An Emic Perspective and Ethnoscience Methods for Organizational Research Author(s): Nancy C.

Morey and Fred Luthans Source: The Academy of Management Review, Vol. 9, No. 1 (Jan., 1984), pp. 27-36 Published by: Academy of Management Stable URL: http://www.jstor.org/stable/258229 Accessed: 28/08/2010 17:20
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=aom. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Academy of Management is collaborating with JSTOR to digitize, preserve and extend access to The Academy of Management Review.

http://www.jstor.org

cAcademy of ManagementReview, 1984, Vol. 9, No. 1, 27-36.

An

Emic

Perspective and Ethnoscience Methods for Organizational Research1

NANCY C. MOREY FRED LUTHANS

Universityof Nebraska, Lincoln

Thispaper is in answer to the call for new, innovativeperspectivesand methodologies for organizational research. Althoughthe approach here falls withinthe subjectivelidiographic/qualitative/insider set of methodologies ratherthanthe objective/nomothetic/quantitative/outsider set, thereis the potential to bridgethe gap betweenthe two sets. Cominglargely from anthropology,the emicperspectiveis explained;the specificstepsfor ethnoscienceanalysesaresummarized; and examplesand implications aregiven.
The last few years have seen an increasing call for innovative and broadening methodologies for organizational research. For example, Hackman, in writing the introductions to a series of volumes on innovations in methodology for organizational research, states: The methodologies usedin research on organizations have been far too limitedand conventional.... Because the need for higherqualityorganizational reis pressing, search now maybe the timeto tryto break throughthe constraints of traditional methodologies and seek new approaches to organizational research (1982, p. 8).

Contrasting Views of Social Science Research

Although there are many issues involved in this awakened interest in research methodologies, Burrell and Morgan (1979) identify one of the most basic. In their distinction between the subjectivist and the objectivist approaches to science, a common theme can be found in the majority of suggestions for new perspectivesand methods for organizationalresearch. As they explain it, the issue revolves around certain basic assumptions about what it is that organizational researchersare investigating and how they should go about it. Is the "reality" of investigation something
'The research leading to this paper was supported in part by the Organizational Effectiveness Research Group, Office of Naval Research (Code 442), under Contract No. N00014-80-C-0554; NR 170-913 (Fred Luthans, principal investigator).

that is imposedon the subject(i.e., externaland objective), or is it a productof the cognitionsof the subject(i.e., internaland subjective)? Obviously,an organizational researcher followingthe subjectivist view will take a differentapproachfrom that of the researcher following the objectivistview. This basic split in orientationsto research can be expressedin a varietyof dichotomies.One such dichotomyis found in the terms"idiographic" versus "nomothetic."Luthansand Davis(1982)describea nomotheticperspective as one that is group-centered and usesstandardized, controlled environmental contexts and quantitative methodsto establishgeneral laws. Theydescribe an idiographic perspective as one that is individual-centered and uses naturalistic environmental contextsand qualitative methodsto recognize the particularand unique experienceof the subject. Another, more limited dichotomy is the qualitativeversusquantitative.Qualitative research impliesan idiographic, particularistic perspective of science.Quantitative research withstatistical analyses of data takes a nomothetic,generalizing perspective of science. Still anotherdichotomycan be found in the distinction between the terms"insider" and "out" An insider's sider. perspective of sciencewouldtake the viewof the organizational participant in research; the outsider'sorientation wouldtakethe nonparticipant "scientific" researcher's view. Therefrequentlyis a subtleshift in extensionsof the terminologyof these differentdichotomies,and the termsoften becomemerged andusedinterchange27

