Missing Relations - Columbia Business School 1995). Cultural differences in values, beliefs, and...

Missing Relations

IncorporatingRelationalConstructs

Into Modelsof Culture

MICHAEL W MORRISJOEL M. PODOLNY

SHEIRA ARIEL

Theysayyouare notyouexceptin termsofrelation to otherpeople....Whatyou do, which is whatyou are, only hasmeaningin relation to

otherpeople.—RobertPennWarren,All the King’s Men

A5 an increasingnumberof organizationsreachacrossnationalboundaries,organizationalresearchincreasinglyaddressesthe questionofhownationalculture influencesindividual behavior(Erez& Earley, 1993; Hofstede,1991). Researchersface a choice amongnumerousdefinitions of culture andmethodsfor studyingit that havebeendevelopedin different academicdisci-plines (Keesing, 1974). As the diversity of chaptersin this volume attests,researchersin different subfieldsoforganizationalbehaviorhavechosendiffer-ent methodologicalpathsandhavearrivedat differentinsights.Thepathchosenby mostmicro-levelorganizationalbehaviorresearchersis to study“subjectiveculture,” that is, to conceptualizeculture asexisting in an individual’s mentalrepresentationsratherthan in externalstructuresand artifacts (Triandis, 1967,1972, 1995). Cultural differencesin values,beliefs,and attitudesareassessedthroughinventoriessuchasthosetraditionallyusedto measurepersonalitydif-ferences.Moreover,justasgeneralizationsaboutpersonalityweremadein terms

52

sk2521

Text Box

Used by permission of Sage Publications, Inc.

RelationalConstructs 53

ofunderlyingdimensionssuchasintroversion-extroversion,generalizationsaboutculturalattitudeshavebeensoughtin termsofdimensionssuchasindividualism-collectivism(Triandis,McCusker,& Hui, 1990).Theindividualism-collectivismconstructhasbeenenormouslypopular, in part becauseit promisesa way inwhich to capture,at the level of individual• values, the cultural differencebetween“looselyknit” and“tightly knit” societiesdescribedby classicalsocialtheories(Durkheim,1897/1951;Tonnies,1887/1957).Althoughwereviewmanycontributionsofthis subjectivistapproachto culture,we describeshortcomingsthat haveemergedin the prevailingresearchprogrambasedon individualism-collectivism.Moreover,we arguethat, in principle, the subjectivistapproachisincapableoffully capturingtheinfluenceofcultureon individual behavior.

In this chapter,we arguethat a betterunderstandingof how individualsareaffectedby nationalcultureis achievedwhenthesubjectivistanalysisofinternalattitudesis supplementedwith the structuralist approachof focusing on theexternalsocialrelationsthat constrainbehavior.’Morespecifically,wesuggestthattheconceptsdevelopedfor studyingpatternsof socialrelationsareuseful incultural research.Somerelevantrelationalconceptsrefer to the form or geome-try of an individual’s relations.Density, for example,refers to the degreeofinterconnectednessamongthe otherpeoplein one’snetworks,and densityhasbeenlinkedto theculturaldifferencebetweenlooselyknit andtightly knit societ-ies (Fischer& Shavit,1995).Otherconceptsinvolve both thecontentandformofrelations.Multiplexity, for example,refers to whetheror not one exchangespersonalfriendshipandinstrumentalresourcesin thesamerelationship,andthisrelationaldi~iensionhasbeensuggestedby severaltheoriststo explainculturaldifferencesin workplacebehavior(Dore, 1983; Gluckman,1967). We reviewthesparseliteratureon cross-nationaldifferencesin relationalvariablesandcon-cludethat thereis a greatdealof unexploredpotential in capturingthe differ-encesamongculturesin termsofthe differingrelationsin which individualsareembedded.However,we also review theoreticalargumentsagainstthe notionthatculturecanbe reducedto apurelystructuralanalysis.

The approachto culture that we advocate integratessubjectivist andstructuralistanalyses.We arguethat internationaldifferencesin interpersonalbehaviorin theworkplacearebestunderstoodin termsof systemsof norms. Inthetraditionof Weber(1922/1963)andParsons(1951),we assumethat a normexistsboth in theinternalsubjectiveattitudesthat individuals hold aboutinter-personalrelationsandin theexternalstructureofrelations.A norm is thebasisofthe local socioculturalsystemthat organizesthe interactionwithin a circle ofemployees.Importantly,wedo not makethe assumptionthatnormsaresharedandorganizedat a societywidelevel, anassumptionfor whichclassicaltheoristshavebeencritiqued(Wrong, 1961).Wemerelyassumethatemployeesin a givenworkplacecontextconstructandsustaina local patternof interaction(Giddens,1984; Swidler, 1986).2Moreover,to a largeextent,thesesystemsarebasedon

54 M~CRO-ORGANIZAT1ONAL APPROACI-ES

prior forms of sociality suchas market, family, legal, and friendshiprelations(Fiske, 1991).Thus,theroleofnationalculture is indirect; its role lies in shapingwhich normsare appropriateto organizea local system of interaction in theworkplace,andthis local systemin turn guidesindividual behavior(Lonner&Adamopoulos,1997).

By reviewingourown studyofNorth American,Chinese,German,andSpan-ish employeesofa multinational corporation(MNC) (Morris, Podolny,& Ariel,1997), we illustrate the advantagesof a sociocultural-system~nalysis over apurely subjectivistor structuralistanalysis.Not only doesa systemslevel ofanalysishelp to distinguishbetweenindividualist and collectivistsocieties, italsoprovidesa morefine-grainedunderstandingof qualitativelydifferent formsof collectivism.Traditional unidimensionalanalysesof collectivismservewellto contrastthe English-speakingnationsfrom most of the world’s othermajorcultural traditions,which accordmoreconcernto socialcollectivitiesand lesstoindividuals(Hofstede,1980, 1991).In itself, this contrastcomportswell with theethnographicandhistorical record.Comparedto WesternAnglophonesocieties,thereseemsto bea greateremphasison social relationsand collectivitiesin theChinesesocietiesin EastAsia, in socialdemocraticnationsin NorthernEurope,in SouthernEuropeannationson theMediterranean,andin mostother largecul-turalgroups.Yet, theethnographicrecordalsomakesit clearthatnotall ofthesedifferingculturalgroupsemphasizethesametypesofsocialrelationsandcollec-tivities. Simply put, not all collectivist societiesarecollectivistic in the sameway. In this chapter,wemakethe casethat conceptualizingculture in termsofsocioculturalsystemshelpsto clarify thequalitativelydifferent forms of collec-tivist orientation.

TRADITIONAL APPROACHESTO CULTURE

SubjectivistApproach

Historical roots. Theapproachof accountingfor societaldifferencesin termsofsubjectivepsychologicalcharacteristicshasa long tradition in anthropology.Aninfluential movementin this tradition wasthe researchprogramlinking cultureandpersonality.EthnographerssuchasBenedict(1934)andMead (1935)wereimpressedby thevariationacrosssocietiesin patternsofsocialbehaviorandthestabilityofthosepatternswithin societies.In keepingwith psychoanalytictheo-ries oftheirday, theyexplainedpatternsofbehaviorin termsof amodalperson-ality type, transmittedacrossgenerationsthroughprimary societal institutionssuchaschild-rearingpractices(Kardiner,1939; Mead, 1935).Thelimitations ofthis approacharemostevident in theWor d War 11-erastudiesofthe “national

55RelationalConstructs

• character”of the United States’ military adversaries.For example,basedonobservationsof andinterviewswith JapaneseprisoneLrsofwar,Gorer(1943)pos-ited that patternsof Japanesebehaviorreflect an anal-compulsivepersonalitytypeinculcatedthroughearlytoilet training. Themost obvious empirical short-comingof thiswork is thatno attemptwasmadeto measuretheunderlyingper-sonalitytype(Inkeles& Levinson,1954).Closelyrelatedto thisempiricalshort-coming is thetheoreticalshortcominginherentin thesimplisticassumptionthateachsocietyis characterizedby a predominantpersonalitytype. In thewords ofWrong(1961)’ suchanapproachimpliesan “oversocialized”view ofindividualsandan “overintegrated”view ofsociety.

Latersubjectivistapproacheswereshapedby thecritiquesof theculture andpersonalitymovement.Researchersshifted from the method of unstructuredethnographicobservationto surveysofrepresentativesamplesof differentsoci-

• eties. Thesesurveysrevealedthe considerableheterogeneityof personalitywithin societies(Cantril, 1965). Moreover,the useofquantitativesurveymea-suresledresearchersawayfrom hypothesesframedin termsofunifiedbroadper-sonalitytypestowardhypothesesarticulatedin termsof more distinctspecificvalue and attitudevariables.The complexity inherentin describingsubjectiveculture led researchersto seekmoregeneral,parsimoniousconstructsthroughstatisticalabstraction(Osgood, 1964), in favor of “thick description.” (Yet,anthropologysoon turnedtoward rejectionof abstractsubjectivistmodelsinGeertz,1976).Two statisticalapproachesarenoteworthy.First, clusteranalysishasbeenusedto form culturalgroupingsof societieswith historical, linguistic,and geographicsimilarities suchasthe SouthernEuropeanLatin societiesandthe EastAsianChinesesocieties(Ronen, 1986; Ronen& Shenkar,1985). Sec-ond,factoranalysishasbeenusedto uncovertheunderlyingdimensions,suchasindividualism versuscollectivism, that accountfor variance among culturalgroups(Triandis, 1967, 1972). The secondapproach,which promisesa moreparsimoniousaccount,hasbeenmoreinfluential.

Recentcontributions. Recentsubjectivistresearchon culture hasbeenlargelyfocusedon theculturaldimensionof individualism-collectivism.This constructgainedprominencein thewake of Hofstede’s(1980)worldwide studyof IBMemployees.Hofstedefoundthat individualism-collectivismcaptureda substan-tial fraction of the varianceacrosscountriesin meanlevels of endorsementofbasicvaluesaboutlife andwork. In addition,Hofstede’sfindings revealedthatthe United Statesand the few other WesternAnglophone countriesscoredextremelyhigh on individualism relativeto the othercountriesin the sample.This finding raisedthe concernthat organizationaland social psychologicalresearchbasedalmost exclusively on thesefew hyper-individualist Westernsocietiesmight not generalizeto themorecollectivistsocietiesin which mostoftheworld’s populationresides.As a result,a wave of studiescomparingsocial

MICROORGAN1ZATIONfL /~ ~IOACH ~S

psychologicalvariablesin highly individualist and highly collectivistcountrieshasaddressedthis question(for areview, seeTriandis, 1995).

n attemptingto investigatedifferencesbetweenindividualist ai d collectivistcultures,Triandisandcolleaguesdevelopedinventoriesto comp~re severaltypesof socialattitudesbetweentheUnitedStatesand morecollectivistAsian or Latincountries(Triandis,Leung,Villareal, & Clark, 1985,Triandiset al., 1990).Othershave investigatedthe consequencesof individualism-collectivismin behaviorratherthanmerely in attitudes.LeungandBond (1984)tested ~ypothesesabouthowthecollectivismdimensionaffectshow peoplehandleconflictwith in-grouporout-groupmembers.In ataskthat requiredeachparticipantk ivide resourceswith a friend or a stranger,Leung and Bond observeda greaterdegreeof in-groupfavoritism, or generositytoward friends, amongChinesestudentsthanamongAmerican students.Likewise, in conflict resolutiontasks, researchershave found a greaterpreferencefor nonadversarialprocedureswith in-groupmembersamongChineseand SpanishstudentsthanamongAmericanstudents(Leung, 1987; for a review, seeLeung, 1997). Earley(1989, 1994) foundthatChinesemanagers,relativeto Americans,were less inclined to exhibit socialloafing in a grouptask,apparentlybecauseofa concernfor groupsuccessthatrivals theirconcernfor individual success.

