should We Study it?

Author(s): Ernst B. Haas

Source: International Organization, Vol. 40, No. 3 (Summer, 1986), pp. 707-744
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/2706824 .
Accessed: 03/07/2011 16:41

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=mitpress. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit 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.

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to International
Organization.

http://www.jstor.org
What is nationalismand whyshould
we studyit? ErnstB. Haas

Benedict Anderson. Imagined Communities:Reflectionson the Origin

and Spread of Nationalism. London: Verso Editions & NLB, 1983.
ErnestGellner.Nationsand Nationalism.Ithaca:CornellUniversity
Press,1983.
Dudley Seers. The Political Economy of Nationalism. New York: Oxford
Press, 1983.
University
AnthonyD. Smith.Nationalism in the TwentiethCentury.New York:
New YorkUniversity
Press, 1979.

Nationalismand modernization

Most essays on nationalism beginwiththe lamentthatthe conceptis as

fuzzyas thestatesofminditis supposedto describearediverse.Studiesof
nationalism pose theproverbial elephantproblem:theanimal'sappearance
seemstodiffer depending on whereitis touchedbya groupofblindpersons.
Our authorsare no exceptionto thisrule. For DudleySeers nationalism
refersto certaintypesofeconomicpolicy,whileforBenedictAndersonthe
termconnotesmanufactured linguisticidentity.AnthonySmithconsiders
nationalism to be a particularideologyof solidaritybased on preindustrial
roots.ErnestGellnertreatsthephenomenon as a distinctly princi-
industrial
ple of socialevolutionand social organization.
No singletermshouldbe madeto bear so heavya burden.As Arthur N.
Waldronnotedin a reviewparalleling myeffort:

I gratefully
acknowledge thehelpful receivedfromRobertH. Jackson,
comments Kenneth
Ian Lustick,andWayneSandholtz.
PeterKatzenstein,
Jowitt,
InternationalOrganization40, 3, Summer1986
Institute
C 1986bytheMassachusetts and theWorldPeace Foundation
ofTechnology
708 International
Organization

A disproportion existsbetweenthewayin whichnationalism

definitely
is used to explainhistoryand politics,and theexplanatory capacityof
theconceptas outlinedby itstheorists. The intellectual
foundations
providedby thelattersimplywillnotsupporttheexplanatory struc-
turesthathave been placedon them.Nationalism in generalis a power-
fuland comprehensible idea. Yet, whileit definesgeneralsituations,it
is notveryusefulin explicating specificevents.In cases wheresuch
eventshave in thepastbeen explainedby invoking nationalism, we will
have to searchforanotheranalysis.1
unlesswe are carefulto specifywhichaspectsof"specific
He is quiteright,
events"we are interested withthehelpof theconceptand
in explicating
which"generalsituations"theconceptis supposedto capture.No concept
cancovereverything. As ofnow,theelephantlumbers aroundwithout doing
muchusefulwork.
The studyof nationalism has verypermeableboundaries.It overlapsso
heavilywiththestudyof modernization, ofmodernpoliticalideologies,of
economicand social history,and of politicalanthropology as to suggest
redundancy. Whystudynationsand nationalism at all? For me thereis no
self-evidentanswer.The description of historicalpatterns requiresno spe-
cialfocus,no theory,apartfromandbeyondthetheoriesalreadyembedded
in therelevantdisciplines.A specialfocusis justifiedonlyifthepurpose of
theinquirysomehowdiffers fromwhatwe alreadyknow.

Nationalismand politicalrationalization

My purposein studyingnationalism is to exploreits role as a typeof

whichhelps or hindersdomesticand international
"rationalization" har-
mony.Thispurposeenablesme to sidesteptheconceptualconfusion com-
mon to mostof the authorsunderreviewand to attemptthe task that
Waldronoutlines.Like mostconceptswe use in thesocialsciences,nation
and nationalismare cognitive we inventto markoffan intellectual
artifacts
universe.Mydefinitions,whichwillfollowlater,haveno moreinnatevalid-
itythananyoneelse's; theyare justifiedonlyby mybasic purpose.My

1. Arthur N. Waldron,"TheoriesofNationalism andHistorical

Explanation,"WorldPoli-
tics 37 (April1985),p. 427.I agreewithWaldron'scoreargumentthatthevagueinvocation
of
"nationalism" in explainingeventsin thenon-European worldis unsatisfactory,becausethe
has beenattachedto people,movements,
"adjective'nationalist' andsentiments ina waythat
is taken(usuallywithout each ofthemmeaningfully
as distinguishing
explanation) fromsome
othervariety."Indeed,as he says,struggle comesfirst, and thennationalism, andto under-
standwhythereis a struggle we mustunderstand is
itspoliticalsource(p. 433).That,however,
notwhatinterests me.I am concerned witha singlegeneralsituation-patterns ofrationaliza-
tionin thepost-Enlightenment world-and manyspecificevents,i.e., thebehaviorof self-
identified
groupsincopingwiththatworld.Nationalism is one(andonlyone)wayinwhichsuch
groupsdo identify themselves.No moregeneralclaimis suggested.
 Whatis nationalism?709

demarcation beginswitha roughdescription ofwhatI takenationalism to

A
mean. statement on the meaning of rationalizationfollows.The twoare
thenconjoinedin an attemptto inquireintotherelationship betweenhar-
monyat the as
national, opposed to the international,levels.
Nationalism is theconvergence ofterritorial andpoliticalloyaltyirrespec-
tiveof competing fociof affiliation, such as kinship,profession, religion,
economicinterest, race, or even language.Nationalismis "modern"be-
cause it stressesthe individual'ssearchforidentity withstrangers in an
impersonal world,a worldno longeranimatedby corporateidentities. All
nationalisms implya principleofidentity based on impersonal ties,remote
ties,vicariousties-all ofwhichare mediatedbya setofcommonsymbols
embeddedin a certainpatternof communication. Successfulnationalism
also impliesa minimum ofsocialharmony, an acceptanceofthevaluesthat
thesymbolscommunicate sufficientto maintainsocial peace and peaceful
socialchange.Legitimate authorityunderconditions ofmasspoliticsis tied
up withsuccessfulnationalism; whenthenationalidentity is in doubt,one
propsupporting legitimacy is knockedaway.
Nationalism is also a civilreligion, oftenin conflictwithbutoccasionally
drawing strength fromrealreligions. Thatcivilreligion containsa setofcore
valuesthat,whether forobjectivist or subjectivistreasons,cometo be ac-
ceptedbythepopulation ofa state;theybecomethedefiners ofselfhood.In
successfulnationstheyremainin thatrole untilchallengedby the next
sourceoftension;no civilreligionis gravenin stone.As longas thecore
valuesprovidetheframework forsocialaction,peopleknowwhatto expect
oftheirfellows,understand and respectauthority, are secureintheirviews
of theschemeof collectivelife.Such a societyis temporarily rationalized
despiteits size, impersonality, and vicariousnatureof impersonal ties.As
Webertaught us, rationalization neednotbe thebureaucratic kind(though it
usuallyis); butitmustbe in a formula thatfitstheconditions ofa commer-
cializedand industrialized society.The kindof identity we seek to under-
standis an issue onlysincetheonsetoftheindustrial revolution.2
The crispestformula forsumming up therelevantconception ofrationali-
zationwouldrunlikethis:rationalization refersto modesofbehaviorthat
reston a materialist ontology, a proceduralepistemology, andan empirical

2. Theconceptofrationalization,ofcourse,is adaptedfrom MaxWeber.Thebestdiscussion

of Weber'softenconfusing and contradictory treatment of theconceptis StephenKalberg,
"Max Weber'sTypesof Rationality: Cornerstones fortheAnalysisof Rationalization Pro-
cesses in History,"AmericanJournalof Sociology85 (March1980).Kalbergis especially
helpfulin showingtherelationship Weberestablished betweenhis fourtypesof rationality
(theoretical,
practical,formal,substantive)and thefourtypesof social action(traditional,
affectual, means-end
value-rational, rational)(p. 1161).Strictlyspeaking,we are concerned
withthetypesofsocialaction.Whatmatters fornationalism is thatitcombines,initsvarious
types,all fourtypesofsocialactioninvariousvolatilemixtures thatcallintoquestionthefinal
victory ofanyrationalization
formula.I amindebted to KennethJowitt forthecrispdefinition
ofthecombinedvaluerationallend-means rationalformula ofsocialaction.
 Organization
710 International

methodology. Magicaland prescientific views of the natureof thingsare

banished;causal connections amongphenomenaare sought,notassumed;
ends-means chainsare subjectto verification bystandardized methods.An-
dersonandGellneroptformuchthesame idea in distinguishing themodem
worldofnationalism fromthe past. Weber's premodern rationalizationfor-
mulasandtheassociatedformsofsocialaction remain relevant even though
onlythe "formal-rational" typeis giventhepivotalrole. Nationbuilding,
infusinga sense of nationalidentity, depends,in myargument, on thevic-
toryof thelegal-rational formoveritspotentialcompetitors. The factthat
thisvictory may,inpractice,neverbe completegivesus thestuff ofstudies
ofnationalism.
Ourtaskis theexploration ofhowandwhya vicariousprinciple oftrans-
personalidentification can give shapeand orderto a societyunderstress.
The underlying idea is thatnationalism can hold a societytogether while
peoplearebeingbuffeted bythestrainsofmodernization. Rationalization by
wayofnationalism, ofcourse,can taketwoforms:peopleunderstresscan
seektoresolveitbyidentifying withtheexisting state,buttheycanalso look
forhelp by secedingfromit. Each courseis predicatedon principlesof
rationalchoice. The desireto live in an orderedsocietywithpredictable
rulesthatsustainone's demandscalls forreciprocalactionson thepartof
one's fellowcitizens.Whether one identifieswiththeexistingstateor not,
thechoiceimpliesthedeliberatesearch for links ofinterest and valuewith
otherssimilarly situated. The "bargaining" involved in the searchevokes
thespecification ofcorevaluesoforder and predictabilityfor thecollectiv-
ity.Each "bargain"is based on theexpectation benefits come,but
that will
notnecessarily at thesametimeforeverybody. Someactormaywellhaveto
delaygratification in theshortrunin orderto gainacceptancefora set of
rules,whichwillproduceotherbenefits fortheone whomakestheconces-
sionsat a latertime.Eventually, ofcourse,thatperson'sor party'sprefer-
ences also have to be met.
A rationalizedsocietyis a societythatordersitselfon thebasis ofrecip-
rocalexchangerelationsamongitsmembers. Its members accepta common
normofbasic fairness.Theypracticecontingency, thatis, theyexpectthat
goodbehavioris rewardedwithgood,and bad withbad. In addition,these
featuresdependon the further expectationthatrewardscan be delayed,
thoughnotindefinitely, and thattrulyequivalentconcessionsin socialbar-
gainingare not necessary.3The bargainsreferto the rearrangement of
wealth,status,andpowerwhich characterizes the modem world. A society
is rationalizedif it managesto practiceinternalbargaining thatresultsin
theserearrangements withoutblowingitselfapart;it practicesreciprocity

inInternational
owesa lottoRobert0. Keohane,"Reciprocity
3. Myreasoning Relations,"
InternationalOrganization40 (Winter1986).
 Whatis nationalism?711

whenitresolvesinternal strainbycontinuous adaptation.

Nationalism refers
to theparticularlegitimating principle thatmakessuchadaptation possible,
theperhapsuniqueprincipleeach modernsocietyseeksto agreeon forits
memberswhilealso markingitselfofffromothersocieties.The formal-
rationalvariantofthisconstruct is byno meanstheonlypossibleone,butit
maywellbe themostsuccessfultype.4
Now comestheparadox.Is a worldmadeup ofrationalized societies-of
successfulnation-states-aninherently bellicoseworld?Since thevarious
nationalselfhoods arearrivedat bymutualexclusivity andoutright hostility,
theanswerwouldseemtobe yes.Is nationalrationalization thereforeincon-
sistentwithinternational harmony? Is itimpossibleto conceiveofan inter-
nationallyrationalizedsociety?The paradoxcan be resolvedonlyif we
breakopen thebox labeled"nation"and inquireaboutthevariousbeliefs
thatmakeup thecivilreligion at variouspointsinitshistory. Ifwe discover
someevidenceofan evolutionary pattern,theparadoxcanbe laidtorest.In
short,a dominant concernwithinternaland international harmony (or its
absence), under modernconditions,justifiesa scholarlyconcernwith
nationalism despiteitspermeableintellectual boundaries.
Spottingsuch an evolutionary patternis not easy. Whattimehorizon
shouldone adopt,a hundredyearsor twenty?One mightsurmise,as the
literature of"social turbulence"suggests, thata breakpoint forthesuccess-
fulnation-state is reachedwhenno additionalinternal deals can be struck.
Domesticlegal-rational formulas forfurther adaptationwhichsatisfy impor-
tantgroupsmayno longerbe possiblewhena certainthreshold ofinterna-
tionaltechnological and economicinterdependence is crossed,thoughwe
cannotspecify thethreshold.Atthatpointwe maybe entitled tospeakofthe
obsolescenceof nationalism and thenation-state.
It is thisreasoningthatleads me to thestudyofnationalism: to discover
howrationalization maycome aboutin timesofrapidsocialchangeand to
explorethelimitsof nationalism as a successfulrationalizer.

