Zripadoceskri . .

< UmVerzlta v Plzni
Fnkultn ... kat . 1w .~l~llltnJ(~h RtudH edra socwlni a kulturni

n Ii


Handbook of Methods in




Cultural Anthropology
Upadoee.k8 unlv
Fakulta hum

.el"Zita v Pmu

Katedra Boct...:l

Tylova 18, 301 2" antropologie tel 5 Plzeti .; 019 / 732 24 66

amtnfch studif a kult



Handbook of Methods in Cultural Anthropology
EDITOR: H. Russell Bernard, University of Florida EDITORIAL BOARD Carol R. Ember, Human Relations Area Files Michael Herzfeld, Harvard University lane H. Hill, University ojArizona Roy A. Rappaport, deceased Nancy Scheper-Hughes, University oj California, Berkeley Thomas Schweizer, University ojCologne

Hanclli~~ 9t~4t1Iods)in




l l l l

l II


H. Russell Bemard editor


A Division of Sage Publications, Inc.
Walnut Creek. London. New De/hi

l l l


Copyright © 1998 by.Alta.Mirarress









All rights reseered. No part of this'book may be reproduced or utilized in any form by any means, electronic or mechanical, including photocopying, recording, or by any information storage and retrieval system, without permission in writing from the publisher.
For information address: AltaMira Press A Division of Sage Publications, Inc.
1630 North Main Street, Suite 367

H. RUSSELL BERNARD I. Introduction: On Method and Methods in Anthropology tt. RUSSELL BERNARD


Walnut Creek, California 94596 USA explore@altamira.sagepub.com http://www.altamirapress.com Sage Publications, Ltd. 6 Bonhill Street
London, EC2A 4PU


Part I: Perspectives
2. Epistemology: The Nature and Validation of
Anthropological Knowledge


United Kingdom Sage Publications India Pvt. Ltd.
M-32 Mnrket Greater Kuilash I New Delhi 100 048

3. In Search of Meaningful Methods lAMES FERNANDEZ and M1CHAEL HERZFELD


4. Research Design and Research Strategies

173 203 235




5. Ethics

6. Feminist Methods
Handbook of methods in cultural anthropology I H. Russell Bernard, editor. p. cm. Includes bibliographical references and index.
ISBN 0-7619-9151-4 (cloth) I. Ethnology-Methodology, I. H. Russell (Harvey Russell), 1940GN345 .H37 1998 305.8'OOI-ddc21 CHRISTtNE WARD GAlLEY

7. TransnationaI Research

Part II: Acquiring Information
8. Participant Observation KATHLEEN M. DEWALT and BILLlE R. DEWALT, with CORAL B. WAYLAND 259

98-25423 CIP

9. Direct Systematic Observation of Behavior



10. Person-Centered Interviewing and Observation
ROBERT I. LEVY and DOUGLAS W. HOLLAN 98 99 00 01 02 03 04 7 6 5 4 3 2 I


I I. Structured Interviewing and Questionnaire Construction


Editorial Production: Carole Bernard/ECS Cover Design: Joanna Ebenstein

12. Discourse-Centered Methods


sSJ_~:~1 (:::-/1

13. From Pictorializing to Visual Anthropology
FADWA EL GUlNDI 14. Fieldwork in the Archives: Merhods and Sources in Historical Anthropology CAROLlNE B. BRETTELL

459 513

Part III: Interpreting Information
15. Reasoning with Numbers W. PENN HANDWERKER and STEPHEN P. BORGATTt 16. Text Analysis: Qualitative and Quantitative Methods H. RUSSELL BERNARD and GERY W. RYAN 17. Cross-Cultural Research CAROL R. EMBER and MELVlN EMBER
549 595






Part IV: Applying and Presenting Anthropology
18. Methods in Applied Anthropology ROBERT T. TROTfER. 11 and JEAN J. SCHENSUL 19. Presenting Anthropology to Diverse Audiences CONRAD PHILLlP KOTTAK
About the Authors Author Index Subject Index 691 Carole Bernard and I lived in Cologne, Germany, for a year in 1994-1995. Thomas Schweizer, director of the Institute for Ethnology at the University of Cologne, had invited me to be an Alexander von Humboldt research scholar there. It was the sort of opportunity scholars dream about: the chance to do nothing but read, write, and reflect, and to do all this while interacting with graduate students and colleagues. No classes. No committee work. Paradise. This handbook is partly a product of that year. I have been interested in social research methods for as long as I can remember, so I took the opportunity in Cologne to read and read, and then read some more. My goal was to become better grounded in the range of methods used by scholars across the social sciences and to understand the role that anthropologists had played in the development of social research methods. The idea for this handbook emerged in conversations about all this bye-mail, with Mitch Alien, editor of AltaMira Press. It seemed like a good time to take stock. The last handbook, edited by Raoul Naroll and Ronald Cohen, had been published in 1970. The content of the discipline of cultural anthropology and the demography of the profession had gone through big changes since then. In 1970, most anthropologists went into academic jobs. Today, most are in nonacademic jobs. Fewer graduate students do fieldwork in small, isolated communities now. They couldn't, even if they wanted to, for such communities are an endangered social species. In 1972, women received just 32% of the Ph.D. degrees in anthropology in the V.S. In 1995, women received 59% of the doctorates. There was also the resurgence of the great epistemology debate that has so long pervaded the social sciences. Each side claims support from an indisputable observation: On the one hand, people construct their own realities, and the process

737 763 775


l l l


, ,




is dynamic, ever-changing; on the other, there are regularities in human behavior and human thought. While rhetorical energy is spent arguing that (a) the first fact renders impossible the pursuit of the second or that (b) the second fact renders irrelevant our worrying about the first. working scholars of all persuasions are out there doing empirical research. The core of the discipline, it seemed to me, was in the fact that nearly all cultural anthropologists choose from the same awesomely large kit of tools. My goal, then, from the beginning has been to put together a handbook that would be useful to academic anthropologists and practicing anthropologists; to interpretivists and positivists; to idealists and materialists. No project of this magnitude can be managed alone. Six colleagues graciously agreed to join this project and serve as a board of editors: Carol Ember (HRAF), Michael Herzfeld (Harvard), Jane Hill (Arizona), Roy ("Skip") Rappaport (Michigan; deceased), Nancy Scheper-Hughes (UC-Berkeley), and Thomas Schweizer (Cologne). When I thought about senior people whose work was respected by colleagues across the field, Rappaport's name came immediately to mind. Tragically, he didn't live to see the end of the project. Right from the beginning, the members of the editorial board contributed ideas about chapters that needed to be included in the handbook and about who might write those chapters. They read the chapters and offered critical advice and support. Three of them (Ember, Herzfeld, and Schweizer) contributed chapters themselves. I am grateful to all. Over the years, I have come to expect nothing less than the best of editorial guidance from Mitch Alien. He never disappoints, never holds back, never pulls punches. I am also grateful to the following colleagues (in alphabetical order) who read various chapters ofthe handbook in draft and provided detailed reviews: Devon Brewer, Karen Brodkin, Edward Brnner, Douglas Caulkins, Garry Chick, Victor De Munck, William Dressier, Dama Dufour, Robert V. Kemper, David Kertzer, Maxine Margolis, and Alvin Wolfe. Special thanks go to Ronald Cohen, my colleague at the University of Florida. His was a pioneering effort in 1970 when he and Naroll put together that first handbook of methods in cultural anthropology. My thanks also go to the Alexander von Humboldt Stiftung, Bonn, and to the College of Liberal Arts and Sciences at the University of Florida for support during 1994-95. I know that I cannot thank sufficiently my partner, Carole Bemard, for her support all along the way and specifically for her work copyediting and producing the final product. But I can try.
H. Russell Bemard Gainesville, FL July 20, 1998

On Method and Methods in Anthropology

This introduction has two parts. In Part I, I offer some remarks about the history and scope of methods in cultural anthropology. Later, in Part 2, I describe in some detail what is in the various chapters.

Part I: On Methods in Anthropology
Method is about choice-the choice of taking a verstehen or a positivist approach; the choice of collecting data by participant observation or in the archives, by direct observation or by interviewing; the choice of making quantitative measurements or collecting oral, written, or visual text. The authors in this handbook deal with all these choices, and more. I will have a lot more to say later about the content of the chapters in this handbook. In this first part of the introduction, though, I want to make clear why this book is important-for all social scientists, not just for anthropologists. Conventional wisdom notwithstanding, anthropology has always been about methods, from the earliest days of the discipline right up to the present. Anthropologists have been prodigious consumers and adapters of research methods, and they have made important contributions to the big social science tool kit as well. I am going to document this and put it in perspective here. There has always been a certain tension between those who would make anthropology a quantitativ~ science and those whose goal it is to produce documents







that con~ey the richness-i~d~ed,. the uniqueness-of human thought and experience. Enc Wolf captured this In his wonderful aphorism that "Anthropology is the

(and last) edition was published in 1951 and was reprinled five times until 197I.
That final edition was edited by Brenda Seligmann and "a committee of the Royal

most humanistic of the sciences and the most scientific of the humanities" (1964:88). Students of cultural anthropology may be asked early in their training to take a stand for interpretivism or positivism, humanism or science, qualitative or
quantitative research. Readers of this handbook will find no support for this polarized vision of

Anthropological Institute" Ihat included Ihe likes of E. E. Evans-Pritchard, Daryl Forde, Raymond Firth, Meyer Fortes, and W. E. Le Gros Clark-and is must
reading for anyone interested in learning about field methods. Strip away the quaint language and the vestiges of colonialism-ua sporting rifle and a shotgun are . .. of great assistance in many districts where the natives may welcome extra meat in

method. Instead, they will find scholars laying out the methods they use in
practlcmg their craft-a craft rooted, for every author in this book, in one of the most essentially empirical traditions in all of science: participant observation fieldwork. Some authors are identified with interpretivist methods, some with quantitative methods for the collection and analysis of data but none dismiss humanism or science and none ask their readers to choose once 'and for all between expressing their findings in words or in numbers.

the shape of game killed by their visitor" (RAI 1951 :29~and the book is still full of useful, late-model advice about how to conduct a census, how to handle
photographic negatives in the field, what questions to ask about sexual orientation, infanticide, food production, warfare, art .. . The book is just a treasure. In the 19205, leading sociologists were concerned with moving their discipline away from an emphasis on social refonn-away from the study of what ought to be and toward the study of what is. If the public were ever to trust social science, said Carl Taylor, then the emphasis had to be on "exact and quantitative expressions

l l l l

Romancing the Methods
John Whiting and some of his fellow graduate students at Yale during the 1930s
asked abo~t .havl2g a se~mar on m~thods. "Leslie Spier infonned us disdainfully,"

and measurements" (Taylor 1920:735). This, he said, required "technologies which will reduce observations to a comparative basis" (p. 753). The technology of choice, said Taylor, was the social survey, a method dating at least to John Howard's monumental, comparative study of prisons (1792). Taylor's idea of what a survey should be was much broader than just questionnaires. "The survey method," he said, "is nothing whatever but the recognized and accepted comparative method of all science," and he concluded that uwhat . . .

recalls Whiting, that this was a subject to discuss casually at breakfast and was not worthy subject matter for a seminar" (Whiting 1982: 156). Try quoting Whiting at a ~onve.ntlo~ of anthr?pologlstS. Chances are, you will discover that everyone
chimes In with a favonte story of the same ilk. It's all well and good for anthropologists to romanticize fieldwork-vulcanologists do it, too-particularly for ~eldwork in places that take several days to get to, where t~e locall~ngu~ge has no htera~. traditi~n, and where the chances of coming down with a senous Illness are nontnvtal. Social research really is harder to do in

some places than in others. But the fact is, there is a long, noble tradition of
concern with research methods in anthropology---quantitative and nonquantitative humanistic and scientific. '

surveys can do and have done in the field of anthropology and ethnology, they can do and probably are destined to do for any body of knowledge or field of research to which they are applied (1920:752-753). Taylor singled out the systematic study of vision, hearing, and pain that Charles and Brenda Seligmann (191 I) had done on the Veddas of Sri Lanka. (Charles Seligmann was an ethnologist and physician.) Their 422-page ethnographic account covered family life, religion, the arts, property, and inheritance-and an 18-page report of the results of some psychological tests that they had used in their study of Vedda senses. Some of those tests had been devised by W.H.R. Rivers, an
experimental psychologist who became interested in anthropology in 1899 when he was invited to join the Torres Straits expedition and saw the opportunity to do

Kathleen and Billie DeWalt quote at length in Chapter 8 what is surely one of t~e most-cited early discussIOns of methods in anthropology: Malinowski's introductIOn to Argonauts of the Western Pacific (1922). It is justly famous because as the DeWalts say, it established the importance of long-tenn participant observ;tion as
a strategic method for field research on other cultures. No peering off the veranda at the natives for Malinowski.

Participaut obsen:ation is an important method in anthropology but, as the DeWalts POlOt out, It IS one. of many methods used in fieldwork. By the time Maltnowskl weut to the Trobnands, Notes and Queries on Anthropology-the fieldw~rk manual produced by the Royal Anthropological Institute (RAI) of Great Bntam and Ireland-was 10 its fourth edition (the first came out in 1874). The sixth

comparative studies of non-Western people (Tooker 1997:xiv). Rivers, of course, developed the genealogical method-highly detailed, egocentered graphs for organizing kinship data. The genealogical tables he produced in his study of the Murray Islanders were singled out by Taylor as an example of "as perfect a scientific compilation as could well be imagined" (Taylor 1920:753. See Rivers's work in Volume VI of Haddon 1901-1935). These works by anthropologists, said Taylor, were examples of research to which all social science could

l l l l l

aspire. Anthropologists continue to this day to work on improved methods for representing complex kinship structures (see White and Jorion 1992).






Rivers continued his work in anthropology and his development of the genealogical method with his research on the Todas of India (Rivers 1906). He developed what he called the "method of indirect corroboration." This method involves "obtaining the same infonnation first in an abstract fann and then by means of a number of concrete instances" (p. 11). Here is Rivers explaining the method with reference to his study of the laws of property inheritance:
I first obtained an account of what was done in the abstract-of the laws governing

pelto (1970; and Pelto and Pelto 1978), by John Brim and David Spain (1974), by AlIen Johnson (1978) and by Michael Agar (1996 [1980]) contain a wealth of information and wisdom. In writing my own textbook (Bernard 1994 [1988]) I depended heavily on the work of all these predecessors and on the work of

colleagues across the social sciences.
Methods, in fact, are us, and the book you are reading-this 1998 handbookreplaces nothing. It adds to a growing body of work about methods of inquiry in

the inheritance of houses, the division of the buffaloes and other property among the children, etc. Next I gave a number of hypothetical concrete instances; I took cases of men with so many children and so many buffaloes, and repeating the cases I
found that my informant gave answers which were consistent not only with one another but also with the abstract regulations previously given. Finally I took real

cultural anthropology and to methods in the social sciences in general. That is one of the really nice things about research methods: There are more and more of them as time goes on.

persons and inquired into what had actually happened when A or B died, and again obtained a body of information consistent in itself and agreeing with that already
obtained. (p. It)

Methods Belong to All of Us
Another nice thing about methods is that disciplines cannot own them. Just as it was never true that only sociologists did surveys, it was never true that only anthro·
pologists did participant observation fieldwork. Of course, these days, everybody knows that everybody does everything. Methods really do belong to everyone and this does not just go for methods-as-lechniques.1t goes just as much for methodsas~approach,methods-as~commitment, and methods-as-epistemology. Neither quan· titative nor qualitative researchers have the exclusive right to strive for objectivity; neither humanists nor scientists have a patent on compassion; and empiricism is as much the legacy of interpretivists and idealists as it is of positivists and materialists. Of course, new methods develop within particular disciplines, but any method that seems useful will get picked up and tried out, sooner or later, across the disciplines. Projective tests, like the Rorschach and the TAT, were used by

Rivers discussed his selection of informants, how he came to know that one
of his informants had lied to him, the pros and cons of paying informants for

their time, the need for getting information from many informants rather than just from a trusted few, and the importance of using the native language in field research. Taylor must also have known about Lewis Henry Morgan's study, Systems of Consanguinity and Affinity of the Human Family (1870). It was a massive, crosscultural survey of kinship systems. Morgan collected a lot of the data on various

Indian tribes himself, but he also sent questionnaires to missionaries and Indian
agents. And Taylor surely also knew about Edward Burnett Tylor's key contribution to the literature on cross-cultural surveys (see Tylor 1889). Contrary to popular wisdom, then, anthropologists have been keen survey methodologists from the earliest days ofthe discipline. Unlike sociologists, however, anthropologists studied small, remote groups of people. "These groups," observed

Robert Lynd in 1939, "were 'primitive,' according to Western European standards, and therefore the older social sciences did not much care what anthropology did
with them" (p. 14). The point is, that by the time A Handbook of Method in Cultural Anthropology was published in 1970, the concern for methods in anthropology was already quite venerable. That volume, edited by Raoul Naroll and Ronald Cohen, was an enormous compilation-I,OOO pages and 49 chapters by 46 authors (Naroll and Cohen wrote 6 of the chapters and participated in several others), including 5 that

were reprinted from journal articles. The chapters in that handbook, as well as
all the chapters on methods in Anthropology Today (Kroeber 1953), and in the Handbook of Social and Cultural Anthropology (Honigmann 1973), are as useful today as they were when they first appeared. The pioneering textbooks by Pertti

anthropologists in the 1940s (Du Bois 1944; Gladwin and Sarason 1953; Henry and Spiro 1953), were quite popular for a while (see Lindzey [1961] and Hsu [1972J for reviews) and continue to be used by some anthropologists to this day (see Paul 1989; Tsatoulis-Bonnekessen 1993; Boyer at af. 1994). The consensus model of culture was developed in the 1980s by two anthropologists and a psychologist (Romney et af. 1986) and has already started showing up in psychology journals (Johnson et af. 1992; Van Raalte et af. 1992). A solution to the ecological inference problem (inferring individual behavior from aggregate data) was worked out recently by a political scientist (King 1997) and will have wide application in sociology, criminology, and demography. Field experiments were developed by social and educational psychologists (see lhe classic volume by Cook and CampbeIl [1979] and Boruch [1997] for a review of more recent work) but some anthropologists also do field experiments (see, for example, Harris et af. 1993).

And of course, participant observation-the sine qua non of anthropological
fieldwork-is no one's property. I've heard talk at anthropology conventions about





how the discipline somehow "lost" the method of participant observation. 1 could not disagree more. Anthropologists continue to make consistent use of participant observation fieldwork, but we did not invent that method by ourselves. Sociologist Beatrice Webb was doing participant observation-complete with real note taking and informant interviewing-in the 1880s and she wrote at length about the method in her 1926 memoire (Webb 1926). Just about then, the long tradition in sociology of urban ethnography began at the University of Chicago under the direction of Robert Park (one of the many participant observers trained by Park was his son-inlaw, Robert Redfield). That tradition continues today in the pages of the Journal 0/ Contemporary Ethnography, which began in 1972 under the title Urban Life and Culture. Participant observation today is absolutely ubiquitous in the social sciences. It has been used in recent years by political scientists (Fenno 1990; Glaser 1996), social psychologists (Weisfield and de OIivares 1992; Smith and Inder 1993), psychoanalysts (Perry 1985; Hirsch 1990; Hegeman 1995), students of management (Gummeson 1991; Weick 1995; Watson 1996), researchers in nursing (De Valck and Van de Woestijne 1996; Woodgate and Kristjanson 1996), education (Woods 1986; Rovegno and Bandhauer 1997), social work (Lawler and Heam 1997) and expert systems engineering (Meyer 1992), as well as by legions of sociologists (see Denzin and Lincoln 1994). Among the salutary results of all this is a continually growing body ofliterature, including a lot by anthropologists, about participant observation itself. There are highly focused studies, fun of practical advice. and there are poignant discussions of the overall experience of fieldwork (Agar 1996, Wolcott 1995, Smith and' Komblum 1996). ) The boundaries between the disciplines remain strong, but those boundaries are no longer about methods-if they ever were-or even about content. Anthr~' pologists today are more likely to study anny platoons (KilIworth 1997) or con; sumer behavior (Sherry 1995) or the mean streets of big cities (Bourgois 1995i Fleisher 1998) than they are to study isolated tribal peoples. Today, the differenc~~,. within anthropology and sociology with regard to methods are more Important th~ the differences between the social sciences. There is an irreducible difference. f~r example, between those of us for whom the first principle of inquiry is that re~U,. is constructed uniquely by each person and those of us who start from the pnnclpla that external reality awaits our discovery through a series of approximations. Th~ is also an important (though not incompatible) difference between those of us w seek to understand human phenomena in relation to differences in beliefs and v~.~. and those of us who seek to explain human thought and behavior as the .~ sequence of external forces. ' Whatever our epistemological differences, however, the actual methods by '!!. we collect and analyze our data belong to everyone across the social sciencas

Anthropology: The Humanistic, Positivistic, Interpretive Science, Etc.
Anthr.opology was built by empiricists, including many who understood that reducmg people to words was no better than reducing them to numbers and who did both with aplomb. Franz Boas took his Ph.D. in physics from the University of KIel, wIth mathematIcs and geography as his minors. When he turned his attention to anthropology, he advocated a historical, particularizing approach, but he did not abandon numbers. Nor did most of his students. Alfred Kroeher, for example, analyzed measurements on 300 years ofdata on skirt length, waist height, and depth of decolletage to look for long-tenn patterns in style (Richardson and Kroeber 1940). Robert Lowie, Clark Wissler, Edward Sapir, Margaret Mead, Ruth Benedict, and E1SIe Clews Parsons conSIdered themselves scientists. One of Boas's students Paul Radio, :amously r~jecte.d his mentor's scientific bent. He accused Boas of being naturwtssenscha/tlzch. emgestellt, or science minded, of treating ethnology as a branch of natural sCl~nce and named his cofreres-Sapir, Kroeber, Mead-as examples of the bad thmgs that happen to cultural anthropologists who follow the ; path of quantification (Radin 1933: 10). ;:;(, But even Radio was an arch empiricist. His passion against quantification was ii):surpas~ed by his cOffi",litment to the· continual collection of texts. It was Radio who ade It clear for all tune that, while theories come and go original texts in Our Informants' 0 n la '1 b ' , '~", , W I nguages. ~re aval a le for every generation to analyze anew, • ptred by Boas s efforts WIth George Hunt (a Kwakiutl) Radin handed Sam 10Wk(W'b ' produced one of ..;-. sua e a mne ago) a pad and a pencil-an act that jlhroPOlogy'S most famous texts, Crashing Thunder: The Autobiography of a ,nnebago lndian--and establIshed a model for emically oriented ethnography low,nake 1920).'



l l l l' l


':~bove all, the distinction between quantitative and nonquantitative must never ",¥Sed as COver for talking about the di fference between science and humanism " of scientists do their work without numbers; and many scientists whose work Ighly " . t' quantitative consIder themse Ives to be humanists. Humanism is often used • synonym ~ h .. . ". . or umaOltanan or compassIOnate values and a commitment to the horati.o~ of suffering. The myth that science is the absence of these values is "pernIcIOUS. Jane Goodall said recently that part of her work these days is not ~~h chimps, but to correct the mistaken idea that science has to be lonat "I ft e. am 0 en asked to talk about the softer kind of science" she says way ?f bringing children back into realizing that [science] i; not abou; g thtn~s up and being totally objective and cold" (Holloway 1997:44). !pust reject a culture that equates objectivity with being cold. Counting the . one way-not the only way-to preserve outrage. We rately in Rwand .~. , a IS ~ , not less, SCIence, lots and lots more, and more humanistically infonned

1 l l


in Sidney and Beatrice Webb's attack on the British medical system (1910).16 BERNARD Introduction 17 I i . all social scientists can be humanists by any or all of these definitions. sexism. all experience is surely unique. of Adolphe Quetelet. And this growth wasn'. . is not the positivism of Auguste Comte. "it will be necessary to crush out emotion. This kind of rhetoric just begged to be reviled. "It is certainly desirable to be precise. tedious. history. then. It is a method used by idealists and materialists alike. Can we. record what others experience? Yes. the Belgian astronomer whose study of demography and criminology carried the audacious title Social Physics (1969[1835]). But just as surely. The practice that we associate today with the positivist perspective in social science. but it's much more fun to be both. just for lOath and the natural sciences. None of these is the tradition that so many today love to hate. as Oberschall (1972:244) concluded. The legacy of positivism as a vehicle for social activism is clear in Jane Addams's work with destitute immigrants at Chicago's Hull House (1926). whether or not they are done in service to explanation or understanding. just 22 Ph. Historically. the need to measure carefully and the need to listen hard. values. of course. "but it is quite as needful to be precise about something worth knowing. Giving birth. by 1975. the tension between science and humanism is wrought by the need to answer practical questions with evidence and the need to understand ourselves-that is. the adoption by anthropologists of the full range of social research methods has accelerated since 1970. Wiiliam F. advocating full equality for women. a winning or losing struggle with iIIness-or to write someone else's story for them. degrees were awarded by U. surviving hand-to-hand combat. but. auditorily. In turning sociology into a science.. he said." said Robert Redfield (1948: 148). It is a truism that one cannot fully know someone else's life without living that person's life. of John Stuart Mill. but the humanistic component forces us to ask if we are pursuing worthwhile ends and doing so with worthwhile means. For some social research'''s these pitfalls are evidence for natural limits to social science. In Charles Booth's account of the conditions under which the poor lived in London (1902). Humanism sometimes means an abiding appreciation of and search for the unique-the unique in human experience and the unique in culture. and routine tasks" (1930: I0). then anthropology was the search for exceptions. while for others they are a challenge to extend the current limits by improving measurement. The Subjection of Women (1869) by John Stuart Mill was a radical work in its day. Adolphe Quetelet.S.D. "it will be desirable to taboo ethics and values (except in choosing problems). When insight and understanding are achieved. There were challenges to Ogburn's prescription. The fact that knowledge is tentative is something we all learn to live with. Permanent Methodological Eclecticism While anthropology has always been an eclectic discipline with regard to methods. The quality of the recording. produced billions of dollars for expansion ofU. and beliefs to achieve insight into the nature of human experience.' The central position of positivism as a philosophy of knowledge is that experience is the foundation of knowledge.~ L_ Humanism is sometimes pitted against positivism. to identify ourselves as humanists or as positivists. to using our own feelings. It is not something apart from social science. and it will be inevitable that we shall have to spend most of our time doing hard. was a committed social reformer. Are there pitfalls in doing so? Yes." We are all free." Further." In his 1929 presidential address to the American Sociological Society.I . That honor belongs to what Christopher Bryant (1985:137) calls "instrumental positivism. to nomothetic or idiographic goals. with its MAD (mutually assured destruction) strategy. and . The arms race that followed. and emotionally. ~- i L_ I. of course we can. of course there are. nor is it even the logical positivism of the Vienna Circle. Eventually. however. To write a story about the thrill or the pain of a successful or failed border crossing. most did not. . as an ethnographer might do--these are not activities opposedto a natural science of experience. which is said to be linked to support for whatever power relations happen to be in place. becomes the key to knowledge.S. This was not a random even!: In 1970. it was Ogburn's commitment to value-free science and to statistics that won the day. This kind of humanism is. Combined departments of sociology and anthropology separated. and the social sciences. living with AIDS . dull. he said. All those engineers and physicists needed courses in English. in some way. A generation earlier. universities in anthropology. in fact. in 1950. the foundation of many clinical disciplines as well as the foundation of participant observation ethnography (see Berg and Smith [1985] for a review of clinical methods in social research). . Whether they rally to the idealist or the materialist flag.. we may call it knowledge. though. 6). They are the activities of a natural science of experience. the Soviet Union launched the first orbiting satellite-Sputnik. In the end. Humanism sometimes means a commitment to subjectivity-that is. there would be no need for a separate field of statistics because "all sociologists will be statisticians" (p. In 1957. ethnic nationalism-in the world. there are commonalities of experience. The scientific component of anthropology demands that we ask whether our measurements are meaningful. and in Florence Nightingale's (1871) assessment of death rates in maternity hospitals. a successful or failed job hunt. If the rest of social science was the search for regularities. science. universities. most anthropologists went into academic jobs. philosophy. Ogburn laid out the rules. of course. to contribute more to tbe amelioration of suffering and the weakening of false ideologies-racism. positivism was linked to the most critical stance. We record what we experience visually.

tracing the "centrifugal and centripetal tendencies" that emerge from anthropology's special location between the humanities and the sciences. all those assistant professors were in place and there were no new jobs in the academy (D'Andrade et a1. their bodies.6% in 1993.000 Ph. and eventually opened. it had recently separated from sociology. "the difference in kind between the presence of the hermeneutic circle in the humanities at one extreme and the lack of it in the natural sciences on the other becomes a mere difference in degree. 1989). the last time the NSF looked (Givens and Jablonski 1996). Not that all those Ph. The attraction of anthropology for new students.D. Readers should come away from this book with a sense of the richness in the methods available to all of us because of the diversity in perspective that we bring to the research craft. ~ut with all thos~ anthropologists from previous years still applying for academIC Jobs." Recently. aboul 11. 413 since 1974. Chapter 18 deals with the range of methods used in applied anthropology. degree. more anthropologists were empl~yed in no~academicjobs than in academe (see Fluehr-Lobban 1991:5). young anthropologists knocked on. our discipline is equally rooted in positiVism-in the "method of hypothesis testing as a general procedure for generating and validating scientific knowledge. With the expansion complete. in sociology. methods for the collection and analysis of visual data. It plays well for all of us because we can all resonate to a good story about how people experience their lives. says Schweizer.-granting departments were filled with assistant professors training th~usands of graduate students-and awarding over 400 Ph. their illnesses.D. not the sociologists psychologists. Durmg all the tough years of the 1970s and 1980s. and Chapter 19 is a discussion of methods for presenting anthropology to various audiences.D. "I I T Schweizer sorts through these epistemological alternatives-positivism. ThIS same thmg happ~ned. of course. Four years later. Program administrators were accustomed to hiring social scientists whose methods and products were known commodities. however. the continued move toward sophisticated methodological eclecticism is permanent. person-centered interviewing. as Schweizer makes clear. the result was a synthesis of qualitative and quantitative approaches-a synthesis that we see today coming into academe from many directions. through prose." In the end. and approaches for transnational research. In the absence of. In total. They rewrote their n. those who had participated in Ihe great 1960s expansion began to retire in the mid-1980s. ethics. say.s were out of work. structured interviewing. degrees per year (Glvens and Jablonski 1996). From 1985-1994.18 BERNARD Introduction 19 Ph. a few thousand new academic jobs for anthropologists. in developing ecotourism -the list is as big as the hunger and imagination of the pioneers themselves.sum~s an~ competed suc~essfully for jobs in medical schools and other parts of the uOlverslty. T~e expansion was irrepressible. . Chapters 2-7 cover epistemology.~ . but that was just the beginning. their children. But ~ost lmpo~antly.D. two approaches-postmodemism and radical constructivism-have developed as altematives to both classical positivism and classical hermeneutics. they were forced to be more explicit about their methods.-granting anthropology programs proliferated.s had Jumped to a heady 195. research design. The jobless rate for anthropologists was 1. and economists." . in rapid rural assessment for health care delivery programs. and methods of historical anthropology. m training executives for overseas assignments. as Schweizerpoints out. but the effects were quite dIfferent. By 1986. feminist approaches. all those new Ph. direct observation. specializing m l~tematlOnal market research.D. ' By and large. They Jomed consult~ng firms and opened their own firms. and posttnodemism and radical constructivism both adopt a relativistic stance that stresses the creative power of scientists to invent 'reality'. 1975). Postmodemism. of that reasoning is as strong as ever in the humanistically oriented segment of our discipline. Part 2: What's in the Handbook? This handbook has 19 chapters by a total of 27 active researchers. methods for the systematic analysis of text. the department awarded its first Ph. It was the anthropologists who had to make the adjustment. Chapters 15-17 describe methods of data analysis including numerical methods. the number of anthropology Ph. an average of 331 academicjobs were listed in the Anthropology Newsletter(Givens and Jablonski 1996). Thomas Schweizer sets the tone for the book in Chapter 2. The method of empathic reasoning and the method of display. By 1970.. Chapters 8-14 cover specific strategies and techniques of data collection including participant observation. . the opeOlngs were (and remain) far fewer than the number of available candidates. anthropologists wer~ produced from 1970-1998 (Givens and Jablonski 1996). D. methods for the collection and analysis of natural discourse.The number of Ph. As anthropologIsts competed for nonacademic jobs. says Schweizer. for the first time in the history of the discipline. On the other hand.D. and methods for testing cross-cultural hypotheses. When I began graduate study at the University of Illinois in 1961. many doors (Koons et a1. degrees awarded annually by institutions that participate m the survey of the American Anthropological Association has remained on average. By 1974. they competed for jobs in industry and ?o~emme~t. Demography being more-or-Iess destiny. methods in the search for meaning. remained undimi~i~hed .D. "questions systematic approachesto the production ofcumulative knowledge (which radical constructivism does not reject).

The strength of this approach. has videotaped interactions between artisans and their apprentices. "Anthropology. The more secure metho." says Johnson. strategies. Each approach has its advantages and disadvantages.fonnants. but the research is not conducted clandestinely nor are the results secret. anthropologists advocated the scientific method in ethnographic research. accurate. Carolyn Fluehr-Lobban seeks to make ethical considerations "an ordinary. but a determinedly empirical insistence on the constant calibration of the methods we use to elicit data with our informants' understanding of what constitutes appropriate discourse. he ~ance? dance~ in all-night sessions." In other words. to the ethical debates during the Vietnam War era. while women tend to sympathize with th: ap~rentice. hermeneutics. We may how they can be analyzed." The World Bank. Instead. the differences the culture. describable social rhetoric. Jeffrey Johnson surveys the broad range of research designs and strategies in cultural anthropology. but selecting between them can be excruciating. In dance. out in expressive culture-in the architectonics of buildings and village layout. The search for convergences. says Fluehr-Lobban. just as buildings and games reflect features of our culture. in games. to the promulgation of the latest PPRthe Principles of Professional Responsibility-by the American Anthropological Association. This may have been because it was just too difficult to make reliable. I share their enthusiasm for taking part physically in the lives of those whom we would understand and their enthusiasm for making cultural anthropology the enterprise in which we all participate to understand and explain human experience and behavior." Johnson shows. and data analysis from the beginning. "shouldn't silence the real differences or lead to shallow. in dyadic interactions.. design (including sampling). in other words. acquired an agricultural plot among the Fang and apprenticed himself to some elders in order to understand agricultural practices. "t~ereby demonstrating the replication of ideologies of hierarchy in adjacent but discrete contexts of social life. and words to our in. Fernandez. Or it may have been the result of simply being tom between the Scylla of precision and the Charybdis of verisimilitude. for the ways that ~eaning plays. The men. and askmg them to comment on the interaction. In Chapter 4." She traces the history of ethical issues in the discipline. is "itself caught up in an identifiable. to operatlOnalize by bodily activity. from Franz Boas's public excoriation in 1919 of anthropologists who worked as intelligence agents for the American military. Among the Switi. we see that the history of our discipline is one of tension between two essentially different approaches to research design. continues the theme of building a bridge across the interpretivist-positivist divide. Claims to cultural knowledge through purely v~rbal operatIOns may be a black-box-opening operation of the most problematic kl~d.d of approximation is to have learned by apprenticeshIp how to do . smce there may be many ways to get anythiIlg right in a given culture." says Johnson. Herzfeld. is directly related to Fernandezand Herzfeld also advise us to take part in physical activities. with a series of concrete examples. by James Fernandez and Michael Herzfeld. objects. for example. and radical constructivism-pointing out where they converge and diverge. he says. Johnson stresses the importance of the ethnographic approach. then. The position taken by Fernandez and Herzfeld-what they call the "militant middle ground"-accepts neither the militantly positivistic ori~tation of some '~ The principles that guide ethical conduct in any research.20 BERNARD Introduction 2/ meaOlng. it should open up the possibility that there are overlapping concerns and rational procedures in each framework that can be cross-fertilized and even combined. Since 1971. so do our reactions to the actions of others reflect our deepest understandings of our culture. false compromises.. He lays out the varieties of experimental and nonexperimental research designs and discusses issues of sampling. one key question to ask about any . are openness and disclosure. "there is little evidence that they considered appropriate design issues when they actually did the research.g of activities. never get It all fight. Thus. postmodernism.. "to remember the connection between theory. between exploratory and explanatory research. but corporeal understanding is critical to understanding. for example. Like all the authors in this handbook. Fluehr-Lobban points out that "secret" is not the same as "Proprietary. and in work. "is its ability to incorporate a wide range of methods. and he followed workers into the mines. for example. "is often conducted in circumstances in which to talk a good game IS not r:alty to play it at all. This is what makes it possible for the people with whom we interact in our research to choose whether to provide us with information or not. social interaction. because how the data were collected." This chapter. The "fashionable distinction between positivistic and interpretive styles of analysis." say Fernandez and Herzfeld.S." they say." Learn. all combining in ways to improve the chances for credible results." Chapter 3." In Chapter 5." And how to find meaning? Look. says Schweizer. what others know how to do in the world. to walk the walk. is at on~e a statement about epistemology and a description of methods for studying colleagues nor the militantly antipositivistic position of others. say Fernandezand Herzfeld. Once again. "It is critical. and intellectual activity. both in terms of measurement and sampling. may contract for proprietary research. he says tend to sympathize with the artisans. not extraordinary. and valid measurements of variables in human thought and behavior under tough field conditions. . "While early British and V." What is needed is "not a contemptuous dismissal of methodology as scientistic nonsense.. among Asturian villagers he did an apprentIceshIp learnmg to manage the scythe. part of anthropological practice. for other knowledgeable people In. not just talk the talk to understan~ the ~eanin. every code of ethics promulgated by the American Anthropological Association has taken a firm stand against secret research. between the desire to explain and the joy of understanding. He has been playing the tapes: with the sound turned off. and designs within a single enterprise.

she says. "Developing awareness of the fundamental importance of ethics as a key component of professionalism. "and as a necessary adjunct to science and research is an essential task of the next generation of anthropologists. "should not be to learn more about ouliielv~s as individuals (although that will happen). "Cross-checking for bias. an . should not be "shy or intimidated about educating members of institutional review boards about the nature of anthropological research and the methods we use" but we cannot simply walk away as if the rules don't apply. anthropology may face the of its tools. WIth people that comes i~ from a genuine interest in their experience." says Gailey. takes place in a social ethnographers have absented themselves physically from their fields' world and is conducted by people who have cultural frameworks that shape their thinking. .. they say. b Mal·tnowski in 1922 and Bourgois in 1995 of thetr fieldwork are stn~tngly tlons y h d . F thing they Transnational studies pose unique methodological problems. This may create a tension that comes with responsibility to various audiences and interlocutors. liThe creative discomfort or tension that results from the adoption of such responsibility is part of one's state of being. She advocates an approach that calls for "integrating feminism into one's relations in daily life" and notes that this is what distinguishes feminism from other critical approaches. ng l i letter. she says." she says. or anthropologists. preliterate people.. touch on the part of the researcher will not necessarily appear as stnkmgly Ant.-Me~tco b?rder culture. "don't vary whether they're studied by psychologists.e~pmg . is not as easy to ensure as one might think.me IS I.S. . (3) gaining the kind of rapport. ~ites~ but today Indeed. some of us maintain contact bye-mail in pl. to migrants gb d t 11 "With studies of the transnow to cultural groups that have no or ers a a.at rapport t? engag . We must." H and particularly h Another feature of contemporary anthropology. e~~n Christine Gailey follows in Chapter 6 with a discussion of feminist methods. to more-or-Iess self-contained peasant Ultimately the objective of ethnography. "As transnational fields are constItuted. "The rights of research participants. and our goal may be to reduce bias by making our assumptions and epistemological claims clear to our 010 ists have always kept in touch with their friends and consultants fr om . "Iies in reflexivity. ifoot the . The go~ls include: (1) cr~~~I~~:~~~: conscious empathy for the lived experience of people m another culture. or Hong Kong. says Fluehr-Lobban. f h The DeWalts layout the goals of participant observallo n and many 0 t e techniques that comprise this strategic method. as method. To do this.. is marked by the study of populations that have more and more fluid borders-from isolated." she says." English or Chinese and English in order to conduct transnallonal studies . Anthropologists. Fluehr-Lobban advises anthropologists against asking students to fill out research questionnaires in classes that students are taking for credit.r~~ as t e ~n . she says. "techniques can be viewed as neutral. ":ith their own internal patterns of long~distance cohesion.t~ saY Quebec: Bombay. acknowledge that "scientific inquiry. of course. or one f ' may require fieldwork at both ends and a knowledge o~ th~ l~ng~. say the Dewalts. but. predIlectIOns. however.oa~~ more countries.here electTlclty h s only recently arrived. Informed consent." . IS S? c 0 e basis for a relatively unchanging theoretical core in anthropology as." Chapter 7.~o provtde th a wide range of theoretical development around that core.Introduction 23 research project. causes and consequenc~s of this lived experience.a~e ~int. With the development of transnatI~~~l stu b:~~ however Hannerz says Ilnetwork analysis may again offer ways of t~m mg a . of the principle of informed consent. how uni~s of study are'constituted and about the social spaces occupIed by vaned actors.I I I I I l . to localized ethnic groups.." But.." There are." she says. for feminist anthropologists. Ethnographers need command of Frenc an . Nor. u final test national and the global. says Hann~rz. "that the met 0 ." Thus.. this "should not translate into a rejection of quantitative techniques-that somehow these are inherently masculinist or antifeminist. intellectualized understanding of the. from unstructured conversallonal mtervtews to ptle l l I 0." says Hannerz. . is: "Would I be willing to disclose the source(s) of my funding to the people I'm studying?" Disclosure and openness have been forced on researchers with the implementation of the principle of informed consent." Gailey traces how feminist thinking has influenced and is influenced by other approaches. durin Hannerz points out that network analysis was promment m anthropolo~yb g the 1960s and 1970s but that it became unattractive to anthropologists as tt ec. cardiologists. is that Ilfieldwork goes on in some :vays. because ethics are part of the methodology. The~ observe that the descnp.ant" Will this contact itself forge a transnational cultural environment? audiences. eV~ha~ters 8-13 on specific methods of data collection begins with Kathleen a~d Billie DeWalt's discussion of participant observation. such ways of k. says FluehrLobban. says. Gailey. • d l l li more focused on itself. A field worker m~st know both Spams an English in any study of the fast-developing u.le. should anthropologists take an "exceptionaiist view" of the principle of infonned consent. from villa es to cities. is a coercive act-one that violates the spirit. abandoning claims to perfect objectivity does not mean pursuing bias. says Harmerz. says annerz. The history of method in our discipline.aces ~. and the conclusions we draw. she warns. says Gailey. Researchers can become aware of their own biases and can com- municate those biases to their audiences. and commumtles. I sely lied to similar. and (4) usmg th.. some special issues associated with informed consent in fieldwork settings. by Ulf Hannerz. This suggests. oping an articulate. how our research is designed and conducted. sorts and survey questionnaires. all kinds of research activities. is about a new field in cultural anthropologytransnational research. in critically examining the links that we make or do not make between our assumptions. but to learn m?re abou~ others''''Ji'r call for "systematic study of the effects of biases. of transnational studies.

as lohnson a~d Sackett:s chapter makes clear. lohnson and S~cke~. although it involves a greater commitment of time and energy. say Levy and Hollan. but as a vital complement to our understanding of culture itself. that the respondent will become annoyed. . Please." It also requires time..L BERNARD ImroductiolJ 15 personal characteristics on the research enterprise. is to focus more attention on human behavior. address the. Census to develop in-depth interviews with Mexican migrant workers in Florida. That is. "being a man or wo~an may be the mo~t significant social fact concerning an individual and obvIOusly should have an Impact on participant observation. (a penismutliatmg rite of passage) is done by Tahitians"-oras respondents who report on th~mselves-"asking him 'Can you tell me about your supercision? What did you thmk and feel about it then? What do you think and feel about it now?''' Some topics that are particularly suited to open-ended. The chall:"ge. that the interviews will become sterile. fantasy." Weller notes that a "top-down" approach can also be valid. In Chapter 10. the respondent usually regrets having the series discontinued. looking for examples of how the interviewer may be distorting and biasing the interview process.m of behavlOr m a communIty to generate a fair representation of people of dl!!erIng age." say Levy and Hollan. by Alien lohnson and Ross Sackett. I can say from my own experience that the temptallOn to study talk rather than action can really be irresistible. and fluency in the local language. if you really need to know. For some behaviors. describe for me exactly how and why supercision . reactions to illness and stress. in which the initial goal is exploration and the development of "a set of items relevant to the area of interest and to the people to be interviewed. including participant observation and mtervlew. responses to death. "moves back and forth" between treating people a. sexuality." From the sort of interviewing that requires learning to say "uh-huh" and "tell me more about that" in just the right places. "It is deeply distorting.ormants who report on their societY-"askinga Tahitian interviewee something lIke." She describes what she calls a "bottom-up" approach. personal data also help us understand social change and stability and a community's responses to internal and external forces. for direct observation. personal morality. argue Levy and Hollan. say Levy and Holla~. It is also a sample of behavior. Most of us. thoughts. This combination of exploratory and systematic methods. more than a report of culturally and socially important facts. "In ethno?raphy. This is not what happens in well-conducted interviews where understanding deepens and where . key questions involved in studying behavior under field conditIOns: What does It mean to describe an activity? Whom should we observe." They recognize the importance ?f uSl. "one might begin with survey results collected by someone else (a national survey.ng . "Survey results. she says. We can also examine the interaction between the interviewer and the interviewee. and so on. "Few ethnographers can speak to this issue as authoritatively as can the DeWalts. significance. This allows us. deals with direct observation. and 8aer (I 996) used results from the U. "the less that is known about an area. ~n other words.where? At what level of detail? How can we sample from the ~trea. is no easy task. open-ended methods. but to how and when they say it. person-centered interviewing." says Weller.) and supplement their findings with more detailed.multIple methods m the field. and st~tus." . the ~ultur~ and SOcIety of the communities in which they live. however. to pay close attention not only to what people say. by themselves. environmental values. The person-centered. A person-centered mtervlew. But. Chapter 9. and dreams. open-ended interview becomes more than just culturalIy standardized discourse." says Weller. in all the re. the census. and choosing between spot obsen:atlOn and continuous mo~itoring of behavior. in Chapter 11. Person-centered interviewing requires excellent rapport. Inexperienced interviewers fear that they will run out of questions. "inadequate to the task of constructmg trustworthy accounts of activity patterns." Studying behavior under field conditions. knowledgeable mformants can report accurately on the mean-what people generally do. The next two chapters are about two quite different approaches to interviewing." The results of this approach can be incorporated into more systematic interviews. but are limited in the depth with which they may explore a topic. are personal identity." noting that. the feelings. record reports of behavior. the more appropriate are unstructured. when. we move to a discussion of highly structured interviewing. "are good at providing a representative picture of what the population may be doing or thinking. in turn. affec. according to Levy and Hollan. ~nd .interviewing.levant contexts of their lives?" They deSCrIbe the p~oblems ofsamplIng. etc. is the study of the mterrelatlOns of prIvate and publIc worlds of people in a community-that is the study of how p~ople's minds and selves are affected by and. "In general. . they say.S. ~ccording to Levy and Hollan. "produces a study superior to one based on either method alone. despite the difficulties. selectmg UnIts ofanalysis.s mf. gender. open-ended work on the same topic.mg. The exploration of private experience (for exarnpl~. there is no substitute. or how much time women in a society have for leisure.." ~hey say uWha~ people do matters. . and experiences of a Hindu Untouchable in response to hIS or her public role and behavior) is important for anthropology's humanistic aims. to children." Kempton et aI. "The depth. what a senes of mdlvlduals eat. But. for example. to religion. Person-centered.S. and honesty of respondent responses tends to grow over the course of a series of interviews. (1995) used this approach in their study of U. Robert Levy and Douglas Hollan discuss person-centered interviews and t~is is followed in Chapter I I by Susan Weller's discussion of structured intervIeWing. but that these methods are. and they offer guidance on several optIOns for recordmg behavlOral observations. "not to work primarily in the respondent's core language. not as a SUbslItute for the study of culture.

She lays out the methods for constructing questionnaire items. Fadwa El Guindi describes the development of methods in visual anthropology-from Felix-Louis Regnaultfilm in 1888 ofa Wolofwoman making pots at the Exposition Ethnographique de l' Afrique Occidentale. In Chapter 13." That. Graham used two tape recorders in her work with the Xavantt. demographic history. and (2) We must work not only T l l . Many anthropologists will want to use va'os were asked to shoot film records of their own culture. but are guided by rules for coding social phenomena in language. The rules. ( ) Historical documents should be read with a sensItIvIty to the SOCIal." This work requires taping and accurate transcribing and close attention to detail. Graham and Famell insist that field notes aren't nearly enough to capture the verbal data on which close analysis depends. including scales and indices. triad tests. certain kin and affines.000 still shots in Bali and New Guinea (which. say Graham and Famell. cogmtlOn" an .h'. and political-economic history. however. va. Gui~di's ~. identity. the other recorder cap1Ured her consultant's responses to questions that Graham asked about how the language y a be i~hibited. seg". or what she calls "fieldwork In the arc. dlanes. say Graham and Famell. are "implicated in broader social and cultural phenomena.' I In recent years.to El. they enter t e archives through the back door of ethnography"-theirwork remams dlfferentfro~ Ihe work of historians. Still. or not analyzable when the investigatIOn IS t?tall . pile sorts." for example. For some historical anthropologists. and the community as well as fonnal elements of texts themselves. of ':~Ices fatthful In diction and mannerism). many anthropologists have adopted the standards of hls~onca methods for studies in family history." These features were not found in elicitations-in the speech of informants historical methods 10 link the past with the present-because. sentence substitution. hfe hIstOry. .:: this case the Navajo--would reveal aspects of coding. to Margaret Mead and Gregory Bateson's work of 1936-1938 that produced 22. but. oral history. often make leaps of mter pretation that "might make the historian wince. and she outlines the steps for creating and testing new scales in the field. and objective voice (analyltc narraltve)." For example." who responded to questions from the ethnographer. but scales develped for one population might not transfer to another. and cultural context within which they were produced. Like Levy and Hollan. say Graham and Famell. says El Guindi. not observable.Englt:h titling of Arabic songs. aesthetics. l l l . Caroline Brettdl assesses th methods of historical anthropology. a researcher can look for features and patterns that. The disagreement between these tw? sch~~ars 1 is profound. Egyptian Birth RtI~a . an? many l' l was being used and about interactions between speakers. d f According to Brettell. polltlca.hlv~s. . In Chapter 12. Using that method. the individuals with whom she or he works. As one recorder played back specific instances of discourse. kinship. and . pohtlcal. property transactIOns. to understand the .ly agree on ~wo th~ngs. In a method called "layering. photographed." . or about. gender relations. and sequentially arranged by: p~ople.stu y 0 historical problems include tax records.'ent-bY-S~gment. t ~ object is more global and comparative. Laura Graham and Brenda Famell discuss discourse-centered methods. El Gumdl u:es several tec ~iq~es "all layered within the film itself' to translate and communicate a cult~r~ to a foreign audience. With a corpus of transcribed discourse. subjective voice (English r~ndltton.n ~xplorations of method in her shooting of a feature on an Eg~ptlan hbl~~l nt~al. church and civil. El Guindi shows the ethnographic filmmaker at work.UERNAIU) Introduction 27 Weller describes the variety of methods available for in-depth studies of cultural domains. the object is to capt~re the emIC vlehw of history-how local people understand their own history-=-whlle f?r other~. because anthropol. to Sol Worth and John Adair's 1966 "native film experiment" in which Brettell discusses the recent case of disagreement betwee? Gananath O~eyeskere (1992) and Marshall Sahlins (1995) regarding the interpretatIOn by HawaIIans of the death of Captain James Cook in 1799. I I~ describing her own experience in filming El Sebou '. 7 l l l l' economic. the documents that anthropologists u:e In the . Worth and Adair says El Guindi.. In Chapter 14. One of the key concepts in this linguistic anthropological tradition is that culture is visible in actual discourse between and among persons-an emergent phenomenon that is "constantly open to new associations and interpretive moves. "together with interpretation of their connections to broader social phenomena [that] distinguish discourse-centered research from other anthropological methodologies. Anthropologists. including free-recall listing. she says. "are discovered through fieldwork and data collection and involve the interaction of such things as the ethnographer. such as relations of power.ng of Arabic statements made by the people in the film. wills. turned ?ut to be an Important idea one that has been modified and replicated in projects m other cultures. WeBer describes in detail some methods for translating and adapting scales to new environments.000 feet of 16mm film and 25. archlv~os of bIrths and deaths police and judicial records.e Yendent on verbal exchange-especiallywhen such research must be d?ne m the la:guage of the investigator.soclal. says El Guindi. devising ne~ methods durmg the shooting itselfto deal with issues and questions she had r~lSed m developmg t~e ro·ect. and the cultural consensus model. and cultural processes that characterize human behavlOr m the past as well as in the present. These techniques include "bottom-of-screen English sUbtltll. showed how film could be used to record the details of bodily movement and nonverbal behavior). ues a scaling device at some time in their fieldwork. and affect. they would prob~b. says Brettell. Graham discovered "features associated with the speech of elders and alternative patterns of secondary sources as well--even photographs. says Brettell. based their project on the Idea t at motIOn picture film-conceived." These moves are not random.~glsts u~e person marking and verb morphology used in speech to." It is the analysis and interpretation of these texts.

Anthropologists have always had a strong tie to texts.r ." We continue this pr. Cross-cultural researchers.28 BERNARD Introduction 29 to identitY the ethnocentrism of the writer of a document but our own ethnocentrism as evaluators of that document." Cross-cultural research has deep roots in our discipline. see thmgs iO and about people's lives we'd otherwise miss and answer qu~stlOns I? ways that help us pinpoint those errors in jUdgment that make us qUll1tessentlally human. can we v~hdl~ generalize?''' In addition. ." Chapter 15. we can be very confident it doesn't mean 0-40%. it will become easier and easier £0 examine subtleties of text that were invisible to us before. to the Human Relations Area Files (HRAF). I h d . "When w. With new technologies con? 100 along (voice recogmtlOn software. And we can be fairly sure it doesn't mean 40-49%. Intrac?lturaI compa~lson involves collecting many examples of the same story (lIke askmg 200 AmerIcans to tell the story of "Oeorge Washington and the cherry tree") in order t. . say the Embers. selecting a sample. One of Edward Bumett Tylor's best-known papers. In fact.'"we don't know if that means 50% or 100%. Many cultural anthropologists try to understand a culture. "construct a story . Today. and doing appropriate statistical tests-are the focus of Chapter 17 by Carol and Melvin Ember.o ~xamine variation in how culture heroes are portrayed. the practice of agriculture. say. The goal for many ethnographers: say H~ndwerker and Borgattl. the goal is to derive and test hypotheses about culture itself from the systematic comparison of cultures as units of analysis. describes the numerical methods ~f data analysis that are used most often by cultural anthropologists. IS to learn what is typical-How do most people conceptualize an event? At wh~t age do most people become parents?-and many ethnographers also ~ope that their field research work will allow them to generalize beyond the few mformants whom they can interview. and the films shot by Margare! Me~d and Gregory Bateson (of the Samoans and the Balinese) remain for new generations of scholars to interpret.and the comparison of stories across and within cultures. And so it goes. most of the recoverable data about human thoU~I~ and 11l1ln~n behavio. by W. say ~andwerker and Borgatti. as a functIOn of variability in experience. was to establish a database of ethnography so that scholars could make systematic comparisons of cultures. years spent in school.). but the texts collected by Franz Boas (from the Kwakmt/) and Edward Sapir (from the Wishram). Ryan and I describe developments in grounded theory (a qualitatively oriented approach to text analysis) and classical content analysis (a quantitative approach). if anyone. to ask "What proportion of societies have dowry?" and quite another to ask "Why do 3% of the world's societies have dowry?" It is one thing to ask "What proportion of societies practice polygyny?" and another to ask "What is the effect of polygyny on total fertility?" Testing hypotheses across cultures presents interesting challenges of measu~e­ ment and sampling because ethnographies-the fundamental data for comparattve research-are not based on a standardized set of questions to begin with. and do.is text of one kind or another.nced as children. The methods for doing this-deriving hypotheses. Debates about materialism and postmodemism and f~nctionalism come and go. w: people expe:ie. For others. .. in 1949. We aIm. HRAF is a consortium of 20 sponsoring universities (and hundreds of others) that support the collection and coding of a continually growing archive of ethnographies (currently about a million . and search for concrete events and Circumstances that shape those experiences. say the Embers.' . was on a method for showing t~e association of customs across cultures. say Handwerkerand Borgatti.e go to the next person. For example. Murdock's goal. P~IC wl~h those systems of meaning"-and data on the individual characteristicsthmgs like age and gender.." In Chapter 16. about the people we worked with: about their lives and the Circumstances in which they have lived'about what those people now think. Other methods that involve the analysis of text include analytic induction (also known as Boolean analysis). or how much if any violence Ethnographers. Cultural schema analySIS mvolves understanding statements like: "Blanche lost her data because she forgot to save her work." If you want to know whether the frequency of extended families is related to. It is one thing.var~. But it was George Peter Murdock at Yale In the 1930s who is most responsible for the development of cross-cultural research today. etc.ocess until f~el that we can. high~speed scanners. "then we should be able to see . A whole set of links are left out. In ethnographIc research." On the other hand. Oery Ry~n and I discuss methods of text analysis. Some ethnographer~ also want to go beyond description and explain how patterns and ~bs:rved ." . published in 1889. This led. l:t5 us. for example." In the end num . schema analysis. say Handwerker and Borgatti. "If an ethnographer says 'extended family households are typical. but they are easily filled in by any listeners who have the background to do so. the comparison of texts across and within cultures. ask descriptive questions and questions about the causes and consequences of cultural variation. Penn Handwerker and Stephen 80rgatti. ." We know that Blanche's forgetting to save her work did not actually cause her to lose her data. "we always find something different.. collect two quite different kinds of data: cultural data-data on the "systems of mental constructions [that] people US~ to ~nterpret themselves and the world around them and of the behavior isomor~ Both approaches involve identitYing themes and testing hypotheses. say the Embers. "explicit reasoning with numbers . coding the variables.abllIty m cultural and individual data emerge and persist or chan"e. enca met 0 s are thus nothmg more than explicit tools of data collection and analysis that address core research questions: 'Did we get it right?' and 'To whom. we begin by collecting data f~om one. feel.the relationship even though we can't use a percentage measure based on numencal infonnation. person. "to explain variability in culture and behavlOr. a~ well as somethmg much the same.