Critiques and current directions. Although the constructof individualism-collectivismhasbeenthe mostfruitful researchprogramwithin the subjectivistapproach,therehasbeenan increasingnumberofcritiquesoftheconstruct.First,compellingtheoreticaland empirical objectionshavebeenraisedin researchonattitudesand valuesagainstthe notion that individualist and collectivist itemscanbearrayedon asinglebipolardimension(Kashima~etal., 1995).Manydiffer-entcomponentsofcollectivismhavebeendistinguished(Ho & Chiu, 1994).Sec-ond, studiesof socialbehaviorhaveindicatedthat not all forms of collectivismarealike. For example,studiesof Latin societieshave emphasizedthat socialinteractionsare characterizedby outwarddisplaysof warm emotion. Even inworkplaceinteractions,a personcreatesa harmoniousfeeling throughwarm andexpressivebehavior,a tendencyreferredto assimpatiain Latin America(Diaz-Guerrero, 1967; Lindsley & Braithwaite, 1996; Sanchez-Burks,Nisbett, &Ybarra, 1998; Triandis,Mann, Lisansky,& Betancourt,1984).This obligationof emotionalexpressivitycontrastssharplywith observationsaboutexpressivedisplaysin manyotherhighly collectivistsocieties.For example,in theChinesevirtue oflen (Hsu, 1985)and theJapanesetradition of amac (Doi, 1962), har-mony is createdthroughpassivityratherthanexpressivity.Hence,althoughtheConfucianandLatin traditionsboth emphasizeharmoniousrelations,they seemto createdifferent typesof harmoniousrelations.Third, scholarsof businessanddiplomatic p ac iceshavefounda needfor morenuancedcha acteriz~tions. Forexampe, observa onsof~n enslycompetiivebehaviorby C~uinesen ~gotiators


haveled researchersto seekconstructsotherthana generallycollectivist valueorientation(Pye, 1982). In sum, the promiseof a parsimoniousreductionofmyriad cultural differencesto a unitary individualism-collectivismdimensionincreasinglyseemsto be false.

Subjectivistresearchershave attemptedto redressthe limitations of theindividualism-collectivismconstructby proposingmoremultifacetedmodelsofculturalvalues.Schwartz(1994)andothershaveempiricallydistinguisheddif-ferentstrandsofindividualist andcollectivistvaluesalonglinesmuch like thosedrawnby sociologistsandintellectualhistorianswho analyzedthesevalueswithdifferentmethods(Bellah, Madsen,Sullivan, Swidler, & Tipton, 1985; Lukes,1973).Triandisandcolleagues(Triandis, 1995; Triandis& Gelfand,1998)havedevelopedameasureofsocialvaluesandattitudesthatdistinguishestheeconom-ic aspectofcollectivism(sharingresourceswith thegroup)from theexpressiveaspectofcollectivism(affective involvementwith thegroup)anddistinguishestheeconomicaspectof individualism (belief in achievementthroughcompeti-tion) from the self-expressiveaspectof individualism (desireto be unique).3Althoughthedistinctiveconsequencesofthesespecificattitudedimensions-ha-yenot yet beendelineated,the first threeof these,on facevalue, seemto capturenormsrelevantto workplaceobligations. With regardto countrydifferences,thereis asyetlittle data.Singelis,Triandis,Bhawuk,andGelfand(1995)foundthatAmericanstudentsofan EastAsianbackgroundarehigheron economiccol-lectivism(sharingresources)thanarethoseofaWestEuropeanbackground,andtheyconjecturedthatthis typeofcollectivismshouldbe highestin asocietysuchasChina,wherethe primarysocialvaluehasbeenneitherequalitynor freedombutratherstabilityofthesocialorder.In addition,Singelisetal. conjecturedthatexpressivecollectivism(affectiveinvolvement)shouldbehighestin communalgroupssuchasthe Israeli kibbutz, whereequality is morehighly valuedthanfreedom.By contrast,economicindividualism (achievementthroughcompeti-tion) shouldbe highestin Westernmarket democraticsocietiessuchastheUnitedStates,wherefreedomis valuedoverequality.We find thesemorefine-grainedconstructspromisingandwill returnto themlater.

However,in addition to the limits of particularsubjectivistconstructs,thereareobjectionstothe subjectivistapproachitself, whichby definitionreducescul-ture to a set of factorsthat exist “inside the individual’s head.”Although cer-tainly anactor’sbehavioris greatlydeterminedby internalfactorssuchasvaluesandattitudes,behavioralsois greatlydeterminedby theexternalsocialsituationsthat theactorfaces.Decadesofdebateby psychologistsoverdispositionalversussituationalinterpretationhaslargelyconcludedthattheproximalcausesofactionare not personalitydispositionsbutrathersituationalfactorsandinteractionsofsituationalfactorsanddispositions(Ross& Nisbett, 1991).Becausethe subjec-tivist approachto cross-nationaldifferencesrestrictsitself to dispositionsascausesofbehavior,it is limited in muchthesamewayaswastraditionalperson-

58 MJCRO-ORGANJZATJONAL.iPPROACHES

ality research.Interestingly,this bias of dwelling on dispositional causesofsocial behavior might itself reflect the Westerncultural orientationof mostresearchers(Morris & Peng, 1994). In any case,the subjectivistaccount,byfocusingon internalcausesof behavior,is limited in principle to capturingonlyhalfof thepicture.

StructuralistApproach

Historical roots. There is a long scholarlytradition of analyzingbehavioraldifferencesacrosscountriesin terms of aspectsof the socialstructure.Marx’s(1844/1972)argumentsconcerningalienationsuggestthat reducedsubjectivequality of life resultsfrom the patternsofrelationsfosteredby capitalistinstitu-tions (Lukes, 1967).Durkheim (1893/1933)attributedcountryandethnicdiffer-encesin suicideratesto the tightnessof the social fabric. Even Weber(1922/1963),althoughmoreknownfor his contentionthatsubjectivevaluesandbeliefscancausechangesto thesocialstructure,alsomaintainedthatoncesocialstruc-turesandpositionsareestablished,theyshapevalues,attitudes,andbeliefs.

A majorresearchprogramtestinga structuralistanalysisofinternationaldif-ferencesin attitudesandvalueswas Inkeles’ (1960)workon whathascometo becalledthe“convergence”thesis.Theconvergencethesisholds thatthespreadofindustrializationand the resulting exposureto similar social structuresin theform of institutions,suchasfactoriesand schools,haveled peoplein differentcountriesto holdsimilar values.In supportofthis structuralistargument,Inkelesfoundthat countrydifferencesin valuesaresubstantiallyreducedwhenlevel ofindustrializationis controlled.Yet, countrydifferenceswereby no meanselimi-natedin Inkeles’ findings.Moreover,increasinglysalientcounterexamplessuchas Japan,highly industrializedyet highly distinctive in its cultural values(Pascale& Athos, 1981),led scholarsto becomedisenchantedwith the ideathatindustrializationinevitably is linked with onesetof values.

After decadesof furtherwork, Inkeles(1978)proposedseveralvaluablerec-ommendationsfor researchon internationaldifferences.Onewasthattheimpactofmacro-levelstructuralvariables,suchasacountry’slevel of industrialization,shouldbeunderstoodin termsof how thesevariablesshapemicro-level socialstructures,which providetheproximal causalinfluenceson themotivationsandbehaviorof individuals.A secondrecommendationwasthattheeffectsofmicro-level structuresshouldbe understoodin conjunctionwith thesubjectiveunder-standingsthat guide people’sresponsesto the situations.As we shall see,thefirst recommendationto focus on micro-level structures is a direction thatstructuralistresearchershavetaken.Our approachalso takesthe secondstepofincorporatingpeople’ssubjectiveconstructsasamechanismin theirresponsestosocialstructures.


Recentcontributions.During recentdecades,structuralismhasreemergedin so-ciology. To a greaterextent thanclassicaltheoristssuchasMarx (1844/1972)andDurkheim(1893/1933),thenewstructuralistshavetriedto reducethecausesofsocialphenomenato thepure geometryofsocialrelations,removingan.yref-erenceto subjectivefactors(Mayhew, 1980; Mayhew& Levinger, 1976).Thismovementhasbeenaidedby advancesin themethodsof conceptualizingsocialstructuremadeby the field of networkanalysis(White, Boorman,& Breiger,1976).Whereasmanynetworkanalyticstudiesfocuson thestructureof macro-level ties amongorganizations(for areview, seeLincoln, 1982),otherstudieshavefocusedon the structureof micro-level ties amongindividual employeeswithin an organization(Burt, 1992). The networks of employeeshave beenlinked to severaltypes of organizationallyrelevant attitudes and behavior(Ibarra,1997).Themostinfluentialaccountsofemployeenetworkshavefocusedstrictly on theform ofrelations,that is, on networksizeanddensity(Burt, 1992).However,othershavearguedthattheeffectsof relationalform dependon con-tent; for example,theformsthatworkbestin relationto peersmight notwork inrelationto thosein power(Emerson,1962; Podolny& Baron, 1997).

Therehasbeensometheoreticalandempiricaluseof relationalconstructsinaccountingfor cultural differences.In particular, investigatingcountry differ-encesin theform of individuals’ relationalnetworkshasbeenseenasa way inwhich to capturedifferencesin themicro-level social situationsthat constrainbehavior.The most prominenthypothesisin this work is the highly intuitiveargumentthat collectivist societieswith tightly wovensocial fabricsshouldbecharacterizedby networksof high density(Gross& Raynor, 1985). This argu-mentis interestingin thatdensitymightoffer awayin whichtomeasurethebasicassumptionthat collectivistsocietiesarecharacterizedby tight in-groups.Also,it would serveasabasisto questionsubjectivistexplanationsfor countrydiffer-ences.Forexample,thegreatertendencyofChinese(ascomparedto American)participantsto showgenerositytowardfriendshasbeenascribedto Chinesecol-lectivist attitudes(Leung& Bond, 1984).It might be thatweall areparticularlygeneroustowardfriends who are interconnectedwith otherfriends (manyrela-tions areat stake)andthat Chinesefriendshipstend to be embeddedin densecliques.

Althougha structuralistapproachto collectivismappealsintuitively, empiri-cal findingsaremixedon the questionofwhetherindividualsin collectivistsoci-etieshavedensernetworks.A review of findings from similar network surveysconductedin different countriesconcludedthat density levels are higher inChineseandIsraelicities thanin AmericanandBritish cities(Fischer& Shavit,1995). Furthermore, a study of interaction patterns amonguniversity studentsfoundthatI-long Kong studentshavegroupinteractionsmorefrequentlythandoNorthAmericanstudents(Wheeler,Reis,andBond, 1989).Although this studydid not measurerelationshipsandstructuralpropertiessuchasdensity,thefind-

60 MICRO-ORGANIZATIONAL APPROACHES

ing thatstudentsin acollectivistsocietytendtowardgroupinteractionsis consis-tentwith apicture ofa clique-likesocialstructurein which aperson’sfriendsarefriendswith oneanother.Yet, the only previousstudy (to ourknowledge)thatmadea controlledcross-culturalcomparisonofnetworkdensitydid not observethepredictedpatternoflowerdensityin thefriendshipnetworksofNorthAmeri-canscomparedto thoseof EastAsians(Kashimaet al., 1995).In sum,themostpromisingoperationalizationof a purelystructuralistanalysisof cultural collec-tivism hasnot receivedconsistentsupport.4

Critiquesandcurrent directions. Despitethe facevalidity ofthe ideathat indi-vidualsin collectivistsocietieshavedensernetworks,comparisonsof the aver-agedensityofrelationshavenotuncoveredaclearpattern.More important,therealsohavebeentheoreticalobjectionsto accountsthat explain actionin termsofthepureform orgeometryof socialrelationswithoutreferenceto theircontentormeaning.Oneobjectionis thatthis accountfails to providea rolefor individualagencyin creatingandperpetuatingthesestructuresofsocialrelations(Giddens,1984).Anotherobjectionis thattheeffectsofstructuralpositionson behavioraremediatedby subjectivebeliefsandnormativecommitments,andthatmight welldiffer acrossculturesandyet the crucial role of subjectiveelementsis unspeci-fied in structuralistaccounts(Brint, 1992;Emirbayer& Goodwin, 1994).Thesecritiquesdo notdisputethatrelationalmeasuresandnetworkanalysis,in particu-lar, offer powerfultools for conceptualizingsocialsituations;theymerelyrejectthe notionthat anuancedaccountof culture andactioncanbe framedpurely intermsofthestructure,ratherthanthecontentormeaning,ofsocialrelations.