4. In Weber'sworkthe "formalrational"typeis usuallyjuxtaposedto othertypesof

rationalization
whichfeaturetraditionandreligion as corefeatures oflegitimate authority.One
wayofreading
(overlysimplified) Weberis totreat"rational"principles ofsocialorderas flatly
opposedtoreligious ones.Does itfollowthatreligionis totallyincompatible withrationalization
thatrelieson nationalism?TerranceG. Carrollexaminedfourideal-typical nation-states
(lib-
eral,Marxist,socialdemocratic,andconservative) inordertodetermine whether eachis ableto
practicethekindofadaptation I discuss,ifitspopulation is strongly
identified withone ofthe
majortraditionalreligions.He concludesthatnationsstrongly identifiedwithliberaland/or
Marxistvaluescannotat the same timepracticeadaptationwhiletraditional religionshold
sway.He confirms thestrongversionofWeber'sthesis.However,he also confirms theweak
versionby showingthatmodernization can be compatible withShi'a Islamand Catholicism.
SunniIslamoffers moredifficulties,
Hinduism is neutral,whileBuddhism is heldto be incom-
patiblewithmodernization.Carroll,"Secularization andStatesofModernity," WorldPolitics
36 (April1984).
 Organization
712 International

Embeddedconventionsin discussionsof nationalism

My purposein studyingnationalismis not everybody's.The studyof

nationalism has beenso elephantine becausestudents acknowledge no com-
monpurpose.Instead,theyhave been dividedby a numberof underlying
conceptualdichotomies. These also characterize thefourbooks underre-
view. As long as these dichotomiesmonopolizescholarlyattention, the
elephantproblemwill remainwithus. Beforecommenting on the books
themselves, I summarize theunresolveddebate.
First,thereis the naggingquestionof whethernations,thebeliefsthat
inspiretheircitizens,thepoliciesthatderivefromthebeliefs,are good or
bad. In the 19thcentury liberalsadvocatednationalism as progressive and
Marxistsdecriedit as reactionary; in ourera therolesare reversedup to a
point.LiberalsstilldenounceGaullistsforbeingnationalistic; butMarxists,
whileagreeingwiththisjudgment, laud thenationalism of LatinAmerican
opponentsof dependencyand of Africanfreedomfighters. Yet theyde-
nouncethe nationalism of Mussolini,Hitler,and Tojo, of the Argentine
military juntas,of MargaretThatcherand RonaldReagan.Whyis one ex-
pressionofsocialsolidarity bad andanothergood?The qualityofthehistor-
ical processisjudgedin termsoftheoutcomesitproducedoris expectedto
produce.The issue,therefore, is notnationalism as such,butcapitalism as
opposedto socialism,democracy,and popularparticipation as opposedto
authoritarian rule.
Ideologyconfusesthe issue, not onlyin the sense thatthe ideological
preferences ofstudentsdefinethephenomenon ofnationalism, butbecause
scholarsdo not agreewhetherthe phenomenon refersto thebeliefsof a
movement or a party,or whetherit is theproperty of an entire"people"
(whoeverthatmaybe). Forone setofscholars,Anthony Smithamongthem,
nationalism is an ideologythatcompeteswithliberalism, socialism,fascism;
it is theproperty of movements in conflictwithothermovements. For an-
othergroup,however,nationalism sidestepsor subsumesotherideologies
by focusingon what a givenunit-a "people" or a "people claiminga
state"-believes of itselfin distinction to otherunits.Andersongoes to
greatlengthsto show how Marxistscan be good nationalists once they
transcendrigidclass analysis.Scholarsfavoringnationalism-as-ideology
thengo on to comment on particularpoliciesrather thansocialprocesses,as
does Seers.
But confusionis by no meansbanishedifwe optforthenationalism-as-
group-identity formula.We mustthenconfront the deep divisionamong
thosewho findan "objective"basis forthisidentity, as opposedto those
whosee identity in self-consciousactsofindividual the"sub-
identification,
jective" basis of nationhood.The objectivistschool definesa nation(in
Stalin'sfamousformulation) as "a historicallyevolved,stablecommunity of
language,territory, economiclifeand psychological makeup,manifested in
 Whatis nationalism?713

a community ofculture."Unlessa givenpeoplepossessesall thesevirtues,

it cannotlegitimately be considereda nation.ErnestRenan,on theother
hand,coinedhisfamous"dailyplebiscite"as thedefiner ofthesubjectivist
approach.People chooseto identify witha givenunitmovedbyincentives
and disincentives;exit, voice, and loyalty,as AlbertHirschmansays,
definetheoptions,an approachalso favoredby Gellner.
Most writerson nationalism liketo workwithan objectivist formula of
somekind.For someitprovidesa handlefordistinguishing thegoodnations
fromthebad. For others,however,a preference forstructural theorizing
overcognitive andvoluntaristic constructs accountsforthechoice.A struc-
turalistlikes to explaina "nationalist"policyresponseas a reactionto
deeplyembeddedstimulilocatedin theinternational economicdivisionof
labor;voluntarists preferto findtheirexplanations in domesticupheavals.
Whateverthereason,thechoiceofformula entailsfurther controversy.
Suppose we opt fornation-as-group identity.How did thiswe-feeling
originate?The literature is dividedinitsansweramongthosewhourgesome
primordial ties-usually labeledcultureifthewriteris skepticalaboutthe
entirecatalogofobjectivist criteria-andotherswhoarefascinated withthe
artificialityof thisfeeling.The second school stressesthe manipulative
policiesused by statesto inculcateit,theimportance ofmarkets and com-
munication channelsinfacilitating it,andthesheeraccidentsthatresulted in
a givenregionwinding up in one nationratherthananother.The distinction
has additionalimplications. ArnoldToynbeeremarkedthat"that havoc
whichthe applicationof the westerninstitution of 'nationalstates' has
workedin [Africaand Asia] whereit is an exoticimportis incomparably
greaterthan the damage thatthe same institution has done in Britain,
France,and theotherwestEuropeancountries inwhichithas been,notan
artificially
introduced innovation, buta spontaneous growth.""Early" (i.e.,
WestEuropeanliberal)nationalism is natural;"late" nationalismis artificial
inadditionto beingtaintedbyauthoritarianism. Primordial-culturalfeatures
legitimatetheearlyWesternnations;theabsenceofsuchfeatures deniesthe
restoftheworldthesamelegitimacy.
Thewideacceptanceofthesedichotomies has takenitstollinexplanatory
rigorand normative prediction. Concernoverwhichentitiesare or should
become nationshas resultedin a scholarlyemphasison nationbuilding
(sometimesconfusedwithstatebuilding)at the expenseof treatments of
nationmaintenance. The literature takesforgranted thatold and successful
nationalisms providetheproperexplanatory model.Newerentities thatare
notyetsuccessful nationsarestudiedintermsoftheirabilitytoliveup tothe
model;and since theylack the objectivequalitiesurgedby some, their
successis heldto be in doubt.
Commitment to thisdichotomy compelsthestudent to dealwiththepres-
entexclusively intermsofthepast.It neglectsthesearchfornovelmodesof
manipulation byelitesthatmight bringaboutnationhood despitetheabsence
 Organization
714 International

ofthehistorically validatedconditions. In addition,it stackstheenergiesof

scholarsin the directionof explainingthe rise of nationhoodbut not its
decline.Fortunately, thegrowthof "micronationalisms" in Europeduring
the1970shas oncemorerivetedtheattention ofsomescholarson thefragil-
ityoferstwhile successfulnations.Thisphenomenon forcesus to studywhat
went"wrong"withFrenchand Britishnationalism to makepossiblethe
growthof movementschallenging Studiesof the policies,at-
its finality.
titudes,and processesthatmaintain a nationbecomesalient.
All thismeansthattheoutcome thedailyplebisciteremainsin doubt.
of
Variousclaimantsto nationalist legitimacy, whether fortheentirepopula-
tionofthestateor for some region within it,continue tocompete.National-
istsstruggleagainstcosmopolitan visionsas well.Andthereis goodreason
to supposethatthesestruggles overlapwiththeordinary ideologicalcompe-
titionof socialistsagainstliberals,secularistsagainstthereligiously com-
mitted,authoritarians againstpopulists.

Four authorsin search of a singletheme

How do ourfourauthorsrespondto theprevalenceof thesedichotomies?

Do theymakeefforts to transcendthemandto recognizetheeternally prob-
lematicnatureofidentity formation?In manywaystheydo. Allfourauthors
areoutto demystify theconcept,to demonstrate thatclaimsto "authentic"
nationhood are untenable, thatthereare no "true"or "legitimate"nations.
Thatis a stepforward. Atleastitmakespossibletheintegration ofnational-
ism withstudiesof modernization and withchangesin intercultural and
international evenifnoneofourauthorscarrythisintegration
relations, very
far.Finally,all fourauthorsconcedethatnationalism maywellbe a rational
responseto certainsocial upheavalsand frustrations, nota throwback to
barbarism.This insightmakes it possibleto treatthe phenomenon as a
speciesofrationalization. Andersonsomewhat
Gellnerdoes thisexplicitly,
hesitantly.Seers wouldsee his workin thesame spirit.But Smithwould
stand the argumenton its head and equate nationalismwith failed
rationalization.
Thatis thegood news.The bad news,however,is thatourauthorsfailto
makeany seriouseffort to acknowledgeor use, leave alone integrate,the
plethoraof existingworkon the subject.Theywrite,each fromhis own
perspective,as ifno previousworkhadbeendoneon thedynamics ofsocial
solidarityand fragmentation. One cannothelp wondering, since all four
authorsare British,whethera smattering of nationalismkeptthemfrom
usingprevioustheoriesand empiricalstudieslargelyof Americanprov-
enance,or at leastmethodological inspiration.
 Whatis nationalism?715

The mostglaringomissionis the neglectof Karl Deutsch. Anderson,

Seers,and Smithcouldhaveenrichedtheirtreatments iftheyhadacknowl-
edgedDeutsch'streatment of therelationship
betweensentiments of iden-
tityand massive socio-economicchange. Gellner'sneglectof Deutsch
is even moreseriousbecause Gellner'simpressive theoryis based on the
same cybernetic assumptionsas Deutsch'sand makescopioususe of the
Deutschianmobilization-assimilation
balance/imbalance withouteverusing
the label. Smithfailedto take advantageof the exhaustivestudiesof
nationalistideologieswhichcontrastedtheWestern"liberal"variantwith
Easternnotso liberalones.6Anderson'streatment ofThirdWorldnational-
ism ignoresthe seminalworkof RupertEmersonand otherson the same
subject.7All theauthorsdisdaintheuse of statistical
serieselaboratedby
othersin orderto maphistoricalsequencesandpatterns.8 None makesuse
oftheelaborateand quitesatisfactorycomparativeandhistorical studiesof
nationalistthought.9

5. Karl W. Deutsch, Nationalism and Social Communication(Cambridge:MIT Press, 1953);

Nationalism and Its Alternatives(New York: Knopf, 1969).
6. The typology of nationalist
ideologiesdevelopedby CarltonJ. H. Hayesinfluenced
the
one used in thisreview.See hisEssays on Nationalism(New York:Macmillan,1926);The
Historical Evolution of Modern Nationalism (New York: R. R. Smith,1931); Nationalism: A
Religion(New York:Macmillan,1960).Amongthemanyworksof Hans Kohn,TheIdea of
Nationalism (NewYork:Macmillan, 1944)pioneered
manyofthedichotomies
thatstillcharac-
terizetheliterature.
Thesharpestjuxtaposition (early)andofnon-Western
ofWestern national-
ismis Elie Kedourie,Nationalism(London:Hutchinson,
1960),andNationalisminAsia and
Africa(New York:MeridianBooks, 1970).
7. RupertEmerson,FromEmpireto Nation(Cambridge: HarvardUniversity
Press,1960);
Self-DeterminationRevisited in the Era of Decolonization (Harvard University,Center for
International OccasionalPaperNo. 9, 1964);LloydA. Fallers,TheSocialAnthropol-
Affairs,
ogy of the Nation-State (Chicago: Aldine, 1974); CrawfordYoung, The Politics of Cultural
Pluralism(Madison:University of WisconsinPress, 1976).The classic statementof Third
Worldnational intermsofcultural
identity encountersbetweenWestern settlers
andindigenous
intellectuals
is AlbertMemmi,TheColonizerand theColonized(Boston:Beacon,1967).
8. Typicalworksinclude:Karl W. Deutsch,"Social Mobilization and PoliticalDevelop-
ment,"AmericanPoliticalScienceReview(September1961);Karl W. Deutschand William
Foltz,eds., Nation-Building(New York:Atherton,1963);LeonardW. Doob, Patriotism and
Nationalism (New Haven:Yale UniversityPress,1964);S. N. Eisenstadt
andS. Rokkan,eds.,
BuildingStates and Nations, 2 vols. (Beverly Hills: Sage, 1973); InternationalInteractions11,
2 (1984),entireissue.Publicopiniondataon issuesrelatingto nationalism
arepublished
regu-
larlyinEurobarometre (Brussels:EuropeanCommunities). Forexamplesofstudiescombining
quantitativeand qualitativetreatmentsmosteffectively see Eugen Weber,Peasants into
Frenchmen (Stanford: StanfordUniversityPress,1976),and MichaelHechter,InternalCol-
onialism(Berkeley:University ofCaliforniaPress,1975).
9. Forexample,BoydShafer,Nationalism: MythandReality(NewYork:Harcourt Brace,
1955); Louis L. Snyder, Varietiesof Nationalism: A ComparativeStudy(Hinsdale, Ill.: Dry-
den, 1976)and Global Mininationalisms:Autonomyor Independence(Westport,Conn.: Green-
wood, 1982).The problemof thecompatibility
of socialismand nationalism
is exhaustively
explored by J. L. Talmon, The Mythof the Nation and the Vision of Revolution(Berkeley:
ofCalifornia
University Press,1981),and HoraceB. Davis,Nationalism
and Socialism(New
York:MonthlyReviewPress,1967).
716 International
Organization

