VIEWS FROM INSIDE AND OUTSIDE: INTEGRATING EMIC

15d ago
5 Views
0 Downloads
7.31 MB
17 Pages
Transcription

Academy ol Maaagem»nt Review1999, Vol. 24. No. 1781-796.VIEWS FROM INSIDE AND OUTSIDE:INTEGRATING EMIC AND ETIC INSIGHTSABOUT CULTURE AND lUSTICE JUDGMENTMICHAEL W. MORRISStanford UniversityKWOK LEUNGChinese University of Hong KongDANIEL AMESUniversity of California at BerkeleyBRIAN LICKELUniversity of California at Santa BarbaraWe analyze forms ol synergy between emic and etic approaches to research on cultureand cognition. Drawing on the justice judgment literature, we describe dynamicsthrough which the two approaches stimulate each other's progress. Moreover, wedelineate ways in which integrative emic/etic Irameworks overcome limitations ofnarrower frameworks in modeling culture and cognition. Finally, we identify advantages of integrative frameworks in guiding responses to the diverse justice sensitivities in international organiiKations.In the study of cognition in organizations, andin social science more broadly, there are twolong-standing approaches to understanding therole of culture: (1) the inside perspective of ethnographers, who strive to describe a particularculture in its own terms, and (2) the outside perspective of comparativist researchers, who attempt to describe differences across cultures interms of a general, external standard. Pike (1967)designates these approaches the emic and eticperspectives, respectively, by analogy to twoapproaches to language: phonemic analysis ofthe units of meaning, which reveals the uniquestructure of a particular language, and phoneticanalysis of units of sound, which affords comparisons among languages. The emic and eticperspectives are often seen a s being atodds—as incommensurable paradigms. In thisarticle we argue that these two approaches toculture are complementary. Drawing on the justice judgment literature, we delineate forms ofsynergy between the two research perspectivesthat go beyond those identified previously (e.g.,Berry, 1990; Brett, Tinsley, lanssens, Barsness, &Lytle, 1997). We first analyze ways in which emicand etic research programs have stimulatedeach other's progress. Then we analyze advan-tages of frameworks integrating emic and eticaccounts—both as middle-range theories of culture and cognition and as applied guides toresponding to diverse justice concerns in international organizations.EMIC AND ETIC PERSPECTIVESThe emic and etic perspectives have equallylong pedigrees in social science. The emic orinside perspective follows in the tradition ofpsychological studies of folk beliefs (Wundt,1888) and in cultural anthropologists' striving tounderstand culture from "the native's point ofview" (Malinowski, 1922). The etic or outside perspective follows in the tradition of behavioristpsychology (Skinner, 1938) and anthropologicalapproaches that link cultural practices to external, antecedent factors, such as economic orecological conditions, that may not be salient tocultural insiders (Harris, 1979).The divide between these two approachespersists in contemporary scholarship on culture:in anthropology, between interpretivists (Geertz,1976, 1983) and comparativists (Munroe & Munroe, 1991), and in psychology, between culturalpsychologists (Shweder, 1991) and cross-cultural781