viewscan be clustered ably. Thus, the contrasting into the terms subjective/idiographic/qualitative/insiderand the termsobjective/nomothetic/quantitative/outsider. Use of any one term in the set often that the othershave conjuresup all the implications for whoevermakes the distinctions.When the implied dichotomiesare acceptedon faith, too often thereis an associated distinction between"good" and "bad" science/research. Wherethe value is applied depends on the "camp" of the particularscholar. Traditionally,the vast majority of organizational researchers havebeen objectivist/nomothetic/quantitative/outsider advocates. hasthe Onlyveryrecently subjectivist/idiographic/qualitative/insider perspective been recognized. The potential and real disagreements among organizationalresearchers over these contrastingapproachesare unfortunateand can lead to neglectof commoninterestsand understanding on both sides. This paperpresentsthe modestproposalthat a generally ignoredresearchperspectiveand set of techniques from anthropologymay help organizational researchers overcomesome of the problemsassociated with these disagreements and also may meet some of the goals for organizationalresearchproposed by advocates of new methodologies. It is recognized that both the subjectivist/idiographic/qualitative/insider and the objectivist/ nomothetic/quantitative/outsider campspresentlegitimatemethodologiesfor organizational research. The authorsalso feel stronglythat these should not be consideredmutuallyexclusiveapproaches. In the interest of forginga rapprochement between advocates of the contrasting views,presented hereare an emicperspective and a specificset of ethnoscience techniquesfrom anthropologythat show particular promisefor being both qualitativeand idiographic in field applications,but with potential for quantificationandnomothetic explanatory analysis.These techniques offer insider,subjective dataof immediate practical utility for practicing managers and researchers and also combinethis with data gathering techniques that can be objectified,and thusbe adaptableto moretraditional methodological analysisand conclusions. A brief review of recent literature discussingmethodologicalissues in organizational will set the stageand serveas a logicalpoint research of departure.
28

Methodological Concerns
Questionnaires arethe most convenientand widely used technique of data gathering for objective/nomothetic/quantitative/outsider studies. Recently, some organizationaltheoristshave been of the overuseof questionnaires critical (Dubin,1982) and the reliability and validityof standardized questionnaires widely used in organizationalresearch & Kerr,1977;Schriesheim, (Schriesheim Bannister, & Money, 1979).Questionnaires also arebeingcriticized by advocates of qualitativetechniquesfor a variety of failings in design, results, and usability (VanMaanen,Dabbs,&Faulkner,1982).Amongthe proposedalternatives,or supplements, to questionnaireshas been the call for observationtechniques (Dubin,1982;Luthans & Davis, 1982).Dubin(1982), for instance,urgesorganizational researchers to examinetheirpastandreturn to someof the morefruitful participant observation thatwereused techniques in classic studies in the field. A growing, but still relatively small,number of organizational researchers aretakingthis advice.Examplescan be found in the works of Bussom, Larson, and Vicars(1981), Kotter (1982), and Van Maanen and his colleagues (1982).In additionto observation techniques perse, the role of qualitativemethodology in general is receiving moreattention.For example,VanMaanen et al.'s (1982) recent book and a 1979 issue of the Administrative ScienceQuarterly show some of the wide ranging possibilities for different kinds of qualitativeresearchin organizationalstudies. Despite these suggestedalternatives,it must be remembered thatcriticsof questionnaires wantmore than just observation and qualitativetechniques. These are just the surfacearguments and proposed alternatives.More importantlyis an expansion of organizational research that will take accountof the "insider's"view of the organization.For example, Pfeffer calls for a distinctionbetweentwo levels of analysisof organizations: (1) the levelconcerned with prediction of actionstakenwithinorganizations (outsider's)and (2) that concernedwith predictingand how suchorganizational understanding activities are "perceived, interpreted, and legitimated" (insider's) (1981, p. 8). As anotherexample,Everedand Louis (1981)specificallycall for attentionto the insider's viewand a linkageof it withthe opposingoutsider's view. They are concernedwith defininghumanaction within settings,the insider'sown definitionof

of the insider, the situation,the motivesandpurposes and contextof the situation.Burrell andthe historical Morgan(1979)also believethat this viewpointneeds to be added to traditionalorganizationalresearch. Weick (1979) specificallypinpointsthis insider's viewpointwithin the study of cognitive process in organizations.He believesthat it shouldinvolve attention to examinationof thoughts, thinkingpractices, and thinkersin organizations.Organizations, statesWeick,can be viewedas bodiesof thoughtand can be describedin terms of sets of thinkingpractices. Such descriptions,then, would emphasizethe dominantruleswithinthe organizationfor combining variouskinds of thinkingpractices.The central researchjob, at the descriptivelevel, would be to identifythe importantcognitivevariables(maps of the organization in people's perceptions, beliefs throughwhichpeoplesee the organization,and acts that providethe raw materialfor "sensemaking"). He wouldliketo see thesecombinedin research projects with organizational variablessuch as size, forin orderto join cogmalization,and centralization nitive and organizational theory. Pfeffer (1981) also calls for an emphasis on cognitiveapproachesin terms of "meaning." This is secondedby Dubin (1982), who requeststhat researchers considerthe intentionality of actorsin the and the meanings organization attachedto behaviors that are observedand cognitive processesthat are measuredby traditionaldata gatheringtechniques such as questionnaires. The present authors are concernedabout what seems to be a wideninggap betweenthe two major orientationsto organizational research.Burrelland Morganhavetakenthe pessimistic stancethatthe two groupscannot be broughttogether,referring to the "disinterested hostility"(1979, p. 36) characterizing their relationships.Dubin (1982) and Van Maanen et al. (1982) emphasizethe lonely course to be run by the researcherattemptingto deviate from the traditionalpath. Everedand Louis (1981), who obviously want to see the two approachesbrought together,stillfeel thatresearchers trained in andcommittedto quantitativetechniquesand those trained in and committedto a morequalitative approach are likelyto clash and not recognizeone another'scontributions to organizational research. Lammers (1975)recognizesthis problem,but he offers a solution. He suggeststhat those so inclinedtake up both approachesat the same time, within the same re29