The puzzle overwhatnationalism

"reallyis" can be illustrated
withthe
confusionthatpermeatesSmith's Nationalism in the TwentiethCentury.It
is impossibleto tellwhethernationalism refersto an "ideal," the actual
modalbeliefsofa "movement,"a typicalhistorical patternofdevelopment,
oran inescapablestateofaffairs, becausetheauthoruses all thesemeanings
interchangeably. He says thatnationalism is "a doctrineofthehistory and
destinyofthe'nation',an entity opposedtootherimportant moderncollec-
tivitieslikethe'sect', 'state','race', or 'class'" (p. 13; italicsmine).
Smith'snotionofthe"real" nationalism-ascontrasted withotherdoc-
trinesthatmerelylook like it-is an amalgamof thewritings of Herder,
Burke,Jefferson, andRousseau:nationalism is a doctrineoffraternalethnic
solidarity,accordingto Smith.He extractsthismeaning fromthewritings of
ideologists;he is less concernedwithwhatnationsand nationalist move-
mentsactuallydo. The task Smithsets himselfis the exploration of the
compatibility ofnationalism (as an ideal,a modalbeliefsystem,a historical
pattern?) withthemajorpoliticalideologiesofthemodernera. He devotes
separate chaptersto religiousreformism, fascism,communism,pan-
Africanism, thecontemporary ethnicmovements inEurope,"beaureaucrat-
ism," and internationalism.
Religiousreformism, fascism,and communism and themodembureau-
craticwelfarestatemake use of nationalist themes,but theyare in fact
inconsistent withtherealthingand actuallyabuse it. Populist-participatory
religiousprotestmovements are otherworldly, escapist,and chiliastic;they
appealto sects.Fascismoverlapswithnationalism, butitstypicalclientele
is notthesameas thenationalist public.Fascismstressesrace,nationalism
the ethniccommunity. Fascism veneratesthe heroand the use of force,
nationalism remainsneutralon thissubject.Communism, of course,is at
bottomcosmopolitan and appealsto classes,notnations,thoughpragmatic
Communists willbendtherulesoccasionallyto advancetherevolution by
exploiting primordial ethnicfeelingsamongthecomrades.Nationalism, by
contrast, is revolutionaryand harksbackto ethnictraditions. The Western
bureaucratic abusesnationalism
state,finally, bypretending to speakforthe
entirepeople even thoughits citizensresentits impersonality, coldness,
elitism,and excessiverationality.
Becausemostpeopleseemtoprefer theemotional reassurance ofirration-
ality,Smithconsidersthecurrent movements forculturaland politicalau-
tonomyin Scotland,Wales, Brittany, Spain, and Corsica(amongothers)
to be authentic expressionsof nationalism. So are pan-Africanism and ne-
gritude.To thebenighted liberalsand Communists who stillthinkthatthe
modernworld'sincreasing interdependence impliesthedemiseofnational-
ism,Smithhas thisto say:
The veryattempt to eradicatenationalism it
actuallyhelpsto entrench
and to provokeitsperiodicresurgence,
further, and itwouldappear
 Whatis nationalism?717

to tryto livewithit,tamingitsexcesses
moresensibleand appropriate
throughmutualrecognitionsand legitimations.... More importantly,
nationalism's and appeal mustbe derivedfromtheconjunc-
persistence
tionofthethreesetsofforcesthatshapedit originally:
long-standing
thebirthofnew secularideals,and thepeculiarchar-
ethnictraditions,
ofmodernization
acteristics (p. 196)
and its social concomitants.
The veryinternational systemtodayguaranteesand legitimates nations.
Externaland internalforcesreinforce each other,ensuring thatnationalism
remainsaliveand well.
Smithwantsto separateauthentic nationalisms fromthosethatarephony.
He comes close to wantingto discoversome mysticalUr-template that
reemerges in timeofperil,liketheEmperorBarbarossa,to shakeotherand
moreephemeral ideologiesintoline.He failsto ask himselfabouttheorigin
oftheseallegedprimordial Nordoes he inquirewhysomesurvive
traditions.
and othersfadeaway. He avoidsthefactthatmodernization is notjust an
evil thattheauthentic combatsin orderto escape fromformal-
nationalist
thattheveryoriginofnationalist
legalrationality, thoughtis also associated
powerfully withthedemandforsucha rationality andretainsthisqualityin
manyplaces.Nordoes Smithrecognizethattheappealsofnationalism have
demonstrably fadedinotherplaces.Nationalism is Janus-faced:itis usedto
advocateas wellas to obstructmodernity. Solelyforthisreasonone should
not separatenationalism fromotherideologies.Insteadone shouldstudy
these ideologiesin orderto discoverwhichtypeof nationalism infuses
them.10
is thetitle.
The bestthingaboutAnderson'sImaginedCommunities
In an anthropologicalspirit,then,I proposethefollowing of
definition
thenation:it is an imaginedpoliticalcommunity-and imaginedas both
inherentlylimitedand sovereign.It is imaginedbecausethemembers of
eventhesmallestnationwillneverknowmostoftheirfellow-members,
meetthem,or even hearofthem;yetin themindsofeach livesthe
image of the communion.... In fact,all communitieslargerthan
primordialvillages of face-to-facecontact ... are imagined.... The
nationis imaginedas limitedbecause even the largestof them . . . has
ifelasticboundaries,beyondwhichlie othernations.No nation
finite,
withmankind.(pp. 15-16;emphasisin
imaginesitselfcoterminous
original)

10. In a reviewofSmith'searlierwork,Theories ofNationalism (NewYork:Harper& Row,

1972),Gale StokesarguesthatalthoughSmithsucceededin erecting themostcompleteand
sophisticated typologyof nationalist thissuccessby
ideologies,he fallsshortof exploiting
statinga theory.Stokes,as I do too,findsGellner'sThought and Change(Chicago:University
of ChicagoPress,1963)to be closerto theory.Stokesis stillconcernedwithidentifying the
featuresor attributes
thatcharacterize the"true"nationandthatdistinguish from
nationalism
otherpoliticalideologies,a taskI deliberatelyforeswear. See "The Undeveloped Theoryof
Nationalism," WorldPolitics21 (October1978).
718 International
Organization

Bravoforthiseloquentreminder thatsize,vicariouscommunication byway

of sharedsymbols,and exclusivenessmarkthenationofffromotherpolit-
ical constructs.
The author'spurposein writing thebook is important forunderstanding
theargument. He wishesto instruct hisfellowMarxistsbytellingthemthat
nationalism is not inconsistent withrevolutionary commitment, is not a
bourgeoisanachronism, and oughttherefore to be taken seriouslyas an
omnipresent historicalphenomenon withan uncertain halflife.His taskis to
explainhowrevolutionary Marxiststates(China,Kampuchea,Vietnam) can
fight old-fashionedimperialistwarsagainsteachother.This,tobe sure,may
notbe newsfornon-Marxists.
The argument goes as follows.The possibility of imagining the nation
arisesonlywhenthreeancientculturaltraitsweaken:a particular script-
languageloses its monopolyon conveyingthe truth,monarchslose the
statusofsemidivinities, and conceptions oftimecease to confound cosmol-
ogyand history. The one development mostresponsible forthebreakupof
culturesbased on theseideas was movabletypein thehandsof private-
enterprise publishersseekinga mass marketfortheirwaresamongpeople
not versedin the universalsacred language.Andersoncontinueswitha
conventional accountoftheindependence movements ofthelate 18thcen-
turyin thewesternhemisphere, linkingincipient nationalism to discrimina-
tionagainstcolonialsby themetropolitan country, yetmakinglittleuse of
theneatproposition aboutlanguageand "print-capitalism." But language
remainsthecore ingredient oftheargument he makesforEasternEurope.
Here, by the middleof the 19thcentury,"marginalized vernacular-based
coalitionsoftheeducated"(p. 78), lookingwestward, founda "model" to
be emulated.The designofthenation-state was theretobe seenandcopied.
Rulersrespondedto thepressureby converting vernacular languagesinto
officiallanguages,thesole legitimate mediumofpublicdiscoursein multi-
ethnicstates.Whilethispleasedsomecoalitionsandalienatedothers,ithad
theeffectofextending thescope ofthepoliticalcommunity byencouraging
hitherto inertgroupstoparticipate politically,
ifonlytoprotest.Atthispoint
two pathscould be taken:the further development of the modelintoa
democratic-populist one, or an alternatethatAndersonlabels "officialna-
tionalism,"inventedin Russiaand Englandand widelycopiedeverywhere
since.Official nationalismdeliberately selectskeythemesofnationhood and
foiststheseon thepopulationby appropriate policiesofeducation,recruit-
ment,reward,and punishment, alwaysmakinguse oftheofficial language.
Linkedwithnotionsofracismand appropriate economicincentives, official
nationalism becomestheimperialism practicedbyEuropeafter1870.(Yes,
dearreader,I am confusedtoo. Anderson'sargument is nota restatement of
thethesisaboutthe"good" and "early"WestEuropeanvariety ofnational-
ismas againstthe "bad" and "late" East Europeanvariety, theargument
madebyHans Kohn,ArnoldToynbee,and Elie Kedourie.The WestEuro-
 Whatis nationalism?719

peanvariety is "bad" too. In whatsensecouldithaveprovideda preferable

model?)Whentheintellectuals ofAsia andAfricacameintotheirowninthe
20thcentury theymostlychose official nationalism, especiallytheMarxist
revolutionaries.Each newlyindependent determined
intelligentsia, to build
its own state,followssimilarpolicies,including imperialism. The futureis
evokedbut notclarified by Anderson;he impliesthatnationalism willbe
aroundas longas thestateremainswithus, buthe hedgeshisarguments.
ImaginedCommunities does notclaimto offer a theory.It is moreevoca-
tivethansystematic. It reliesmoreon highlysubjectiveinterpretations of
nationalistpoetrythanon statistics of socialmobilization.Someofthevig-
nettesof SoutheastAsiannationalists and theirthoughts are marvelous, as
aretidbitsofinformation on languagepolicy.Buttheensembledoes notadd
up to a coherentargument. Nextto nothing is said aboutnationalism inthe
countriesthatprovidedthe firstmodels-France, America,Britain,Ger-
many.Onlythe imitators are treated,and Andersonfailsto explainwhy
certainfeaturesoftheexemplarnationswerechosenand notothers.Point-
ing out historicalcontinuities is a legitimate task,thoughit providesno
specialwarrant forfocusingon nationalism.
The lateDudleySeers,as behoovestheleaderofthemajorbridgehead of
thedependentistas in Europe,theSussex Institute forDevelopement Stud-
ies, equates nationalism withself-reliant development, delinkedfromthe
world's core. In The Political Economy of Nationalism, an ephemerallittle
tract,he recantshis formercommitments to orthodoxMarxismand Key-
nesianismand atonesfordecades of-he thinksmisguided-workas a de-
velopment economist fortheUnitedNations.Nationalism,truetothecreed
oftheCepalistas(he workedcloselywithRaulPrebischandOsvaldoSunkel),
is development planningthatseeks autonomy forthestatefromtheworld
demands.Marxismand Keynes-
economyand catersto egalitarian-populist
ianism,in his words.
bothfailto takedue accountofnon-material motives,especially
nationalism-theurgeto promotethepresumedinterests ofa group
withculturalcoherence,probablyshowingat leasta degreeoflinguistic
and ethnichomogeneity, and usuallyinhabitinga politicalunit,or
nation-state(thoughsometimes appliedto a groupofthesamekindsub-
mergedwithinone or morenation-states).

