CONCEPT CALCULUS: MUCH BETTER THAN

WHATISIDENTITY(ASWENOWUSETHEWORD)?
JamesD.Fearon
DepartmentofPoliticalScience
StanfordUniversity
Stanford,CA94305
email:jfearon@stanford.edu
DRAFT{Commentsappreciated
November3,1999ABSTRACT
Thepaperundertakesanordinarylanguageanalysisofthecurrentmeaningsof\iden-
tity,"acomplicatedandunclearconceptthatnonethelessplaysacentralroleinongoing
debatesineverysub eld ofpoliticalscience(forexample,debatesaboutnational,ethnic,
gender,andstateidentities).\Identity"aswenowknowitderivesmainlytheworkofpsy-
chologistErikEriksoninthe1950s;dictionaryde nitions havenotcaughtup,failingto
capturetheword’scurrentmeaningsineverydayandsocialsciencecontexts. Theanalysis
yieldsthefollowingsummarystatement. Asweuseitnow,an\identity"refertoeither
(a)asocialcategory,de ned bymembershiprulesand(alleged)characteristicattributesor
expectedbehaviors,or(b)sociallydistinguishingfeaturesthatapersontakesaspecialpride
inorviewsasunchangeablebutsociallyconsequential(or(a)and(b)atonce).Inthelatter
sense,\identity"ismodernformulationofdignity,pride,orhonorthatimplicitlylinksthese
tosocialcategories.Thisstatementdi ers fromandismoreconcretethanstandardglosses
o ered bypoliticalscientists;Iargueinadditionthatitallowsustobetterunderstandhow
\identity"canhelpexplainpoliticalactions,andthemeaningofclaimssuchas\identities
aresociallyconstructed."Finally,Iarguethatordinarylanguageanalysisisavaluableand
perhapsessentialtoolintheclari cation ofsocialscienceconceptsthathavestrongrootsin
everdayspeech,averycommonoccurrence.1 Introduction
Inrecentyears,scholarsworkinginaremarkablearrayofsocialscienceandhumanities
disciplineshavetakenanintenseinterestinquestionsconcerningidentity. Withinpolitical
science,forexample,we nd theconceptof\identity"atthecenteroflivelydebatesin
everymajorsub eld. StudentsofAmericanpoliticshavedevotedmuchnewresearchtothe
\identitypolitics"ofrace,genderandsexuality.Incomparativepolitics,\identity"playsa
centralroleinworkonnationalismandethniccon ict (Horowitz1985;Smith1991;Deng
1995;Laitin1999). Ininternationalrelations,theideaof\stateidentity"isattheheart
ofconstructivistcritiquesofrealismandanalysesofstatesovereignty(Wendt1992;Wendt
1999;Katzenstein1996;LapidandKratochwil1996;BierstekerandWeber1996). And
inpoliticaltheory,questionsof\identity"marknumerousargumentsongender,sexuality,
nationality,ethnicity,andcultureinrelationtoliberalismanditsalternatives(Young1990;
Connolly1991;Kymlicka1995;Miller1995;Taylor1989)
Comparedtorecentscholarshipinhistoryandthehumanities,however,politicalsci-
entistsremainlaggardswhenitcomestoworkonidentities.Duetoin uences rangingfrom
MichelFoucaulttothedebateonmulticulturalism,thehistoricalandculturalconstructionof
identitiesofallsortshaslatelybeenapreoccupationforbothsocialhistoriansandstudents
1ofliteratureandculture.
Despitethis vastlyincreased andbroad-ranginginterest in\identity," theconcept
itselfremainssomethingofanenigma.WhatPhillipGleason(1983)observed15yearsago
remainstruetoday:Themeaningof\identity"aswecurrentlyuseitisnotwellcapturedby
1SeeBrubakerandCooper(1999)forsomecitationstothisvoluminousliterature. Forameasureof
thespreadof\identity"inacademicdiscourse,Ichartedtheprogressofthewordindissertationabstracts,
whichcannowbesearchedon-linegoingbackto1981.Thenumberofdissertationabstractscontainingthe
word\identity"almosttripledbetween1981and1995,risingfrom709to1,911.Thisincreasehasoccurred
entirelyinthelasttenyears. Theaverageincreasewasabout12%peryearfor1986to1995,whileitwas
roughly at at-2.3%for1981to1985.Someofthisincreasecouldbeduetoanincreaseinthetotalnumber
ofdissertationsabstracted.Ihavebeenunabletogetthese gures, butIdidtrysearchingyear-by-yearfor
aneutral\controlword"{Iused\study"{togetaroughestimate. Bythismeasure,thetotalnumber
ofabstractedincreasedbyanaverageof.64%peryearfor1981-1985,and4.4%peryearfor
1986-1995.Thusthenumberofdissertationsabstractsusingtheword\identity"hasbeengrowingalmost
threetimesfasterthantherateforallabstracteddissertations.
1dictionaryde nitions, whichre ect oldersensesoftheword.Ourpresentideaof\identity"is
afairlyrecentsocialconstruct,andarathercomplicatedoneatthat.Eventhougheveryone
knowshowtousethewordproperlyineverydaydiscourse,itprovesquitedi cult togivea
shortandadequatesummarystatementthatcapturestherangeofitspresentmeanings.
Giventhecentralityoftheconcepttosomuchrecentresearch{andespeciallyinsocial
sciencewherescholarstakeidentitiesbothasthingstobeexplainedandthingsthathave
explanatoryforce{thisamountsalmosttoascandal.Ataminimum,itwouldbeusefulto
haveaconcisestatementofthemeaningofthewordinsimplelanguagethatdoesjusticeto
itspresentintension.