' tend to converge toward an of the mix of professionals makmg up aKte~~ should be ever mindful of the organizational norm. Another difference between applied and basic anthropology is the speed with which results have to be delivered. grant proposals. behavior. I am indebted to McDonald (1993. in NOTES l l l l' l 1. "is a book field. government bureaucrats. much of the research done by applied anthropologists is done as part ofa multidisciplinary team. textbooks. understanding ofhumanity:s diversity of expertenc. American Ethnologist o~ rd-there were no colons in the l l l And Finally '1 f n of the current state of This handbook is the ~atest an: . just as it does Kottak shows that content an stye c ani th titles of the first three articles in 't' He compares for examp e. 1994) for pointing me to Nlghlmga e song. commerce. Finally. I. The role of the anthropologist in a research project. ea (fig ~l I h ge.D. in popu Iar wn mg. validity. the role of the anthropologist and the informant can be reversed. these vanous ' uSlOg . In traditional ethnography. students. the widespread research to make a difference. and. an issue of the In Chapter 18. an d was repnn . 'bTt We all want our accuracy. potentially significant.ecent issue are more playftd and titles and lIttle jargon-Wh. 0 . fr \984 and those from an issue m 1997. "Cultural anthropology. the general public. conducting research. applied anthropologists have an even greater burden of responsibility because "the human." methodological toolkit. the annual . ::~' ~ba:~et:Il~:~.s to fend for themselves. "Applied anthropology calls for reciprocal learning and sharing of expertise in identifying a problem. ~on~ol." AnthropologIsts. fe :nd scientific methods for This handbook leaves no doubt that t e umanIS IC doubt about the need for doing these things are improving all the tIme.m~:~i~:t:'~ pre~ision." says Kottak. and our own anthropology has fully joined-as a consumer an as a co and development of methods in the SOCial sCIences. the speed with which research often must be done. and articles for the popular press.o:ucedto the full range of the art in anthr?pologl~al met Oh s. informants-Conrad Kottak describes the breadth of writing in the discipline. m d' more hterary. on presenting anthropology to diverse audiences-to our colleagues." says Koltak. al research. has become generalized to rapid ethnographic assessment.f~t.o~ cOd:. h was published in 1920 as part of the IS University of California's series on archaeology and Thunder Autobiography ofa Winnebago Indian was pu et~~o~o7i.r~~ult~ ~~ that ethnography took a year.n 1926 by D. yes. and government. "it isn't possible to write a book for more than one audience. ~a. f want to contribute we study. responding to reVIews. And fend they did. The first edition of Blowsnake s autobiOgrap Y .'terary forms learning to write for particular audiences. and the experience of others.'I~~I~~. Robert Trotter and Jean Schensul survey the methods used in and developed by applied anthropologists. e . contract reports. All of us across t e SO~I. all the methods of social research-and this means that they require very broad training in research methods." Also. Internet). tradebooks. .::~!. and ecological consequences of applied research are immediate. We are learning better." He draws on his personal experience. It leaves n~ It leaves no doubt that constant vigilance regarding ethical conduct of ~esearc ~tributor-thediscussion In Chapter 19. It used to be part of the anthropological canon The titles i~ th: earlier issu~ ar~ mo. social. Since 1994. and sometimes critical to the life and survival of communities. in academic writing. and using results.. ' .Introduction . and the ethical responsibilities implied by the conduct of research that can have direct impact on people's lives are some key differences in applications research. Some 0 .~ ~~fore These demands are for greater demands on our methods than we ever I . says 0 a . ted in 1983 by the University of Nebraska Press. defining a researchable question. ' I' I i nored 2. opening thousands of new jobs in industry. 11 ues in those l l ! .lhe tltles . Applied anthropologists use all the methods of anthropology-in fact. A from t e academy. say Trotter and Schensul. developed jointly by agricultural economists and anthropologists. but the distinction Koltak draws is a mirror for the scientific and humanistic emphases that characterize anthropology. but. The collection is used by researchers across the social and biological sciences. say Trotter and Schensul. He offers practical advice on 1 .JU 31 UERNARD pages) on 365 cultures (and counting) around the world. the most important form of communication in biological anthropology is scholarly articles in peer-reviewedjoumals. Just as anthropologists are methods used m empmca l ant ropo oglc . Co.g work. The method of rapid rural assessment. .'t~om~eono~ena to the prediction. all anthropologists are obliged to consider the ethical implications of their work. d r 'th the media and so on." By contrast.:o::~~~h:~:. . reliability.1 ~~:~~. increments to the collection (about 40. Appleton and e The book titled Crashing · .000 pages a year) have been electronic (CORaM. special potential contnbutlOn of thelf field. The Embers anticipate that this will encourage even more comparative research as time goes on. Cultural anthropologists are also expected to write articles. writing monographs. The filling and closing of the academic market for anthropologists in the early 1970s left a generation of new Ph.~~n~:e~i~t~:~h~:St~~~::r~:~~o~~n~~i~~~F:~::~~~:. explanation. "In general. and modus operan I where so much anthropology goes o. way n "the discourse. scholarly articles. team research on applications proJects.

K. Qualitative Methods in Management Research. 1953. VA: American Anthropological Association. and D. Fenno. L. 1991. University of Wisconsin Press. Essays in Honor o/George A. . H. and H. Human Organization 55(J):58-66.D. diss. Positivism in Social Theory and Research. In Search Cambridge Umversity Press. Booth. Frands L. Twenty Years at Hull-House. CA: AltaMira Press. Chicago: University of Chicago Press. Randomized Experiments/or Planning and Evaluation: A Practical Guide. G. In The AAA Guide. New York: St. Gary. P. Brypant. and T. Thousand Oaks. Consorte. A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data. and John P.Richard F. 304-315. 2d cd. 1989. De Vos. Hartley. Dead End Kids: Gang Girls and the Boys They Know. Communication Problems on an Oncology Ward. 1997. G. Paul Radin. JamesM. Cambridge: Schenkman. J. Michael. Kroeber. Handbook o/Qualitative Research. ed and Ir. PragmatIcs ~nd In D. Who Are the Whites? Imposed Census Categories and the Racial Demography of Brazil. Spain. Berkeley: University of California Press. 72:451-462.. Elizabeth B. Christopher. Sam. 1973. New York: Macmillan. CA: Sage Publications. Carolyn B. C. Du Bois. Academic Opportumty In Anthropology 1974-90. E. S. Business and Professional Ethics 10(4):57-68. and Seymour B. Stalking Employment in the Nation's Capitol: A Guide for Anthropologists. Fleisher. PS: Political Science & Politics 29:533-537. Cross~Cultural Issues in Interpersonal Psychoanalysis.32 BERNARD REFERENCES Introduction 33 Addams. Kathy E. 1985. 1978. ed. l. 0/ Respect: Selling Crack in El Barrio. Robert F.. and James S. Nonnan K. V~n de Woestijne. New York: Boyer. ed. eds. I. Paul R. Professional Ethics and Anthropology: Tensions Between Its Academic and Applied Branches. Exploring Clinical Methods/or Social Research. DC: The Washington Association of Professional Anthropologists. eds. Berg. CaroJy~. Koons. H. The People of Alor: A Social~Psychological Study of an East . lohnson. (University of California Publications in American Archaeology ond Ethnology 1[7]). Gummess on . Social Forces Hegeman. Developmental Changes Within the Structure of the Mammal Domain. Mark 1998. John. Lang. Evert. American Anthropologist 77:753~773. Denzin. The Challenge of Campaign Watching: Seven Lessons ofParticipantObservation Research. Research Design in Anthropology: Paradigms and the Testing of Hypotheses. Brim. K. 1995. 1996.. Beatrice Hackett. A. M. Marvin. Hammel. Lincoln. 1990. Adam. P. Vol. D. 1994. Cora Alice. 1995.. 1973 [1792]. Bemard.I. San Diego: Academic Press. and an Account of Some Foreign Prisons and Hospitals. Madison' . John. Truk: Man in Paradise. 1920. J. Martin's ress. Holloway. John ~. McDaniel. L. 1902. ed. Stanford: Stanford University Press. Psychological Techniques: Projective Tests in Field Work. King. K. 1953. 1996. ~. New York: Wenner-Gren Foundation for Anthropological Research. England: The University Press. Cambridge: M. Alien W. Thousand Oaks. The Professional Stranger: An Informal Introduction /0 Ethnography. Honigmann. Watching Politicians: Essays on Participant Observation.. 1997. Gladwin. Hillsdale. CA: Sage Publications. Psychological Anthropology. and C. Boston: Houghton Mimin. Killworth.. Kroeber (and others). 1993. Chicago: University of Chicago Press. 1901-1935. Rinehart and Winston.. University of Cambridge.T./~' Indian Island.. David N. . Adkins. In Anthropology Today. New York: Macmillan. Prisons and Lazarettos. PatIent EducatIOn & Counseling 29:131-136. 1995. Stein. . Boruch. Alfred c. NJ: Analytic Press. Health and Mental Health among Mexican American Migrants: Implications for Survey Research. Pp. Smith. CA: Sage Publications. Byrne. Marguerite. 1953. Press. and K. 1944. Minneapolis: University of Minnesota Press. Washington. Agar. ed. Developmental Psychology 28:74-83. T. F. CampbclL 1979. NJ: Analytic Press. ~. 1974. ~nd K. ed. Life and Labor of the People of London. Boundaries and Construction. 1996 AAA Survey of Departments. Hsu.Philipp~ I. 1996 [1980]. University of California. Thomas. Berkeley: Institute of Governmental Studies. Scientific American (October):42-44. Willett. lules. ~oy. 1972. Boster. Sarason.1994. 2d cd. Jablonski.. Mason. Alfred. and B. Walnut Creek. 1992. Baster. A. 1996. Thomas D. Research Methads in Anthrapology: Qualitative and Quantitative Approaches. Thousand Oaks. and Yvonna S. Quantification in Cultural Anthropology: An Introduction to Research Design. De Valck. HJllsdale. L. J991. lane. lohnson. Blowsnake. and 1. Arlington. ~ryce. 1975. Henry. Kempton. Princeton: Princeton University Press. Montclair. 1997. CA: Sage Publications. 417-429. 1997. Ph. Bourgois. Chicago: Rand MeNaHy.. Beverly Hills. David. R. Russel!. Handbook o/Social and Cultural Anthropology. NI: Patterson Smith. Cook. 1985. Anthropology Today. 1994 [l988J. Harris. Reports ofthe Cambridge Anthropological Expedition to Torres Straits. A. and Melford Spiro. Cambridge. Mervis. Fluehr-Lob~an. 1926. Howard. Culture and Power in the British Army: Hierarchies. Glaser. Quasi~Experimentation: Design and Analysis Issues/or Field Settings. Baee. Boyer. Charles. New York: Halt. Give ns. Environmental Values in American Culture. eds. Pp. Profile: lane Goodal!. eds. The State o/the Prisons in England and Wales: With Preliminary Observations. D'Andrade. 1996. The Autobiography of a Winnebago Indian. Haddon.

6th ed. 1993. Seligmann. Gunesekara. Oberschall. Knowledge/or What? The Place a/Social Science in American Culture. Originally publIshed In 1835. eds. 1951. 1989. 1 1 1 l l Paul. and Cybernetics 22(5):983-991. New York: Appleton-Century-Crofts. 1922. a. Taylor. Pelto. 1992. Charles Gabriel. W. Anthropological Records 5(2): 111-153.Inder. Obeyesekere. A Handbook pology.. In The Establishment of Empirical SOCiology. Man. The Apotheosis ofCaptain Cook' European Mythmaking in the Pacific. C. The Veddas. Helen Swick. The Subjection of Women. Kroeber 1940. Oberschall.~B. D. Social Interaction in Same and Cross Gender Pre-School Peer Groups: A Participant Observation Study. In Handbook of Pe . John. Notes and Queries on Anthropology. and G. Annual Review of Anthropology 18: 177-202. Publication of the American Sociological Society 25:1-10. 1911. Naroll. K. Reprinted 10 translation in 1969 from the 1842 edition as: A Treatise on Man and the Development of His Faculties.. New York: D. History. and W. Meyer. Montreal: McGill-Queen's University Press. Smith. New York: Harper & Row. Projective Techniques and Cross· Cultural Research. McDonald. I. American Journal of Sociology 54(3):181-190. Green. Seligmann. Malinowski. Culture as Consensus. 1997. New York: McGraw-HilL Redfield.z. 1995. P. Marshall. Batchelder. Introduction. 1933. Lynn. K. Richardson. Robert A. 2d ed. eds. Princeton: Princeton University Press. 1972. Gananath. 0/ Method in Cultural Anthro- Nightingale. 1997. Westport. Systems afConsanguinity and Affinity o/the Human Family. Anne 8. 1969 [1835]. American Journal of Sociology 25( 1919-1920):731-756. {nex. Dutton. Pelto. Weller. Robert. 1871. CA: Sage Publications. William F. 1870. EliJabeth. How to Apply the Anthropological Technique of Participant Observation to Knowledge Acquisition for Expert Systems. . New York: Cambridge University Press. Educational Psychology /3: 29-42. Malinowski. McDonald. John Stuart. London: Routledge and Kegan Paul. London: Longmans. N. Anthropological Research: The Structure ofInquiry. 1985. The Folk-Ways of a Scientific SOCiology. FL: Scholars' Facsimiles and Reprints. 1869. S. The Managers of Social Work: The Experiences and Identifications of Third Tier Social Services Managers and the Implications for Future Practice. The Early Origins o/the Social Sciences. SmIth. New York: Macmillan. 1948. 1927a. Raoull. S. Myers. Lewis Henry. P. 1939. M. Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea. Sex and Repression in Savage Society. Kornblum 1996. 1986. Together With a Proposal for Organising an Institution for Training Midwives and Midwifery Nurses. Bronislaw. essai sur~e deve. Paris: J. Bronislaw. It Pertti and G Pelto 1973 Ethnography: The Fieldwork Enterprise. Contemporary Marketing and Consumer Behavior: An Anthro· pologica/ Source book. IEEE Transactions on Systems. In Exploring Clinical Methods for Social Research. Berg and K. Cambridge. Lynn 1994. Florence. OU. 1970. The Father in Primitive Psychology. L. 1927b. Lincoln: University of Nebraska Press. 1. Psychoanalytic Anthropology. . Rivers. Sherry. Quetelet. Lindzey. The Social Survey and the Science of Sociology. and A. John F. and W. 1997. 1995. CT: Praeger. ed. Hearn. New York: Natural History Press. and Ronald Cohen. Beverly Hills. Physique sodale. 1920. Pp. Introductory Notes on Lying-In Institutions. and 1.s. Morgan. Princeton: Princeton University Press. 1930. Bronislaw. 1992. Adolphe. Mill. DC: Smithsonian Institution. and Patricia M. Mary A. Systems ofConsanguinity and Affinity ofthe Human Family. and Dianna Bandhauer. The Women Founders of the Social Sciences. ed. Pertti 1. Robert. Three Centuries of Women's Dress Fashions: A Quantitative Analysis. Jr. 1978. Rovegno. New York: Harper & Row.:Xial a~d Cult~ral Anthro.34 BERNARD Introduction 35 Lawler. Thousand Oaks. with C. 2d ed. Pp. A. 2 .. 1906. Lyod. 319-332. Pelto. New York: Norton. The Art of Social Science. Chicago: Rand McNaIly. American Anthropologist 88:313-338. New York: E. Anthropological Research: The Structure 0/ InqUiry. John Honigmann. In the Field: Readings on the Field Research Experience.. R. How Natives Think: About Captain Cook For Example. Using Participant Observation to Construct a Life. Journal o/Teaching in Physical Education /6:401-425. Sahlins. Pertti. Smith.. Gardner. 1993. Malinowski. 1970. Romney. A. Chicago: University of Chicago Press. Carl C. 187-251. Tooker.ology.'oppement des[acult~s de l'lomme.~ . The Todas. Washington. Gainesville. Perry.. Ottawa: Carleton University Press. The Institutionalization of American Sociology. CA: Sage Publications. Appleton. Norms of the School Culture ~hat Facilit~ted Teacher Adoption and Learning of a Constructivist Approach to PhYSical EducatIon. Pp. 241-288. H. Anthony. Brace. Carolyn D. Radio. Ogburn. British Journal of Social Work 27:191-218. H. 1961. Bailliere et fits. and A. Lewis Henry Morgan. The Method and Theory l l 1 1 I l 0/ Ethnology. New York: Harcourt. Englaod: The University Press. Royal Anthropological Institute of Great Britain and Ireland.

1898. E. 1980. London: Longmans.. 1995. E." International Journal of Nursing Studies 33:271-284. R. Inside Schools: Ethnography in Educational Research. Barbara. Wolcott. How Do Managers Think? Identity. Hoebel. 1996. Edward B. W.W. and Beatrice P. Whiting. Webb. 1992. Brewer. Teaching ofPsychology 19:J80-182. Exploring Ethnic Contact with Stories and Stere- otypes: The Thematic Apperception Technique and the Adjective CheckJistas Supporting Interview Tools.. Weisfeld. Green. View from Spring Hill. Weick. A Participant-Observation Course in Applied Adolescent Development. CA: Sage Publications.36 BERNARD Tsatsoulis-Bonnekessen. Morality and Pragmatism in Managerial Theory and Practice. Peter. Pp. White. The State and the Doctor. Englewood Cliffs. 1992. 1993. 1889. Journal of the Royal Anthropological Institute of Great Britain and Ireland /8:245-272. Beatrice. 1995. D. In CriSiS in Anthropology. Judy L. 1910. Webb. Tylor. Kaiser. Standards for PsychoculturaJ Research.. 1986. and D.M. E. Roberta. Sensemaking in Organizations. 1963. Karl E. Glenn E. and S. New York: Garland. Woods. Currier. Tony J. D. Madison: University of Wisconsin Press. Webb. Walnut Creek: AHaMira Press. eds. Thousand Oaks. Sidney.• and Paul Jorion. 1964. London: RoutJedge & Keegan Paul. NCAA Division 11 College Football Players Perceptions of an Athlete Who Consults a Sport Psychologist. Linder. R. " .5-164. A. Brewer. Journal ofSport and Exercise Psychology /4:273-282. 15. PART I PERSPECTIVES Woodgate. Harry F. My Apprenticeship. Kristjanson. Wolf. NJ: Prentice-Hall. New York: Longmans. 1992. Management Learning 27:323-341. Current Anthropology 33:454-453. Webb. Douglas R. J. Beatrice 1926. Green. and Karen de OJivares. Representing and Computing Kinship: A New Approach. B. . and Linda J. The Art of Fieldwork. Watson. 1982.. Van Raalte. On a Method of Investigating the Development of Institutions Applied to the Laws of Marriage and Descent. Anthropology. Cultural Anthropology Methods 5(2):1-3. . American Diary. 1996. A Young Child's Pain: How Parents and Nurses "Take Care.

relativism. Then I address three key problems of anthropological research: (I) the problem of reality. (2) the relationship between interpretation (including empathic verstehen) and explanation. I present the developments. historical reconstruction. and some strengths and weaknesses of these methodological frameworks. I. epistemology addresses the scope and justification offactual knowledge that anthropologists have established through fieldwork (Werner and Schoepfle 1987:Vol. Ch. The core methodological problems of anthropological research are: (I) how to produce valid descriptions of other people/societies/cultures. I introduce the positivist framework as a representative of the scientific approach and present hermeneutics and postrnodernism/radical constructivism as exemplars of the humanist approach in the social sciences._- In this chapter I discuss the nature and justification of anthropological knowledge. and truth. I). true belief') in general. and (3) the problem of theory construction and theoretical progress. main tenets. and comparative studies on human 39 l 1 . These core problems are broken down and translated into lower-level methodological problems in debates within and across different methodological frameworks. In cultural anthropology.THOMAS SCHWElZER • Two j 1 i Epistemology The Nature and Validation of Anthropological Knowledge . and (2) how to establish comparative and theoretical syntheses of these descriptions. "reasoned. Epistemology and Methodology in Anthropology Epistemology is a subdiscipline of pbilosophy concerened with the nature and validation of knowledge (that is.

SCIen h' . even practitioners of this critical school adhered to fieldwork and now produce ethnographic descriptions that differ in focus. and hermeneutics (including postmodernism d d'cal constructivism as recent currents). thick description versus establishment of societallaws). positivism. Bernard 1994) and how they write ethnographic reports-had been umfymg bonds across rival methodological currents. This quest for empirical knowledge can be applied to any society and culture. and. Wilk 1996).fic versus humanist methodologies has been most mtense m anthropology.. Tsing 1993. researchers and researched can't easily take for granted a common cultural background. including the anthropologist's." In cultural anthropology. in D'Andrade (1995:194-199). Clifford 1988). Kondo 1990. however. and linguistics in departments and with a commo~ hlStO":. Depending on their adhere. However. Steedly 1993. The most basic distinction in the methodology of the social and behavioral sciences and the humanities is the split between scientific and humanist medlodologies (Bohman 1991.but there was general consensus on the value and conduct?f gat?erm g ethnogr~phlc knowl~ edge on the ground and presenting data on actors 10 theIr local a~d regIonal co. Hoilis 1994. empirical disciplines like anthropology. humankind. The scientific method involves the claim that there is no difference in principle between the goals (description and explanation) and the conduct of research in all disciplines. so the clash of p. Marcus and Fischer 1986. Werner and Scho:p~e 1987. produces some minimum commonality of understanding. the theoretIcal splits had ~nly slight consequences. In such cases. for the universal logical forms by which humans reason. As used here. after intense programmatic debate. epistemological reasoning at the core of philosophy to the more concrete methodological level of how to conduct research and what the goals of inquiry are. Methodology includes discussion of method-How shall we proceed?-and of the principles of theory construction-What are the goals of inquiry? What shall the knowledge that we want to produce on OUr subject matter be like? What does the· concept of theory mean? A methodological framework. Epistemological questions become translated into more manageable methodological ones that can guide empirical research. . see the evidence for emotional universals in Ekman (1980). but not in prindple. . social. all competing anthropological schools were In some way mter~ ested in comparison of ethnographic cases. 1986.ntext in reports. open-ended mtervlews. anthropologists like Melford Splro (1986. Dumont 1992. although they disagreed about the systematics of comparison.A current assessment of the subdiscipline shows that substantivist concerns and formalist models have become blended even at the theoretical level (Netting 1993. . Most debates on theory and method in anthropology are methodological. Note the similarity to the methodological a~d theoretical debates on formalism versus substantivism in economic anthropology m the 1960s. I conclude this introductory discussion of anthropological epistemology ~Y pointing out that there are two main problems of methodological debate In :m . sociology. In discussing methods and theories. In the United States. t'. At first. Lindstrom 1990.from t IS ~pecIa oca Ion between the sciences and the humanities and the ensumg proP7 nsIty to adop~ the different science or humanist agendas has been cap~ured by Er~c Wolf (1964. as examples. prehiswry. and systematic techniques (Agar 1980. method refers to the procedures of acquiring knowledge on a subject matter. is a particular school of thought that makes some claims and sets up prescriptions on how to conduct research in a discipline. on the other. monographs of the past or from descriptions produced by anthropologISts folIowmg other methodological frameworks (for example. I I r The centrifugal and centripetal tendencies emerging .nce to vanous methodological frameworks. Still. on the one hand. This tended to separate economic anthropologists at the level of programm~tic statements while at the empirical level of monographic description. the humanistic method involves the claim for special goals and procedures of research in the humanities that take meaning and history into account. Gupta ~nd Ferguson 1997 in general). when anthropologists investigate what some call the "Other"-people who speak a different language and belong to different historical and cultural traditions. the ~o:r~field orientation connects cultural anthropology. then. are the main representatives car I ientific versus humanist methodologies respectively. Lang 1994. In debates about the aims and conduct of research in human. biological anthroology. The problem of comparative synthesis. Yang 1994. Epistemology discusses the general problems of how we can know something and what knowledge is. while theory refers to substantive results of a more general nature on the state and structure of the subject matter. Plattner 1989. adherents of the different agendas debate the g~aIs of inquiry and draw different rules for empirical research. postulating different goals for the research agenda and favoring different strategies on how to achieve the goais (for example. Kuznar 1997). is still there. Further. the most agreed-on point is that researchers in a particular field lack a uniform methodological and theoretical framework. whereas methodology is a more specific discussion of the goals and procedures of inquiry in particular disciplines. and behavioral sciences. and psychology shift the focus from very general. from. anthropologists disagreed about SpeCI~C procedures.88).j 40 SCHWEIZER Epistemology 41 cultures past and present. who observed that anthropology is "the most humanIst of the sCIences and the most scientific of the humanities. But fieldwork ~ractI~es­ how ethnographers collect data by participant observation. hermeneutic understanding versus scientific explanation. and this creates special problems for generating valid knowledge. 1990) argue that our belonging to one species. . Generating valid anthropological knowledge becomes more difficult in practice. Gewertz and Errington 1991. recent postmodernist challenges of ethnographIc authOrIty and wntmg seemed to end anthropological consensus on fieldwork practices and the monographic tradition of ethnographic writing (Clifford and Marcus 1986.

d' t differences or lead to shallow. More specifically.ei~v~~. society). the When Auguste Comte and his disciples coined the concep~ 0 pO~I.~~:a~~. d rt l Since every framewor teo s 0 i'i contingences of their creators an suppo ers. (2) At a more theoretical level of anthropological synthesis. n meant collecting and vahdatmg actua now d b the traditional classic. positivist stance. In the theory section.a~u~~~. I take into account the broader epistemological and methodological reasoning in the philosophy of science and the neighboring human.~~e~. Thus..e~es . t I look for convergences acrosS r > . od I l l l l l l . positivists disagree on the specifics of their methodology (and many dislike the term "positivism" as well). :~~~ss~~~~~~~:~set:oe:Pl~::~i~n of common understandings in historically based cultural traditions. we must answer the question of . I discuss the differences in the framewolics as well as the similarities. pro- (Kutschera 1982. First. 3.tlv:. StegmUller 1969.~~i~~it::{:~~. .conne. pohttcal.nd I l.ctlO.- 43 anthropology: (I) At the ethnographic.J historical/political context m whtch tt tS embedded..~~~::.. an pos . but complementary (as are verstehenand explanation). and behavioral disciplines. false compromises.. among others. h h . h t when It was esta tS e meamng In the nIneteent cen ury. etween that stresses the creative power of scientists to lfivent rea I y. f ameworks in 1. The growth of cumulative knowledge in the natural sciences is the model for this framework.n. 1979b.: ~~~~r~IW.~s~ocial E . f th frameworks Positivist 'r' In the following. I ' " framew~rks and put less emphasis on the historical.~~:i~Sm..:. fi /k ledge by sctenttfic meth0 s. .e.:r ~~~. t eredt the a particular framework an J.. Second. ")i\:~t~~~~~h~::~e:~~~:c:. . Positivism proposes a unified methodology fordifferent branches ofthe sciences and the humanities. how to compare ethnographic cases across time and space and how to arrive at valid theoretical generalizations. . ..~~r:~:~t~~~~t~~~:~ ~:~ f :~~dti~~ °tm CO"dme'mUli:~:~~~:~~c~~ec~:~:.~~~~s~~:'~:~::'e~ho:oIOg:C:~ . true knowledge h?d been hOb~cu~e ge~s Thus classic .~~:~::o~. ting parts of tbese Thus I focus on ideas as the more endurmg an In . I propose the new structuralist theory concept as a rich and precise framework for describing and analyzing the structure of generalizations and their change in the discipline. number crunching.:ntl~~ r~~ The Science Approach: Varieties of Positivism Divergent Methodological Frameworks Below the basic split between scientific and humanist approaches. like the contrast between verstehen and explanation (see below).s Postmodemism questions systematic approaches to the. I roaches d 'srn and (4) radical constructivism are relative y new app.Epistemology 42 SCHWEIZER '. However.~~~~~~~ ~~~~t:~~~.:ss~~:~~. " genesrl'tal. . powers of the state and the church and their metap ystca a T l . In methodological discussions. .m t~ive" nineteenth century.> . two. these fundamental problems are split into lowerlevel problems..~ po. ifferent meanings are scientists and social reformers. Iostea . :o:~:.te \and henneneutic ideas have been developing for two centur~es. 1. three. f .i:. it had an enlightenment and antimetaphYSl~al sPlftt · dPosl the . my goal is to assess methodological frameworks and methodological problems in anthropology and to sketch a proposal for solving these problems. . I would like to make the point that positivism is less homogeneous than is sometimes thought and that scientific and humanist approaches are different in focus. b "f simple and detenninistic . and acceptance of thbel~tahtuds qbyuo · ur:ps . 1973.'. Today the teon "positivism" is often associated negatively with n. I draw on analytic philosophy of science as a metatheory ':. . A vartety of somettmes very d associated with positivism. . In the following. Bunnin and Tsui-James 1996. This s~ollld~'t ~~e~. :dkal con. . ~~:~o~:~~~~~e::~~~~::~~~:~~:i~~~~. describing them as rational procedures for acquiring anthropological knowledge.~~..om . or four-depending on one's standpoint-main schools frame specific discussions of epistemological and methodological issues in anthropology and related disciplines.l C the rival framewor 0 " I for understanding meanmg m t e uman establish a speCial meth~d~ ogy f hermeneutics is the interpretation of texts. I characterize the mam tenets 0 ese.:.'~~t~. . we must consider how to produce valid descriptions of the Other (person. S I ·oncenIT." "--eworks are complex belief systems allowing for mternal vartatton.et~:~IOgical agenda. and btkOgraPdhlc. more observational level. I discuss some key problems of ethnographic description and theoretical generalizations that have been put forward by the different methodological schools. 1979a. I introduce the different methodological frameworks as background traditions of research. culture. "I't" these ar:. I outline the different methodological approaches used in anthropology to address these issues and some related lower-level problems encountered in pursuing valid ethnographic descriptions and producing sound theoretical syntheses.~ow-mi~~~~ data c~llection. social. . . It s ou d in each 'possibility that there are overlapping concerns and ~ational prace ures framework that can be cross-fertilized and even comblOed. .. It poses the discovery of general laws as the ultimate goal of scientific inquiry and advocates the scientific method of hypothesis testing as a general procedure for generating and validating scientific knowledge. readers should be aware of this possible hias). Herrneneulics (the mkorte "./j'2.

Positivism in the past has . The newly invented methods of formal logic were applied to philosophical and other scientific statements to clarify language. depicted today by its CritICS.n~ver been as. not the least being respect for the erudite collection of The second meaning of positivism relates to the ideas of logical positivism as developed in Vienna dUring the 1920s and '30s by scholars in philosophy and some related disciplines (see Camap (1963] for a personal account. these people wanted to tree philosophy and the empirical sciences of any metaphysical.fi~~t%:s and concepts therefore necessary standards: clarIty of langu g ( I' b r tional are tools for communication) and validation of truth Calms y a he of logic and empirical inquiry (i~c1Ud~~g a ~:~:~~ ~~ :~~ :~:~titfiC assumptions of arguments and'dtests s)o ~o~t ~I :~::.1 h Formal analySiS well as later positivist ideas were taken ~nto . means ' U as is sometimes I. :~I:t~:bates and theoretical diversity.e p~sitive heuristic of hypothesis falsificationism has to be supplemente yam ~o::Uld ~Iso statemen~ th~h~r~~~s:~~n~e~t '~fh~potheses empiricist doctrine. re ardin the different grounded and . Popper concluded that positive evidence ("confirmation") and the inductive method (the search for rules that lead from limited observations to the establishment of valid generalizations) are not at the heart of science. a SIS . Some 0 t e leas 0 . critical rationalism. b they all depend on t e ru es 0 humafiltles. Eventually.IS ~:~h~I~~ricas and ~ome parts of Europe domin~nt mode OffPhhll~~oPhlc. . Against this falsificationist vl:w'd·some P I °thSOePtest situation is always more . ecause f" . g th d ." positivism was not at all a politically neutral or conservative doctrine. minimum and 2. most notably . Rather.:~hg. But the positivist movement. t th's odology. negative evidence ("falsification") and deduction are at the core. In fact.i?t~nse debate °9f67~~thtO~0:~~~.~?~Jytlca~/ ~:o~. d has furnished . construction and theory devel~p~ent. when choosing between rival hypotheses. Geier (1993] for a historical one of the "Vienna Circle". and emptrlcal testmg.in pillars of this as a tool of rational reconstruction fian. The clarity of rational debate and critical spirit among its adherents were outstanding.. 1972) regarded himselfa critic of logical positivism. Even if all present evidence is confirmatory. along with related currents like evolutionary theory. easurement lex· a wrong conclUSIOn eau Imp falsiry a?y other in is theories and observatl. a wrong consequence disproves the truth ofthe premise. 199~. Blaug 1992. narrow'~minded able diversity within . Janik and Toulmin (1973] on the zeitgeist. Nevertheless. and Achinstein and Barker (1969] on logical-positivist ideas). to detect and refute metaphysical claims. framework that leaves a lo~ of r?. And we should keep the one that has survived serious attempts at falsification and therefore has proven less false than its rival. SOCla.96~ So ~he on~-sided critical view of statements in future apphcatlOns (papm. 1969b--or a similar approach of the IllventlOn of h~poth '. It also led to the recognition and better practice of scientific principles. h. according to rules oflogic. In the positivist view. aIso called "the hypothetlco~deductlve me 0 . 1969b.rents of analytic methcommunity can check the eVI ence . . Failure of prediction inevitably leads to a refutation of the generalizing statement. . Because our tests are limited. The strict h·1 hers of science. there could always be a refutation in the future. As with earlier positivists. for example. but Willard Qume . I. and early as and ASIa. .~~~itiVism.om or In ral term than positivism) has had a "Analytic philosophy of science (a better gene. . provided the foundation for several social and behavioral science disciplines. t es and refutations" (Popper validation and use the method 0 . Id .conJe~ ur eses 10 ical proof.a~~::~i't98~. Viennese philosopher Karl Popper (I 969a. .~u o. I nd behavlOral sCiences an significant impact on so~e S?CIa a . In the spirit of critical rationalism. .. The Positivist Thrust There are three main points to remember from this short sketch.o~al statements: ~~~~e whole test situation and is guided always based on a hohstJ~ u~derstandmg d velo ment otential of theoretical by inductivist (probablh~tlc) Judgments of t~.. it fell into disrepute when it turned out that its antimetaphysics were based on a rather metaphysical creed of progress. ahn I f logic and empirical '. these standards apply to any d!S~iPIi~e'b~nh~:ii~~~I:~~:n:::t~der~e no difference among the natural. the scienti~lc method :e~s ~. currents of POsItIvism (Albert I . We can never be sure that a (universal) theoretical statement is true. The ideas of the logical positivists spread to other parts of Europe (Berlin.c hilosophy which became the The fourth varia~t of p~sltlvl~m . (1961·Ch 2) pomte out tht a .:~. Prague. OXford. and to create better science. speculative assumptions not founded on observation or confirmed by empirical tests. because. some of the movement's key early ideas were refuted from within. he developed critical rationalism (a third variety of positivism) in opposition to but connected with logical positivism. Cambridge) and the Americas before and after World War /I. I d. Backhouse 1994. Kuipers 1996). among others). T~ere IS consl er positivism/analytic philosophy of sCIence. demanding that all theoretical statements be based on observation couldn't be maintained because the universal truth claims of theoretical statements always transcend the limited observational evidence available from tests. we should always select the one that has a higher information content by being more general and thus more challenging due to its wider range of application.44 SCHWEIZER Epistemology 45 empiricaIIy based "social facts..nes Economics has seen the most methodological self~reflectlOn III these dlsclpl~ . Iy that the hypothesis at hand is wrong. Hempel 1966.

nineteenth-century and logical positivism informed the methodological outlook of earlier leaders like Franz Boas (1961: 260-269). A major problem for philologists has been to understand and hole unravel the meaning of ancient terms and to grasp the . to do so would miss the subjective.l.. Positivist Impact on Anthropology Within anthropology. Nade!'s (1951) almost philosophical treatise on "The Foundations of Social Anthropology" is exceptional in its explicitness of adopting a logical positivist stance. A. Sometimes these contrasting goals of inquiry are described by the oppOSItIOn of nomothetic (law seeking) versus ideogrophic (describing the. d rt ture Bible. Alfred Kroeber(l948:Sect. we should remember that this framework has never been really tested with regard to theory building (compared to its successful application in economics and some parts of sociology and psychology.. As Wilhelm Dilthey. In addition. ots in the classical tradition of Greek antiquity. So. as hermeneutlc the interpretive activity itself is embedded in and dependent on hlStoncal traditIOns. In this view. meaningful. put it (1900: 144): "We explamnature: [but] we understand the living of the mind. and some parts of social anthropology and in neighboring fields like biological anthropology and archaeology.m~anmg of texts. It also informed discussions of observational techniques: Many anthropologists adopted the empiricist part of the positivist agenda by enhancing the quality of data collection in the field. And Radcliffe-Brown's (1957) general ideas on comparative anthropology as a "natural science of society" were innuenced by positivist thinking. R. l' Thus. see the discussion in Brodbeck The Humanist Approach Hermeneutics Hermeneutics (from Greek hermeneutike. In the analytic perspective. Siegfried F. economic. some mostly German philosophers and associated humanist methodologlSts developed the larger vision that hermeneutic rules of text interpretation could be extended l l l l t~ . The search for laws. . 1968). when hermeneutics provided many with a new challenge and reoriemation. In contrast. at least implicitly. George Peter Murdock (1949). natural scientists are only mterest. because interpretation rests on historically framed preconcepttons: The notion of the "hermeneutic circle" is invoked to descnbe the constant dialogue between preconceived backgrollnd knowledge on the text as a whole ~nd newly acquired insights into the meaning of its parts. the study of humans and their meaningful products In special historical circumstances should describe the particular and pr~ceed along the lines of text interpretation and empathic understanding (verstehen. although he never actually established societal laws. as a result of his training as a philosopher in the Vienna Circle before becoming an anthropologist. . hermeneutics was first a valua~le c~llection o[heunstlc rules and Interpretive hints on how to read and interpret hIstoncal texts. usmg an approach caIled phenomenology. particular). v: Beginning in the late nineteenth and early twentieth . . It ongmaIly flounshed m the Its ro 't'es which were preoccupied with the exegesis of historical texts hke the humant I . Radcliffe-Brown (1957. society. the outside).art of interpretatio~) has ." Philosophers hke Edmund Husserl. Bohman [199I:Ch.of human subjects and lead only to trivial results. when critics debunk positivism as an infertile perspective in anthropology. Some analytic philoso+ phers of science have attempted to demonstrate that even in the humanities historians. for instance. however-the crucial theoretical agenda of the more recent and more sophisticated brands of positivism-was never the prime goal of "positivist adherents drew the methodological conclusion that the search for general laws m nature is the goal of the natural sciences (using experiments and observattons fro~ anthropology" in the past (perhaps with the exception of Murdock [1949]).become general guidelines for research in the humanities. implicitly use laws of human nature or quasi-laws (claiming validity for certain areas and epochs only) in explaining historical events and processes (Hempel 1965-this view is not accepted by all historians and philosophers of history. Scientific. hermeneutt~s met~od~loglsts ~?ted. [1968:Sect. a refined positivist agenda-including the search for generalizations-is being pursued mainly in the subfields of demographic. The key features of meamng and historical situatedness were considered to be in sharp contrast to lhose of the natural sciences whose objects do not produce meaning and whose resear. In the hermeneutic perspective. philosophical and jurisdictional treatises.. histoncal documents. and others. all humanttles . Their successors adopted some ideas of analytic philosophy of science.Epistemology SCHWEIZER 47 3. Thus. positivism mainly had an impact at the level of general methodological orientation of anthropologists. and most anthropologists adhered to a positivist methodology up to the 1960s.chers can neglect history.ere conceived as being preoccupied with the study of meanmg ~reated by human bemgs in particular historical settings.centunes. Currently. Therefore. mea~ing the . and then the tide turned to interpretive and postmodemist methodologies.. I]). Robert Lowie (1959:279-291). an influential proponent ofthe humanities framework. In the sense of adopting the actor's point of view). research-oriented pOSitiVIsm seeks to establish laws (logically sound theoretical generalizations that have survived serious attempts at refutation). In accordance with these ideas· the humanities shouldn't try to estabhsh general laws. 7). and historicaIly specifi~ nature . and culture and can be used in scientific explanations. 5]. ecological. an I era f the past often written in foreign languages and stemmmg from distant epochs ~Geldsetze~ 1989). invariant aspects oftheir subject matter. such laws rule nature.ed in the unchanging. cognitive. attempted to show how everyday life is structured l l l . among others).

2) has explored insights from Gadamerand Schutz on bridging different traditions to arrive at an in-depth methodology of the unfolding of knowledge in fieldwork. it stresses the historical embeddedness. I take culture to be those webs.aning and h~stor~cal situatedness. Postmodemist thinking in philosophy and the humanities argues against universal visions of science and systematic approaches to knowledge. there can be no argument against positivism as . The relativistic stance of postmodern philosophy has been adopted by postmodernist anthropology (see Clifford and Marcus [1986]. when There are varieties of postrnodernism. Fox [1991] and Ma. interpretivists like Geertz silence the voices of the Ot~er. Henn. I): In their writings. In philosophy. Postmodernism and Radical Constructivism Hermeneutic Impact on Anthropology In contemporary anthropology. ~nd secular ~ystems of beliefs in different societies. Paul Ricoeur (1981) and other continental philosophers worked on projects similar to Gadamer's to extend hermeneutics to cover any understanding of meaning in the perspective of text analysis. DolgIn et al.reus. however. This line of thinking had a major impact on the work of Alfred Schutz (1964) in sociology. and homogenize the partial ethnographIc expenen~es ~y producing abstract. Strictly speaking. it is a continuation of henneneutic concerns with ~e. hermeneutical ideas were taken up by Clifford Geertz (1975. Building on the hermeneutic model oftext interpretation. when a whole culture or society is seen as a collection of texts-that is.eneutic~ ~layed a promment role m ethnographies portraying senses of self: emotl~ns." Geertz sometimes uses the text model in a narrow sense: ethnographers have to rely on field notes and transcripts of the stories told by informants to arrive at an ethnographic interpretation and synthesis (as in the Moroccan merchant's story analyzed in Geertz [1975:7-9]).977 ). with Max Weber. pectations fail. and Clifford [1988] as classic expositions. to ::Ive the puzzles. 1983:30-33). but it has been drawing mainly on henneneutics as a springboard for establishing a conceptually less constrained and more open and pluralistic framework of th~ught. Understanding results in a dialogue between these different traditions. Cognitive coherence is finally achieved or reachie~ed by fittl~g together different pieces of tradition. Clifford 1988:Ch. metaphorical sense of "culture as an assemblage of texts" (Geertz 1975:448. . Husserl's successor. In this chapter I won't go into furth~r details of present postmodern. stressing that interpretation has to bridge different traditions: that of the past to which a text belongs and the tradition of the interpreter that entails assumptions on which meanings to expect.. the incompleteness. In this respect. Lyotard [1984]. The task of the anthropologist is not only to unravel the meaning of written and verbal messages she or he has collected in the field.' In his model of ethnographic understanding. But its main Use is in the wider.one of man~ tra~i~ions if we hold to the idea that there is no metatheory. Seidman [1994]. and Western rationality in particular. It is interesting that Ihe postmodernist propoSItIOn of the hlstoncal limits of knowledge is a generalizing statement at the metalevel. and following attempts at resolution by ethnographers who try. an ensemble of socially produced and shared meanings. 1983. Geertz's (1975:5) programmatic statement on "thick description" as the goal of interpretive ethnography echoes Dilthey's separation between interpretation and explanation: "Believing. Michael Agar (1986:Ch.st .1VISIOns of reality. but also to understand what whole cultural scenes and the whole culture are all about semantically (Geertz 1983:Ch. which uestioned interpretive accounts of Ihe Other and found claSSIC InterpretIve ~thnOgraphies wanting (Marcus and Fischer 1986:Ch. Since the mid-1980s it has given way to postmodernist thin~ing. Ironically. and Welsch [1996] for appraisals ofdifferent positions in an evolving literature). developed a radical version of hermeneutics that tried to eliminate any remnants of scientific/technocratic rationality. [1994]. Ethnographic knowledge grows through a series of cognitive breakdowns. By reconsidering Greek hermeneutical tradiHons. His thinking was influential in the human and moral sciences. whl~h are art ofheteroglossic fieldwork. that man is an animal suspended in webs of significance he himself has spun. totalizing monologues.losophlcal hermeneu~lcsstrongly Influenced interpretive and symbolic anthropology (for exampl~. and the limited validity of any system of knowledge in general. see Safranski [1994]). 1995) and his followers in attempts to establish an interpretive anthropology. Henneneutics as text interpretation is less prornlOent now than It was m the 1970s. religIOUS. The contrast between understandIng and explanation is still debated by different methodological frame~orks and WIll be addressed below. 2. On the contrary. too (see Derrida [1967]. only multIple and part. 1.Heidegger's pupil Hans Georg Gadamer (1965) expanded hermeneutics into a general theory of interpretation for the humanities. that is itself a claim for a universal truth and thus contradicts its own premise. 3). and the analysis of it to he therefore not an experimental science in search of law but an interpretive one in search of meaning. Ideas ofph. 1988. then there must also be room for the positivist game ?f science. Mareus and Fischer [1986]. If "anything goes" is the metarule for the pursuit of knowledge as some pastmodernists hold. Heidegger attempted to reach a deeper level of insight into the essential nature of things and human existence. Martin Heidegger. in anthropology the postmodermst View IS often combined with a dismissal of positivism (see. but less so in the social sciences (on Heidegger in his time. Friedrich 1992). Rosenau [1992].. Seidman and Wagner [1992]. Dirks et al. [1994] as newer discussions).48 SCHWEIZER Epistemology 49 by rules of meaning.

More complex methodologies have been proposed. sufficient understanding of what scientific research is all about." ha~ ~merged recently as a new methodological framework that . It questions ~he . I introduce the background and basic ideas of these developments. shifting the focus from the test of Isolated hypotheses to the development of whole theoretical networks (Lakatos and Musgrave 1970). I' History of Science and the New Structuralist Theory ofScience . this new structuralism attempts to integrate conceptually the "normal" development and testing of hypotheses within a leading theory (paradigm) with the overthrow and succession of scientific theories during times of scientific revolution. I will elaborate on this below. in the sectlon on key methodological problems. The focus on subjectivity of experience is a comm~n thread with henneneutlcs.exIstence of a world out there and conveys the instrumentalist view that all SCle~tlfic kn~wl~dge is just the imagination of a scientific community and has no relatIOn to o~Jectlve realit~ at all. sociology. this structuralist theory has no relation to anthropological structuralism). systematic methodological analysis that specifies the necessary conditions of scientific concepts with the reconstruction of specific pieces of research to gain a deeper. Periods • Wolfgang StegmUller (1979a:79). After Kuhn's work.recelved sC.lentlfic procedures does not imply a tacit beliefin reality. and anthropology of science have become Important corre~ltves of a ?ormative philosophy of science (see Franklin [1995] for an anthropologIcal overvIew). and Balzer [1997] for critical appraisals. This has been emphasized in a conclusion to the Kuhn debate reached by l . But the significance of the methodological distinctions and the value of the rules of method laid down in these treatises must be judged against the specific background of the questions and answers established by a particular scientific community." when hypotheses are tested and falsified in the framework of o dominant theory (paradigm) are punctuated by periods of revolutionary change in a h . Balzer. Balzer and Moulines [1996]. l l l l l General Versus Specific Philosophy of Science l The Challenge of History and the Rise of Sophisticated Methodologies In the last part of this assessment of methodological frameworks I sketch some recent developments.makes claIms sImIlar to postmodernism (Glasersfeld 1995). A structuralist theory ofscience has emerged as the main (and powerful) approach for rationally reconstructing what scientists do. y y f "nonnal science. In recent year~ history. Hlstonans of science like Thomas Kuhn (I 962) showed that some of the best natural scientists didn't always follow the strict methodologIcal ruies prescrIbed in positivist textbooks of scientific method. that arrive ~t synth~ses of former debates and lead to solutions that I later propose. mostly within the analytic philosophy of sci. While radical constructivists emphasize sound observatIOn (and ther:byestablish some common ground with positivism) this respect fOf. some of the more naive methodological prescriptions of positivism had to be modified or abandoned entirely. Applying a set-theoretic fonnalization of the theoretical core of any science. (See StegmUller [1976. 1979a]. dominant theory-the so-called paradigm shifts~periods when researc practIces are changing and don't confonn to the strict rules of scientific method.nce. It has been most productive to combine general. approach and thereby use the hermeneutic insight (Gadamer 1965:281) that It s hard to Judge events of the present compared to distant events of the past. In partIcular. In the theory section of this chapter I follow the structuralist concept of theory that overcomes some of the difficulties older. Moulines. their dismissal of realism has few con~equences for research and can be neglected to a certain degree. Key Problems in Anthropological Epistemology and Methodology Three issues have figured prominently in discussions within and between the scientific and the humanist methodologies in anthropology: (1) the problem of l l l • . "Radical constructivisffi. I focus instead on the more developed hermeneutic ideas as examples of the humani~t. Later. First. including positivist methodologIes encountered when they were confronted with and challenged b th hIStory of science. It ~hould b~ mterpreted Instead as adherence to the norms of scientific communities IrrespectI~~ of whether reality exists or not. I tackle the reality and relativism issue addressed by all methodological frameworks. and Sneed [1987]. This framework integrates many of the earlier insights ~f analytic philosophy of science and of history of science in a new syntheSIS (including Lakatos [1970] as a less formalized forerunner). c~rren~ The conclusion to be drawn from this on-going research is that the precise concepts and results of analytic philosophy of science (and of any other epistemological/methodological framework) can be used as guidelines and critical yardsticks to discuss methodological questions in any discipline. Kim [1991].Epistemology 50 SCHWEIZER 51 methodology. Given the commitment of radical constructlVlsts to established scientific procedures. but debated most intensel b postmodernlsts and radICal contructlvists.

Even Weber recognized in his work that scholarship is embedded in and depends on a background of ethics and normative judgments. Still. Sahlins [1995:191-198] crIttqumg Obeyesekere [1992]. However. On the other hand. the verstehenlexplanation problem.) Whether we follow Weber's advice to separate factual and normative statements analytically (as I would do for the reasons he gives. But. These problems pose factual questions. Realism. and times. these questions have to be answered first." If we're too commttted: Weber cientific "objectivity" (or intersubjectivity. on the other. and Truth (Key Problem I) Is there any objective reality? Can there be objective knowledge? If so. unintended and unwished-for consequences. Most adherents of the scientific approach adopt the stance that there is a real world out there. (2) the problem of understanding and explanation and their relationship. we still have to solve the reality problem. one could still conduct positivist-style research (as a solutIon to . A fourth problem. factuai and normative statements should be kept distinct and should never be confused. generalizing research? There is no solution to these problems on which all the methodological frameworks agree-in fact. the better we can change reality in principle by manipulating the causes producing the undesired or desired effects. and has special relevance for law~seeking approaches: How can we construe a concept of theory that's productive for empirical. . In the discussion. Weber seems to have held the pragmatic view that when we put too much emphasis on nOInlative commitments. in what sense? How do we establish truth?-these are the leading questions of this section.52 SCBWEtZER Epistemology 53 realism. we might leave ~ictim to wisbful thinking and distort "reality. and even if we let our normative commitments guide empirical research closely. see below) aside and . but there is less controversy today about ethical prmclples than about the practical applications of ethical norms. and the problem of theory con~ struction. then we d?n't need an elaborate theory suitable for producing generalIzatIOns across dIfferent cultures. In Weber's view. and the factual knowledge that enables our successful moral intervention has to rest on sound logic and true facts. the kernels of truth in Weber's position are (I) that normative and factual statements are different. he included norms and values as part of the subject malter of the social sciences and knew that research is guided by normative rules oJmethod. and (2) that moral commitments that aren't rationally controlled can distort one's analysis. relativism. 1917) for value-free SOCial SCIence. there's some consensus that these points are indeed key problems of contemporary anthropological and social science methodology. In my view.problems tw~ ~nd three) without accepting a realist position in problem one. Relativism. However. Furthermore. in the analysis part of a scientific investigation. Therefore (and this is the core of Weber's doctrine of value~free social science). To be sure. The first problem is epistemological and asks whether it is possible to have true knOWledge of the world: Is there a world out there and how can we know anything about it? T~e second problem is methodological and addresses the goals (descnptlOn/explanatlOn) and procedures of empirical inquiry: What do we mean by verstehen and explanation and what are their consequences for the pursuit of empirical knowledge? The third problem is methodological. but wanted to keep separate the language of facts. the problem of value judgments and applied knowledge centerered on Max Weber's plea in the first part of this century (1904. be confused with n?rmative sta~ements ("What should be the case?") that can only be deCided accordmg to religIOUS. in the pursuit of scientific research. factual statements ("What is the case?" and "Why is some~ thmg the case?") that can be checked by logic and empirical validation shouldn't. political. The problem was a focus of intense debate in the past (Brodbeck 1968:Sect. However. blaming the latter for distorting the historical record for emotional reasons. and (3) the problem of theory construction and theoretical progress. . we might disregard logic and factual evidence that work agal. the humanist approa~h). changing causes can still be difficult to achieve in practice and can also raise further. almost nonrational decisionist approach to ethical questions) or whether we strive for a closer connection between both types of arguments (as adherents of advocacy wouid do. on the debate in general. for example. see Scheper-Hughes [1994]). and truth. He didn't consider these points problematic. there's no question that a theoretically informed social and political practice is better than practice without theory-which raises the question of what a good theory is and leads to a key problem of method. but without agreeing to his extreme. (Rawls's [1971] theory of justice is an example. the wish to do good necessitates the ability to do good. on the one hand. too. I move from discussion of the realist position to procedures of empirical validation. . societies. see Borofsky [1997]). if we confine anthropology's task to describing partIcular cultures and societies (in essence. or other dogmas beyond rational reasoning. But there are many more links of reasoning between factual and normative statements possible than Weber could know and we can engage in a much deeper rational discussion of nonnative and ethic~1 issues than he had envisioned in his decision~making approach to ethics.thereby fall feared. and the language of values and nOInlS. StegmUller [1973:46-M] is both a rational reconstruction and a devastating critique of this and other aspects of Weber's thinking on value judgments. their position toward these problems establishes their methodological differences. as radlca~ c?nstruct~vlsts do.n~t ?ur normative convictions (see. the better our factual knowledge. I1). Then I address radical alterity as an extreme example of relativism and a possible obstacle to cross-cultural understanding.