As an econo-
Untilthe 1960s,I too tooklittleaccountofnationalism.
mist,I naturallyconcentratedon materialmotives:peopleworkedto
earnmoney,and thelevelofourincomedetermined howwe spentit.
Moreover,likemanyofthoseeducatedin theAnglo-Saxon cultural
tradition,
I saw nationalismas fundamentallyirrational.
Fortunately,
withthespreadofinternational contacts,ofmediasuchas newspapers
and television,and ofeducationtherewas a growing of "in-
realisation
terdependence", whichwouldbe completewhenall foreigners sensibly
learnedsomeEnglish.(pp. 9-10)
 Organization
720 International

Now, he believes,talkof interdependence is merelycode forthe global

hegemony ofthesuperpowers. Internationalism, he argues,is anotherword
forAmericandomination, latterly mediatedbythelegitimacy ofmonetarist
economicdoctrineas well as by politicaland military intervention. The
argument is too familiar to requirefurther elaboration.
How can the nationalist liberatehimselfif he is saddledwith"a small
population,seriousethnicdivisions,locationclose to a superpower, few
naturalresources,a culturally subvertedbureaucracy, highconsumerex-
pectations, anda narrowtechnological base?" (p. 91). HereSeerstalkshard
sensethatdeparts from the orthodoxy ofdependency thinking. Nationalist
leaderscan diversify their trade contacts and exploit the geopolitical needs
the
of superpowers. Dom Mintoff's Malta is his exemplar. The International
MonetaryFund is notonlytheexecutivecommittee of international capi-
talism:its staff can be argued with and persuaded in many circumstances,
becausetheold canonsofBrettonWoodsare indisarray.Across-the-board
planningis a waste of time.So are nationalaccountsthatare based on
Keynesianassumptions.Patrioticappeals can be used to persuadethe
populace to put up withhard times.Appeals to patriotism shouldtake
and
theplace of irresponsible incompatible promises made by nationalist
leadersto their diverse followers (Allende's Chile is his exemplar ofwhat
notto do). And-heresy of heresies-hard-headed fiscal policies are abso-
lutelynecessary!
Nationalism forSeersthenreallymeans"as muchautarky as youcan get
away with." He means not only economic but also cultural and linguistic
autarky. Sincecosmopolitanism in taste and attitude means being a lackeyof
a superpower-and entails lack of economic development-a self-reliant
cultureis also likely to be an economically successful one. He seriously
the
weakens punch of this argument by constantly stressing the seductive
wilesofcosmopolitanism.
So muchfortheThirdWorld.WhataboutEuropeand Britain?He advo-
cates "extendednationalism." In fact,thetripartite divisionoftheworldis
he
no longeraccurate, says, with the waning of the neocolonial consensusin
WesternEuropewhich, until recently, had succeeded in perpetuating the
defunct empiresbyother means. A new legitimate protectionist consensus is
forming; the European Community should become a geschlossener Han-
delsstaat;BritishLabor shouldEuropeanizeitslegitimate protectionist im-
pulses.Other culturallycognate clusters of countries should follow suit.The
worldshouldbe made up of moreor less closed economicblocs, each
following itsownextendednationalist holygrail.
In contrastto our otherauthors,Gellnerin Nations and Nationalism
offersa full-fledged theoryof nationalism, embeddedin a theoryof the
universaltendency toward industrialism. Nationalism and thenation-state
providethe sole legitimate form of political organization withintheglobal
trendtoward industrialsocieties. "Nationalism is primarily a politicalprinci-
 Whatis nationalism?721

ple,whichholdsthatthepoliticalandthenationalunitshouldbe congruent.
Nationalism as a sentiment,
or as a movement, can bestbe defined interms
ofthisprinciple.Nationalist
sentimentis thefeeling ofangerarousedbythe
violationof the principle,or the feelingof satisfaction aroused by its
fulfillment.
A nationalistmovementis one actuatedby a sentiment of this
kind"(p. 11;emphasisinoriginal).Again:"nationalism is a theory ofpolit-
ical legitimacy,
whichrequiresthatethnicboundariesshouldnotcutacross
politicalones,and,inparticular,thatethnicboundaries within a givenstate
... shouldnotseparatethepower-holders fromtherest" (p. 1).
The theoryuses threevariables:(1) thedistributionofpoliticalpowerina
society(access limitedto geneticallydefinedor quasi-hereditary status
groupsversusupwardmobility);(2) easy access to a "highculture"of
literateand sophisticated
communication via a systemof publiceducation
versusverticallysegmentedsocial groupseach attachedto a local "low
culture";(3) ethnichomogeneity as opposed to ethnicheterogeneity, as
defined bylanguage.The combination inwhichthesevariablesoccurinany
givensocietywilldetermine whatkindofnationalidentification-if any-
willcometoprevail.Agrarian societiesareinnocent ofnationalism; thepush
and pullcomes onlywhenthetrendtowardindustrialism manifests itself.
The difference betweenagrarianand industrial societiesis bothcybernetic
and semantic.Agrariansocietiesare prerational becausetheyfeature"the
co-existence withinthemofmultiple, notproperly united,buthierarchically
relatedsubworlds,and theexistenceof specialprivileged facts,sacralized
and exemptfromordinary treatment." In industrialsocieties,"all factsare
located withina single contiguouslogical space . . . statementsreporting
themcan be conjoined and generallyrelated to each other . . . one single
language describes the world and is internallyunitary.. . . there are no
special,privileged, insulatedfactsor realms,protected fromcontamination
or contradictionby others,and livingin insulatedindependent local spaces
of theirown" (p. 21). The difference betweenthetwois also expressedin
thecontrastbetweenbeliefsthatnotonlyclaimto be "true" butthatalso
providethesole criteriaforjudgingall "truth"(i.e., dogmaticuniversalist
religion),and beliefsthatadmitthecontingent natureof truthclaims,the
possibilitythattruthis not revealedforall timebut unfoldsgraduallyin
conformity withrelativisticcognitivecriteria(i.e., thepost-Enlightenment
scientifictradition).Nationalismis typicalof thetransition to therational
tradition.It expressesthe societalthrusttowardhomogeneous perception
and homogeneous social organizationand behavior.It can be triggered by
religiousreformism thatstrivesforcoherenceanduniversality, though itwill
eventually come intoconflict withreligion.Nationalists maymakeuse of
prerational symbols,buttheywillrejectthemas soon as theytakepower.
Nationalismis not the resurgenceof submergedprimordial longingsfor
ethniccommunity. Nationalismis a consequenceof the "objectiveneed"
forindustrial rationality.
722 International
Organization

This pointis crucial.Gellnerrejectstheoriesof nationalism thatstress

eithertheobjectiveor thesubjectiveschoolsofthought becausehe consid-
ers themcomplementary ratherthancontradictory. Modernization causes
thelongings andadjustments thatproducenations,nottheobjectivecriteria
ofculturalidentity or assertionsofpersonalwill.Noteverypotential nation
becomesa realone becausemostpeopledealwiththeirlongings andfrustra-
tionsby simplyassimilating withothersinitiallyunlikethem,individual
determination to "pass" and priorculturalaffinities permitting.
The theorycan be summarized schematically. The schemeis mine,not
Gellner's.
Situation1. Agrariansocietyis nonrational and nonnational because its
social structureis characterized by discontinuous communication patterns
and heterogeneous symboliccontent.A few specializedhereditary status
groups,organizedhorizontally, professa commonhigh-literate cultureand
ruleovera vertically organizedilliterateand isolatedpeasantry attachedto
variouslow cultures.
Situation2. An exogenousforce(foreignconquest,a new religion,a
suddenincreasein theavailability of capital,a technologicalchange)trig-
gersindustrialization/modernization. Peasantsare "mobilized"by moving
intoarmies,factories,cities;theyacquirea tasteformiddle-class amen-
ities.Theyaspireto some of theperquisitesof thehighculturebutfailto
achieveanything likepoliticaland economicequalityduringtheearlystage
ofindustrialization.
Situation3. The newlymobilizedbutunassimilated beginto act outtheir
frustration.Conditions theysucceedeventually
permitting, in acquiringthe
linguisticand numericskillsneededto pass intothehighculture,thereby
losingwhateverculturalfactorsdistinguished themearlierfromtheruling
groups,aftersucceedingin theirdemandsforgreatereducationalandpolit-
icalparticipation.A successfulandreasonably contentednationis born.The
pattern followsthe"statefirst, nationsecond" sequenceofWestEuropean
experience.
Situation4. Thingsare muchas inthepreceding case. Educationalaccess
improves, thelowerorderssucceedingainingaccess topoliticalpower,the
societyas a wholeis characterized by muchculturalhomogeneity. How-
ever,thepeoplelivein manysmallstates,whichthoughsimilarintermsof
themainvariables,nevertheless because of theirsmallnessand weakness
are unableto represent thecultureto therestoftheworld;dispersion into
manystatesfailsto givetheculturea symbolic roof.A nationalist
sentiment
existswhichexpressesitselfintheformofnationalunification byforce.This
is theGermanand Italiancase, "nationfirst, statesecond."
Situation5. The newlymobilizedbutunassimilated beginto act outtheir
frustration.But preexistingculturaland powerconditions preventeasy as-
similation.Culturally butmobilizedpeopleare excludedfrom
differentiated
thebenefits Theyrespondby rejecting
of industrialism. thehighcultureof
 Whatis nationalism?723

the rulersand creatinga rivalhighcultureby drawingon theirown low

culture.This takestheformof a nationalism challenging thebeliefsofthe
rulers,secession,orthedisplacement ofrulersnowseenas alien.Thisis the
"nationfirst,state second" patternfamiliarfromthe modernhistoryof
EasternEurope.
Situation6. In certaincases (Gellnermentionsthe Islamicworldand
Africa)things do notworkoutthatcrisply.Whatifthefrustrated, mobilized,
butunassimilated areunableto optforthehighcultureoftherulersbecause
theyrejectportionsof it on culturally and religiouslyconditioned grounds,
and yettheycannotgo back to theirpriorlow cultureseither?Whatifthe
culturalheterogeneity is so greatthatportionsof the alien highculture
appearnecessaryas a meansto holdthesocietytogether, as in Africa?In
sucha situation thenationalisms thatdevelopare notrationalbecausethey
cannotcome to termswiththehomogenization of truth-findingnormsthat
prevailin industrialsociety.Religious-culturalabsolutismand rationalrel-
ativism,becausetheyare compelledto coexist,cannotresolvetheconflict.
Situation7. In someotherwise rationalizedsocietiesinwhichthelogicof
situation3 has largelyprevailed,certaingroupsremainwhose members
cannotassimilatebecause theirculturaldistancefromthe majority is too
greatand the majoritywould not permitthemto assimilateeven if the
minority wishedit. Such situations representflawednationsthatareinfora
lot oftrouble.
Whatdoes all thisportend?Gellner,in a turgidchapteron typologies of
nationalism, concludesthatonlythree"typical"nationalist situations mat-
terwhenhis three-variable modelis put to work.Situations5, 6, and 7,
because theythrowintodoubttheuniversaltendencytowardrationaliza-
tion,are thesalientones forGellner.He deniesthatthereare fundamental
differences betweenlate and earlymodernizers-nation-builders, thatthe
WestEuropeanpatternis uniqueand all theothersdifferent. But his very
modelsuggeststheopposite.He exploressmoothas wellas discontinuous
rationalizationvia nationalizationbutcomesto theambivalent conclusion
thatwhilenationalism everywhere oughtto becomeweakerandevendisap-
pear,maybeit won'tafterall because ofthecontinuedviability ofpseudo
highcultureswhichare onlydressedup lowcultures.Perhapshe shouldnot
have dismissed"primordial"culturalself-definitions so completely before
makingthe sweepingargument thatall nationsare artifactsof modernity,
owingnothing to theiragrarianoriginsand everything to theimperative of
assimilation.
Havingsummarized Gellner'sargument, I now wishto commenton it.
First,however,I wantto makeclearthatI am in agreement withhis ap-
proach;thoughI wantto extendtheexamination to thefurther questionof
thequalityof international relationsimpliedby thecoexistenceof various
kindsofnationalism. Gellneris on therighttrack;butthetheory (sometimes
he callsita model,sometimes a scheme)is morea hintthana persuasiveor
724 International
Organization

completesetofpropositions. He maybe forgiven (withtherestofus) forthe

inabilityofthetheoryto predictmuch;he shouldnotbe so easilyabsolved
forhisfailureto providecoherentexplanations ofthepast.
The theorylacksclarityat crucialpoints.Gellnerleavesus indoubtas to
whetherthereare one, one-and-a-half, or two separatemodelsof national
development. He distinguishes betweena lateandan earlystageofindustri-
alization,and also a "verylate" stage.It is notat all clearhowthemodels
relateto thestages.Partofthetroubleliesinhisresoluterefusalto quantify
anyofhis statements aboutratesofchange.Hence we cannotbe surewho
and whatis identified witheach stage.Nor is it clearwhichmodelcan be
appliedto whichcountry.
Further, I cannottellwhether thesamemodelexplainsall sevensituations
or whethersituation6 is a specialcase notcoveredby thesamemodel.If
so, the exceptioncovers all of Africa and the Middle East. Japan
and Latin Americaare hardlymentioned;China not oftenenough.Is
thereno nationalism to be foundthereor is it irrelevant withrespectto
rationalization? Situation4 strikesme as extrinsicto the modelbecause
"culture"is heregivena different meaning thanintheremaining situations.
In situation 4 it meansintercultural comparison(GermansagainstFrench-
menor Englishmen), notconflict amongculturalsymbolswithinthesame
society.More seriousstill,thetheorycannotexplainwhy,aftera country
goes through theexperienceof situation 2, it shouldtilttowardsituation 3,
as opposedto4 or5, exceptintermsofuniquehistorical circumstances. The
retrodiction is episodicratherthansystematic.
The theoryis also incomplete.It worksbestforthe "nationfirst,state
second" pattern;it explainsthe secession of the mobilizedbut unas-
similated.Gellnerloses interestin situationsin whichthe mobilizedare
assimilatedbutby beingengulfed in a singlehighculturecometo identify
witha politythatis aggressively exclusiveand demanding. Whyis French,
British,and Americannationalism (leave alone Japaneseor Russian)no
longerinteresting aftereverybodybecomesa happynationalist? In short,
Gellnerneglectsthe phenomenonof successfulsymbolmanipulation in
rationalizedsocieties.Althoughfascinatedwithcommunication patterns
duringtheearlystagesofmobilization andindustrialization, he showslittle
interestin how thesameprocesscan be used to explainthepersistence of
nationalsentiment in thelaterstage.
The differences capturedby the seven situationsrequiremorenuance.
The situationsimplythatthereare degreesof rationalization, as Gellner
freelyadmits.If so, thedegreesoughtto be describedandexplained.To do
so, however,requiresattention to doctrines and ideologies,thevariousand
competing symbolicmeansthemobilizeduse to come to termswiththeir
newcondition.Gellnerpaysattention to thisonlywhenhe discussesIslam.
He insistsin a chapterdevotedto ideologythatitis all "falseconsciousness"
(despitetheanti-Marxist thrustofthebook),notworthy ofseriousstudyor
 Whatis nationalism?725

attentionbecause ideologiesare merelythe sillyjustifications nationalist

troublemakers dreamup tojustifytheirexistence.It is apparently thiscon-
victionthatenableshimtoignorethelaternationalism oftheWestEuropean
statesandofJapan.Itjustifies hisignoringthesubstantialideologicaldebate
in thesecountriesas to theidentity, purpose,and characterofthenation.
The debatein turnimpliesseriousinternaldisagreement aboutthatvery
identityand suggeststheabsenceofthekindofcoherent nationalmyththat
situation3 implies.
His muddling ofmodelsand stagescaststheargument as a gestaltrather
thanas sequentialmacrohistory. On theotherhand,Gellneradds a number
offeatures thatDeutschneglects.He does a wonderful job ofdescribing how
religioncan defineidentity; he showshow religiousidentities shade into
culturalones andhowethnicidentities form.The typology
acquirereligious
ofnationalism is a laudableeffortto showthedifferences betweenhistorical
processes,a step Deutschdid nottake. Finally,Gellneravoids theiden-
of languagewithethnicity
tification whichtendsto giveDeutsch'sversion
a monocausalcharacterthatwas notreallynecessaryformakingcredible
thecoreconstruct balance.
ofthemobilization-assimilation
How can we buildon theworkofDeutschand Gellner?We musttakeas
centraltheiranchoring of nationalismin thelargerstudyof how people's
perceivedidentities changeduring theprocessofmodernization, howpeople
becomecarriersof highculturesextending beyondtheirpurelypersonal
experiences.Like Deutschand Gellner,we mustrejectromantic notionsof
Urvolker whose unchanging naturereassertsitselffromtimeto time.We
mustaccepttheirinsistencethatnationalism, thoughmanufactured and in-
ventedintheformofdoctrines, ideologies,andpolicies,is a rationalwayof
organizing impersonalsocietiesaftermass social mobilization getsunder-
way. Hence the identification of nationalism withthisor thateconomic
policycan be ignored,as can theinsistence on thedomination ofclass over
nationas an organizing concept.And specificnationalist ideologiescan be
studiedas potentialrationalizingagentsinsteadofbeingtakenat facevalue
or merelydebunked.But thistask remainsto be done. I now sketchan
approachfordoingit.