Althoughwehaveseenthatpurelystructuralapproachesto capturingculturaldifferencescanbecritiquedon empiricalandtheoreticalgrounds,thereareotherrelationalanalysesthat incorporateboth the form andcontentofrelations.5Forexample,manyethnographicscholarshavecharacterizedcollectivistsocietiesintermsof themultiplexityofrelations,that is, thetendencyto havetwo typesofrelationshipswith the samepersonsuchasa coworkerwho also is a relative.Studiesoforganizationsin collectivistsocietieshavepointedto the‘importanceof relations that combinesentimentalsocioemotionalcontentwith pragmaticexchangeofresources.For example,multiplexity hasbeennotedin studiesofrelationsamongJapanesebuyersandsuppliers(Dore,1983),in relationsofJapa-neseemployeeswith theirsupervisors(Rohlen, 1974),and in peerrelationsofAfrican workers (Gluckman,1967; Kapferer, 1969). Although thesescholarshaveassertedthat the multiplexity they observedis greaterthanthat whichwouldbefoundin comparableWesternsettings,we knowofno previousstudiesthat haverigorouslytestedwhethermultiplexity is higherin collectivistsocieties.

Anothercontent-orientedareaofrelationshipresearchfocuseson theaffectiveclosenessor socio-emotionalintensityof socialrelations.As notedearlier,stud-jes ofLatin culturalsettingshavestressedthehigh level ofaffectivity in relations


(Triandis et al., 1984). However,thereare theoreticalreasonsto believethataffectivity is not alwaysassociatedwith collectivism.The classiccomparisonoftheUnited Statesand Germanyby Lewin (1948)arguedthat the more individ-ualist North Americancultural context allows closenessin friendships to beachievedmorerapidly thanin theGermancultural context.

Yet anotherapproachfocuseson the durationofrelations, linking collectiv-ism to relationsof high longevity. Comparativestudieshavefoundthat NorthAmericanrespondentstendto have briefer friendshipsthanthosereportedbyChineserespondents(Gudykunst,1983).Although for a given individual thedurationof relationsmight be associatedwith their affectivecloseness,modellevelsof durationandaffectivity maynotbeassociatedacrosscountries.Indeed,NorthAmericanrelationsmaybecomeaffectivelyclose in part becausenonnspermit one to exit quickly. Consistentwith this picture of superficialor non-bindingfriendshipin theNorthAmericansetting,studieshavefoundthatAmeri-canstudentsengagein briefer and more frequentsocial interactions(Wheeleretal., 1989).

In sum,althoughculturaldifferencescannotbereducedpurelyto thestructureof socialrelations,analysesincorporatingtheform andcontentofrelationsmayoffer new insights into how individual behaviordiffers acrosscultures.Mostlikely, different relationalconstructswill be usefulfor distinguishingdifferentcultures.What is neededis a level of analysisfor conceptualizingculturethatintegratesthe causalroles playedby the externalpressof socialrelationsandtheirsubjectivemeanings.

AN INTEGRATED APPROACH:SOCIOCULTURAL SYSTEMS

- — -~

Organizationalscholarstraditionallyhaveapproachedthetopic ofculturecarry-ing thebiasesoftheiracademicdisciplines,with psychologiststendingto reducecultureto thesubjectivecontentsofindividual thinking andsociologiststendingto reducecultureto the propertiesof social structuresand institutions.Increas-ingly, however,socialscientistsconcernedwith culturehavecalledfor a level ofanalysisthat encompassestheimprint of aculturein the individual’s thoughtsaswell asin thesocialstructuressurroundingtheindividual.Ontopicsasdiverseassocialbehavior,mentalhealth,andeducation,researchershavelookedfor waysin which to framehypothesesthat integratetherolesofsubjectiveandstructuralfactors(for reviews,seeFiske,Kitayama,Markus,& Nisbett,1998;Wertsh,delRio, & Alvarez, 1995). Our own researchattemptsto integratesubjectiveandstructuralelementsin an analysisofcultural influenceson workplaceinteraction(Morris, Podolny,& Ariel, 1997, 1998a,1998b).Ourpointofdeparturein devel-

62 MICRO-ORGANIZATIONAL AV PROACHES

oping thisanalysiswasthework ofParsons(1951)on how socialaction is struc-turedby systemsof sharedvaluesand norms, althoughourassumptions,con-structs,and methodsareinfluencedby recentwork in both the subjectivistandnetworkstructuralisttraditions.

Beforerecountingthepointsmadeby Parsons(1951)on which wewill draw,it is worthaddressingthequestionof“Why Parsons?”CritiquesofParsons’proj.-ectarewell known,andby invoking selectedParsonsianconstructsasapoint ofdeparture,wedo notwish to imply an endorsementof all thefeaturesofParsons’stance.Much like the culture and personalitytheorists,Parsonsconceptualizedvaluesasinternalizedin childhoodandassharedon a societywidebasis.Not anempiricist, Parsonspresentedonly illustrative anecdotesratherthan systematicevidencefor his claims.For decades,Parsons’work hasbeencriticizedfor itshighly consensualandfunctionalistview ofvaluesand for its lackof empiricalunderpinnings.6In ourview, however,suchobjectionableaspectsoftheParsonsproject do not undermineits singularcontributions.To a greaterextent thanprior or subsequenttheorists Parsonsdelineatedthe role of norms in inter-personalconduct.For Parsons,norms arestandardsthat an individual, or “ego,~~relieson in judginghow to acttoward agivenotherperson,or“alter.” Proposalsat this level of analysiscangeneratehypothesesaboutbothspecific subjectivesocialvaluesandspecific featuresof social relations(althoughParsonshimselfdid not draw or test suchpredictions).Second,some of Parsons’descriptiveclaims aboutparticularculturesstrike us aspromising ways in which to inte-grateobservationsby ethnographersandcross-culturalresearchers.

Parsons’proposalsabout particularnorms were framed within a patterntypologythat resultsfrom answersto two dilemmas.Thefirst dilemma,termeduniversalismversusparticularism,centerson whetheran egoshouldmakedeci-sionson thebasisof generalprinciples(i.e., criteriaall actorswould be expectedto usein the situation)or on the basisof idiosyncraticaspectsof the situationincludingaspectswith meaningorrelevancespecificto an ego,analter, andtheweb of surroundingrelations.The seconddilemma, termedascription versusachievement,concernswhethertojudgealterson thebasisofwho theyare(i.e.,thegroupsandcategoriesto whichtheybelong)or on thebasisof what theydo(i.e., their performances).Importantly, Parsonsarguedthat answersto onedilemmaare reachedonly in combinationwith answersto the otherdilemma.7Four qualitatively different resolutionsof thesedilemmaswere illustratedbyParsonsin termsoffourmajorcultural traditions,asfollows: -

I. The universalist-achievementorientationis exemplifiedby theprevailingnorma-tive standardsfor interpersonalconductand pattern of behavior in the UnitedStates.Relativeto othersystems,an egois expectedto conductbehaviortowardanalter in theway that servesthe ego’sinterestswithoutbeinggreatlyconstrainedby


thepresenceor absenceofprior relationsto thealter or by thesocialcategoriestowhich-thealter belongs.In Parsons’(1951) idiom, treatmentofanotherpersonisbasedon “performancesindependentofrelationalfoci” (p. 183).

2. Theparticularistic-achievementorientationis exemplifiedby Chinesesociety.IncontrasttoscholarswhohaveportrayedAmericanandChinesesocietiesasculturalantipodes,Parsons(1951)regardedthetwo societiesassimilar in theirorientationtowardachievement.In bothsystems,anego is stronglymotivatedto achievesuc-cess.However,unlike theAmericantraditionofgettingaheadby leavingor over-turningthesocialorder,theConfuciantraditionrequiresachievementwithin asta-ble orderingof powerrelations: “achievementshould . . . be focusedon certainpointsofreferencewithin therelationalsystemofsuperiority-inferiority” (p. 195).A key determinantofan ego’sactiontowardan alter, then,is whethertheegohasan establishedhierarchicalrelationto thealter.

3. For Parsons(1951),Germansocietyservedto illustrate a universalist-ascriptivesystem.In this system,an ego’streatmentofan alter dependsnot so much on adirect relationto thealter as on thealter’s formal positionwithin an impersonalsystemof classification,that is, “on classificatoryqualities, . . . on statusratherthanon specificachievements”(p. 192). Parsonsarguedthat this typeof systemalso involvesa typeofcollectivist orientation,but theconcemis for broadercol-lectivesthataredefineduniversalistically(e.g.,categoriesofoccupationornation-ality) ratherthanparticularistically(e.g., thehandfulofalterswith whom theegohasarelationship).

4. Finally, for Parsons(1951),Spanishcultureexemplifiedtheparticularist-ascriptivesocial system.In this system,an actor’streatmentof anotheris notstructuredby-instrumentalpurposesorby bureaucraticcategories;sociality-ispursuedasanendin itself.Thatis, “emphasisis thrownin theexpressivedirection”(p. 199).In short,this final cultural systemcenterson concernfor aparticularisticallydefinedcollec-tive (e.g.,one’scircle offriends)that carrieswith it a socialsystemofaffectivelycloserelationships.

The forgoing review of Parsons’(1951)descriptionsof interpersonalinter-action in NorthAmerican,Chinese,German,andSpanishsettingsillustratesthelevel ofanalysisatwhichwewishto work. AlthoughParsonsianideashavebeenout offavor during the pastseveraldecades,therearea few scholarswho haveworkedto refinethis level ofanalysis.Contraryto Parsons’emphasisonabstractvalues,Garfinkel (1967)arguedthat, although abstractcultural valuesmightexist,theyarenottheproximal causesof socialbehavior;morespecific,concretesubjectiverulesareat work. Indeed,whatwe find mostvaluablein Parsonsarethemoreconcretenormsor standardsfor behavior.Contraryto Parsons’notionthat normativesystemsare internalizedin childhoodandexist in a societywideequilibrium, Swidler(1986)arguedthatconsensusof valuesandpatternsofrela-

64 MICRO-ORGANIZATIONAL A9P~~OACHES

tionsariseandexistwithin delimitedcontextsof adult life. This argumentis inkeepingwith theParsonsianassumptionthatsystematicityarisesnot throughthepressurefor cognitiveconsistencyin an individual’s beliefs but ratherthroughthepressurefor complementarityof interactionwithin a socialcircle.

How arenorms developedwithin a specific social context?Fiske (1991)arguedthat, acrosscultures,thenorms governingparticularsocialcontextscanbe seenasconstructedfrom the elementsofbasicmodelsofsocialrelationssuchasthosecharacterizingmarkets,hierarchies,and friendships.Fora givensocialcontextsuchasthe workplace,cultural traditionsdictatehow thesemorebasicnormsareenacted.This is not to saythatall oftheworkplacesin agiven societywill developan identicalnormativesystem.On the contrary,thesesystemswillvary dependingon particularaspectsof thesocialcontextsuchasthe organiza-tional structureand the natureofthe work. In addition, it is likely that certainnormswill prevail in certaintypesof workplacesregardlessof the country; forexample,authoritynormsaremorelikely to prevail in amilitary regimentthaninan artist’scooperative.Nevertheless,it still is possiblethat countrydifferencesexistwhen otherfactorsareheldconstant.In the next section,we statepredic-tions aboutceterisparibuscountrydifferences.We reporttestsof thesepredic-tions in an internationalorganizationthatdeliberatelyholds formal structureandwork tasksconstantacrossdifferent countriesofoperation.

A FOURFOLD PROPOSAL

Our level of analysisreflectsrevisedconceptionsoftherelationsamongculture,norms,andbehavior.We suggestthatcoordinatedworkplaceinteractionrequiresa systemof normsgoverninginterpersonalbehaviorand that nationalculturedetermineswhich norms arecarried into theworkplace.Dependingon thenor-mativesystemthat is established,workplacerelationsresemblethosein a mar-ket, a family, or anotherbasicarenaof socialinteraction.Importantly,thesenor-mativesystemsshouldbemanifestboth in subjectivesocialvaluesaboutselfandothersat workandin thepropertiesofsocialrelationsin theworkplace.We fol-low Parsonsin offering a comparativedescriptionofNorthAmerican,Chinese,German,and Spanishsystems.Unlike Parsons,wedo not regardtheseculturalsettingsasexemplifyingfourbasictypesof socialsystems.Theyaremerelyfourof many cultural traditionsthat differ in the norms enactedin the workplace.Accordingly,welabeleachproposedsystemin termsof its historicalgrounding.