Empirical Validation H n we validate empirical statements? The scientific approach was invented ho~c:t as a rational means for systematically generating validated knowledge. when it com~s to interpretation of texts or discourses. do not accept the realist background theory but adhere to strict methodological procedures of logic and observation. th corirical hypotheses and theones and thereby to Increase e tests 0 f emp . Since all our knowledge is fallible and depends on observations that can be ultimately false. But it's hard to think of a rational (that IS. re ~wo methodological standards are minimal and thus ne~essary for reachln. and some interpretation based on previous preconceptions are always part and parcel of data collection and analysis. belief in it is backed by pragmatism: it's a simpler. that all experiences and ensuing data are of equal value and that we can't improve our knowledge. Thus. and when they do. . and it makes life easier to assume an objective reality that accounts for intersubjectively shared experiences. . 6). I " I' that growth of systematic knowledge in anthropology. as a rational discourse. we can't prove the truth of this claim and convince a skeptical relativist who thinks that everything "out there" isjust a product of our imagination. Neither naturalistic descriptions of ethnographic experiences grounded in open-ended interviewing and participant observation nor the results of systematic interviewing based on pile sorts or triads tests. but nonetheless compelling. . but are our self-made instruments of thought: this has been clearly seen by the idealist.Cs a particular cultural tradition of rational discourse.ghe~ authorities (although intuition may be important in the discovery of sCientific Ideas). We can compare the Infonnatlon content of different statements. d.fi u dures of validation. and thereby increase our understanding of the world. test them as to their truth. falsification is crucial. capture reality in raw condition. d the public support it has achieved IS due mamly to ItS effiCiency for SCience an . 2:Ch. lntersubjectivity means that the concepts and procedures applIed by a researcher should be public so that other members of the scientific commun. some reactivity. Some selection. logiC and empirical evidence and can't be finessed by recourse to mtultloo an~ hl. . We also have to make decisions and act as if there is a world out there.ysls and j. one-to-one representation of the world (as the "fax" model of socialization assumes. these mini. for example. . our own ideas. there is som~ recognItIOn . they are not forced upon us. ThiS entails the Id~a that in science. individually and culturally variable preconceptions (concepts. . the infonnation given should enable others to rephcatethe study. This does not mean.um standard~ of clanty and empirical validation. however. In this respect. theories) are part of our experience of the world and structure our expectations (see D' Andrade [1984] on culturally constructed things).. While these ideas have been given prom!~ence In the analytic tradition of the philosophy of science. that there is something to remind us of the fact that our ideas may be mistaken. of the claims of truth and the logical structure of arguments. As Popper points out: "Theories are our own inventions. However. Although we can't prove the realist hypothesis. hypotheses. In everyday life.lty can understand and control the logic of arguments and the weight of the emplrlc~1 eVidence. . . the ultimate impossibility of understanding the Other (person culture society) is sometimes mentioned as an obstacle to the . common sense and shared experiences tell us that there's something objective besides our individual experiences.I~~~~ t~ other cultures and its use isn't Iimit~d to Westerners. the re atlVlst calm l l l l l ~l l l l l l l l l . the real and the ideal are mixed in human perception and representation (Putnam 1981. However. like empirical tests. So. truth claims must ~e J~~tlfied bY. . . 10 t e ~ .stmodemlst texts are difficult to comprehend due to the metaphorical language and Imagery preferred by their authors but no doubt they use rational arguments to conVince others. (-) Procecnttclsm . The diffusion of dt. Empi:ical procedures for checking and validating scientific knowledge are specific for each discipline (see. In addition to this universal distinction rooted in the common nature of all humans. and (2) is therefore not a direct. and. see Strauss [1992:9]). we know that there is a reality. Werner and Schoepfle's minimum standards for ~thnog­ raphy [1987. S ondence of our preconceptions with reality. The objectivity (read: mtersubJectlvlty) ofSCientific knowledge rests on these standards.of these standards in the other methodological frameworks. sound knowledge The scientific approach IS deslgne to provl e s r pro ducmg .g thiS ob'ective: (1) Clarity of language is meritorious because It enhances ana.comm nities as controls of the validity of empirical statements. 11]). Human beings seem to make a difference between individual experience (inner states) and something objective out there as the trigger or object of experience (Kutschera 1982:401-405). Its standards have A. d 'd tong . must be estabhshe . doctrine. In sCl~ntl lC . logic and evidence count. The Impossibility of Radical Alterity Radical alterity (Keesing 1994:301). we routinely evaluate statements to determine whether they're purely subjective or point to something shared or real. Lenk 1995:Ch. But some of these theories of ours can clash with reality. radical constructlvlsts. more parsimonious hypothesis. And that is why the realist is right" (I 969b: 117). Some henneneutlc and po. we must be cautious and understand that our experience of the world Is (I) framed by learned cultural conceptions. sCientific) approach to knowledge that does not adhere to.Vol.54 SCHWEIZER Epistemology 55 Realism The so-called realist claim that there is a world out there that exists irrespective of our imaginations and subjectivities is a metaphysical.:. In principle at least. But.

The next' key methodologICal problem tackles more concrete issues of how to conduct researcli.IOI]). explaining a certain relaxation of behavior as the consequence of alcohol consumption).charjt~bJe way. in fact. meanmg " of all the . T~e construal of radical alterity is an extreme position. but these are contrasts which.on [1984:IOQ-. I can clarify the more complex concept of empathic understanding (verstehen in a strong and special sense) that interpretivists advocate as the method of verstehen in the humanities. However. Determinative understanding: knowing the meaning of an event or process by referring to its aim or the pattern that it belongs to (for instance. It opens up a range of different. or forever beyond rational resolve-then I reject conceptual radicalism. because one knows its causes (for example. the chief witness cited by Geertz in the quote above. can differ massively-to the extent ofbemg mU. for this seems (absurdly) to ask us to take up a stance outSide our own ways of thought. Verstehen/understanding can refer to the following (Kutschera 1982: 8Q-. Searle [1995]. when he wrote about "scientific understanding" and characterized sociology (1972: I) as "a discipline that tries to understand social action interpretively and thereby attempts to explain its process and consequences causally. this prmclpled reason agalDst systematic research in the humanities is gone. The argument is self-defeating. the ordinary and only feaSible strategy used by ethnographers in the field.stand the Other (say. Trouble comes when we try to embrace the idea that there might be more c?mprehensive differences. . didn't see a stark contrast between these concepts. StegmOller 1979b). 2. and person to person of kinds which we all recognize and s~ruggle With. understand the Other has to be rejected as logically unsound. m Davlds. He envisioned an integration of both. it might seem absurd to look for relationships between interpretation and explanation. If 1 don't. on relativism in general see Krausz [1989].plam and understand. The second task is to discuss the relationship between both approaches and to problematize the idea that the distinction of interpretation versus explanation is congruent with the distinction between the human and the natural sciences. d behavioral sciences pursue the search for laws in culture and society and try to :Plain cultural and social phenomena by recourse to laws. because I've just manag:~ to understand the cad. se: the lucId diSCUSSion of the tenn "hippopotamous" used in a strange sense. making it difficult to assess their specific claims. In hermeneutic methodological treatises (for example. Against this background. The first task is to nravel the meaning of the concepts of verstehen and explanation to better ~nderstand the kind of methodological procedures and theoretical objectives underlying both. in principle. Of course there are contrasts from epoch to epoch. m prmclple. When reading Dilthey or Geertz. The statement that 1 can't. bU~ the principle of charity is the only practical way out (p: ~9]. According to Donald DavIdson (1989:159-160).m I owe these fundamental insights on the ImpOSSibility of radical altellty (which he calls "radical conceptual relativism"). and Harre and Krausz [1996]).j 56 SCHWEIZER Epistemology 57 ~ue t~ o~r different socialization and different individual experiences we can never. Charitable translation is. under. e l r meanmg IS totally dl~erent from the meaning of comparable terms that 1 can Use to express hIS or her Ideas. (The relevance of Davidson's philosophical ideas on anthropological debates has been further explored by Bowlin and Stromberg [1997]. from ~ulture to culture. But difficulty should not be made an obstacle in principle. Weber for one.2. We can never be sure that we capture the meaning of ~trange words m the fight sense (the principle of the indetenninacyoftranslation Qume [1960:Ch. 4 True. . Positivists in the social example. Understanding a/meanings: knmving the significance of terms or gestures (for . partially related concepts that need to be distinguished. or the languages associated with them.terms that the Other uses. liD. 3. with sympathy and effort. If by radical conceptual relativism we mean the idea that conceptual schemes and moral syst~ms. we can ex. mtelpersonal and cross-cultural communication can be difficult. Gadamer 1965). who stress massive methodological differences between both kinds of disciplines and claim verstehen as an exclusive method for the study of culture and society (see the quotes above). Causal understanding: knowing why an event happens.leal Other. What we m~st do IS try to interpret and translate what the Other says and writes In a sympathetlC.84): I.tually unintelligible or incommensurable. This tacitly presupposes that 1 first have to completely understand the Other's words to make the outrageous claim that I can nev understand him or he:. to who. when one observes a sequence of actions in a particular culture and knows that it is perfonned by the actors as an initiation ritual). 8 ut once it's refuted. Dilthey 1900. So these texts often have a rather opaque quality.. some utterance or text) presupposes that 1 know the 1 can't tell that th ." Varieties of VerstehenlUnderstanding The meaning of the Gennan word verstehen (translated as "understanding" in English) is less dear than often thought (Kutschera 1982:ch. I I Interpretation or Explanation? (Key Problem 2) Intelpretation-in the sense of empathic understanding (verstehen)-has been proposed by hemleneutlCsas the method of the humanities. the different senses of verstehenlunderstanding aren't always distinguished. when one khows the content of specific words used by a speaker in a particular context or a writer in a text). which is the claim that this argument denies.

to t "e lassification of the different meanings ofverstehen m the prevIOus section)..: heart of the verstehen approach 10 past).58 SCHWEIZER Epistemology 59 4. t k Gaining this knowledge would enable the ethnographer to actIOns h:n~:nd adopt the native standpoint to a certain degree. as participant observer) are in. what role a certain job plays in a production scheme). arc Igger. . Lastly. as proposed by henneneutic philosophy (Gadamer 1965. interpretivists propose Ihat the researcher engage in a thought experiment and try to grasp the actor's point of view by acling as though he or she were in Ihat situation.) Basically. knowle ge . Genetic understanding: knowing how an event has become transmitted as an outcome of a temporal sequence (for example. Rational understanding: knowing the reasons why an actor chooses a certain behavior given his or her preferences and his or her knowledge of the probability of outcome (for example. ThiS calls for a larger synthesis of the data on meaningful action across different types of events and would contain empathic understanding of particular events only as a step in a larger hermeneutic project (see 3. (-) ~he kind of situation. some of these concepts are process/product-words that refer both to the sequence of generating knowledge and to the results of this process. why an actor buys or sells shares in the stock market depending on his/her calculation of profit and risk). ourselves. Interpretive ethnographic writers mostly adopt a 2. 1979b]. Thus. She or he would capture the thinkings/feelings actors associate with the whole situation and the particular ethnography. In t~es~ ~escf1Ptl?n~. l l l l l l l l Verstehen as Empathic Understanding What does "the operalion called verstehen"(Abel 1948) really mean. empathic understanding is invoked when we want to know the reasons or purposes underlying the meaningful behavior (including speech) of other people in the present or past. The goals of interpretive anthropology. this Ihought experiment has to rest on validated ethnographic or historical background knowledge or risk distorting the decision siluation of the actor and mistakenly take the selffor the Other. Then they draw on the background of actors. More concretely. amic concepts).:hat we see p~op e ~ ~hat it is that we need to know in order to participate. 6. they trace th ac'tions taken by the dramati~ personae and the collective outcome of often divergent strategies of actors guided by different interests. (3) the reasons. .:nes is at tht. languag~ (the experience·near concepts mentlOne y ee z . ~ scription as the correct prediction of what the actors do and the ability t~ ~ct a: the natives do in a particular situation: "We record . Ricoeur 1981) and interpretive social science (Geertz 1975)? (In this discussion. These different usages of verstehen/understanding don't exhaust the whole semantic field. many of these different concepts can co~occur in research. because it favors observatlon·near. db G rt [1983'56j)andtneslodeplctthe l l 1 l . b' and not . interpretive eth~ogr~phers . She or he ca~ ~~:P~:scribe a particular situation "throug~ native eyes"-although thiS d will always be incomplete and fallible.c d what they say about what they do and then try to mfer from t~IS recor. h t h s to "understand the native's point of view. . and (4) the intentions of actors ~ a minimum. Goodenough (1994:264) states Ihe goals of such an em." He emphaSizes t e ve:s e en meanmore than the einfilhJen (empathy) part of empathic und~rstandmg and part t ts the whole task by distinguishing between expenence~near and recoe~~e~cce_distant concepts. I he ~oal~ include understanding the whole fabric of society and the ~hole ~nsem e 0 cultural texts along the lines ofhermeneutic text interpr~tatlon. Functional understanding: knowing what something contributes to the working of a larger system (for example. as we will see later. StegmUller [1969:360-375. I possible by giving information on concrete cases. let's look at ethnographic research and break down the task of reaching empathic understanding into three parts: 1. and they are not restricted to the humanities. Use of native terms and quotes from key actors creates a sen~e of authenticity in these reports. 7. h t " oint of view" as closely as naturalistic style and try to convey t e na Ive s p . however. In this case. restricted to understanding meaningful action in particular events. empathic understanding means that the researcher starts with certain events or behaviors of actors and tries to elicit and document the "subjective" definition of the situation the actors (and the ethnographer. t' ~ Hows Geertz's (1983:56-57) discussion of what It This charactenza Ion 0 .search. Description of results. Also. ethnographers usually sketch the larger context as a stage for m~lvld~al actl~ .. explained above are par and precondition of gaining empathic understandmg: one ~ould need to learn ~ know (1) the meaning of utterances (the num~erlngs he:e refer. To be sound. 5. In what IS going on and do so in ways that those people accept as showmg we are knowledgeable. WilIiam Whyte's (1943) snapshots on the hfe of .for eX~scriPtion in termS of concepts that are close to observation (mcludmg natlv~. In his view. In contrast to rational understanding. . how the custom of decorating Christmas trees in the present goes back to traditions of the North European . intentional understanding does not invoke the idea of an efficient means/ends scheme-any purpose will do. Intentional understanding: knowing why an actOr chooses a certain behavior (for instance. decision situation faced by actors as closely as possible. understanding the meaning of words and rules governl~g particular cultural SCt. . . The discovery process." t d Several of the senses of understanding.. In fieldwork. In fact. below).. th~ir sen~esofthe SItuatiOn they r: in their thoughts and feelings and their strategic chOIces. why an actor engages in physical education in order to achieve better health).. and Kutschera [1982: 132-149]. I am infonned by the rational reconstructions of Abel [1948J. Thi~ type of report IS a so called phenomenological.

in the text. . methodologisls of the humanist approach ollen mention the so-called hermeneulic circle as a special precondilion for acquiring knowledge on humans. Lila Abu-Lughod (1991:149-157.investigations would be enhanced if we could engage In generalIzatl?ns ~cross dIfferent lime frames and detect stability as well as cha~ge aI each polOt In tIme. New York. 139-162. We have to know the meaning of the parts (say. At the descriplive level. feel. 3. it's nographles of the partlc~lar. world) in a d. In ethnographic reports. 4). the one that better predicts what the actors do in a particular situation is preferred. PP: 49-51~. As Geertz notes (1995: Ch. 3. multiethnic context in Sambia.mgful action. ethnographic generalIzatIOn could caplure the general and t~e particular and need not homogenize ne. mt~~estmg ~nd complementary to generalizing approaches to portray the particu- 1980:Chs. class. In this sophislicaled way. 2) text . 10 understand Ihe meaning of the parts. the pro~lem of generalization looms large in ethnographic case studies and IS not easily solved-but it must be. Clyde MitcheII's (1956) analysis of Ihe Kalela dance in an ".Ived experiences of concrete actors at particular times. we can enhance the validity of ethnographic knowledge and better assess the typicality of interpretive state· ments by combining qualitative case material with I) verbatim recording and transcriptions of native consultants. then. I) IS that a concrete case should reveal some general cultural truth -how. reg.Nde~bu In the same country. In ethnographic descriptions. oIherwlse. I conclude this section on empathic underslanding with the folloWIng pomts: I. 1986. empathic verstehen arrives at ideas as to how the different parts of a text hang together and exhibit some coherence. The stories they tell are selected as examples of some cultural or so~131 tendency and. HermeneuIic work on texts or discourse. 2. Romney et al. and Clifford Geertz's (J975:Ch. ~an display lantles of 1. the interpretivist has discovered by empathic understanding. undertaking. Werner and Schoepfle 1987:1. we also need to know the Whole of Ihe text as background knowledge. the Javanese think. But. It belon~s to th~ c?ntext. Most researchers who adopt this naturalistic strategy of ~es~r. both should be logical and.nalysis of transcriptions. of a case-centered verstehen approach. In text interpretation. 6) a Javanese town at the apex of political conflicts are classic ~nd con~i~cing examples. . Ihe dIfference between the particular and Ihe general does not become revealing. local commumty. agamst ~eneralizationsi~ anthropology and has launched a project of writing "eth- Recently. behaviors. the point of Gccrtz's notion of thick description (1975:Ch.on. It s basically a procedure for finding out what other people think and feel an? how their subjective (but to a certain extent shared) beliefs structure rnean. ~ discovery of (interpretive) hypotheses. The Hermeneutic Circle Along wilh empathic understanding. and 3) numerical data on the frequency of ideas. In the context ofjustification of results. Thus. explicit generalizations can be conducted as controlled multilev~1 p:oced~res that. 1993:lntroduction) has argued Geertz ~ shortschrIfts on Ihe Javanese or Ihe Balinese). words in a texI) 10 understand the meaning of the whole.lbmg others in concrete situations are not content simply to describe Indlvl~ual cases. the interpretive statements should be logically consistent and in accord with the :acts. Bernard 1994:Ch. Victor Turner's (1957) descriptions of s~c~al dramas as ~equences of conflict and conflict resolution among the matnhneal.oCletal mtegration and observation (family.urpose of demonstrating by concrete examples how the Other acts and reacts ~n natural situations. is justly praised for its combination of actors' verbatim quotes and ethnographic observation ofactions and events these people are involved in. in discourse analysis. apart from rendering inf~rmation on central tendencies.fferent lev.tudied.g~~orhood. In and reify the different persons ~nd un~Is studied. the products of empathic understandmg have to conform to ordinary standards of scientific validation. What is a hermeneutic circle? Gadamer (1965: 25G-283) and other proponents of hermeneutics point to a special melhodological dilemma on which each understanding of a text (or utterance) depends. offered to the reader as typical of the larger system descnbed and as revealing of salient andlor contested issues in the society or (sub)cll!turc s. control of the adequacy of interpretations is often difficult. ethnic group. But. Furthermore. Trying to grasp the native standpoint is the gist of the verstehen approach. 7. it's hard I~ beheve how she can depict the individuality of particular OIhers without renderIng mformatlOn on the general ethnographic and hislorical context Ihat gives each ~elecled persona her or his particularity. When there are rival interpretations (see the example of dissent in text interpretation presented in SlcgmOllcr [1979b:46-53]).60 SCHWEIZER Epistemology 6J a street gang in Boston.IIn most ethnographic research this is a useful starting point. Many recent mterpretlvlstlpostmoderOlst monographs quoted above (most notably Kondo [1990] and DumoDt [1992]) contain ethnographic vignettes that serve the same ~. Gene~a!ization. Some sort of generalization is unaVOIdable.~perslOn at the d. for example. as fieldwork methods texts point out. . but it's not necessan y the only technique of data gathering and the end of ~nalysis." She is righI in criticizing Ihe flat and abstract genera~lzatlOns apparent m much of interpretive elhnographies (for example. However. Empathic understanding is mainly a powerful heuristic strategy for detectm~ patterns of meaning in texts and discourses.rban. 6. Philippe Bourgois's (1995) in-depth study of crack dealers In East Harlern. Thus. In ethnographic applicatIOns. and observed events (Agar study of a rItual 3. case material illustrates some general POint. and act. the one that makes sense in more instances is preferred over the other (Ricoeur 1981 :21 0-215) or. state. . has to shift between these two opposite sides of one interpretive sensltIv~ way.els of s. Th~ v~lue of such .

Let p be the descriptive statement specl . They aren't restricted in their truth claim to specific places and lImes (although one could use quasi-laws that are restricted to certain regions or epochs to ~xplaln phenomena within these boundaries). evolution) is often part of ordinary work in the natural sciences. that Sawahan IS strattfi~d. q. this isn't a vicious circle that we cannot evade or even a principal problem that we cannot solve since we must always distinguish between background knowledge (assumptions) and claims that can be checked. one can then logically conclude that p IS true. Thus. ~::cri tive statement(s) specifying that the antecedent condition q of the lawhke state:ent is the case.any~ng . Pitt 1988. We could' use the lawlike statement from cross-cultural researc~: "If ~ co~mun~ty has intensive agriculture. causal (sense 3) and functional understanding (sense 6) are definitely part and even genetic understanding (sense 7) (for example. of this distinction. then it has social stratification" (q--p) . terms. the word explanation has different meanings.. called the explanans e th(th exth~:gn~:i:. Some also apply to the natural sciences. As an example of a deterministic law. The Concept of Scientific Explanation How can we explain explanation? Like verstehen!understanding. f 109 . above). The answer. [ have to clarify the meaning of explanation itself. as some leaders of hermeneutics claim. It applies to any knowledge-making situation. statements are true or at Ieas t we 11 con firmed descnptlve I . some overlapping with the senses of verstehen discussed above (HempeI1965. every element of background knowledge can be criticized (but not all at the same time). Stegmaller (1979b) demonstrates this by comparing an interpretive debate on a medieval poem. . A valid explanation presupposes that the lawlike statement and a~co~p. to an astrophysical explanation. is represented by two types of statements: at leas)t e lawlike sentence of the con d·lllona . The social sciences fall between these extremes or be long to either the humanities or the sciences depending on one's methodological standpoint.62 SCHWEIZER Epistemology 63 Crilics ofhermeneutics (Stegmaller 1979b. he or she has to classify events (sense 2). Kutschera 1982:Ch. the meaning dimension is unique to human studies. It s ca e pheno~enon d (Latin ". Deterministic laws or lawlike statementsdalm that the consequence happens without exception if the antecedent condItIon IS fulfilled. So. .fy. but not as validated as laws. the difference in kind between the presence of the henneneutic circle in the humanities at one extreme and the lack of it in the natural sciences on the other becomes a mere difference in degree. whereas this distinction is often blurred in the humanities. Bartelborth 1996). but there's a qualification. The lawlike statement can be much more complex than ~he e resented here that postulates only one condition. Given th~ lawhke ~~le~ent(s) and the statement(s) on the antecedent condition(s) as premISes of an explanatory argument. Here. th t th· exp . In the Javanese community of Sawahan that I observed. take the baSIC demographIc equation: population change in a certain time period = blrths-deaths + 1 l I I l I ·l l l I l l - . a student of nature has to understand the meaning of words in scientific texts (sense I. then p" and one (or more .p. social. reco~ ex lanation is the prototype of scientific explanatIOn. In words: if q and (q Imphes p). rtam phenomenon X .' The conditional form of determinIStic laws rules out historical inevitability. and behavioral sciences. in" 1. And the fact that human subjects can rellect and purposefully change their behavior must also be considered. The pragmatIc distinctIon between laws that are well confirmed and lawlike statements that are formally similar. The descnptlve statement' "The village Sawahan is a stratitied commuOlty IS the explanandum. He're~ a simple example. rests on this fact. The next question is whether meaning and reflexivity exclude the search for laws in the human. Second. The conditions specified in a law or lawlike statement of the deterministic form shown above are sometimes called "real reasons" for the occurrence of t~e phenomenon that is to be explained. Stegmaller's (I979b:58-68) main finding is that the distinction between background knowledge and interpretive hypotheses can be drawn sharply in the natural sciences. because the consequence is only predicted to happen if the antecedent is true. g wh"'-question that asks for the occurrence or some feature of a Iona/IOn-see In . We needn't make to~ m~ch". .979. It answers the D P k. Before we look at this. 1 '. We might wonder why there is social stratificat~on. . Stegmaller 1969. with a closer look. The different senses of verstehenJunderstanding outlined above aren't restricted to the humanities. However. this dilemma is not exclusive to the humanities. 11 I focus on DN (deductive nomologi<:al) explanation (also called the HO speclfica Yf' explanation after Carl Hempel and Paul Oppenheim who first schema 0 tnlcted it in 1948 [Hempel 1965:Cb. ( 1\ (q-p))-p.the eX~laining). [n principle.orm "·f I q.hat is to be explained). including hypothesis testing in the natural sciences. a 11 IS d ce X is the case Since we are looking for an explanation 0 p. Understanding as a Special Method of the Humanities Now we can pose the related and larger question: Is understanding peculiar to the humanities? The general answer is no. In formal . 2. . 10]).I~ cO~Junctlon with the descriptive statement on the truth of the antecedent conditIOn S~waha~ has intensive agriculture" (q) to deduce the explanandum P. [ confine the discussion to scientific explanation. Albert 1994) make two points. sClentlfi~ generalizations is that they claim condmonal truth for an unhmlted set 0 phenomena-that is. First. . The gist of laws and law like statements. More . then p. and in this respect hermeneutics is right. on the other. a. on the one hand. all cases that fulfill the antecedent condition.

. In addition..'t lead to a fertile classification. belong to epistemic reasoning. Kuipers (1996J. . typically USIng laws or lawlike statements) is description (singular statements specIfyIng what the case is in particular entities). depending on the current state of knowledge in a domain. In this new perspective. and other social and behavioraI sciences. we can first use (empathic) understanding to discover hypotheses on the subjective meanings ?f actors and then consider meanings as well as the conditions of the larger context In which actors are embedded in more encompassing explanations of social and cultural phenomena. Hence. both procedures aren't polar opposites (StegmUller 1973. not all the natural sciences can conduct experiments (astronomy.. we're switching to reasonseeking or epistemic why-questions that ask for good reasons to believe that the phenomenon specified in the why-question is the case.' . They use the tenn "cause" in a much more restricted sense for real reasons in laws of succession (when there is temporal asymmetry between earlier causes and later effects in detenninistic laws). including anthropology. (3) A more productive methodology in social sciences attempts to integrate verstehenand explanation. and behavlOral sCiences ave a : imp:ct on empirical research in anthropology. At first we take 'model' in an informal sense . Lawlike statements and laws of a statistical form that postulate a high probability to expect p when q is the case. But this statistical law does not allow us to predict for individual eases whether a boy or girl will be born. h h d .mg devices in human thought (D' Andrade 1995:Ch.Slmple distinction between the natural and the human sciences according to explanatIOn and verstehen disappears. statlS~rw9 ten . "causal analysis" covers both types of reasoning. Over time.~al models . . Even if one were s~eptical of the ·t· . ta et e well-established findings from comparative resear~h on th~ cau. Such questions are answered by epistemic reasons. but take history into acco~nt as well (for example. Explanation is possible in the medical SCiences. generalizing model. A lot of research in analytic philosophy of science attempts to extend the scope of ON explanation and cover more deviant cases in more encompassing rational reconstructions of explanation (Pitt 1988. NettIng 1993) o~ the cognitive anthropological finding on the pervasiveness of sche~ata as orgamz. The necessary minimum conditions for adequate scientific explanations have been specified as a result of these general inquiries. and they are not confined to the invariant aspects of their subject matter (as Dilthey had it). . As an example. Bartelborth 1996). then the consequence happens in most cases. d h . while the antecedent conditions are often called "causes" and the consequences "effects. The philosophical thinking on DN explanatIOn. see Moulines (1996:7].ples.." (Barteiborth 1996:30). . (Note that this is a pragmatic statement. ·ng and the minimum standards of scientific explanation In conJunc Ion lIcal reason! . social.. and ON explanation is considered a special case of a broader concept of explanation. t· . A valid co~trast to explanation (embedding a case in a larger.sal Imks bet:v~en population growth. philosophers of science hold to the distinction between real reasons and epistemic ones. Can we combine them? Understanding and Explanation: From Opposition to Inlegration How are understanding and explanation related? (I) Conceptually. take the well-established statistical law on the sex ratio at birth: this backs the expectation that there will be slightly more males than females at birth. economics. we can only posit that if the antecedent is given. and political integration in human societies (Johnson and Earle 1987.1- Epistemology 64 SC!lWEIZER 65 immigration-emigration (the population change as dependent variable is explained by Ihe conditions in Ihe right-hand part of the equation). economic intensification. Whereas real reasons are reasons of phenomena. a representation of something. (2) The dichotomy of verstehen/explanation doesn. Above. Weakening detenninistic laws to statistical ones leads to a different type of reasoning and leaves the standard model of DN explanation. As exam. I979b). natural history in biology or astronomy).th th the search for laws in the natural. Instead. However. also plays a role in the natural sciences." In contrast. and Balzer (1997:200-232]). In social science methodology. Anthropology need not adopt this special tenninology. Then we're no longer answering explanation-seeking why questions by giving real reasons and using deductive logic. for example). t agenda in the humanities there is a growing body of lawlIke statements POSI I V I S ' k h in anthropology and the other social and behavioral sciences. interpretation. So the . epistemic reasons are reasons for phenomena. The theoretical models include laws as elucidated above (see Bartelborth (1996] ded discussion). in social sciences there are frequently so many exceptions to a Jawlike statement that we have to give up the whole idea of a deterministic relationship and can only claim a statistical connection. but the difference between epistemic and real reasons can be revealing.. In this broader view. research efforts could lead to better approximations by reducing measurement errors and by including additional conditions that enhance causal understanding and prediction. In thiS sense It IOcIu es . We could argue that there is always some error or chance fluctuation in empirical data and that therefore empirical data only approximate detenninistic laws (on approximation in general. mee antea or ::a. psychology. socm} stratlficau~n. "explanations are essentially unifications of our knowledge (and] we should conceive them as embeddings of an E in a model M. The crucial methodologIcal question now is how to integrate the goals and procedures of understandmg and explanation. as well as theoretical models. I explanan durn . 6). 1 tried to show that both approaches rely on rational procedures and have merits. in several of the senses distinguished above. It does not rule out the possibility that in the future we could establish deterministic laws on the same subject matter)..

however. and that the general can thus be abstract. in-depth explanations. and cross-historical generalizations. 1996]. and social restrictions of the context that enables and constrains individual action. and Ensminger and Knight [1997] on how actors use and change institutions. and information on the concrete decision situation of actors. 2. the emergence of new nonns and institutions as an effect of present actions. l l . But this is a promising attempt to integrate verstehen and explanation in a more encompassing explanatory theory." He remarks that thIS Isn t meant to be the mere description of factual occurrences.. which is a neW development in analytic philosophy of sCience. systematic growth of knowledge. G5rlich 1998). Gadamer (1965: 11) also points out that theory looks for the general. l l l l ~l Theory in Hermeneutics In a paper entitled ''In praise of theory. too." the philosopher Gadamer (1983 :~3) tells us that the root of theory is "[t]o see that what is. Then they try to capture common symbolic understandings shared by members of a culture. In my vIe. Then'l present the established view 0 t eones In ana y • IC In case su t hilosophy of science tbat has framed posi~ivist ~iscussion about what a ~heory IS. The refinement of the rational actor model refers mainly to the fact that the economizing scheme has to incorporate social constraints to avoid the flaw of the undersocialized actor and has to include social approval in addition to material well-being as maximizing goals (Lindenberg 1990. Theory Construction and Theoretical Progress (Key Problem 3) Here. When hermeneutic anthropologists compare cultures (mainly comparisons between the ethnographer's "I" and the culturally distant Other. Cashdan [1990]. This theoretical approach challenges the older. although it's very powerful and challenging. I discuss how to construct theory and in what s~nse there can be the~retical rogress-that is. 1 t· . This leads to infonnation on actors' goals and expectations. They also trace the purposeful and unintended collective outcomes of individual choices on the larger system. false distinction between verstehen and explanation and the accompanying idea that both procedures exclude each other. interpretive anthropology seems content to describe particular cultures. Siegwart Lindenberg discusses heuristic strategies on how to cross-fertilize these two approaches to arrive at more precise and more realistic. 44). These social theories use empathic verstehen as a bridge between a very general and abstract economic and social theory. this concept has greaJ potential for rational reconstructIons of the~retlcal work to any science. Following some of Weber's early concerns. l l I l l l. but of something deeper. theoretical aspect comes to the fore when interpretivists try to tell us what it means "to be Javanese. Finally." etc. ~e main point of this section. Ensminger [1992]. on the other.:". First. Nevertheless. in a major reconstruction. including cultural preferences.. f h . Geertz [1995:Chs. Netting [1993] on the rational core of small-scale agriculture). Lindenberg [1989. see Coleman [1990]. It's also clear that anthropological applications of rational choice theory have to draw a lot more on cultural background traditions molding goals and actions in variable cultural and social contexts (Schweizer 1996:Chs. hidden. Thus. one can study the embeddedness of actors in previous institutional and cultural arrangements and. they don't seek universal characteristics of all cultures or humans per se. on the one hand. contrastingly. Working in the spirit of hermeneutics. 1992. though. they investigate the actor's view of his or her situation. a pattem that we can only "see" when we try to get rid of our prejudices (p. t dl·es . By integrating both kinds of information. reval mg III ~ey arrive at more grounded. some generalization is involved at the level of individual cultures when interpretive ethnographers try to capture the essential qualities of a culture. 2. IS to mtroduce the s~ructurahst th~Ory concept. they use contrasts to discover the essential features of the unique cultures that are compared. . not just the surface deSCrIption of a particular case. We needn't accept economic theory in full or at all. 1990. Coleman and Fararo [1992]. society at a certain time). 1996). political. they use a refined rational actor model (as a law or lawlike statement) to explain the actors' strategic actions. This general. This embeddness and emergence is borne out. Lindenberg and Frey [1993]. Then analysts study the economic. but still sufficiently general explanations (1992. cross-cultural. Again. some interesting methodological ideas have been put forward by rational choice theorists in sociology (most notably Esser [1991. Better theories capture subjective meanings (the actors' views) and objective constraints (the institutional context and shared beliefs . 1996). see. on rational choice theory in general. and think of Levi-Strauss's [1963] SimIlar Idea of "structure"). for example. are elements of other concepts of theory (see below. In a diachronic perspective. 1996]. 3. and Bohman [1991:67-76] for its logic of explanation). by anthropological applications of rational choice theory (see. 1 look at theory m herP t' which contains some prehmmary Inslghts and IS useful for generahzmg meneu ICS. sociologist Hartmut Esser (1991) demonstrates how the verstehen-oriented social theory of Alfred Schutz (1964) corresponds with rational choice theory and can be incorporated in rational choice explanations. I will propose this new theory concept that could gUide explanatory research in anthropolOgy and solve some of the methodological problems generated by cross-societal. The stipulations that theory aims to represent something general.1. 1993.3] on Morocco and Java). .66 SCHWEIZER Epistemology 67 In this respect. these analysts begin with the "logic of the situation" at the level of individual actors: using verstehen.

3. the general conce~t of utility I~ economic theory whose significance depends on and is established bY"applymg some kind of maximization hypothesis). of course. but the same Idea should hold: An interpretation that makes sense of more cultural data should be the better. 1993) pleas for ethnographies of the particular (see the section on empathic ungerstandmg. b/: '.lOn ~nd empirical validation. It became more awkward when it had to cope With theo~ change in a diachronic perspective. Balzer 1997:Ch. In questIOn. Theories are conceIved as networks of laws. intuitively speaking. falsified or somehow checked. compared to rival ones. (p.. amongothers.ngs an~ hlstoncal tradItIOns are of prime importance in every cultural and socJaI settmg. How does a theory come into contact with empirical data and how. there is no well-de. can It be refuted or confirmed? Different adherents of the analytic framework dissent The concepts of theory established by the different currents of analytic phil- Elements of Theories in the New Structuralist View Within the structuralist theory perspective. so-called structuralist view of theories has been proposed as an alternative (StegmUller 1976. 70) The main difference among concepts of theory established by rival analytic philosophies of science relates to whether theories are logically connected sets of statements-that is. Basically. 19-:9a. their formal approach builds on Suppes's set-Iheoretical axiomaHzation of scientific theories [1957:Ch. the form 0 f revealed a system of uniformities that can be expressed In emplflca generally. Balzeret aI. They differ in generality and contain observable as well as abstract. the better the interpretation is. Hence. or beneath them. they hold that theOries are complex structures that pose difficult problems of logical ded~ct. I never heard '.a he~eneu~ic text i~terpretation: The more parts of a text that can be covered partIcular interpretive hypothesIs. ASIde from the idea that ethnographers should derive interpretations ~f cultures that challenge fac~ual evidence and that a more encompassing interpreta~ tlOn that makes sense of a bigger part of textual evidence is preferred. . Thus. whereas the "Tnontheoretlcal concepts can be measured independently of the theory (Stegmuller 1979a:Sect.es are usually introduced when previous study of a class of phenomena has 111 eor " I1 aws.eloped theory or growth of systematic knowledge in hermeneutics. 12]). I about this.rml~ed by applymg the theory in question (for example. the fonnaJ structure of theories should be dl~tmgUlshed from their substantive interpretation. (p. POSitivist philosophers of science set themselves the task to arrive at a more predse and comprehensive concept of theory. Rules of correspondence work as bridges between the formal structure and the empirical applications. as statements can they be true or false. Franklin Tjon Sie Fat (1990) has used a preliminary version '~. as it were. theoretical concepts. This version turned out to be very clumsy when philosophers of science tfled to ratIOnally r~construct particular theories. this is not to deny that it is very important for science to make statements-things that ~an be true or false.ctation and key hermeneutic insight tha~ meanl. An in-depth introduction to the structuralist theory concept would require the reconstruction of particular anthropological theories. 2). but I will introduce some of the structuralist concepts and suggest how these concepts relate to anthropological generalizations. But. culture-specific generalization. 9) Theory Concepts in Analytic Philosophy ofScience oso~hy o~ sCience share the common idea that theories should be generalizing d~Vlces al~~~ at il:uminating the particular by reference to Some broader princIples. . to afford a deeper and 111 cories then seek to explain those regularities and. constraints. . or theoretical principles. by means of which the theory then explainS the empirical uniformities that have been previously discovered. T 0 th'IS en d.a 'h eory rate understanding of the phenomena . The general positivist vision of the integrative role of scientific theories is underlined by Carl Hempel (J 966): . a nonstatement. henneneutic anthropology is inf?rmed by th: general expe. general methodological work and detailed reconstructions of specific theories in different branches ofscience have led to the discovery of subtle and precise concepts for describing theories and yielded new insights on the use of theories in the empirical sciences (Balzer et al. or not. and usually also predicts 'new' regularities of similar kinds. of the following kind: that a given domain of intended applications may actually be subsumed under the theory's principles (laws. that can be verified. of course.tr' 68 SCHWElZER j} Epistemology 69 This search for the essential quality of cultures is equivalent to the search for comprehensive interpretation that gives clues to the different parts of a text . theories are not sets of statements. Hence. Vllses Moulmes (1996) explains the logic: According to structuralism. if ever. So iar. The first alternative is the received statement view 0/ theories as proposed by Carl Hempel (1965. Moulines 1996:7). 1966) and Karl Popper (l969a. have then to be checked. ccu more a e phenomena as manifestations of entities and processes t hat I'le beh'In d thos es t ~nsru h . I cannot discuss this in detail here. and links). 1987. The statements made by means of scientific theories are. explicit generalizations will become less important in hermeneutic anthropology. What structuralIsm maintains is that theories are not statements but are used to make statements-which. The latter are called "T-theoretical" when they can only be dete.tlc theoretical laws. above) open up the possibility that even at the level of particular cultures. Balzerand Moulines 1996. These are assumed to be ~overned by c aractens. 1987~ Balzer and Mouline' 1996. AbuLughod s (1991.tluch about competing interpretations in interpretive ethnog_ raphy. taking the explanatory schemata of phySICS (among others) as exemplary cases of elaborate theories. I969b).

while Alexander Bierstedt (I997) introduces structuralist concepts in his assessment of evolutionary theory in anthropology. In his recent textbook. We can (following StegmOller [1979a:33]) "distinguish between three kinds of 'nonnal scientific' progress: (I) theoretical progress. . structuralism defines a theoryelement T as an ordered pair of the fonn T = <K. or it can turn out to be approximately true.m~re setting. d l l l l l l l l 1 The Uses of Theories . This claim can be falsified. (theory-nets) that integrate several little theones on the othe . Rather. The claim that the paradigm:s theoretlc~l st~ucture IS e: ma not be able to actually f "reality") course sub]"ect to possible failure: the SCientific commu ity y. thiS I~ true ~~: Si~gle theories in the intuitive sense have encounter in scientific literature. only more or less productive. successful empirical applications--{)f the theory (as a subset of the set of intended applications). and methodological prescriptons on data structures and measurement. "Nolll1al work" consists of specializing the principles to cover new domains.. When the general theoretical principles of the formal core are extended to and specialized/or different domains. to remove these cases from. but a certam structure an about phenomena of a certain domain). whereby there is overlap of the set of intended applications of rival theorienhat enables transfer of the successful applications from one theory to the other in case of theory dislodgement with reduction (Diederich 1996:77-78). The basic idea of structuralism is that theories are conceptual structures. f h f h' d h fact that they are all specializatlons 0 t e conceptual rame. Members of a scientific community having or holding a theory (StegmOller's pragmatic concept for the use of a common theory [1979a:29-32]) share the fonnal structure-including the principal laws expressed in this fonnalism and a set of paradigmatic. to be viewed as aggregates of severa (s " Th's reflects the fact that most scientific I 1 alled "theory-nets. They induce empirical statements on intended applications and lower-level laws that can be refuted. (2) empirical progress.C I f r the simplest kinds of theories we theory-element.SC1. t e same fundamental law.Episremology 70 SCHWEIZER 71 of the structuralist theory concept in his formal reconstruction of kinship alliance theory. and to expand the set of true.. since It IS. i I l l. The received concept of theory in analytic philosophy of science had already revealed the insight that scientific theories are hierarchical systems that differ widely in their generality_ At the lowest level of theoretical abstraction.war . no statement at all. d rich 1996:76): formal core of a theory is not subject to falslficatlOn le e 'tendable to further cases is. Wolfgan~em::t:r ~s ch~racteristicdatastructures contrast between "little theortes (theory-e P h nsive "big theories" and test procedures) on the one hand and more compre . and extended The structuralist literature contams onna h ' ts and it has detected t of theory-elements or t eory-ne . consisting of a refinement of the net of theoryelement cores [work on K mentioned above. TSj. 1 y) let alone the theory itself.. the c~mmon holds together the whole array d' t' ction between the T-theoretical and k second the common IS In . As concepts. (B I t I 1987' Balzerand Moulmes discussion of these concep s even more elaborate theoretical structures ~ ~er e ~997'48~9) draws the main 1996). 11. o~am 0 not falsif the respective claim of the apply the theory in a certam sttuatl~n ~oes. o~e °ome. that a particular phenomenon can be characterized as behaving according to the theoretical principles contained in K and thus belongs to the set of intended applications I-which has to be checked. t e. C the sense of common theory-element or When researchers holdmg a. Usually. an t lr . Diederich 1996. we get a theory-net that links several theory-elements in a more complex theoretical pattern. structuralists distinguish aformal core K and an open set of intended applications I as two essential facets of any theory. intended applications. e d t array of more specl3 a of the hierarchy an a vas 'I determines a new theory-element. The failure to h d ' f IOtended app IcatlOns. of .:. Basically. KI~ 19 .er. (pp.calIo~ an(~. the latter IS not eve~ pass y su( Je~ a heuristic to generate claims itself. But reconstruction of theory change with and without reduction is still a thorny issue (StegmOller 1979a:Sect. . What degrees of specialization. roughly.' 91' Schmid 1996:238-240) who hold Critical rationalists (Popper 1969b. .th~ory m f h f 1 rinciples and when they are theory-net) embark on the speclahzatt. theories can't be true or false. ' theory (since It may e app. his "instrumentalist view" of the statement view of theortes wouldn t accept t F r them even top-level theories as tools for producing generalizing statements. consisting of a successive increase in the set of intended applications: I 1 C ll-t-ll and (3) progress in confirmation by which assumed elements are transfonned into firm elements of I.on 0 ~ eor~ tca P confinnation whereas the checking intended applications there IS fals1fi. I> whereby K is the fonnal core and 1 is the set of intended applications of the theory-element (introduced above as the most basic distinction). . d h appty the theory to certain cases an t us ~ay \. empirical applications. We may say t ' t "democratic" sort of entity." Scientific revolutions can be reconstructed as intertheoretic relations. As Moulines (1996) elucidates: f . Each spe~:\::s' 'i~ the hierarchy is. H~we. fonnal models.. 0 .. Balzer and Moulines 1996). It IS a axiomatic than o~hers. it can be true. 'bl b' t to falsification. . b lied to thiS situation atcr on . not statements. The empirical claim ofa theory element is. the T_nontheoretlcal layer. first. These aggregates ar~ c d'ff rent degrees of generality within the same conceptua theories have laws 0 v~~ a:l ~xioms of a theory are axiomatic but some are . -. 'otitic theories can actually be reconstructed as Doe Some "real-life" examples 0 . ji I definitions examples. lQ-l1) . A theory IS no t: re is a single fundamental law "on the top" '11 ws (and constraints) with different strongly hierarchical system. . .imes a great number of) theory-elements. t be urged (by reslS ance 0 . They contain general principles. . to l strictly speaking.

earch At a higher level of 2. . theoretical idea (say. Anthropologists are primarily devoted to carefully observing and classifying their objects/subjects of study and establishing empiricalgeneralizations(ifthey're interested in generalizations). if there weren 0 D crl'b' t!lese patterns doesn't rule out attempts to detect suc~ patte~s. because structuralist research on theories contains formally precise and rich concepts for dissecting empirical theories and their evolution. tt needn't be very Ixe d an ' alent rules or activity pa erns . an open set of intended applications. qualify as theories in the strong sense. es . ial or anization and cogmtlOn : : ec?logy) others than to the being mtermedIate). These ideas lead to testable hypot eses a a researchers When ethnographic record and are illustrated by exemplary cases. What remains to be clarified is what kind of theories are used in anthropological research. the ritual cycle among the Tsembaga of Papua New Guinea (Rappaport 1968). the strong sense of structuraltst IOO't altfy as theory UI mg m .~~: ~:a::eI~:~~~O. . or to the nature oft~e jnstrinsic nature of soct .. Sometlmes . 't doesn't engage in a systemauc bfi lds When a sctenttfic communt y I' f special su te '.etal and croSS. And there is always refutation at the level of intended applications and lower-level laws. . economiC 0 . how human na ur .ISnonc to arrive at cross-Cu IturaI'. ad hoc. Genera Iza Typically. t )" necessitates discovering cultural eltmo I s "describing a culture (or so. d h whereas the last two types of generaltzauon philosophy of science dtscUst~e oaf ~~~hropoIOgical theories in the proper sense. h tic descriptions of different These typologies are much more than m~re sc ~ma e testin of isolated types of societies. including exemplary cases. BU~aI and cultural reality. but it changes the perspective on what a theory is and what it's useful for.. rts of culture and society (for example.. 0 e informed by some anthropology are generalizations at the middle lev~1 that ar . econ?my lanatory success about some pa ". In this chapter. I've introduced this new perspective on theory and theory building. political systems of highland Burma (Leach 1954). on and Earle 1987. t (ns 10 ethnographiC cases. . the first two types of gener~l­ and theorettcal prmctp . Hypothesis construction as part 0 comparall~e res. . ked and how they social and political structures.an t oug man collectives would break down empirical systems anthropologists study. Cultural and social types as theory-eJem~nts. makes a lot of intuitive sense in anthropological theory building. and there is more aggravated. m a . and the historical events connected with Captain Cook's stay at Hawaii as the most recent controversy (Sahlins 198 I.n e ~ t: ct valid for a tot of other of hypotheses that claim to be true an are. So a theory concept that stresses theoretical principles and empirical applications as necessary and equally important components and that highlights the important role creativity plays in connecting theoretical concepts and empirical examples is closer to actual research practice and thus more revealing than earlier and alternate concepts. tlheo~et. the discipline is still in a natural history stage of theory building. f rces t e the environment. if . theories ate not immunized against criticism. And the previntracultural or intrasocletal vanatlon (which c fi Y d . and patterns oftho~ght a. critically assessed in Borofsky 1997). but a successful theory won't be overthrown hy a few failed applications." in general. I' t'f . 1992}. jzations wou qu. are devoted to the construc IOn th~n . There aren't many deep theoretical principles (but see below).~atic ~ircumstancs l. and methodological prescriptions on data and measurement.. So. establish cultural and SOCial types {for example. and one that doesn't lead to successfui applications will eventually be abandoned. This i~t~et ~~~x~!~~~tOry schemes in 3. even verstelzen approaches ~n ethnOg~~t. structuralist rational reconstruction of the role of theories comes closer to scientific practice and thus seems a more valid description of what scientists actually do. soc'.Epistemology 72 SCHWEIZER 73 theoretical principles are falsifiable. rendered plausible.girl As 1 argued .:a:::~~::~:~i~nal level and because many of these hypotheses shc case y . aim at eneralizations at the above. In addition. The new structuralist view doesn't lead to totally different empirical research in the discipline. the structuralist concept of theory as a combination of theoretical principles. cross-cultural research aims at dlsc?verm t~s "~e: hindered that are true for many cultures. Joh s Maryanski and Turne. t d resent systematically compare ethnographic and historical cahs~StOf"haelpaSY~t~:s~ they . Leach and Leach 1973).~ l:n re checked in the mold society). b 'ld' ..cle Yd h ht-that give some coherence to the _shared patterns of behavlOr . There has been a lot of reanalysis of renowned cases in the light of alternative theoretical frameworks. Theories can be more or less productive. It has led to a much more detailed and realistic view of theory and theory change than its predecessors. Howe~e~.cross-. However. empirical applications and testing are crucial for research.ve than th g .graphY l:ads to the discovery level of particular cultures. ~u d cooperation Ethnography 't s me shared expectatIOns an . and they're much more mtegrat. and chaJJenged by rival theories via fitting or resisting ethnographic cases. Since most anthropolog~sts .'. t?e ge~eralJzauon as arativists work Theories in Anthropology What is the state of theory construction in anthropology? Generally.are ~ k is even more finding differences than similarities. as part of a case st ... udy At the most elementary level. A higher level\ of synthes~s ISe~c~~:::~yWa~~~v~no. Nuer segmentary lineages (Evans-Pritchard 1940). We should remember that in the l structuralist perspective.~ structures that on establishing whole exp anator)' sc em . cl l' h potheses gan generatization. ore interested in they're isolated. mg an be s stematic). 1995 versus Obeyesekere 1992.r:. in anthropology theories are introduced. w n't be much discovery of empirical genera lZa tons search for r:'gular!tle:'i~~e~~ th~ classification below. Some examples include the Kula (Malinowski 1922.

structuralism (see Tjon Sie Fat [1990] on structuralist kinship theory as an attempt in this direction). 306) as "social approval. nature to be fou~d m th~ l~t~ratur~REEM For each aspect. political. ~r~t e 109. dUcio social goodS. I996])dbel g represents some additional theoretical ~. among others. Both would yield a theory-element. o~ s to the core. every theory includes characteristic 1 I' l F . Anthropological theory-nets.ion dismissed m ea[\~~O. D'Andrade 1995). while a lot can be consldere as exem . C~ e . with the help of rational choice prinCiples I . On the empirical side are intended applications. 1 :Vcr. social. we have theory-nets. bridge assumptions expecting. such transparency in favour of subjective p~obabJlltleS~ndhbY~~tro f"e:aluating. Well man and Berkowitz 1997). . whereas the majority of ethnographic descriptions are static. including paradigmatic examples. . then we have homo econom~cus S uld be reason to make the economics textbooks.Othe assumption of market model more complex. symbolicanthropology.' the basic remises to account for new omams phenomena and the speClahzatlo~ o~an I990)P Taken together. economic. Coleman and Fararo 1992. and cognitive/emotive phenomena that Telate to and aTe valid in a lot of ethnogTaphic cases belonging to a certain type (for example fOTageTs. The theoTetical agenda underlying these typologies is that the diffeTent types TepTesent nonrandom. These authors attempt theoTeticalreductions ofthe massive ethnogTaphic Tecord investigated and they try to integTate 10weT-level hypotheses. In most contexts. t· lend formal and their impact on an actor's strategIes (Wasserman Ig bra and statls ICS h Schweizer 1996:Ch. Becker 1976.ChS 1 2]" Lindenberg and Frey [1993]. I t. fundamental configurations of demographic. if "restricted" reter~ la somt': : Y Sf U ~s'that maximizes utility given price technical meaning of chooslOg the pac age 0 goo .llustrated in Lindenberg's exposition on . We could also reconstruct other theoretical Currents including those of the past.~ a theory of action. db ~:':"~lFor (1) rationa~ P1entioned above. mosl often encounh. ho~eve~. Furthennore.p e 1. Esser[1993.:. maxlmlzmg ( e used . I e 1 1 cognitive anthropology (Holland and Quinn 1987. In addition to the models and intended applications.:e t::t~~'p~~g7C~l discourses.. . Gfirlich 1998).~ and "maximizing" refers 10 the goods. but let me offer some ly to the four theoretical currents the structuralist theory concepts app . For this generalizing task. cho. Some theoretical structures in anthropology (and related disciplines) qualify as theory-nets in the structuralist sense. evolutionary ecology (Smith and Winterhalder 1992). hypotheses.. social network analysis (Wassennan and Faust 1994.. . For example.n IS Isol~te Y ta Th~ attributes are: resourceful. cultural ecology. 4). To a degree." (p. postulating causal developments like the transfonnation or emergence of types in an evolutionary framework. d rticular domam IS mce lY . d d r tions of the theory covering a tical principles (Betzig behavior set up domams of mten e app Ica r d h empirical cases and containing a lot of specla \ze t eore 1997). Grap h t eory. a certain configuration specifies constraining as well as enabling conditions that influence the historical tTajectory of certain types (How does a Big Man collectivity become a local gTOUp or chiefdom? What happens to hunter/gatheTeTs when they become sedentary? How can we explain the move from foraging to agrarian modes of subsistence?). I s and reworking them so that they can be theory-net.nm:~~~~'omic applications of the theory of prices .Epistemology 75 74 SCHWEIZER ~:. as examples of contemporary theories.W~llman and theoretical principles establishing an overall relatIOnal pers~e t ong(and weak ties Berkowitz 1997) and specific hypothese~ on the pattemm~:d sdalaskiewicz 1994.. like evolutionism. structural-functionalism. Mathematical game theory. for example by at . '~"> [1990. ecological. The typology is refined in light of new empirical and theoretical evidence obtained by research.. cultural materialism. When there are specializationsofthe principles at the top for different domains of application. Using theoretlca~ p~mcl. . crosssectional investigations. in-depth empirical generalizations. 4.). Strauss and Quinn 1994.t8 suuctures and testmg· . Theoretical progress in constructing these typologies is difficult insofar as the intended theoretical statements are often diachronic. if we have to be made before the"core c~~ b -ctin'" as being completelY informed abuut interpret "rcsourc~fuln~ss and. and reproductive food acqUiSitIOn among g . f the economic model to noneconomlc ncems the extenslOn 0 d . as theory-elements or theory-nets in the structuralist view.. agraTian communities.:red in and budget constraints. evaluating. I gy (2) natural selection would Similarly. . 1996]. but a more realistic version del of "economic man" (not the bogus 1 1 l I 1 ~ . ( 996)· apphe to a pa . D' Andrade and Strauss 1994. etc.' presented in Coleman .. d h net There are Social network analysis (3) can also be reconst:ucte as a t c~i~~. and - bints hoW devices I won't go into details. "0 although different m history . in the theory of evolutIOnary eco. t:' ere. ~x~~luat~l "as reference functions for consumer alternatives and prices.. . the types are theoretical idealizations that can be illustrated by empirical cases and that are approximated by other cases belonging to a given type. a e 1994) The set of intended representations of social structure (Wasserman and aust . chiefdoms. Ensminger 1992. principles o~ stTa~eglc actl0p~~ryO applications of rational choice theory. Lindenberg ~. ': we take e bd" dg. Wasserman and Galaskiewicz 1994. this establishes a {for example. anthropology would need more longitudinal datasets describing cases at different points in time. Ensminger and Knight 1997. Think of rational choice theory (Coleman 1990." with shadow pnces under t e ea 109 0 . how to explam socta ac ton kin the five most relevant aspects of human A core theory of actlo.o 's ecialization of these establish the topmost principles at the theoret~~~t~~~e~th~rers. 1 The core of the theory in anthropological theory-nets are general principles that guide research and formal models for deducing statements that can be checked empirically. restricted.a. lot of Principles to patterns of . of work noW co . and it should carefully analyze the available record on historical junctions when one type changes to another one to establish more validated. u. etc.