How to studynationalism

The enterpriseof nationbuildingand nationmaintenance is farfromfinal

andcomplete.How thencan we talkabouttheachievement ofdomesticand
harmony?The studentsof "early nationalism"could confi-
international
dentlyengagein the enterprise because theywerenotyet
of retrodiction
inWest-
challengeofethnicstirrings
compelledto faceup to theintellectual
ernEurope;nordid theyfeelobligedto commenton theglimmers of na-
tionalobsolescenceimplicitin the movement forEuropeanunification. I
726 International
Organization

wishto exploitthecompetingideologiesofnationalism as cluesto patterns

of domesticand international and I mustpay attention
rationalization, to
pastandfuture uncertainties. is notfeasible.Myaimcannot
Yet, prediction
be higherthanOtisDudleyDuncan's:
ofourforecastswe will,withincreasing
As ingredients methodological
continueto prepareprojections,
sophistication, trendextrapolations,
modelsimulationsand developmentalconstructsso as to provideas
broadan arrayas maybe usefulofthelogicallypossiblepathwaysto
hypotheticalfutures.... There will be no pretensethatwe can gradu-
allymoveforward to theperfection ofmethodsofanticipating whatwill
actuallyoccur,forsuchperfectability [sic] is notlogicallypossible,aes-
thetically
appealingor morallyinspiring. Whatwe mayhopeto im-
prove,ifnotperfect, is our senseofresponsibility formakingknown
theimplicationsofourknowledge.1"
The thoughtsofferedhere are hypothesesabout possiblerelationships
betweenrationalizationand nationalism,not predictions.The research
necessaryfordemonstratingsuchrelationships in thepast is notyetcom-
pleted.Wereitcompleteitstillwouldnotsufficetopredictthefuture,either
of newand old nations,or forecasting
in termsof thedomestictranquility
theshapeoffuture orders.The resultsofthatresearchwill,at
international
best,openup reasonablespeculationaboutpossiblefutures andaboutpasts
thatwerefarfrominevitable.
of thecore termsnation,nationalism,nation-state,
I shall offerdefinitions
nationalistideology, and national myth.I shall thendevelop a typologyof
ideologies,followedby a discussionofindicators
nationalist and measures
forobservingthe existenceof a successfullyrationalized To
nation-state.
concludeI shalldiscusshowthevarioustypesofnationalism can be usedto
anticipatedomesticand internationalharmony.

Definitions12

A nationis a sociallymobilizedbodyofindividuals, themselves

believing
them(in their
thatdifferentiate
to be unitedby some set of characteristics
own minds)fromoutsiders,striving to createor maintain
theirown state.
Theseindividuals havea collectiveconsciousnessbecauseoftheirsentiment
ofdifference,or evenuniqueness,whichis fostered bythegroup'ssharing

The StateoftheArt,"as quotedwithapprovalby LloydFallers,

11. "Social Forecasting:
SocialAnthropology, p. 134.Fallersoffers studiesofnationbuilding
contrasting inTurkeyand
Ugandain orderto isolatethepatterns of syncretismthatemergedwhentraditional values
collidedwithWesternones. He illustratesnicelythedisciplinerequiredto avoidretrodiction
thatmerelytellsthehistory ofthecountry andto escape thehubrisofpredictingthefuture.
12. Fora morefullyarticulatedbutsimilar linkedtopropositions
setofdefinitions consistent
withmygeneralargument, see E. K. Francis,InterethnicRelations(New York: Elsevier,
1976),pp. 381-405.
 Whatis nationalism?727

ofcore symbols.A nationceases to existwhen,amongotherthings,these

symbolsarerecognizedas nottrulydifferentiating thegroupfromoutsiders.
A nationis an "imaginedcommunity" because thesesymbolsare shared
vicariously withfellow-nationals overlongdistances,thusproducing expec-
tationsofcomplementary andpredictable behaviorfromfellow-nationals. A
government is not consideredlegitimate unlessit is thoughtto represent
such a group.A nationis a groupof people who wish to practiceself-
determination. Nationalismis a beliefheldby a groupofpeoplethatthey
oughtto constitute a nation,or thattheyalreadyare one. It is a doctrine of
social solidaritybased on thecharacteristics and symbolsofnationhood. A
nation-state is a politicalentitywhose inhabitants considerthemselvesa
singlenationand wishto remainone.
I call attentionto the factthatthisset of definitions restson theprior
notionof social mobilization. Nationand nationalism implya situationin
whichpopularawarenessof,and some degreeofpopularparticipation in,
politicsprevail.I emphasizethatmanyof the 160-oddstatescurrently
in existenceare notnation-states.
Nationalsentiment is a beliefamongintellectuals andotherliterate groups
thattheyconstitute a nationandoughttopracticeself-determination at some
timeinthefuture, eventhoughthecondition ofevenpartialsocialmobiliza-
tionhas notyetbeenattained.The conceptis necessarybecausewe haveto
recognizesentiments of solidarity-perhaps potentialsolidarity wouldbe
more accurate-in situationsof literaryself-consciousness, as in Eliz-
abethanEngland,theearlyrisorgimento, in RussiaunderAlexanderII, or
in Brazilin the 1820s.
Such an elitesentiment mustbe sharplydistinguished fromnationalist
ideologies.A nationalist ideologyis a bodyofarguments and ideas abouta
nationadvocatedby a groupofwritersand acceptedby a specificpolitical
movement.Nationalistideologiesembodypoliticalprograms.They arise
onlyaftersocialmobilization hasgoneonlongenoughtohaveresulted inthe
availabilityof mass publicsattentiveto the message.They referto the
specifically"nationalist"contentof whateverideologiesare in political
competition. Hence theyprovidean additionaldimension fortalkingabout
thecontentofliberalism, conservatism,socialism,andfascism.Therewere
no nationalist ideologiespriorto thelate 18thcentury.
Typically,nationalist ideologiesmakeassertionsaboutkey contentious
aspectsofthesolidarity beingurged.Sincetheychallenge, advocate,orseek
to come to termswiththe impactof modernity, all nationalist
ideologies
mustbe concernedwiththe validityof the core values of thetraditional
culture.Revolutionary ideologiesseek to getridoftraditional values;syn-
cretistideologiesseek to amendor retainthem,differing on theextentof
interculturalborrowing whichoughtto be fostered.Ideologiesmakeasser-
tionsaboutthenation'sclaimto historical uniqueness,to theterritory that
thenation-state oughtto occupy,and to thekindsofrelationsthatshould
728 International
Organization

prevailbetweenone's nationand others.Nationalist ideologiesalso contain

constitutional and institutionalprogramson how the nationoughtto be
governed.Finally,theseideologiesadvanceideas on thehistorical mission
ofthenation,ranging fromquietself-perfection to conquestor therestora-
tionof somegoldenage.
Continuing strifeamongrivalideologiesclaimingdifferent missionsand
differentinstitutionsfortheirnationis proofofunsuccessful rationalization.
Such strifeprovidesevidencethatthesociallymobilizedare split,thatthey
cannotagree on the characteristics thatmakethemdifferent fromother
nations.Theycannotreachagreement on theuniqueinstitutions thatought
to governtheirstate.The societydoes notseemable to cometo termswith
thestrainsof modernization. A certainideologymaysucceedin capturing
the statefora limitedperiodand thenenact its program.But if a rival
ideologytakesoversoon thereafter and scrapsthepoliciesofitspredeces-
sors,we are entitledto wonderwhether a morepervasivenationalism ever
reallycharacterized theentirepopulation, whether an acceptedcorebodyof
valueseverexisted.
I reservethe termnationalmythforthe situationin whichtheclamor
amongideologieshas beentranscended totheextentofresulting ina coreof
ideas and claimsabout selfhoodcommonlyacceptedby all the socially
mobilized.Put differently, the nationalmythrepresents thoseideas, val-
ues, and symbolsthatmostcitizensacceptdespitetheirbeingdividedinto
competingideologicalgroups.The mythrepresentsthe overlapamong
ideologies.It is possible,of course,thatthebearersof a specificideology
capturethe state and eventuallysucceed in imposingtheirbeliefson
everybody. Theirideologythenbecomesthemyth.Something likethishap-
penedinFranceafter1870,inRussiaafter1917,inJapanafter1945.Itis also
possiblethatno singleideologyeverwinsfinally andthatthemythis made
up of itemson whichrivalideologieshave compromised, as in theUnited
Statesafter1865or in Belgiumsince1970.In eitherevent,evidencethata
nationalmythprevailsis also evidenceof successfulrationalization.
My definitions seek to sidestepthe conceptualand historicaldichoto-
miesthatplaguethestudyofnationalism. Armedwiththesetermswe do not
have to worryabout the objectiveas opposed to the subjectivebasis of
nationhood, authenticas opposedto illegitimate nationalisms, the virtues
of the older typesand the vices of the latecomers.The termsoughtto
covereverysituationand providea value-neutral way ofcomparing them.
Whetherwe approveor disapproveof the historicalresultsis reallya
questionofwhether we are in agreement withtheparticularformofration-
alizationthatcame about; it need not be a judgmenton nationalism as a
describablephenomenon.
Finally,mydefinitions are capableofbeingoperationalized through sys-
tematicobservation. Paul Valerymayhavebeenquiterightwhenhe wrote
that
 Whatis nationalism?729

Historyis themostdangerousproductevolvedfromthechemistry of
theintellect.Its properties
are wellknown.It causes dreams,it intoxi-
cateswholepeoples,givesthemfalsememories, quickenstheir
reflexes,
keepstheirold woundsopen,torments themin theirrepose,
leads themintodelusionseitherofgrandeur or persecution,
and makes
nationsbitter,arrogant,insufferableand vain.'3
However,we need not stop withhis observation.We can determine just
howarrogant and howvaina givennationmaybe, comparedwithitselfand
withothersat variouspointsin theirhistories. Moreover,we can determine
whythearroganceprevailsby comparing thecompeting ideologiesand by
discovering whichbeliefsemergedas thenation'smyth.Notevennational-
ists act on the basis of pure will or intuitive
romanticinsight.They are
constrained, like everyactorin a collectivesetting,by the rulesof sat-
isficing,bytheneedto calculatetheopportunity costsofmaking alliancesor
stressing thisor thattheme,by theimperative of recognizing
relationsof
strategicinterdependence withotheractors. Even the formulation of a
nationalistideologyinvolvesrationalchoice.

Nationalist ideologies compared: a typology

An incomplete readingofmodernhistory suggeststheexistenceofseven
manifestations of nationalist
ideologies,fourrevolutionary and threesyn-
cretistinemphasis.Despitetheirdifferences, all sevenhavecertaincharac-
teristicsincommonthatdistinguish themfrompremodern ideologies.Allare
populistic;theyall derivetheirappealfromtheclaimthat"thepeople" ofa
certainterritory(nota class, or statusgroup)have an innaterightto self-
determination. All are progressivebecause theyrejectall or some of the
historicalpast;theybelievein theefficacy ofhumanintervention to change
history forthebetter.Andall arerationalbecausetheydiagnosea challenge
andprescribea response;theyembodydistinct notionsofcause andeffect,
endsandmeans;matching meansto endsis notusuallyrandom,emotional,
passionate,willful,or romantic.
But the differences betweenthetwo maintypesalso mustbe stressed.
Revolutionary ideologiesinsiston drasticinstitutional
change.Certaintypes
of social groupsare to be removed.Relationsamongremaining and new
groupsmustbecometotallydifferent. The old elitemustgo anda newelite,
compatiblewithpopulismand progress,musttake its place. Syncretist
ideologiesare unwilling to be thisdrastic.Theydo notrepresent a sharp
breakwiththepast,onlysomecompromise withit. Theyoftenrejectthe
values of modernity, thoughtheyseek theincorporation of its techniques
and someofitsinstitutions.