782Academy oi Management Reviewpsychologists (Smith & Bond, 1998). In the literature on international differences in organizations, the divide is manifest in the contrast between classic studies based on fieldwork in asingle culture (Rohlen, 1974), as opposed to surveys across many {Hofstede, 1980). Likewise, inthe large body of literature on organizationalculture, there is a divide between researchersemploying ethnographic methods (Gregory,1983; Van Maanen, 1988) and those who favorcomparative survey research (Schneider, 1990).The conceptual assumptions with which Pike(1967) defined the emic and etic dichotomy aresummarized in Table 1. Emic accounts describethoughts and actions primarily in terms of theactors' self-understanding—terms that are oftenculturally and historically bound. For example,emic studies of justice perceptions in NorthAmerican organizations today might center onsuch constructs as "age-ism" and nondiscrlmination, whereas studies of Japanese workplacesmight be couched in qualitatively different constructs, such as amae and gimu (see Kashima &Callan, 1998). In contrast, etic models describephenomena in constructs that apply across cultures. For example, a country's level on the cultural dimension of individualism-collectivismmight be linked to the prevalence with whichmanagers reason about justice in terms of theequity rule (i.e., rewards received should be proportional to contributions).Along with differing constructs, emic and eticresearchers tend to have differing assumptionsabout culture. Emic researchers tend to assumethat a culture is best understood as an interconnected whole or system, whereas etic researchers are more likely to isolate particular components of culture and state hypotheses abouttheir distinct antecedents and consequences. Although, of course, the emic/etic contrast is, inpractice, a continuum, this dichotomy hasplayed a central role in the metatheory debatesin many social science disciplines (see Headland, Pike, & Harris, 1990).'' Some scholars have used the terms emic and etic inways that depart irom Pike's definitions (see Headland et al.1990). A narrower usage refers to the contrast between culture-specific versus culture-general constructs. This missesthe essence of the distinction, because culture-specific constructs do not necessarily resonate with cultural insiders'self-understandings, A broader usage refers to the underlying interests of understanding versus control (Habermas,OctoberEtic and emic approaches traditionally havebeen associated with differing research methods. As Table 1 summarizes, methods in emicresearch are more likely to involve sustained,wide-ranging observation of a single culturalgroup. In classical fieldwork, for example, anethnographer immerses him or herself in a setting, developing relationships with informantsand taking on social roles (e.g., Geertz, 1983;Kondo, 1990). Yet, emic description also can bepursued in more structured programs of interview and observation (e.g., Goodenough, 1970).Methods in etic research are more likely toinvolve brief, structured observations of severalcultural groups. A key feature of etic methods isthat observations are made in a parallel manneracross differing settings. For instance, matchedsamples of employees in many different countries may be surveyed to uncover dimensions ofcross-national variation in values and attitudes(e.g., Hofstede, 1980), or they may be assigned toexperimental conditions in order to test the moderating influence of cultural setting on the relation among other variables (e.g., Earley, 1989). Insum, although the two perspectives are definedin terms of theory, rather than method, the perspectives lend themselves to differing sets ofmethods. Given the differences between emic and eticapproaches to culture, it is not surprising thatresearchers taking each perspective have questioned the utility of integrating insights from theother tradition. A common tendency is to dismissinsights from the other perspective based onperceived conceptual or methodological weaknesses (see reviews of this tendency in particular research areas by Harris, 1979, and Martin &Frost, 1998). On one side, emic accounts basedon ethnographic observation are often discounted on the basis of inconsistency across1971). Although there may be a correlation in some researchareas between the emic versus etic perspective and orientations toward control (e.g., in studies of "organizationalculture"; Martin & Frost, 1996), there is no necessary link andno strong correlation in the literature on national culture—our focus. The association between perspectives and methods isnot absolute. Sometimes, in emic investigations of indigenous constructs, data are collected with survey methods andanalyzed with quantitative techniques (Farh, Earley, & Lin,1997; Yang, 1986). Likewise, ethnographic observation andqualitative data are sometimes used to support argumentsfrom an etic perspective (Nelsen & Barley, 1997: Sutton, 1994).