. we ave 10 re people say and do. and cultural consensus. and activity in particular cultures or . while adherents of the paradigm are expanding the set of intended applications (Hage and Harary 1991. covers a h Inte retation (meaning of words. hb d' means of SClentl lC v a . acUo~s. 1 empJrlcaltrut Yor mary t h n should somehow be typlca considerthatthe stories and cases produc~d by. i: r . I conclude by pointing out some crucial features of theory construction in h "m ortance of meaning. The set of intended applications includes native taxonomies.nderstandmg . explanatory schemes that integrate sUbjeC:iv~ m:~n~ngs the uncovering of a hidden c The concept of theory in hermeneutlcs IS her e ~e general pattern underlying message in a text or discourse and the searcc or so I'ent patterns of thinking. (I) In discussing positivism. as 'fi so l'ldat'lon This is apparent when we . and of lbe humanisl approach focuse~:a~s~ ~ap~ured and represented by he~eneuti~ anthropology. We need not integrate all our hypotheses. history. At the theoretical core. communications. D'Andrade 1995). I've stressed the internal diversity of this analytic framework and the fact that the theoretical part of the positivist agenda is still underdeveloped in the discipline. marketing.n IS 0 bedding of a phenomenon in proven. rational choice theory has been used in economics. cognitive schemata. Theoretical integration is the crucial task ahead of us. . cognitive anthropology (4) is characterized as a mix of principles and procedures from linguistics.~ are the main strengths of the _dering rich descnptlons of na ~th 'ramework rules out the possibility of the . I 'pervasIVe an presen In . background knowledge and tested this distinction parallels the . Thus. except to engage in empirical bypo. . Research should attempt to refine the understanding of theoretical principles of these theories and to expand the set of intended applications.. Thus. It seems to me. and computer science (Strauss and Quinn 1994. anthropology.n more encompassmg ry and can e corn me t Both procedures are comp Iemen a . sh~wn this doctrine of radical alterity impossibility of understandmg th~ 'lyon charitable translation of what other to be 10gical1y flawed. (2) My discussion of hermeneulics as the main exponent . and anthropology. but theoretical work shouldn't stop at this level. Conclusion: Toward an Integrative Agenda I consider the methodological frameworks to be partial and rationally grounded routes to the generation of systematic empirical knowledge on human societies and cultures. ~ rtT d as the examination of the re atlons lp r and the two can even be cross. but It . and dlScour~e:~1G::t~. sociology. etc.. anthropology. reflexivity.s no w~ re eatedchecking.. cl the whole of a text or discourse) IS dialogue between unders. Regarding the key pro em d't' It assumes that there's a real world out ~list perspective as background tra . evolutionary ecology is connecting biology and anthropology.hes es on the world. . organizational studies. There . It uncovers patterns of mea~l. t Ive the usual problems of logiC and o hypotheses on purposeful action. . of the background knowledge can be chec s e~al sense of grasping lhe actor's Empathic understandmg (ve:s~ehe" m th~ anthro ological and social science point of view) is a useful heUristiC apP~o~nd leads to ihe discovery of interp~etive research. etc. (3) Generalization at the level of single cultures and societies is often the first step of theory building. (I) There are some explanatory schemes that qualify as theoryeh::ments and theory-nets. producing better theories tbat enhance causal understanding of and orientation in the world and that lead to better. (2) There are empirical generalizations and isolated hypotheses that should be integrated into more encompassing explanatory schemes. d d' and explanation IS reJec e betweenempath1cun erstan mg b b' d '. d d exp Ianation shows een verstehen an h i ' 01 method I introduced a revise betw bl s of ant ropO OglC . Scientific explanation IS eluctdate n~xt as. that the best of our current theories must transcend narrow disciplinary boundaries.~n: rt that it's difticult to establish there lhat it's conce ptual1y con~true m P:u. 1997. Schweizer and White 1998). sand ohjective constraints. lbat aren't speCIfic for humamst researc . Social network analysis is an interdisciplinary brand of structural thinking in several social science disciplines (sociology.e 1 Ize b I ) othe (see e ow . w theses generalizing across 'he science ("positivist ") framework alms at arger syn ... among others. thIS quest t en e t hile the concept of theory In feeling. d rp t' al1 fields of scholarshIp." whole set of different and related con~epts The term "u. psychology.) in a genera sense tS ' s The hermeneutic circle (an on-gOln.. cognitive psychology. 1996. then.ocus on sa I concrete cases.landin g the parts ~n In the natural (and soelal) sciences. Anthropologists should embark on theoretical syntheses of the ethnographic record. h However nel er" I' h' hellJ\eneutlC approac ' . This distinction 15 not a senoUS::~ in rlnciple at'teast.~ d rawing of polar opposites some theoretical model.g including the SOCial and natural SClence. geography. not specific to investigatIOns In the hum:m les. but we should try hard to establish networks of validated hypotheses that lead to theory-elements and theory-nets. usually containing ~ws: ~ t d as a misplaced contrast. What is most revealing is that these evolving anthropological theory-nets are cross-cutting disciplinary boundaries by integrating parts from different disciplines for the sake of better explanations. etwee~lem however because any element hypotheses. h I ads to a . Epistemology SCHWEIZER 77 applications includes the classic case studies of this approach..difference. political science.n. .ex s.. and cognitive anthropology has reemerged in the 1990s as an anthropological variety of the broad cognitive sciences field. . theoretically informed applied research.). oricat embeddedness arno?g i the actor's point of VleW an ~ studies of texts. .~~~ e ~en implicit claim has to be for the larger setting that is i~vestig~te . As to the ultiroate validatio~ ~nd to eliminate false hY6t~:rs~:v.

(' t hIS perspective. of arguments in themselves. e Ica structures ese t eoretlcal structures can still b e ' . The gist of the thesis of radical alterity is the claim of a universal impossibility to understand other people. 5. analytic philosophy. close inspection of methodolo ic I f g a problems and solutions reveals . and Martin Heidegger. . embers of a are stnvmg to refi th '. J use the word "hypothesis" as a general term for unvalidated. cognitive anthropology) achie ~h DIce t eory. One reader of a previous version ofthis chapter related this claim to another cognitive limitation-the almost impossibility of any humans to understand how cruel humans can be. much effort has gone into critiques and not enou h' 0 0 oglC~l schools. methodological debates in nth I ac IS POSSIble and advisable. different frameworks. in particular Michael Bollig. 79 NOTES that some of the differences in k· d J thank the editor and the following people for valuable advice: Joachim GOrlich. that the standards of ethnograph. 6. a e le dwork tradition · eSIS are common ISsue t 11 1~ad 109 to convergences at feast of problems and . theories are conceptual structures . reports. Rereading the writ in s of . s 0 a anthropologists. 4. as { shall say. 1·1 11 e t e theoretical core cannot be . In the past. 10 0 past debates ( understanding versus explanation. they can also have different and more elaborate forms than the logical implications (deterministic if/then statements) connecting two phenomena that are mentioned in the text. I d· d omams are dlstmgUlshed as important elements f . and this leads to a logical flaw. but a serious practical limitation to understand the cruel Other. m case studies. there are three worlds: the first is the physical world or the world of physical states. most notably empathic . e prInciples and to falsified. To keep the terminology concise. Karl Popper. Although one can try to causally understand how such events come about. 10 arger and ncher reveals that there has been a refinement gf revlOU~ anthropologists clearly o met thinking.em View DJ theories. transcripts of interviews. .) . In anthroSome explanatory schemes that cross cut donc~Pltua Ized as theory-elements. They can even be considered . Hartmut Lang. IS time to stress th t th fi I and a concern for comparative synth . ' n eses 0 the ma' d th eoretlcalldeas that have accumulated· th d' " SSlve ata and selective In e Isclplme. Standing on the sho!d afs hgamed a real mcrease in causal u ers 0 t ese schol h . can conduct constructive theoretl·cal sy th f ars. I use the term "text" only for documents produced for storing information in a written form {like letters. hypot h eses from cross-cultural research d h . t. places. Clarence Gravlee.l d . . Thomas Widlok. cultural and social types can be c ~. the recent massacres in Rwanda. never cover everything but a more i· t . or even individual cruelties. eela aws can be fals Th and apphcations only) and theory-nets (that· dd. Peter KaPpelhoff.. among others). n egratlve appro h· . This empirical turn (compared to the more l l l l ~l l l -.. eory-elements(cores applications include specializations of th m a HID. it is the world of possible objects of thought: the world of theories in themselves. I concede that there is a point in these instances where rational understanding ends and where it's impossible to grasp how humans could act like that This. . e theoretical prin . t heoretIca! proJects. and an open set of intend d . ad hoc theory-nets into more encompassing theo'r atn . intended applications and sp . and their logical relations. sometimes even of sOlutions across Furthermore. partial and selective. while ·· ISClP mary bound . members of my colloquium on anthropological theory and method. as useful steps· I . Michael Schnegg. field notes. observational records." which is much in accord with actual research practices in empirical disciplines. e empmcal applic r 6 M ' sCientific community holding a theory .... a Ions. and of problem situations in themselves. e-specl IC thick d . The established 0. Too next step after necessary criticism) It" g mto constructive efforts (the . Rudolf Carnap. This doesn't question the legitimacy of studying "culture as a text" in a metaphorical sense. for instance. Christoph Brumann. e ecome much more understanding." to the theoretical core and . On the different logical forms see.fi . 3. and leading figures like Auguste Comte. however. . anes In the socIal and . . be haVlOral sCiences (like eVOlutionary I .e.ln an interesting study. statements is contrasted with the new t t I' P SltIVlst View of theories as ' s rue ura 1St nonstate . work analYSIS. extend the set of intended applications WI·I h me . In formal models. In his recent exposition of the theory concept. is not a logical matter {as the thesis of radical alterity discussed in the text). and that the discipline h ." 2.t e available theory-elements and Th h . ant ropologlsts today ' . search for laws) only establish differences of de re escnphon versus the supplementary and can be incorporated g. . Integration will . contammg general principles . [logical] positivism. thinking of the Holocaust. fiction. and epochs.cu/tur . Dovring (1954) points to the beginning of quantitative text analysis in the hermeneutic tradition when rival schools of Protestant biblical exegesis turned to word frequency counts to settle a theological dispute at the courts in Sweden in the -eighteenth century.m. I l l l l 1 . The focus on ideas in my presentation can also be grounded in Popper's (J972:154) pluralist philosophy (although we needn't draw on his conception): "In this pluralistic philosophy the world consists of at least three ontologica{ly distinct sub-worlds. Useful background information on the philosophical currents discussed in the text can be found in Bunnin and Tsui·lames (1996) and in the Encyclopedia Britannica {see the entries on the keywords philosophy of science. e. and Barbara Zschoch. the second is the mental world or the world of mental states. conditional statements.0d?loglcal and epistemological · . . Hypotheses can postulate deterministic or statistical relationships between antecedent conditions and consequences. emphasized the differences between th a rep I 0 ogy and related subjects overly eoretlca and meth d 1 .. whereas laws and Jawlike statements are hypotheses that have been tested and confirmed. or of ideas in the objective sense. or. and the third is the world of intelligibles.l).78 sell WElZER Epistemology different cases. The crucial genera IzatlOns fro .) and "discourse" for verbal interactions. Suppes (1957:Ch. socla[ nettheoretical task ahead is to integrate vel' e :tatus of theory-nets. sClenhfic theOrIes. eco ogy ratIOnal ch· h . c escnptlon hav b soph IShcated. etc. . Balzer 0997:48~O) includes data structures and approximation in his definition of "theory. Clp es to Ifferent pology. .

Pp. Scientific Explanation. The Economic Approach to Human Behavior. e s. Gary. Cambridge: Cambridge UniverSity Press. L 'ecriture et la diffirence.es~\1991. Race. m. Ortner. The Proftssional Stranger. Pragmatic an Structuralist Theory 0/ Science. . . In Recapturing Anthropology. d D 'd M Schneider. eds. B i d C Ulises Moulines. Hans. . am Tl g . e s. . '. 1977. CA: AltaMira Press. Anthropology. The Operation Called Verstehen. . nd Cultural Models. th and InterpretatIOn. rh P edicament o/Culture . 1994. Abel. I 1__ d 1996 The Bfackweff Companion to Philosophy. eds. Kemnltzer. 1991. Jacques. h" hy 0/ the Social Sciences. CA Sage Publications. May. Oaks CA: Sage publications. R' h d A Shweder . New Philosophy o/Social Science. am Tl BourgoiS. A thro ologist 99: 123-134. e . . Boulder: Westview Press. . 1996. and Stephen F. Berlin: de Gruyter. e r ClilTord. 1967.· . . '. 1986. Cambridge: <. Writing Against Culture. and Peter G. Agar. . Balzer. 1994. Germany: Luchterhand. Barker.Epistemology 81 80 SCHWEIZER theoretical core interests of structuralist philosophy of science in general) comes with an attempt to include the social sciences and applied natural sciences in the consideration of an appropriate theory concept. Laura. Blaug. d 1992 Rational Choice Theory. Vo!' 5. . Cambridge: Belknap Press. 2d ed. Oxford: Blaekwel!. Language and Culture. Pp. Berkeley: University of California Press... d 1994 Culture/Power/History. L Achinstein. 010 Cambridge: Cambridge D'Andrade. d Diachronic Aspects of tructur IS . Qualitative Research Methods Series. Ray.. 1997. . n earnap.. 1980. IL: Open Court. Freiburg: Alber. C I ry Prices. 2d ed. Ulises Moulines. I . Alexander. American Journal a/Sociology 54:211-218. Oxford: Oxford University Press.• and Thomas l. Ideen Ub ere. SluIlgan: In his Gesammelte Schriften. 1948. EditIOns du SeUlI. 1969. Frankfurt: Peter Lang Veriag. Wolfgang. Walnut Creek. .. Stram erg. 1993. lames. lames. . eds. Princeton' Princeton University Press. . e . Tsm-lames. Davidson. 1996. 1997 Cook Lono. Agar. S stems In Culture Theory. and loseph D. An Architectonic for Science. Speaking o/Ethnography. REFERENCES hI' Current Anthropology . Die Wissenschafl und ihre Methoden. Cultural Meanmg Y C b"d e' Cambridge University Press. The Myth ofthe Subjective. Readings In the P I osop \lrOdbeck. Berkeley: University of ClifTord. New DirectionsinEconomic Methodology. 23-43. . Chicago: University of Chicago Press. Writing Culture. L V' eds Pp 88-119 am n g . 1984. Structuralist Theoryo/Science. 1990 Information Costs and us oma Cashdan. ed.. h ed Pp 139-240. 1988. Wolfgang Balzer and C. f . . h I The Philosophy o[RudoI[Carnap. TlIbingen. Vol 2. Wolfgang a zer an . Peter. Pp. e s. Hans. Becker. -84 La Sal le. Franz. Kritik der reinen Hermeneutik. Bowl m. p '. Sneed. eds. and Robert A. American n 'P. Pp. The Methodology 0/ Economics or How Economists Explain. ed. Derrida. Bohman. . P in Trl a an C b'd e' Harvard UniversIty resS. lames. Th S 1990 Foundations o/Soc/al eory. 75-82. Theodore F. . f Notre Dame Press. 159-172. Albert. 1987. .. 'nebeschreibendeun zerg le DiIthey Wilhelm. . 'b' d Peasant EconomIes. Netherlands: Reidel. . Albelt. Donald. d I' derndePsychologie. Bunn . N' holas and E. Ray. and George E. 0 ford' Clarendon Press. Thomas. ed. . Ulises Moulines. IC ar O'Andrade. x . Geoff Eley. Bartelborth. . . and C. 1900(1894).1995. Berlin: de Gruyter. . 1976. . . Mark. New York: Columbia UniverSity Press. Ulises Moulines. Fararo. Santa Fe: School of American Research Press. Svmbolic . ' . 1967. m Davidson Donald. Marcus. Robert. Lila. lam. Werner. Writing Women's Worlds. ~ o. 1984. 259-278. London: Routledge. ed. an ~VI . Culture. In Risk and Uncertamty A. Backhouse. . Research Methods in Anthropology: Qualitative and Quantitative Approaches. I . e me. 1989. Inqulf/es In 0 r u . S al' m In . Nicholas 8. Fox. Berlin: de Gruyter. 1994. Bierstedt.Roy. 1997. Abu-Lughod. . and Claudla tra~ss. The Development o/CogmtlveAnthrop gy. Marktsoziologie und Entscheidungslogik. Lila. Human Nature. Thousand Oaks. and Sa mS. J Dolgin. Sherry B. California Press. Gennany: Mohr. Schtlpp. ds 1986. . . Diederich. Russel!. Roger E. Balzer. The Legacy of Logical Positivism. James S. New York: MacMillan. Janet L. Balzer. C. Baltimore: Johns Hapkins Press. . afls. In Re atlV/S . New York: Academic Press.. Borofsky. 1997. Oxford: Polity Press. Dordrecht. Richard G. John R. In Structuralist Theory 0/ Science. Bernhard Groet uysen. Pp. P. M . 1992.. Teubner. H. London: d 1968. Wolfgang. S d 1994 Human 0 Ives a D'Andrade.ograp y. Boas. Darwins £rben und die Vielfalt der Kultur. David S. IC MacmiJlan. Bernard. Abu~Lughod. Obeyesekere. University Press.ol . Michael H. Betzig. 1995 In Search of Respect. Dirks. C b 'dge: Cambridge University Press. Neuwied. 137-162. eds. Pp. t T. Michael Krausz. Elizabeth. 1961 [1940]. . Michael H. e s. Thousan " Coleman. Phlhppe. 38:255-282 . 1996.· b 1997 Representation and Reality in the Study 0 . ed. Paul R d If 1963 InlellectualAulob. Notre Dame: University 0 . d Coleman.ambridge University Press. Wolfgang.

Hartmut. Public Opinion Quarterly /8:389-394.. Hartmut. l l l l l l . Interpretation and Vision: A Critique of Cryptopositivism. Kim. . Oxtord: ac we . 1940. Clifford. . eds. Lawrence AltaMira Press. 1983. Hans-Georg. New York: McGraw-HII. dC UI' M rnes eds Pp 83-1t3 StructuralistTheoryofScience. Dorinne K. Works and Lives. CA: A. Chicago: University Kuhn.. 1995. and Jack Knight. . Method and Appraisal in Economics. 1994. The Rationality of Everyday Behavior. Kutschera Franz von. Local Knowledge. C b ' d . Esser.. Keesmg. Geertz. Michael. Lutz. Germany: Mohr. . White. arttn. Amsterdam: RadoPI. 1982. 301-312. Jean. mre. Geertz. oger. . Making a Market. Ekman. eo .g an Leach. 1987. Ensminger. Johnson. Gupta. u Jons. Glasersfeld. . !le P h d Naomi Quinn eds. The Kula. OOrlich. New York: Garland STPM Press. Samh. Networks. . Manfred. eds. Ernst von.I . . Notre Dame: University of Notre Dame Press.. Latsis. Fox. I d Crff" NJ' Prentice-Hall. hi dB London' Athlone ress. Tilbingen. Oxford. an . urma. Rec Immmg clen tfie I I Kuznar. Stanford: Stanford University Press. I'arieties of Relativism. 1970 [1962}. Goodenough. Berkeley: University of California Press.Epistemology 82 SCI\WEIZER k Barary. Thomas S. Germany: Rowohlt. Quantitative Semantics in 18th-Century Sweden. Pp. 1995. eds.. . ed. ed. Esser. M . am n ge. 1966. cd. Geertz. Clifford. Munich: Ehrenwirth. Kritik des Strukturalismus. Cambridge Umverslty . 1998. Altered Contexts. 127-139. Hempe1. 1989. . n KU1pers. Lob der Theorie. 1989. Franklin. e H e Per an e Harr . Annual Review Anthropology 24: 163-184. 1997. Robert Borofsky. E. 1993.. ur d'le et }moI oglsc Lang. Cambridge: Cambndge University Press. Geier. Der Wiener Kreis. Leach. Twisted Histories. Press. Spiro. V' N w York' Simon and . ! 1970 Falsification and the Methodology ofScientiflc Research Programmes. 1997. 1990. Th A F 1996 Truth Approximation by the Hypothetico-Deductlve et 0 . Deborah B. Geldsetzer. 1992. . Schuster. TUbingen. The Interpretation 'fi 0/ Cultures. Pp. Germany: Mohr. \994 The Philosophy o/Soclal SCience. Edmund R. Wissenschaftstheone bndge' Cambridge . .Wo1fgangBalzeran . an . . 303-331. Holland. Jean. Santa Fe: School of American Research Press. S . Bo-Hyun. 1995. Science as Culture. Kolner Zeitschriftftir Soziologie und Sozialpsychologie 48: 1~34. d tJ e Growth of Knowledge. Grund/ragen der Erkenntnistheorie. Cultures of Science. Toward a Working Theory of Culture. . M h d I · .. 262-275. Political Systems of H. . 1991. Holhs. Berlin: Relmer. and Exchange. 1991. and James Ferguson. Recapturing Anthropology. . fi·· . Pp. Rationality and Society 5:7-31. Cambridge: Cambridge University Press. 1988. . Evans-Pritchard. E. London: Hutchinson. The Nuer. 1993. eds.. '1 Robert Borofsky. Wlttgenstem Sienna. Crafting Selves. Ensminger. Aspects of Scientific Explanation (Illd Other Essays in the PhilosopJI. Paul. 0) clence. I3I k It d Michad Krausz. Die Definition der Situation. 83 Dovring. m el. P . In AssessingCultural Anthropology. 1996. Gewertz. d At Musgrave eds Pp. 1994. 1991. In Criticism and the Growth of Knowledge. an d Step hen Touhnin. Wahrheit und Methode. Tlte Structllre o/Scientl IC Revo It' ~ 0/ I i Friedrich. 1983. of Chicago Press. Joachim. Clifford. Cambridge: Cambridge University Press. a . Cultural Anthropology 7:211-231. 1948. Cambridge: Cambridge University Press. In Assessing Cultural Anthropology. Clarendon Press. Ises ou 1 .. Hartmut. Bridewealth. eds. . 1992. Chicago: University of ChIcago Press. Gadamer. . . 1970. 1991. C am University Press. Krausz. Cambridge: Cambridge University Press. Oxford: Clarendon Press. A11 an.Cambridge:Cambridg e ntVerSl y r s . Radical Constructivism. Hartmut. Chicago: University of Chicago Press. Imre. 1973. and Alan Musgrave. R M \994 Theories of Culture Revisited. Leach. New York: Free Press. New York: Basic Books. and Frederick K. Reinbek. bridge' Cambridge University Press. Cam . Cultural Models in Language and Thought. The Evolution of Human OCle les. 1991. New York: Harcourt Brace. Berlin: de Gruyter. Jamk. a S' Anthropology . Anthropology.1996. Kondo.V ag" ran . . r St nford' nd Timothy Eark. Esser. U· 't Pe's ge Per an d Iran !l• • 'dF kHarary. Gadamer. Imre Lakatos an 91-195. 1987. . Alltagshande/n und Verstehen. Current Anthropology 38:1-24. 1980. Cntlclsm an 1 Cambridge: Cambridge University Press. AI! en W . Errington. . Frankfurt: Suhrkamp. Richard G. Philosophy o/Natural Science.rS . Clifford. Akhil. Exchange in Oceania. lakatos. The Face of Man. and Edmund R. In Kinship. Paul. Geertz. a Stanford University Press. 1997. Visayan Vignettes. Alfred L. Anthropological Locations. Karin. . ed. In Handlexikon zur Wissenschaftstheorie. 1992. \976. Carl G. . • . Jerry W. Jean-Paul. 1975. Eng ewoo 1 s. Kroeber. Cambridge: Harvard University Press. e . Dumont. Hermeneutik. ed. 1954. Hans-Georg. 1965 [1960}. Dorot y. Berlin: de Gruyter. .IslandNetworks. and Clan Exogamy. Changing Social Norms: Common Property. A Rational Choice Reconstruction of the Theory of Action. Helmut Seiffert and Gerard Radnitzky. Thomas Schweizer and Douglas R. London: Falmer Press. Rom'la~ 1965. Pp. 1996. Between War and Peace: Gift Exchange and Commodity Barter in the Central and Fringe Highlands of Papua New Guinea. . New York: McGraw-Hill. After the Fact. Walnut Creek. . Ward H. Car . Relativism. Lakat os. he Praxis '. 1954. .

Susan C. RationaliUit und Theoriebildung. Moulincs. Siegwart. Muster sozialer Ordnung. John. The Hague: Nijhoff. Kimball. Hans. Networks.I. Alexandra. Thompson. 1992. 1959 [1937]. Pastmodernism and the Social Sciences. Marchi. George E. 1986. Theories o/Explanation. R. Objective Knowledge. 1992. After the Critique of Ethnography: Faith. The Popperian Legacy in Economics. Fischer. New York 1959: Basic Books]. Siegwart. 40-54. Structuralism: The Basic Ideas. ed. The Kalela Dance. Robert Borofsky. 1994. Robert Borofsky. 1-13. Hilary. eds. Karl. Philosophy of Science. Social Structure. The Method of Decreasing Abstraction. "L~JchaeI. 1989. In Structuralist Theory Science. Lowie. IL: Free Press. Hope. Robert H. and Social Revolutions: Prerevolutionary France and Russia. Levi-Strauss. Anthropology as Cultural Critique Chicago: University of Chicago Press. 1963 [1951]. Nicholas Bunnin and E. Netting. Stanford: StanfO~ University Press. Berlin: Reimer. Culture as A Theory of Culture and Informant Accuracy. 1968 [1952]. Glencoe. White. lames S. Householders. ed. 1981. Batchelder. Rationality and Society 1:51-77. Alternatives. Cambridge: Cambridge. Interpretation und ReaUliit. How Natives Think.ois. Smallholders. George Peter. 1992. Lindstrom. 1995. Oxford: Oxford University Press. London: Falmer Press. 3-20. London: RoutJedge & Kegan Paul. Neil de. In Assessing Cultural Anthropology. ed. Minneapolis: University of Minnesota Press. 1988. '_~:' '" c: ~d Douglas R. The Social Cage. Gananath.1homas. 1993. S. Robert McC. and Relative Prices: A Broader View of Rational Choice Theory.2. Ulises MouJines. 1996. Collected Papers 1/: Studies in Social Theory. University Press. Post-Popperian Methodology of Economics. Cambridge: M. Conjectures and Refutations. . iffe-Brown. Cambridge: Cambridge University Press. but the Greatest of These Is Charity. Oxford: BlackweJl.84 SCHWEIZER Epistemology 85 Lenk. Planner. bndge University Press. ~hall. A Natural Science Lindenberg. DC: Smithsonian Institution Press. F. Glencoe: Free Press. A. Bronislaw. 1971. Lindenberg. London: Cohen & W'f. Munich: 'i['- ~us: . Truth. Cambridge: Harvard University Press. ed. ed. In James S. Ulises. Willard Van Oonan. Ein Meister aus Deutschland: Heidegger und seine ZeiJ. Lindenberg. Frames. eds. Papineau. Willard Van Orman. 1922. . ed. Pigs for the Ancestors. Pp. 1993. Lamont. Manchester University Press. 1963. New York: Free Press. . Pp. Homo Socio·oeconomicus: The Emergence ofa General Model of Man in the Social Sciences. 290-324. Boston: Kluwer. 1956. 1960. Embodied Knowledge: Thinking with the Body in Critical Anthropology. Popper. A.. Argonauts of the Western Pacific. Reason. American Anthropologist 88: Pauline Marie.. Wolfgang Balzer and C. Economic Anthropology. t 998. Fararo. transl. Mitchell. Berlin: de Gruyt~ Murdock. New Haven: Yale University Press. • A. . 1969a [1934]. The Post-Modern Condition: A Report on Knowledge. 1992. Deficits. Thousand Oaks. CA: Sage Publications. In The Blackwell Companion to Philosophy. 0/ Society. Pp. Popper. WeBer. Lyotard. Marcus. 1981. Marcus. J. and William H. 1996. 1996. eds. Kinship. Tsui-James. 1996. Karl. 1995. The Logic of Scientific Discovery. and • Cambridge: Cambridge University Press. Acta Sociologica 36:191-205. In Rational Choice Theory. Manche~. 1994. Obeyesekere. eds.13-338. 1996. Knowledge and Power in a South Pacific Society. Stanford: Stanford University Press. The Foundations ofSocial Anthropology. Structural Anthropology. Roy A. and Michael M. Siegwart. and Charity. and Exchange. Rhodes-Livingstone Papers 27. Amsterdam: Radopi. New York: McGraw-Hill. Malinowski. 1949. Journal of Institutional and Theoretical Economics 146:727-748. and Jonathan H. Pp. Cambridge: Harvard University Press. 1957 [1948]. Structure and Function in Primitive Society. Coleman. Pin. 1988. Siegwart. Clyde. Paul. Neil de. Coleman and Thomas J. A Theory of Justice. Maryanski. Press. London: Routledge & Kegan Paul. RDdiger. Marchi. 1961. From a Logical Point of View. Lindenberg. Princeton: Princeton University Press. The Apotheosis of Captain Cook. Chicago: University of Chicago Press. 1984. 1989. rt. 1969b. Stuart. The History ofEthnological Theory. Turner. Karl. and Bruno S. ed. Jean-Fram. 1968. 1986. Pp. R.Stanford: Stanford University I' Nadel. George E. David. In Assessing Cultural Anthropology. Claude. LogikderForschung. New York: McGraw-Hill. 1994. 1964. ed.T. Lindenberg. Popper. Frankfurt: Suhrkamp. and History. Constitutionalism Versus Relationalism: Two Versions of Rational Choice Sociology. pp. Princeton: on University Press. John B. New York: Rinehart & Company. C. Social Production Functions. 1972. Nancy. Siegwart. liffe-Brown. TObingen. ughes. J. Frey. Jon Clark. 1992. 1990. Hermeneutics and the Human Sciences. Joseph C. Putnam. P. New York: Basic Books. Germany: Mohr [Eog!. Quine. 299-312. Quine. Washington. 1990. Word and Object. Oxford: Clarendon Press.

Cambrtdge University ress. Richard A. Stuttgart: Reclam. Anna Lowenhaupt. U' ity of Chtcago Press. 'l1li " c Spira. Neue Betrachtungen Uber Aufgaben und Ziele der Wissen· schaftstheorie. Wissenschaften. 4. Steedly.IP. OswaId.86.hu~~:r Weber. Spiro.. Greenwich. The Construction 0/ Social Reality. 1-64. It haca. 1994 Gi'is Favors and Banquets. 1994. In Human Motives and Ca/turaJ Models. . nlvers . Steven. . Smith. . Germany: Mohr. Claudia. 1995. 1992. Johannes Wmcke/man n . Barry. . P e . 1957. 1973. 1943. In the Realm of the Diamond Queen. . . Wolfgang. ects. Co rnell University Press. A Cognitive Theory of Cultural Meaning. . . 1993. Die ObjektiviHi. and David G. Stegmilller. Cambridge: CambrIdge University Press. Pp. . Pp. Richard R. W0 . Economies and Cultures. Thousan Wcrner. ax: I h' Methodologische Schriften. Cambridge: Cambridge University Press. 211. 1990. Patrick. eds.D. 1997. The Structure and Dynamics afTheories. New York: Free Press. ' Cambrtdge. : . Berlin: Springer. Hanging Without a Rope. 1992. an '. NI: Prenllce·Hall. Thousand Oaks. APpliCGtl°Sns. Melford E. CA: Sage Publications.:. Wolfgang. ran . Manchester: Manchester University Press.ccerman. ass . ! I Wel'ch. 1986. CA: Sage Publications. Models and Motives. Stigler. 1957. Cambridge: Cambridge University Press. Wolfgang. h . eds.U1. Claudia. Oaks. The Pas/modern Turn. Probleme und Resllltate der Wissenschaftsthearie und Analytischen Philosophie. Cultural Anthropology 1:259-286. Vol. I y and loseph Galaskiewicz. 284-300. Pp. Yang. thesis.. Gilbert Herdt. Berlin:*pringer. Pp. 1987. Stegmilller. ed. Anthropology. l l -1 l l l l l . 2 vols. Frankfurt: Fischer. Princetan: Van Nostrand. Strauss. Berlin: Springer. Advances in Social Network 'IIII.. n IS F kfurt· Fischer. an . ed. The Structuralist View afTheories. 1993. . . Social Network Analysis: Methods and nnan Stan Iey. StegmOller. 1992. Victor W. Claudia. M 1972 Wirtschaft und Gesellschaft· TObingen. • 'J'.t sozialwissenschaftl. 1979b. . Seidman. Wellman. cd. On the Strange and the Familiar in Recent Anthropological Thought. Turner. Oxford: Blackwell. 1990. and Bruce Winterhalder. . Princeton: Princeton University Press. . Wolfgang. Steven. \996. Strauss. If Eric 1964. A Cognitive/Cultural Anthropology. Ph. New York: McGraw·Hill. StegmOller. Pas/modernisn and Social Theory. Pp. d d G Mark Schoeptle. . Introduction to Logic. Witk. 47-61. Shweder. John R. e. In Cultllral Psychology. 2d rev. Systematic Fieldwork. Mary Margaret. Wagner. Social Structures: A Network Approach. MaJ<· \917 [ hI M th dologische Schriften. 1979a. Pp. New York: Airline de Gruyter.. eT: lAI Press. eds. h 1968 Der Sinn der Wertfreiheit der soziologischen uod okonomtsc en Weber.. Weber. Street Corner Society. 1997. cd. 1904 [19681. Princeton: Princeton University Press.. tan e • . Walther von der Vogelweides Lied von der lraumliebe uod Quasar 3 C 273: Betrachtungen zum sogenannten Zirkel des Vers!hens und zur sogenannten Theoriegeladenheit von Beobachtungen. Tsing. s ":0 Ana/: . In Rationale Reklnstruktian van Wissenschaft und ihrem Wandel. Eric Alden. Seidman. VoL I: Wissenschaftliche Erkliirung und Begrundung. 1969. Englewood Cliffs. 1996. Chicago. Wolfgang Stegmoller. p.. Y . 1-20.. .icher und sozt. Johannes Winckelmann. Berlin: Springer. In Probleme und Resultate der Wissenschaftstheorie und Analytischen Philosophie. Franklin E. eds. Tjan Sie Fat. d S D Berkowitz eds. and Naomi Quinn. Wolfgang. Representing Kinship: Simple Models of Elementary Structures. Erkenntn1S. Vernllnft· Frankfurt: Suhrkamp. Strauss.:. Evolutionary Ecology and Human Behavior. lames W. and Naomi Quinn. ed. '. In \s e 0 22~277. . StegmOller. 1994. Cultural Relativism and the Future of Anthropology. M . . Leiden: Faculty of Social Sciences. Schism and Continuity in an African Socief/. ax. . Wh te Wi11iam Foote. and Kathenne Faust 1994. Wolfgang. In Assessing Cultural Anthropology. Suppes. Cambridge: Cambridge University Press. 1994. Melford E. Boulder: W~stview Press. Robert Borofsky.Epistemology 87 86 SCflWEIZER Searle. Ray 0' Andrade and Claudia Strauss. 1976. May f air el.

we should hope to follow any systematic course of action is not clear in the dictionary definition. While much has been written about the virtues of various metalanguages of method. Why. and the very contingency of choice-imposed by the varying needs of different disciplines-belies the common assumption that methodology is neces- • . methodological specificity will.' In a theoretical phase in which anthropology is still recovering from the detennined resistance of classic structuralism to both historical specificity and textual philology. there are nevertheless good reasons for using etymology-which is both historical and. sarily dictated by a logic that stands outside culture and history.~ 1!i5!!! Three In Search of Meaningful Methods Fieldpaths: On the Meaning of Method Lost as a dead metaphor in the word "method" is' the Greek hodos. . road.. I . path. "way. given the contingencies and uncertainties of life' s constant negotiations of meaning. philological-as a reflexive methodological instrument. from currere. be a clarification of the goal to which the path is thought to lead and of what is to be reveaied in achieving that goal We begin here by etymological reIlections on some of the tropes buried in the central tenns under discussion. The problem is compounded in the analysis of meaning by the fact that the very object of analysis is also the medium for the exercise of the method. at least traditionaIly." Webster's dictionary (1959:1548) tells us that method is following some systematic course (Latin cursus. 89 . This is an epistemological question requiring resolution for particular cases. "to run") of action. in most cases. none-certainly in the fields ofsemiotics and semantics-has been plausibly upheld as context independent For these reasons.. lAMES FERNANDEZ MICHAEL HERZFELD .

under what :~a~ts recognize it. ca d" . This is especially true for the mterpretatlOn of matenal remains. arguments used to center on whether lineage segmentation was an accurate description of social reality or "merely" a folk model (Peters 1967. rather than by literal-mindedly reJect~ng the model Jt r~presented as "'inaccurate" and as divorced from the empirical dom~m. meaning is not exhausted by a listing of referenti~l correspondences or denotations. and md~ed vlrtu~lIy every semiotic domain of culture would lose Hs evocative power. Miceli 1982). Gable et al.in social interaction and its attendant cultural configurations. Bond and GilIiam 1994' EIHaj In press).) I l l 'I I l l • . to grapple with the innuendoes. of "how it worked. and. shaped by and shapmg of personal understandings. because the "natives" supposedly did not have "our" capacity to detach the.we th echo of an etymological root that we can recognize mean that our : t. does their puzzlementwhe~. and it would be a poor eth~ography that failed. would be an open mVllatlOn 10 engage m linguistic and other forms of cultural incompetence. The literal . we can see how individuals manage the observed pattern of political alliances by using the idea of segmentation as a legitimation for intended or already accomplished acts of violence. where any cntlque of factuality must appear to be corrosive of power-precisely the point that VICO w~s at pains to make in his New Science (see below). pronouncedly reflexive. Moreover. lied "trailing clouds of etymology"-a motif to which we shall return. The inquiry compels reflection on its own organizing tropes so as not to fall prey to unrecognized and unanalyzed path dependencies-that is.1O we can learn a great deal more by examining how the various interpretations of that site are used by interested factions and individuals (Bruner and Gorfam 1984. to teleological expectations that are concealedwithin the presuppostions of the methodology Itself. while pointing out the obvious is-<Jbviously?-a primary task. The critical tenn is use..90 FERNANDEZ f HERZFELD In Search of Meaningful Methods 9/ Attention to etymology also shows that the anthropological study of meaning ~anno~ . and for what purposes (according to their own or others' attributions of motive). in some absolute sense. lIke an Anstophanes joke painstakingly dissected in the apparatus criJiCllS. we are particularly invoking what J." To be sure. (A good example is the significance of our own cultural and "rational" distaste for the klOds of analytic disorder just mentioned: They are violations of a categorical order. eXCuses. . the object of empIrIcal InvestigatIOn IS not sernanticaccuracy. however discomfitting they may be and howeverdeeply they offe~d our own methodological desire for order. providing paths by which inquiry can proceed in order to discover and understand that which is not so obvious is crucial to method in the study of meanmg. once people became too directly conscious. ~s a result. empirically identifiable precisely because they have been dlsaggregared from the processes of routinization that prot~ct th~m from. To take a particularly salient example. for these ignore the temporal mstabIlIty of semantIcs: Meanmgs rarely stay the same for fong. and the primary methodological challenge is to assemble as full an account as possible of the contexts of use for each item to which local commentators attribute referential meaning. Second. writing of ere.~ Hero k h m about this mean that the connection is. rather than asking which narratives about a historical site are Hcorrect. the study of meaning is also. as a consequence of its dependence on bemg able to respond to such indefinable but undeniable problems. Salzman 1978). too? Or. 1n the most reductionist of terms. Perhaps the most obvious locus for this kind of distinction between literalist and symbolic underst~ndings of fact lies in the management of history. but the use that informants make of symbols and models. not l COndItiOns. mste~d of worrying about whether a social group '"really" pracltces lineage segmentation. there are obvious aspects of meaning (Rappaport 1979: 173-221). and factionalism. Austin (1971). represent the irruption of an awkward reality into the symbolic order of science-an instructive irony. challenge in everyday life. more significantly. Shryock 1997). much that is meaningful . It could also be asked who used it. the anthropologist treats the claims of referentiality (and other forms of pragmatic essentializing) as themselves elements of the social universe of . or at least there seem to be some. b t Without necessarily reviving Levi-Strauss'. In this way. Once It became pos~ible to treat it instead as a metaphor for political uncertamty (Meeker 1979:222). In this approach.mpttons ~f pure reference. but obviousness is itself culturally contingent (Douglas 1975. is not at all obvious even to the actors. 1992.content attributed to the narratives and the sym'bols deployed m thelt perpetuatIOn (see Handelman 1990) are interesting primarily for the uses to which they are put (see Roberts 1997). L.symbol-production. Much in such analyses will not be reducible to a verbal account of "what these symbols. mean:' There are many intangibles. Thus. and other connotative meanmgs by whIch people construed the intimacies of everyday existence. Indeed. as such. given that symbolism is so often conceptuallyopposed to science. Herzfeld 1991.rest on assu. " and "unconscious" models-m WhlCh the former were tmpoverlshed ""consclOUS . any method for the study of meaning will entail reflecting on the tropes that appearto frame its argument in an interesting and intelligible way. avoldmg these cloudy issues. visual and olfactory sIgns. full SlgOlficance of the SOCIal phenomena I." immediate interpretations to the real.n question-we may certainly assume that much of what we acknowkdge as poetic "resonance" in speech. For example. and this made it possible to see segmentation as a pragmatic resource in the marshalling of historical memory for political purposes (Dresch 1986. ironies. ThIS also made Jt possible to recognize the constitutive role given to m~taphor by effective and interested social actors. Third.' (1963) unfortunate IstlnctlOn e ween . loyalty. notably archeological sites because these physically imposing cultural objects force us to confront the fallac~ of misplaced concreteness head-on. in some absolute sense. Thus.

we do not thereby dispose of the ever-present possibility of contingency in the actual working out of human interaction. argued that when thinkers lose s ht o~ the material..2. ~ !fFtrameworks as ethnographIc data In a c~mparatIve project engaged by both our )('h Informants (who are usually.ani. on the other. by performance. Vico (and.. Method is itself a project that . lSnstant calibration of the methods we use to elicit data with our informants' #. we recast our theoretical .ute~ the criteria of relevance for its own objects of discovery. any method must be in some sense both a path and the project of following that path.. it has clearly strayed from the path of empirical inquiry and fallen into the trap of pure self-reference-thevery danger against which Vico's etymological subversions were Intended to provide a provocative and persistent warning' Method implies systematization of procedure usually leading to the goal of greater clarity or parsimony in the explanation of human affairs. then.»s understand the range of coderstanding of what constitutes appropriate discourse. What IS mV~lved here IS what m his New Science (1984 [1744]).' . for example. and corporeal metaphors out of which their abstract i~:as arIse.'~. Vico. trying to ma~e. IS often lost m the mesmerizing abstractionism of intellectuallifl and has to b~ ~escued from it. Sewell 1996. we may actual..ued th.ons to the human body and its parts and to human sensahons and human passIOns" (Vico 11:2. taken for granted as inevitable and universally self-evident. and :ellectual activity. ' Recognizing. Following a path of investigation is a methodological performanc. . Lakoff 1987). and sa:red wealth terms (Maori). often arrog~nt narcissism and solipsism-in the tropes of en~lrely local and embodied sensatIOns and experiences as these are understood and articulated by the people we study... ' 0 nson Arguments like this-whether of Vico or in the latter-day cognitive approach to metaphor-should be partIcul~rly congenial to anthropologists.~.1 ~~n place of vague ge?eraliza~ions about m~aning •. for Bourdieu [1977) and Jackson [1989]).+wb t people "mean by meaning" (see. often indexed in English by passive-voice constructions indicating an avoidance of scholarly agency and accountability. There is a delicate balance here by which a method must be purposive without being deterministic-that is.d~p~~d and their di.2 [1984:248J in Herzfeld [1987. the presence of pure randomness and contingency in the undertaking of inquiry through the use of method.fference from the tropes characteristic of o~~:: frames ?fmtelhglb.92 FERNANDEZ / HERZFELD In Search of Meaningful Methods . Indeed. after all. At the same time. of the degree to which it may be overty influenced by a teleological dynamic which overcommits it to final causes and a priori endpoints of pathlike reasoning (Goldstein 1976. describable social rhetoric.i:::'. for. with . Vice argued that "in all languages the majority of expressions for inanimate thmgs are made by means of allusi." in this active (or "performative") sense: The choice of the path we follow determInes and in some sense guarantees the "truth" that can be asserted as a consequence of following that particular path. see also Fernandez 1993. Tilly 1997). anthropologists have argued in recent years that good ethnogr~phic field methods must provide a clear account of the field paths followed to obtam the data out of which the local knowledge and the eventual ethnography built on it are constructed (Sanjek 1990:398-400. who deployed etymology agains~ ~t~ no?"atlvlzmg uses by the official establishment to develop a critical. for a . as in Salmond's (1982:82-86) companson of European and Maon frammg of knowledge claims in territorial terms (Europea~): on the one h~nd. it is dynamic and actively adaptive to CIrcumstances rather than preordained and given." . This . . to fulfill what is taken and received as mtelligible. 169-170]). He arg..:~. I996}-a position that exposes the fashionable distinction between positivistic and interpretive styles of analysis as itself caught up in an identifiable. ·~. that the concept of "method" IS Itself ~n~es?e~ In a complex Cb' tory of shifting significatIOn permIts a determmedly empmcal mSlstence on the .13. When method appears to have that effect. which anchors the dynamic of the trope~ In the prototypical corporeal and sensation-anchored domains of human expenence (the experiential gestalts of life) (Lakoff and Johnson 1980' J h 1987. It has been similarly argued that historical method must be critically aware of its path dependency. sense of us) and ourselves. and for the reasons that Vico was the first to recognize. and... at the same time.at we must remember these crude origins of abstract reason If w.. social interaction. and consequently subversive method. the eighteenth_ century NeapolItan ph Ilosopher Giam battista Vico called "poetic wisdom"-wisdom about the experIential sources of our understanding and about the ways in which knowledge is represented and organized by various devices of the imaginationO~~ as both to frame and. allegedly immanent in what it is used to investigate and thus self-fulfilling. and. has esca ped our corporea IIlIstoncal selves.wisdom. This old point about poetic wisdom has been more recent Iy .hty can be rea~i1y demonstrated. they also lose a sense of the experiential grounding of their knowl d Superficial intellectual enthusiasms and actual loss of knowledge are the effe:tf:~ outcomes. a osition of Jakobson's [1960] "poetic function" with local concepts of me. The probmg of these allusions is basic to Vico's notion of poetic ~Isdom. in the sense that method conslJt. for it enables them to lodge ~he sources of academIc abstractions-the scaffolding of learned theorizing wIth all ItS attendant. And the particularlr e on which they ..~~}' 93 I:f. Thus. In particular. Thus. By juxtaposing our interlocutors' wisdom and theoretical capacity with the formal knOWledge of academic writers. while we seek to condition and control through pathlike procedures. Its claims to transcendence and sophistication . defamIhanzmg. Herzfeld [1985]. social. it is also constItuttvely performative.• The "frames of intelligibility" (Geertz 1973:36) that we use are thus not t0 b . more recently.ly be able to develop more useful tools of analysis to decipher local meanings: The SImIlaritIes and differences that emerge may help.e a~e to a~OId the Ignorant and ultimately degrading conviction that our thought. developed In the cogmtlve theory of metaphor.

'on f ~ . That dimension is evoked in the subtitle of the collection that has stimulated so much recent inquiry into cultural poetics in anthropology. ams answerable to what M allOWing V. L. in another context. ::f$f'"lDd to make . oft en turns out to be For poet"c w' d . But the creative defonnation of norms-what. Parkin 1982) _"semiotic ethnography" [Herzfeld 1983]).t" . lSa vantaged may b SOCJa solidarity (for example wI'th. to quote Lim6n again. on the .\Wile also sometimes contributing to cultural change.matnmg open to Con. .stmgUlShable entities. . affectations of numer. co (Tagliacozzo 1983) 'd' . I ' . so to speak to account 'or thPel: yan thereby challenges parsimonious . And. w en a customer osten~ c?~sjdered sociaHy disruptive.ns by the development of use or action theor' f e: In thIS It anticIpates hIStOrically . there may be no actual competition present in a given situation. tome cultures being more agonistic" than others.'cal or monetary imp " merchant pretends not to wel'gh prod fi reClSlOn-such as when a . .jf::societies' conventIons to ways that. l --. It attends to the dynamic ofm " y 0 meanmg m the abstract. Colby et aI.94 fERNANDEZ I HERZfELD In Search of Meallingful Methods 95 Jms epen on Its re . Eighteenth Brumaire (]978 as they please. here perhaps' [1852]:Section I) that. and only to presuppose that they were. Writing Culture: The Poetics and Politics of Ethnography (Clifford and Marcus 1~86). . P jects In the socJaI and cultural affairs of those studied (Same 1963) includ' th ' .iJioetics (Herzfeld . Such assumptions too easily . The first. on the m that the various traditions of semiotic and symbolic inquiry in anthropology 'de.. But we answer i::~. And notable recent instances of the cultural poetics approach (Crain 1989. 1997a) is the study of h~w social a~tors play on their f?. We also must recog. . Dolgin et aI. 1981). 1977.li'i In pursuing a social poelics. as far greater complexity in the processes b.." Contestation thus lies at the root of the realization of human existence.1". to detach our methodology from assumptions (such as '('1"':~-\" . recognizes the presence and influence of ra" .k:~~ i In The. . uce care ully or h tatlOusly avoids looking at the scale-demon : ..mn g (S configured. even when we Jive in societies where the dominant ideology is a denial of any such thing.ew IS that poetic principles guide 11 "r. they do not make it just ~ tm~en~Y-thecombinationofpredictabilit and j . 1971]). preemptive 10gic wouldo:I~~:' ~. in focus.. that is.j:. can e a poetIcs of pre~ P an antiempirical.rellectthe social norms of our own cultural milieu and thereby illustrate perfectly what we havejust said about the dangers of solipsistic methodology.. is pounded in a recognition of the role of cultural form in social life. In etenslble relIance on the not..an extravagance out of restraint (whether social or aesthetic). the war of position as L. wh~ ~s produced IS an understanding of the poetic wisdom is sensitive to that corn I c me. Here we are obliged to recognize distinct but interrelated traditions of tanding: semantic anthropology. 0 pure relerence. d'" les 0 meanmg as oppose d to an empmcally ' . af)(. mg e presence of goals a .rnea.1985.ii!l'. and a long and complex tradition ~o(r!letorical analySIs m several semIOtic domams and mtellectual tradltJOns-soclal . .. ':. C Ion ID t e social world is as instrument (Gibbs 1994) Indeed t th I IS an abstracting and concept-forming • . that also reduces their servICes (HerzfeId 1991: 168-174). "on the lestheticallysalient and culturally embedded textualities and enactments" (p. The notion of xl t accounts.~\\~:.'mon calls it (evoking Gramsci).'n ' I • e a way of creating Socla or a pro f<e"lOn ' ) ' h ' a whle people can help each oth er-a f<arm of class on the basis of sac' J ' dep~ndence on often hostile market mechanis I~ capl~al." e r own hIStory. .. dlose that have dogged studies of "Mediterranean" or "Melanesian cultures") about ~. In fact. But the praclitioners of cultural poetics desire in their interpretations to focus. projects. ~'~~~i\~ .. I l ~i I I l I . But It entails caution about th " SImmanent m the procedures and its empirical cia' eda PflOdrl pres:nce of final causes in its owQ social Interaction it studies It rem' Y mprovlsatJOfl that characterizesthe '. " rm~n~e-t e. " 0 e extent that ther b ' clslOn. It is vital."wbere the competition over the control of good form is palpable. Conversely. By this we mean that th . and symbolic anthropology. ' h' . " Social and Cultural Poetics as a Method for the Study of Meaning in Performance We noted above that method most usuall work (powerful simplicity) of explanation with th Y s toward greater parsimony as the study of meaning is concerned wh t . Building on several strands-the recognition of cultural form (for example.dl. Boon (J 982) has called "the exaggeration of culture"-must always be a response to people's understanding of wbat Bourdieu (I984) has theorized as the value of "cultural capita!.' . The tenn "cultural poetics" surely does not deny the contestatory and agonistic politics of everyday life.t it is possible to perform non performance. r pammonY-for th' d ' . ~W!JakObson's concept of "poetic functio~" (1960).e caveat that frequently enough. modes of hierarchy lit of inclusion and exclusion) in interaction (Fernandez 1977a). • Instead. especially by the economically d' d per ormancesofsuch careful impre. d a eJ/ec/{ve and aj]ective social intere mm engaged m a r 'h ' much if not more a poetic instrument as 't . LImon 1994) seek to keep the practical politics of everyday life. IS om IS notthe stud f '.. the traditions of . ' action. of course. mcrease theIr power or status) i~. Some critics have charged that this approach can only work in agonistic societies . that there have also been two distinct but also interrelated deployments of 'c wisdom: social poetics and cultural poetics. Austin [1962. . while "men m.~" or "action" theory in semantics (analogous to practice theory in social 7~!}ailthropology and largely derived from the work of J. these are not necessarily se arate " e. eanmg In actual perfo h h W le slgmficance is conveyed in actual social lit. And this requires us to t: IS t : heart of the matter that should ocus on sIgns themselves. r(peacock 1975. Our v. effecti::ratefidlsdam for calculation. t 4) of H '-. ms In obtaining cheaper goods and The conveYing of significance in action' h concern us. to be boastfully modest and demure. If successful. on the one hand (Crick 1976. social poetics. which is clslon. . elf re uctlveness as It .

intended to animate if not Bourdieu 1977:113). We can also approach physical monuments as ornaments of the body politic. inchoate in their. thereby treating the most obviously material manifestations-architecture. the lingering association of poetics with a purely textual reading of the Aristotelian discussion of rhetoric and approaches that are styled "materialist. they have faced two criticisms: that they are too language based or logocentric and that they either neglect or ignore the dynamic interdependence between meaningfulness-the significance present in symbolic activity we might k practitioners o into recognizing th~ technical features by which we grasp s . this approach ironically speaks more directly to the materiality of experience than do many of the f I I I say-and power relations. has . We nevertheless wish to emphasize that the reflexive and recursive nature of language are main challenges to th~ complexity through which we must find our way and that. The transgressive move of using poetics as fundamental to a social science in a sense explicitly separated from that of poetry is. for.I I constitutive of major shifts in the balance of power (Bailey 1973. the medIUm of language must often be our principal guide. this alerts us to the possibility of a Jakobsonian classification of the uses of metaphors and other tropes. exercise of power. analogous to those of the body personal. allow us to track probable causalities. that its practitioners have often felt obliged to add the words "and politics. careless uses of the term) "poetics. Yet both narratives and carefully observed sequences of events can. even raw power. in curiously similarterrns." where a clearer definition of poetics itself should have made that phrase redundant. V:e further argue that political dynamics. The writings of F. From the first. Harris [1974]). at the suggestive in this regard. hut failing to recognize its existence.base natures. p d usually logocentric sense by which the term poetics is commonly understood an uld not obscure the deeper and more general semiotic properties and processes sh0 k . but the underlying principles-as must follow from Jakobson' s communicative model-were unequivocally social. can usefully begin from the concept of "poetic function. G. and ." Cultural poetics has been so heavily associated with textuality. the bodily beings logocentricallydenoted by the social pronouns. Bloch [1975] and. J \ This last point requires some elaboration.I). and opposition-idioms ideally SUited to. the "socially produced. fonnal properties emerge more from the social relationships among actors than from the sometimes "fuzzy" semantics of their language itself (see also . for example-as tropes of collective identity that is also personal and familial. it is inherently also about power. which is to say the cultural preoccupations of that life." This is especially applicable in the sense that the embodied nature of human knowledge is embedded in relations of causality-especiallyas perfonmativity-that are often channeled precisely through those phenomena that self-styled materialists dismiss as merely symbolic and therefore epiphenomenal (for example. pragmatically. for example. as opposed to the rhetorical. and perhaps originating in.96 FERNANDEZ I HERZFELD In Search of Meaningful Methods 97 social life. the approaches of social poetics and cultural poetics constitute a contemporary and comprehensive program--eonstellation of paths of inquiry-for the-study of meaning. The apparent concreteness of language is a major source of this fallacy (misplaced concreteness). is equally unjustified methodologically and may lead to false assumptions about the very nature of knowledge. The unconscious may be analytically opaque. for they show that an inadequate perfonnance can be least. 1994. (Fernandez 1974). The common. Methodologically. His definition was ostensibly a purely linguistic one and all the material adduced to explore it was taken from the verbal domain (Waugh 1976). in part. and this overdetennination is nowhere more evident than in discussions of (or. inasmuch as poetics is about creativity. was much concerned with the poetic uses of symmetry. Moreover. to say the least. Taken together. It also suggests that similar principles might obtain for the entire range of human semiotic domains. Jakobson (1985). we should not forget that even the most determined (and determinist) materialists gain access to the majority of their data through conversation and archival consultation. Indeed. the play of power. Meaning is never reducible to verbal form or restricted to language itself. restrictIve.ts lOescapable rhetorical or poetic component. audience). however. We speak of a "body politic. 1983. which become significant especially as they are recognized by the social actors in question. . Still. destination." as originally formulated by Jakobson (1960). and by extension a cultural poetics. . we would argue." and we can trace its vicissitudes in the disciplines and constraints imposed on individual bodies in its name (Cowan 1990). worse. Poetic Function A social poetics. narratively mediated and relatively unconscious ideological responses of people" to recurrent problematics and challenges to their situation (1981 :Ch.the h C lace of the rhetorical and the expressive in human affalfs. chiasmus. se~ also Paine 1981): We "do things with words" (Austin 1962) and all manner of other signs. that would be sufficient to ground the inchoateness and ambiguity of so much figurative language in the observed details of social interaction. defined in terms of social actors (sender. Bailey are . for example. They seek to probe the "political unconscious"-in Iameson's phrasing. the appositional aspects of social life. We tried to make clear above that poetics and politics have been mutually complementary from the first. o Such an approach illuminates the way that the main entities of social life. however we label it. parallelism. obtain social identity and the possibilities of extrabo~lly SOCIal experience by poetic processes of predicating physically sensate signs on themselves. dramatic statement has tended to confine the term to the verbal and to exclude it from the raw. f human interaction that a social and cultural poetics seeks to evo e and appraise.

the apparently superficial play on notions of sexuality offers an arena for play~ng?n an~ contesting conventional expectations about other social relationships. these are often cumbersome or recondite and pose a practical challenge to comprehension and interdigitation with the verbal (see Famell and Williams 1990. is about that playfulness: how people master the rules to demonstrate that the rules have not in that authority occur. We might argue that their relative absence from the literature has been diagnostic of their very success at insinuating themselves into the realm of what Mary Douglas has called "self-evidence"(1975)-the common-sense. But we must recognize that nonverbal signs are important in the flow of meaning-that they may. entails videotaping sequences of dya~lc . in whom it generated a more abstract understanding of the concept of complementary opposition. indeed. betwe~n social relations and aesthetic elaborations. In Newfoundland. We thus evoke the theme of play and playfulness (as opposed to planfulness) (Femandez 1994) as central to meaning -. supply the words being used.asons and under conditions very different from those of modem anthropology. FameIl1994).or use-oriented theories of meaning and the constant social and cultural transformations for which they attempt to account. the question of elicitation becomes vastly more difficult. 1972. then pla~Ing them with the sound turned off to other culturally knowledgeable actors and askIng the latter to comment on the interaction. (This consideration also undoubtedly undergirded his occasional collaboration with Levi-Strauss. play a far larger role in the constitution of sense than will ever be possible to recuperate from the data available to us. Gesture and posture can't be reduced to verbal equivalents and must be indicated by notational systems of various kinds."p~sed c?nformism). Preziosi 1979. A complaisant view of meaning privileging reference over use would have been of no practical use in understanding how social actors playfully manipulate the truth claims of seemingly fixed symbols to satisfy their own interests. In this sense. In other words. Here. respectively. for example." This expansion of etymology enables us to expand the notion of poetics to a more technical sense than originally suggested by Vico and is more consistent with the thoughts of Jakobson. The Elicitation of Nonverbal Meanings in Symbolic Interaction There is always the tendency for discussion of methods. Vico's perspective. The Play of Meaning .98 FERNANDEZ I IIERZFELD In Search of Meaning/ill Methods 99 powerfully iconic of that life. Herzfeld 1985) in a manner sometimes explicitly recognized in the associated terminology. At this p~i~t. ences and hVIng structures. Often used by the powerful to legitimate their power. to become logocentric. VieD hlm~elf eXhlblt~ ~ certain playfulness in his subversive use of etymology-his tracmg of semlOtlc and tropological origins.) inherent in all human institutions rather than their capacity to resist forever the erosion of time. When we turn away from language. Note that we say '"semiotic and tropological origins" rather than merely "the origin of words. although conceived for very different ideological ". but-by the same logic-often also conducive to a loss of referential meaning. but by no means confined to. reproduced the contrast in the social us~s to whkh these two types of room were put. IllOs anticipatedthe present-day interest of anthropologists in action. When linguistic models predominate. methods usually entail the elicitation of taxonomies and the systematic recording of violations of those taxonomies-"matter out of place. we espe. but in the act of reclassifying and of challenging or cons~antly changmg world taken-for-granted aspects of a culture that are precisely what anthropology should defamiliarize and bring into focus. 1992b. in turn. consistent with our view that meaningful knowledge of the IS found not in classification itself and the planfulness that It represents (and that may be mdeed a suppression of active knowing in favor of an ". between artisans and their apprenticeships). it may also be used subvefSlvely or playfully to undercut that power. By noting the instability of meaning in all symbols-including. Blier 1994) are per- role of meaning creation in the construction and maintenance of political authority and an illustration of the subversive activity by which cultural dynamics and change haps the m~r~ obvious "diagrams" of the poetic resonances between bodily experj. Poetics IS partIcularly attuned to the playfulness of social life-indeed. linguistic elements-Vico provided both a method for analyzing the Parallels. namely. particularly when etymological method is invoked. Pocius (1979) showed how the distribution of more and less hierarchicaIly organized rug designs in formal front-room and familial kitchen spaces.• i . Another example is provided by rhymIng couplIngs that reproduce sexual couplings (Dundes et al. and reconstItuting eXIstmg classrfications. One . play reveals fonn as it revels in it-sometimes in the moment of recasting or destroying it. i creation and as diagnostic of it. etymology is a kind of diachronic pun: It undercuts received and conventional meanings. mastered them.approach. as in architecture ( Of. archltectomcs) (Femandez 1977b.cially wish to focus attention on the notion of play."teractions (in this case. better. This is. This avoids facile equations of bodily movement with segmental speech units and permits greater 1 I ] ." in Douglas's formulation (1966). recently attempted by Herzfeld. This was Vieo's understanding ofetymology-a device through which the discovery of the instability of the term-referent relationship suggested the "civic disability" (Struever 1983) answer some specific questions about the nature of the interaction. to the play on the conventions ofform.