searchproject. Another suggestedsolution offered by Luthans and Davis (1982), Evered and Louis (1981), and Burrell and Morgan (1979) is that researchgo back and forth or proceedsequentially from one to the other. In this paper, the authors propose still another solution to the growingdilemmaposed by the need for new methodologies,on the one hand, but the potentialfor polarizationon the other. An "emic" research perspective borandethnoscience techniques rowed from anthropologyare suggested.This perspectiveand accompanying techniquesseem readily to organizational research adaptable and, alongwith an "etic," more analyticalapproach,may have the potential to help merge the subjective/idiographic/qualitative/insiderand the objective/nomothetic/quantitative/outsider approaches.

An Emic Research Perspective

The terms "emic" and "etic" in anthropology were originallyintroducedby a linguist, Kenneth Pike, who coinedthemusingthe suffixesof the terms phonemic and phonetic, familiar categories in linguistic analysis.Loosely,thesetermsdistinguished sound structure,as analyzedby a linguist (phonetics) from the meaningof the sounds to the native speaker(phonemics).The termemic has sincecome to denote a generalorientationin researchcentered on the native, that is, the insider's or, as anthropologists call it, the "informant's"view of reality. Thus, the emic approachemphasizesnative or respondent categoriesand meanings in general and native rules for, or respondentbehavior, in particular. Etic designatesthe orientationof outside researchers, who havetheirown categories by which the subject's world is organized. The analyticaldescriptive categories of the outside researcher generallyare organizedwith a view to explanation in the broadersense traditionally used in organizational research.What the emic-eticdistinctionproduces, in its most extremeinstances,is the type of divisionin methodologicalapproachthat presently characterizes organizational research. It neednot remain such, however.The methodological approach suggested here may bring the two opposing approachescloser together. Extreme of the emicviewpoint adherents insistthat the subjectand not the researcher is the best judge of the adequacyof the researchand analysis. The subject'sacceptanceof the resultsof the research is

and sufficientvalidationof them the only necessary (Frake, 1980;Sturtevant,1964).But extremeadherents of the etic approachbelievethat the researcher is the best judge of the adequacyof the description or analysis.The subject'sopinionmaybe interesting, but it is not really relevant(Harris, 1979). Most anthropologists fall somewhere between these two extremes,utilizingboth emic and etic apand analyproachesto completetheirtotal research tical designs. As Pelto (1970) indicates,there is an "imbedded emicism"in all anthropological research at the fieldworklevel, in which native viewpoints, meanings,interpretations, and so on are givengreat importancefor understanding behavior. However, movinginductivelyup the levels of analysis,the anthropologistbecomes increasingly etic in approach as the importanceof universalcategoriesfor comparisonbecomespredominant. Eventually,emic categories are fitted to etic concepts so that general propositionsabout human behaviorcan be tested. Obviously,whicheverapproachis taken at a given time (emic or etic) will dependon the research questions beingaskedand the stageof the research being conducted. Theemicperspective to research is compatible with an idiographicdesign. An emic orientationsuggests that researchbe conductedwith particularindividuals,focusingon theirunique,individual "insider" viewpoints. The appropriate methods for such researchare basically qualitativein nature, using observation techniques of datagathering and ethnoscience techniques from anthropology. However, once again it should be recognized that anthropologistsdo not treatthis approach as the end of the research.They stress that the insider's viewpoints mustlaterbe translated into outsider'scategories for purposesof nomothetic analysisand generalizations. In other words, the position taken here is that organizationalresearchshould include more emic (subjectivist/idiographic/qualitative/insider)perbutthesethenwouldgenerally spectives, be translated into etic (objectivist/quantitative/nomothetic/outsider) terms. Both approachesseem essentialfor a completeresearch perspective, but the emic perspective has been largely overlookedin organizational researchand thereforeis the one given specific attention in this paper. In addition it is time to go beyondgeneralprescriptions and "advocacy"statementsabout the emic perspective and providesome actualresearch techniques thathavebeensuccessfully
30