In developingthe fourproposals,westartwith ethnographicfindings to sub-stantiateourdescriptionofthenormsemphasizedin eachsystem.Then,westatepredictionsin termsofspecificsocial valuesidentifiedin recentwork onsubjec-


tive culture (Triandis& Gelfand, 1998)andin termsof thepropertiesof socialrelations.Theseareceterisparibuspredictionsaboutthenormsthat characterizework relationswithin particularculturaltraditions.As weshallsee,thesepredic-tionsdepartfrom thoseofthetraditionalsubjectivistargumentthatcountriescanbe arrayedon a unidimensionalindividualism-collectivismdimension.Thesepredictionsalso departfrom thoseof the structuralistargumentthat countriesvaiychiefly in termsof thepureform of socialnetworks.

North American Work Relationships:Market Orientation

A feature ofNorth American culture notedby manysocialobserversandcrit-ics is the tendencyfor peopleto enteror exit socialties accordingto themarketstandardof whetherit profits their individual achievementgoals. Thepopularconceptionof “networking” captureshow this normplays out in thecontextofwork organizations; one managesone’sportfolio ofrelations toward the goal ofupwardmobility, seekinginstrumentalrelationsand necessarilymaking roomfor them by discarding old onesthat have outlived their instrumental value(Baker, 1993). This view of networks as serving achievementgoals is madeexplicit in a leadingnetwork theory of performancein organizations(Burt,1992).From deToqueville(1848/1945)onward,observersofAmericanindivid-ualismhavedescribedtherelativeeasewith which instrumentalwork relationscanbe establishedwithout the prior basisof friendship or family connection.Yet, individuals also are relatively willing to break ties to achievesuccess.Fromthetime oftheexpandingfrontieronward,theparagonofthesuccessfulAmeri-canhasbeenthepersonwho leavesthe groupor disruptsthe socialorder,andAmericansremainrelativelywilling to reshapetheirsocialnetworksto pursuetheir professionalgoals (Bellah et al., 1985). The valuesthat support theseactionshavebeenlinked to theindividualist Anglo-Americaneconomic,legal,andphilosophicaltraditions(Lukes, 1973).

Now, let us considerwhatpredictionsfollow from the proposalthat NorthAmericanemployeestendto conducttheir interpersonalrelationsaccordingto amarketnorm. Whichdimensionsofsocialvaluesandrelationswill be theoneson which NorthAmericanemployeesstandout relativeto thosein othercoun-tries?With regardto socialvalues,wecanpredict that economicindividualism(i.e., the belief in achievementthrough social competition)shouldbe highlyendorsedby adherentsofamarketorientation.With regardto socialrelations,wecanpredictthat thedistinguishingfeaturesof amarketsystemwouldberelationslow in multiplexity (becauseinstrumentalexchangedoesnot requirea baseofsocioemotionalconnection)andlow in duration(becausepeoplearefreeto exitrelations).

66 MICRO-ORGANIZATIONALAPPROACHES

ChineseWork Relationships:Familial Orientation

Thenormsofsocialinteractionin Chineseorganizationshavebeendescribedasfamilial collectivism(Bond& Hwang,1986).As in a family, employeesmakesacrificesfor thegroup.Interestingly,however,this Confucianconcernfor har-monywithin thegroupdoesnot precludeplacinga strongvalueon achievementin competitivearenas(Hsu, 1953). Rather, it allows for achievementin partthroughrelationships,suchasfilial relationsto powerfulmembersoftheorgani-zation(Ho, 1976, 1998),andthroughtheconnections(orguanxi)that resultfromtheseties (King, 1991).The structuralmanifestationoffilial valuesin thework-place is suggestedby researchon Japaneseorganizationsthat have developedundersimilar Confucianinfluence.Comparedto Americanemployeeswhopre-fer same-statusfriends, Japanesemanagersseekhigherstatusfriends (Nakane,1970; Nakao 1987).Studiesof Chineseorganizationshavenotedthe carewithwhich employeescultivaterelationsto thosein power(Redding& Wong, 1986).

The predictionsthat follow from this Chinesefamilial orientationcan bestatedin contrastto thoseof the American marketorientation.With regardtosocialvalues,Chineseemployeesmayendorseachievementthroughcompetitionin general;however,Chineseemployeesshouldnotcondonecompetitionwithinthe in-group.Rather,Chineseshouldendorsethesharingofresourceswithin thein-group (i.e., economiccollectivism). Finally, attitudestoward superordinatesshouldbecharacterizedby a filial loyalty anddeference.With regardto proper-tiesofsocial relations,thereshouldbeahigherlevelofmultiplexity anddurationof relationsbecauseachievementis soughtthrough a stablesocial structure.Also, filial patternsof interactionshouldbeevidentin closesocioemotionaltiesto morehighly rankedcoworkers.Thesedifferencesnotwithstanding,it is impor-tantto notethattheAmericanandChineseorientationshaveasmuchin commonastheyhavein contrast.Both systemsinvolve a concernwith achievementandcondoneinstrumentaluseofrelations.

GermanWork Relationships:Legal-BureaucraticOrientation

ManyobserversoftheGermanculturalsettinghaveobservedthat workplacerelationshipsappearto be boundedby formal categoriesand rules (Bomeman,1992; Hall, 1990).A secondobservationis that affective expressionin work-placerelationsis somewhatmuted,or at leastslow to develop,comparedto thatin othercountries(Lewin, 1948).Bothof thesefeaturesmight reflect that inter-actionsareguidedby an impersonalbureaucraticstandard.In asystembasedonsuch a norm, an ego decideshow to act toward a given alter accordingto thealter’s formal status,not accordingto what the alter, as a specific individual,


offerstheego (as in themarketsystem)or accordingto thefilial relationestab-lishedbetweentheegoandthealter(asin thefamilial system).

Given this characterizationof Germanworkplacerelationsasguidedby alegal-bureaucraticnorm,onwhatdimensionsofsocialvaluesorrelationsshouldGermanemployeesstandout?First, comparedto theprevioustwo systems,thereis lessreasonto expectendorsementofthe valueof economicindividualism orachievementthroughcompetition.Second,acentralpredictionofthis systemis alow level of expressivecollectivismor empatheticinvolvementwith the workgroup.Relatedto this secondvalueprediction,therearestraightforwardpredic-tions aboutpropertiesof workplacerelations.More specifically, informal rela-tionsshouldfollow thelinesofformallyrequiredinteraction,asopposedto inter-action that spansformal categoriesand ranksand that might result in othersystemsfor reasonsofopportunism,filial loyalty, or loyalty to old friends.Infor-mal relationsalso shouldbe characterizedby lower affective intensitythan inothersystems.

Spanish Work Relationships:Affiliative Orientation

Anthropologistswho havedescribedwork relationsin thecontextof Spanishculturehavepointedto strongnormsofwarm sociabilitytowardthepeerfriend-ship group(Murphy, 1983).Similarly, aswehavereviewed,psychologicalstud-ies ofsocialbehaviorhaveobservedan expectationofwarm andfriendlybehav-ior amongcoworkers(Triandisetal., 1984).It is instructiveto considerhow theimplicationsof an affiliative norm differ from those of the legal-bureaucraticnorm. In both systems,actionstowardothersdependsonwho theother is ratherthan what the other does,that is, on characteristicsratherthanperformances.Yet, it is a differenttype of characteristicthatmatters.TheGermannormturnson characteristicsof the alter that aredefinedby the formalorganization(e.g.,rank),whereasthe Spanishnorm turnson characteristicsthat aredefinedin rela-tion to theego(e.g.,centralityin onesgroupoffriends).Theconsequenceis thatwhereasfor Germanemployeesit is appropriateto adjusttheir friendshipnet-works astheirjob titles change,for Spanishemployeesit is not normativetochangetheirpatternsofinteractionaftertransfersorpromotions.

Now, let us review the distinctive featuresof workplacesocialvaluesandrelationsin the Spanishsystemof interactionbasedon anaffiliative norm. Withregardto subjectivevalues,Spanishemployeesshould stand out as high inexpressivecollectivism(affective involvementwith coworkers)aswell ashighin economiccollectivism(sharingresourceswith coworkers).With regardto fea-turesof socialrelations,theSpanishworkplaceshouldbecharacterizedby rela-tionshigh in affectiveintensityandin longevity.Thedurationofrelationsshouldbe longerthanin othersystemsbecausetherelationsarenot delimitedby instru-

68 MICRO-ORGANIZATION! L. ~PPROACI-IES

mentalpurposes(asin theNorth Americanand Chinesesystems)or by formaljob categories(asin theGermansystem).

A CONTROLLEDCROSS-NATIONALCOMPARISON

ThepredictionsthatwehavederivedaboutNorthAmerican,C linese,German,and Spanishrelations were testedin a surveyof employeesin an MNC, theCitibankconsumerbank.Twopoliciesofthis firm makeit particularlyappropri-ateasa siteto testourhypotheses.First, employeesarehiredfrom thelocality inwhich the bank is situated.This policy of “embeddedness”allows for localnorms of interaction to arisewithin the workplaceeventhough the bank is aNorthAmericancorporation.Second,thereis a “global” strategyofstandardiz-ing organizationalstructureandproductsacrosscountries.Theformalorganiza-tion chartandjob categories,thephysicallayout ofthe bankbranches,and thefinancialservicesprovidedandsold aresimilar acrosstheretail operationsin thedifferent countriesthat we san~pled.This policy of maximizing cross-countryvariationin thehumancompositionoftheorganizationwhile minimizingcross-countryvariation in theformal structureand practicalcontentofwork virtuallycreatesa naturalexperimentfor investigatingthe effectsof cultural norms onworkplaceinteraction.8

To testourhypothesesaboutthe fourcountries,wegathereddatafrom retailbank employeesin North America,Hong Kong, Germany,and Spain. Withineachcountry,we selectedseveralareasfor greatestequivalenceon ecologicalvariablessuchascity size anddensity. Within a selectedregion,all employeesabovethe level of part-timetellers and below the level of areadirectorsweresampled.Employeesanswereda surveythat presentedquestionstappingsocialvaluesandsocialrelations.Thesurveywaspresentedin the official language(s)of eachcountry.Very high responserateswere obtained.For full details aboutsampling,procedures,and materials,seeMorris et al. (1997).In what follows,wereview themeasuresofvaluesandrelationsusedto testourdescriptivepro-posalsaboutdifferencesin thenormativesystemsthat arisein work settingsindifferent countries.

Measuresof socialvaluesdrewprimarily on the itemsdevelopedwith univer-sity studentsamplesby Triandisand Gelfand(1998)to captureeconomicandexpressiveaspectsof individualism and collectivism,althoughscaleitemswerewinnowedandalteredto fit ourresearchsetting.Otheritemswereadaptedfromotherpreviousinstrumentsor weregeneratedbasedon suggestionsby culturalinformantsto captureotherdimensionsofsocialvaluesrelevantto ourhypothe-sessuchasfilial loyalty anddeferenceto superordinates.For eachdimensionof

RelatiOnalConstructs 69

interest, a scalecouldbe formed from severalitems that clusteredtogetherineachofthe fourcountries.

Measuresof relational featureswere basedon thosein previousstudiesof• employeerelations (Podolny & Baron, 1997); however, the measureswererefined and extendedbased on input from Chinese,Spanish,and Germanresearchersto avoiddisproportionateemphasison issuessalient in the UnitedStates.Thesurveyaskedemployeesto list (usingcodenames)individualswithwhom they interactduring thework day. Specifically, respondentsnamedtheotherswith whom they sharepersonalfriendships,the others with whom theyexchangetaskadvice,andthe otherson whose“buy in” or powertheydepend.Ourfocusis on thefirst typeofrelationgiventhat it is mostopento volition andleastdeterminedby the structureof the organization.Respondentswere askedquestionsabouttheir relationsto the alters,aboutattributesof the alters,andaboutthe relationsperceivedamongthe alters.Foreachrespondent,we calcu-latedpurely formal relationalfeaturessuchasthe sizeanddensityofthe friend-ship network.We alsocalculatedtheextentto which friendshiptiesoverlappedthe adviceexchangenetwork, the extent to which they overlappedthe power-dependencenetwork,andtheextentto which theyfollowedtheverticalandhori-zontal lines of the formal organization.Finally, we calculatedthe averageaffectivity andlongevityofeachrespondent’sworkplacefriendships.

EVIDENCE FOR DISTINCTSOCIOCULTURAL SYSTEMS

In this chapter,ourgoal is not to comprehensivelyreviewtheresultsofourstudybut rather to highlight selectedresults relevantto our currentargumentthatincorporatingrelationalconceptsenrichesresearchon cultureandorganizationalbehavior.A first point is simply that the goal of incorporating relationalcontructsbringsus to a fruitful level of analysis—normativesystemsthat areconstitutedby a combinationof subjectivevaluesand relationpatterns.As wehaveargued,this level of analysisis particularlyhelpful in that it distinguishesdifferenttypesofculturalcollectivismthathavebeenconflatedin simplerframe-works. We illustratethis by steppingthroughthe subjectiveandrelationalvari-ablesin ourstudyanddescribinghowthedatafit predictionsofouraccountandofpreviousaccountsofculturalcollectivism.