The method requires a set of dyads that can eliminate random effects. given that the local term for apprentice in Herzfeld's Cretan field site means "foster son" as well. it is above all historical. irony. Our reconstructions of meaning thus starts from the recognition that before we can say what the meaning of a given action or object situation might be. The second operation is analogous to the pragmatic turn to observations of actual change in use in ongoing action. respondents also accommodated a variety of sensibiJities (for example. it would be an impoverished anthropology that excluded from the domain of culture everything not reducible to verbal equivalents. which could then be compared with longer-term testimonies by former apprentices and artisans. in circular fashion. the inclusion of a father~ son dyad-in which the relationship was significantly wanner-helps to "materialize" the significance of kinship in affective tenns. Note. as it was not in Turner's case. we emphasize that the Vichian notion of poetic knowledgelwisdomsuggests the outline ofan appropriate methodology. Such circularity of argument weakens the claims of symbolic analysis to have escaped solipsism (Schwimmer 1978). visual as well as verbal. we suppose that all meaningful acts have a history and that the process of their formation can be reconstructed-we have propositional paths that enable us to appeal to at least two central foci: use and change over time. see Stark 1996). language moved the analysis beyond a purely referential and language-based model of meaning.' Returning to diachronic aspects. In this context. thereby reducing the effects of shyness and embarrassment (on which. but if we recast his phylogeny of human semiosis as an ontogeny of semiotic forms-if. While such elicitation is to some extent intrusive. However. Such study gives substance to the idea that cultural form is always in a state of emergence (Bauman 1984. in the absence of notational systems analogous to writing. thereby demonstrating the replication of ideologies of hierarchy in adjacent but discrete contexts of social life. Their very evanescence makes them ideal vehicles for indirection. moreover. This assumption ofethnographic presence and direct observation of actual use undergirds inspired studies of tropes and symbols. and respondents were invited to put their own words into the actors' mouths as a means of determining conventional expectations about such relationships. we may even conclude that meaning is itself an inappropriate tenn and that "meaningfulness" is the more significant and appropriate form because it implies an ongoing temporal and emergent process of successive usages that themselve~ have a range ofacceptabitity. Difficult as it is to specify methods for the study of nonverbal meaning. in other words. Bruner 1994). The difficulty is that it has tended to rest on contested claims of ethnographic authority (Sperber 1975. since all contemporary human communities whose symbolic universes we attempt to enter are composed of language-using subjects. It is the restoration of the centrality of time in all interpretation by following the paths that particular usages follow over long periods of time. Stimulated by the etymological impulse." lOO FERNANDEZ / HERZFELD In Search of Meaningful Methods 101 access to ideological presuppositions entertained by the actors themselves. Clifford 1983) or on putting forth formalized systems that often enough. 1985. it fits increasingly with a worldwide familarity with the technology that makes it feasible. Recovering Meaning in Use and Emergence Openness to nonverbal meanings brings with it a host of problems. justified their own logic. some social actors are willing to provide their own glosses). Vico's own use of etymology was notoriously unreliable. For example. Some sessions induced elaborate reminiscences. The first of these two operations concerns the recovery of simple image-schemata as these repose in primary experience or process. and other nonreferentialmodes ofmeaning. a criticism made of Levi-Straussian structuralism (see Leach 1970. This is the "argument of images" (Fernandez 1986. that the "diagramming" of social relations in the Jakobsonian sense allows us to recognize homologies of political relationship across two or more spheres (here. Although anchored in the concept of the image (which has usually been taken to be visual but need not necessarily be so-we shall briefly explore other iconicities below). The repetition of the same scene with different voiceover and subtitle instructions to the viewer. 1976). . too. ambiguity. the origins of nonverbal signs have usually proved harder to trace than verbal etymologies. labor and gender). we must first establish the range of acceptable alteration in the form of the act both at present and over time. and this helped establish ways in which informants could conceptually square their long-term recollections of ideal-typical relationships with the immediacy of experiences that were often very different. the basis of interpretation must become one of observed use. it works primarily with verbalized tropes invoked as the basis of claims to local knowledge. But the adequacy of the lexicographic model of referentiality is brought even more into question by the flows of nonverbal meaning often tied up in symbols in which the absence of lexicographic methods challenges (or shOUld challenge) the illusion of analytically reducible reference. Symbolic analysis has sought to escapethe purely verbal. of which an early and extended example comprises Turner's studies ofNdembu symbolism (Turner 1967). Unless this observational anchorage is made explicit. But if this kind of interpretation is to be applied to tropes and symbols not expressed in language or supported by verbal exegesis. Especially through this comparison of the elicited responses with the verbal commentaries from a rich series of artisans' oral work autobiographies. Preliminary inspection of Herzfeld's material from Crete shows that male commentators tend to sympathize with the artisans and female with the apprentices. attention to eye contact as well as gesture). the regrettable appearance of arbitrariness or invention-a projection of referentiality onto an analysis that is not intended to convey it-becomes the basis of criticism (Sperber 1985).

from culture to c~lture and time to time. 1992. But only long-term participant observation can e~ectIvelY rev~al the significance of different tempos. in dress or performance styles. context. Cret~n artisans. precisely because they follow social norms too ngldly. at a crude level. which protects us from the tendency to essentialize our own categories by invoking as explicitly as possible the primacy of use and action. however. for example. But the study of the meanings attached to or promoted by these differences is usually lacking. both etymologically and pragmatically. As a modest example. ~any societies for those who.s with res~ect to table manners may provoke anger or contempt and thus very significant SOCial meaning. This is clearly a complicated matter-but only the most determined Cartesian would claim that it was less empirical than the summing-up of such complex malters in the form of a referential taxonomy devoid of the amhiguity that lies at the heart of all experienced social life. Analysi~ ma~ thus.een lietymological cognates" among modes ofsocial c.n to ell er c~e orization depends not only on what they do and what relatIOnshIp they have With t. observation of the shlftmg relatlOnshl~ betw. and character of permissible invention. marginalize themselves from the fellowship of or~inary so~ial in~erchangeand thus from its manifold possibilities for inventive meanmg creatIOn. for example. but often identifiable.' In a truly Vichian mode. . . In societies with no recorded history of l l . Etiquette books register. in gossip for example." "wimps. infer what factors influence informants' judgments of each other and how predictable their effects are thought to be. It does so by keeping reification at bay. Mole 1990. Moreover.as li." '"hacks. A historical awareness of what has transpired may provide actors wtth qUIte flexible templates for present action as the range of ways t?at ~urner's (l?72. for example. partiCUlarly in soclelles m Which an Ideology of opposition to authority is explicit. . 197~) usocial dramas" can be played out indicate. see. for example. for example.t' 'gn I·nfluence" or '"degeneration of morality.that there is a. both complicates and strengthens description.over•. . We may thereby. but only by following their evolution through several editions (see. Sudden variation. entail listing conventions and mapping them onto a SOCial gnd or a temporal chart in order to establish empirically the range. b th the rigidity or permissible range ofbehaviors and reflect the acceptable rate of l:ng-term change. both openly elicited and ~upposedly "natural" (a r~vealing term in itself. that they usually retain for long periods of time only those apprent~ces who have :hown sufficient daring and resourcefulness to undemune the artisans fo~al authont. Some of these ranges are codified for us.jog. or acceleratlO~ of behavlor excites the search. There is some evidence that abrupt vanatlOn attests to superficial rigidity of normS and that a challeng~ such n~rms can conslltute status-gaining acts of daring. then. m whIch the methodological and cultural artifice entailed in such invocations of nature is clearly exposed). Indeterminacy: Tapping the Historical Consciousness of Permissible Variation Tracking variation over time in the uses of signs makes us aware not only of emergence of meaning but of a considerable degree of variation (Fernandez 1965) and even indeterminacy (Giddens 1984) in the meaning of common practices and their accompanying and associated symbolic components. iiTime servers. for example (Kroeber and Rlchardson 1952 [1940]). 1995) can we trace both the authors' learning path and the deeply hidden flexibility of the culture itself. The proliferation of such books and the telegraphic instructions travelers find in guidebooks often occasion mirth from local p~ople precisely because such formalization usually leads to the excesSIvely hteral observance of rules: True mastery can only be shown by bending those rules wlthm the shifting limits understood. deny that they encourage insubordination in their apprentices. flux of interpretation and meaning. for it reminds us that the empirical task." Actors teeter between SOCial . is the grounded inspection of lived experience (Greek empeiria) rather than an exercise in abstract classification far removed from the social contexts in which we gather OUr data. the difference indexes the other factors that contribute to the shifting. for meanmgful explanatIOns such . it can be checked against informant responses. . negotiation. and variations have greater potentiality for meaning production both for the participants as well as the ohserver. In the anthropological literature there has long been awareness of ranges of permissible variation. When a single action appears to constitute a solecism in certain actors but not in others. only by those who inhabit the culture's t? '1 ~ ~ l I: i I I lI intimate spaces. mvesttgation of the kinds of semiotic stability necessary. Etiquette books have their modem equivalent io the transnational world in the form of handbooks designed to help entrepreneurs negotiate their way through foreign cultures. to the effective maintenance. The use of hlstoncal consciousness 10 understanding the range of possibly meaningful action and the scope of the indeterminate is always problematic..omp~rtment becomes an. The converse al~o holds true. a too-sudden change in the way of doing th.y-:-a paradox that exemplifies the role ofethical and political ambiguity m many soclelles where agonistic relations are prominent. temporal dimension here in relation to which meaningful~ ness 10 action can be denved.ose who are judging. So it is vital to link the tempo of perfonnance (Bourdieu 1977:6-7) to the rate of change in the prior form of the act: Some acts are much more resistant to change than others. or in the graphing of the differential distribution ofnormative behaviors generally (LeVine 1973). . .In Search of Meaningful Me/hods 102 FERNANDEZ / HERZFELD 103 This move. The eVldencesu~gests.orel ." iiburcaucrats" are labels reserved m. Boehm [1980j. but also on ever-shifting percepti?ns of what conslttutes ~ permissible distortion of convention-an appraISal often mfluenced by the actors own mastery of play. Empirical observation shows more. In a given culture or society.h t boorishness or approbation for social daring: TheIr altnbutlO. or usurpation of power. supposedly.

since any threat to the or der of words logically entailed an equivalent threat to the order ~f things. . b d . capable of pure referentiality-thatthe relations among words are hke. We argue. Most famously perhaps. tha. Recording the ways in which even such "fixed" relations as those of kinship can be renegotiated in the course of political realignment (for example. similarly. notions of nonverbal etymology have been around for a long t. It is all too easy to reduce that knowledge to dismissive remarks about formulaic representation.. as in the cliches from which most headlines are built. in such cases.lOns of class and status. But even here. are perhaps a more useful illustration of what und~rl1es. and they certainly leave enormous areas of ambiguity unresolved. There were for the Hawaiians. idea of an etymology beyond language may strike some as a contra IC I. the denotata of which. in pointll1g up Imphcat. e . and interpretations made ofthem are inevitably subject to present interests (see also Borofsky 1997. Among . Modem nationalism is another arena in which.on . ?nl y the faintest traces connect them to each other and to the explicit"c~nfiguratlOns..104 FERNANDEZ / HERZFELD En Search of Meaningful Methods 105 symbolic forms such analyses are hardly feasible. . "the facts" become inextricably entangled with constructions of the past necessarily derived from present concerns and leading to a serious questioning of the very distinction between the supposedly analytic categories of history and myth (see Hill 1988. I et mologies are a special case of the more gener~1 rhetOrical pheNonve~:of s~lf_backgroundingor self-evidence that typifi~S all SI~n1ficatlOn ba~ed nomen '11 ne claimed to be timeless-ICOnlC1tles. a point that was re~lIzed In an important way by the anthropologist E.I~1S . through the common error of misplaced concreteness. Such iconicities are truly constitutive of cultural c. for example. of power. t rms but this offers a peculiarly ironic illustration of the problem. tic continuity with a supposedly fixed pomt m the pas~ (Herz e hngUtsS6_73). our insistence that a social poetics is inevitably political. architecture.~ hesitantly begin to emerge in what Scott (1990) has called the hidden tr~nscr.hey are translated into verbally explicit forms of protest and revolt. In the case of linguistic etymologies. Art histortans r~co. to be sure.the Fa~g of Gabon. the traces are more clearly mar~ed. cultural. the seemmg esemblance. Shryock 1997) also sheds valuable light on the ways in which the meaning of the past may be calibrated.s ~ which challenges-including claims of unscholarliness!-are mounted ~= d't' Th. ~~ .medcommo n se~se m be part of a critical anthropology's methodological commitment to meanmg.at I~~~uage . were represented as equivalent to essentialized objects-institutions. especla y 0 • d k h on r I .n 0 ~~mOlogy exclusively verbal. Sahlins 1995) about the Hawaiian understanding of Captain Cook's murder centers on just such a question of recovering and assessing interpretations from indigenous myth and ritual practice.ze the rhetorical intentions in the making of such links eaSIly.. Ardener(197l). Roberts 1997).apltal and the If97a. and sUll more of gesture (to name just a few examples). context-free assertions of the truth or falsehood of historical assertions are not so much untrue in themselves as simply unhelpful and uninformative. the Sahlins-Obeyesekeredebate (Obeyesekere 1992. and the aura of antiquity and deep knowledge surrounding such etymological linkages may be also clear enough. the reenactment of m~!hological events is a kind of collective remembering. The debate suggests the magnitude of the problem in recovering just what occurred in the events surrounding Cook's death. Academic discourse. often has a reductive effect on the information its users seek to convey. but this overlooks the fact that modem journalism is equally subject to formulaic design. an effort to establish a past for words. . These represent the outer edges of iconicty. figuring the soclOpolttlcal of peasant reslstance' they may e engage 111 con landscape long before . the problems 9ftranslation and the establishing of "coevalness" (Fabian 1983) are so great that they may defeat the best of intentions. Truth to tell. Rappaport 1994). values. W. Parmentier 1996. as well as to note ~:~~. nations. with extraordinary rapidity. Of course.. or at least most recently. that-as With poettcs ust challengingthat long-cherished assumption of a self_procla. to the contrary. as was customary in folkorists' accounts of songs and tales about historical events. Icon~ 1Sf" the things they supposedly denote-would logically make the notto. Only a pnor :s~mPt.~h. Standard etymology was indeed precisely a fonn of essentialism in action.onthat facticity can only be established through language and . J. and space. Its proponents sought "the truth" (to elumon)-the goal that made Vico's intentionally subversive use of etymology so dangerous. as neither informant nor analyst will have the knowledge of the relevant antecedent events in a form that can easily be matched to the sense of documentary facts available-albeit from differing perspectives-to both. ritual precedents for such a death. Steedly 1993. Gom~rtc~ s (1979) well-known studies of visual "rhetoric" (his term). ' Synchronically such nonverbal etymologies may be accesSIble by a kmd oft"angulation of link'ages in one area of culture: architectonics. we can identify parallel structures in the archltectoOlcs of .gn. to the political exigencies of the present. What is involved is an etymology of fonns that are not necessarily or invariably verbal: The attempt here is to discover essentialized meaning in the past.' . but the intractable problem of deducing what these precedents meant and what behaviors they might provoke or determine remains. of which provides a refuge from analytic scrutmy an ma es t em ~~a:~rt~~~s~f political propaganda and especially to c1ai~s of racial. it may sometImes suffice to c alm~ h asserts them and under what peculiar circumstances. for social actors.lme j De Jorio (1832) provides a minor instance of the attempt t~ con?ect the VI sua rhetoric of modem Neapolitan gesture with the forms of gesttc?latton repr~sented in ancient statuary (see especially Kendon 1995). 1996. when he pomted to the "blank banners" of resistance. When we move to the etymologies of art.

s howev~r infantilized they experience and maturity of IOtentlOn of ethnograp.bar between the symbolic and the material. village layout.' can be as Jenkins argues (1994). d d t abstract practical . h Iy 'Intellectualist approach. ' actively discounted or even lSappr d' k say or say-how (see real-world know-how (savoir faire) rather than aca emlc n~~. Gero [1991]. These tensions are likely to be especially acute in societies where the physical body is itself notably an object of strong. or the "body work. d tr'u gle for objectivity in the pursuit of meanmg edge. in fact. and dance forms (Femandez 1982:Ch. Verbal exegesis is unquestionably beset by ambiguity. bnngs the ethnograp e~ s s dl'rectly into line '" f h th graphic expenence more martlculate 10lponderables 0 tee.ml~g Coy 1989. because of t at SI ua IOn. The tension between what is observed and the specificities of informant exegesis is a constant in fieldwork. on the one hand. " attendant Imphcatlons 0 Sltua e e. f madequacy an 1 . so that variations among different exegetical texts may provide.. " iven society can be studied to important. Sys1ematic Aspects and Limits of Meaning Acquisition In informant exegesis. But this tension. overcoming the dearth of verbal commuOlcatlOn to learn a y g 1 . 'd uch It also has the compared with local methods of inculcatIOn and experience asds ' t res on the I' h are rather wor Y crea u advantage of forcing anthropo oglSts-W 0 f t ' s in which verbality is whole-to be attentive to. But It IS the . Apprentlceshlp. a relationship sketched out in a pioneering way in Abner Cohen's Two-Dimensional Man (1974). Bourdieu 1977. d b 'ng What they seem to earn. . in which meaning is reviewed in many contexts. as we might call it-the struggle for objectivism. rescu~s 1S p. Je~k\~S ~rgu:i~el~ ~ f~~ombi~~ion of frustration and example to e~po~e the in the mlOd ). gender. For it is 10 . if not to adopt. a model-for (Geertz 1973) action. in turn. particularly in such complex matters as the interpretation of meaning. Dougherty Anthropology is often conducted in circumstances 10 which to talk a g g " t not really to play it at all" h ' fro tration with the often This circumstance.:unattainable sense obJective. it is difficult to escape the tense relationship between politics and meaning. at the very least. Again. on the other. Gardin [1980]. Concentricities between the deployment of specific dec- orative motifs in domestic house plans paralleling their distribution in the territories of dominant and subaltern populations. everyday life For the ethnographer. . Kondo 1990). IS tr y. FIeldstUdy. but must also.esses (Delbos and Jorion 1984. and political commitment to the established privileges of always-already body-based. see also Leone [1988]. including those of class. may-if ethnographic parallels are any basis for judgment-be highly suggestive for deciphering the management of cultural capital and thus for the relationship between domestic and civic power (Herzfeld in Gardin and Peebles 1992. IS the km 0 cUOOlOg . but a means of access to the ways in which seemingly rigid codes provide protection and justification for a wide range of cultural invention. ingrouporiented and ingroup-adapted biases. The mod~1 0 socla Iza I 'ot corres o~d to the actual passivity and an abs~nce of intentlOnahty that d~. Willis 1977.' d f ' o r better era nee e 0 In many cases. object of that anthropological knowl g I t' nship in the two systems of meanmg.In Search of Meaningful Methods 106 FERNANDEZ 107 I HERZFELD dwellings. t' on by contrast implies both a contextualize the ethnographer s own le. not greater or lesser approximations to the truth.1984]. timony to or exeges1s 10 . on the other-is a condition of both infonnant exegesis ~nd rejection of o~ectl~~:ianthropological interpretation.ac :od ame is Pirsig 1974. . but we also insist that the ideal-as. on the one hand. n thin at all-in short. they must learn by seeing and domg an not y saYftl . socially organized sentiments of embarrassment (such as in southern Europe). d thereby ennc a pure . This is a replication of structures (Yogt 1965) that suggest investment of persistent common thematic meanings in the deployment of bodies in multiple ways through space. and thus as consequential-is no less real than what is conventionally assigned to the latter category. a set ~ prac flce d'lness artisanship d' oved 10 lavor 0 h a n . and the out-and·out self-assertion and consequent [1977.5' J k on 1989). d K d II 1982) may be made to feel. is a condition of social life. or com~arison of ~~~hrO:e:~~~~ol~geica\insight from a crass subjectivism-or fr~m JenklOs. it is thus epistemologicallYUns~un~~~~ :cell:~c: with it~ The metaphor of apprenticeship and the accessIOn 0 ad W 1991) can be f ' t d laming (Lave an enger . But it's also anchored in indexical relationships to significant parallelisms. . n~t c 1 een characterized as a process of although anthropolo~lcal r~tsearc~~:sd:~t~~ct~on is methodologically useful and socialization to anot er cu ure. . a source 0 no". .' ( R d-Danahay's [19951 critique of Bourdleu a thinly disgU1sed obJectlv1sm see ee r --. susbsistence conditions. There is a refle~lVe mterr~r"alO renticeship in which the body. the turn to practice-oriented theories has fatally undercut the old division of I. . the source of its claim to produce knowledge. and Watson [1994]. This struggle of social subjectivi.. 'I may appearhmd~~~culty of escaping social subjectivity and to be in some u~l~ate represents t e . 4). I thod with observed local practIces that. games.'im. is the operative model here. It . for example. This insight may also help archaeologists overcome the absence of explicit verbal information. and technology). for example~:ldhood socialization. Herzfeld (1991) has argued that architecture represents the clothing of the body politic and suggests in that architectonicmetaphorthe linkages and tensions between external display in its various fonns and interior familiarities.thIS tenslO:da: ~os( closely models the politics of it'. This tension in the sUbJe~t testimony an eve . since the forms of ~pprentlceshlP m r~. the messy suspension between a desire for sincere and accurate description. and its presence in field observation as a supposed inadequacy is. no nl su ress their questions with those of repressed craft apprentices who must not 0 y pp 'I of h 't t' find practlca ways and backchat.

one of the most cerebral description~.of the social role of metaphor: It predicatesa sign of known sel!hood on a necessarily inchoate other.~ldwork :. to local masters. thereby demonstrating hiS ~n:pt~ess~was a valuable object-lesson. a~. complete with a culture-hero. .ll P as ~ source of descriptive insight (see above). ~s a ~onsequence. even within one's own culture (hardly an unproblematic concept). Claims t~ cultur~1 knowledge through purely verbal operations may be a black-box-opening ?peratlOn of the most problematic kind. " SIS. he sought help in learning the bodily work ~mong Astunan vIllagers (Spain). It is in this reaching back or below language that we can most critically understand Vico's "poetic wisdom"-not. The possibilities through participation of reading the mostly unarticulated and also diverse and only partially lntenneshing world views of English villagers (1993 :Part l) are necessarily limited by social conventions of silence and introversion.'t s 0 f ac t ua I apprentIc: s . The more secure method of approximation IS to have learned by apprenticeship how to do (the essence be it noted ofpo. our infonnants' usual primary concern).Operience. We would do incalculable violence to the reality of social life if we insisted on perfect representations of social experience as the criterion of admissibility (see also Leavitt 1996). as Jackson (1989) has noted in his own attempts to grapple with the life worlds of African peasant women (although this account presupposes a somewhat problematic concept of empathy). but as an empirical recovery of the ways in which aImed at gathering data (which is not. in their very closedness. ··)Th·· lDe. .aning or embodied meaning (or "body·in-the mm IS IS sure Iy ro riate to social and cultural anthropology. Plore generically. Indeed. Lakoff 1987) have attempted . political hierarchy.enterIng Into the mIne with the various officios and following them to their vanous work sites. Such an approach works through what Femandez (1974) identified as the basis . the body registers all knowledge before it is processed as increasingly abstract "disembodied"representation. even though they may also. however. Johnson. responses to literary fonn may also be a valuable source of insight. But the way m which Herzfeld's verbalized enthusiasm for the project so dIsgusted the shepherd-who d:clared that while he had made a nOi'ma (a nonverbal g~st~re) Herzfeld had Inappropriately responded with speech.od ~e sought t~ reh~e the mmer's experience in its most corporeal tenns by . be revelatory of what those world views mean for social actors as they try to decipher each other's intentions and inner lives. . and gender.:ences.108 FERNANDEZ IIIERZFELD In Search DJ Meaningful Methods 109 knowledge from observation and hands on participation very much like th described by Jenkins (1994) for the day-to-day business of fieldwork. app~enticing ?imself as a mower learning to manage the scythe. Evans-Pritchard's (1940: 13) celebrated accountof Nuer-oSls. for much meaningfulness in social : : .or ~xampIe. to operationalize by bodily activity what others know how to do in the world (Fernandez 1992a). danced at times in the all-night cycle. and I . in effect. Coy (1989) .'e . r' h rlS 0 nVla Izmg t e latter. as a realizatIOn of those of the local infonnants: There is a real . we have found it useful not only for rapport b~t Simply for ethnographic learnmg to try and apprentice our bodily selves. k f t . When ":or~mg on the rehglOus movement of Bwiti. not necessarily as a conscious act of self-apprenticeship should not be confused with direct knowledge of the representation itself th~ugh critics can be viewed as registering a cultural model of the plausibility of fictional accounts of psychological inner states (Cohen 1994. but through relatively disinterested fonns of emulation (see Jackson 1989: 134-135). fi. especially when unless we are al~o prepared-as we must-to assume a degree of comm~nality i~ all hu~an expenence under conditions of strong physical engagement. We come to e~penence that engagement. clearly betrays the frustration of trying to fonnulate verbally unartlculated knowledge. pso embodied and implicit and is not cerebralized accordmg to academic •f. But it's important not to romantlclz~ on: s own experiences. Examination of the discourse about that experience (' ·.several different field situations. does Levi-Strauss's (1963) equally intellectually problemalIc comment on conscious models. For actors who may be thought to share a range of cultural experiences. we mIght say. th b" . that is. Fernandez acquired an agricUlt~r~1 plot and s?ught not only advice on agricultural practice from elders but partl~lpatlOn and dlr~ctlon of his physical efforts in his apprentice practice. dlsclplme necessary for participation in some of the easier dances in the complex dance cycle and. and as both phenomenologists (Merleau-Ponty 1962. Herzfeld 1997b). The absence of verbalization in the learning process IS not a complete barner to understanding. others have documented the advantages and lim. Csordas 1994) and the cognitive linguists (Lakoffand Johnson 1980. ' In. Herzfeld was les~ successful at persuading a sheep-thief to take him along on a r~ld. a method of attentiveness to bodily 'd meanmg. 1987. after all. Fernandezalso sought physical participation III agrIcultural activity. . of the. for by working alongside our infonnants we can partiallyand sometimes only inchoately-perceive the grounds of their SOCial experience across differences of culture.. as an origin tale. a. perhaps more useful even than the partIcIpatIOn In communalhouse-building work parties through which Herzfeld did nevertheless come to acquire a more physically engaged sense of community. In his The Corporeal Locus of Meaningful Activity .. common though the ground and moment may be. Among Fang villagers. e proper su ~ect-matter of poetIcs). All human knowledge ofthe world beyond immediate sensory experience is mediated by signifying processes located as Vico noted. popu I ar and cntlcal responses to fictional accounts) may at least furnish mfonnatJon about the culturally acceptable representation of such experiences This As Rapport (1993) remarks. h here to pinpoint what is. to explain (in the manner of an "etiological myth") the mundane practices of our workaday world. there are severe limitations. however.

these all imply what can be called progressive transfonnations of character. requitement. by their agonistic interaction. we have indicated-some paths to follow and techniques to use in the project of recovering the meaningfulness of expressive fonus. and power analysis.e ~~icular societies (see HDdge~ 1964. was most obvious in classlcal evo Ut1onJs~ we regard here as classic and t5 . 1975. for examp e. cepts and emotlOns 0 f universe or the range 0 con .. But this dramatic or narrative trope of social science understanding. h The feudally denve upper." or the fonnal analysts of gestur f xpressio n not easily captured by is left.' E Vico whose VIeWS I ' . they obtain continnation. Europe and the co-occurent idioms in language are plenllfully avallabl e . ./~~entary Forms of the Religious ritual space/social space teleology found m T . denial. but they raIse quest. . ful metllods narrated a developmenta ery to civlltzatton. t Wbl·ch social actors wish to give . rt .op. · .. or denouement of their affairs. . of particular analysts such . whlc IS per t h n we encounter it in the field. in its dramatic developments om "'The. emphasizing a mode of fieldwork that focuses on the mediations of corporeal experience and that locates what has been called "the mind" where Vico located it-in the body. elt er We may not necessarily buy in~o thoe :.1I0 FERNANDEZ / HERZFELD In Search of Meaningful Methods III to demonstrate in more fannal ways. to various states of more or less enduring resolution. and fools (Klapp 1972). This rather easily maps on social fonnattOn an can c·lety and vice versa is . I been a major element in anthroThe master trope of teleology haSh cc amh~pS one thing that makes it bard to 's own self~constroct1on. as found m 0 I h ve all sbown how the Douglas (1970). The rejection of such fieldwork as unempirical merely because we presently lack a full conceptual apparatus for describing such processes we regard as a refusal to engage descriptively with the imponderabilia of meaning. t· with no verbal eqUivalents. Paths to and Techniques of Meaningful Inquiry into the Agonies of Everyday Life A central form by which life is made meaningful is narrative emplotment-the characterization of human interaction and of the persons involved therein in dramatic tenns. denouements of a kind. f l' 1 arrangements on so teleological mapp10g 0 spa la t t" ons of the working out of the Durkheimian in the sense that they are ~eco~: . . . . . ther opposlllona ax Life (Durkeim 1976). . Spatial figuratlOn tbe side of the road" (Stewart 1996)that lSda tog be analyzed taxonomically. t" modehng of emo Ives expression. I j"fe However t ey 0 meaningfulness of socta 1 . Characters are ensnared in various mixtures of tragic Or comic expectations vis-a-vis tbemselves and others. and they are consequently involved in interactions that lead.a I ether marginal. d I mental narratives themselves. ven d' . . for clearly there are many areas 0 e f '-. <essive soclal formatlOns·1 ment but the tropological pbases. villains. b h hich these successive stages mIght be g t w logl·cal in the study of meaning rests ic synechdOChic. in the body's preobjective placement from which the self negotiates its life world. t· the United States produce t e railroad extension an~ w~stem town settle~~~h~nother side-or even a "space . . We can provide-indeed. or significant transfonnation. Bourdieu (1977) bas expressed dlssalls.on and the meaning of space. . r t" n with such a reduct100lS t society.01. ~at 0 omies in their context of use of meaning to be alert to and describe suc t axo~omies are plainly central to the ax These tropes ahnd d not exhaust either the semantic (Douglas 1966. It onomies Bourdieu (\973). ICO s a d 1 f a society's narrative sense of ItS pace notions of the tropes as part an parce 0 00 .. Spatial Take the matter of spallal commUnlCall for class. linguistic models. I expected to pass. 1992). r econstructing or deconstructmg the Nisbet 1965). 0ppoSlt1Ons (notably but not exc . t rfou nym . But to leave matters there would be a poor excuse for and a poor contribution to meaningful methodology. We are taking a militantly middle-ground position here (see Herzfeld 1997a). can be found in much less explicit fonn in models of human or social or cultural development. e and the like does not dispose of wbat "gustemes.h in space and time. and !rOOlC. In the process of the denouement. .on distinction between thIS SIde of the tracks.: 0 :f common sense in respect to both kes clear our obligation in the study taxonomies that conslltule the loca . \. Even the Imguls lC . Less dramatically.1 lO~ societal division and sOCletal progressIOn. He pomted up not on . and Needham (1962).. savag infonnative in an~ ISCUSSto ion over historical time of consciousness 10 ~uc­ del of progress!Ve transformat . metaphonc~ metosociocultural stage of deve. social interaction is seen as the struggle for desired or feared characterization as heroes. n of m e a m n g .ac 10 .. poI0 gy . d town lower town 1 n . In one fonn or another. Iy the key tropes of a gIven mo . linking the short-term performance of ritual time in Van Gennep's celebrated tripartite schema to the /ongue duree of historical event structures. It has been noted for Western literary culture tbat male autobiographies tend to be more deterministically teleological than female (Jelinek 1980).as VICD . lusively those 0 gen e . V' . '. given the new interest in "auto-ethnography"-the comparison of "their" constructions of their cultures with "ours" of our own and theirs (ReedDanahay 1997)---this kind ofgeneralization should be distinctly susceptible to crosscultural comparison and to sensitivity to changes across time and between contrasted contexts. status. Turner (1974) has taken this dramatic trope and made it into a phased method of social analysis. orthy of commen w e fr ognize as dlstmctIve or w. or at least its teleological implications of ultimate denouement or resolution. . This theatrical trope-basically the idea of social role playing-has long been persuasive as a plausible formulation for grasping the essentials of human interaction. .es: the key complementary expr d r) obtaining in a given physical layout of domestic and public space may f . [n starkest [onn. wareness 0 th e ropo . .

ptlve .:d kno'. sacred enclosures and the haram of the h yplcal ~odels of space (for and internal domestic ornament as ex . In addition. Uses of space r an especially useful perspective th: ~mme .Part I.es that may hav b h ~~ 1O. 1997) or the translation of social celebrations into a heightened sense of bodily wellbeing-and to relate such correspondences to larger patterns of social change or stability. t t' ug as s PUrity and -pro ec Iveness of oat" I . . tend to lose their ability to convey meaning with any real immediacy and. Clr more explicitly investigate the lack Of:'t bet comp ex~ty: In. Indeed.y ~ston achelard(l964). on the one hand. not ' I1lCnlllllg to idcolo. on the bodily experiences entailed in that life. meanmgs of space in a highly politicized context H " 0 e . vl. It IS the comparattve per· of how meaning organizes the percept~ spafce.tants of a once· . ISCUSSlOn of narratives I I refugees from Algeria with those of the rem~' ~mparmg the: me"-. e ec Ive y s Own how t r' .'11I01st'c In.t the shifting spectIve.n ID t e exotIc a~ss~n. p e soc.mportant respects). This is not to downplay the importance in anthropology of ethnography enriched by long-term observation and attention to the intimate details of everyday routines (thick description [Geertz 1973]). S~hama 1995) and on lexicon as well as deeper lying le'ss I' . b IS to ocal ntual. the organization of public space ~ot~ t~P 993. informants' inability to provide a clear gloss on some symbolic form may index a moment of important-if temporarily unfathomable--conceptual change.Ones of Jewish g shared domestic compound has ffi t' I Inhm Mushm mhab. appears to be'no le~ ~ us a t ?u~h." ynamlC ~nterplay between state ideologies le. such concerns highlight the inadequacy of Durkheim's reified dichotomy of sacred and profane. . . f erne. and the complexities of corporeal .5 0: about historic sites. an important segment of which addresses the "structure of the conjuncture" or transformative event in history (Braudel 1980. On transformation there exists a significant li(erature. tensIOns between external pressIOns 0 conflictin 'd I ' cu Iture) provides an accessible aren r . r' e . agency bers of supposedly simple societies Th e rea . t at permits a clearer understanding . moments of much more intense-because mOre palpably exceptional-meaning. Ion 0 space. g J eo ogles of national a lor eXamlOmg contest t" between the state (or religious communit and th .. a concept ultimately In more recent work on landsca e Ra 0%e.es experienced by the mem- 1956). ' ' D anger 's as applicable to the self. a IOns of meaning .:. :hese may be Ordinnry people 0 I 'I n y nltn )lite ' An openmg up to the poetics method ro 0 d d .~ .. Social life oscillates.tlOns lodged in the . 6 :WS :. p I ~s and hngUlst'c meanings. into bodily disease (Kleinman et al. everyday experiences. Their intensity arises from increased uncertainty and unresolved complexity-from an increase in ambiguity. 113 l I ! IIId practice complicate local and nonliterate social relations as much as they do the life of highly industrialized. ."ledge about Here we can explore embodiments in 11th . and thus also provide and the politics of everyday I. postures of Poetic Wisdom in Ethnographic Inquiry and Interpretation Revitalizing Moments The suggestiveness of the trope of path orientation in meaningful inquiry should not determine method altogether..c't oppos.societies.i:o . claims to open-endedness themselves contain the seeds of their own disproof: Even ambiguity can be reified IS dogma.deologies ofmeaning tions between "complex" .:~~:e embodi. once learned.th'. .~ urb~n-mdustnal.e~n presc~.~otal ~~~lal fa~ts (supSOCIeties). Such events may serve as moments of social revitalization or devitalization (Wallace posedly found in their pristine oppositional per~o t' as contexts in which we can examine DU own classlficatory desire for closure than of th supposedly necessary equivalences bet~~~n::t.or to the "untouchability" s as I e argument of Mary Do I ' .c. As Vieo showed. as we saw earher In the d' . grounded in a common physic I h' ere. not of referential stability. But we wish to emphasize the importance of particularly heightened moments of meaningfulness seeking to understand the special vitality that they embody. . Sahlins 1981).spallal posture or derived from Mauss. again. recent willingness of anthropologists t p p ~n e here I. I .atmg earher and ethnocentric distinc.In the relatively as pale or distorted sites for the fragm~n~:. geographically extensive social formations. recovered through narrative: Bahloul (I t~6) c e do~estlc sphere. between established and more-or· less inflexible routines that. This approach suggests a method of tracking the changes in accustomed (normative) .S seen .Ia~e interests. This is an organic trope (Nisbet 1965}----of vitality in social life-that focuses our attention.:se exp!. and the calibration of ideal-~Y ~ lell It Impinges 011 their example. or the identification of particular events as diagnostic of systemic changes (Moore 1987). Meaningfulness derives from the special reinforcement or abatement of the intensity or transformation of social relations. Here the methodological challenge is to recognize through its external signs or narrative realization the embOdiment by the social subject of social experience-thetranslation of social suffering. such settings. .een t e result of our ureaucracy(Herzfeld 1992' Malkk' 1995' B IOna or mteroatlonal factor in urban ghettoizatio~ dyna~ic ' 't ~rnettlJ 997) . readiness for inteaction in the world the h b't (~h P?ral. literally so}----than that intended b toe lC~ ofSpoce-and more grounded tie architectonics to political and th . we can and actual practice or use thereb als b . personal (Lawrence and Low 1995) This k h expenence In the world are explored meaning ofspace much mor~ respo~~~e t~ methods for the study of the (indeed. Conversely. exp 'c.nd "s"mYI .s.t meanings lodged' h' and socIal.t.0 0. In IstOry. but ' .et. for example. on the other hand. L 112 FERNANDEZ I HERZFELD I~ In Search of Meaningful Methods formalism and has posited a much more corn lex tern .

f om the genealogical arboretum . o. ' seemingly without limit. A social poetics must consider method itself under this rubric. has assumed the status of a kind of elementary form of reference (Fernandez 1973) in seeking to understand in a methodological way the motivations of metaphoric predication itself. for examp e. .. This trope. I ". The study of meaning is bound to have to consider the nature of surprise-tu/ness. not only with respect to the conjunction of . a~dl~~~~~e . tropes of thickness. It is as a means by which to grasp those displacementsfrorn unacceptable experiences to more satisfactory experiences that metaphoric predications are often used (Fernandez 1974..<irlC~ 1996. All these "figurations of social thought.h their cultural history and their role m 'odologist). their quality space. m term 1 f tee (Fernandez 1998) and bird imagery ... of the tropes of every ay. It can be understood as a trope anchored in the primordial gestalts of human experience as this takes place in our gravitationally anchored three-dimensioned world (see Lakoff and Johnson 1980). d n C eluding but not pnvl egm . decides to act in the world (Osgood et al. account for differences in the quality of life from moment to moment and phase to phase. When we say that a ~e 0 ing qua semantics and assessing its significance. or of depth. but also with respect to the associated corporeal commonplaces of social interaction and the confinnation or challenges to these presuppositions. I · such as. som::I~et~~nto analyses of the figuration of methodological l A poetics approac 1S no 'd ~.from family trees and roots and ot~er spe~:~::~a:i_totemicidentificationofsocial . at any rate.In Search of Meaningful Methods 114 FERNANDEZ / HERZFELD Jl5 structures of behavior as a consequence of such events. . . And the notion of quality space has long been given by philosophers a special place as a fundamental tool of understanding of how the mind places its experiences in corporeal terms (Quine 1981). g that of the methargurn . Relevanthere. and their followers) must recognIze Its own ental m th d is meaningful. as a form of understanding that uses a choice of tropes to organize its clarifying activity in the world. t::e 'fonn of anthropomorphisms of . are ana yses 0 r . differences that constitute an important part of life's changing mean~ ingfulness. 1957). These surprise-full moments. as in "thick description" (Geertz 1973). The questIon What l~t I. and it is their presence and evocative associations that a poetic approach takes into particular account. I 1 Conclusion . whether in this master t~ope (see also e . are fundamental in the arguments of methodology and methodologists. Iives. The presence 0 Ir s area~ tro e of power in other cultures. Of equal interest are other tropes of utility in the rhetoric of methodological argument. ent but as we have sal .~o.~." like the tropes of game. We find the tree trope m a Wide vanety . . and text (Geertz 1983).m~kc to b~ a human?" corresponds · fatalities of their lives.t \'ke to be a tree or a bird or a bat to the widespread attem~t to [igI"re Ollt ~::~. "raw" and "cooked" data (the latter being somewhat -.~. d cursive note by concluding with a Why end on an exclusively refleXIve an re d I I~cated almost anywhere tho .u b that is a human animal-are understanding of what it is to be human-to e.has deep classical antecedents and IS a wldesp t d' PmytholOgical form in Robert I th se so vividly presen e m . drama.rvdav understanding. Even culinary tropes organize assessments of data: "thin" or "meaty" ethnography. indeed. to tree diagrams in logic and hngUls. 'b'l't' s of cultural self(N agel 1974). The Nichomachean Ethics . dexes the allemptto d0 so. 993) B t the poetic posS! I I le various kinds (Guthrte 1 . Our emphasis here is less on the structures involved than on the corporeal resonances arising from the reinforcement or weakening of 0 Id associations or the emergence of new ones. consequently.e between Popper and Foucault (if not ~t the sohPslStce~ :~ figuration and metaphor. . and that seem to emphasize path dependency without regard to the associational structures involved. -. l 1 Qualitative Inquiry and the Quality Space of Life The trope of quality space for the temporal complexities of corporeal-mental life has long been recognized as a basic time-space conversion and is used in social sciences to plot the vectors ofjudgment by which the body in the mind weighs and asssesses its experiences and. thIS IS not a confuSIOn 0 mean -1 I . ~erely" metaphoric~1. as in deep or penetrating analysis and of movement (as in "social movements"). we are as Max Black (1962) noted. It is hardly posSlb e Gardner's (1983) New Gumea ethno~ p fthe qualities and perimeters of their 0 to understand the New Guinea pe?pl~ s sen~~ enerally without taking into account ... tive references that many men and women the bush or 10 the hand.0 og~ mes beyond these authors poetics of methodology itself? An honest m. 1 ems of bo.. 1997). . are pnme figur~ I lexilies the fulfillment and the may grasp as they reach to und~rsta2 t ~e . th~n birds in many cultures. of social and intellectual circumstances. Often thIS figUring out . ra hie film Dead Birds. 1986). ' ld be hard to claim that such assessments SUSpect). the prevailing and changing systems of structure. This interest in surprising or arrested moments or events in culture in respect of embodied experience contrasts with other approaches to meaning that engage primarily with routine. and "processed facnualltY:r~i ~ta~~u although the device of labeling them as · oulS 'de ' l the cntena 0 lay . . .lc~ an ked out extensively in The Golden groups with trees-a set of assoc13t~o~~dwors fi urations of the human condition : Bough (Frazer 1890). Id d interested in studying it to look It is the mark of a man experienced m the wor an for no more precision than the subject matter a~~~:~tle.. nngs "1 . under scrutiny the associated commonplaces .

POSSI e e~ause we re~~gnize that such freedom is always relative.elabora~lOns of pro~edures of discovery are hardly sui generis and cannot be satIsfied with the purSUIt of reputation and esteem among methodologists alone (Taylor 1985). accessible to analysis as a tangible aspect of the indeterminacy as ~ell as the cultural values and conflicts that charaete~iz~d her chosen ethnographic situation.ures and processes here. certainly recognized a contingent sense in which truth claims could be established. k' longer confine themselves to reputedly isolated commUnities but wor III ::reaucracies. It is also worth noting that fieldwork is itself always. the immediate. in effect. confinning the accuracy of its perceptions beneath the surfaces of propriety and cultural defensiveness.CIt y on this methodological issue in the context of analyzing large-scale internatIOnal organizations. ~ethod In ethnographIc practice and in ethnographic interpretation and ~xplanatlOn" shares somethmg fundamental with the human condition that constitutes ItS o~n ~ubJect Method does not escape the human condition. and subject to. it can be viewed as a coping res~onse to t1H~ Inchoate in the s~n. is to be situated somewhere and somewhen 10 particular.al system. We ay "as fi 'bl"b s ree as. a sampling that local actors can always accuse of unrepresentativeness-thereby. highly productive. Precisely because we view methodology as a fonn of socially embedded practice. but to ?e inside practice is also the only way to come to know. in which the partiality of her sample was Itself made explicit and consequently. they have b~en forced to confront the inevitable partiality of their research and to turn it to practIcal advantage. The practical issue-and methodology is about (i~deed. to be judged an adequate approach to the human condition. He and others have repeatedly shown how highly structured theories of theorists and metho~ologists are themse!ves artifactual effects of a lack of practical mastery (Bo~rdleu 1977) and a neglIgence of the social conditions that make social science pOSSIble (lenkms 1994). methodology in anthropology. and international organizations. should itself be a COpl~g p. a synecdochical activity. Inchoateness may not seem a promising target for methodology. and it should not mystIf~ Itself wIth the hope of such platonic transcendence. whose path we have followed so assiduously here. one of the few anthropologists to have commente exp '.hon~st meth. Th~t . . It must take account of the relation between the commitment to formal procedure. and the facts of the contmgency and open·endedncss of much of life as it is lived-an inchoateness that moti. in any event. . as open to reflexive analyses as any other practice. In this respect. and enunciations of value" (1995:42). there are cultures m WhICh It IS eIther unrecognized or explicitly dismissed as untenable. Moreover. and nowhere more clearly tban in Zabusky's methodological statement. it's worth emphasizing that methods are enmeshed in what one of us has calied "the politics of significance. indetermmate. Vico. In other words. . as we have ~Iready noted.od must .se that it seeks explicit procedures and practices as free as possible from the conditIOns of the life that motivates it. is that any methodology. This point has been made wIth some mSIstence by Bourdieu (for example. d I" I As StaciaZabusky. On the other hand. we have tried to identify some explicit proced. which is a version of the logIcal error.va~es much .dis~inction is itself contingent and culturally specIfic. ethnographically: the soc. remarks of her work in the European Space Agency: "To be mSIde practice . our "middle ground" posture requires us to acknowledge . The vIew we take bere. Thus. context.vitably reflect on the material and social metaphors on whIch ItS practIces and ItS Ideas about its practices are based (Herzfeld 1987). It is often the condition of the very possibility of method. And see the discussion by Jeffrey Johnson [this volume] of Zabusky [1995].which is characteristic of the practice of any methodology.ine. however. If not always '~ the. . any methodology that claims to be supracultural-that is con~ext-mdependent-is an exercise in circularity and self-deception. human sciences generally. should embrace the actuality ofthe contmgen~y that beset~ tha: condition and should recognize the often necessarily unpredIctable ways m WhICh ~eople cope with such contingencies. (We should also note that the intensity and intimacy of even the most unstructured ethnogr~phic work represent an implicitly statistical assessment as much as any numerical sampling device. evasions." in a sense expanded to a greater degree of provIslOnalIty than that intended by Popper). Methodology in anthropology. and it certainly corresponds in its refusal of arbitrary closure to such natural-science understandings as complementarity (Bohr) and uncertainty (Heisenberg). Methodology I~ Itself motivated by contingency. not denial of thIS nch actualIty. 1990:Book I). choosing a site where the limits of that value in practice would become palpable precisely because of the intennediary position of that site was melhodologicaliy appropriate and. for his assertion of the necessary /xIlh 1'- These concerns notwithstandi~g. ~mpl~lcal denial ("falsification. since the system is replicated and reproduced through agents everyday actions. Given that her focus was on cooperation. Its . the importance of establisbing some norms and the virtual impossibility of 'l' Z'Ing them against the corrosion of time and doubt. of misplaced concreteness." in which objections to non-Cartesian methods are often linked with surprising fixity of purpose to nationalistic and other political ideologies (Herzfeld 1997c).ractlce-~ way of cop109 with human copings with the inchoate.L L L 1I6 FERNANDEZ / HERZFELD In Search of Meaningful Methods 1/7 mea~ing qua social impo~ance. A method i~ a km. is) practice-is simply that an accurate and empirically sound representatIOn of social life requires some means of recognizing the "fuzzy" and "indeterminate" social matrix out of which emerge graspable and explicit meanings. to ignore it would be decidedly non empirical. which is. An . perhaps..cl of practice in. science laboratories.figuration in thought and practice (Femandez J 974). For her own work: she chose a "crossroads" department. to say that a method is meaningful presupposes a social and cultural.. Now that ethnographers b stall .

. Probing.IS t ere. Herzfeld h.p:~:~~~::ci:f:~~~y ::e ~ot. our goal in this essay has been to suggest w~ys. ~ut s. unconvmcmg y. the social dramas. Through participant observation. . reduced to the demarcation of separate functions that he\d~oft~n .n:~r~t:~~~:~:. cultural identity. In a sense.." but we might now more reasonably ask. an quite vana~\Y a~ .r .about .118 FERNANDEZ I HERZFELD In Seorch of Meaningful Methods 119 imperfection of human knowledge... 3. and so forth. through imagined reconstruction and other techniques. t of ursuin that a roach.'::. b : processes and ' .~~s.) Tracing the more obvious narrative emplotments of social life.urchase on this slippery shape-shifting (Fernandez 1974) by we can o~ y ge i deed as ects-not icons. f [. is largely devoted to the wo. w . 6. For some time. making it possible to establish systematic features of interpretive differences by class. '. Herzfeld was once challenged at a conference. Anthropologists have indeed long been engaged m replacmg the C\:~~shioned segmentation of the discipline into sep~rate fields of stu~y. ultimately grounded though it was in theological preoccupations. Parmentier 1994) the symbolic aspect is recognized only as the most explicitly meaning-conveying. m ~ db rmativity in both reductive SOCial sCIence and equally X ~m~~ber~n~e~i~~:~e:a~~ ~ove :~:~~i~c:~~~e~:~r:~iCeins~t:~ions. [t· t' rty as a model 0 the socla eloquent demonstration of the madequacy 0 re eren la I f . with careful attention paid to the social relationships .rk OfS~~iO~O~lsts. (We add parenthetically that. trope to the loose use 0 ognizable There is in fact a techmcal reason or pre ernng .. It is our view that the distinction in anthropology between the ~wo ap~roa. Anchoring these images and tropes in bodi Iy experience and its Vicissitudes. differential qualities of life. 3.ymbolism of a given population as a finite subset of ItS cUltu~a\ equipment. ~~:~i~~ a SOCial and cultural poetics as indicated. 2.' [ h ' I [rm of precIsIOn m other areas 0 t e trope are usually those who demand a more nurnera e 0 These stages in the research process permit evaluation of what once were called the "expressive aspects" of mundane activity. in which the ethnographer both internalizes and attempts to render explicit formal norms of social practices and interactions of the more routine kinds. quality space. "When is culture not expressive?" In a semiotic anthropology of partially Peirceanderivation (see. industrial. h d· I the most paSSIOnate a ergy 0 "metaphor" and it is doubly IroniC that those w 0 ISP ay . The key offender turned out to be "trope"! lroOlcally. relating this "poetic function" (the "play of meaning") to political norms and practices. indexes.'1 early '970s although casting a wide net. bU~~ :~e.. . and so on.. of aB socm interaction. Singer 1984. It would e [.mbolis': (see also Eco 1976). Concomitantly.. I we would add-to typify modem. Fernandez 1965. the figurations. and in most cases he asked authors to switch to a first-person mode. .s onenbeen struc · Y d l . for the images and organizing tropes that define specific norms as practical projects in the world and on which they may depend for their realization. Herzfeld 1983. f d· g responslblhty for ana YSls an effects of passive-voice constructIOns as a means 0 eva m data. Recording commentaries on conformity and deviance in these norms synchronically and diacronically. 5. for example. and symbols. deVI'ces and m measunng Y ' scholars arc interested in narrative narrative . da ) hich has appeare smce t e approach in anthropology The journal Poetics (Amster m. . does not depart radically from positivistic models of falsifiability except in its anticipation of Foucault 's explorations of the relationship between discourse and power-and this is the crucial. This both reSISts IconICIly. Elicitation and ethnographic learning on a model of apprenticeship. our OW? se us t~fi~at. enabling condition for the emergence of a methodologically explicit social and cultural poetics such as we have sketched here. then. and hierarchi zed rights of choice and decision. obtaining between commentators and their subjects and anthropologists and informants. t action and expenence.~~S~':b:~ci. anthropologists spoke of "expressive culture. iri contrast to the relational meanings of indexicality and the seemingly natural meanings of iconicity. we now propose the following stages in the research process as a plausible path for the attainment of poetic wisdom: 1. NOTES l. II t . For s~ch a p~rspecllve reqUlrl~~ is ense with similar reifications of the meanmgful m. differing proportions on these three aspects of meanmg. gender.hiS use ~f ~nn:~:s~~~l~ ~oo~~ . ~~~I:e p$)'Chologi~ts and psychologists' only occasionally do anthropologists pu . age.: is also an epistemological consequence. h .'. Western society. As editor of American Erlmalogist (1994-98).on. n 1S . with their oppositional or agonistic workings-out and denouements and their sometimes discomfitting entailment in encompassing tropes of revitaliza· tion. new recording devices such as video permit the exposure of field materials to multiple observation.. defined as the maintenance and contestation of systems of privilege. it recognizes the shiftiness of . d scale is a shift from primarily indexical ("we are aB identllIeS on~h\~:':::c ("we are all alike") claims. Ever attentive to the provisionality of our acts and deeds. in addition to the well~tried resources of narrative analysis. the corporeal locus from which embodiment they derive their meaning and exploring the system of entailments which constitutes the cultural logic present in social interaction and the qualitative anchor of meaningfulness.~'o~~~~~. frames the fundamental ynamlc 0 de. the shift from local social relations to national and dis<:ipline! k b Ihe 2. Ily recwords. but of finding the semiotic implications. in which the symb~hc IClated") to pnman for contesting this view or interpreting it for speCIfic """vides the groun s H e Id 1997a) Thus we can say that the emphasis falls r~F~(=e . d advantages a d . no ~gUrin: out th:. one tO~ :h~Che~:i~:~~:~~. 4.

v. ntation as Power. is predominantly engaged in semiotic and ' literary criticism of verbal texts. New York: Routledge. Farnell [1994. intellectual activity could not be separated analytically from bodily sensation. Edward M. Towar a Colby. . an ex. Gaston. The Poetics of Space.. Current Anthropology Borofs y. 1984. of Modernism. on 0 Bloch. Social Construction 0/ the Past: Bond Genrge . Bachelard. . cd. . Black. It: Waveland. Ardener. Benjamin N. The Witch-Hunt. Harmondsworth. N h G ece Princeton: Prmceton Cowan. B tamma/zba rC I ec ur . Distinction: A SoclOl rftlque OJ e u Harvard University Press. Marcus. L. ed. lames. Play./n Other Words: Essays Towards a Reflexive Sociology. 1965. suzann~ rAes °h~t' t ai FJpression. Bailey. eds. an . JoeIle.rth J dgement o/Taste. How to Do Things with Words. and George E. London and New York: Tavistock Publications. Amencan t no '. At the other extreme. 1980.story Chicago' University ofChica~o Press. P' 1977 Outline of a Theory of Practice. Maunce.S. e . Ithaca: Comell University Press. Introduction. 1994. Symbolism in Complex SOCIety. and Sahlins. 6. The Architecture ofMemory: A Jewish-Muslim Household in Colonial Algeria. F. Obeyesekere. Clifford. American Anthropologist 96:397--415. 38:255-282. Cambridge: Cambridge University Press. . The Kabyle. 1974. d r.n ~p . 'G 1994. House. 199 . Dance and the Body PolitIC In ort ern re . University Press. Bamett. W . Compara lye ~. 1994. d F d d David Kronenfeld. Two-Dimensional Alan: An Essay on t le . . Indifference. E. ~~nst~p . Bourdieu. 1973. G. h' R tations [(12):118-146. Monographs. d M ta hor in p t 1994 The Anatomy of Architecture: Ontology an e P sUer. '" . even thoug~ their actions may suggest it-and few anthropologists have bothered. New York: Orion Press. 5. 1997. In Te:"t. J. On Ethnographic Aut onty. 1990. aoe DefinitIOns ~ and Angela Giiliam. American Ethnologist 7: 1-26.L In Search of Meaningful Methods 121 120 FERNANDEZ I HERZFELD 1962. and Story. c. . England: Pengu lO • • d N Y rk' . 1995] and Farnell and Graham in thisvolurne). Bourdleu. ed. 1937-1962. " C" . Austin. The Wil/iam James Lectures. IrVieD was right to suggest that all verbal expression was grounded in the immediacy of bodily experience.. Pierre. " B k I . lames. J. for a rather different approach. Ithaca: various technical methods their consequences in the maintenance of social order and in the formation of public opinion. was supported by Award #9307421 of the National Science Foundation. . Bahloul. Academic Press. A Plea for Excuses. In Social Anthropology and Language.. Pierre. Richard. 1971. Other Scribes: Symbolic Anthro~ology in the Boon James. A. d N w York. Both journals thus share a basic concern with language. Ed war d M . Michael N. Pp. arguing that vision itself was "immured in the body" and that. Englaod: Polity. p. 1983. Stuart Plattner.d Bourdieu. Bauman. Reason. 1983. Represe A 1982 Other tribes.~ep Ethnography Descriptive. '. Broner. ' although their epistemological orientations are radically divergent. Berkeley: University of Cahforma Press. 4 Abraham Lincoln as Authentic Reproduclion: A Critique of PostBroner. ts Cambndge and New • t" Study 0' Cultures Histories. Prospect Heights. I A thr ology of Power and Cohen Abner.. Cohen. U iversity of California Press. Edwin.. In Philosophy and Linguistics. What the linkages mean for social actors is quite another matter-Austin's phrase "trailing clouds of etymology" nicely captures the problem: The fact that we know that "having an accident" suggests the impersonal forces implied by the Latin verb accidit tells us absolutely nothing about the degree to which the social actors are themselves conscious of any such linkage. Re IglOns. 79-101 London: MacmilIan. comell University Press.. 1984 [1977]. b . Colin Lyas. The Triumph ofMorality. DiaIOguCdN~rrat~~. 1990. b'd . n . Masada. In Rules and Meanings.d ~:S:i~:~~~e~c: of Ethnographv. L.A. The UN Security Council. Cambndge an ew 0 .' Y rk' Cambridge University Press. and Reality. the evidence for links of this sort The problem is philological and psychological as well as social and cultural. Eh logist8'442--450. 1984. and Phyllis Gorfain. 198: . er e ey. 4. This research. epresen . conducted in 1993-95. On H. Self Consciousness: An Alternative Anthropology of Identl!. C lrure' The Poetics and Politics 6 n~lng u Clifford. . Poetics Today (Tel Aviv). ernan ez. Pierre. Cambridge: Harvard University Press. Bailey. F. 1964. L d n a .. for this reason. lane K. Mary Douglas. Ithaca: Corn ell University Press. Convergence of Cognitive and Symbolic Anthropology. London and New York: Routledge. W. Verbal Art as Performance. . Debate and Compromise. and Genocide in Rwanda. 1996. Anthony P. Vico'sown method was strictly etymological: He traced political abstractions back to terms denoting physical images. 1973. or. lerre. this grants the body a primacy emphasized in a recent cognitive linguistics work (Johnson [1987] and see also. Cam n ge. h' 1980 Exposing the Moral Self in Montenegro: The Use of Natura hm . American Ethnological Society. d 1975 Political Language and Oratory in Traditional Soclely. The Tactical Uses ofPassion: An Essay on Power. or are adequately equipped to seek. Models and Metaphors: Studies in Language and Philosophy.' REFERENCES Ardener. Chicago: University of Chicago Press. Femand. Max. . Cam n ge. Braudel.. I an e . Cambridge University Press. 1981. 10. Bailey. The Politics ofInnovation.. G. James W. ok' R bert 1997 Cook Lono. Cultural Anthropology /2:551-578. Bourdieu. also only occasionally a locus for anthropological work. F. Austin. Oxford: Blackwell. eds. 1971. o .