andwouldseemto be especially usedin anthropology applicableto organizational research.

EthnoscienceTechniques
Commonlyused termsin anthropologyfor emic research techniques include ethnosemantics, ethnographic semantics,ethnographic ethnoscience, formal analysis, and componential analysis. Ethnosemantics, ethnographic semantics, and ethnographicethnosciencereflect slightly different in a technicalsense,but they are essentialemphases ly interchangeable terms. They all referbasicallyto a consciouslimitationof researchto the analysisof verbalcategorieselicitedfrom respondents.Formal analysisrefersto an analyticalstep following emic elicitationof data in whichthe data are represented in termsof formalset theory.Componential analysis is a particular techniquefor analyzing the attributes, or " components," of contrasting setsof lexicalitems or words.Theentirerangeof techniques is mostoften underthe covertermethoscience.This encompassed term is what is used here. comprehensive Specific ProceduresUsed There are many ways to begin eliciting "'emic" data from a subject. Most of the proceduresare disarmingly(and deceptively)simple. Perhaps the most directand powerfulis suggested by Hunterand Foley (1976). They start with what could be called the "emicquestion."The procedure simplyinvolves asking a subject what he/she is doing, listing the responses,and then following up each item on the list with a furtherquestionsuchas: "Whatkindsof questionsdoes it makesensefor me to ask you about
-

?"

This eventually will produce a host of

questionsthat can be used to pursueeach topic further. The responsesalso can be used to begin questioning others involvedin the same activitiesin the same culturalsetting. Spradleyhas operationalized many of the techniques of ethnosciencein a seriesof books (Spradley, 1979,1980;Spradley & McCurdy,1972)thatgivedetailed and easy-to-follow instructions for basic ethnoscience research. Althoughsomeorganizational researchers undoubtedly are familiarwithSpradley's work, his six steps can providea useful framework for presentingethnoscienceresearchtechniques.In brief, these steps are as follows: 1. AskingDescriptiveQuestions.Thesequestions defineimportant cultural settingsin the respondent's

termsand also can be used to find out more about a sequenceof importantevents. They move from to specificin a set pattern.Spradley general illustrates how to begin using "native language"to minimize eticinfluence andmaximize valuefromthe questions. Thesedescriptive questions,systematically pursued, of the intersection focus on all the possibilities of the nine categoriesof space, object, act, activity,event, time, actor, goal, and feeling.For instance,an event by time questionwould be phrasedto find our how an eventof interestfalls into particular timeperiods. An actorby actor questionwould requesta description of all the "actors" in a culturalsettingor event of interest. 2. Making a Domain Analysis. A domainis any symboliccategorythat includesothercategories,all of whichshareat least one featureof meaning.Domains consist of a cover term that names the category, a seriesof includedtermslinkedby a semantic relationship,and a boundary.A domainanalysisis much more difficut in practicethan it would seem from a descriptionof the task. This is principally because investigatorshave so many of their own presetcategoriesthat it is difficultto continuequestioningrespondents for theirown. Thereis a tendency to assumetoo early (especiallyin one's own culture)that thereis no more informationneeded,and thus the researcher can miss a great deal of important data. The unfamiliaraspectof domainanalysis is likely to be the semanticrelationship connection. A semantic relationshipis the connector between subsetsand the domaincoverterm. Rohlen's(1974) Japanesebank study providesan example.He suggests that a domain termed "dismissal' can be isolated. Theft, breakinga majorlaw, and extremely unrulybehaviorarelikely"included terms"under this domain.Thesetermsareall linkedto the domain cover term by the semanticrelationship"is a way to," becausethese activitieswill lead to dismissal, but almostno others.Othersemantic relations might be "is a kind of " or "is a part of." Spradleylists severalsuchsemanticrelationships for whichhe proposesuniversal applicability and whichare usefulfor beginningdomain analysis. Weick (1979) also gets into a type of domain analysis with what he calls "relationalalogorithms."Theseare preferred ways that peoplecombinecognitionsand bits of information with relational words. The domain analysis described heremighthelpthe researcher arriveat exactly what these alogorithmsmight be.
31