SubjectiveValue Measures

Let us beginby looking for evidencefor our distinctionsin employees’sub-jectivevalues.Figure3.1 showstheextentto which employeesin thefour coun-

70 MICRO-ORGANIZATIONAL s’YPROACHES

tries endorsedeconomicindividualismor abelief in achievementthroughsocialcompetition (e.g., measuredby level of agreementwith statementssuch as“Competitionis the law of nature” and“Winning is everything”).As expected,employeesin the North American settingadheredto this valuemore thandidemployeesin theotherthreesettings.The differencewasnot reliablein relationto Chineseemployees;this is not unexpectedgiven that Chinesenorms favorcompetitionin generalbut not within the in-group.Figure3.1 alsoshowsa spe-cific within-group competitionscaleconstructedfrom items referring specifi-cally to competitionwithin thecircle ofone’scoworkers(e.g.,“Without compe-tition betweencolleagues,it is not possibleto havea goodcompany”).Here,asexpected,employeesin the North Americansetting were significantly higherthanthosein theChinesesettingaswell asthosein theGermanandSpanishset-tings. Overall, the resultsin Figure 3.1 establishthe familiar point that Ameri-cansareextremein endorsingcompetition,andtheresultsalsoclarify that sometypesofcompetitionarehighly endorsedin Chinesecultureaswell.

Figure 3.2 reportsseveralscalestappingdifferent formsof collectivism orsocial solidarity. The first scaletappedsolidarity and sentimenttoward one’ssupervisorin theform offihial loyalty(e.g., “A goodmanageris morelike apar-entthanafriend,” “If my supervisorhadarivalry with anothermanager,it wouldbe inappropriatefor me to becomefriends with that manager”).As expected,Chineseemployeesendorsedthis valuedimensionfar morethandid employeesin theotherthreesettings.Here,ouranalysisdrawsattentionto asubjectivevaluein Chinesecollectivism that differs markedly from German and Spanishcollectivism.

Also shownin Figure3.2 aretwo dimensionsthat concernsolidarity with thepeergroup of coworkers.Theseare two scalesbasedon dimensionsdistin-guishedby TriandisandGelfand(1998),which we label expressiveco/leejivism(e.g., “The well-being of my coworkersis very importantto me,” “I feelgoodwhenI cooperatewith others”)andeconomiccollectivism(e.g., “I like to sharethings and shareadvicewith my ‘neighbors’ at work,” “A groupof coworkersshouldstick together,evenif sacrificesarerequired”). Expressivecollectivismwaspredictedto be relativelyhigh in Spain,andour findings supportthis pre-diction. High levelsof economiccollectivismwere expectedin all threecollec-tivist settingsfor different reasons.Economiccollectivismwaspredictedin theChinesesettingaspart ofthe syndromeof group-levelachievementstriving, inthe Germansetting aspartoftheobligation felt to thefellow membersof one’sformal unit, and in the Spanishsettingasabyproductof one’s intenseaffectiveconnectionsto coworkers.Thesepredictionsreceivedmixed support.As canbeseen in Figure 3.2, Chineseand Spanishemployeeswere significantly higherthanAmericanson economiccollectivism;however,Germanswerelowestofall,an unexpectedresult.

71RelatiOnalConstructs

‘a’ 3.5

0.4-a-4—. 3.4

0)

S0

SU’ 3.2

-d~ 3.1

•0

.3.00C)

4.0

C0.4-,

3j5

E000. 3.4

0

(9I 3.2C

-c4-’

~3.0

N. Amer Chinese

Cultural setting

Cultural setting—a~rzrcc~flsr~ r~I&r&WrMS~r s~rr~r&~~

German Spanish

N. Amer Chinese uerman Spanish

FIGURE 3.1. CountryMeanson Scalesfor EconomicIndividualism(achievementthroughcompetition)andWithin-GroupCompetition

72 MI ~RO~ORGANZN IIONAL A7W AC JES

32

30

28

CT’>~ 260

-j

~ 2.4

U-

2.2

20

1.8

N. Amer uninese uerman SpanishV

Cultural setting4

5.0

4.9

4.8

> 47I

C)a)4.6

oC-)0) 45

44

N. Amer CflIneSe uerman ~pan~n

Cultural setting

4.6

45

U)

> 44

C)C)= 430

§4 C).2 42

S0

410C)

Lt~40

39

N. Amer ~nincse uernian Spanish

Cultural setting

FIGURE12. Country Meanson Scalesfor Filial Loyalty to Superordinates,ExpressiveCollectivism(affective involvementwith the in-group), nd EconomicCollectivism(beliefin sharingwith/sacrificing for the in~gro ip)

RelatiOnalConstructs 73

Let ussummarizehow well ourproposalstandsafterreviewingtheevidencefrom subjectivevaluescales.Althoughnot fitting thedataperfectly,ourpredic-tions fit betterthanthose from the unidimensionalindividualism-collectivismframework(i.e.,NorthAmericancultureon theindividualistpole,GermanyandSpainnearthe center,andChineseculture on the collectivistpole; Hofstede’s[19801construct).The clearestindication is that, looking acrossthe differentvaluemeasures,AmericanandChineseemployeesin ourdataare fairly similarto eachother.This comportswith theParsonsiannotionthatAmericanandChi-neseculturesaresimilar in importantways; they arenot the antipodeanoppo-sites that sometimeshave beendepictedin researchwithin the individualism-collectivismrubric. Wherethe evidencefrom subjectivevalue variablesfallsshort is in supportingourpredictionsabouttheGermanculturalsetting.We pre-dictedthat Germanemployeeswouldbecharacterizedby economiccollectivism(sharingresourceswith the in-group)but not expressivecollectivism (affectiveinvolvement).Counterto ourpredictions,Germanemployeeswerevery low inendorsingsacrificefor the in-group.It mightbe thattheirformalrule orientationmeansthat theyshareresourceswith the in-group only to the extent that theirjobsrequireand,hence,do notsubjectivelyconstruethisassharingbut merelyasproperrule following. On theotherhand,it mightbe thatourcharacterizationofinterpersonalnorms in the Germansetting is out of dateor inapplicableto thebankingindustry. If wehadonly subjectivevaluemeasures,thenwewould notknow. Fortunately, our characterizationalso can be testedwith relationalmeasures.

Relational Measures

In reviewingtheprofiles for eachcountryon relationalmeasures,we beginwith the pure formal relational featuresandthenmove to featuresthat involveform and content.As Figure 3.3 illustrates,therewere no reliabledifferencesacrossthefour settingsin theaveragesizeanddensityoffriendshipnetworks.Inotherwords,thereis no supportfor the notionthat thedifferencesbetweencul-turalsettingscanbe reducedto pure socialstructure,that is, to thegeometryofrelationsthat surroundindividualsin differentcultural settings.

Moving on to the relational featuresthat involve form andcontent,we firstassessedmeasuresofthe multiplexity or theoverlapof differing relationalcon-tents.In particular,we assessedtheproportionof theadvice-exchangenetworkwith whom eachrespondentalso sharedfriendships.Second,we assessedtheproportionofthepower-dependencenetworkwith whom therespondentsharedfriendships.As canbe seenin Figure 3.4, thesetwo measuresshow the samebasicprofile acrossthe four settings.As expected,NorthAmericanemployeeshad lessmultiplex networksthandid employeesin theotherthreecountries.

74 M ~RO-ORGANIZAT1ONALA ROt C IES

30

0)N 2.0

4

1.0

N. Amer Chinese German Spanish

Cultural setting

.3

4A

U’~ .20)

.1


Culturalsetting

RGURE13. Country Meanson PurelyFormalRelationalMeasuresNOTE: Size of network is the numberof people listed as personalfriends at work.Density is the meanproportionof other friends to whom friends areconnected.Theseweretheonly variableson which no countrydifferenceswere reliable.


N. Amer Chinese

N. Amer Chinese

Cultural setting

.5

-5

.4.

-4.

.4.

U’

4)

H4)

C)

4)C)

-o

.3

0.3

.4

.4

German Spanish

Cultural setting

.3

U’4)

H .3-4)C.)4)-o ~C4)

4)

0

‘~ 2‘-4

4)

0.2German Spanish

IFIGURE 3.4. CountryMeanson MeasuresofMultiplexityNOTE: Thesescoresaretheproportionsof othersin a respondent’sadvice-exchangetiesand power-dependencetiesfor which theothersalsowere listedasfriends.

76 MICRO-ORGANIZATIONAL APPi~(;ACiIEs

Whereastheaforementionedmultiplexity measurescapturethe alignmentoffriendshipswith othertypesor contentsof informal relations,we also derivedpredictionsaboutthe extentto which friendshipsfollow the linesof the formalstructureoftheorganization.A first measure,theaveragerankoffriendsrelativeto therespondent(—1 = lower, 0 same, 1 higher) wasexpectedto revealamoreupwardorientationamongChineseemployees(becausetheyseekfilial tieswith powerful coworkers).As can be seenin Figure 3.5, Chineserespondentsdid in fact havethe most upwardorientation,although they were not signifi-cantly higherthanSpanishemployees.The nextmeasurewas theproportion ofsame-unitcoworkersbefriended,which waspredictedto behighestamongGer-man employees(becausetheir friendshipsare boundedby formal categories).Germanemployeesdid tendto countasfriendsahigherproportionoftheirworkunits thandid NorthAmericanandChineseemployeesbut not Spanishemploy-ees(middlepanelofFigure3.5).A final measureoftheextentto which informalfriendshipfollowedthe linesofformally prescribedinteractionwasameasureofhow closely a respondent’sinteractionfrequencycorrespondedwith the offi-cially requiredlevel. As predictedfrom thenotionthat Germanfriendshiprela-tions follow the lines of the formal organization,we observedthat Germanemployeesshowedthe leastinteractionbeyondthat officially required(bottompanelof Figure 3.5). Onthis last measure,the sharpcontrastbetweenGermanand Spanishemployeesis noteworthy;theaffiliative norm in the Spanishsettingresults in nearly twice the level of nonrequiredinteractionswith friends thandoesthe legal-bureaucraticnorm in the Chinesesetting. Overall, the proposalthat interpersonalrelationsfollow thelines of formal categoriesandrules findsconsistentsupportin thethreemeasures.Comparedto employeesin othercoun-tries, German employees are oriented toward same-rankcoworkers, hadbefriendeda high proportionof theotherswithin theirunits,andinteractedwiththoseothersaboutasmuchaswasofficially requiredby theirjobs.

Finally, we measuredthe averageaffectiveclosenessand duration of theworkplacefriendshipsin the four settings.For thesefeatures,we hadpredictedthat Spanishemployeeswould standout becauseof an affiliative norm thatmakes friendship an end in itself. As expected,we observedthat Spanishemployeeshad friendshipsthat were significantly higher in affectiveclosenessandin durationthanwerefriendshipsofemployeesin theotherthreeculturalset-tings (Figure 3.6). Importantly, thesetwo variablesdid not haveidentical pro-files. Consistentwith Lewin’s (1948)arguments,relativeto Germanemployees,North Americans had friendshipsthat were affectively closer but not moreenduring.

Overall, results from our measuresof relational constructsgreatly help inunderstandingthe differencesamongthe norms of workplace interaction inthesefour cultural settings.Interestingly,the purely formal features—-sizeanddensity—didnot vary. It might bethat the intuitively appealingthesisthatcollec-


4)

Lz~‘-40.34

4)

4)3..44

)

U

.3

.2

.4

0.0

.40)

00)

‘4-0) .3m

‘4-~ .20

0

8~o~ .1 ii,.. ~ chh~cse German

cu~Isetting

•U 1.4

I.)

-o00

l.a

00

C)

42 .6

FIGURE 3.5. CountryMeanson Measuresof theAlignmentof FriendshipTiesWiththeFormal StructureNOTE: The measuredepictedin thetop panelaveragesa rating of the relativerank ofalters(this is zeroif theaveragefriend is at therespondent’sown rank andis positive ifthe averagefriend is at ahigherrank). The measuredepictedin the middle panelis theproportionofsame-unitcoworkerswho arelistedasfriends.Themeasuredepictedin thebottom panelis derivedfrom thedifferencebetweena ratingof level ofinteractionwithanalter anda ratingof the level officially requiredwith that alter.