13. CollectedWorksofPaul Valery,ed. JacksonMathews,vol. 10 (New York:Bollingen

Foundation, 1962),p. 14.
730 International
Organization

Revolutionary ideologiestherefore tendto be moreinternally coherent

thansyncretist ones. Theyembodya sharpersenseoftechnicalrationality.
Theyarewillingto tradeoffvaluesquiteruthlessly, whereastheirsyncretist
rivalsare oftenhesitantand inconsistent in theirchoices. All nationalist
ideologiesstressthe shortrunover the longterm.None appreciatesthat
overthelonghaulanysetofmajorinstitutional changestriggers unforeseen
and possiblyunwantedconsequences,inconsistent withthe values being
urged.But syncretist ideologiesare muchmorelikelyto suffer fromthe
uncertainties of the long run. Revolutionary ideologiesare consistentin
urging inclusivepopularparticipation, whether voluntaristicormanipulated.
Syncretists tend to fudgethe issue of participation, alternatingbetween
voluntary modesand manipulation, betweenelectionsand repression, be-
tween individualrightsand the obligationto submergeoneselfin the
nationalist struggle.
FollowingCarltonHayes and Hans Kohn, I dividethe revolutionary
ideologiesof nationalism into"liberal" and "integral"variants.Each, in
turn,must be subdivided.Liberals break down into "Jacobins" and
"Whigs,"integralists into"Marxists"and "fascists."
Jacobinliberalsrejecttraditional valuesand institutions completely and
wishtoreplacethem;Whigsalso rejectthembutlookforreplacements more
cautiously.Both believe that liberal societieshave manyinternational
affinitiesandoughttoborrowfromoneanother.Jacobinsderivethenation's
claimto historical distinctivenessfromallegedethnicand linguistic homo-
geneity;thustheyprofitfromwhateverprocessesof homogenization and
centralization mayhave been triggered by earlierregimes,as did Robes-
pierre,Danton,Cavour,and the 1848Frankfurt Parliament. Whigsprefer
thelegitimacy of historicalcontinuity to ethnichomogeneity, as did J. S.
MillandNehru.Bothtypesagreethatthearea occupiedbywhatever group
is defined as "thenation"is theproperterritory ofitsstate.Bothalso agree
thatrelations amongliberalnationsoughttobe peaceful.Towardnonliberal
politicalentities,however,Jacobinsand Whigsare willingto use forceto
makeothersprogresstowardliberalism, to aid in theirliberation, and to
colonizethemfortheirowngood.Jacobins aresomewhat moreaggressive in
theirproselytizing zeal, as shownby WoodrowWilsonand ThomasJeffer-
son. All liberalnationalistsadvocate representative democracy,natural
rights,and the freeparticipation of all citizensin government. Jacobins
believethatthehistorical missionofthenation is notmerely continuous self-
perfection butalso theglobaldiffusion of the creed. Whigs preferto confine
matters to continuousself-perfection.
There has been an electiveaffinity, to say the least, betweenliberal
nationalism and late 19th-century imperialism, withAmericanintervention
in Mexicoand theCaribbean,and withAlliedmeddling intheRussiancivil
war after1917.The powerof thecreedwas obviousin theoccupationof
Germany andJapanafter1945,as wellas intheconstruction andconductof
 Whatis nationalism?731

NATO. It is farfromobviousthatliberalnationalism also has an elective

affinityfora peacefulworldorder,despitethepresenceofsomethemesthat
mightlead one to suspectit.
Whenwe examinethe integralist familyof ideologies,such suspicions
vanishfast.Fascistsand Marxistsalso rejecttheold orderand itsvalues,
thoughfascistssometimes pretendto retainsomeofitssymbols,as in Hit-
ler'splayingwithpre-Christian themes,Mussolini'sappealstoRomangran-
deur, and Kita Ikki's to Shintoideas. Borrowingfromotherintegralist
societiesis praiseworthy. Whatgroupof people is to be selectedas "the
nation"?Marxistsopt fortheparticular class, or coalitionof classes, that
resistsimperialism in a specifiedterritory.
Fascistsuse a racialcriterion or
arguments abouthistoricalcontinuity or both.Bothideologiesadvocatea
totalitarian
modeofgovernment bya vanguard oftheelectwhichincarnates
thenationas a collectivity. Bothassumethatthenationmuststruggle for
survivalbecause it is constantlythreatened by attackfromhostileexternal
forces.Fascistsglorify warandself-assertion as partofthenation'smission.
Marxistsacceptwaras inevitableas longas imperialism continuesto live,
buttheyglorify onlywarsofnationalliberation. For thefascist,themission
ofthenationis toassureitsownsurvival;fortheMarxistitis theushering in
ofa classlesssociety.The ideaofa harmonious internationalorderis aliento
fascistsandMarxists,excepton minorissuesintheshortrun.As longas the
contrastbetweenliberaland integralnationalism remainsas starkas the
historicideologiessuggest,the idea of a rationalizedworldseems far-
fetched.Butthen,thesuccessorsofStalin,Mao, Hitler,andPrinceKonoye
seemto have sensedthatideologicalpurityis notalwaysrational.
The threekindsofsyncretists disagreeon howmuchoftherevolutionary
ideologiesoughttobe acceptedintheircountries. "Synthetic syncretists"-
MohandasGandhi,Leopold Senghor,K'ang Yu-wei,Mazzini,theGerman
romantics-consider manymodernvaluesas desirable,providedtheycan
be mixedappropriately withtraditional values to be retained.However,
suchideologiesfeelthatnotall traditional valuesare worthyof retention,
particularlythosecloselyassociatedwitha diffuse agrarianorder.Synthet-
ics wantto borrowvalues,alongwithinstitutions and techniques, fromthe
earlymodernizers. Theirclaimto nationhood fortheirowncountries rests
on historicallongevity. Theydemandonlytheexisting stateforthenation's
home.Theyseek peacefuland cooperativerelationswithothers,afterthe
survivalof theirnationseemsassured.Democracymayor maynotbe the
featuredformof government; various formsof authoritarian rule by
theelitethatunderstands the propermixtureof values is morecommon.
The historical missionof thenationis to bringaboutitsown survivaland
protection, whichimpliesheavy borrowing fromnonindigenous cultural
sourcesin orderto succeed.Amongthesuccessfulsynthetic syncretists we
couldlistsomeof theMeiji reformers, modernSouthKorea, and perhaps
thepostindependence regimesin IvoryCoast and Senegal.
732 International
Organization

TABLE 1. Attributesof revolutionarynationalistideologies

Ideology

Liberalism Integralism

Dimension Jacobin Whig Marxist Fascist

1. What should be reject reform reject reject sub-

done about the core outright gradually outright stance but
values of the tradi- retainsymbols
tional culture?

2. What ought to be borrowingis good; liberalsought borrow from borrow from

borrowed fromother to borrow fromeach other other Marxists other Fascists
cultures?

3. What is the ethnic and historical class defined as race or histori-

nation's claim to linguistic continuity resistance to cal continuity
historical homogeneity seen as ethnic imperialism
distinctiveness? or linguistic withinspecific
homogeneity territory

4. What territory whatever area is occupied by the whatever area is occupied by the
is properlythe group defined in 3 group defined in 3
nation's?

5. How should the spread liber- make peaceful permanentstruggle,need forsur-

nation relate to alism by ex- contributionto vival, endemic hostilitytoward
other nations? ample and by war the expansion of others;
if appropriate; liberalism;

foster peaceful relations among fightimperialism, glorifywar and

all liberalnations; fightother but do not self-assertion
nations or colonize them in order glorifywar
to advance liberalismglobally

6. What is the proper representativedemocracy totalitarianrule via a vanguard

institutionalstruc- natural rightsfor individual group incarnatingthe nation;
ture forthe nation? citizen rightsinherein nation, not in-
dividual
7. What is the continuous continuous bringabout assure survival
historicalmission self-perfection self-perfection classless of race
of the nation? and the global but no explicit society
diffusionof the mission other-
creed wise

Examples FrenchRevolution, Mill,Nehru Stalin, Tito, Mussolini,

Jefferson, Ho, Mao Hitler,Eastern
Wilson, Cavour, Europe in the
German liberals 1 930s, Kita lkki
 Whatis nationalism?733

TABLE 2. Attributesof syncretistnationalistideologies

Ideology
Dimension Synthetic Traditional Restorative

1. What should be many modernvalues distrustmodernvalues; reject existingtra-

done about the core are good and usable; verycautiously borrow ditionalvalues in favor
values of the tra- mix with good tra- of restoringvalues of
ditionalculture? ditionalvalues a past golden age
2. What ought to values as well as onlytechniques and onlytechniques, not
be borrowedfrom techniques and institutionsneeded, institutionsor values
othercultures? institutions not values

3. What is the historical culturalsuperiority religiousrevelation,

nation's claim to longevity as evidenced by scripturalauthority
historical historicallongevity;
distinctiveness? race
4. What territory usually, but not area of existing area occupied by people
is properlythe always, the existing state to whom the revelation
nation's? state was made

5. How should the cooperativelyand peace- ambivalently,because of hostilelyand distrustfully;

nation relate to fully,aftersurvivalis continu-iigfear for need forstruggle!
other nations? assured survival vigilance

6. What is the variable various corporatede- theocracy

properinstitutional vices to limitpopular
structureforthe participationand
nation? legitimateleadership
of traditionalgroups
7. What is the assure its own survival assure its own sur- restorethe golden age
historicalmission and self-perfection vival
of the nation?

Examples Gandhi, Senghor, Stein, Maurras, Meiji Tilak, Khomeini,

K'ang, Yu-wei, Mazzini, reformers,many Con- Al-Banna,Slavophiles
Afghani,Lutfi,German fucianists,Indian
Romantics Muslimreformers

"Traditionalsyncretists"
distrustnonindigenous valuesand have no in-
tentionofintroducingthem.Theyremaindeeplyattachedto religioussys-
tems thatpenetratethe local culture-Confucianism, Islam, Hinduism,
Catholicism.However,such people are quitewillingto take over nonin-
digenoustechniques,suchas technology, scientific
education,literacy,and
modemarmies.Moreimportant, theyarewilling
toadaptindigenous institu-
tionsto theextentnecessaryto incorporatethesetechniques, forexample,
through theintroduction
ofconscription,compulsory publiceducation,and
evenaspectsofindustrialism.
Traditional persuadethemselves-
syncretists
usuallymistakenly-that theycan borrowtechniques andinstitutionswith-
734 International
Organization

out also acceptingthe values thatgo withthem.Theirclaimto national

distinctiveness restson an argument forhistoricalculturalsuperiority
over
theirrivalswhichoftentakesa racialform,as intheworkofCharlesMaur-
ras, some of thelate ChineseConfucianists, and of Rabindranath Tagore.
Theyclaimas thenation'srealmtheterritory oftheexistingstate,butthey
are ambivalent aboutthenation'srelationswithothercountries becauseof
theirstrongfearthattheymaynot survive.Assuringthatsurvivalis the
nation'smainmission,whichimpliesan indigenousculturalrenaissance
alongwiththe introduction of nonindigenous institutions
and techniques.
The modeofgovernance preferredby traditional is corporatism
syncretists
designedto containand channelpopularparticipation whilelegitimatingthe
leadershipoftraditional groups,as clearlyexpressedby one oftheearliest
"latecomers,"Prussia'sBaronvomStein.
"Restorativesyncretists"rejectforeignvalues and institutions; they
merelywanttheforeigner's techniques-hisarmiesand factories.In fact,
theytakethepositionthatthevaluesactuallyprofessedby theirowngov-
ernment are alreadydangerously corrupt and mustbe replacedbypureand
authentic indigenous values.Theyare "restorers"becausetheywishto get
ridofforeign moralandinstitutional accretionsandbringbackthepurity of
an earliergoldenage. Theyrelyon religious revelation
andscriptural author-
ity-the vedic textsforTilak, the Koran forKhomeiniand Hassan al-
Banna, the ChristianBible forthe Slavophiles.Who is the nation?The
peopleto whomtherevelation was made,irrespective ofwheretheymight
live. How shouldthe nationbehave towardothers?It mustbe eternally
vigilant,trustnobody,and be readyat all timesto defenditsspiritual trea-
sure. Restorers,by virtueof theirstance,mustexpectwar and violence.
Theirchiefmissionis torestorethelostgoldenage,whichtheyattempt todo
by institutinga theocratic dictatorship.Theycan no moreenvisagea har-
moniousworldorderthancan theintegralists.
The history of mostnationsis a storyofcompetition amongtheseseven
ideologies.Can theybe transcended to giveus a consensualnationalmyth
thatwouldwaterdownand sloughoffsomeofthecontrasting themes?Or,
as in the case of some of the integralist and liberalexperiences,is the
nationalmyththesamethingas a victorious ideology?In ordertoinvestigate
thesepossibilities,and theirimplications fora rationalizedworldorder,I
mustfirstshowhow to "measure"a nationalmyth.