3. Asking StructuralQuestions.These questions builda usefuldescriptive pictureof the cultural scene or event of interestand are a procedure that finally begins to provide information with potential for quantification and comparison. Structural questions usually require more explanationthan do simple descriptive questions.They often are in the form of examples. Spradley discusses special techniques in askingsuch questionsand distinguishes necessary five majortypes of structural questionsand several subtypes, all of whichhavedifferentpurposes.Structuralquestionsreachfurtherinto the structure of a respondent'sknowledge.Followingthe exampleof the Japanese structural bank,further of the questions type called"covertermquestions"mightpursuethe domain of "dismissal"by asking: "Are there differentkinds of dismissal?"or "Are theredifferent waysto be dismissed?" or "Whatareall the different steps in dismissal?" 4. Makinga Taxonomic Analysis.In a taxonomic analysisthe researcher selectsa particular domainfor extensivequestioning.The goal is to determineall of the inclusiverelationships that can be found for that domain. The work of Burton (1972) demonstratesthisprocedure. His research dealtwithEnglish role terms,and one aspectof his studyused the domain of occupationtermsfor in-depthanalysis.He was particularly interested in the correspondence betweenthe meaningof occupationnamesand his respondent's judgmentsof themin relationto prestige as an attribute. His firstleveltaxonomyhas the form shown in Figure 1. This investigationof occupationscontinuedwith a techniqueknown as sorting (discussedin detail later), the resultsof whichhe submittedto a multidimensional scalinganalysis.The three-dimensional representation of his dataverifiedthe hypothesis that the criterion of prestigehad beenusedin the respondents'sortingof 60 occupationnames.In taxonomy terms, his hierarchial clustering,greatlysimplified, createda taxonomyas shown in Figure2. 5. Asking ContrastQuestions.Thereare several different types of contrast questions. Their basic point is that the meaningof a symbol can be discoveredby findingout how it contrastswith others in the samedomain.Taking just one of these,"rating questions,"the researcher can get information about valuesplaced on sets of symbolsby asking respondentsto makecontrastson the basis of whichterms are best, easiest, most difficult, worst, most in-

Figure 1 A Taxonomic Analysisa

Role Terms Occupations Kin Terms S gers and Other Disreputable Characters Artists Other Occupations

aFrom Burton (1972).

othercriterion or whatever teresting,most desirable, is preferred.This would be the type of questionto use to pursuekinds of tasks employeespreferover others. It often creates scales in which items are rankedalong the dimensionchosen. The important pointis that theseareemic scales.Theyderiveexclusively from the categoriesof the respondents; they arenot responses to scalesor to categories presetand defined by the researcher or on face validity. Such ratingand rankingquestionshave been used profitably by anthropologists to discover"native" stratificationsystems,the criteria on whichtheyarebased and the units subsumedin the rankings(for example, see Silverman,1966). 6. Making a Componential Analysis. A componential analysis is a systematicsearch for the atof a symbol.Comtributes of meaning) (components in the form ponential analyses usuallyarerepresented of paradigmsthat schematically distinguishall the membersof the contrastset in the domain of concern and show the multiple relationshipsbetween them. In makinga componentialanalysisthe emicorientedresearcher would take all the membersof the contrast set of interest and discover how theycontrast with each other on differentdimensions.The purposehere is to find out the attributesand create the contrastsin the set. Returningto the Japanese bank example, the researcher interestedin doing a componentialanalysisof categoriesof employeesin this organization wouldfind a number of dimensions of contrast that would have to be investigated. Employeesmay be members,quasi-members, and nonmembers of the bank. Eachcategoryhas certain attributesthat createits meaning.Thereare distinctions in the mode of recruitmentof different employeecategories and in theirmeansof selection,and there is a cross-cuttingdimensionof sex that adds further complications in determining attributes
32