Spanish

culuami setting

Spanish

P4. amer unmcsv ~ Spanish

culuami setting

78 MICRO-ORGANIZATIONAL ~JPROACIIES

3,7

cj~

cj~

0

c-)

0

3.6 -

3.5

3.4 -

3.3

3.2

Cultural setting

8.5

8.0

‘-4

U

0

7.5

7.0

6.5

6.0

5.5.

5.0 ______


FIGURE 3.6. CountryMeanson Measuresof Affective Closeness(I = only ascloseasneededto work together, 5 — oneof my closestpersonalfriends)and Duration(in years)


‘4

6~

‘46,6

66,

44

6,6

44‘66

44

44.66*

4444‘V

6*

44

“6

44

*6

44


tivism is associatedwith density(Fischer& Shavit, 1995) is true offriendshipsoutsideoftheworkplace,butthebasicquantityandshapeofinformal interactionin this work setting is fairly constrainedby the structureof the bankbranches.Whereculturalnormsdid shapeinteractionwasin howfriendshipswereorientedin relation to other typesof informal ties (i.e., multiplexity), how friendshipswere oriented in relation to the formal structure,andthe affective intensityofwork friendships.Whatmakestherelationalevidencefor distinctculturalnormsimpressiveis thatthesedifferentrelationpatternshadto beestablishedagainstacommonsetofconstraints—thephysicallayoutofbranches,theformalstructureofcommandandcontrol,andthejob design.In asense,theserelationalmeasures

demonstratewhatGiddens(1984) labeledas“structuration”—thecreationandeventualcrystallizationof a structuredsystem of interactionby individualscarryingnormsandassociatedhabitsof interaction.

HOW VALUES AND RELATIONSINFLUENCE BEHAVIOR

Thusfar, wehaveprovidedevidencefor ourdescriptiveproposalsoffour socio-culturalsystems.However,wealsowantto briefly illustratehowtheseproposalsgeneratepredictionsaboutoutcomesthatfollow, that is, aboutindividualbehav-iors that are importantto the functioningof organizations.We limit our discus-sionto one outcomevariable,butweusethis to illustrateourview ofthedistinctrolesthat subjectivevaluesandrelationshipstructuresplay in influencingindi-vidual behavior.

The outcomethat we focuson in this discussionis the obligation that anemployeefeelsto helpothersatwork. Theorganizationalliteratureon obligationcontainsa basicdivide betweentheoristswho maintainthat employeesarepri-marily obligatedin verticalrelationsto thosein powerandthosewhoemphasizean obligation in horizontalrelationsto peers.Theemphasison verticalrelationsis evidentin Weber’s(1947)descriptionof bureaucraticorganizationsin post-WorldWarI Germany,whereobligationtotheleaderwasbasedon theperceivedlegitimacyofthehierarchicalsystem.Thisview concerningthedirectionofobli-gation persistsin leadingAmericantheoriesof administration,although moreemphasisis placedboth on theinstrumentalrationalityof obligationto those inpower(Simon, 1945;Thompson,1967)andon a socialexchangeprocess(Blau,1955).A contrarytraditionhasarguedthat in manyorganizationalcontexts,theprimaryobligationslie in horizontalrelationsto peersratherthanin verticalrela-tionsto authorities.RoethlisbergerandDickson(1939)observedthat workersinabankwiring room felt strongbondsofobligationto cohesivepeergroups.Theinsightthatthesocialrewardsofhorizontalrelationsprovidean importantsource

go MICRO~ORGANZAT1ONA~ (P1 ROACHES

of work motivation andjob satisfactionsparkedthe humanrelationsschool ofmanagementtheory(e.g., Dalton, 1959).

Despite the large volume of researchconcerningobligation 7n vertical andhorizontalrelations,surprisingly little workhassoughtto establishtheboundaryconditionsdeterminingwhetheran employee’sprimary obligationwill occur invertical relationsto those who providepowerand opportunityor in horizontalrelationsto coworkerswho provide friendshipandsocioemotionalsupport.Themajortheoreticalposition on this questionis Merton’s (1969) argumentthat afocuson horizontalrelationsdevelopswhenan employee’soriginal focuson ver-tical relationshasbeendiscouragedby frustratedmobility aspirations(Kanter,1977).We haveargued(Morris etal., 1997, 1 998b)thatsucha descriptionmightbe accuratefor adynamicthat occurswithin theNorthAmericannormativecon-text but not in all cultural contexts.In otherwords, whereasprevioustheoristshavetakenit asgiventhatthe defaultdirectionof employeeobligation is towardthosein powerratherthantowardpeers,wesuggestthat this dependson thecul-turally boundsystemofnorms.

Theverticalorientationis consistentwith themarketnorm in theNorthAmer-ican setting,aswell aswith the familial norm in the Chinesesetting,becausesocialenergiesaredirectedtowardtheend ofachievement.Moreover,theGer-man legal-bureaucraticsystem,with its emphasison formal status,also shouldleadto a verticalorientationof obligation.A differentprediction,however,fol-lows from the Spanishaffiliative norm. Ethnographiesof Spanishworkplacesstressthat honorobliges fulfilling obligationsto the friendshipclique but notnecessarilyrequestsfrom thosein power(Gilmore, 1982, 1987).To illustrate, arecentethnographyof a blue-collarwork groupin Spaindescribeda systeminwhich it was normativeto treatmanagerialrequestswith suspicionyet deplor-ableeitherto withholdafavorrequestedfrom within thepeergroupor to complycheerlessly(Murphy, 1983).Thesignificanceofhonor,loyalty, andfriendshipindeterminingobligation is echoedin descriptionsof white-collarcorporateset-tings (Alvarez& Cantos,1994)2

In the four-countrystudy that wehavedescribed,a measureof obligationtoeach alter was taken in a vignettetailored to the retail bankingcontext thatassessedthe likelihood that the respondentwould volunteerto help that alter.This variableallowedusto calculatethe averagelevel ofobligationto thepowernetwork andthe averagelevel to thefriendshipnetwork.Becauseourconcerniscomparative,we focusedon thedifferentialobligationto friends. Consistentwithexpectations,Spanishemployeeshad a muchgreaterdifferential obligation tofriendsthan did employeesin the other threesettings(Morris et al., 1997). Tocheckourunderstandingof this patternof outcomes,wetestedfor evidenceofmediatingandmoderatingvariables.

An influential strategyof analysisin the subjectivist tradition of culturalresearchis to test whe~hermeasuresof cultural values mediate (or come

ational Constructs 81

ween)nationality,on theonehand,andanoutcomebehavior,on theother.Inworkof LeungandBond (1984)on in-groupfavoritismorthework ofEarley

) on social loafing, an importantfeatureof the resultswerecorrelationsteencollectivismscoresand the respectiveoutcomevariables.’0Althoughositionallyheldgeneralvaluesandattitudesmaynot be theproximal causes

behavior,they should havea tighter associationthanmoredistil variablescountry)that havearemotecausalconnectionto mostoutcomebehaviors

interest.We conducteda mediationanalysisin our dataand foundthat thentry effecton obligationorientationis mediatedby subjectivevalues.Thatis,enourmeasuresof individualismandcollectivismareenteredinto a regres-

model simultaneouslywith thecountrydummyvariables,theyremainsig-cantpredictors,whereasthe dummyvariable for Spainno longerpredicts

jerentialobligationto the friendshipnetwork.A similar finding hasemergedanotherrecent study of country differences in obligation (Cialdini,

)sinska,Barrett,Butner,& Gornik-Durose,1998);hence,thedependentvan-of obligationcanbeaddedto the list of behavioraltendenciesthat are in-

by individualist versuscollectivistorientations.Although the subjectivist tradition hassucceededin identifying values as~diatorsofcountrydifferences,thesubjectivistapproachhasnotbeenstronginintifying moderatingvariables.Eventhemostwidely knownstudiesofsocial

organizationalbehavioracrossculturesthat take into accountmoderatingAors (e.g.,Earley, 1989;Leung& Bond, 1984)typically havebeenconductedlaboratoryparadigmsthat bearonly anabstractresemblanceto theactualcon-

of organizations.Within theselaboratoryparadigms,slight variations inceduresoftenhavemeantafailure to replicatepatternsofculturaldifference

~eung,1997). Oneway in which to analyzethis problemis to assertthat thevist traditionhasnot hadvery good tools for analyzingthemoderating

variablesthat really matter—thesocialcontextsthat conditiontheactivationofLbjective rules,ultimatelyresulting in culturally varyingpatternsof action in

organizations.Thepotentialfor fine-grainedmeasurementofthesocialcontextsthatmoder-

ateindividual actionmightbewhererelationalvariablesmaketheirlargestcon-tribution.Relationalvariablesoffer away in whichtomeasuretherelevantsocialcontextsthat surroundaparticularactiondecision.Differencesbetweennorma-tive systemsshouldbesharperwhentherelevantmoderatingconditionsareiden-tified. ConsiderEmployeeA who is askedfor a favor by a friend and whosefriendsareinterlinkedin a denseclique.Now, considerEmployeeB who is alsoaskedfor afavorby a friend butwhosefriendsdo notknow eachother.How willthis differencein densityof theirnetworksaffect the intensity obligation felt?Doesthis, in turn, varyas a functionofthesurroundingnormativesystem?

Following from our analysisof why Spanishemployeesare differentiallyobligatedto the friendshipnetwork,we reasonedthat in the Spanishcontext,

82 MICRO-ORGANIZATIONAL APPROACHES

Ci2

0

•Low0H

Density4-.

z0

Density

CulturalSetting

FIGURE 3.7. DifferentialObligationto FriendshipNetworkasaFunctionofCultural Settingand theDensityLevel oftheRespondent’sNetworkNOTE: This figure illustratesaninteractioneffectin whichhigh densityofthefriendshipnetworkcreateda greaterdifferential obligationto thefriendshipnetwork in theSpanishsettingbutnot in theothercultural settings.

where doing a favor for a friend is a matterof honor, densenetworksshouldincreasethe pressureofobligationto friends. By contrast,if Americanemploy-eesareobligatedto friendsprimarily asa functionofhowthe friendshelp theirown advancement,thenAmericansconverselymayfeel lessobligatedin adensenetwork than in a network that is less dense(in Burt’s [19921terms,a non-redundantnetwork).As canbeseenin Figure3.7, this predictedinteractionpat-tern was observed.High friendship density makesSpanishemployeesmoreorientedto the friendshipnetwork, whereasAmericansmovenonsignificantlyin theoppositedirection.Whenthis interactioneffect is enteredinto aregressionmodel with the main effect of country, the interactionremainssignificantbutthemain effectdoesnot. Hence,the interactionunderliesthemain effect. Thisexampleillustratesthatmeasuresofrelationalstructurepickup situationalvari-ablesthatmakeagreatdealofdifferencein behavioralresponses.Understandingthesesituationsallows a way in which to bettercontextualizefindings aboutculturaldifferences.

N. American Chinese German Spanish


CONCLUSION- / WO~aX ,,W ~ SflW*~ 4att$052x0%a,w~arn A~0XS 4~1Ce

This chapterhaspresentedselectedfindings from a researchprograminvesti-gatingthe proposalthat cultural differencesin organizationalbehaviorcanbeunderstoodin termsof normativesystemsof interpersonalinteraction,that is,systemsconstitutedby subjectivevaluesandby relationalstructures.Althoughthis level of analysisis lessparsimoniousthana purely subjectivistor purelystructuralistapproach,it offersamorecomprehensiveunderstandingof howcul-ture influencesindividual behavior.Whereasthe subjectivistand structuralistapproacheshavebeenregardedasrivals, we arguethat neithersufficesalone.Moreover,thetwo approachescomplementeachother.Subjectivemeasurespickup theindividual traits andcognitivestructuresthat serveasmechanismsin theinfluencesof macro-levelvariablessuchascountry.Relationalmeasuresiden-tify micro-levelsocialcontextsthatconditiontheseinfluences(for asimilaruseofsubjectiveandrelationalvariablesin predictingfinancialrisk taking, seeHsee& Weber,1999; in predictingjusticeperceptions,seeMorris & Leung,2000).