. 1982 . Rinehart and Winston. . Mary. Durkheim. Fcrnandez. Purity and Danger: An Analysis o/Concepts ofPollution and Taboo. h logist 67· 902-927. eds. Directions in Cognitive Anthropology. London: Macmillan. ' . Pp. Man (n. Femandez. Paul. Janet W. Symbolic Anthropology: A Reader in the Study o/Symbols and Meanings. 1992b. 1989. 2d ed. Eric. Brenda M. \993. Paris: Editions de la Maison des Sciences de I'Homme. urrenl andez. Emile. 1890. 1. W 1977b Fang Architectonics. Pp. Unpublished ms. . . Austin: University of Texas Press. Alan. E.) 29:929-974. No. C Ant r o p O ' M. Leach. Path-ologies of Iberianist Thought. 1985. P I . Crain. American Anthropologist 99:306-320. Farnell. Dundes. New York: Holt. Fcmandez lames W. On Trees of~no.. D. D. . James W. Genevieve. Steven. James . ' Crick. The Significance of the Course Events Take in Segmentary Systems. Eco. Collection ethnofogie de la France.. Architectonic 1nquiry: ReView ArtlC e. James. Umberto. . and David Schneider. 1832. American Ethnologist 20: 179-184. Berkeley: University of Cahforma Press. 1990. Translating Truths: Nationalism. 1998. . Dougherty. Gable. Friedrich. 1 Crashes in the Kraal: The eoses 0 Im . Femandez. eds. . In Directions in Sociolinguistics: The Ethnography o/Communication. Malcolm.: e ge 0 e for the Moral Imagination. . f Metaphor in Expressive Culture. e . New York: Columbia University Press. In The SOClQl LIfe o/Trees. 1994. d Th Bull Who 992 What I Learned from The Parrot s Egg an e a. 1992. The Nuer: A Description of the Modes of Livelihood and Political Institlltions 0/ a Ni/otic People. and Paul lorion. Ethno-graphics and the Moving Body. H if Stanley Diamond. American Ethnologist 13:309-324. and Black and White Histories at C 0 Ioma Ethnologtst /9:791-805. ~ . Pp. EI-Haj. Naples: Fibreno. by Mockery. Paul. James. the Practice of Archaeology.s. p. -. 'J W fcrnandez. lames W. American Conjecture. Ga' ey. Spielerisch un Planvoll: Zur Theorie der Tropen in der W 1994. La lransmissiondes savoirs.982 Switi' An Ethnography ofthe Religious Imagination in AfrIca. 1976. lames W. ·1 d P 209-216. 0/ Semiotics. The La b an SCrlp. Csordas. The Strategy of Turkish Boys' Verbal Dueling Rhymes. lames W. A Theory · t Canberra: Australian d Orid Williams. E. Janet D. Symbolic Consensus 10 a a fcrnandez. 1986 Persuasions and PerJormances. and Anna Lawson. 130-160. Dolgin. Johannes. and B. The ISSlon 0 fern I 15·11'1-145. . Bloomington: Indiana University Press. 1977. New Literary History 8:459~83. Nadia Abu. 1d f S If and Other in Culture: Models Rival and Maurice Bloch. London and Boston: Routledge & Kegan Paul. eds. S f l' e Binding and Turn Taking in Being with the fcmandez. Kemnitzer. The Elementary Forms o/the Religious Life. . Urbana: University of Illinois Press. 19 . Oxford: Clarendon Press. 1972. How Classification Works: Nelson Goodman among the Social Sciences. Christine Other. &1-110. 1989. 1996. 1975. Mary. La mimica degli antichi investigata ne/ gestire napo/erano. 1974. 1976.. and the Re-making of Past and Present in Contemporary Jerusalem. 1984.' Prioceton: princeton UOlverslty Press. Dresch. Douglas. London: Alien and Unwin. philadelphia: University of Pennsylvanm ress. Albany:. fernand~z. Thomas 1. The Golden Boug: . d W eping Poetics.122 FERNANDEZ / HERZFELD In Search of Meaningful Methods 123 Coy. . S. Delbos. . 1995. and Song in Kaluli Institute of Abong Sound and SentIment. Brenda M.' D niel and J N Peck Explorations in Anthropology and Literary S~udie~. Andrea. AmerIcan W 1965. Time and the Other: How Anthropology Makes Its Object. Sir. Culture In ClIlture/Contexture: friedrich. Mary M. Evans-Pritchard. an d· . Explorations in Language and Meaning: Towards a Semantic! Anthropology. Ritual. Quito: Corporaci6n Editora Nacional: Ediciones Abya~Yala. lr s. An Avian and Aphrodisian Reading of the Odyssey. C I . 1997. w " 1 fcmandez. Historische Anthropologie 2(1): \-19. 1966. . . 1 'oal and Torres Strait Islander Stu les. P . Do You See What I Mean?: Plains Indian Sign Talk and the Embodiment of Action. . 1976. London: Berg. Gainesville: UOlverslty of Flonda Press. F I 92 On the Uses of RelatiVism: ac s. B. Paul. American Ethnologist. Embodiment and Experience: The Existential Ground a/Culture and Self Cambridge: Cambridge University Press. London: Routledge & Kegan PauL Douglas.. Culture in Poetry and th e P oetry toy I . 1986. c and y J . Ozkok. h A Study in Comparative Religion. In Dialectical Anthropology: Ess~s l~ onor o. Fabian. E.. 1994. In press.f'. cd. Laura cds. Breoda M . Working Papers in the Tradtllonal Arts. Advice to the Perplexed Ethnographer 215-226. I S ·otica 89(1-3)· fcmandez. Michael. eml .: AnthropO ogy· poet in Motion: Being Moved by Amusement. ExpreSSIon. Jerry W. New York: Columbia University Press. . \ Farnell. lames George. Douglas. Bloomingtoo: Indiana UOlverslty Press. 10 A an ge of Soundbites . State University of New York Press. ' F ng Reformative Cult. F mandez ames. a entlOe a . Edinburgh: Edinburgh University Press. I i . 1996. lames. '. Gumperz and Dell Hymes.~a:~i~'i~9t::a~sturi~nCountryside.. I 'Ph'ladel hia' 1nstitute for the Study of Human ssues. The Play o''Tropes in u lUre. I WiHiamsburg. . 1940. John 1. Implicit Meanings: Essays in Anthropology. Richard Handler. lames W. 1983. 37-38. memoria popular y proceso politico en la sierra ecuatoriana. Mary. Aothropologie. e. . Apprenticeship: From Theory to Alethod and Back Again. frazer. de Jorio. New York: John Wiley.

New York: Basic Books. 1980. J p t Urbana: University of llhnOls Press. 1993. Roman Jakobson. 1960. ' I 'alion and The Bodily Basis ofMeaning. 1980. The Wrightsman Lectures. Don.. Villains. Berkeley: Kleinman. Fredr . assisted by Brent Vine. Michael. h 'd e' M I T Press. In Collected Essays oJ ed. 1996. U' sity of Chicago Press. Historical Knowing. Bloomington: Jelinek. Hill Jonathan. In Local Knowledge: Further Essays in Interpretive Anthropology.( _ . 1987. Chicago: University of Chicago Press.. . A. Iowa CIty: n I . HerzfeJd. e. Veena Das. ' '. Arthur M. Chicago: University of Chicago Press. Language. Metap ors . '. HIstory. 997 Social Suffering. 1995. ' . Ithaca: Comell University Press. Cultural Intimacy: Social Poetics in the Nation·State. Anthony. Cambridge: Cambridge University Press. Mark. mas A. Michael. f h We Live By Chicago: Umverslty 0 lakofT. 1964. and Margaret Lock. . Chicago: nlver . 20-35. No. Jean-Claude. Michael. 1985.. Pigs. 9. and lane Richar ~on. Semiotica 46:151-166. 1972.. Harris. Cambridge: Cambridge University Press. od n MargaretT.. Giddens. Gibbs. . 1988. Michael. Guthrie. Herzfeld. Rethinking Hlstorya~ .<1. 4/:107-1 38 . Leon J. Ken a. Paul Secord. Michael. 1974. 1997a. Timothy. Dead Birds. Cambridge: Cambridge University Press. A Place in History: Social and Monumental Time in a Cretan Town. . Oxford: Basil BlackweJl. New York: Random House. Craftmg Se ves. 358-372. 1987. h . E. Gardin.-' :'. Engendering Archaeology: Women and Prehistory. eds. The Social Production of Indifference: Exploring the Symbolic Roots of Western Bureaucracy. Looking Both Ways: The Ethnographer in the Text. eds. Jackso le . Representations in Archaeolo Bloomington: Indiana University Press. Clifford. 1994. publications. Tho lakobson. Herzfeld. and Fools. O. New York: Oxford University Press. 350-377. ' t P ess I uiry. Faces in the Clouds: A New Theory ofReligion. Clifford. . Cam n g. Wars and Witches: The Riddles of Culture. Clifford Geertz. . ep . Roman. n's Dress Japanese Workplace. 1997b. d 1I t 982 RefleXIVity an le w . Cows. Mlchae .. Joan. In Style In Language. a Quantitative Analysl~.) Jcnkins. . 1991. Herzfeld. Princeton: Princeton University Press. onathan. Goldstein. Austin: University of Texas Press. Ver ad s~gnhen Rudy eds. Bloomington: Indiana Umversl y r . . linguisticS and Poetics. p. Pp. 1992. . Pp. Early Anthropo. f W d 1952 [1940] Three Centuries 0 ome Kroeber. 1990. Anthropology Through the Lookjng~Glass: Critical Ethnography in the Margins oj Europe.s. The Interpretation a/Cultures: Selected Essays. 1990. Chicago: University of Chicago Press. f Eve day Life. 1994.1 Records 5(2): 111-154. . the Sixteenth and Seventeenth CenturIes. Adam. New York: Basic Books. Women's AutobfOgrap y. Reflectl Diego: Aegis. MichaeJ. magm ' . Gero. Krystyna Pomorska an Minneapolis: University of Minnesota PreSS. a le M n. L. 1983. Stewart. The Body in the ~ind: 29'433-455. Portrait ofa Greek Imagination: An EthnographiC Biography of Andreas l\'enedakis. Andrea De l Anthropology 7:371-390. . San Klapp.ogy m H ge. U· 'ty of Pennsylvama Press. 1992 on the a s . E. 1973. . Heroes. Handelman. 1976.E in Criticism. an e . 'lc 198 t The political Unconscious: arra Ive . U iversity of Iowa Press. .' . ? Reason.y . and Understanding. Oxford: Berg. 1983. Karp. power. 1992. dId ntily' Ethnogenesis in the Americas. Marvin. Berkeley and Fashions. 1979. N t' as a Socially Symbolic Act. 1984. . ssays Indiana University Press. Blurred Genres: The Refiguration of Social Thought.c. d ' P 249-273. d F' Id ork In Explaining V· I Isua 'ons ofthe American Character. Hill. England: Polity Press. Michael. Verbal Time. P Los Angeles: University of California Press. Peebles. Social Archaeology. Pp. The Poetics of Mind: Figurative Thought. New York: Routledge. rId t 'ty in a University of California Press. Beverly Hills. b 1 S' Jakobso n. Raymond W. The Constitution ofSociety: Outline ofthe TheoryofStructuration. Kendon. . Herzfeld. . . I ameSon. Ithaca: Cornel! University Press. Geertz. H. 1985. and Martha B. The Poetics of Manhood: Contest and Identity in a Cretan Mountain Village. orio: The First Ethnographer of Gesture. nq Roman. C. . I 1997c. Verbal Art. 1991. Herzfeld. philadelphia: ntverSI _ . I-Ierzfeld. .. and Discourses 0) en I Kondo Dorinne K. Cambridge: Cambridge University Press. ed. 1980. Gardner. . R d' of Empiricism and Ethnographic . 1 . } 492~ . 1983. f S· nificance Social Analysis t~i-. Jean-Claude. Cambridge.In Search of Meaningful Methods 124 FERNANDEZ / HERZFELD 125 Gardin. and Christopher S. and Mark Johnson. Archaeological Constructs: An Aspect of Theoretic Archaeology. • . An~hropolog Re rinted in The Nature of Culture. CA: Sage Human Behavior. Fieldwork and the PerceptIOn 0 ry r"- JOhns~n. George. Models and Mirrors: Towards an Anthropology of Public Events. . Man (n. Robert. I ' power Gender. Ivan. The Sense ofOrder: A Study in the Psychology ofDecorative Art. Sebeok.--told. Gombrich. Geertz. hael 1989 Paths Toward a Clearmg. dM th'/ndigenousSouthAmericanPerspectlves . New York: McGraw-Hill Films. . . Chicago Press. Princeton: Princeton University Press. " . Anthropology and the Po ItlCS 0 Ig . Estette. I' . ed.

Edmund Ronald.ology. ed. . c Books.er mg . Built Environment and Spatial Form. 1979.V. I . 1997. 1994. LeVine. Quine W. Richard 1. Chicago: University of Chicago Press. Thomas. Honolulu: Bishop useum I jn the PaclifilC .1 . and Dangerous Things: What Categories Reveal about· the Mind. Situated Learning: Legitimate Peripheral Participation. Theory of Practice. . Edinburgh: m urg University Press. Belie/. 1996. John. 1973. London: Nicholas Brealey. Ed' b h Rapport. Merleau-Ponty. Levi-Strauss.. Prmceton. Purity and Exile: Violence. Roy A.1975. Leach.e . Osgoo • . . Review of Anthropology 19:453-505. 1994. and I:. 1962. Chicago: University of Chicago Press.Consciousn~ssan ange: ym Perspective. a . Rodney. Monograp. American Ethnologist 14:727-736. d' . Anthropological Society of Washington Series. ao e~mg m . Phenomenology ofPerception.Bour dle~ps . 109 on. Rohert A. C am rt ge. Approaches to SemIOtics. . 1974. Behavior. 1995.. The Journol of Philosophy 34:435-450. 1992. HI I" Studies in Semiotic Anthropology.JamesL. 1981. 1995. 1976. Anthropological Quorterly 69:220-230. Beijing: Foreign Languages Press. 1988. t New York: Oxford University Press. Dancing with the Devil: Society and Cultural Poetics in Mexican-American South Texas. . Bedouin of Cyrenaica.. Meeker. . 1990. Auto/ethnography: Rewriting the Self and the SOCial. 1982. Histories. Claude Livi-Strauss. oom\'Il"". Liisa H. Silvana. . Denise. . 1962. 1996. Cambridge: Cambridge University Press. Meaning and Feeling in the Anthropology of Emotions.. Fire. Malkki. Gerlad L. . Michael. Cum be Reborn: An An ean t 1tI0 r . New York: John Wlley. Tannenbaum. 1997. H l' .. ~. Modern Masters. New York: Morrow. 1965. am. and Setha M. Low 1995. . Marx. Sally Falk. Nagel. Lawrence. f S . Moore. The Kabyle and the French: occidentaGlisffi in . Memory.273-284. 1972. Nigel. Mind Your Manners: Managing Culture Clash in the European Single Market. the Wind' On the Rhetoric of 1dentity Parmentier Richard J.. Cambridge: Cambridge University Press. Architecture. London: Industrial Society. . Language an eam.rRhetoric philadelphia: Robert. Mind Your Manners: ManagIng Business Cultures in Europe. Structure and Sentiment: A Test Case in Social Anthropology. In Occidentallsm: mages OJ 61-84.:. Zen and the Art of Motorcycle Maintenance: An lnqlllry mto Values. C arner. Culture.' . Kar!.pertence. P' t n University Press.n. 1979. Literatllreand Violence in North Arabia. El' Ch S bolicAnthropologym vo utlOnary d Peacock. Cultl/re and Communication: The Logic by which Symbols are Connected: An Introduction to the Use of Structuralisl Analysis in Social Anthropology. Signs in Society: . Theories and Things. Michael E. Miceli.A. Institute for the Study of Human Issues. M . Claude. Hooked Rugs in Newfoundlan~: The Representation 0 OCla ~ Structure in Design. Maurice. I H d' . Madison: University of Wisconsin Press. A. . d M . Robcrts. Mark P. George . d' . Cambridge: Cambridge University Press. Rccd-Danahay Deborah. N 49 The Hague and World and Its Semiotic Organization. and National Cosmology among Hutll Refugees in Tanzania. John. hs No 22 London and 'd d 1982 Semantic Anthropology. 1 Pocius. Mole. Language. 1978 [1852]. od Percy H. Developm . University of Chicago Press. g' The Origins of the BUilt Prcziosi.r M aning' Urbana: University of llhnols Press. DavI . New York: Viking' Press. Rodney. Press. John. SOCJQ ang an . . George. e. Needham. Anthro.nti.'ew York: Academic Press. Explaining the Present: Theoretical Dilemmas in Processual Ethnography. . politically Speaking: Cross·Cultural Stu les oJ . 1970. Chicago: Aldine. and Personality. PClCrs. The Eighteenth Brumaire of Louis Bonaparte. nnce 0 '.r the West lames . 1991. 1993.~ J S . . Deborah E. Harvard University Press. and Eticnne Wenger. 1st ed. American Ethnologist 23:514-539. Mole. 1995. 1987. Africa 37:261-282.). Learning in Doing. t 'Indiana University Press. 1987. . r' Ri hmond CA" North At antic " Rappaport. Leach. Ecology. 1979. DC: Smithsonian Institution Press. o. ()beYcsekere. 1994. . d E I g ophy of History Chicago: Rappaport. r_' Oxford' Basil Blackwell. . Joanne.. Jean. R~bert M. Edmund Ronald. Mole. London: Nicholas Brealey. b'd . UCI. John. Jose Eduardo. Chicago: University of Chicago Press. I Ch e d History' Aspects of the Western Theory of Robert A. . 1 . anan . . Leavitt. 1979. New York: Basic Books. t.S. 1963. Lave. Meaning. 1957. Diverse World-Views in an Englisll Village. J . The Recovery of Meaning: Historical Archaeology in the Eastern United States. . International Social Science Journal 49:373-385. 1981. r Captain Cook' European Afythmakl1lg G ath 1992 The ApolteoSIS OJ ' . an d Re IglOn. Oxford: Oxford University Press. . Paris: Mouton. new ed. 1974. E rys 1967 Some Structural Aspects of the Feud among the Came . Women. Journal of American Folklore 92. Measurement OJ e .In Search of Meaningful Methods 126 FERNANDEZ IIIERZFELD 127 Lakorf. m . The d Charles Egerton. en . Palermo: Sellerio. Structural Anthropology. .Donald. Mind Your Manners: CIl/ture Clash in the European Single Market. Rced~Danahay. What It Is Like to Be a Bat. . In nome del segno: Inlrodllzione aI/a semiotica della cu/tura. Leone. Washington. . Limon. New York: Humanities Press. Pirsig.. Oxford: Berg. 1o.

It is the articulation of this "process by which we came to know It" that reflects Ibc elements of research design. which we came to know it. skills.JEFFREY C. a systematic argument enjoys a star~ spangled legitimacy. JOHNSON 8. the observations produced by how a study was designed are fundamental to the proper assessment empirical evidence: "We always want to reject evidence if it can be explained by Ibc deSign of the research or by a large number of small. In Agar's statement above.Some things. we get the Impression that a credible argument should be systematic and based on a process dlat informs us about how researchers came to know what they know. Michael A. In such worlds. not as the only possible representation that our field can offer. a mode that ensures we can represent our representations in credible ways. experiences. We need a ''''ay to argue what we know based on the process by . methods of investigation." ':¥fJ. like perceptual errors. his or her motivations. who is to be helieved or lrusled? Are cilia themselves. For Stinchcombe (1987:23). independently of how they were conceived and collected. unorganized causes. That's what I seek.IJ Four Research Design and Research Strategies I Wc need a powerful mode of argumentation. for many it's crucial to know something about the author. Agar (1996:13) Introduction In a complex world of competing argumenls. that hinder our observation may be beyond our or 131 . proper evidence for making a case? Although some may he swayed by the elegance of a trcll-written essay. and so on before passing judgment on the conclusions. but as an essential lever to try and move the world.

. 22. 1997. Tiny Cbar1es. On Anthropological Knowledge: Three Essays. '. . Bergm Vieo Glam aUIS a. Philosophy ofthe Human Sciences. Launching Europ~: An. I. Cambridge: Cambridge University Press. ed. Linda R. 1976.. Fables of Power. 1983. Giorgio. Princeton: Princeton University Press. ed. New York: Knopf. ed.. In The Historic Turn in the Human Sciences. Roadsfrom Past to Future. Vogt. . 1978. Salmond. A. pnng le . Atlantic Highlands.\(Hhieal Realities: Structure in the Early History ofthe Sandwich Islands Kingdom. England: Saxon House. S . Shryock. Singer. London. . Domination and the Arts of Resistance: Hidden Transcripts. 1977. . Kathleen. . Paul E. 65 Structural and Conceptual Replication in Zinacantan Culture. . Sartre. Roger. Jean~Paul.. Merriam. Andrew. Willis. .Ethnography of European CooperatIon In Space Science. Turner. 1. Rethinking Symbolism. MA: G. McDonald. Stacia E. Schism and Continuity in an AfrIcan Society.S. Anne. For Example. Burst & Co. New York: Cambridge University Press. . Man's Glassy Essence: L(plorations in Semiotic Anthropology. V ICtOL . 1985. A Space on the Side of the Road: Cultural Poetics in an "Other" America. Stark. b . Ant hony F. Representations 4:108-127. · S d 1994 Memo . Power and the State: A Method. Mary Margaret. Marshall. 1985. Sewell. Thomas G. Communicative Interaction.' . ed . Taylor. Ru ble . The Yearbook of Symbolic Anthropology. Vico and Marx: Affinities and Contrasts. 1997. an . Tagliacozlo. Farnborough. I. Paris: Editions de la Maison des Sciences de I'Homme. Dan. ASAO Special Publications. Landscape and kfemof)'. David Parkin. Toronto: University of Toronto Press. lthaca: Camell University' Press. 1990. Schwimmer. . 65-87. e . Struever. Roman lakobson 's Science ofLanguage. . Princeton: Princeton Umverslty Press. Search for a Method. VictOr. . London and New York: Academic Press. Steedly. Santa Watson. James C. lthaca: orne Turner. Three Temporalities: Towards an Eventful Sociology. Chicago: University of Chicago Press. New York: Knop( Schamn. Sperber.. Learning to Labour: How Working Class Kids Get Working Class Jobs. NJ: Humamues. Evon. Pp. 245-280. Theoretical Landscapes: On Cross-Cultural Conceptions of Knowledge. C 1\ . 1995. Hanging Without a Rope: Narrative Experience in Colonial and Postcolonial Karoland. The Hague: Peter de Rldder Press.A. Trans. Does Complementar. WilIiam Hamilton. 1972 (1958). .128 FERNANDEZ IIIERZFELD In Search of Meaningful Methods /29 Sahlins. Highlands. 1959. Stewart. At antIC Turner. 1996. New Haven: Yale University Press.' Historv and Opposition under State SOCialism.!. . Nationalism and the Genealogical Imagination: Oral History and Textual Authority in Tribal Jordan. 1981. 1983. Waugh. Ann Arbor: University of Michigan Press. Salzman. t 1984 [1744] The New Science ofGiambattista Vico. Sperber. 1995.' 'd M ax. trans. Montreal: McGill-Quccn's University Press. Anthropologist 58:264-281. Vo\. Opposition Exist? American Anthro_ pologist 80: 53-70. 1967 The Forest of Symbols: Aspects of Ndembu Ritual. lthaca: Cornell Umverslty Press. In Semantic Anthropology. ed. Pp. Cambridge: Cambridge University Press. 1975.' Fe: School of American Research Press. Bloomington: Indiana University Press. 1. Eric. American Anthropologist 58:264-281. Dan. . Simon. 1990. . American Z 19 . Zabusky. C ' . fi Id Webster's New International Dictionary of the English Language. Lanham. Berkeley: University of California Press.. . Sahlins.. NJ: Humanities Press. and C. . . No. H Fisch. Historical Metaphors and . No. Wallace. F. 1996. Sanjek. Fieldnotes: The Makings ofAlllhropology. C. How "Natives" Think: About Captain Cook. 1996. Nanc)'. 1993. . Marshall. University Press. MD: Rowman and Littlefield. Ann Arbor: University of Michigan Press. I. Charles. 1978. 1982. 1963. 1956 Revitalization Movements: Some Theoretical ConSiderations. Ithaca: Carnell Umverslty Press. Scott. 1995. 2d ed. Philip Carl. 1. Monographs. . Princeton: Princeton University Press. . 1974 Dramas Fields and Metaphors: Symbolic Action in Human SocIety. and with an introduction by Hazel I3ames. VictOr.. 1984. . Milton B.

others may reqUIre a deSCriptIOn ofthe research problem and site but require less detatl about the. validity.s . Members of such competing groups-such as large-scale commercial producers. The value of empirical evidence can anI be properly evaluated by understanding the details of how the research was cond/t c d According to Pelto and Pelto (1978:291): "Research design involves combi " . sampling. measurement. and recording are at least partly wlth10 our control. Design.S. Environmental organizations launched campaigns to ban nets in tuna fishing because dolphins are often caught incidentally in that fishery. however. that their positions are credible. Fishers petitioned the federal government to fund a study of pinger effectiveness. ideological. such as tragedies or acts of nature that dISrupt fieldwork. One result is that infectious childhood diseases that killed thousands of young people a century ago are today only a memory in industrialized countries. The laundry-list component is important. chance factors often lead to great discoveries h' or unexpected find1Ogs. policy emerges from interactions between groups of differing political." This statement Illustrates at least two Important elements of research design. mng the essentta so investIgatIOn mto an effectIve problem-solving sequence. r:liability. The outcry for a ban on nets in tuna fishing is a famous recent example. Some conservationist groups claimed the study was biased in that ::. It IOvolves detads ahout gett10g into and out of the field situation travel arrangements. and economic backgrounds. while luck plays a role in research planning for I ' suc Iuc k · IS not WI·th·m th e rea m of research design (Kirk and Miller 1986). herrneneutic ~nd postrnodem approaches make little explicit reference to ethnographic desigd Issues..L L 132 JOHNSON Research Design and Research Strategies 133 . the fishing industry proposed the voluntary use of "pingers"an underwater acoustic device~to keep porpoises from their nets.'crlptlOn of the proposed project paymg attention to the research design logic of sCIence (for example. etc. ccc~1 v~lves the methodological and analytical details that contribute to the credibility. ' On a practIcal level. ~n idealized pl~n gives guidelines for linking theory to the methods of data collectIOn and analySIs that yield either valid or "defensible" results I "defensible" in addition to "valid/' which I normally use. method~ of data collection and analysis. Still. This case led to a systematic test of a technology that might ameliorate the problem. All funding agencies expect a well-orgaOlzed outltne of the proposed project-one that meets the design expectations of peer reviewers and agency personnel. A good understanding of the research problem : and the res~arch SIte allows us to plan for some contingencies. hypotheses. and so on. They have evidence. The study used the classic control/treatment design in which catch rates for a set of nets with pingers were compared to catch rates for set of nets without pingers. Wildlife conservationists petitioned the V. to make readers' aw~s: tha~ I am broadening the. on the other hand.lJmer. or plausibility of any study. traditional application of research design to include t~e V3f1ety of research strategies found in anthropology today. contributed to Florida's totally banning fishing nets~even though no marine mammals were threatened by the use of nets in Florida waters. Thus. waters).ical facts . There has been similar concern over the incidental catch of harbor porpoises by net fishers in New England (Schneider 1996). the control net caught 10 porpoises while the treatment net caught none. but well-written examples from ethnography may provide "moral evide " to deaf ~ith current social problems. moving people (including politicians) in that !!. A distinction needs to be made between what's sometimes called the laundry-list compon~nt of researc? and researc~ design. good research design is essential in the competition for researc~ grants and contra~ts. First.. and there was no finn evidence in the literature about it. a plan cannot deal wIth the unantIcIpated or unknown realities of research. believability. Some thi~g~. and real estate developers-believe strongly in their positions. Without some unbiased means for assessing the evidence. getting proper government permissions.orh t e objectives of the study to be realized.. making contacts at the field I. In the first experiment. Nevertheless. In this chapter I ~oncenlrate on "elements of design related to the production of valid results or a beltevable ethno<. There is much variation in what funding agencies and foundall. In f~ct. the truth is only be a matter of who has the most political dout.~an't (Seidman 1994: 134). federal government in 1991 to declare harbor porpoises a threatened species. see also Plattner [In press J).~ control. Interpretive.S. research design involves an a priori plan or strategy for all phases of th researc~ (such as data collection and analysis) including. Media campaigns in the V. By definition. showing pictures of dolphins being caught in nets (generally not in V.S. I " . environmental groups. The Need for Design Evidence for the power of research design is all around us. commodity producers.. social. The invention of the simple control/treatment design of clinical trials allowed researchers in this century to evaluate competing therapies and to select the ones that worked best. for some researchers. a well-articulated project de"gn helps to promote the effective conduct of research" whether one starts from a positivist or humanist perspective (Ellen 1984:158). '1 f' . but there is 'no ' research desIgn cIJ:'stal ba~l. The effectiveness of the device. Porter 1995).: 'te arranging for living accommodations. often anecdotal. was in question. The lessons learned from controlled experimentation are applied today to the policy arena where groups are in conflict over resources or because of social inequalities (Johnson and Pollnac 1989. Second. One agency may require a detailed de. th: production of the final product (like an ethnography). like site selection." graphic account. Thus the plan of research is a statement that concentrates on the components that mu t b 'd ' Se present m or er ' . In response.ons expect regard109 r:search design. ':lidity.

the treatment net caught only 1 porpoise while the control net caught 32. If research design gets relatively little attention from anthropologists. Pelto and Pelto's (1978) book on research methodology in cultural anthropology. but this series of studies illustrates how the elements of research design help muster evidence in light of competing beliefs and philosophies.000 fishing nets. the effectiveness of pingers compared to not using pingers). validity. will not have very much meaning and that the characterization of a selected number of cases must necessarily be the material with which you operate. Boas. (from Orans 1996: 128) This response is important for at least two reasons. A complete elimination of the subjective use of the investigator IS of course quite impossible in amatter of this kind but undoubtedly you will try to overcome this so far as that is all possible. Ironically. Sechrest 1973. despite his concern for scientific . k' methods of fieldwork and data collection (Ellen 1984. infonnant accuracy. This time. What should we make of this apparent dearth of specific treatments of research design in cultural anthropology? I don't think we should make too much of it because the important elements of research design-reliability. The logic of the research design contributed to the production of credible results. . to do a quantitative comparison of the slmllanty of attitudes among the adolescent girls in her study. Bernard (1994) has elaborated in more detail on issues of design.. She wrote of her doubts about the comparability of c. Thus. the treatment nets were placed in areas known not to have large numbers of porpoises. In each successive study. 127). Some environmental groups were still concerned that evidence with more statistical power was needed.as I 0 concerned with the aims ofscience and with methodological rigor. I believe she thought her mentor. were more a demonstration of the value of ethnographic writing-his "unusual literary sense" (Lowie 1937:231)-rather than of methodological details of proper ethnographic fieldwork (Ellen 1984). Mead wrote to Boas WIth her concerns about possible violations of scientific principles in the data she had collected to that point. Again the evidence was impressive: The treatment nets caught 2 porpoises (I was thought to be deaf). but his treatment is necessarily limited. other social scientists have written volumes about it. What is most surprising is Boas's response to Mead. As Orans (1996) describes it. Statistical work will require the tearing out of its natural settj~g. Yet. would Ultimately hinder the comparability of data from different ethnographic sources (see Moran [1995] for a recent discussion of this issue and see Ember and Ember. and operationalization of theoretical concepts-have been present in the writings of cultural anthropologists even before Boas. thIS volume). but only the pingers on treatment nets would activate once placed in the water.Research Design and Research Strategies 134 JOHN SON 135 ~ . A good example of this tension between the stated early concerns for the methods of science and the actual use of such methods in ethnography comes from correspondence between Boas and his student Margaret Mead durmg h~r first fieldwork in Samoa. without that settin~ may . the Boasian idea of data leading to the construction of theory). Boas 1920). She had concerns-and. fishers were blind as to which nets were control and which were treatment-a classic double-blind experimental design. The issue is still under debate. Although the power of experimental design is evident. or even the need. but Mead was attempting to exert her authority without necessarily following the research procedures advocated by Boas and others. Malmows I. objectivity. So another study was conducted placing experimental treatment and control nets in the same proximity. this concern for context and meanmg over methodological rigor. particularly for those in search of theoretical foundations (that is. First. it demonstrates the differences between the stated scientific objectives of ethnographic work as advocated by Boas and the actual practice of ethnographic research. meth . He writes: Iam very decidedly of the opinion that astatistical treatment of such intricate behavior as the one that you are studying. however. ~~ l l I I .as~s a~d about her ability. 1973. HIS earlIest ~:ntributions. investigators tried to control for as many extraneous variables as possible so that the hypothesized effect could be assessed (that is.have ~o meamng whatever. Malinowski. and Research Design in the Scientific Tradition Boas and most of his students advocated a natural science logic in the collection of ethnographic materials and a true concern for the collection of reliable data that could lead to the production of valid theory. Orans says: "What she wants is pennission to present data simply as "illustrative material' for the representativeness of which one will simply have to take her word" (p. Some early exceptions include Brim and Spain's (1974) book on hypothesis-testing designs. some particular aspects of be havi or which. od Boas was more explicit about his methods of data analysis than about his . Spindler and Goldschmidt 1973). Both control and treatment nets were outfitted with pingers. Lobbying efforts by fishers yielded more funds for a larger. would feel similarly-as to whether a valid comparison of this type could be made given the selection process for her sample The constraints of field research may lead one to stray from the Idealized prescriptions of a research design. while the control nets caught 25. There appears to be a perception that a systematic treatment of the data will have to be abando~ed to preserve context and meaning. concern for its application in anthropology-particularly cultural anthropology-has been limited. which has several chapters that address issues in research design (LeVine. '''j Boas. and Naroll and Cohen's (1973) A Handbook of Method in Cultural Anthropology. given his task of describing the range of methods available to anthropologists. more comprehensive study involving more than 10.

~~n. for it is my conviction that so long as prime theoretJc~1 c~nc7rns ~~ ~iS study of culture are an attempt to record and understand the native s view t d ure and the obiective and historical realities of culture. of fieldwork. Further. systematic'l and. !be current lac:.~:r~ :e overriding d f culture like those of the natura sciences an . by the late nineteenth century. sInce . Boas's final sentence in his response to Mead illustrates that even at this early stage the issue of the subjectivity of ethnographic research was of concern. this debate has its parallel in sociology where schools such as ethnomethodology and symbolic' interactionism developed in response to the largely quantitative macro-level focus of the discipline.L 136 Research Design and Research Strategies lOHNSON /37 Second. There was a faith. 64-Q5) 0 LeVine (1973) and others (Johnson 1990) make the point that th~. It is a strange paradox in the development of field methods that the scientific study of other cultures has been built upon such a foundation. Thus. A good exa:nple of WIth respect to hyp Rh wgll' (1967) published in the Spindlers s senes this is a book by Thomas ys. ". village" mentality of some and the fact that careers were made by dis. . I find thiS s:an~: comfortable. the tdea of researc ers pu I n . The idea that quantification detracts from context and meaning in the ethnographic endeavor--evident even in the time of Boas-and a failure to· understand that systematic methods-whether quantitative or qualitative-help') minimize the subjectivity of the investigator have impeded the development of well.te.t. I have taken the pOSlu. h 11 tt' gall t elf egos limited in their ability to do so. whereas the British tradition was more nmatter ofsubjective experience. then methods for flheldhs u Y cult . (pp. In ~ng h a d proposed hypotheses are inappropriate to the cultural setting under stu y. h.S. less we II -k nown cu Itures has certainly In t Ite theOries replication efforts (Johnson 1990). psychologists. traditions. one liner that reflected the attitude of the tImes. f . Boas's concern for contextual meaning over the statistical analysis of data was prophetic.panse was Deem for methodological rigor was probably at Its .er~s among many sociologists. . th blem Laboratory and survey researchers have some flexibility to change e pro . museums and government agencies. There was a wide belief among British anthropologists that fieldwork could not be taught to new recruits. that of rep IIca IOn. 'I h designs because 0 I S has often limited the attractiveness of more lonna researc I' . However. of a field methods course and one concernmg a · Wagley s teac 109 b ' concerning d nt at Berkeley asking for advice before going to th~ field. fi d' th t the fieldwork often dictate the need to change the problem focus or. an ethnographercou n t · t' another important goal of science. personal experience for the anthropologist.:or cl (LeVine 1973:184). Kroe e~ s paduate stu ea terse. anthropology was already established in universities. I I IDthropo og 'h d for more systematic methods and design detatl. Freeman 1983.d Bernard (1994) relate stories about Kroeber's recommendattons Agar (1980) teaching and conduct of ethnographic research. while early British and V. ..S.t le:~:~~i: the unlikely there can be a rigorous. 61) of formal training in methods and research design in anthropol~gy. In the American-> tradition texts provided what was regarded as an objective body of data. f . .. Thus. . across the British and V. ~ven m t . . t tarns on field methods. there are surprisingly few details about actual methods anthropologists used in the field. delineated research strategies in anthropology. " . particularly in the British tradition. however. This belief supports to somehow salvage the research. h . the huge investment in time and resources hmld. A further irony is that the one thing that might have lessened potential subjectivity biases-the use of standardized methods-was rejected outright because meaning might be compromised. Orans 1996). As Vrry (1984) sees it: In Britain the claims that anthropology not only studied a distinctive body of data but also that it possessed a sophisticated methodology to collect these data.:~~a~n~ca~e the essence of cultural anthropology to students or gene~al . Id Ok in one basket" may have limited the a priori fonnulation of problems In fie . Rightly or wrongly.ea~ers.. but could only be experienced by individuals in the field. . t e nee There is much anecdotal evidence for a belief. partlcU ar Y pl~yed down t e ne~thesis_testin approaches (LeVine 1973). and the complexity of a holistic approach.d since I have indicated here that research in culture involve~ ~ gre~t d~t. In the stortes: one regarding the .formal pr:s:~~. focus and study populattons In hght 0 emerging p r o . Mead's position on these various elements of research design provided fuel for the continuing discussions about the validity of her original findings (Brim and Spain 1974. and economist. ~. I' k The "my natIves or my realistically be expected to replicate someone e se s wor . u~lt~. the preeminence of contextualization has been a consistent issue in ethnographic research and has often clouded issues in'. Williams writes: · nl someone wholly involved and fully immersed in fieldwork can I b~ll~e.cove~lnhgbnewd · or describing exotic . This was less necessary in America where. . . that awareness of the potential biases associated with the subjectivity of the investigator could be dealt with in some reasonable way. research design. beyond a few first principles and illustrative anecdotes. These micro-level approaches are attempts to get at a better understanding of meaning in everyday life (Cook 1994). when co t f search methods and design in the lIterature I y many treatmen sore .on t . there is little evidence that they considered appropriate design issues when they actually did the research.S. was an important factor in the establishment of anthropology as a discipline. hi ntofapartofte uman will have to reflect the end purpose ofmakmg a woe acCOU experience. e ".peak In late 1960s. in a trial-by-fire method of training for ethnographers. (p. ( It's tempting to explain this as the consequence of the intensely personal nature. . blems but field workers are . 'fic way The rea Itles 0 commitment to studying spectfic problems In aspect . anthropologists advocated the scientific method in ethnographic research.na:~e of fieldwork in tenns of its requisite huge investments in time and geograp Ica °fC~tS. But in spite of claims to scientific methodology.

Types of anthropo og OR DIMINISHING THREATS: "! .. authority. s are what they appear to be or ~~volve a general search for nHdity is concerned with wheth~.. systematic strategies tend to involve the search for explanations of phenomena and the pursuit of theoretical foundations. In searching for such foundations. A discussion of the basic arguments as related to epistemology.lDITY CONCERN FOR: THREATS TO BELIEVABILITY MEANS F UTERARY MEANS FOR DIMINISHING THREATS: METHOOOLOGICAL bes more generallY to make pro roaches are used to develoP hypothes~e:~~od topics. whether quantitative or qualitative. RATIONALES ~~C~'O. where exploratory reh can be !..he primary focuS ea Gl.9tl~ this enterprise..A:WIL~S.C:.OCAl. it is the purpose of research design to ward off as many threats to validity as possible. systematic work is objectivist. I think. As a result. These approaches are by no means mutually exclusive in approaching a research problem (see section on Research Design in Systematic Research. For any given research problem. Exploratory res grounded theory ideas of eanmgfu1 theory development of more m n components. say." Figure I is a taxonomic characterization of the different types of research strategies found in contemporary cultural anthropology. It is not.'N. This is similar to : : Exploratory: Exploratory ~p~ jnte retation of less welt-un descriptive research leads to for circumscription. below). The figure distinguishes' between strategies within the realm of interpretive studies and those using systematic strategies that have more of the elements of science. and the like are beyond the scope this chapter (see Schweizer in this volume). the assessment of any theory involves research designs more heavily concerned with the means-the research process. an~ Slra: s (l967). The historical tension between interpretive and scientific approaches in anthropology has given way to an outright rejection by some anthropologists of science and its logic of design. The light line connecting the two categories indicates their complementarityand interrelatedness in that a design may include both within an overall research design framework. descnpllon.UE MEDIATED FINDINGS OR CREATED FINDINGS CONCERN FOR: THREATS TO VAl. 1 ical research stra e Figure I.F~O. research design and its logic have been associated with science and an underlying. reality. be misleading. rather than simply the way the study was written or argued-since the validity of study results depends on the scientific soundness of the research design. INTERPRETIVE at POS~ L:'~ IN SEARCH OF: UNDERSTANDING MORAL TALES l. causality or prediction. explanation can I 1he truth (Cook and campbell I . each entailing a specific design strategy. Ive testing elementS R earch designs 10 roaches generallY IOVO loratory research. This leads to designs that involve concern for a higher degree of methodological and analytical detail. and control over possible sources of error leading to a valid assessment of a given theory. To say that the research. there is a need for objectivity. In this line of thinking. This is a highly simplified representation. d Research Strategies Research Design an /38 JOHNSON ANTHROPOLOGICAL RESEARCH 139 Contemporary Design Issues in Cultural Anthropology SYSTEMATIC There is an ongoing debate in cultural anthropology concerning science and its role in contemporary research. t gies and their features. Explanatory: Explanat~ry app been proposed in the literature or d their primary purpose IS to e I the best approximation to this mode are detennined a priori a~. postmodernism would. Suffice to say that traditionally. Many examples of research in anthropology fall within the two extremes of the continuum. Epistemoiogically. EPISTEMOLOGY: SUBJECTIVIST VAl.RN. but the tenn "design" itself smacks of the very formalism that is being rejected. that interpretive approaches lack some fonn of research plan. Its practitioners are ultimately interested in research findings that approximate an external truth. In its most extreme fonn. design logic of science has been replaced by something that is recognizable as the' research design logic of. objectivity. replication.~~I~URCESOFERROR REPLICATION UNDATIONS THEORETICAL FO PURPOSE: VAUD ASSESSMENT pURPOSE: ccoUNT OR STORY A BELIEVABLE A EPISTEMOLOGY: OBJECTIVIST THE TRUTH FINDINGS APPROX. es .as er an and measures. here ed that have been infonn by e~p l'minate threats to vahd~ty.-------IN RULING INTERPRETIVE ANTHROPOLOGY EXPLORATORYDESCRIPTIVE EXPLANATORY More SEARCouHT~~ AL HYPOTHESES " ~~~~T. Under the systematic distinction are the two primary categories of exploratory and explanatory approaches. of theory that may already ha~e of a given design or just one of ma Y .E.N~D~ET. A more appropriate teno that would encompass the diversity currently found in cultural anthropology might be "research strategy.. the researcher is a field-workeras-writer. beliefin objectivity and explanation.R~D~E~SI~G.