(Rohlen, 1974). Any componential analysis of employeeswould have to considerthe interrelationships of all these attributes.A typical, simplecomponential analysis paradigmmight have the form shown in Table 1. The rows contain the attributes associatedwith a particulardomain or subset of a domain. The columnsshow the dimensionsof contrast betweendomains or their subsets. Other EthnoscienceTechniques The above discussionof the least complexaspects of Spradley'sethnoscience"manuals," with additional illustrativematerial,providesa generaloverview of the specific procedures of ethnoscience methodology. There are, of course, other specific ethnosciencetechniques. TheUseof Lists.All the techniques in ethnoscience begin with lists of one kind or another. These lists usuallyare obtainedby unstructured interviewing to make certainthat the categories(symbols)of which they are composed are as emic as possible. SortingProcedures.Items derivedfrom the lists often areusedin sortingprocedures. Sortinginvolves puttingnamesof the list items on cardsand having the respondents categorize themon some basisof interestto themor to the researcher. Respondents may to dividethe cardsinto as many simplybe requested piles as they think appropriate. The researcher then questionsthe respondents to learnthe basis for this sorting. The researcher also will try to elicit cover terms that will characterize the individualpiles in some manner. Triadsortingis a specialvarietyof forced choice sortingin whichthe respondentis given threecards at a time and asked to pick the two that are most similar to each other, eliminating the one least similar.Whenthe choice is made, the researcher attempts to learn the basis for it. This is one way in