Indicators and measurement

The neglectofsystematic indicators anddisintegra-
ofnationalintegration
tionis one of themainfailingsof theworksreviewed.Thereseemsto be
measurement
littleexcuseforthisas faras thequantitative ofdegreesand
typesofsocialmobilization are concerned.Conceptsanddataareplentiful.
The best studiesof nationbuilding,suchas EugenWeber'sPeasants into
 Whatis nationalism?735

Frenchmen, use themto greatadvantageand do nothesitateto mixthem

withqualitativeobservations.Wheresystematic publicopiniondata on
popularexpectations ofthenation-stateexist,ofcourse,thepoliticalconse-
quencesof social mobilizationcan be directlyassessed. Whenwe do not
havesuchdatawe areforcedtouse qualitative observationson a systematic
comparative basis.
Whatkindsof questionscould we ask of a societyin orderto makea
judgment ofitsdegreeofrationalization undertheumbrellaofa nationalist
myth?We wantto knowwhethersuch beliefs,despitetheprevalenceof
divergentattitudes, consensualtoenablethestatetofunction
aresufficiently
Moreover,we wantto knowwhether
to everyone'sbasic satisfaction. this
happystateofaffairs ethnic,religious,
obtainsdespitelinguistic, class,and
statuscleavages.The following questionsare appropriate indicators:
forpoliticalsuccessionthatis regularly
Is therea formula observed
withoutengenderingcoups and civilwars?
Arethecore valuescommunicated by thepublicmediaand theschool
systemgenerallyaccepted?Is thereevidenceofmajormovements or
thesevalues?
eventschallenging
Is theeconomicpolicypursuedby thestateperceivedas equitableirre-
ethnic,or religiouscleavages?
spectiveofregional,linguistic,
policypursuedby thestateper-
Is thecultural(especiallylinguistic)
ceivedas equitable?
consensusdespitetheexistenceofcompeting
Is therea foreign-policy
partiesand changesofgovernment? (N.B. Thisneednotmeanthatpol-
icyremainsunchanged;it meansthatchangesin policydo notengender
strife
thatcovarieswithothersocialcleavages.)
Two key indicatorsare impliedby thislist,one quantitative, theother
qualitative.Whatis theincidenceofcivilstrife, sincestrikes,riots,coups,
conspiracies,and full-scalecivilwar are themostvisibleevidenceof dis-
Is thelanguagepolicypursuedbythestatea sourceofsatisfac-
satisfaction?
tionor dissatisfaction?
Civilstrife
is a verytrickyindicator becausewe cannotassumethatevery
strike,everyriot,everyassassination providesevidence
ofa politicalfigure
of deep-seatedand widespreaddissatisfaction. These eventshave varying
depending
diagnosticsignificance, on theextentof socialmobilization, the
degreeto whichthe statepenetratesthe society,thenatureof theurban-
rural,high culture-lowculturesplit. FollowingTed Gurr and Muriel
McClelland,we mustdistinguish between"turmoil,""conspiracies,"and
"internalwar" as providing a scale ofviolentevents.14Turmoil, ingeneral,

14. Ted RobertGurrand MurielMcClelland,PoliticalPerformance

(BeverlyHills: Sage,
1971).
736 International
Organization

is nota validindicatorofdissatisfaction withthenation-state. It maysuggest

alienationthatfindsexpressioninrivalnationalist ideologies,butwe cannot
be surethatthedemandsputforward do notremainwithin theconfines ofa
sharedmyth.We mustbe certainthattheriotsandstrikes aremotivated bya
setofsymbolschallenging theofficial onesbeforewe can interpret thisform
ofcivilstrife.
Whatabout the incidenceof conspiracies(terrorism, mutinies,coups)
directedagainstthegovernment? Theseare notalwaysvalidindicators ofa
desireto makeor unnmake a nation.I distinguish betweenconditions that
obtainwhena nationis yetto be builtand challengesto an existing nation-
state,theconditionof nationmaintenance.
Let us assumea situationofincomplete socialmobilization ina stateruled
autocratically bya narrowly basedelite,thoughthereis no doubtthata state
existsand is recognizedto exist.Conspiracies thenconstitute evidenceofa
desireto createa nation-state ifone segment oftheeliteseemstoimposeits
visionofthenationon othersegments oftheelite.Conspiracies also consti-
tutesuchevidenceiftheconspiring elitewishesto makea nationoutof a
targetpopulationdistributed amongseveralstates.Conspiracies, however,
arenota validindicator ofsucha desirewhentheymerelyserveas a method
for displacingone set of caciques when another,as in most of Latin
America's"national"periodduringthe19thcentury. For instance,theen-
demicconspiracies ofSantaAnna'sMexicoarenotevidenceofnationalism,
buttheconspiraciesofthereformers of the 1850sprobablyare.
Now letus assumea situation inwhichsocialmobilization is faradvanced
andan effective stateexistsas well,which,untilthepointatwhicha marked
increasein conspiraciesoccurs,had enjoyedlegitimate authority. The in-
crease in conspiraciesis validevidenceof a nationalist desireto splitoff
fromthatstateor to takeit over,iftheconspirators advancea nationalist
ideologydifferent fromthepreviously prevailingmyth, as inthecase ofNazi
Germanyor Lenin's Russia. An increasein conspiraciesis also a valid
indicatorwhenthe conspirators represent groupswho feelleftout of the
benefits bestowedby the stateand its nationalmyth,usuallybecause the
conspirators are "different" forreasonsof religion,ethnic,linguistic, or
statuscharacteristics.ExamplesareIndiasince1947,Northern Ireland,and
Biafra.
Internalwar, underconditionsof completesocial mobilization, always
proves the failureof a nationalmythand the prevalenceof competing
nationalist ideologies.Underconditions of incomplete social mobilization,
however,thisis notnecessarily thecase. Internal warheremayjust be the
"normal"way of resolvinginterelite forspoilsand provenothing
conflict
one wayor theotheraboutnationalism.
Languagepolicyis a verysensitiveindicatorof satisfaction withthe
nation-state, providedwe keep in mindtwo verydifferent situations, rec-
ognizedmultilingual states(Switzerland, Belgium,India)thatmakeno pre-
 Whatis nationalism?737

tenseabouthavinga singleofficial languageused in all publicbusinessand

imposedbythepublicschoolsystem,and statesinwhichseverallanguages
are (or havebeen)used vernacularly, thougha singleone is imposedas the
officialmodeofcommunication. In thefirstcase, one wouldhaveto know
abouttheextentofbilingualism amongtheeliteand aboutpeople'swilling-
ness to function
in skilledprofessions in a languageotherthantheirnative
one. Multilingualpolitiesare based on varioussubtlecompromises; the
willingnessto forgoa singleofficiallanguagedoes not necessarilycome
aboutwithout The character
strife. ofthesecompromises providestheindi-
catorsforobserving thefashioning ofthenationalmyth. 15

The morecommoncase involvestheimposition of a singlelanguageto

taketheplace ofa numberofvernaculars thatmayor maynotbelongto the
same languagegroupas theofficial one. How can peoplebe persuadedto
abandontheirnativevernaculars in favorofmodernHebrew,Bahasa, and
Swahili?Whatincentivesare givento thosewholearnthelanguage?How
are dissenterspunished?What makes the work of officiallanguage
academiesauthoritative? As recently as 1863,one-third ofFrenchelemen-
taryschoolchildren couldnoteffectively communicate in standard French.
We maywonderaboutthesituation intheSovietUnionnow.Yet thereis no
evidenceof disaffection fromthe nationalmyth.In short,the mannerin
whicha singlelanguageis imposedon people providesclues about the
threshold beyondwhichtheimposition ofa languageis perceivedas a viola-
tionof nationalself-determination.It is a mistaketo thinkofanylanguage
policysimplyas imposition by the bruteforceof the centralizing state.
Speakersof minority languagesmayhave excellentreasonsforcomplying
withthepressuremerelyto maximizetheircareeropportunities; theymay
have equallyrationalgroundsforopposingthe pressure.Everything de-
pendson the policyand economiccontextin whichthe processoccurs.
Rationalchoice criteria,not presumedattachment to primordial cultural
values,explainoutcomes.16
Definitionsand indicatorsare toolsto enableus to say something about
therelationshipbetweennationalism and therationalization of societiesin
thethroesof modernization. Theyare also tools thatshouldthrowsome

15. On Switzerlandsee Carol L. Schmid, Conflictand Consensus in Switzerland(Berkeley:

ofCalifornia
University Press,1981);thisbookscontainsamplequantitative evidenceon lan-
guageandnationalist
symbols.On Belgiumsee ArendLijphart, ed., Conflict andCoexistence in
ofInternational
Belgium(Berkeley:Institute Studies,UniversityofCalifornia, 1981).Kenneth
Jowittshowshowthesevariablesmanifest themselvesin quitedifferent forms in "market"as
opposedto "ordered"societies,inwhichinterpersonal
competition is legitimatedintermsofits
contribution
to theorganicunityof theentiresociety(as in Leninistpolities).See his The
LeninistResponse to National Dependency (Berkeley: Instituteof InternationalStudies, Uni-
versityofCalifomia,1978).
16. For a convincing
demonstration thatgame-theoretic can illustrate
formulations theout-
comesof encounters betweenlanguagecentralizers and speakersof minority languagessee
and CataloniaafterFranco"(Paperreadat theannual
DavidD. Laitin,"PoliticalLinguistics
meeting oftheAmerican PoliticalScienceAssociation,New Orleans,August1985).
738 International
Organization

lighton theconditionsunderwhichthepattern breaksdown,whenrationali-

zationat the level of the nation-state seems no longerto work,and the
searchforalternative politicalconstructs begins.
Thistaskhas notyetbeentackled.All I can do hereis to offera research
agendaraisonne.Threequestionsarise:(1) Whatarethetypicalprocessesof
of an industrializing
rationalization society,in termsof the growthof a
nationalmyth?(2) Must successfuldomesticrationalization occur at the
expenseofinternational harmony? (3) Underwhatconditions does thefail-
ureofdomesticrationalization lead to nonnational ofgovernance?
formulas

National mythsand successful rationalization

I take rationalizationto be "successful"if a nationalmythcomesinto
existence,ifconspiraciesand internal wardo notchallengethatmyth,ifa
consensuallanguagesituationis legitimated by themyth,and ifthemyth
providesa coherentsecularalternative to earlierfragmentedandincoherent
patternsof beliefand conduct.The mythmustcontainthe nonreligious
assertionsthatare designedto givecertainty anddirection to thebeliever's
mindandofwhichAndersonand Gellnerwrite.It maycontainthepatriotic
appealsthatSeers considersvitalfortheimplementation de-
of self-reliant
velopment policies.Successfulrationalization,to use Gellner'sformulation
again,impliesthesubstitution ofa singlecommonhighculturefortheprevi-
ous flourishingof manyfragmented low culturesin thetargetpopulation.
Alternatively,therationalnewcultureis a unified amalgamoflow-andhigh-
culturethemes.
How mightsucha transformation be explained?We mustgo backto the
characteristicsof the people involvedin the modernization process.The
mobilized(eliteand mass)maydiffer fromone anotherinwhether theyare
easilyassimilatedintothe highcultureor whethertheyare differentiated
fromit. On what basis can the mobilized-unassimilated differfromthe
mobilized-assimilated? Whatvariablesmayappearto derailthewholepro-
cess? We mustconsiderthetimingof theeventsthatmobilizepeopleand
renderthemdissatisfied, thevariousdimensions ofcleavagethatmaydivide
theassimilated fromtheunassimilated (suchas race,socialstatus,religion,
language,ethnicity), thespatialdistributionofthetargetpopulation, andthe
ideologiesthatmotivatetheactors.

Timing.Timingof stateinterventionin theprocessofsocialmobilization

can makean enormousdifference is
in theoutcome.If social mobilization
notyetfaradvancedand ifthestatehas appreciablecontroloverthelevers
of symbolcreationand resourceextraction, can be
thenrationalization
achievedbyrelaxingthecriteria educational
ofaccess to high-culture facili-
tiesand statussymbols,thusadmitting theunassimilatedto nearequality.
Theelite,ineffect, co-optsthenewlymobilized
successfully beforetheycan
formulate a counterideology.
 Whatis nationalism?739

Co-optationis mostreadilypracticedwhenthestate"builds"thenation,
whenan effective stateanimatedbyan elitepossessedofa nationalist ideol-
ogyfollowsthepoliciesthatresultin thegradualincorporation of theun-
mobilizedintosociety.It is easierwhenthetargetpopulation is notdivided
byobviouscleavages,especiallyethniccleavages.Anditis facilitated bythe
use ofideologiesthatcan actas dramatic motivators. Co-optationis likelyto
fail,however,whenthe targetpopulationis not homogeneous, whenna-
tionalsentiment precedestheexistenceof a powerfulstate,whenthereis
sharpcompetition amongnationalist ideologies.Contrary toAnderson'sand
Smith'sarguments, underthepropertimingconditions, "officialnational-
ism" can be a veryeffectiverationalizing agent.Its failurein Eastern
Europe,Africa,and the MiddleEast is attributable not to its "official"
characterbutto the absence of the additionalfacilitatingconditions.The
failureofanyformula a societyis almostguaranteed
forrationalizing when
socialmobilizationis compressedintoa singlegeneration andwhenthereis
neithera strongstatenora pervasivenationalist sentiment,as in mostof
Africa.