. . such on the ends of research-the ethnographic or literary strategies.'" of cooperation in Europea. are more radical in their sweeping rejection of scientific method and design logic (see Schwandt 1994). The core activity is still social description of the other.~~. still adhered to Some logical empiricist methodology and maintained a degree of beliefin ethnographic authority. recent int~rpretlve para?~g(~\are 1991'S5}-or in critical theory. Epls em. for conveyi~g the u~derstan Img g~m\ acter to exploratory and As stated. There is. (p. the exploration rather than umdJrectlOnal exammatlon (p. mo In contrast and to the field-wor k er-as-wn·ter. However. ratlOnales I' . some interpretive work is more similar in nature to the exploratory or descriptive strategies found under the systematic side of Figure 1 than to some of the more radical forays into. with research on cooperation by "experimental" PSYC~Ologrtlsts.these interpretive paradigms depiction of the truth.. b k' (1995) ethnographic study the form of mutual postmodem studies.. symbolic anthropology. An important implication here is that scholars who follow this line of inquiry are searching for local rationales rather than nomothetic theory or universal foundations and may be more interested in conveying a moral tale of some ty~e rather than a usually literary.L 140 JOHNSON Research Design and Research Strategies 141 Interpretive strategies.i~~:n~a~~~s~~~t~:g:~:~~:r:~~~~: \ic anthrarop~Oy~~~:b~~rred :c~~~t· ~~ ~:~~ry~a~S~re Oft~ fOCUSproe~ There is a fundamental belief that the intersubjective. 242) Seidman 1994). question. the work ofZabusky 1995). to name a few. interpretive anthropology. there is less focus on the means ~ are value medIated of data collection and analysis as found in the systematic research. and changed in any given context often dery objective study and explanation. such as early interpretive anthropology. Rabinow (1986) observes: At first glance lames Clifford's work. The other for Clifford is the anthropological representation of the other. . cross-cultural settmg.he . Fo o:~gk:~d of ethnodescription" tradition of Geertz. Geertz (like the other anthropologists) is still directing his efforts to reinvent an anthropological science with the help of textual mediations. hermeneutics. etc. geoglra P. interpretive inter_ action ism. Discussions about this debate can be found in Seidman (1994). In contrasting Geertz and early interpretive anthropology with some of the later postmodern turns of such ethnographic writers as James Clifford. 0 a. de constructionism. the rather simple characterization of research strategies found in Figure I attempts to recognize the variation inherent in the range of work found in contemporary anthropology by placing "interpretive anthropology" adjacent to "exploratory/descriptive"(see. postmodemism. everyday meanings and how they are produced. however modified by new conceptions of discourse. rnent. and. In a short me th 0 d0 Io g section graphiC authonty y . cultural. author. and there is more nog . Sh e a Iso d'ISCU sses . tructured an I unstruc tur data quality.. epistemology. the use 0 f se mi s... the purpose ofresearch strategies under value-free account ( . £ n inte retive to trustworthiness (Kincheloe and McLaren 1994!. Thus. postmodemism. we find the writer-as-field-worker(Denzm ':h:Ot~ano~u:ingle th~t th:~::~:a~I. many interpretive studies are c oser In c ar h treme . particularly in the ethnographic context where the ethnographer is often the instrument of measure. y er . more specifically for anthropology. by Kuznar (1997). volv~s both an~~:~~~~~a. .~i~~ ~n~I~~~sno~h~hi~~~.e:r~t:~f:h~~~:~:r:~:~: . d· focused more on t rea s to validity.eveds m so s the challenge of . or text.~dmlts t~~~ contrasts her study h . 0 f her work and the . with findings story.exts. for example. s e Iscusse h' lIy conducting partic'ipant observation research in this rather c. h' he s stematic mode than to some of t e more ex descnptlve researc m t y . A variety of names are used in the lexicon of social scientists that can be associated to varying degrees with an interpretive strategy. and constructivism. that they question a researcher's ability to maintain objectivity. differ from systematic approaches in . ::t~e~~~~.~:~. an? even et~nograp rimaril concerned with threats 1994) While systematic analytical paradigms are p y h t to bel'levability . on the one hand. polyvocaI t. there is implicit concern for such Issues h roug ethnography. h h s as montages. Practitioners of almost all interpretive paradigms are searching in one way or another for some understanding (verstehen) rather than for some explanation of social phenomena. seems to follow naturally in the wake ofGeertz's interpretive turn. Phenomenology. since it is plausible account or stOryl~~ude of plausible accounts rather than just a single true thought paradigms are subjective. and Faia (1993). .rco~:t:~~:r~~::'tI~. If we talk. the.~~ring . 46). . He can invent his questions with few constraints. problems of work109 m d .or se ec t109 dispersed. like that of others in this volume. Geertz.omplex. some or all of the ontology. the "thick cultural and SOCial context (social.s as~: tha~ produ~t.~I~I~:r~~~e~~~dt~. of observatIOn and this blurs the distinction between what IS an descnptlO meaning 1°£ ?u7 extreme forays into experimental eththis distinction even further. A good example ofthlS IS Za us y.Impo 11' ance o f.) in her a~altlS. a major difference. 'ven created. Further. . in one way or another.n spac~ science t~at s. .~nterpretive t~~. . such as postmodemism and constructivism. he must constantly feed off others' texts. on the other. evoca Ive re writing strategies that mclude sue approae fictions (Denzin and Lincoln sentations. for example. onentatlOn . political. however. Zabusky clearly eh. maintained. Thus. and methodology of systematic approaches. tically and techa mgulS ed I'nterviews site and the group she studied.. more recent approaches. IC~he . This means that Clifford is simultaneously more firmly in control of his project and more parasitical. on the other hand. threats -as in "Do you believe my story. em P caos~:. Although some of the older interpretive strategies that emerged from the scientific tradition in the social sciences. ~ ed on the production of a behevable or . ' and and the effect of her role as ethnographer on mfonnant re atlon~ s for Although Zabusky doesn't talk specifically about design or a oU: conc~:t the potential threats to validity. OIcally diverse SOCial milieu. .

in that she was varl'ously I' I todry. For further discussion of research strategies in the interpretive mode.: These examples offer only a brief glimpse of the range of possible strategies in . structural equation models. mtervlewer. with additional ethnographic i~Si. thus allowing for a less . resentatlve of the genre that reJ'eets expl t' ' & : wo ways. For others. f healing power. mUltiple tests of a theory. m~se Along with this shift came the freedom not b .. Behar dISCUSSes the multiplexity . an t e S . over IOto hIStory of the principle shamamsm with its adroit albel't . In the following pages. ISsertahon to the flesh and blood of .. both qualitative and quantitative. at least in h IS View.~{. nograp Icencounterwlth the Y . although ethnography has been variously associated with a number of qualitative methods. and economics often led to rather standard designs. ceount ~ s amaOlsm. 1 n onnant mtervtewmg reliance on her own mem~ry . . First. Other examples' ropography" and Behar's (1993) f use of we. IS more extreme in its rejection of sysiematic Tre. . craditional methods sections are replaced by discussions on how to read the work i or on the particular methods used in writing the ethnography itself (see. The bad news is that the open-ended character of ethnography contributes to a less well-focused discussion of research design issues in ethnographic approaches. It is repI ' ana lOn In lavor of co . has influenced the nature of design in anthropology. .~~~~~~~~ka':~r:o:d~ashioned aknd theoretically unsoPhist:a~:~~~dT~a~e~~t~ e eXlve wor . as I learned that princiPI~ not o~ nf. Thus. Although this process might be equated to a method. I choose to work with' d'f~ t mg to rUf~e the placid surface of this primitivism it conjures into life-namet t~e:~. the design of research is driven by featuresofthe analysis.:~'~ l~:~tage in her co. s arnos sees It "I found If . (1987) historical and ethnographiCga g p~n~lPle was also central to Taussig's in South America This work' . collector. promise of yielding system from chaos do no h~xplanatlOn WhiCh. ~. ucmg t e ethnographic In this vem include Panourgia's (1995) p~o duct. -. sociology. In sociology. xiv) .unconscIOUS use of th . Whereas the analyticaltechniques most often used in psychology. concerned with ISsues of bias r restrictive ethnographic narrative AWltoh mghRsystemahCally. by interpretivists in anthropology. (p. e~re~enltatlOn of facts. This allows for flexibility. and wheres of a given project. For some. nveYlOg a moral tale.. see Fernandez and Herzfeld (this volume). ~. for •example. . but in profoundly different ways. ts purpose IS not a traditional attem t a real. like psychology. triangulation. . 1ft In the \ ratlve .ttve.J3 · In contrast to Zabusky. has recently published an ethno ra eSlgn ISsues. in anthropology the eclectic nature of ethnography leaves the design of research more open ended. . As he states: Research Design in Systematic Research: The Challenge of Making a Case In some social science disciplines. it is both a process and a product (Van Maanen 1988). multiple regression models.work In anthropology th example. ~ . Ra~os (1995). including the use of multiple designs within a single ethnographic context. Ethnography.' e ~eJe~ts the "anthropologi captures the "flavor" of her eth h' mtersubJechve understanding" th . . ~heir alchemical natural order. increased chances for various types of validity. with the strategies and methods of ethnographic p"sentation and With the refleXive character of the ethnographic enterpnse. editor. between synchrony and d'lac h rony Thus the . . translator. nvo ve as "prie t ' . Panourgia's discussion on the use of the parerga). • Part of the confusion stems from a lack of consensus on what ethnography really IS (Iohnson 1990). h' b re IS tradition.ue IOn IS argely devoted to discussions of her m wntmg the ethnography d h h' .nd they m her" Athenian Anthwoman in the telling of that wo. referred to as the anthropological method by Wi1\iam Foote Whyte (1984).n\ conflatl?n o~ modernism and the of montage. but political interpretation and ~ t expla~atlOn where facts are considered "realness.nConn.. Instead. rid' e ISCUSSlOn of research' deSign and methods of data collection a' . interp"tive se work is concerned more. dISsertation. ' pnnclp e of monta " ge as a means. but from Putumayo . 6).Research Design and Research Strategies /42 JOflNSON 1. for better relating the lessons of h'lstOry. interpretive work is an exploratory CIIlerpri with an implicit concern for methodological issues. it's better to think of elhnography as a strategy in which a variety of methods can be used in the quest for knowledge (Pelto and Pelto 1978).. . colonialism. 12). y om terror. analyst acad' The idea of a montage as ~n or :~il. To some.lIaboration with a single of roles. b. mto the self-conscious meanderings of e11' > makmg forays axis of analysis from the skeletonlik: dexlve anthropology m order to shift the ethnography" (p. there is a bod ofinte' . . e n t e 01 and the new developed over the last 30 yea A PR emologlcal emphasis in the field that has · rs. and terror IS Important In at least t . and validity or with the need for k' to e. ThIS contrast betw e h d reflects the increased variation in e ist . There are generally no ethnographic "analytical techniques" driving the design. Ramos emphasizes th s mlg t e found In work in the systematic the literary strategies used in prod . and peddler" (p.~ed ~n a rewnte of her 1972 austerity" of her original work in favor of an ". DUg amos dIscusses i ~ ." Second Taussig uses the . Analysis-of-variancemodels and multigroup comparisons (factorial designs) may dictate the whos. r. whats. I focus primarily on research designs in systematic . . and may involve applying more than one '1 It: nar~ As against the magic of academic rituals of ' .search. e magic 0 history and its -'1 . transcnber.ols~eur. an d vanous sources of data her intr d l' . The good news is that ethnographic research is amenable to a wide range of research designs. and path analytic models (all related analytical techniques) have influenced the design of survey research. and the potential for high levels of innovation and creativity. independent of their . . e hemergent and reflexive nature of data and · . ethnography should involve mUltiple methods. Thus.

d t rmine the manner m . al statements are derived from theory. multiple regression and text analysis). The levels at which theoretical concepts are measured (for example. groups to be studied). nominal or ordinal).. computer intensive text analysis._.-i . Figure 2 shows that the research process involves a simultaneous concern for the development of empirical statements from theory (for example. similarly distinguish between exploratory or descriptive studies that seek to describe and determine patterns in ecological data and those studies that specifically seek to predict or test hypotheses. .. and the application of appropriate types of analysis must all be considered as a part of the t EXPLANATORY Empirical StatementS Data Analy5ls t Data Collection Study . etc. is the foundation of good research. d . cost. data collection. analysis. the data can be reanalyzed. The of realistic constraints (for example. this volume). given the large number of computer analytical packages available for analyzing text (see Bemard and Ryan. .. for example. hypotheses). f will partIally e e S. the only complete remedy for design or execution errors is repetition of the experiment" (p. scope. however. Rather. . "Statistical analysis and interpretation. and there is a critical need to link theory. given some set so as to produce the most valId assessment POSSI eo raphical setting. meaningful and reliable measures). design. This is particularly true today. Theoretical knowledge is derived either from earlier studies Or from exploratory work.. and data analysis (for example. t p . "are the least critical aspects of experimentation. and interpretation in a coherent fashion. Thus. Currently. There is no substitute for a good theory. provide infonnation essential for constructing comparison groups. qualitative versus quantitative). although this is usually he articular structure of an empirical statement or h~poth~sis For example. ( tmc corn an . wh'. 189). On the other hand.) howempmc Data Collection Data Analysis Interpretnlion t Interpretation Conclusions EXPLORATORY not the case. but can also be used in hypothesis testing and explanatory research. Theory Development Theoretical KJ10wledge ---+ Study Objectives I 1'--- The figure shows that the overall research process is more than just a matter of study design. As we will see in the examples to come. . Figure 2 illustrates the relationship between exploratory and explanatory approaches within the ethnographic context. the types of sampling strategies used. the most common model has expJoratory research informing and complementing explanatory research.-. exploratory research is often an essential component of the explanatory research process. t methods of data collection and analysis. As with research in community ecology. Design.ch theoretical concepts are operatIonahzed esign. facilitate construction of structured questions or questionnaires. h be [1987] provides an excellent d. research deSIgn IS more than JUs r k 11 the elements of research together It involves constructing a logical plan that m s ~ble of some theory. redoing a year-long ethnographic field study because of such errors is quite another. ------/45 Research Design and Research Strategies 144 JOHNSON '-' research design. design (for example. or provide information necessary for producing a sound probability or nonprobability sample. Community ecologists. the operationalization of theoretical concepts (for example. No amount ofsophisticated statistics. or elegant writing can salvage a poorly designed study. This contrast between explanatory and descriptive or exploratory approaches is comm6nly made in non experimental disciplines in both the natural and social sciences." he says.scussIOn 0 d eventually analyzed.). ethnographic research can be purely exploratory or descriptive-involving a research process focused on producing better theory-or purely explanatory. the qualitative analysis of text and discourse is no longer restricted to either interpretive or exploratory approaches. Relationship between exploratory and exp anatory app ethnographic research process. Redoing an experiment because of fundamental design errors is one matter. data collection (for example.-Design Study Objectives I roaches within the overall Figure 2. Hurlbert (1984) emphasizes this in a classic paper on the design of field experiments in ecology. in that jf purely statistical or interpretive errors are made. Exploratory research may contribute to the production of reliable and valid measures. g g .

quasi-experiments.evel~p. a · (1979) make a sim nar distinction but refer to these kmd.:xtdi~~:~: Howevert. In natural settings. n . I (but see Harris et al.-~oOk and Campbell riments as (for example. Mensurative designs. companson groups e~ . J h son and Murray compare an the control (the pier without the FADS ) pIers.or olumn were alterrates. Two fixed fishing structures (piers) were pret~sted . when selection interacts with other factors to create erroneous findings).nant::vISeyteicxapleUrl~mit:n~:es:~. He t . even in a non laboratory setting. members agamst the potentia e e . (1982) in constructing a typology of research designs. .mental and observational approaches is siro ilar to This distinctton b~~wee~expe~1 rlbert (1984) distinguishes between two classes one in e~ologlcal fi:nn:t~h:~. Then. researchers leave themselves open to one of the worst criticisms of all-ofbeing "not even wrong" (Orans 1996).Research Design and Research Strategies 147 146 JOHNSON purpose of research design is to ward off as many threats to validity as possible and to help one eliminate competing hypotheses. a multitude of influences can threaten the validity of any conclusions. war. however. . Random assignment maximizes the probability that experimental groups are equivalent on key variables prior to the introduction of an intervention.mvO ve m ev evaluation research are most likely to use this deSIgn.c~~:l~:~~ft~t~~ ~oas~. True experiments are. Rather it creates an open forum that can contribute much to important theoretical and methodological debates. Experiments involve the random allocation of subjects to groups and afford the most control over distorting effects from extraneous factors.st ~aniPulative experiments.II::~: fisheries development use of fish aggregation eVlces . particularly fieldwork. f the types of comparative designs found then. 0 n detennined catch rates. It requires careful attention to detail and. often. FADS.. differences in catch projects.mg ~r ~:.o rev~luate the and Murray (1997). or interaction effects involving selection (that is.~tl:~ ~n~~~~:~~~~g.:::r:~~: hereas mensura d . . Observational studies involve neither random assignment of members to comparison groups nor the manipulation by the observer of independent variables. bies 10 space an lme an -.~:::~ anthro~ol~gical experimental and natural expenmental deslgns can b pp a ch Johnson field settings. and what I refer to as natural experiments. -. Without such attention to good design and methodological detail. or the Ut mg o. allocation a~d the m~niP~~~~::~~~P~. In other words. I borrow terminology from Kleinbaum et al. ma~~:ul~::~:7xp~l:i~~~:~. Thus. rare to anthropo ogy. are often impossible in anthropological fieldwork.. the measurement procedure itself caused a change in the dependent variable). allows for the most valid assessment of the causal relationship between the independent and dependent variables or response variables. duced phenomena ferent levels of exposure to some naturally ~ccl'. i . the effects of reactive measurement (that is.Cextraneous factors. o~ natur:1 e:I~~ment and "passive-observational studies. . A recognition of limitations doesn't invalidate a study's results. of course. there is no perfect design that can control for all possible extraneous effects at once..:oo. t multiple comparisons (for example. ~n~. umbrella-li~e ~n.es not afford t~\sa~ec. natural expenments are Slffil ar 0 .~:~~~~s. What separates quasi-experiments from true experiments is the lack of random assignment of group members.lty salT~p bl g b producing homogeneous comparative controlling for confou~dmg f var~: es d~ces comparison groups that are rep regroup~. with careful attention to deSign and e~~og. observational study designs. These are basically true ofexpenm ents . are observational and charactenstlc 0 in field studies in anthropology. and artificial manipulation of independent variables (also known as explanatory variables or study factors). In groupS. ..nit~~a:~~~ o:i~~~~~ ~:P:~:i. ~ expertrn ents 10 l) a d the manipulation of independent vanables. u(s. natural disaster.~r~:~.its suspended to t ~n~~~~:ed si~ultaneously nately placed at the piers and tndlVldual fishers were h . W T lin J scheme While random assignment al S 10 probability or nonprobabl. . He re ers treatment versus contro . unmeasured variables that might affect the dependent variable). True experiments involve random assignment and afford the best chances for controlling for things like: the effects of extraneous factors (that is. without f . with all other variables or factors controlled for.~~:." Anthropol~glsts .1 t quasi-experiments except tbat the Finally. for exdample. comparison groups differ because of the way they were selected and not due to the treatment). a lack of design and methodological detail makes it next to impossible to fairly and adequately assess the validity of any study's conclusions such that "rightness" or "wrongness" may not even be debatable.. Random allocation produces equivalent comparison groups. Outside the laboratory. Designs of this type. These and other sources of error are all potential rival hypotheses and randomized experiments are best at eliminating the threats of rival explanations. an admission concerning the potential weakness of a given design. nned rather than manipulation of independent variables occurs natural~y o~~:::~~ the basis of difay artificial or directed. 10 1 • . Nevertheless. . the effects of selection (that is. particularly in evaluation resear~h and d.~~i)uai~-e:~:.:i~~t~:~~us allocation. ven in uasi- n: ~~~~I:~~t:. but it do. Nonrandom assignment lays an experiment open to validity threats and reduces our ability to make causal inferences. Included are experiments. Ran om tests. random sampling 0 d unls:~r~in J meets the restrictions of some statistical P t ~ion as does random assignment of group sentatlve of such groupS. t hich involve simply the measurement to the secon? as mensU'datt!ve eXPdera':~:gS~ :umber of comparison groupS. the principles of experimentation are instructive and are a guide for understanding potential sources of error. volving random asslgnmen. ·th the FADS) and d d during randomly selected times at both the treatment (t e pIer WI .

such designs can have increased external validity and generalizability. Discussing "common sense knowing" in evaluation research. . More details can be found in these and other sources (for example. va:lOu l' ' f theoretical constructs and other considerations such as the op~rauo. hich e--Contro1 Study .:ture ofthc old and the new under conditions in which it is easy to make the mistake of attributing to the program results which would have be.ition for valid comparison. (1997) used a cross-sequential design. It IS assume t h d on one or more variables. if done properly. Often referred to as a survey study. x~er1m d and groups are compared. When one is interested in explanation. It would also help if the anthropologists were to study the school for a year or two prior to the program evaluation. d f etors are not controlled directly. the examples that follow are comparative observational designs. thus. 0 111 re r... s study designs. f" te stwlthoneormore . I . Quasi. and an anthropologist is usually in a very poor po. (1982) refer to as observational studies. which involved cross-sectional research on a cohort of children carried out over time. It is important. emphasis added) 'R h Designs Relevant to Anthropologists Examples of BasIC esearc ational Desi ns Study this is a longitudinal design where individuals are roups subjected to different treatments or ough time May mvolve comparISon g Design: Often referred to as a..na lza~o~~re I stress the importance of sampling. though. His idea is that "to know is to compare" is fundamental to explanatory work in anthropology: The anthropologists have never studied a school system before. control of variables during analysIs. f I lation Stratified samp mg IS d' f this type alloW for the slaUs lca popu . Campbell (1988) gives an important critique of ethnography.• l>'" expose . om the same underlying population. Designs of this type lack direct control over independent variables and.sign. s Ire. panel study. (This would be hard to schedule. s of observations is made both prior to some Design: One experimental group In whIch a sene treaunent and after the treatment. group does nol. .. are better or worse at dealmg With eac .. . however phenomenaJly absolute it appears. 149 From a statistical standpoint. • r . In anthropological fieldwork.. They have been hired aficr (or just as) the c:. However. have more potential problems with various types of internal validity and with the ability to assess time order effects and causality. Johnson et al.:pcrimcntal program has got under way. e variable} compares a group 0 f cases 111 w ~s'. members have some characterlstlc . This has apparently not been considered. designs that don't involve random assignment including quasi-experiments-are considered observational (Cook and CampbeU 1979).) All knOWing is comparative. (1982) and Cook and CampbeJl (1979). these designs and others can be used in tandem to test or explore components of a theory (such as combinations of time series and repeated measures designs particularly applicable to long-term fieldwork). Table I describes examples from observational and quasi-experimentalstudy designs discussed by Kleinbaum et al. Posttest only nonequivalent groups de. the importance of comparative thinking in ethnographic work cannot be overemphasized. (p. in combination with external.. of some type and posttest observatIons are made.'~rioss_SeclionaJ Study lA' it enerally involves a random sample of a target . re d t P sttest observations are made an group gets the treatment while control group oes no. like' true experimental designs. except for the new experimental pt:ogram. members Design: A variant of the cross-sectiOnal eSlgn d 'th a comparison or control group whose exposed to some variable of interest) is compare Wl members are not exposed to the variable of interest.Research Design and Researc11 Strategies 148 JOHNSON TABLE 1 . internal. . I The group receives a treaunent Design: Pretest observations are m~de on a sll1g e group.-- While the purpose of experimental design is to ward ofT as threats to validity. to contrast quasi-experiments to what Kleinbaum et a!. For example.. g d quate sampling of comparison groups.s determined without random .en there anyway. . and are inevitably studying a mi:. d on both groups.\ followe d th r . Expenmenta . . . groups in which the characteristic ~ .ment while the control Design: Experimental and comparISon or con E . Due to their predominance in anthropology. eSlgns 0 Although stu y a .. d to different condItIOns. 372. Posttest observatIons are ma e Pretest/posttest nonequivalent groups design [ '. Most research designs in the explanatory mode.. b P test observatIOns are ma e d allocatIOn ofgraup mem ers. etc. d that both groups come Ir of interest is absent.sign trol ou are determined without random gr p. I d arison or contra group DeSIgn: Expenmenta an camp . control versus treatment). I . there are several types of validity-face. 0 groups are compared. construct. members of the groups are ma c e Statlc-Group Comparison . . Robson 1993). but we might regard the current school ethnographies as prestudies for new innovations still to come. . often used to enSure a e . ental group receive allocation of group members. It would help in this if the anthropologists were to spend half of their time studying another school that was similar. In one way or another. Often. .Experimental Designs One group posttest only design . statistical conclusion. . are comparative (for example.gn' For some study factor (hke an outco~ . Interrupted time series design . The most common designs used tradi_ tionally by anthropologists have been observational in nature. as their own student experience and their secondhand knowledge of schools involve such different perspectives as to be of little comparative use. in their study of preschool children. d ' in which a treatment group(s) (that is.

mg. esearc process It' s . me ude hypothesized f synchronic.roblems wilh generalizing due to the nature of th . the aval.150 JOHNSON Research Design and Research Strategies ISI thinking through how validity threats have . e approXImatIOn to the t th lt .g. Bemard 1994). ? TABLE 2 Hist -Ch eSlgns T t?ry ange due to unmeasured or unobserved factors . b errors. r TABLE 3 Threats to External Validity ' Selection Problems with generalizing due to th e se I nonrepresentative) ectlOn process for study subJ'ects ( . the pretest/posttest nonequivalent groups design controls for some internal threats to validity. External validity is concerned with the .. Potential errors and bi.. will help lead to the production of solid evidence. changes due to how individuals were tested (instrumentation). whether probabilistic (Babbie 1990) or nonprobabilistic (Johnson 1990). For example. particularly with respect to accuracy and specificity. es mg-Change resulting from experience ained b Instrumentation-Change resulting from v g.. face-to-face interviews or mail-out surveys). How reliable are your measures in tenns of precision. can hamper the generalizability of one's findingS. e. In the quasi-experimental case th'" (that IS. chance). expanded to other sett. Internal . a IS WIt the 1" b'I' The threats in Table 2 deal 'th' genera Iza I Ity of research findings WI extraneous fa t h . data recording errors. e on a gIVen measure. I S means changes between pre. Each of these threats may hamper a researcher's ability to asseSS the contribution of a hypothesized effect to any changes observed. or how the data were analyzed (for example. such as problems stemming from biased samples or research in atypical or unique settings.. th y subjects as a consequence of measurement Regression-When selection of art" arymg e ~ay study participants are tested . The selection of any of these designs or the development of some hybrid design depends on the overall design of the research itself. and type of Bnalysis used are other important factors.antics. statistical conclusion validity). d ausa lrectlon IS ambiguous I slon of Treatment-Chan ge ue to one group re celvmg . nonresponse. I elects dealing with differences simil . (1982) offer a similar discussion of the strengths and weaknesses of observational designs in terms of controlling for threats to both Internal and external validity.lability of a sampling frame. The nature of the groups or characteristics to be compared-in tcnns of such things as the size of the comparison groups in the overall population. in that they are actually measuring what they Bre intended to measure? Attention and concern with all the potential sources of error. but it's problematic with respect to controlling for changes due to how groups members were selected (selection ). hidden or clandestine populations)-all influence the choice of a sample design. ' vations or data (for a more in-depth d' . cluster sampling.and external validity . e m stu y partiCipants over time d ..and e Ing can be expand d t . e study settmg (e. mfluenced and will influence obser . changes litalUra due to tion the selection of individuals with extreme pretest measures leading to regression toward the mean (regression). or .. sensitivity. is essential. and interviewer effects (see Pelto and Pelto 1978. and changes due to local events not a part of the study (history).operationalizationoftheoreticalconcepts. But it's not always easy to know who or what you want to sample and to know enough about these sampling units to derive a valid sample. Kleinbaum et a1..h quaSI-expert mental designs in a Ing WIt each of the threats to validity that are .ts to Internal . your JO to contain these' b • e IS loreanned Ta les 2 and 3 give examples of th .. Compensatory Rivalry-Participants' pere " (~ y toward glvmg all groups the same treatment ep IOns lor example ' th reats ) that affects perfonnance not a part 0 the treatment Threats to Internal Validity in Quasi-Experimental D ' Some Comments on Sampling Many probability and nonprobability sampling designs are available for any given research problem. In research design forewarn d' r . how the data were collected (for example. validity is concerned with th : o~ quasl-expenmental designs.ue to factors unrelated to expected effecls SelectlOn-Observed effects due to n d SI' oman om assignment of b e e~tIO~ by Maturation Interaction-Predisposition of el mem ers and nonequivalence of groups d s ected group members to grow apart AmbigUIty about Causal Direction-Whe t' 'ffu . resolution. ISCUSSlOn of how th t . informant inaccuracies.' Impact study conclusions see Cook d C b ese ypes of vahdlty can creep in at various steps' in the r an h amp ell 1979).h approximation to the truth . p lClpants are atypIcal or extrem . threats to external validity. presence or absence of a hypothesized effe t c ors t at may account for the invalidity). or aSSOciatIOns whether diachronic ~ round in Tables 2 and 3. stratified random sampling. Other sources of potential bias include sampling error (that is. the frequency of characteristics of interest in the population. Cook and Cambell (1979) detail how each of the Table I are better or worse at de 1" . and multistage sampling. . subsequent Mortallty-Changes due to participants d ' e regressIon toward the mean Maturation-Chang' d " roppmg out of the study . . Measurement. a11 or a portion of treatment meant for D another group Compensatory Equalization of Treatments-Tendenc . me~sures Will become less extreme and there will b . discussed in Cook and Campbell (1979 re. Careful attention to sampling. the use of imprecise measures.. These include systematic sampling. h' :V In the research setting. Similarly. n lme-or er and c i d ' . whether stemming from how the study was designed. contrast validity with pasttest.'ngs th t . and consistency? Are they valid. but this way of th "Ink" . setting atypical) HIstory-Problems with generalizing to either th e past or the future ~W P e~ . the ability to identify members of the population (for example.

In one whether classes or categories of p mgfi. Miles and people. ' a ~ane with respect to their values 00 . If you don't use random sampling.summa~~ dl~cu~~~ons of sample sizes (see. then sampling requirements may be relaxed.. think of selection in terms of some type rr'ncre~smg ~xtemal validity. ' samp mg. opening the way for the development of new test statistics particularly suited for the problem at hand (Noreen 1989. purpoSlve sampling.h k iJmlts and One that represents the po I f f . mterest. for of SllIllpling isslIes). mtervrewed memb f Eh· environmentalist group) and dry I . Johnson and Murray 1997). ty of mtema' and external threats to . pu atIon. This increases the chances or detecting potential bias and also makes replication feasible. Johnson 1990. design (including sampling). uaSl-expenmental case to use a static-group co t· d . 0 e nographlc mfo t . Some have defended Mead(Shankman 1996).r~~Sible to deve. Kempton et aI (1996) d . others have pointed to the biases and flaws in Freeman's argument (Marcus 1983. and data analysis from the beginning. much of the bias in the sample is a matter of the logic used in the original selection of sample seeds and any statistical analysis of the data must be concerned about violations of assumptions for the particular statistical test to be employed (for example. · . IOlormants or consultants (Wemer d S h fl rman s. These new approaches seem particularly well suited for the imperfect world of ethnographic research.ound m the po I · h lor a treatment effect across a wide f I . save Come to be a ' d· . I atlOn smce they involve elements of ' 0 mvo ves some form of d I· . c c asses . it does inform the glOg logic behind this model can be e t d d b classes wlthm the population.ole. random selection of locations in who . researcher if an effect holds across wide . . the model ofdeliberate sampl' h are random samples. n a le [1990] for detailed discussion Cook and Campbell (1979) discuss two sam . such as construct validity. Replication is extremely important to external and other types of validity. random w Iks . researcher to say something about h th f e. times. you should be explicit about how you chose the sampling units. events. . T us.a~o~u atIOo as a ~h. Ember 1985). . though . ' When generalization to a target population is th es ~. In most cases. If you are interested in generalizing to a given population. The x en e eyond the q . approaches or for the selection f th ' . ~ !" . ( lOr example. popu at~on of mterest. Such matters are particularly gerntane for observational designs using various social network approaches (see Johnson [1994] for a review).e case.' ere are a vanety of solutions. ea. Or /53 validity.' In the.lop a sampling frame from intercept sampling snowball I.' chosen to represent the range of su h I l"'" or events.L /52 JOHNSON Research Design and Research Strategies The selection of units of analysis whether settin s . popu atlOn allows the WhIle this might not be generalizedtoO~ e erect holds m a variety of settings. is directly re lated to how they can be analyzed. is important for understandln a varie g. but it is particularly importa~ fo . In that case. Each of these a r a . . ssoclate wIth qualItative . inclUding . a researcher may not be interested in generalizing 10 a population but may just want to know whether two subgroups obtained from • snowball sample differ with respect to some variable of interest.. assumption of a random sample from a population or skewed. d ' mpara Ive eSlgn environmental issues Kempton t . Ifgeneralization is not a primary goal. if you can use a random sample. sparse. This usually entails a . Bemard [1994] for . The next section shows how concern for the elimination of potential errors and bias through design and attention to methodological detail applies to discussions about the findings of Margaret Mead and Derek Freeman in Samoa. and not allow for generalizations about~ oach~s?as P?tentral problems. ese mo els don't n 'I' selectIOn and are consequently less powerful than ecessan y Involve random approach. where the rather restrictive assumptions of parametric analysis are often difficult to meet. households. me us 109 a selectIOn proc cl . This not always. and most do PoPU unknown errOr even if the meth d . . do it! No matter what the sampling method.t~OItS. pay careful consideration to possible violations of assumptions for a given statistical test. target classes ofunits ersons. p aces times .ran om se ectIOncnteria NonprobabIlity sampling method h Ich to mtercept respondents). c eanmg s op owners (who depend on toxic For Some populatioos it ma be im ' which to draw a sample. observational studies. are delIberately . or small sample sizes). But it is critical to remember the connection between theory. ' quota samphng. re On samplmg theory and example. independence of observations or random sample from a population). . Recent developments in randomization and computer-intensive methods of statistical analysis involve less restrictive assumptions concerning the data (for example. define a sampling universe Or f i r a ' ~ 0 ~ectlve. We mostly population. both in terms of measurement and sampling. a requirement for classical statisti:'1 ~uI:tlon and IS generally. Random sampling has been a primary requirement in the proper application of parametric statistics. validity in quasi-experiments Th d plmg models for increasing external .--- Huberman 1994). In some cases. random sampling procedures includ' d' . . you should strive to . samp I 109 across a range of groups th t . ere IS a vast IIteratu .or et~rogeneity. partIcularly key an c oep e 1987. h ers 0 art First (a radical chemicals for their business). T~ generali~e to a target IS essential if we are to generalize to a whole 0 e. because how the data were collected. testing (including both extremes and th range dOl c lasses ~n the set of all possible classes e mo a c ass) m th I . e ure Wit nown eITor random sample of some kl·nd Th ~u a Ion 0. Mead Versus Freeman: Research Design as Mediator Oerek Freeman's (1983) criticism of Margaret Mead's work and her findings in Samoa has led to reactions from anthropologists who come from different epistemologicaltraditions. random sampling of some kind is essential. How samples are chosen is an important element of any research design. the sample has to be represe~ta:~~~.

We certainly cannot hold Mead to the design standards available today. of sexua behaVIO no compart son . g ihe interplay of exploratory and show how. adolescent traumn.olescents m Samoa and the United States. and th differences in their ideological positions (Ember 1985). no matter the mix of limited later criticism of her wor\.. Iunits and adolescent troubles. The comparison group.. f II factors and the expenment may b ' t 'nformative and If care u Y . She : . given the time betwee Mead's and Freeman's studies. Freeman contended that some of Mead's informants lied to her and that Mead's commitment to a particul ideological position caused her to evaluate evidence incorrectly. e of the issues discussed so far. She could have chosen comparison groups that were as similar as possible in order to rule out the effects of unmeasured variables as much as possible. the differences in locations of their work.Ypol ~ 'ngthe n1'ght" Melbin (1987) . I ps are construe e . Mead used what can be referred to as a static group comparison design with a conjectural treatment group. and he worked at different field sites than did Mead. Mead made an assertion concernlOg Tb dal to Mead's argument. Meadcould have compared girls living in the households of native pastors to those who did not. en Yanous sources of data that w~re w ere was a lack of companso ns bet . has little control over a11 poss . and fo. . In field experiments. in common with Brim and Spain. a halcyon society. Mead makes ~ere is a lack ofwell~deftned samples fO~:u~t~: ~oWledge of the frequency of eventdsd:w. In lieu of the between-culture comparisons. differences in stressexpenence ya . In her use of a conjectural treatment group. She could then have tested the proposition that exposure to competing standards of sexual morality leads to higher levels of emotional distress in adolescents. exposure to Western culture. might have been due to onc or several extraneous (unmeasured) factors and might have had nothing to do with the independent variable. b ·lity Mead's research design limited her a I For Brim and Spain. useho ld and gIrlS hYmg Wit elr eo le and events. population and overstate the . not IllVO ve ran . many years after Mead's. noting how slight modifications in design and method could have thwarted later criticisms. relationship between the size ofres. rou s were chosen. Mead relied mostly On herself as an instrument to measure the variables of interest. to test the. h th· own family. even for less powerful :a~~i~". Mead could have made a withincase comparison that would have suffered less from problems with possible sources of error. any observed difference between the two groups with respect to the dependent variable. These examp es abies is a function of how comparative. t can e qUl e l ' . Orans's work was.r Grans. the design of the stul~y sc~~fi. I dom assignment of subjects to . . field expenmen s h th ses derived from and comconstructed can provide formal tests of .. More recently. For example. Still. There were possible problems with interaction between selection and the effects of extraneous variables. Orans (1996) reanalyzed Mead's field notes and correspondence and once again found that her depiction of Samoa as a halcyon society was at odds with his own impression of Samoa Field Experiments ible extraneous . American adolescent girls. there d:as . The following examples Illustrate som arative designs of vanous · th t incorporate comp examples are reviews of stu d les a t d 'In order to control for as .· Implicit in this proposition is the overall theoretical notion that culture is the m4ljor factor contributing to human behavior. ortwn of girlS in her study. More ke valid claims and pOSSIbly ethods would have improved her ch~ces to :caresearch. . But. of course. it is instructive to review her work through a contemporary design lens./54 JOHN SON Research Design and Research Strategies /55 The criticisms and counter-criticisms are difficult to assess. did not. There were no equivalent measurement procedures for the two groups. For example. That is.ed (Campbell 1975).lden:pansons among the different units. . however. . distribution of a sue . Was compared to a conjectural treatment group. Samoan adolescent girls. In IS wor as much more agonistic.In l~ ~l:~afor an ethnographers' hypotheses or ·D\ethods. Orans found itemized problems with Mead's research design. attention to issues of research design ~nd draw the conclusions she dId. mak: any systemat: ~oam. during Mead's day. Grans pomts out that Mead ma~e Similar to the observatlon1bY B n~ r·nbet:e en girls hving in a native pastor s .eral . Nevertheless. Brim and Spain (1974) recognized several problems in the design that could have affected Mead's ability to draw valid. I t'lon of independent vantype s in which noneqUlva ent grou 'bl and the manlpu a I many extraneous factors as POSSl e.- bunches to be rejected as we as Research Design in Anthropologica~ Practice: Systematic Research StrategIes . For exa~p e.prop 1 ment and operationahzation of key There is a lack of speclficlty JO the dey: :easurem ent on which to compare conceptS. Martin Orans did fieldwork in Samoa. the expenmenter. then her conclusions might have been flawed. Sev. Mead assumed some things about American adolescents without collecting comparable data. h· kon co omll ' plementarywith ethnography. Some of his experiences were incongruent with Mead's descriptions.oups.1t sbe h~d a tendency to understate the assertions about 11the ranty h events In a \tlOn. conclusions. If Samoa was not.gainst threats to validity (Robson explanatory approaches can ald III g g 1993). proposition that exposure to Western civilization increases adolescent trauma.

One point Was given . 75).. while the key experiment in:o~:ed dIrect pe:s onal contact among the ' Illustrates nicely the importa f n.- Melbin speculates that the variation in e I that the other three experiments inv I ~ s~ ts may have been due to the fact subjects. Young and Garro could then test the second hypothesis relating to the similarity in beliefs between the two communities. Returned keys' ch for keys dropped unwrapped . stressing physician accessibility as the most important determinant of . Had Melbin conducted only th h ' come to very diffi ' e elpfulness and sociability of .Western populations. The research design included the comparison of two Mexican communities that were similar in terms of cultural traditions and economies but varied in tenns of access to Western medical services.come across an implied need fo hI" m.- Control and Treatment in a Two-Community Comparative Design Oae afthe central concerns of medical anthropologists has been to better understand relationship between health-related behaviors and native perceptions about DJness.S. hav fulness and sociability lag I'n th d I and penl.n IS Implemented. . involve lawlessnessnct.::. He designed a clever experiment a. ease etu " 'h In p. mong other thin isolated settlements.'n th 'I t ey Were sent. s examp e also shows how setting. as . and three points if th' n s or eys returned wrapped Contrary to expectations nl'ght tl'm e envelope contained a personal note .ons. 156 JOHNSON ~- Research Design and Research Strategies theorized that the night was a frontier not ' nineteenth century. The hypothesis was that' the day.. h VISIt over a 24 h .o such mteractions.ber of studies have suggested that such a congruence was the primary detenninant aflreatment choice among Third World peoples. our peno covering day behaviors among the different tim Acre w. The idea was to see if th -. opposed to statistical inference. have a reputation for hel ' .ers Were n o t · more amIable than day-timers in thClr key-returning behaviors' in ti t h Meibin's three other tests s~pporta:d' t~ ey were the least "helpful. Tenned the "conceptual-incompatibility"hypothesis. ht . 0 md a key is reSIdents of the night would return ~ e p (p. astle (keys dropped in the maiJbox m: Wit an address encased Were delivered by the V. erent conclUSIOns regarding ' l ' nlg -tImers Thi I rea d 'y multIple tests can be in carp t d' . .us d~:~i e a . the researchers collected triads data and what they call term-frame data on infomants' perceived similarity of illnesses. M b' el m conducted four tests of the f e ' " ability. e as to what time of day they were picked' In all. ~nvolve novel hardships. . One of the primary purposes of the research design was the elimination of competing hypotheses-the hallmark of good research design-and ~1he testing of the primary hypothesis is an example of descriptive inference. mu lIple tests and measures (Stin h b a smgle test. An important issue in this area of research concerns the factors influencing the use of Western treatments among oon. Using a standard chi-square test. Posta" ServIce to the address on the keys w'U coded so they could be identili d I 1 postage due).ere a dIfference in key-returning to es. . ora e mto a research deSign within a fi~ld In all research. sen IS 0 tamed pr' t '" · ments I lke Melbin's individuals ofte . 326 keys were picked u of h' were also scored for the manner i~' wh . This established. The keys had the request to ::psl on aRverage more often than those of . a num. e eve opment of r ' and involve a variety of interest gro IJ' po ICles to exploit and regula up con IctS. . Young and Garro took a different stance. Ironically. e We -bemg of e ' . ' n I e studieS where in fanned con t' b ' xpenmental partICIpants. fi unhke the western United States in th' e am eatures in co A ey provIde escape and opportunity tal t ' mmon. One explanation views use tied to congruence between a client's medical beliefs and scientific medical I'theory: the higher the congruence. in experiramifications and c~nsequences of n pa~lclpate without knOWing about it The tl I expenmental out lOroug ]Iy before any experimental de' o . .ture :~latmg to helpfulness and soci_ locations during each two-hour field " III W Ich keys were placed at similar d . Young and Garro collected data on the number of illnesses that had occurred during the previous two months and the treatment each had received.w l~h 220 were returned. ccordmg to Melb' "T fi . helpfulness at night. the two communities seemed to differ in their use of Western medical services. there should be a concern for ethics and th pelrl. but particUlarly in field e x ' . They had to establish differences in treatment choice behavior in the two communities before they could assess any hypotheses concerning differences in beliefs. An important element of this position is that traditional medical "beliefs are not a barrier to choice of physician treatment. The town of Pichcitaro had restricted access (a 20-minute bus ride from Uric ha). e mal two poi t fi k m an envelope. Each of the keys were up and from what location. the more likely the client will choose a I physician's treatment. IOr 0 partiCipatIOn. Thus. consist' decentralized authority. This example' I ' nee 0 not relymg on . in statistical terms. have fewer sta." However. I . Descriptive inference is an approach highly suited for much anthropological research. with the exception of folk curers. comes mUst be considered sl. physician use. d an mg t. the authors found a significant difference in the frequency distribution of treatment altematives between the two towns.ments lIke the one described above U rk . Frontiers have c rt'. Young and Garro' s (1982) investigation of treatment choice in two Mexican munities is an example of a static-group comparison where the presence or nce of the treatment is based on selection criteria not directly under the control (the researchers. while the town of Uricho had easy access. but having the key experiment he may have c corn e 1987). the authors have more interest in the null hypothesis of no difference in beliefs than in the alternative hypothesis ofa difference in beliefs between . . From a random sample of approximately 10% of the households in each of the towns.wlde~ range ofbehaviors. e hypotheSIs of more sociability and l 157 L . Later. Young and Garro tested the two main hypotheses in sequence.

experienced recreational fishermen.11'. This is analogous to treatment IIId control groups without the random assignment of subjects to experimental units . to the production of impressive ev'd h 19n and analytical Issues contributed \ < ' I eoce t at casts doubt h . by cOlltrast. the rosters of sport fishing clubs in each region were sampled at random. both culture and language were held constant while experience with fish was varied. a visual inspection of the g a h' I no other conclusion than that ther: p ICI~ rlepresent~tions of the data could lead to was ut e or no dlfferen ' b i' " ce to e lefs between the (see Figure 3) . t' . ~ ~ an the term-frame interviews. . ' erence. Young and Garro (1982) communities. On the basis of the data from the triads st d d little reason to reject the "null hypoth . ~~t' '~n. Overall design framework for the Young and Garro stu dy. . four consisting of experts and one of novices. Each of Ibe four expert groups comprised \5 subjects chosen at random from a larger sample of recreational fishermen. 15 college undergraduates who had . ote that the anal' d h YSIS use to test the ypot eSls concerning similarities' b i' f ' stat' t' I' 10 e le S IOvolved de c . conceptuai-mcompatibility" hypoth . 2. I rmants. IS leads us to th I' su stantla variation apparent in th e use 0 f a phYSICian's . Thus the groups to be compared consisted of five groups ~f 15 subjects. Individuals . through extensive ethnographic background ICSC""'h. Individual informants classifying organisms on the basis of form or function? Boster '. abundance. a consequence of differe t' I men etween the two d n la access to such treat c m. Th~re are. .lIId Johnson used a static group comparison design to compare several groupS of expert fishermen with a group of novice lishcnnell. Is unrealistic to suppose that Young and Garro could have randomly assigned JQunity members to the different comparison groupS in order to control for jiGnfounding variables and then subject their informants to the treatments of interest. Despite th h" s np Ive 101erence not £" e aut ors claims of find. we see . S~ahng. Using multidimensional compared the belief data and found stnkin . language. Recent developments in statistical r e sense of statistical P Ilarlhes to aggregatedJ'udged-s' ' I ' . Comparative Design and Ethnobiology j. Four groups-from North Carolina. threats to validity in this study.'II£IIS'" shared similarities with regard to a number of important characteristics. Panicipant obs~rvatlon leading 10 the selecllon ot communities. and economics while varying physician accessibility b. Boster and Johnson (1989) explored this issue in an ethnobiological study of fish. a. Ethnographic interviews yielding the construction of disease and (eon-frames. it's difficult to know the To ensure that experts were. These students were the control group. ~ t . Using a questionnaire to gain background information. of course. All the groups were shown cards with artists' renderings and the common names marine species commonly found from North Carolina to Texas.. Static-group comparative design. there was no real way at least h ng no significant difw e~ t~e study was conducted.\lIId where the treatment is implied rather than researcher directed (that is. and Texas-were sampled to examine the 'i~V effects of different kinds of experience since there are regional variation in species . N on t e vahdlty of the h' ests.i . ro. Were .~~t. I 'i1!. Because respondents weren t randomly assigned into comparison groups. we can only assume that beliefs were similar . Young and Garro. but a greater in-depth exploratory understanding and an explicit of design can enhance our chances for the production of valid explanations. IS Ica mference. This 'Isn't perfect. although nonequivalent in the quasi-experimental . Ibe least amount of recreational fishing experience were selected from two Introductory anthropology classes. or to the availability of physicians in Uricho. Explanawry I. They conclude: g SIIDt antIes m the medical beliefs of . . 1462) responses of the two groups of info CSIS °Th"'o significant differences between the -.d\5Cussion i The authors' careful attention to research des' . produced groups that. (p.Research Design and Research Strategies /58 JOHNSON /59 the two communities. H d ' n oung and Garro's case..ent. test of primary hypothesis using descriptive inference Figure 3. to assess the extent to which any diffe ~ b bT r nces were slgntficant in th . uSlon that the treat e cone t b b samp les.Ity.lImobiologists have long debated whether folk biological classifiers are natural historians who compare animals on the basis of their morphological characteristics 'or pragmatists who compare on the basis of the utility of organisms._ . '5:': ExploralOry 1.. p oceduresallow us to assess the simIml anty matrices between th t an werker and Borgatti this 1 d e wo communities (see . an Hubert 1987) I Y . occurs without orrespon iog degrees of variation in reside t' tt' d n 5 a Itu es and beliefs about illness.. ~. In the comparison. natural " :jdifferences in experience with lish). in fact. ". Th'IS d"Istlnctlon d t wo communities IS in rt ' regar to anthropological research. West Florida. particularly with without narrowly restricting it to anal t' Iypot eSls-te~tlOg research can be done y Ica met hods usmg statistical inference. In lieu of equalization through ilDdomization. v o ume.'. The selection of control group subjects. in that h h' 'po ant.. control for culture. involved a purposeful selection procedure in which potential subjects were screened for recreational fishing experience. uencesofconfounding variables on physician utilization and beliefs about illness. East Florida. Given a lack of pretest observations.~"..

use to test theIr knowledge (see Figure 4).. (p. coherence. y. t. it was crucial to interview a wide range of couples who. reasoning as to a "lasting marriage" (see Figure 5). ~thnogmphic interviews yielding Ilsl of fish and belief-frame . a measure of morphol . s on t e owledge base .'nto a t e IS 0 tamed from extensive ethno-_~ ' W e /1 er and Romney (I988) Finall sen ence-frame colop I etlOn task described by · · . individual as so deeply unique and personal as to not be researchable in tenns of cultural universals. Based on in-depth interviews with 22 informants. Quinn was careful and diligent in her selection of informants. Based on an in-depth analysis of informants' discourse about marriage. In contrast. hold culture and language constant while varying 2. Although not generally representative of either the regional population or of the population of the United States. religious affiliations and ethnic and racial identities. Static-group comparative design. and that these models can be developed from the discourses of cultural members. at the aggregate and individual er experts and novices'. Boster and Johnson . That is. In my view. Test of hypotheses using both descriptive and statistical inferences. beliefs about graphic interviews were turned . or social class. experience with fish 2. JU gments of fish. o~gh such deSIgns are less driven by . d etermmed. . sort (see Weller this volume and W II JU ged slmIlamy of the fish-a free pil h ' . Quinn claims that her sample of informants represents the regions' population in tenns of the high degree of recent in-migration to the area from regions outside the South. Quinn views culture as being shared-that there are cultural models for a variety of domains that are widely . nom. . 399) Exploratory Explanatory I. Her sample is an attempt to capture the range of diversity found in the region. Quinn (l996)'attempted to build a cultural model of Americans' reasoning about marriage. ) Usmg use. . the consistency of her findings in this diverse sample of infomlants makes her case stronger (see Johl1son 1990). religion. As she puts it: All of my interviewees were residents of the same middle-sized southeastern city or its immediate environs. Were closer to the morpho_ logICal characteristics of fish (taxa . were not just from one region of the country or of only one ethnicity. J.L 160 JOIINSON Research Design and Research Strategies were asked to perform an unconstrained' d ". Issues of validity in this case are not as overriding as they would be in a purely explanatoty study. AI~n IS more a matter of discovery and exploration than the testing ofhypoth an established theoretical frameworkesthes. held in common. " There is a body of literature that views the interaction of culture with the . / . urt er. their neighborhoods and social networks. However. and her diligence certainly contributes to the potential validity of her model. Although the model appeared to be widely shared among informants from Quinn's sample and data collected from other studies on marriage. • ere stili IS need to . Quinn produced a cultural model incorporating a number of causal links in informants' Exploratory Research and the Deve1opment of Cultural Models Often the primary objective of research desi . 161 An example of research in this mode is Naomi Quinn's (1996) development of Americans' cultural models concerning marriage. all were native-born Americans who spoke English as a first language.c Istance) or th f h . Screening of potential control group members. nces etween pair f fi h used statIstIcal and graphical meth d s 0 IS. or even sharing. they were selected to maximize diversity with regard to such obvious differences as their occupations and educational backgrounds. and all were married during the period of their interviews. . further research in the explanatory mode is now warranted. finding commonality in the face of diversity provides stronger evidence of a shared cultural model (John son and Griffith 1996). Beyond these constancies of cultural and marital experience.'sta b oglCa SlmJ antics Was .~tatemems. statistical and descriptive infere e uses 0 fis (behefs about mformants use form or function flo / 'fince: the authors concluded that whether' . Because she was interested in a model that was shared. this is similar to Cook and Campbell's (1979) model of deliberate sampling for heterogeneity as one of several means for warding off threats to external validity. r c aSSI IcatIon depend h kn f th a e mformants and the methods d . and the duration of their marriages. a num ber of deSIgn details in the prop d I pay careful attentIOn to er eve opment of new theories and models. In principle. e er and Romney 1988) F h e t e use and functional characteristics of th fi h b ' ./ . research still has to be designed to test this model across settings and researchers. d O S to evaluate wheth ' . USing taxonomic d. although of the same culture. a. all in first marriages. d' levels.

language. Develop a cultural model of marriage.nine other behaviors that were outside the realm of the direct sharing of a single syringe by two or more IOUs. The design involved three communities that differed ID hISt?ry. share drug-mixing containers (cookers and spoons). Rubel et al.haviors and such things as producing valid models of seroconversion. and share the actual drug solution itself. Participant Observation and the Search for Validity As seen in Figure 2.. (1985) report on a study of a folk illness known as sustD. demographic. language. In most earlier research on injection drug users (IOUs) and HIV risk. education. These findings are undeniably important for larger epidemiological work that examines elements of IOUs' be. the primary risk factor was viewed in terms of direct needle sharing. neighborhoods and social networks.:rences provl e llCith d standing of the risk factors associated WIth the dIfferent types of ''''''ater un er h' th " . 1. This is a good example of the application of exploratory res~arc tU e . 2. In a subsequent study. ethnic and racial background. Statistical tests of g~oup diff. A major cOIl~ponent was the comparison of \DUs who engaged in both direct shanng and Of the . and marital status while varying occupation. Case-Control Study Design: Susto. Based on participant observation among \DUs. Termed "indirect sharing. religion. Rubel et al.162 Research Design and Research Strategies JOIlNSON 163 Exploratory 1.. Koester et at (1996) used these additional distinctions in sharing to look at the prevalence of injection-related HIV risk behaviors among several subpopulations of injection drug users (see Figure 6). The proposed design co~ld have been conducted in a single community. ." these nine behaviors can promote the transmission of HIV among IOUs who. found in many cultural groups throughout North and South America. I'th IDUs who engaged in indirect shanng only and those who '-"'~ct shanng w . One sUbsample was of individuals who complained of sustD duri~g the fieldwork or who had admitted their condition to relatives or curers. The ultimate aim of the study was to show the relationship between vanous SOCIal forces and sustD susceptibility. Koester (1996) and his colleagues (Koester et at 1996) offer excellent examples of the role of participant observation in more clearly defining the set of HIV risk behaviors surrounding injection drug use. ' d d :"":"~r shared directly nor indirectly. . Canuo! for culture. and relationship between orgamc dIsease and sustD. The authors were interested tU three pn~ary hypotheses relating to role performance and the presenc~ of ~he illness. + ~ra s~dY l i . Folk beliefs surrounding sustD attribute loss ofa critical s. Cross-sectional comparative design comparing 3 groups delennined on the basis of sharing behavior. 2. l l l Figure 5. In-depth ethnographic interviews with purposeful sample of informant selected for diversity. 3. --)- Explanatory (yet to come) Design srudy to test model. In a series of papers. Exploratory Explanatory 1. Koester (1996) identified. Thus.. -.. often share water for mixing of drugs or for rinsing syringes. although not sharing needles directly. Exploratory statistical inference to gain a better understanding of risk factors. Overall design framework for the Quinn study. A major component of the study was the comparison of \DUs who engaged in both direct sharing and indirect sharing with IOUs who engaged in indirect sharing only and those who Figure 6. carefully selected communities that were as similar as possible in termS of forms of government and gender-specific role expectations. -/ . A Folk lllness Here we look at an example of a study design used to investigate the extent to which disease is molded by culture. Overall design framework for the Koester and Koester et a1. '~~~~~~n of better measures of potentially important explanatory vanables. and culture but had similarities in social. Selection depended on l l l l 1 . I ubpopuiations of injection drug users (see Figure 6). share cottons for filtering. most large epidemiological studies of IOUs focused mainly on direct sharing behaviors in attempts to understand seroconversion rates and other risk factors. psychlatnc impairment and sustD. Participant observation leading to better understanding of IOU's sharing behaviors. . exploratory and descriptive research are often essential components of an overall explanatory research design.ubstance orforce due to a frightening experience. but the authors felt that the gener~hz­ ability of the results would be enhanced with a multiple case-control study deSign. and economIc factors. studies.

Her selection of variables allowed a comparison of different levels (for example. Catholic versus Protestant) across the four variables. Based on extensive participant observation. But the researchers' awareness of the problems. Rubel. there has been a shift in religious preference over the last 70 years. For some researchers. Thus. Using her experience as a participant observer. In addition to the I '. W IC 0 matched with males and females with fema~~~t:o dgroup bemg SIck. 164 JOHNSON Research Design and Research Strategies the conditi~~ by men in the community and may also have affecte . Use of standard statistical inference in lest of hypotheses. l. and controls.. and Social stress 2. within case control for ~sickness" status. increases our confidence in their conclusions. a on tearl!er perform social roles. . m?Te social 165 L L 'L L I ~ ~e~~~~'. Psyc?iat.'eal~. an association between susto and an individual's perc~ption oft~e a~. communitIes.ric impairment was operationalized using the ~. a.nd m ekachhof the comm~nities. organic dIsorders.f. Parallei to these economic changes. Goldin (1996) wanted to understand the relationships among religious affiliation. eoce 0 . economic ideology.UadO.~::~ . con ro group ~ontrol group members were selected from the pool of pat::::lts at ~Slck ~eoPI~. ethn?gr~phic research~:o~~~ Exploratory I I. Patient records provided the informati~~r~~ect~. study. Multiple tests are always much more convincing than a single test (Stinchcombe 1987). As she describes the process leading to the selection of the best model: The results of my study. Matched pairs Were mad n ~s~stados an~ ~ontrols were design allowed for a variety of comparisons includ~n~lthm co". She made an earnest attempt to control for as many biases as possible and. P311icipanls observation to help in constructing grealer generaIizabililY.~~:~ "'.si~ ~. there was a relationship between susto the suffering of more organic disease signs. measure of social stress. she applied different statistical controls in each of the competing models.n~om~ peoPI. ~~ ~:'.' i~dlVidual's inability to by one of the researchers.u~t:o~a:~~ess. and economic status in a Guatemalan township (Almolonga).s~anel .. combined with the strength of their multiple case·control study design.:~~i~~:i. must be interpreted within the constraints of the data collection methods. usU. First. This nder bjt . The stigma of susto among males and the greater social stratification encountered in one of the communities are possible threats to the validity of their conclusions.:. age and gender in matched pairs while Y~ing psychiatric impainnent. using the data collected during the survey. et al. Using standard methods ofstatistical inference Rubel et a1 found th t th 10 fact. the shift from Catholicism to Protestantism helps account for economic change. This study is important because of the authors' candor about the potential threats to validity they encountered in conducting the research..~~II~ess (higher-i. of course. sick people were com red t ' .~~~~:1i~.'mun~l1es only. and by matched pairs both within . tie members must also be sick. 2. These groups will be referred to as th:=. Figure 7. Although there was no association betwe"n susto and psychiatric impairment. Overall design framework for the Rubel et al. qualitative approaches were used to suggest different mechanisms and relationships that might be operating within Almolonga. Another problem involved the exist d ~he reportIng stratification in one community than ex ected . occupation. Goldin developed a survey which she applied to a random sample of 10% of the heads of households in the township (n = 57). Md culture b.e comfortable with the co::a~~~~~t~:~n~h~~. conducted statistical tests of the four competing models.from t~e com. This is an excellent example of a study design that incorporates within·study replication or multiple tests of a theory. oC'a tress Gauge developed Explanatory I Case-control sludy design. males were matched in terms of age. as Protestantism is more compatible with capitalist ideology and the accumulation of wealth. a. Using path analytic modeling.ever. Th~researchers were careful to make the control group as c~mpatable to th . This provided for an evaluation of the explanatory power of each. felt munities.:~ef~~ or her performance of critical social roles. L I L. ~hyslC~ans.~~~. ca:~t~~~~~ c~I~~~~. and eConomic fadors While' varying history.of •. compare across 3 cultures allowing for st~~. Participant observation ~ lending background for selection of cultures. was operationaiized :~~. language. Then.'ptomology and heaith problems were operationaHzed ' .~~~~ltC :::~'::~t. Her study design incorporates quantitative and qualitative methods in the overall ethnographic enterprise.~~:~ Multimethod Ethnography and the Comparison of Models Many peasant societies in Central America are experiencing dramatic economic and cultural change. One consequence of these changes is incre'l"ing economic differentiation. Because the asustados were "sick . control for social demographic. IhIDtcs ma e t e final selectIOn. of the condItIon by women. asusla os groups as possible.. Goldin constructed four plausible models that might account for what she observed while in the field.~::~:i~~dat~~.f~a~en~ Sy". Ho.