Figure 2 Results of a Multiple-Dimensional Scaling Analysis Based on Sortinga

Technical Planning / Outdoors

Skilled Workers

~~~Business\\\
Professions

Laborers

Trades and Industrial

A aFrom Burton (1972).

JK

whichcomponentialanalysisdata can be elicited.A number of other variations on sorting also exist (Pollnac, 1975). The Use of Frames. Eliciting frames, sentence frames, or substitutionframes are constructedto elicit the kinds of informationSpradleyrefersto in his categories of descriptivequestions, structural questions, and contrastquestions. Sentenceframes aresimply"fill-in-the-blank" The typesof questions. researcher variesthe key elementin the sentenceto see how respondents vary their responsesto the restricted framework. Frames areconstructed by listening to naturalconversationand selectingphrasesto testwith "native"respondents to be certainthatthey make sense. Usually they are only sentence fragments.Whenusedwithinthe samedomain,sentence framescan be combinedand recombined to see what patternsemergein responses. Returningto the Japanesebank example,to find out more about types of employees and their attributes,the researcher usingethnoscience techniques mightdesignsentenceframeswith the form "Quasimemberswork in offices." Becausequasi-members (Rohlen, 1974)do only supportwork suchas custo33

dianshipand cooking, the investigator wouldexpect that some word such as "never" would be elicited as a response to thatquestionframe.Question frames also can be used as tests of information. The researcher, for instance,couldvarythe abovequestion by stating, "Quasi-members sometimes do office work." The responsefrom the subjectwould be to correctthis misstatement andsupplythe properterm ''never" in place of "sometimes." This frame can be variedby substituting the termsmembers or nonmembers for quasi-members.The verb could be changedor the location for workcould be changed. Therearemanywaysin whicha sentence or substitution frame can be varied to elicit contrasting,but detailed,informationof a limitedculturaldomain. the reasonfor these procedures Remember, is to remain in the realmof emic data, and not to impose the researcher's etic categories on the data gathering process. For this reason, direct questions about in the bank organizationwould not be membership made. Theywouldrunthe riskof contaminating the resultswith etic categories. The techniquesdiscussedso far includeboth verbal and nonberbal elicitingprocedures. The different

Table 1 Basic Form of a ComponentialAnalysis Paradigm

Dimensions of Contrast Cultural Domain Cultural category Cultural category Cultural category I attribute, attribute, attribute, II attributed attributed attributed III attribute3 attribute3 attribute3

varietiesof sorting are nonverbaltechniquesthat often producecategoriesthat subjectsdid not previouslyacknowledge or realizeconsciously.Theyare sometimessurprised at the resultsand may havedifficulty giving verbal explanationsof their sorting decisions.Statistical suchas multidimentechniques sional scaling and other multivariate analysesmay be used to discern the patterns involved in the choices. Johnson (1978) and others (Kay, 1971; Pollnac & Hickman, 1975; Sanoff, 1971) provide useful guidancein applicationof statisticalanalysis to the data gatheredby the ethnosciencetechniques discussedabove.

Conclusions
Clearlythere is a wide varietyof uses for informationgatheredin organizations from an emic perspectiveand ethnoscience techniques. Suchdataprovidea concretebeginning for answering the concerns of researchers interested in using a subjective/idiographic/qualitative/insider approach.The benefit hoped for from such research,as indicated in the first part of this paper, is more knowledgeof subjectunderstandings, perceptions,cognitiveprocesses,meanings,and intentions.Thisis precisely the kind of informationthat the emic perspectiveand ethnosciencetechniquescan produce.Second, such information couldbe usedto developbetterquestionnaireinstruments designedto tap the subject'sperspectivemore closely than in the past. This would be particularlyhelpful in cross-culturalresearch. Preliminary intensiveworkwith individual or small, manageablegroups of employeescould be used as a baseto derivegeneral listsof domains,taxonomies, andso on to testmoreeconomically (in termsof time, effort, and money) than is possible with larger numbersof employeesof the same generalcategory withinthe organization. In-depth interviewing would not have to proceedfrom the beginningwith every
34

organizational member. Perhaps instruments could be developed with even broader applicability across organizations depending on the specific research questions of the study. An emic perspective and ethnoscience techniques also can profitably address a number of content issues in organizational research. Some examples would be: 1. Comparison of manager andsubordinate ideason any numberof dimensionsof interest, such as commitmentor satisfaction.Includedwould be ratingsand rankingson these dimensions. 2. Comparison of actualobserved behaviorwith the verbalstatementsabout behavior. 3. Investigation of the job designandcharacteristics or organizational structure in relation dimensions to actualemployee-relevant dimensions.Included would be ranking of all aspects of these dimensions. The resulting categories, perceptions, rankings, and so on from these content areas are as potentially "countable," and thus quantifiable, as any other kind of data. The richness of the patterns discerned may be clarified/expanded by factor analysis or other multivariate statistical techniques. For example, paired results of responses of managers and subordinates could be correlated and analyzed both quantitatively and qualitatively. The precision of quantification would be balanced by the context and richness of the qualitative data. Ideally, the emic perspective and ethnographic techniques should be applied to a total organization study in all of its rich complexity. This, of course, is not realistic for most researchers. The next best strategy is the more limited, but still valid, application of the ethnoscience techniques presented in this paper that can produce, at minimum, a number of emically derived variables for further research. The perspective and techniques described here provide a way to gain a holistic view of smaller, specific cultural scenes of interest to the organizational researcher.