A final benefit of this sort of cross-culturalstudy is that it revealsculturalassumptionsthat are implicit in formal academictheoriesof organizationalbehavior.We have arguedthat the assumptionof Mertonand othersaboutthedefaultdirectionof obligationshould beunderstoodas culturally bound. Moregenerally,theremight be manyassumptionsin organizationaltheory(OT) thatarevalid in thecultureswherethesetheorieswere developedbutnotelsewhere.ConsiderWeber’s observationsof the “modern” bureaucraticform in post-World War I Germanhospitals,where“all purely personal.. . emotionalele-ments”areeliminatedin the interactionsamongemployees(quotedin Gerth&Mills, 1958,pp. 215-216).Givenour findingsconcerningaffectivity, onemustwonderwhetherthis description,so pivotal in the intellectual history of OT,would have arisenif Weberhad madehis observationsin a Spanishhospitalratherthanin aGermanhospital.

NOTES

1. We usethe termstructuralismtheway in which it is usedby sociologists(e.g.,Mayhew),not theway in which it is usedby cognitiveanthropologists(e.g.,Levi-Strauss).That is, we referto theview that actionis causedby socialstructures,not theview thatthinking is shapedby cognitivestructures.

2. This is not to deny that thesesameindividualsadhereto differentiatedsubculturesand haveotherculturalalliancesand identitiesthatarefragmentedandcontextdependentratherthanunifiedandcontextgeneral(Martin, 1992).

3. Triandis and colleaguesreferredto thesedimensionsas vertical and horizontalaspects.This terminologywould beconfusingin the currentcontext, sowe substitutethemoreconcretedescriptorseconomicand expressive,respectively.

84 MICR&-ORGANIZAT ONALi ~ OACI-lEs

4. Our discussionfocuscson analysesofegocentricnetworksin which the uniquenetworksurroundingeachindividual is theunit ofanalysisbecausethis methodis mostlyamenableto cross-culturalcomparisons.Anotherapproachto network analysi’is to studyall relationsthatexistwithin agroupof people.One issuewith this methodis the bound-ary problem of choosing which group to study. Given that the in-group/out-groupboundarymight differ acrosscultures, this seemsto particularly complicateanalysesof collectivism. Another issue with this method is that anearperfect responserate isrequired.Someofthe methodsthatresearchersusewhentherearemissingdataprobablyaredifferentiallyvalid acrosscultures.For example,onetechniqueis to symmetrizerela-tions, that is, to infer that A hasaparticularrelationto B andthat 13 hasthesamerelationto A (Scott, 1991). Although the assumptionunderlyingthis techniquemight approachaccuracyin egalitariansocieties,it is highly problematicin hierarchicalsocietiesthatfosterasymmetricrelationssuchas the exchangeof filial loyalty for paternalprotection(1-Jo, 1976, 1998).

5. Muchofthis researchis rootedlessin networkanalysisthanin the-more--proccssua!.approachto relationshipstakenby researchersin communicationsandpsychology(for areview,seeDuck, West,& Acitelli, 1997).

6. Descriptiveshortcomingmight be excusedby the fact that Parsonswas primarilyattemptinga contributionto meta-theory;his writings consistchiefly of a priori argu-mentsconcerningdilemmasbetweenparticularsocialvaluesthat limit thepossiblecom-binationsofvaluesthatcouldfunction in homeostasisasa socialsystem.Yet, evenschol-arswho haveappreciatedhis descriptiveproposalsaboutparticularvaluesystemshavebalkedathis insistenceon thegrandersystemof systemsthatencompassesthesepropos-als, which posits inexorabledilemmasand trade-offsby the dozenthat would seemtoarbitrarily limit thepossibleformsof socialsystemsand,hence,thepossibledirectionsofsocialandcultural diversity.

7. Although Parsons’model often is portrayedas adimensionalanalysis,he arguedvery clearlythat thesedilemmasarenot to be interpretedasindepimdent’dime~.sioi~.Cer-tain patternsof answersto thesedilemmasariseas cultural forms, but not all combina-tions arepossible,anda culture’sanswerto onedilemmacannotbe understoodin isola-tion from its answerto the other dilemma. For example,interwovenwith patternsofanswersto thesefirst two dilemmasareanswersto the dilemmaof affectivity—whetherto expressandacton affectivereactionsto othersor to control one’saffectivereactions.Parsonsdiscussedthreeotheraxes, or what he called “pattern variables,” but he inte-gratedtheseotheraxesinto his discussionof the four-categorytypology that arisesfromthe intersectionofthesetwo dominantaxes.

8. The “experiment” provides a conservativetest of the hypotheses,of course,becauseculturaldifferencesaredilutedby theCitibankorganizationalculture.If apatternof predictedcultural differencescan be observedin thecontrastbetweenemployeesofCitibank I-long kongandCitibank Spain, thenwe would expectthe samedifferencestobe much sharperin the contrastbetweentwo compl4ely local firms.

9. Althoughweknowof no previouscomparativestudiesof networksandobligationsin theworkplace,resultsfrom studiesof networksandsocial supportalsoaresomewhatconsistentwith our argumentconcerningthe primay of obligationto the friendshipnet-work in Latin contexts.1-lollinger and JaIler (1990) found that in the Latin country intheir data(Italy), expectalknswith reg rd to friends’ obligationswerenot conditionalon

85J?elatiOflal Constructs

thekins’ proximity. This differedfromthepatternin othercountries,wheretheobligation• offriendswasinverselyrelatedto theproximity ofkin.

10. Thesecorrelationstendto beweak,but this is reasonableto expectgiventhat the• socialattitudesaremeasuredat sucha high levelof generality.The specificactionrules

thatdeterminea givendecisionabout,for example,allocationofresourcesoreffort aremuchnarrowerthancultural valuedimensions.

REFERENCES

Alvarez, J. L., & Cantos,C. M. (1994).From escapismto resentedconformity:Marketeconomiesandmodernorganizationsin Spanishliterature-~In B. Czarniawska-Joerges& P. G. de Monthoux(Eds.), Goodnovels,bettermanagement:Readingorganiza-tional realities.Chur, Switzerland:Harwood.

Baker,W. (1993).Networkingsmart:How to buildrelationshipsfor personalandorga-nizationalsuccess.New York: McGraw-Hill.

Bellah, R. N., Madsen,R., Sullivan, W. M., Swidler,A., & Tipton, S. M. (1985).Habitsoftheheart: IndividualismandcommitmentinAmericanlife. Berkeley:UniversityofCalifomiaPress.

Benedict,R. (1934).Continuitiesanddiscontinuitiesin culturalconditioning.Psychiatry,1, 161-167.

~.B1au,P. M. (1955).Thedynamicsofbureaucracy:A studyofinterpersonalrelationshipsin twogovernmentagencies.Chicago:Universityof ChicagoPress.

Bond,M. H., & Hwang,K. K. (1986).The socialpsychologyofChinesepeople.In M. H.Bond(Ed.), Thepsychologyofthe Chinesepeople(pp.2 13-266).HongKong: OxfordUniversityPress.

~Bomernan,J. (1992).Belongingin the two Berlins: Kin, state,nation. Cambridge,UK:CambridgeUniversity Press.

- Brint, 5. (1992).Hidden meanings:Cultural content and context in HarrisonWhite’sstructuralsociology.SociologicalTheory, 10, 194-208.

Burt, R. (1992).Structuralholes: Thesocialstructureofcompetition.Cambridge,MA:HarvardUniversity Press.

Cantril, H. (1965).Thepatternofhumanconcerns.NewBrunswick,NJ: RutgersUniver-sity Press.

Cialdini, R. B., Wosinska,W., Barrett,D. W., Butner,J., & Gornik-Durose,M. (1998).Compliancewitha requestin two cultures:Thedqjferential influenceofsocialproofandcommitment/consistencyon collectivistsand individualists. Unpublishedmanu-script, Arizona StateUniversity, Departmentof Psychology.

Dalton, M. (1959).Menwhomanage.New York: JohnWiley.deToqueville,A.(1945).Democracyin America.NewYork: VintageBooks.(Originally

publishedin 1848)Diaz-Guerrero,R. (1967).PsychologyoftheMexican.Austin: UniversityofTexasPress.Doi, L. T. (1962).Amae:A key conceptfor understandingJapanesepersonalitystructure.

In R. J. Smith & R. K. Beardsley(Eds.),Japaneseculture: Its developmentandchar-acteristics.Chicago:Aldine.

86 MICRO-ORGANIZATION; APPROAC ES

Dore,R. (1983). Goodwill andthespirit of marketcapitalism.British Journal ofSociof.

ogy, 34, 459-482.Duck, S., West, L., & Acitelli, L. (1997). Sewing the field: ‘he tapestryof relation-

shipsin life and research.In S. W. Duck (Ed.), handbookofpersonalrelationships(2nded.,pp. 1-23).Chichester,UK: Wiley.

Durkheim,E. (1933). The divisionoflaborin society.New York: Frc’~ Press.(Originallypublishedin 1893)

Durkheim,F. (1951).Suicide.New York: Free Press.(Originally publishedin 1897)Earley,P. C. (1989). Social loafing andcollectivism:A comparisonoithe UnitedStates

and the People’sRepublicofChina.AdministrativeScienceQuarterly, 34, 565-581.Earley,P. C. (1994).Selfor group?Cultural effectsof training on sdf-efficacyandper-

formance.AdministrativeScienceQuarterly, 39, 89-i 17.Emerson,R. M. (1962).Power-dependencerelations.AmericanSociologicalReview,27~

3 1-40.Emirbayer,M., & Goodwin, J. (1994). Networkanalysis,culture, and theproblem of

agency.AmericanJournal ofSociology,99, 1411-1454.Erez,M., & Earley, P. C. (1993).Culture, self-identity,and work New York: Oxford

University Press.Fischer,C. S., & Shavit,Y. (1995).National differencesin network density: Israel and

theUnitedStates.SocialNetworks,17, 129-145.Fiske, A. P. (1991).Structuresofsocial life: Thefour elementaryformsofhumanrela-

tions. New York: FreePress.Fiske, A. P., Kitayama,S., Markus,1-1. R., & Nisbett, R. F. (1998).Theculturalmatrix of

social psychology.In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbookof

socialpsychology(4th ed.).Mahwah,NJ: LawrenceErlbaum.Garfinkel,1-1. (1967).Studiesin ethnomethodology.EnglewoodCliffs, NJ: Prentice1-lall.Geertz.C. (1976). From the native’s point of view: On the natureof anthropological

understanding.In K. Basso& 1-1. Selby(Eds.),Meaningin anthropology(pp. 221-237).Albuquerque:University ofNew MexicoPress.

Gerth,H., & Mills, C. W. (1958). From Max Weber: Essaysin sociology.New York:Oxford University Press.

Giddens,A. (1984). Theconstitutionofsociety.Oxford, UK: Polity.Gilmore, D. D. (1982). Anthropologyof the Mediterraneanarea. Annual l?eviews in

Anthropology,11, 170-205.Gilmore,D. D. (1987).Honor, honesty,shame:Male statusin contemporaryAndalusia.

In D. D. Gilniore(Ed.), Honorandshameandtheunity ofthe Mediterranean.Wash-ington,DC: AmericanAnthropologicalAssociation.

Gluckman,M. (1967). Thejudicial processamongthe Barotse ofnorthern Rhodesia.Manchester,UK: ManchesterUniversity Press.

Gorer,0. (1943). Themesin Japaneseculture. New YorkAcademyofSciences,5, 106-124.

Gross,J. L., & Raynor, S. (1985).Measuringculture. New York: Columbia UniversityPress.

Gudykunst,W. B. ( 983). International communicationtheory. Beverly 1-lills, CA:Sagc.


Hall, E. T. (1990). Understandingcultural d(fferences.Yarmouth, ME: InterculturalPress.

Ho, D. Y. F. (1976).Ontheconceptofface.AmericanJournalofSociology,81, 867-884.Ho, D. Y. F. (1998).Interpersonalrelationshipsandrelationshipdominance:An analysis

basedon methodologicalrelationism.AsianJournalofSocialPsychology,1, 1-16.Ho, D. Y. F.,& Chiu, C. Y. (1994).Componentideasofindividualism,collectivism,and

social organization:An applicationin the study ofChineseculture.In M. Kim, H. C.• Triandis, C. Kagitcibasi, S. C. Choi, & G. Yoon(Eds.),hidividualismandcollectiv-

ism: Theory,method,andapplication.ThousandOaks,CA: Sage.Hofstede,G. (1980).Culture‘s consequences:Internationaldlfferencesin work-related

values.BeverlyHills, CA: Sage.Hofstede,0. (1991).Cultureandorganizations.NewburyPark,CA: Sage.Hollinger,F., & Hailer,M. (1990).Kinship andsocialnetworksin modemsocieties:A

cross-culturalcomparisonamongsevennations. EuropeanSociologicalReview,6,103-124.