Dimensionsofcleavage.It wouldbe lovelyifwe coulddemonstrate thata

uniquecombination themobilized-unassimilated
ofvariablesdifferentiating
fromeachotherandfromthemobilized-assimilated couldbe associatedwith
theemergence ofa nationalmyth.We are unlikely to findsucha combina-
tion,despitethecontinuing effortsofsomestudents ofnationalismto single
outethnicity andlanguageas thatuniquecombination. No singledimension
or combination of dimensions can be shownto be neces-
of differentiation
sary,sufficient, to explaintheformation
or necessaryand sufficient, of all
nations.Dimensionsofsaliencediffer withthetiming oftheprocessofsocial
mobilizationandwiththespatialdistributionofthepopulations involved.In
statesand societieswhose populationsdo not differmarkedly on ethnic
grounds, themobilized-unassimilated haveto differfromtheassimilated on
morethanone dimensionin orderto have the incentiveto formulate a
nationalistideology.Simplyspeakinga different languageor beingof a
differentreligionfromthe assimilatedis nota strongenoughincentiveto
demanda nationofone's own. However,in ethnically homogeneous states
and societies,themobilized-unassimilatedneedto differ onlyon thedimen-
fordemanding
sionof social statusto have an incentive a nation-state.

of thepopulations
The spatialdistribution
Spatial distribution. involved
interactswiththe timingpattern.Rationalization is impededif ethnically
diversepopulationsliveintermingledorincloseproximity toeachotherand
iftheprocessesofsocialmobilizationare suchthatthesymbolic andextrac-
tiveresourcesof the stateare taxedbeyondits capabilityto adapt.This
happenswhenseveralmobilized/unassimilated ethnicgroupsmakesimul-
taneousclaimsforgreaterequalityand participation. Theirgeographical
makesthemcompeteagainstone anotherandagainstthedominant
situation
740 International
Organization

theattempted
groups.This was one factorimpeding of the
rationalization
Austro-Hungarian in Africa
and Russianempires;it hindersrationalization
now.

Nationalist ideologies as motivators.It seems likelythat the contentof

particularnationalist ideologieswould make a difference withrespectto
eventualrationalization. The difficultyis againthefactoroftime.Forten-or
twenty-year periods,one mightsuppose,anyofoursevenideologieswould
be effective motivators to adjustto a newkindoflife.Butwhathappensif
the continuedprocessof social mobilization, whenlinkedto thekindsof
frustrations associatedwithethnicor religiouscleavagesor competitively
locatedgroupsofthearoused,undermines thelogicoftheerstwhile success-
fulideologyand givesriseto a newroundofdifferentiation? We knowthat
thishas happenedon manyoccasions-in Belgium,Japan,and Turkey,to
namebuta few.
I hypothesizethatin situationsof ethnicspatialdispersion,wherethe
mobilizing eliteinheritsa weak state,the onlyideologycapable of truly
motivating peopletowarda newlifeis integralism. The Marxistvarianthas
historicallyprovedto be themoreeffective one. We cannottestthehalflife
offascismsinceno fascistregimehas survivedlongenoughto completethe
processofrationalization, thoughFranco'sSpainmay,inadvertently, come
close. If, underthe circumstances stipulated,rationalizationis attempted
through relianceon a liberalor a syncretist-synthetic ideology,thechances
are thatthenewlymobilizedbut stillunassimilated groupswillseek their
salvationin secession,notin loyaltyto thenewdispensation.
I hypothesizefurtherthat liberal ideologiescan serve as effective
rationalizing agentsonlyin societiespossessinga strongstatepriorto the
onsetof massivesocial mobilization, endowedwitha populationthathas
becomelargelymonolingual by thetimeof fullindustrialization, and pos-
sessingtheresourcesforsatisfying mass demands.
Whataboutsyncretist ideologies?The compromises betweensecularand
sacredvalueswhichtheyall attemptsaps theirpowerto rationalize. They
tendtofounder becausetheydo notknowhowtodistinguish thepublicfrom
theprivatein religion.Secularnationalist ideologiesmakethisdistinction
successfully: religionmayflourish as an organizedpursuitwithoutbeing
conjoinedwiththe publicrealm,thoughit mayalso go intodecline.The
syncretist formulas are unwilling to makethisdistinction, insistingthatreli-
gious values and institutions retain public relevance.At the same time,
however,these formulas, by allowing the introduction of some secular
valuesand institutions, tendto undermine themselves and to setthescene
forviolentsocialconflict. Restorative syncretism, by seeking to confinethe
impactof themodernto technology and the military,may be able to buy
time.The efficacy of theformula, of course,dependsfatallyon thetiming
factor.I hypothesize thatsyncretist formulasmaychanneltheprocessof
 Whatis nationalism?741

modernization fora while;but the forcestheyunleashwill successfully

challengetheseideologiesandproduceeithera liberaloran integral
national
myth.

Must successful rationalizationresultin internationalstrife?

Supposethesehypotheseswereaccurate.Whatdo theypredictforthe
relationsamongsuccessfulnation-states? Ignoring structuralconstraints on
foreign-policy choice,theimplications ofthehypotheses forthequalityof
internationalpoliticsare notpleasant.Nationalist mythscontainclaimson
othernations,even afternationalself-determination is achieved.The inte-
gralistmythssuggeststrife, constantstruggle.
insecurity, The Jacobinvari-
antoftheliberalmythlegitimates theimperialism ofthevirtuous, evenifit
stopsshortof acceptingthe inevitability of internationalstrife.Synthetic
syncretism is peaceful,buttheothertwovarietiesofsyncretism areat least
ambivalent withrespectto relationswithothernations.Seers remindsus
thateven in the area of international economicrelations,all thetypesof
nationalism containtheseed forpeacefulas well as bellicoseinternational
contact.
The conventional wisdomhas it thatRousseauwas right:domestichap-
pinesscan be boughtonlyat thepriceofinternational unhappiness. Matters
areevenworseifwe considerthatunsuccessful domesticrationalization, as
inthelateryearsoftheHabsburg,Ch'ing,Romanov,andOttoman empires,
also engenders internationaldisharmony. The argument shouldnotbe over-
stated.True,thepost-Enlightenment periodhas beenone ofactiveinterna-
tional discord. At the same time, that period has also been one of
unprecedented increasesin economicand social welfarein theface of the
steepestpopulationincreasesin thehistory ofhomosapiens.Nor can itbe
convincingly arguedthattheprosperity oftherichesthas beenboughtat the
priceoftheimpoverishment ofthepoorest.In short,thepossibility ofdem-
onstrating a link betweentypes of nationalism,patternsof domestic
rationalization,andtheincidenceofinternational shouldnotbe seenas
strife
theequivalentof a doomsdayforecast.Even ifitweretruethatthehigher
standardsof welfareof the post-Enlightenment worldare a resultof the
competitive prowess of the successfulnation-states-thus implying the
prevalenceofa basicallywarlikeworldorderanchoredin mutually antago-
nisticmercantilisms-the internationalanarchywas notentirely nefarious.
Rousseauwas not wrong,buthe overstatedhis case by exaggerating the
extentofinternational unhappiness.
Nevertheless, the hypothesissuggestedby thisline of inquiryis not a
cheerfulone. It presagesthe continuation of conflict-proneinternational
relationsas longas thenationalmythsnowextantor beingbornreflect the
seven ideologies.Whetheras successes or failuresas rationalizers, the
sevenideologiesspella troubled worldorderevenifSeers'spredilection for
sturdyregion-states wereto becomereal.
742 International
Organization

Can failed domestic rationalizationlead to

internationalrationalization?
I now standRousseauon hishead: can we imaginethatdomesticunhap-
pinesswilllead to international
happiness?Sincethishas nothappenedyet,
we are engagedin a gedankenexperiment.17
Suppose thathithertosuccessfully rationalizednation-statesencounter
difficulties
relatedto technology, welfarecommitments, and international
economicinterdependence. Suppose furtherthatthe same difficultiesbe-
devileffortsof stateswho have notyetfashionedsuccessfulnations.Print
capitalismmayhave fosterednationalintegration. But managedcapitalism
and statesocialismmaynotsucceedin maintaining theintegrationifall of
thefollowing conundramustbe faced:excessindustrial capacity,uncertain-
ties about investment in high-technologyindustries,how to splitup the
resourcesofthesea, deal withinternational debt,andprotectthebiosphere
frompollution, whilealso remaining committed to higherstandards ofper-
sonalwelfarefortheircitizens.Can thesimultaneous pressureofdomestic
turmoiland international interdependencelead to politicalconstructs
that
are quitedifferent fromwhatwe knowand thattherefore implya different
kindofworldorder?
I hypothesizethatbecause of thesepressurespreviouslyacceptedna-
tionalmyths deteriorate. Previouslymobilizedpeopleareno longersuccess-
fullyassimilatedor co-opted.Justas in a previousage theysoughta new
identityin nationalismonce theold identities
ceased beingusefulrationaliz-
ers,thenewlydisoriented mustsearchforan alternativetonationalism.Just
as theirpredecessorscould select itemsfromthe menu of nationalist
ideologies,thenewlydisoriented have choices:autarky, regionalism,inter-
nationalregimes,mininationalism via secession,or a new globalidentity
based on class or religion.

Autarky.Not all countriessufferequallyfromthe cross-pressures of

Some mayhave resourcessufficient
postindustrialism. to enablethemto
meetthese challengeswitha minimum of domesticdisruption and little
dependenceon othercountries,particularly if the volumeof welfarede-
mandscan be reduced.The pressureswillundermine thenationalmythonly
to theextentthatdomesticremediesare unavailing. By no meansis every
manifestation
ofinternational
interdependence likelyto subvert
nationalism.
Largecountrieswithintegral
mythsare mostlikelyto withstand thesearch
foralternative
orders.

17. I havediscussedtheargument thatfollowsmorefullyin TheObsolescenceofRegional

Theory(Berkeley:Instituteof International of California,
Studies,University 1975);"Why
Collaborate?"WorldPolitics32(April1980);and"WordsCanHurtYou: or,WhoSaidWhatto
Whom about Regimes," InternationalOrganization34 (Spring 1982).
 Whatis nationalism?743

Candidatesforthischoiceincludea transnational
Globalidentities. com-
mitment to the creationof a global classless society(Marxism)and the
acceptanceof a global religiouscreed stressingeitherserviceto others
(Christianity)
orthetranscendence ofpolitics(Hinduism, Buddhism).Smith
is absolutelycorrectin showingthateach of thesecan be institutionally,
normatively, and conceptuallyan alternative to nationalism.
However,in
thepracticeof modernpolitics,noneof themhas in factbeen immuneto
nationalism. The reconversion of believersintothe purerformsof these
identitiesdemandsnotonlya rejectionof nationalmythsbuta consistent
rededicationto theoriginalcosmopolitan contentofthesecreeds.Although
the possibility
cannotbe precluded,credibility remainslow. The current
prevalenceoftraditionalandrestorative syncretisms worksagainstthissolu-
tion.The innatesecularismof thefourrevolutionary ideologiescan hardly
be saidtofavora turntowardtranscendence orcompassionate service.The
contentofMarxistintegralism, especiallyin itsantidependency guise,is no
moreconsistent withsucha choice.

Regionalism.The desirefortheregionalintegration
ofpreviously
success-
fulnation-states,
particularly
inEurope,clearlyowes somethingtothedete-
rioration
of earliernationalmyths.However,noteventhemostsuccessful
schemeshaveyetresultedina regionalidentity.
integration Suchan identity
wouldnotconstitute an abandonment butthesubstitution
ofnationalism ofa
largernationforseveralsmallerones. Alternatively,
regionalarrangements
thatfunctioneffectivelywithoutleadingto themergerof sovereignties
are
hardto distinguishfrominternational
regimes.

Mininationalisms.Contemporary effusionsof longdormant ethniciden-

titiesinWesternEuropedo challengethefinalityofthefamiliarnation-states
andcertainly questiontheformer myths.ButI cannotconvincemyself that
theyconstitutesolutionsto thecross-pressures
ofpostindustrialism.
Auton-
omousWales,Brittany, or Euskadiare no morelikelyto deal successfully
withthe turbulenceof the welfarestatethanare the governments from
whichtheywishto secede. Rationalchoice postulates(as well as recent
events)convincemethatwe aredealingwitha temporary phenomenon, not
a seriousalternative
to the nation-states.
existing

International regimes.Such entitiesare alreadyquitefamiliar.Without

concerning ourselvesnow withtheirorigins,I simplyhypothesize thatthe
widespreaddemandformanagednationaleconomiessustaining highlevels
ofwelfarecannotbe metin thecontextofsovereign nation-states.
Meeting
thedemandscallsfornewregimes,andthereconstruction ofexistingones,
withtheconsequencesoffurther undermining theautonomy ofthenation-
stateas it seeksto cope withdomesticsourcesoferodinglegitimacy. How-
ever,countriesanimatedby liberal
and syntheticmyths are mostlikelyto
 Organization
744 International

feelthispressure.Furthermore, widespreaddemandsformanagednational
economiesstriving for rapid industrialization(implying someredistribution
of wealthfromNorth to South), also cannot be met in thecontextof the
sovereignnation-state. Meeting these demands also calls forthecreationof
variouskindsofinternational economicregimes.Integralist as wellas vari-
ous syncretistmyths feel the logic of thispressure. Finally, everybody de-
siresthe benefits of technological innovations. Conversely, the unwanted
consequencesof suchinnovations can be avoidedonlythrough acts of in-
ternationalcollaboration and administration. Maximizing benefits while
minimizing costscalls forthecreationofnonnational practicesand institu-
tionsthatmaytendto undermine further therationalization formula ofthe
nation-state.
Notethatthelogicofthesehypotheses doesnotpredictworldgovernment,
or regionalintegration,or a stronger United Nationsas thelikelyoutcome.
Ifthenation-state loses itsprominent positionas thefontofeffective social
harmony, thealternative maybe all or none of these possibilities.Alterna-
tiverationalizationformulas mayimplydecentralization as wellas centrali-
zation,or bothat once (thoughfordifferent demands and issues).All that
can be affirmed withconfidence is that noneof this is likelyto happenuntil
thelogicof the mobilization-assimilation balancehas runits course,until
happinessrelying on thenation-state has everywhere beentried,untilsocial
mobilization is completein all countries.Whetherwe like nationalism or
not,itseemstobe a necessarystagethrough whichpoliticalmanhastopass.
Even iftherewereno otherreasonsforstudying nationalism, thatconclu-
sionalonejustifies the quest.