In summary, there seem to be four especially important advantages in taking an emic perspective and using ethnoscience techniques in organizational research: the datacan is subjective, 1. Althoughthis approach be objectified(translatedinto etic categories). but has nomothetic 2. Thisapproachis idiographic, potential. data dependsmostlyon qualitative 3. Thisapproach for datasuitable quantifiable but also can produce traditionalstatisticalanalysistechniques. is aimedat the insider,the 4. Althoughthis approach outsiderlogically enters the researchprocess. The intent of this paper is not to persuade organizational researchers to abandon their current

and immediately begintakingan emic methodologies perspectiveand using ethnosciencetechniques.Inand set stead, the intent is to providea perspective of techniquesin responseto the call for new methodologies, yet not widen the gap between subjective/idiographic/qualitative/insider and objecto approaches tive/nomothetic/quantitative/outsider research. Althoughthe emic perspecorganizational obviouslyfall in the techniques tive and ethnoscience first methodologicalset, they also seem to be able to serveas a bridgeto the moretraditionalmethodological set.

References
Burrell, G., & Morgan, G. Sociological paradigms and organizational analysis:Elements of the sociology of corporate life. London: Heinemann, 1979. Burton, M. L. Semantic dimensions of occupation names. In A. K. Romney, R. N. Shepard, & S. B. Nerlove (Eds.), Multidimensional scaling: Theory and applications in the behavioral sciences. New York: Seminar Press, 1972, 55-73. Bussom, R. S., Larson, L. L., & Vicars, W. M. Unstructured, nonparticipant observation and the study of leaders' interpersonal contacts. In J. G. Hunt, U. Sekaran, & C. Schriesheim (Eds.), Leadership: Beyond establishment views. Carbondale, Ill.: Southern Illinois University Press, 1981, 31-49. Dubin, R. Management:Meanings, methods, and moxie. Academy of Management Review, 1982, 7, 372-379. Evered, R., & Louis, M. R. Alternative perspectives in the organizational sciences: "Inquiry from the inside" and "inquiry from the outside." Academy of Management Review, 1981, 6, 385-395. Frake, C. 0. Cultural ecology and ethnography. In A. S. Dil (Ed.), Language and cultural description: Essays by Charles 0. Frake. Stanford, Cal.: Stanford University Press, 1980, 18-25. Hackman, J. R. Preface. In J. P. Campell, R. L. Daft, & C. L. Hulin, What to study: Generatingand developing researchquestions. Beverly Hills, Cal.: Sage Publications, 1982. Harris, M. Cultural materialism: The struggle for a science of culture. New York: Random House, 1979. Hunter, D. E., & Foley, M. B. Doing anthropology: A studentcentered approach to cultural anthropology. New York: Harper & Row, 1976. Johnson, A. Quantification in cultural anthropology. Stanford, Cal.: Stanford University Press, 1978. Kay, P. (Ed.), Explorations in mathematical anthropology. Cambridge, Mass.: M.I.T. Press, 1971. Kotter, J. P. The general managers. New York: Free Press, 1982. Lammers, C. J. Towards the internationalization of the organization sciences. In G. Hofstede & M. S. Kassem (Eds.), Euro-

to organization theory.Assen/Amsterdam, pean contributions

Holland: Van Gorcum, 1975, 25-42. Luthans, F., & Davis, T. R. V. An idiographic approach to organizational behavior research: The use of single case experimental designs and direct measures. Academy of Management Review, 1982, 7, 380-391.

Thestructure of inquiry. research: Pelto, P. J. Anthropological

New York: Harper & Row, 1970. Pfeffer, J. Management as symbolic action: The creation and maintenance of organizational paradigms. In L. L. Cummings & B. M. Staw (Eds.), Research in organizational behavior (Vol. 3). Greenwich, Conn.: JAI Press, 1981, 1-52. Pollnac, R. B. Intra-culturalvariabilityin the structureof the subjective color lexicon in Buganda. American Ethnologist, 1975, 2, 89-109. Pollnac, R. B., & Hickman, H. M. Abduction and statistical inference of interaction patterns: An analysis of data from Peru, Uganda, and Iron Age France. Sociologus, 1975, 25, 28-61.

Rohlen,T. P. For harmonyand strength:Japanesewhite-collar in anthropological perspective.Berkeley,Cal.: organization

University of California Press, 1974. Sanoff, G. Quantitative analysis of sharing and variability in a cognitive model. Ethnology, 1971, 10, 389-408. Schriesheim, C. A., & Kerr, S. Theories and measures of leadership: A critical appraisal of current and future directions. In J. G. Hunt & L. L. Larson (Eds.), Leadership: The cutting edge. Carbondale, Ill.: Southern Illinois University Press, 1977, 9-45. Schriesheim, C. A., Bannister, B. D., & Money, W. H. Psychometric properties of the LPC scale: An extension of Rice's review. Academy of Management Review, 1979, 4, 287-290. Silverman, S. F. An ethnographic approach to social stratification: Prestige in a central Italian community. American Anthropologist, 1966, 68, 899-921.

35

interview.New York: Holt, Spradley, J. P. The ethnographic Rinehart & Winston, 1979.
New York: Holt, Rinehart, Spradley, J. P. Participant observation. & Winston, 1980. Spradley, J. P., & McCurdy, D. W. The culturalexperience: in complexsociety. Chicago: Science Research Ethnography Associates, Inc., 1972. Sturtevant, W. C. Studies in ethnoscience. In A. K. Romney & R. G. D'Andrade (Eds.), Transcultural studies in cognition. American Anthropologist Special Publications, 1964, 99-131.

Van Maanen, J., Dabbs, J. M., Jr., & Faulkner, R. R. Varieties of qualitative research. Beverly Hills, Cal.: Sage Publications, 1982. Weick, K. E. Cognitive processes in organizations. In L. L. Cumin organizational mings & B. M. Staw (Eds.), Research behavior (Vol. 1). Greenwich, Conn.: JAI Press, 1979, 41-74.

Nancy C. Morey has a Ph.D. in Anthropology from the University of Utah and is currently a doctoral candidate in management in the College of Business Administration, University of Nebraska-Lincoln. Fred Luthans is George Holmes Professor of Management in the College of Business Administration, University of Nebraska-Lincoln.

36