Thisisthemainpurposeofthispaper,todistillastatementofthemeaningof\iden-
tity"fromananalysisofcurrentusageinordinarylanguageandsocialsciencediscourse.
Themainresultsareeasilystated,althoughafairamountofworkonalternativepossibili-
tieswillberequiredtoreachthem. Iarguethat\identity"ispresentlyusedintwolinked
senses,whichmaybetermed\social"and\personal."Intheformersense,an\identity"
referssimplytoasocialcategory,asetofpersonsmarkedbyalabelanddistinguishedby
rulesdecidingmembershipand(alleged)characteristicfeaturesorattributes.Inthesecond
senseofpersonalidentity,anidentityissomedistinguishingcharacteristic(orcharacteris-
tics)thatapersontakesaspecialprideinorviewsassociallyconsequentialbutmore-or-less
unchangeable.
Thus,\identity"initspresentincarnationhasadoublesense. Itrefersatthesame
timetosocialcategoriesandtothesourcesofanindividual’sself-respectordignity.There
isnonecessarylinkagebetweenthesethings. Inordinarylanguage,atleast,onecanuse
\identity"torefertopersonalcharacteristicsorattributesthatcannotnaturallybeexpressed
intermsofasocialcategory,andinsomecontextscertaincategoriescanbedescribedas
\identities"eventhoughnooneseesthemascentraltotheirpersonalidentity.Nonetheless,
\identity"initspresentincarnationre ects andevokestheideathatsocialcategoriesare
boundupwiththebasesofanindividual’sself-respect. Arguablymuchoftheforceand
22interestofthetermderivesitsimplicitlinkageofthesetwothings.
Insection2belowIjustifytheenterpriseatgreaterlength,arguingthatforcontested,
complicated,orunclearsocialscienceconceptswithstrongrootsinordinarylanguage(i.e.,
mostofthem),acarefulanalysisofordinarylanguagemeaningsshouldprecedee orts to
legislateade nition forparticularresearchpurposes.Section3considerstheinadequacyof
dictionaryde nitions of\identity"andverybrie y tracesthehistoricalevolutionofitsnew
3setofmeanings. Section4beginstoaskaboutthecurrentmeaningof\identity"bytesting
possiblede nitions againstexamplesfromusage.Thetrailleads rst totheformulationof
aidentityasasocialcategory,and,insection6,toidentityasdistinguishingfeaturesofa
personthatformthebasisofhisorherself-respectordignity(andmore).Inbetween,section
5developsapotentiallyvaluabledistinctionbetweenroleand\type"identities. Sections7
and8drawoutsomeimplicationsoftheanalysisfortwoissuesofconcerntosocialscience
usersofconcept.Insection7Iusetheresultsoftheordinarylanguageanalysistoconsider
howidentitiesbearontheexplanationofactions(politicalandotherwise). Insection8I
brie y extendtheanalysisof\identity"appliedtoindividualstocorporateactorssuchas
statesand rms. Acentralargumentinrecentinternationalrelationstheoryholdsthatstate
interestsaredeterminedby\stateidentities."Themeaningofthisclaimobviouslydepends
onthemeaning\stateidentities,"whichIarguemightrefertoanyofseveraldi eren tthings.
Section9concludes.
2 Whybother?
Giventheintenseinterestinidentityandidentitiesacrossabroadspectrumofdisciplines,
onemightinitiallyexpectiteasyto nd simpleandclearstatementsofwhatpeoplemean
2Theaddedvalueofthisstatementofthecurrentmeaningof\identity"isnotthedistinctionbetween
\social"and\personal"sidesperse.Thereisalongtraditionofscholarsdrawingaofthissort,
contrastingvariousformulationsofindividualorpersonalidentity,ontheonehand,andsocialorgroupor
collectiveidentityontheother.Whatisnovelintheformulationderivedhereisthespeci c contentofthe
twosidesofthedistinction(whichcanbeandhasbeen lled inmanyways).
3Foranexcellentandmoredetailedsemantichistoriesof\identity,"seeGleason(1983)andMackenzie
(1978).
3whentheyusetheseconcepts.WhileIhavenotdoneanexhaustivesearch,Ihavenotfound
thistobethecase. Overwhelmingly,academicusersoftheword\identity"feelnoneedto
explainitsmeaningtoreaders.Thereaders’understandingissimplytakenforgranted,even
4when\identity"istheauthor’sprimarydependentorindependentvariable.
Thisisperhapsnotsosurprising. Inthe rst place,whiletheoriginsofourpresent
understandingof\identity"lieintheacademy,theconceptisnowquitecommoninpopular
discourse.Sinceweallknowhowtoemploythewordandweunderstanditinotherpeoples’
sentences,whybotherwithde nitions orexplanations?Second,inpopulardiscourseidentity
isoftentreatedassomethingine able andevensacred,whileintheacademyidentityisoften
5treatedassomethingcomplexandevenine able. Onehesitatestotrytode ne thesacred,
theine able, orthecomplex.
Ofcourse,onecan nd briefde nitions andclari cations inmanyplaces. Theserun
thegamut,fromsuggestiveglossestosomefairlycomplicatedandopaqueformulations.Here
aresomeexamples,culledmainlybutnotexclusivelyfromtheareasIreadmostin(political
science,internationalrelations):
1.Identityis\people’sconcepts