Hsee,C., & Weber,E. U. (1999).Cross-nationaldifferencesin risk preferenceand laypredictions.Journal ofBehavioralDecisionMaking, 12, 165-179.

Hsu, F. L. K. (1953). Americansand Chinese:Two waysoflife. New York: Abelard-Schuman.

Hsu,F. L. K. (1985).Theselfin cross-culturalperspective.In A. J. Marsella,0. DeVos,& F. L. K. Hsu(Eds.),Cultureandself(pp. 24-55).New York: Tavistock.

Ibarra, H. (1997).Pavingan alternativeroute: Genderdifferencesin managerialnet-works.SocialPsychologyQuarterly, 60, 91-102.

Inkeles,A. (1960). Industrialman: The relationofstatusto experience,perception,andvalue.AmericanJournalofSociology,66, 1-31.

Inkeles,A. (1978).Nationaldifferencesin individualmodernity.ComparativeStudiesinSociology,1, 47-72.

Inkeles,A., & Levinson,D. (1954).Nationalcharacter:The studyofmodalpersonalityandsocial systems.In 0. Lindzey (Ed.), Handbookofsocialpsychology(pp. 975-1020).Reading,MA: Addison-Wesley.

Kanter,R. M. (1977).Menandwomenofthe corporation. New York: BasicBooks.Kapferer,B. (1969).Normsandthemanipulationof work rules. In J. C. Mitchell (Ed.),

Social networksin urban situations:Analysesofpersonalrelationshipsin CentralAfrican towns.Manchester,UK: University of ManchesterPress.

Kardiner, A. (1939). The individualandhis society. New York: Columbia UniversityPress.

Kashima,Y., Yamaguchi,S., Kim, U., Choi, S-C.,Gelfand,M., & Yuki, M. (1995).Cul-ture,gender,andself: A perspectivefrom individualism-collectivismresearch.Jour-nal ofPersonalityandSocialPsychology,69, 935-937.

Keesing,R. M. (1974).Theoriesof culture.AnnualReviewofAnthropology,3, 73-97.King, A. Y. C. (1991). Kuan-hsi and network building: A sociological interpretation.

Daedalus,120, 63-84.Leung,K. (1987). Somedeterminantsof reactionsto proceduralmodelsfor conflict reso-

lution: A cross-nationalstudy.JournalofPersonalityandSocialPsychology,5, 898-908.

ViiCRO-OF~GANIZAT ONA~’. ‘~f LOA (1 iV5

Leung, ~. (199’7). Negotiationandrewardallocationsacrosscultures.In P. C. Earley &Y. M. Erez (Eds.),Newperspectiveson international industrial/organizationalpsy-chology(pp. 640.675).SanFrancisco:Jossey-.Bass.

~ewg, K., & Bond, M. H. (1984).The ivipact oFcultural ~ollectivisrnon tewardalloca-tion. Journal ofPersonalityand~ ocial Psychology,47, 793-804.

Le un, K. (1948). Social-psychologicaldifferencesbetweenthe United StatesandGer-many. In G. W. Lewin (<d.), Resolvingsocialconflicts. London: Souvenir.

Lincoln, .1. R. ( 982).Intra- (andinter-) organizationalnetworks.kesear=~hin the Sociol-

ogy ofOrganizations,1, 1-38.Lindsl~y,S. L., & Braithwaite,C. A. (1996).Youshould“wear mask”: Faceworknorms

in cultural and intercultural conflict in maquiladoras.International Journal of

Intercultural Relations,20, 199.225.Lonner, W. I., & Adamopoulos,1. (1997). Culture as antecedentto behavior. in J. W.

Berry, Y. I-I. Poorting & I Pandey(Eds.), l-Iandbookofcross-culturalpsychology,Vol. 1: Theoryandmethod(pp.43-83). Boston:Allyn & Bacon.

Lukes, 5. (1967).Alienation andanomie.In P. Laslett& W. Runciman(lZds.), Philoso-

phy, politics, andsociety(3rd ed., pp. 134-156).Oxford, UK: Basil Blackwell.Lukes,5. (1973).Typesof individualism. In P. P. Wiener(Ed.), Dictionaryofthe history

ofideas:Studiesofselectedpivotal ideas(Vol. 2, pp. 594-604).New York: Scribner.Martin, J. (1992).Culture in organizations:Threeperspectives.NewYork: Oxford Uni-

versity Press.Marx, K. (1972). Economicand philosophical manuscriptsof 1844. In P.. C. Tucker

(Ed.), The Marx-Engelreader(pp. 53-103).NewYork: Norton.(Originally publishedin 1844)

Mayhew, B. (1980). Structuralismversus individualism: Shadowboxingin the dark.Social Forces,59, 586-596.

Mayhew,B. H., & Levinger,P.. L. (1976). Sizeandthe densityof interactionin humanaggregates.AmericanJournal ofSociology,82, 86-110.

Mead, D . (1935).Sexandtemperamentin threeprimitive societies.New York: WilliamMorrow.

Merton, R. K. (1969).Social theoryandsocialstructure.New York: FreePress.Mor is, M. W_& Leung, (.(2000).Justicefor all?Progressin researchon cultu ~alvaria-

tion in the psychologyof distributive and proceduraljustice. Applied Psychology:

An InternationalReview,49, 100..132.Morris, M. W., & Peng, . (1994).Culture andccus~:AmericanandChinese~ttributions

for s .ci A andphysica events.JournalofPersonalityandSocialPs.vchology,67, 9 9-971.

Morris,M. W., Podolny,.1. M., & A :1, S. ( 997,November).Theties thatbindin dijftr-

eni cultures: A cross-nationalcomparisonofthe determinantsof employeeobliga-tions to coworkers. Parerpresentedat the Citibank Behavioral ResearchCouncilConference,New York.

Morris, M. W., Podolny, .1. M., & Arid, 5. (199Th). Culture and normativesystems

of injbrmal interoedonat work: A studyofNorth American,Chinese,German, and

‘~latiOflal Constructs 89

Spanishoperations of a multinational consumerbank. Unpublished manuscript,StanfordUniversity, GraduateSchoolof Business.

M. W.,Podolny,J. M., & Ariel, S. (1998c).Cultureandtheorientationofobliga-

tion amongemployeesin an internationalfinancial institution. Unpublishedmanu-script,StanfordUniversity,GraduateSchoolofBusiness.

Murphy, M. (1983).Comingofagein Seville. Journal ofAnthropologicalResearch,39,376-392.

Nakane,C. 1970. Japanesesociety.Berkeley:University of California Press.Nakao, K. (1987).Analyzingsociometricpreferences:An exampleof JapaneseandU.S.

businessgroups.JournalofSocialBehaviorandPersonality,2, 523-534.Osgood,C. E. (1964).Semanticdifferential techniquein the comparativestudy of cul-

tures.AmericanAnthropologist,66, 171-200.Parsons,T. (1951).Thesocialsystem.New York: FreePress.pascale,R. T., & Athos, A. G. (1981). Theart ofJapanesemanagement.New York:

Simon& Schuster.Podolny,J. M., & Baron,J. N. (1997).Resourcesandrelationships:Socialnetworksand

mobility in theworkplace.AmericanSociologicalReview,62, 673-693.Pye,L. (1982).Chinesecommercialnegotiationstyle. Cambridge,MA: Oelgeschlager,

Gunn,& Ham.Redding,G., & Wong, G. Y. Y. (1986).The psychologyofChineseorganizationalbehav-

ior. In M. H. Bond (Ed.), Thepsychologyofthe Chinesepeople(pp. 2 13-266).HongKong: Oxford UniversityPress.

Roethlisberger,F. J.,& Dickson, W. (1939). Managementandthe worker. Cambridge,MA: HarvardUniversityPress.

Rohlen,T. (1974). For harmonyand strength: Japanesewhite-collar organizationinanthropologicalperspective.Berkeley:University of CaliforniaPress.

Ronen,5. (1986).Comparativemanagementandmultinationalmanagement.New York:JohnWiley.

Ronen,S., & Shenkar,0. (1985). Clusteringcountrieson attitudinal dimensions:A re-view andsynthesis,AcademyofManagementReview,10, 435-454.

Ross,L. D., & Nisbett,R. E. (1991). Thepersonandthesituation:Perspectivesofsocialpsychology.NewYork: McGraw-Hill.

Sanchez-Burks,J., Nisbett, R. E., & Ybarra, 0. (1998).Relationalschemas,culturalstyles,andprejudiceagainstoutgroups.Unpublishedmanuscript,University ofMich-igan,Departmentof Psychology.

Schwartz,S. H. (1994).Beyondindividualism/collectivism:New cultural dimensionsofvalues.In U. Kim, H. C. Triandis, & C. Kagitcibasi(Eds.),Individualismandcollec-tivism: Theory,methods,andapplication(pp. 85-119).ThousandOaks,CA: Sage.

Scott,J. (1991).Socialnetworkanalysis:A handbook.NewburyPark,CA: Sage.Simon, H. (1945). Administrativebehavior: A study of decisionmakingprocessesin

administrativeorganization.New York: Macmillan.Singelis,T. M., Triandis,H. C., Bhawuk, D., & Gelfand, M. J. (1995). Horizontaland

verticaldimensionsof individualismandcollectivism:A theoreticalandmeasurementrefinement.Cross-CulturalResearch,29, 240-275.

90 M U 00k ‘AN Z/ 1CN.’ ~<AFPR0~Cl-il3S

Swidler, A. ( 986). Culture ~naction: Symbolsand strategies.AmericanSociologicalReview,51, 273-286.

Thompson,J. (1967). Organizationsin action: Social sciencebase.~of administrativetheory. New York: McGraw-I ill.

Tonnies, ~. (l%7). Communityand society. New Brunswick, N.l: iYansactionBooks.(Originally publishedin 1887)

Triandis,H. C. (1967). Interpersonalrelationshipsin internationalorg. nizations.Journal

ofOrganizationalBehaviorandHumanPerformance,2, 26-55.Triandis,H. C. (1972). Theanalysisofsubjectiveculture.New York: .:‘ohn Wiley.Triandis,H. C. ( 995). Individualismandcollectivism. Boulder,CO: .Vestview.Triandis,H. C.,& Gelfand,M. J. (1998).Convergingmeasurementof horizontalandver-

tical individualism andcollectivism.Journal ofPersonalityandSocial Psychology,74, 118-128.

Triandis, H. C., Leung, K., Villareal, M., & Clark, F. L. (1985). Allocentric vs. idio-centrictendencies:Convergentand discriminantvalidation.Journal of I?esearchinPersonality, 19, 395-415.

Triandis,LI. C., Mann, G., Lisansky,J.,& Betancourt,H. (1984). Simpatiaas aculturalscript of Hispanics.JournalofPersonalityandSocialPsychology,47, 1363-1375.

Triandis,H. C., McCusker,,C., & Hui, C. 1-1. (1990). Multimethodprobesof individual-ism andcollectivism.Journal ofPersonalityandSocialPsychology,59, 1006-1020.

Weber, M. (1947). The theoryofsocialandeconomicorganization.New York: OxfordUniversity Press.

Weber, M. (1963). Thesociologyofreligion. Boston: Beacon.(Originally publishedin1922)

Wertsh, I. V., del Rio, P., & Alvarez, A. (1995).Socioculturalstudiesofmind Cam-bridge,UK: CambridgeUniversity Press.

Wheeler,L., Reis, II. T., & Bond, M. [1. (1989). Collectivism-individualismin every-day life: The Middle Kingdomandthe meltingpot. JournalofPersonalityandSocial

Psychology,57, 79-86.White, H. C.,Boorman,S. A., & Breiger,R. L. (1976).Socialstructurefrom multiplenet-

works.AmericanJournalofSociology,81.Wrong,D. H. (1961).The oversocializedconceptionof manin modernsociology.Ameri-

can SociologicalReview,42, 32-56.

Missing Relations - Columbia Business School 1995). Cultural differences in values, beliefs, and...

Documents

Transcript of Missing Relations - Columbia Business School 1995). Cultural differences in values, beliefs, and...