1999Morris, Leung, Ames, and Lickel783TABLE 1Assumptions of Emic and Etic Perspectives and Associated MethodsFeaturesEmic/Inside ViewEtic/Outside ViewDefining assumptions andgoalsBehavior described as seen from theperspective of cultural insiders, inconstructs drawn from their selfunderstandingsDescribe the cultural system as aworking wholeBehavior described from a vantage external tothe culture, in constructs that apply equallywell to other culturesDescribe the ways in which cultural variablesfit into general causal models of aparticular behaviorTypical features of methodsassociated with this viewObservations recorded in a richFocus on external, measurable features thatqualitative form that avoids impositioncan be assessed by parallel procedures atof the researchers' constructsdifferent cultural sitesLong-standing, wide-ranging observation Brief, narrow observation of more than oneof one setting or a few settingssetting, often a large number of settingsExamples of typical studytypesEthnographic fieldwork; participantobservation along with interviewsContent analysis of texts providing awindow into indigenous thinkingabout justicereports (Kloos, 1988) and for inheriting misconceptions from cultural insiders (Marano, 1982).On the other side, etic accounts based on surveydata are often dismissed because researchersremained at a distance from respondents, potentially insensitive to how respondents were affected by their questions (Geertz, 1983).Yet, not all arguments against integration arestaked on critiques of either approach. Separatism has been defended as a means to protectless well-institutionalized traditions from beingassimilated by mainstream traditions. Writingabout organizational culture, Martin argues that"pressures toward assimilation would undermine a perspective's inherently oppositionalstance . . . threatening its conceptual and political integrity" (1992; 187). In sum, both partisanand protective agendas have led scholars to advocate keeping emic and etic insights about aphenomenon somewhat separate.However, not all previous scholars hold thatemic and etic approaches should be kept apart.Some have suggested that researchers shouldselect between approaches, depending on thestage of a research program. For example, it hasbeen argued that an emic approach serves bestin exploratory research, whereas an etic ap-Multisetting survey; cross-sectionalcomparison of responses to instrumentsmeasuring justice perceptions and relatedvariablesComparative experiment treating culture as aquasi experimental manipulation to assesswhether the impact of particular factorsvaries across culturesproach serves best in testing hypotheses (e.g.,Greenfield, 1996).In a more explicit selectionist proposal. Berry(1990) endorses a three-stage sequence. In thefirst stage, initial exploratory research relies on"imposed-etic" constructs—theoretical conceptsand measurement methods that are simply exported from the researcher's home culture, In thesecond stage, emic insights about the other culture are used to interpret initial findings, withan eye to possible limitations of the originalconstructs, such as details that are unfamiliar ormeaningless outside of the home culture. Onthis basis, then, the constructs in the model arefiltered to eliminate details that cannot be measured with equivalence across cultural settings.The factors that survive this filter—"derivedetic" constructs—are culture-general dimensions of persons, such as value orientations, orof their environments, such as economic or ecological factors. In the third and final stage, theresearcher tests an explanation constructedsolely of derived etic constructs.Brett and colleagues (1997; Lytle, Brett, Barsness, Tinsley, & Janssens. 1995) describe anotherproposal based on a three-stage sequence.These scholars differ from Berry in sharply dis-

Academy of Management Review784tinguishing cultural factors from ecological andeconomic factors in analyzing cross-nationaldifferences. Also, they suggest that etic constructs may not always require measurementequivalence. However, as in Berry's model, theend state is an explanation drawn in terms ofetic constructs; emic insights guide the initialsteps but are not retained in the final explanation. The sequential selection models of Berry(1990) and Brett et al. (1997) have been influentialin guiding psychological and organizational researchers in their approaches to culture. Yetthese analyses only begin to explore the synergies between perspectives. Although they address the role of emic insights in refining eticexplanations, they say little about how etic insights stimulate emic investigation. Althoughthey address the interplay between perspectives within a given research program, they donot analyze long-term interplay across researchprograms within a general research area. To laythe groundwork for a long-term analysis, wenow introduce the research area of justice judgments.lUSTICE lUDGMENTJudgments of justice occur whenever authorities in a group allocate resources or rewardsamong its members. For instance, when a manager gives a larger bonus to an energetic youngsalesperson than to her more senior and experienced colleague, observers will evaluate thismanager positively or negatively, depending onwhether they judge the rewards to be in balancewith the employees' respective contributions.Managers and others who wish to be perceivedpositively need to understand how observers arrive at justice judgments. This is not a trivialtask, however, because it is not always selfevident what is fair or balanced; justice is not so Brett et al. (1997) describe a second form of emic-eticinterplay in research conducted by a multinational researchteam. Team members rely on emic understandings of theirrespective local cul