1920. J.. m the evelopment of a sound survey '. -1 . o m re le excl . As anthropologists. as Agar says. M. D. 72) l REFERENCES Agar. R. n eocles wlthtn the township. "move the world. Cook. the life hIstones of a nonprobabilistic sam 1 f' r. Crosr-sectional survey design the development of four competing models. qualitative data analysis can nOW be a powerful mode to test theories. Overman. Quasi_Experimentation: Design and Analysis for Field Settings. Research Methods in Anthropology: Qualitative and Quantitative Approaches. 1994. owever when a larg . i. on are more refleXIve ID character and more . Praxis. I . and possibly Exploratory L Panicipant ob$ervation leading 10 Th~s . CampbeH. Belmonl. e specl lcallon of app' d. . January. tended to support model study IS an example of multimethod ethn . Prediction. 2d ed. 1974. C. Boas. H. "Degrees of Freedom" in the Case Study. D. e newer for~ys ~nto experimental and Chicago: University of Chicago Press. MD: Rowman and Littlefield. American Anthropologist 91(4): 2. Campbell. 360-376. CA: AltaMira Press. and D. Cook. characterizing general trends with' AI I Y these mechamsms 10 terms of in terms of these general trends I~ d ~~ ~ng~.ets. ed. ropnate vanables. A. ' combmation of exploratory and e I ography m wh. New York: Academic Press. and D. all combining in ways ds to improve the chmces for credible results." It is mostly through attention to these concerns that anthropology and anthropologists will have the opportunity to. Bemard. S. lohnson. Comparative political F' 19ure 8. 1993. 1975. 1989. 2d ed.. strategies. Chicago: Rand MeNally. M. The Y e strength of the ethnographIC approach IS ItS ability to lOCorporate a WIde range of JIletho . AAA Newsletter. Overall design framework for the Goldin study. . l . d of the fOUf competing models. My conclusions must be interpreted Indeed. and J. E. In Methodology and Epistomology for Social Science: Selected Papers E.d d ~ tural anthropology only scratches possible within an ethnographic co~t: t nTh eSlgns. D. 1. CA: Wadsworth. S. With advances in computer technology. and economic ideology. and designs within a single enterprise. F. 1994..ch there was a and tests of models with data :~I~n~todry approaches-that is. T. Campbell. H. we should take full advantage of both our current understanding of research design and these new developments to produce a "powerful JIlode of argumentation. . occupation. New York: Holt. T. comparing across different levels of religious affiliation. Criteria of Social Scientific Knowledge: Interpretation. ave arrived at a v d'e< erroneous. l l Boster. I interacted with several·I'nd'vo.' slstent with my general characterizatio d h e IS ones t at were mean- l 1 competing models discussed above H n an W 0 were the basis for suggesting the . conclusion. 1988.nd IOU t that there are exceptions to them • . Rinehart and Winston. Qualitative Knowing in Action Research. The Professional Stranger: An Informal Introduction to Ethnography. The Methods of Ethoology.uslYe y on. economic status. Walnut Creek. conc USIOns (see F Igure 8). Explanatory t. qualitative data ec eft usmg a cross-sectIOnal design. Br~m. Boston: Beacon Press. 1979. Spain. and cO'-"binations of designs other ethnographic forms of presentatt . Participating observation aiding in the development of survey insuument and sample design. The stud ~hows Cl lcatton an~ assessment fosters triangulation that contributes t Yth dhOW the use of mulltple 0 e pro uctIOn of valid I 'methods .. Agar. 1996. 1980. D. . Studies 8(2):178-213. Lanham. Summary This review of research design and strate ies in c I the surface of the research designs h :. T. Form or Function: A Comparison of Expert and Novice Judgments of Similarity Among Fish. Survey Research Methods. Pp. Translated Woman.. criteria (Johnson 1990) she ml'ght Ph e 0 I~ ormants WIthout specified selection . 866-889. T. say. 1990. e representative sample of C as the one that characterizes the ge'nerCal ~e Cd rent ~ata s. . Babbie. ' ery lnerent. 1 I ua s who had Uti h' t ' h .• Research Design and Research Strategies 167 166 JOHNSON with believable and moving representations rather than the production of ed valid accounts or conclusions. advances in computer-intensive methods for testing hypotheses have the potential to expand the range of designs possible. American Anthropologist 22(4):311-321. Her research design illustrates th d without attention to sampling Had G Id: anl?edr m relymg on a single method . Behar. . . T.nst~=mm~spe~ . Research Design in Anthropology: Paradigms and Pragmatics in the Testing of Hypotheses. (p. puticu1arl in the imperfect world of fieldwork (John~on and Murray 1997). R. The combination helped Goldin in th . Applications of statistical procedure that allows for' statistical controls in the evaluation of the models. the township was aggressively pursued th d'ff. Similarly. 2." neem survey approach was used to evaluate the viabitit of .

lronsmith. Lincoln. Anthropological Contributions to th e Study of Social Networks: A Review. 1994. Hartnoll. eds. 1994. Morgenstern. 1978. Early Education and DevefJohnson. and R. Qualitative Research Methods Series. Pollution F . Kruskal. Pp. 1982. Zhang. Pp. Thousand Oaks. S. N. E ' . K. . . S.. and H. Kuznar. Chica~o: Aldine. Madison: University of Wisconsin Press. . Dekker. R. Vo!. 1996. A. Not Even Wrong: Margaret Mead. Reliability and Validity in Quantitative Research. L Johnson. Suow. and the Samoans. Freeman. emography of Brazil.• R. L L.ho '/ ~i. om matlonal DataAnalysis.. Thousand Oaks. M. and B. Johnson.. enzln and Y.. Walnut Creek. . QuaJltative Research Methods Series . R. The Proeess of Drug Injection: Applying Ethnography to the Study ofHJV Risk Among IDU's. 1994. " . ElJen.. . Derek Freeman. Wiley. F. D. M.. '". ' Johnso~ 1. In Handbook ofQualitative Research. nces. ]ohnson. Moran. M. g Ecological FleJd Experiments. Miller. Assignment Methods in C b' . ~. L. F. Griffith. and M. Anthropological Myth. In A Handbook 0/ Methods in Cultural Anthropology.. S. Pelto.1.22-23. Huberman.: Naroll. Rethinking Critical Theory and Qualitative Research. Thousand Oaks CA' Sr npay's. c. G. J. 1995. Social Forces 72(2): Hubert. Cambridge: Cambridge University Press. Kleinbaum. Pp. New York: The Free Press. Cambridge: MJ. J. Pp. Boulder: Lynne Rienner. 1973. March 27:3. 1984. Kupper. u IcatlOns. 1-19. D.m ore me Managemenl/2(3). Ellen. Melbin.. Cambridge: Harvard U:ive~. and G. London: Routledge Press. . In Thousand Oaks. Pseudoreplication and Desi n of . MD: University Press of America. c. New York: John c.. The Pertls o/the Pos/modern. Strauss J967 The D' . CA: Sage Publications. H. . Panourgia. LeVine. and D. ho Are the Whites: Imposed Census Categories and the Racial D 451-462. . 143-158. Kirk. L. Belmont. The Comparative Analysis of Human Societies: Toward Common Standards for Data Collection and Reporting. Pp. N. S Pp. Selecting Ethnographic Infor . J. L. 1997.ood Safety. eds.GaIaskiewicz eds J h ' . W. A. The Prevalence of Additional InjectionRelation HIV Risk Behaviors Among Injection Drug Users. S. E. and Y. Naroll and R. In AIDS. CA: AltaMira Press. Lincoln.. Drugs and Prevention: Perspectives on Individual and Community Action. R. Vol. Kempton. Thousand Oaks CA Sage P bl' "!ants. CA: Chandler and Sharp. e FIeld of Qualitative Research. 2d ed. eds. Qualitative Data Analysis. M. L L L L mber. 113-151. Models of Economic Differentiation an Quanlllallve Anthropology 1-2(6):49-74. Murray. Grans. c. CA: Lifetime Learning Publications. Kingston: . 183-195. N. and J. lsCovery of Grounded Theory: St~ategies for Goldin. New York: MarceJ Hurlbert. Evidence and Science in Ethno r ' .. Margaret Mead and Soma . 133-148. and A. I. H 1937. R. 0 'ren. R. Qualitative Research. M. Koester. Whitcher G Develnpment of Social Networks in Pres. . 1995.i~Y~~~. Introduction In Ethno ra h i e ' F. 1-12 LO~don' Acalfd p. Knowledge. New York: Rinehart. Byrne 1993 W . I' !I' . /38-158. Cohen. What's Wrong with the Social Scie? . Wish.' h A ~ A EffectIveness In Development Potential and Pitfalls.e. and C. Reclaiming a SCientific Anthropology. and A. 1996. onsorte. 1996. a agmg marme Con"icts S . Beverly Hills. 8. 22. C. C. M. Evaluatin F D nac an . ed. age un lcatlOns. E.. Multidimensional Scaling. Rhodes and R. and Y.'~' : WassermanandJ.. J. cds. Fables of Identity. 1986. Boster. CA: Sage Publications. ed. Cohen. McLaren. T. Koester. l. . Lang. 1973. L. Lanham. 1987. and P. M. J.Devices in Developing Fisheries. 1997. B. B. E. P.. pecla JssueofOcean . Pp. W. Press. 1987. C. 1990. 1983. Denzin and Y. Fragments of Death. Hartley Culture. 1989. CA: Sage PUblication~. Pogg. Booth. New York: Columbia University Press. Night as Frontier: Colonizing the World After Dark. 1993. . R. Journal 0/ Harris M J G C . 1997. A. an d . American Anthropologist 87(4):~~~i:et1ectlonson the Mead-Freeman Fal8. ICMRD. Research Design in Anthropological Field Work. A. G. . Pelto. Noreen. and R. and the DIstribution of Miles. New York Times Book Review. S. J.. Epidemiologic Research: Principles and Quantitative Methods. 1984. Pollnac eds 1989 At n e. ' . J985.. The opment 8(4):389-406.g and Unmaking of an Glaser. S. f' ". K. and J.. 1978. Lowie. o nson. pResearch:AGuidetoGenerafConduct R. Handbook of Qualitative Rese~rch K D g. CA: Sage Publications.L L L L 168 JOHNSON Research Design and Research Strategies 169 D enzln.~odteat. and M. J. Marcus. L. J. Computer~lntensive Methodsfor Testing Hypotheses. 1983.. D. . The History of Ethnological Theory. Lincoln 1994 Enterin th . emlC cess.. W. H. In Advances in Social Netwo kA I . L. New York: Columbia University Press. Anthropological Research: The Structure ofInqUiry. A. Human Ecoiogy 24(1):87-110. Journal ofAcquired Immune Deficiency Syndromes and Human Retrovirology 12:202-207. 1996. A Handbook ofMethod in Cultural Anthropology. Environmental Values in American Novato. 1. ProJects: Theory and Praxis In F. L. J. Thousand Oaks. eds.. N gKincheloe. L: R: 1996. 2d ed. G. d Cultural Change. eds. T. Ecological Monographs 54(2): 187-211. 'PoH IS ~gregatlOn .J. One Man's Head. CA: Sage Publications.Controversy. 1996. .

The History of Samoan Sexual Conduct and the Mead-Freeman Controversy. Chicago: University of Chicago Press. The Postmodern Turn: New Perspectives on Social Theory. In A Handbook ofMethod in Cultural Anthropology. Robson. 1996. O'Nell.es In . Naroll and R. K. C nd A K Romney. Ruhel. Method. In Constructing Knowledge: Authority and Critique in Social Science. h' F' Id' A Guide from Experience. F. Systematic le war . Communltles. Furlow. o. Vanatlon to the Olce Young. 1913. Schneider. Culture Contradictions: The Case of America's Reasoning about Marriage. Spindler eds. 118-138. Constructivist. A. Representations afC Social Facts: Modernity and Post-Modernity in Anthropology.Research Design and Research Strategies 170 JOHNSON J7] PlaUner. C and '. 1996. 1988. Sanuma Memories.e . Researchers. SU$/o. '. Schoepfle. New York: Columbia University Press. . r l l l l Schwandt. an . Pp. eds. An Example of Research Design: Experimental Design in the Study of Culture Change. W. d G M. In' Anthropology Between Science and the Humanities. J. CA. Trust in Numbers. a. . eds. Goldschmidt. 1993. In Ethnographic Research: A Guide to General Conduct. 1987. R. 1995. Constructing Social Theories. CA: F 1984. Rabinow. Thousand Oaks. 1995. Collado·Ardon. MethOds Senes. Tyler. Thousand Oaks. . I. Marcus. J. M. CA: AltaMira Press. Seidman. American Anthropologist 98(3):555-561.. 1982. New York: Columhia University Press. A Folk l/Iness. W. A. T. 1987.' P inceton: Princeton University Press. eds. Pp. 1973. Space SCIence. F d L . Pp. R. C. London: Sage Publications. 1994. 196-209. Naroll and R. Pp. Learmng/rom t e le . Madison: University of Wisconsin Press. Cohen. \9 67 . Scientific Anthropology at the National Science Foundation. 10 A Handbook of Methods in Cultural Anthropology. Experiments in the Field.. . CA: Sage Publications. London: Academic Press. Lincoln. Chicago: University of Chicago Press. Shamanism. Clifford and G. Quinn. Real World Research: A Resource for Social Scientists and Practitioner- . Porter. '. Launching Europe: An t nograp Y Zabusky. Shankman. 1996. 1987. . 1985. We\ler. 1991.. eds. Whyte. M. A. S. T. In Handbook a/Qualitative Research. eds. L. 1984. and R. Taussig.. D. E. CA: Sage PublicatIOns. Van Maanen. Oxford: Blackwell Publishers.. Walnut Creek. 1994. Chicago: University of Chicago Press. In press. George SplOdler an oUlse . C. nd Medicine /6:\453-\465.' . Weruer 0. L. R. Qualitative Research 8 Systematic Data C0 11ectlOn. R. J. . Pels. Sage publlcatlons. In Writing Culture: The Poetics ond Politics ofEthnography. Berkeley: University of California Press. L. P. N. \ Anlhr~pological. C. Colonialism. Vo! 2. P. S. G. Sechrest. Scientific American (September):4(}-42. Y k H It · Id Melhods '" rhe Stu 0 W'l\iams T. S. 1986. Rinehart and WlOston. J . Interpretivist Approaches to Human Inquiry. 1\0 Thousand Oaks. 234-262. l V . 78-95. Ellen ed. and W. Berkeley: University of California Press. A History of Field Methods. R. . ed. A Post-modern In-stance. . Tales of the Field: On Writing Ethnography. Alarming Nets. A. New or: 0 .' Ch' of Treatment in Two Mexican L Y Garro. A. Nencel and P. N. Spindler. 21(}--219. SoclOI SCIence a Eh h of European Cooperation In S E 1995.. Pp. J. 198 . Cambridge: Cambridge University Press. . Newbury Park. S. Denzin and Y. l l l l l. S. Ethos 24(3):391-425. Ramos. and the Wild Man: A Study in Terra and Healing. pp. Cohen. 35-62. . Stinchcombe. F' Id 1<. Sage publications. Urry.. Princeton: Princeton University press. . dy if Culture In the series Stud.

Access to resource persons may have changed dramatically. For example. When the anthropological researcher and persons studied respect and trust one another. there are positive feelings and outcomes on both sides. or even been denied. However. The anthropologist's methods may be questioned or his or her position in the social group may be challenged.have been /73 L L L L . not extraordinary. in Vietnam in the 1960s and 1970s. and in the People's Republic' of China in the 1980s. Matters of ethics are an ordinary. However. the art of negotiating and repositioning oneself resolves to ethical principles and choices.L L L Five CAROLYN FLUEHR-LOBBAN • L L Ethics L L Introduction Considerations of ethics are often raised after some incident has occurred or when an anthropologist feels discomfort about the conduct or progress of fieldwork. the anthropologist may wonder what has gone wrong and why. Or the anthropologist may receive funding from a private or public foundation without haVing fully considered the conflicting demands and responsibilities between the funder and the people studied. ethics and professional responsibilities . the anthropologist may not have been well prepared as to the social and political environments of the people to be studied. So. part of anthropological practice. if an anthropologist has had little or poor training for the field. During the research. These occurred in Latin America in 1960s. Research goals may appear suddenly unattainable and. there have been some celebrated cases of alleged ethical misconduct involving major political events of the day. confused or desperate. it will be difficult for him or her to resolve the everyday dilemmas that are part of the practice of the discipline. when the field situation changes and the anthropologist's position becomes ambiguous.

Boas. A prolific writer and experienced ethnographer. like other human beings. while erop Yntatl'ves of scientific institutions m the C ~. The issue eventually pologists working . nongovemmental. Statement on Ethics of the Society for Applied Anthropology 1963.Ethics 174 FLUEHR-LOBBAN J75 sensitive and controversial topics. and Boas. ernments as represe h h themselves to lorelgn gOY f iog on scientific research. Until recently. in 1967. l I I research.n :~or gr"duate students to strengthen OWS for undergraduate study abroad ~n e d . -. or the lIt e 0 intelligence activities (State". as a cover for AAA's first statement on ethlcsmakesexp.velOPing the national Security appreciation to Senator Boren for.as govemmen adversely affected. focus of t Ii~it . Boas. and saw its role as creating a document for general use by applied anthropologists. l' that counterinsurgency research tU of the Anny's Project Ca~elotl after reV~eatl~:s ased social science research. area studies scholars with mlhtary an about the risks of the relatlO~shlP of forelg~ . In a letter published in The Nation.eference to the avoidance of the use f anthropologist. The first formal statement on ethics was developed by applied anthropologists in 1949.:.tudieS Association. may be reluctant to talk about it for fear of "having done something wrong. and Proposed Statement on Professional and Ethical Responsibilities 1983). Moreover. It involved Franz Boas.'p.me . employers.ere tempting. cr'ttt'cal world areas (Preliminary . t elf. The Latin America was a major.. the SfAA has amended its guidelines twice. bave arisen dunng the years 0 . anthropologists. the ISSU~ o~ anth PI d d the AAA drafted the first ment has retained its potency and S1gTnhlncanc~. widely known as a pacifist. significant changes have taken place since 1994." Sometimes.J Q. and as sent for the purpo~e 0 c~rr. have also done the greatest possible 1 shaken the truthfulness of SCience. fundamental issues of research-such as the relationship between anthropologist and research collaborators. The following resolution of the intelligence-gathering organtZallons and pnonlIes.. in 1963 and 1983 (see.S.a s u AssoctalIon.\~ Bo~.. In this chapter on ethics and methods in anthropological research.tddle a~t St d' es Association in expressing Studies Association and the Latt~ Am~rtc~~ 'nud. but sensitivities ~ay ~ave subsided. : ~obeen declining and the Department Although funding for intem~tlOnal stu l~ a criticized by virtually every major of Defense monie~ . (Boas 1919) History of Ethics in Anthropology The history of ethical discourse in cultural anthropology intersects with national and international politics and the changing contexts and paradigms of fieldwork. or voluntary organizations. .hts l~a ers IP t bout the administration of the Education Act.UJ l :~t o~rce to Isse scientific inquiry. orne to my hands that at least four men who B acddent incontro~erllble proof ~as C to cd as government agents. The Society for Applied Anthropology (SfAA) Committee was chaired by Margaret Mead. The first ethics "incident" took place in World War I. 1by the anthropological estabhshment . and I shall try to balance the broad historical picture of crises about ethics and anthropology with recent developments in the profession. Nevertheless. The American Anthropological Association (AAA) produced its first statement on ethics nearly two decades later. intro~uced on anthropological work. the Latin Amencan Stu let' I StSOdelnt app'licants were cautioned . 1991 The NSEP provides funds legislation sponsored by Senator ~a.I of the name of anthropology. just raising the issue of ethics in fieldwork causes a defensive reaction. because ethical conduct may be perceived as overlapping with morality or personal principles. th' 'n 1967 fonnal statement on e les 1 . lIi6~ ds made available through the A more recent example 's that of e e~~ ~ program (NSEP) for training Department of Defense's ~ationaI sec~~~~ies u~~:~SEP was established through in foreign language and internatIOnal ' . Various incidents have precipitated crises that have forced anthropologists to hold a mirror up to ourselves and to be reflective about the consequences of our actions.ica:a. one of the most famous American anthropologists.' career was nt~~e subject even though the same situaUon might persisted for BO~s did not resu~e~orld War II while he was still alive. indeed. . the NS. The person being queried feels that he or she is being accused of wrongdoing rather than just being asked whether ethics have been considered in the fieldwork plan. I want to focus on how to make ethical considerations an integral part of the ordinary. . perhaps the founder and designer of the unique holistic brand of American anthropology. T ey ave united States. ro 010 g ists in the employ of govemThroughout this century.:~:ion to the activities of several anthrofor having dra. mbers and poten . U !-. .'ent 1967:S.. d . objected to the wartime activities offour anthropologists who had combined intelligence gathering with their was censured formally by the Anthropology For this letter of protest. the Middle East area studies assoclatlOn-lnclUdmg the A d' A c'ation and the Asian Studies Studies Association. Further. more defensivemaneuveringthan an affirmative tackling of ethical issues. tnheeewa'ke of the U. . E Studies Association joins the African The Board of Directors of the M. saciety of Washington and lSolat~~ l~ ge. We also share thetr senoUS concerns a l l l "r.:n the general pu ~c i~:el\igence gatherers.~. Since that time. However. Guidelines 1994--95 Pilot Grant: progra~. Boas also had a strong civic sensibility and was a frequent contributor to numerous nonanthropological periodicals. day-to-day practice of our craft. and the protection of informant confidentiality-are all significantly present in the statement. t atlOnal e ucatlOn In ) the national capactty In In em N f nal Security Education program . . after discussions that began in the immediate post-World War 11 era. f oncems that the area studies Middle East Studies Association conveys the types 0 c associations had: . historical concern with matters of ethics has been more reactive than ' proaetive. h . Department IS was In .

aa~onditlon for the conduct ofresearch '"'l. after heated confrontations in print and in public meetings. became more common and tended I d ngovernmenta agencteS . Other strong principles reflected the tenor of the times. erS (P 'nciples of Professional Responsibility 9 take appropriate measures.. suggesting that th~ I ~ren 'tself Discussions of difto challenge the traditional holism of an~ ro~o Og~ ~ractieing anthropologists. allegations that anthropOlogists had engaged in counterinsurgency research. were reluctant t~ See any modific~~todn have summarized some of the responsibility in research. published in Mesa Newsletter 1994. these individuaJs must come first. Defense and intelligence community for a major foreign acea research. to the propriety ofthese actions an (3) when an anthropo th the ASSOCiation may mqUlre ID . Perhaps the AAA took no direct aclion in the National Security Education Program because it was mindful of the need for the funding of international research. and aware of the more direct action and interest of the area studies associations. . recalling t ~ ea lthe principle that the first created.. and some turned to anthropologists w~re unabl~. In this code. In particular. 1992. . col~eague~ OglS .a 0 National Association of Practicing A~throa new profeSSIonal orgamzatton. dhost governments anthropologists . However. more anthroin 1986.rofesstonal. It also can create danger for students and scholars by fostering the perception of involvement in military or intelligence activities. the broadly based reaction to the recent Department of Defense's NSEP is a reminder of the seriousness of the issue of the use or appearance of the use of anthropology or any discipline as a cover for intelligence gathering. between aca ernl~ ~n ut in various forums includ~ continued throughout the 1980s. . [The MESA Board] deplores the location of responsibility in the V. for govem. . ferences. t d'ff. h h d days in which the PPR was Many academic anthropologists. just as the entire American society was tom asunder by issues of politics and morality in the Vietnam war.l. again contract research.' contract or proprietary research Vietnam War the immediacy and mtenslly of 199~~~'ghout the 1980s. . The Vietnam War era was the most important turning point for anthropology and other social sciences.. and intelligence agencies. and the involvement of the CIA on the National Security Education Board. ." They attempted. It caused us to take matters of professional ethics seriously. this time in Thailand and Southeast Asia. t?e the PPR to make it more accommodatmg to demicanthropologists. even if more than half of anthropologists weren't female..) L L L L L IL L I The AAA. a strong first principle Was enunciated that lasted for a generation: "In research the anthropologist's paramount responsibility is to those he studies. but this language illustrates that much has changed in ethics as well as gender since I971. unsucces~ 0 They eventually created their precipitating charged debates wtthm the pro es~." Today.tod ~c. din of the AAA in 1902.rnenta ~n no Vclandestine research for governments. The end result. In the decades after L (Partial text of resoJution adopted by MESA Board. practitioners. . It demonstrates the continuing relevance ofBoas's admonition of 1919 to avoid such ties to defense. governmental or not. 1 71) students or o.1S to the p~ople sth~oleOI~ ists expressed during this time gees were significant enough tensions between academiC and apphed ant (Fluehr-Lobban 1991b).e ~is demographic transformation meant that professional work m th~ app ~e te . while it cautioned its members to negotiate and understand clearly their relationship to their own and host governments. These debates and dlscu~Slons ev~ e \ : su "dialogue for a new era" volume on ethics and the profeSSIon of an ropo gy. n act. that even the appearance of conducting clandestine research hurts the good reputation of anthropology and its these selected core prmctples fad 1 . Once again. It sought to r~vtse the interests and needs ofP.o . and may limit academic freedom. (2) that in relations wl~h o~~i~::. real and imagined. precipitated a crisis in the field. They w~re came 0 I f which I organized and ing symposia at the annual ~AA ~eetmgsk s~vra 0btitie of my 1991 edited chaired. Ethics /77 I L L L f /76 FLUEHR-LOBBAN . When there is a conliict of interest. Between Applied and Academic Anthropologists TenSIOn . (I) that no secret or clandestine research be carried out.. to develop a code of conduct.. the resolution was revised 'and passed by referendum vote of MESA membership in 1993. did not take any specific action nor pass any resolution regarding the creation of NSEP. for the first !tme smce the '~doun f g d mia than within it (Fluehr-Lobban pologists (51%) were employed outSI eo aca e · f I bbying force Wt m . This connection can only increase the existing difficulties ofgaining foreign governmental permissions to carry out research and to develop overseas instructionaJ programs. was the formation of a Committee on Ethics and the drafting and promulgation of the first code of ethics by the AAA in 1971 (see Statement on Ethics: Principles of Professional Responsibility 1971).o~'to the complexities of contract own set of ethical guidelines that were more sut e t~e ~nd of t~~ dramatically'. . pologists (NAPA). (Fluehr-Lobban 1991 a). practicing anthropoIOg~Stht~ intchreeaAseAdAin ::~~:~~~::. education and training program of students and specialists. d turn Many professionally tramed cided with the post-Vietnam War econo~~~e~~nempioyment. Anthropologist was pitted against anthropologist. in particular its location in the Department ofDefense.S. military. should not be require~ t~ CtO~~i:~I:~ :ctionsjeopardizes peoples studied. . the reference to "he" would be unacceptable. and to provide a mechanism for enforcement.:~ 7 and proprietary research.. Thts c?mto neutrahze the tssue of secre . nonac~ th t the "first responsibility of the the practicing anthropologtsts we:e concerne t ~ a s be met under the terms of anthropologist to the people studted could n':ul~ ~ ~evise the PPR in 1984. as to constitut~ someth mg.

de r~infor~es and may even reinvigorate the " With. h where anthrover propnetary researc . co~flicts between the i~terests of ~~ders 0 rese:~\ robably more mythology ~n the same ways that Presumably. In the end. to mediate disputes between coIleagues over matters of plagiarism.s concerns For example. ical fossil and historical record. and matenals they s~~dy and to 'b'. The most important procedural .' The commission discussed cha~ges in the PPR and reached consensus on a number of fundamental issues. f n to the pro I. l' fto ethlca Issue h ' ry ethical responsibIlity to t e pe p . explicIt. are ~nvo ~e I . Based on her experience as former member and chair of the AAA COE. proprietary res. that the AAA code didn't address theIr subgroup des of professional ethics for the CO when the new code was drafted. These codes had been deve o?e m re. h to respect the we . one t t ~ hasizes ethics education. in many professional codes ~nd IS purpos~ u y 'mpiistic Practicing anthropolo. h' hhas evoked such passion in the The issue of secret or clandestme research. 1996:18).Ethics 178 FLUEHR-LOBBAN }79 Some subjective sources of tension between practicing anthropologists and academic anthropologists also surfaced during the the 1980s.stUdied. ~~ br~~o But in commenting on the . Janet Levy concluded that the grievance mechanism should be reviewed and possibly eliminated because ofthe complex procedures involving multiple levels of review and the slowness and general lack of resolution of intraprofessional disputes (Levy 1994). the ne~a: been the hallmark of American anthropology. Some academic anthropologists asserted that their research was "pure. h . Ethics educanon ISc ..modification ofthe former first .sof d so that an ongoing discussion of ethIcal pre been applied. and included AAA members from archaeology and physical anthropology.era lOn 0 .out bel?g.' Professional ArchaeologIsts. they argue.or a spe . the AAA decided to reexamine systematically the underlying principles and practice of professional ethics within the discipline. .' of humans and nonhuman pri' . and to ~onserv~ t?e archaeolO. W le . And applied anthropologists may have experienced some isolation from the traditionally academically oriented discipline as they. e fi st time the new code of the AAA speaks A major development IS that'hfOr}h _fireld def:nition of American anthropology. facts. not a single anthropologist has been censured since this function of the COE was adopted in 1976.. other weak· nesses in the PPR were revealed. code of ethics makes no distinction whatsoever between so-called pure and so-called applied research. anthropologists conductmg pur~ r~e~ tha~ fact) were not constrain~d by thelf h n ~~ revised code extinguishes anthropologists conductmg propnetary researc are. then executive director of AAA. by 1995.. I was a member of this commission. and personal differences. period. 'lOates.'t ~ of the anthropologist to the people . After 2 years. ~~R Thus the anthropologist is now :. e l ___1 . draft code." This is cies. "(-Anthropological ~esearchers ave pn~~e eo le with whom they work. polOglSts contract t elf servtC b ' t ntly criticized in prevIous amount of funding. s we confront m t e lour . As . employment. es conflicting interests. What has happened was that the COE was primarily caIled on." whereas applied research is compromised by the client-researcherrelationship. Practicing anthropologists have contributed ideas and their perspective to the new code. •• integratIve. NAPA judged thIS phrasmg to e tofa SI rch fr'equently affecting indi. and the Arc ~e ~'d harm" is a basic principle The general admonition in th~ new co . as the job market has improved somewhat.th dIverse and somen~ b rved but the individual pracof potential conflict. However.rule can . . has gradually evolved to concerns 0 c c'. sangction individual anthropologlS2ts5to one tthhaatt tmh: PPR had been in effect no er the nearly years d t d the "teeth" of the professional code ha no entioned above.fied period of time and . research is research. hole . The procedure included receiving and screening cases. Call for Revision of the PPR. ov thropologist had been ce~sure .emg c' enjoined to do no arm. •It also reflects a reac 10 Id' sponse to the realization or complamt '. For example. err~ elops and adapts to changing circum. and Issues on w l l l l l l . the PPR was revised and a new code of ethics was drafted (see Fluehr-Lobban 1996). gists.Stances for anthropological research. with the Executive Board being responsible for the final judgment and any action taken. past. In such cases viduals and groupS w. Id'n many types 0 resea . A Commission to Review the AAA Statements on Ethics was formed in 1994 and continued its work through 1997. es to an agency . a cornerstone pnnclple 0 t e l l b .fts from. Clandestine research has een C~~S1: ~een an area where applied codes and statements on ethics. . A Code for All of Anthropology Beyond the tension between academic and practicing anthropologists. a complex grievance procedure was provided for the AAA Committee on Ethics (COE)." take place as the prolesslon cv IssUes can . ha has a rievance procedure with ability e is that the code sh.e~o~~~. time after time. pioneered new venues for applying and using anthropological research. Academicanthropologists may have not have fully understood or were skeptical about the types of contract-based research with which practicing anthropologists were involved.'. then president of AAA. the Soc~ety for Applied ~n~l:~~a~~~~titule of America. The present revised. when jobs for anthropologists were scarce. and chaired by James Peacock. These tensions have eased at present.earc II a es to be aware of possible and practicing anthropologists have urged thel~ co eag~ and the people studied. as weIl as applied and cultural anthropologists who have traditionally been at the center of ethics discourse.groups. the:e wereh I the NAPA. and be prepared to make titioner must make carefully con~ldered ethlc~~h those choices are based (Jordan clear the assumptions. the Society of • AAA (the PPR). In 1994. . It was initiated by John Comman. hollsnc approach that I' c t' f codes of different anthropologIcal .e 0 c~~ice. no absolute .

nst.(t) t d' d? If the research is being the source(s) of my funding to the pe~ple/group s. are openness and disclosure. . . or any field.~~~ or herself. rch? and their potential regulatory role " y 'mpose upon my resea about the conditions that ~ert~1n fundmg so~rcesh~al ~ould be willing to disclose Have I answered the questIOn In the affirmauve. 'the stcolonIa . h will be put? Have resO v .::. The new code does call for the open and accessible publication or. and source(s) of support for research projects with funders. The new code contains specific language regarding informed consent.d lines prace ures 0 Cl a. Have I considered various sources of fUnd~ngappropr~~~. m lexities and sensitivities of conducting reseSl". to eventual publication of results. in anthropological fieldwork. in. The present draft code of the AAA is clear and unambiguous on this fundamental idea: "In both proposing and carrying out research. from initial design. a nece.- Potential Research Projects • ..'% be studied~s:~~~:~~a:i~fsc~:s:~:: ~i:~:~:'~~::. have I Iscusse WI 11 b art of the public the research results will remain private. they should be cautious about the selective use of their data for whatever good cause is being advocated. to pursuit of funding.' rations. potentia 'I con ft't etween ec' nto my own satisfaction. ' . Ethics should be considered first. "might have been the advice of anthropologist-mentors in pa. .L Ethics 180 FLUEHR-LOBBAN /81 " L L L I L L L L I distinctions between pure and applied research-quite simply. . Finally. not last. 'n is slated for very funded bv a nongovemmental orgamzauon.~. b~t I you as fa ood deal of satisfactorily before you begin your research.~. you WIll save yoursel g . the new code addresses the issue of advocacy in anthropology and whether this is an ethical responsibility. this absolutist position untenable today. irrespective of private or public funding.~~~~::~~s ~i~:~:~ some groupS. "Just come up with a good idea Using Data from Research . 'f k and answer them The questions raised here are not SImple ones. h Id h' k bout the following: Have I In the last phase the anthropologISt s ou t In a I I ed ' .iJ~~t~ons ~re sufficiently serious. andtropologists and mentors. th logist should ask himFrom the initial concept of a rtsearch ~~~J~~~ ~:~:ctro::. or If the fundI g h' h . which means that anthropology has situated itself within the broader scientific and professional community of scholars and scientists in the United States and increasingly in the world. . The basic guidelines for the ethical conduct of research in anthropology. which might be negotiated in advance.h 10 . like to think that they have complete freedom in the conduct of their research. through actual research.accommodat:the legitima~e. I a~quall'nte~ mysbelafrwdsl?thl~:~i:~:::~e.t~~~~~=:~. Thinking about ethics should be a part of every research plan from beginning to end. It was agreed that advocacy is very mucb a moral choice that may be made by individual anthropologists. s u le . There are many regulations that students and researchers become aware of only when conducting research. WIt mem e d if ossible? Have '. . research results within a reasonable period of time. anthropologists must be open about the purpose(s).e following. in order to ensure that secret research is not sanctioned. . ~olweverW'tehseteCmo:arid not to mention studying American populauons.h b rs of the ethics committee. .hout violating ethical or herself the followi~g: Can ~ ca~al standards? If there are questions about standards or compromlsmg. . but it is not a professional dul)'.non. persons studied or prOViding information. The informed consent language is intended to encompass both the ethical and legal intent of this broad principle.'tution with some representatives of the group I plan todstu Y'f t?b I ethn'.s! d go and study It. le s 0f'terest In the people-culture researched?. research is researcb" and standards of ethical conduct in research are the same. and with relevant parties affected by the research" (Draft Code 1995:AA1).~I thought over the conduct 0 my resear .. of my consider abandoning the project? Sources of Funding In the next phase. and probably other researchers. or eventua Y ecome a P d t t? domain? Have I satisfied myself that the research is proprietary an no secre ' ~ L L 1 L L Consideration of Ethics at Every Stage of Research and Publication Anthropologists. d' d 'th the funder the degree to w 'c specific proprietary research. but a little advance preparation might allow them to make more informed and better decisions. colleagues. When anthropologists choose advocacy. the anthropologist should ask th.i?terests of the An Ethically Conscious Model of Anthropological Research The foregoing discussion of the history of ethical discourse and concern within anthropology should demonstrate to the relative newcomer to the field that anthropologists and their professional associations take ethics seriously. should I naUona reVIew 0 =~. questio~sr~~.e~.c or . This will be discussed in detail below. considered the potential use(s) to whIch my researc b th \'e t funder and .pro ess~o American ri hts to cultural and material this-fO~::. potential impacts.

A parallel tendency sees anthropological research as unique and therefore not related 10 ideas or ethical practices that developed outside of our discipline. In secret research.182 FLUEHR-LOBBAN Ethics anxiety and discomfort in the field. It often moves 10 a personal.. p~sent revisions. Ideally in this situation. Most people. Some anthropologic~1 research may be evaluat~d as bei?g of minimal risk to research participants. hut is research conducted for the client en.." where l l 1 l " I l 1 l l l 1 1 1 . m nonetheless responsible to various I'stakeholders~ who may be representatives of . us' h '? Research Review Boards eVlew Board~ (IRBs) that are charged with the responsibility to review all research 'd l' proposals that mvolve the conduct of research with human sub'ects ) Th we . \I Proprietary and Secret Research Students and resea~c~ers may be. A large amount of anthropological research today is carried out in the United Slates. Perhaps because our research is comparative and most often conducted outside of the United States. Physical anthropology researc. • matl?n to make informed choices. operate m good faith when they have the necessary' f.e openness and dis~losure weren'r practiced. Little anthropological research remains solely in the public domain. If you're open with uinforrnants. There has been some resistance by anthropologists to being called before review boards. Again.ry research is normally negotiated be~een client and researcher for a ~peclfic proJect'"length ..ormaIlY ~ot owned by the researcher._senl.. depending on the experiences of board members. Anthropologists would do well to consult with their IRB and pethaps open a dialogue with members about the nature and extent of federal and institutional regulation of their proposed research. ThIS IS an especially sensItIve subject for anthropology since it was allegatJOns that secret research had been conducted in Southeast Asia that led to the cnSls that produced the first AAA code of ethics P~oprieta. Researchers from outside the United States should acquaint themselves with comparahle national or institutional review boards. such as "Save the Children" ~~ access to the published results of the research is not restricted. today much anthropological research is conducted in an interdisciplinary environment where ethical standards must be unifonn.. openness and disclosure of research intentions funding and outcom~ of research is normally not practiced with the people studied FU~her pUbl~catJOn of research results is usually restricted and not available to th'e g . irre:pec~~ . digenous peoples. 11 bl'sh J • e gUl e mes re ongma yesta I cd for research in the biomedical fields but anthropologists are also. This view mistakenly adopts an exceptionalist view of anthropological research as being fundamentally different from other kinds of research. cl?ser relationship with human beings. The present code s~s:g"The researchers must ~a~e the mtentlOn and expectation to disseminate publicly results ..~. bound by these regulations.confused by the difference between proprietary and secret ~esearch. Professional codes of ethics with:~~~ AAA. state or national historic commissions.•. public through usual avenues of distribution. ~II ~nlversities and colleges are mandated by federal guidelines to have Institutional . These guidelines have been enacted since the publication of the last general Handbook of Research Method in Cultural Anthropology (Naroll and Cohen 1973). Indeed. from! 971 to th. that is free of ethical responsibility. or anthropologists. such as the observation of public behavlOr.Ol you giveJ~:r~ the right to say no to your requests for infonnation. have stood firmly and consistentl a ains~ anthropologISts engagmg secr~tresearch ofany kind. " :ch of whom has a special and sometimes conflicting interest with the other. . practice of anthropology. class)~ed as a type of biomedical research requiring infonned . though they work primarily with material culture. My own experience as a member and later chair of my home institution's Committee on Research with Human Participants is that review hoards are userfrienlil}' places that act more in an advisory capacity than as "courts of law... and/or their sponsors. thereby having what we may view as our special field methods subjected 10 sctutiny. are of cultural difference. propnetary research is conducted for nongovernmental organizations 5 as the World Bank. and federal guidelines apply because the anthropologists are recipients of federal funding. and the tights of research participants don't vary whether they're studied by psychologists. . Archaeologists. be.)1'ay. Review boards themselves vary considerably. There is no subfield of anthropology. . IRBs may be sensitive or insensitive to issues in cross-cultural research. cardiologists. or nonprofit groups. research is research. an initial infonnal inquiry might make the fonnal request go more smoothly.. requiring explanations of why anthropological research does not lend itself to the use of consent fonns. or a like one for cultural research.. of the research wlthm a reasonable period of time" (Draft Revised Code 1995). ua y. Such questions and answers might also b h proper doses of prevention that avoid disaster when communication and tru t : t ~ /83 down becaus. the anthr?pologlSt can :" play a constructive role as broker.?f research. Some may take a strict or narrow interpretation of infonned consent. The researcher might inquire how lhe research proposal would best address the expected standards of research practice while bringing up some of the particular cultural or linguistic issues that are a usual part of anthropological research. Nor do the responsibilities of the researchers vary substantially. Proprietary research ~. we tend to believe that the rules don't apply to us. negotiating and achieving compromIse among the various interested parties. nor any part ofthe i. and terms of publication.

' guiding spmt IS that of openn~ss h "t f 'nformed consent means that the • in social science t :ikely research outcome with the me researcher actually dlscu~ses t eh t ~. and disclosure in research practice. however. ' not one based on the fear of alleged misconduct or wrongdoing. ~ ed consent requirements have been form" summary document ensunng that m o.lflly.m:' I risk is demonstrated or where met. redesigned. Ideally. in medICal researc . including personal identity information.. rather than the mec ams le ~~p .muc 0 • f informed consent and its ' h' . d b the subject or au onz consent document. . h' h might be impacted by the studied or otherwise identified as haVing m~res~~t .L L L L L /84 FLUEHR-LOBBAN Ethics . Don't be shy or intimidated about educating members oflRBs. and do negotiate issues or areas of possible disagreement with committee members. and eventually approved. (3) modification of I and 2. L L L Low-Risk. researchers are given a thumbs up or down approval or rejection. or a reciate its relevance to their own . As such. provl 109 Jnlonna ' . The legal doctrine of informed consent grew out of the post-World War n world that was shaken 10 its moral and ethical core. improved.~ purposes.r:formel prac~c~\o ~~:~u~e acro~s t~e res~arch. By extension. Unobtrusive Research Much of anthropological or social scientific research may appear to be of low or minimal risk. . 'd" l' tion ownmg or con ro bemg studied. Ust. or their equivalents. federal regulations and institutional review may be exempt in cases of low or nil risk to those studied. In administering tests.e mlt~::b' eclives of research. interviews. to e slgne .ence to informed The present revised code of the AAA ma es slgm I . L L The first two might be considered unobtrusive research. American Anthropologlca Assocla IOn. research using low-risk methods. the researcher than the research part~clpatt. (2) a "short . such as using questionnaires or interviews. . ro riate to our methods and informed consent. h informed consen~ opens up a tw~.. . whether the research is biomedical. 11' S may vIew mforme consen m < I I lif research.nt~rm !ication of a form designed more to protect consent.~ notonlybytherevel~lIonofN~zl eth~cs it scientificlspe19ct97~h:plt~:c~yp~:~I. The President's Commission for the Study of Ethical Problems in Medical. where no observable risk to those studied can be shown. about the nature ofanthropological research and the methods we use. sd develop an open relationship. f the following ways: (I) written informed consent must be demonstrated m ~ne 0 th' ed agent. advance the informed consent of persons Anthropological researchers must obtam I~ t 11'lng access to material being . ' I ' t' ' bri ge . h the documentation of I d . that sponsor researc . /85 I.'. such 85 questionnaires. 3. uSI~g ~hsp~n a~d 2. pro~ortionately to the dimimshed informed c~nse~t r~~~~~~~:. Federal RegISter 1994). :.t informed consent has expan e .~ formed consent will depend research. This has become a virtual canen of ethics and research in all fields. with insignificant or no potential hann to participants. duress. . thus the studler an t e s u le Iof communication that. j case the responsibility of the external revIew com~1 t L I Informed Consent One of the areas where there may be disagreement about research methods and ethics is the subject of informed consent. In the latter case. .ay ~t~nnn:nd ideas. deceIt. Further It IS consent for the first time. a research design that is flawed or insufficiently addresses issues of ethics can be discussed. y . anthropologists should not exempt themselves from practicing full disclosure simply because they are using low-risk methods. . or social scientific (see Fluehr-Lobban 1994). It is understood t~at the degree an~ . or tests in which agreement to participate effectively constitutes Consent. negotiated. whe. or 0 e (Rules and Regulations.' Few anthropologls~s know thlSdhlstory t' :::'ther mechanistic fashion because. This is the spirit of infonned allows for a contmuous floWhOf I. argue. and anthropologists should instead look for venues for research outside of the classroom. Biomedical and Behavioral Research (1980-83) adopted the view that consent need not be obtained for research characterized by the following conditions: I.. combined with behavlOr :ha~~\:~pwePresearch (Fluehr-Lobban our traditionally close relationshIp wl~h th :." 0" 'Ily developed for medical research. ome. review of publicly available information. havin rovided the hIstone scientific-social c"onsent was added to ~ 'd between the two. behavioral. n .:.ng consent 0 ~n ~oice with~ut undue inducement or representative. b . : the new code of of the . all part of taking a proactive approach to ethics.. 1~ IS U . has a coercive dimension. nfortunate because the geneSIS 0 . the observation of behavior on public places where questions of privacy do not exist. in my experience. once opened. able to exercise fr~e power 0 C th r fonn of constraint or coercion" any element of force. ) immoral and unethIcal research with humans . . Using students to fill out questionnaires within the classroom setting. fraud. ~ cement informed consent has 1994). for example. It is worth excerp f109 here'. . the potentially coercive relationship between test-giver and test-taker should be recognized as one where a power differential exists. . In the latter 'ttee or IRB increases obtaining informed consent would mvalIdate .or ega 'i frankly.re:nder~t~:d that the informed consent upon the nature of the proJect. For purposes of fed~ral regulatIOnfan ~:d~~idual ~r a legally authorized been defined as "the knowl. . . atrocities but by the collaboration ofscience with . d for the acceptance of the spirit of In an article in Human Orgamzatlon. h h f it has devolved to signing a form .. For federally regu ate agencIes. d participant. The dialogue is what is important.

~ particular written or signed form.occurring in. course. in Egypt (1982-84). n anguage. understanding that certain . such as the rise of Islamist movements in Egypt. and I may have exerted a double standard in this respect. p e were SUSpiCIOUS as to why two Amertcan . or in develo . Informed consent does 0 . irrespective of culture or e. The contempora r tor. and full losure of who we were and what we intended to do during our 18 months in the . this selected review of my fieldwork experiences is not without its i I i I will draw on my own 6 years of a thr . afforded by informed consent as are A~ nts are . I .< conflicts. spanning 25 years research when the first code of anth. ho~e that the answer would be protections..(CCt who preceded awareness of the importance of theoretical and practical ethics.hroP?IOgy have drawn on the or Westem concept that cannot be fully e I' gd' that mformed consent is a U. well-educated families. few anthropologists :'ve it is FoundatIOn or National Institute of Health) Al y the National Sctence ~iewed their own research as relatively bent~ as n~ted. no know of the benefits of an open m t '1 xi"~tenc~ an~hropological researchers Some even think in terms of a "c~ve~a:~ l~~ a:to~ShIP With research participants. In the Sudan in 1970 peo I e po t~ICS of research were my training . It is the quality or"ht necessarily Imply or require IS relevant (Draft Code 1995:8. 0 an rug abuse Questions of the universality of I'ncorm dPment-related projects overseas. tS not a W estem concept Although cultures va . filience. but I fear that there may b na.tcans. . they believed th y fd' posmg httle ~lSk to research However.. many anthropologists mformants. uropean SOCIal anthropologists .. I struggled with the ethical dilemma of being open with national research boards.•llowing a relationship to unfold that may even permit a participant's withdrawal . . and Tunisia during later stages of my research career.o 01' \.ac~. Sometimes they withheld information.otherwords. admittedly. honorable Americans. owever. orkmg With this historical tie to U. to allow negotiation of terms of iresearch as it progresses through its various stages. Arguments against using informed ~slate across languages and cultures?. or resolving the dilemma by not conducting research.)ta~ :. once people understood the nature of my inquiry into some.. lii' For my generation of anthropologists trained in the 19605 and 1970s. not CIA agents. With the shift from an adjudicative code to an educatiOnal one. I became increasingly cognizant of highly sensitive research subject matter. ~U In retrospect. a double standard.i.3). without the mecha . about one's research methods go I xP alOe m cross-cultural research. e a t s pomt. postwar histo . In up Its cultural and legal history in the United State:~consent m a~thropology brmg may be reluctant to apply inCormed . openness. Ion. "consent as an ethical I I of Its percetved specific connection to th U 't d S or ega concept because informed consent. . earn wor mg m U S al h I d d research proJects. l l P~bl~:~~:l comple~' ~o . with research participants.'.less . and cultural sensibilities about research and privacy issues. or I suspect. they asked that certain information not be . training We received in graduate school was inadequate to carry out informed and . The CIA .:. ar~ enlItled to the same rights and Does the concept of informed conse~/~:racltce. I decided not to carry out the research on the subject and to postpone .. In another vein.etr m. I am acutely aware that higher-class families were more 'iJ" sensitive about protecting their respectability and privacy than were lower-class. 'only antidote to this public perception was honesty. communtty-based projects with is often a part of a multidisciplinary try rea Itykl~ th~t the anthropological researcher . and research funding and ~. and openness were beneficial as we made friends with the members of the 'reSe board.. In my j lIeldwork.We waited six months for our visas in the United States and three more months ""pennission by the National Research Board to conduct research. for example.:sensitive family law cases and issues. the ~. and sources of fundi h . C " e consent can be pos d t th' . I came to uitderstand that being open about research is a way to keep open the lines of com~aiUnication throughout the course of research. I 'it. Informed consent doesn't "'lRJlSlate literally and directly into Arabic. and. try." Individuals or families might be compromised by public knowledge oftheir personal . " ~ During subsequent research in the Sudan (1979-80). . J. th' culture endorses dishonesty or deceptionryEm . and those '. the hope is that professional ethics and active discussion of ethical l l l .. in and '[\Inisia (1990). perhaps. Bamesc(~~. we developed greater familiarity with the language. Honesty ' a s.des of communication. many other languages.. A.~. not the format. scale (in. t e consent. protected Americans.e altonshtp that anthropologists have :e:~~~Je~~tonbeca~se ~maI hers had come all that way to leam Arabic and study their culture. with their research collaborators (Wax 19~5.from research is more ethical and probably results in better research. that Much anthropological research has esca ed ~ d I . ry. In the end. But tmaking honest attempts to talk about the nature. leamed basic colloquial Arabic. I began my doctoral heated debates among anthropologists ~ve~~~cal e~?~cs was promulgated... . much contemporary anthro ot ~ou protect their subjects from harm areas of applied social science and in research is . The waiting. ' umversa Iy recogn'z d . t' e nt e tates. tscusSlon. we may legItimately ask .Ethics /86 FLUEHR-LOBBAN /87 process may be dynamic. pnnclple that transcends culture a d l i e as a slgmficant ethical W .' mentioned as the agent behind us... the intent of of its application is becoming more m~ le US Cl of forms and legalistic implications .S.ntltled to the same protections that humans. disguising my interest in fundamentalist Islam with other research objectives. and those ground. Sudan.. and three different African countries to ~nfo:Ot~~\~~1 research. . ' many who oppose the application of inform . own dose of self-criticism. used or pUblished. ng. or altogether avoid research in the country I had chosen.e. rather than an American university fellowand my husband (a fellow anthropologist) and I felt awkward and untrusted. l l l l -. whether non-American research partici a . and word spread that we were arch 'iClually decent. '.S. However.." balanced ethical decision making. cultural relativist theme.

Inforr:e~' c ecomes rn~re ~f a . manner IS that informed consent bec I an open. nor abuse the confidenc:~e tWh b S? as not to violate the trust A' a as een gIven to you n unconscIOus or unspoken paternalism m . with intimate g people. that they did not understand fully th . I fear that info~~ca or cultural anthropology requires. n sOCJa' or cultural anthropology.be developed from DNA samples taken by researchers. ~ es 1989) concluded that the informed researcher regarded resistance to :~I~~n et~een researcher.that retreating from openness may result in some form of deception. ~ However. /88 FLUEHR-LOBBAN Ethics dilemmas will becom .. Con llctmg research in . agency or home institution to Use I'n~orm d y m~y be reqUIred by thClf funding · J • " e consent .protection for the rather than bilateral and protects th onshent obtamed m this way is unilateral e researe er against eha fi . relIective of the norms of American life. the application of the ethical and legal principle of informed consent has been . First. public meetings of potential participants are scheduled. askmg for a sign?ture on a ance of the form may be intimidatin A y not be I~terate or the official appear. You can obtain informed ralSmg re evant issues that inform and thereby empower the participant.arms • MoSt. consent form. I suspect that the proper communication amounting to informed consent was lacking in the celebrated case where "theft" or appropriation of human genetic material was alleged.. . must carefully consider the ethical implications of every phase of our research project. collaborative . ~nd researched. using a form or some verbal form I be consent.. through the process of . rges rom partIcIpants e mtent or outcome of h l n a research relationship with non W r~searc . when mechanically applied researcher than the researched. r su ~ects' are not . . J informed consent forms. k psychologISt conducting cross'cultural research with Guatemalan worn consent form had become a ben. Much of the history of informed consent has been linked to protecting the rights of individuals. but as producing better research results.I'f not aJJ ..receiving approval to carry out their projects. Controversy over ethical and legal issues arising from the Human Genome ~Diversity Project has centered on the potential abuse of relationships between i\powerful Western scientific bodies and relatively powerless indigenous populations . and indigenous peoples of America.S.ng forms. Some potential researchers may feel uncomfortable with the degree of openness that is being discussed here. local language. The value of d' . anthropologists and non-Western scholars have been the leading spokespersons asserting the difference between Western and other cultures in the matter of the greater value placed on collective rights in non-Western societies. This should not present any insuunountable obstacle to obtaining informed consent. serva Ion wou ld miti t '. under historical 'tions of actual colonialism or perceived colonial-like agencies like the erican Bureau of Indian Affairs. including the activities of jeSe8fChers. necessarily passive. solution is to engage in a open dialogue of the risks and benefits of the proposed research with the research participants themseives and. . which powerful instrument that you must OCcur m translatIOn.. where opposing viewpoints can be openly expressed. . The g assertion of control over the res gh e ~orm as a POSItive demonstration of . k~Pt some of these issues the "My Tribe" syndrome where anthrca I lS~ourse: ThIS IS what used to be called was best f<. In some cases. opo oglSts might have felt that they did what peop e.>r "their" 'I I . indeed postcolonlal world. anthropologists. for negotiations with representative and responsible agents of tribal and ethnic groups can be combined with individual consent..: openness and disclosure amount to informed consent. e an m egra part of undergraduate and graduate education' /89 Informed Consent Without Forms' Acknowledging the Risk of Paternalism Almost invariably. t I anthropology. are actively restructuring their relationships with states and I manner of external institutions that impact their lives.(about who owns and controls human genetic materials. Most V.. The person being stUdi~d ma awkward. researchers should be aware that the ga e .medIcal_ • b10 ogleaI research that interfaces with h .The era of colonialism has passed. the Western researcher ma ieelestem. paternalism was not uncommon and may have ctenzedhuman relations between anthropologist and subject (Asad 1973:16). in effect an official airing of the proposal where the required . The obvious •'c. However. m uch ofanthropological researcb. and Canadian Tribal Councils have autonomous research teview boards to which anthropologists and other researchers must apply before . Typically. The purposes of this rather lengthy discussion of informed consent are twofold. it is also true that research ~e oogm as ~ res~t of long·term residence with conveys a greater intimacy than woulden con . to engage with them as collaborators. I L i (f' U L I I L It is true that anthropological research can b h' relations of confidence and trust de I ' e Ighly personal. ideally. ensuring active community involvement and monitoring.ucte in th. However. These can and should be regularized as an ongoing part of the course of research. when the subject of informed c .'lequesting permission to conduct research. That intimacy is a . omes a natura part of th d I research project and relationship with th e eve opment ef the consent without using forms by " os~ you study.agamst uSl. which .elatlvely powerless par. This is a clear and positive trend in anthropological research.L L L L L i i'. often r. utilizing methods of particl'pant ob 't' eed. Including cell lines that can . but we must acknowledge .is increasingly recognized as constituting not only better ethics in research. the terms and conditions of the research if' plan are negotiated. This is why we . where large community·based studies are proposed and negotiated. eare environment Du b' .+. ticipants. the first objection they raise is the u °Fsent IS brought up with informed consent does not require forms Ind se 0 a consent form. from being fully aired in anthropologi I d' ay have.th care established.

or plagi~rism in th\P~:~~ss.. Therefore..:~~r. d . . and a host of other roles and statuses (see Appendix to this chapter [po 195]). or funding With the persons/c om an fabrication of evidence. the spirit of informed consent has a certain potency as a summary concept for research. ' t AIDS. 2. Whether as acade~lc researchfers'myer~la~ 0:r~:: ':". charged with reviewing the statements on ethics.~t:.° Y d publication of the results of research. I 1 . as a member of a family. and rofessional responslbllllles P . a community. '.searc . affected by our research are treated in the ~3. . and materials they study and to the people with whom they work" (Final Draft. t mationaI dtSCUSSlon 0 C George Ulrich ofthe Univemty of Copenhagen. Anthropological research can place the researcher/scholar in complex situations. inclu~ing the socl~1 . means openness and . degrees. development projec related studies in cross-cultural research.. the discourse about ethics in anthropology. Anthropology Newsletter <'." The modified version said.o~~ Soci~l Anthropologists in Harcelonato diSCUsst~he r~ce:~t~~~~~~. modified the original code. Acting impartl~ny. In the context ofthis assessment. scholar. l' a d as a necessary adjunct to SCience an :I%~i~~~:~.:t tion . . In the Principles of Professional Responsibilities. Infonned consent..~ I principles that should guide all research.. An of ethics and to consider mtemahonal perspec IVes.kers In governmental or nongovemm 'd' h' eth'lcal choices and . responsibility to those into whose confidence you were taken. :s::~~ i~o:Pe~~:::i~tt~. relatively late for the biomedical. t of research face I' 1 . . it results in better researchers and better research.~~:~ I 19: I I "common challenges and similar responSibilities when :he y cond d \ t . partlclpa 10~ In I f conthe study of ethnic minorities in nation-states are just a few ~xa~p es 0 d .led °fu aseYs .~" '~ Anthropologists irrespective of national oflgm or coun ry h: our . disclosure with participants. in its fullest interpretation.~~~~~d~:t:. the anthropologist's "first responsibility" was to the "people studied. f hw'thin funding agencies possibly negotiation ofthe terms and conditIOns 0 researc I d f the 'nation-states where the research is to be carried ~ut.. In e I ~:~oi~~ited to the 4th Biennial Conference of the Europea. for example. r. To do no harm or wrong. IOS0 a~ a~ AAA Draft Code of Ethics. t s. .Ethics /90 FLUEHR-LOBBAN /9/ a recent development.'. ' • • . falsification of ~. and to one's discipline.ptlOnorknowmgm m~nities studied is a sine qua non of social :. In ever more c . l~ d' ' of ethics and international Ethics Network was created to faCIlitate ISCU. each with its own set of contextual obligations. to science. I ~ hers need to conSI er t elf applications of anthropo ogy..re~a.uch t t~iS is ~ossible under the conditions of research. duration. :. The informed and ethically conscious researcher will recognize the multiple layers of responsibility that can obtain in research. t' .. ~ :.an~ ethics.:: SOCla r . Weighing the kinds.~s~~. . These are genera h i ' t Studying these principles and dlscussmg thelf 1996) Concluding Remarks The Preamble to the revised AAA code of ethics states that the anthropological researcher. nine ddition to broadening and comparison of research expeflenceswould be a welcome a . e about ethics and research extends to the intemational a. or teacher is a member of many different communities. • uct researc m anthropological rese. A .. . the AAA Commission. the anthropologist must be guided by the following general ethical principles of conduct: I.~e~~~~:t~~:::.SlO. ~ how anthropologists make ethical decisions in research IS belng carr.~~~t~~~c:t~tc~c:~ education and tramlDg ID anthropology. isre resentationofone'sresearchgoa1s .'~ . h' I d'l mmas and require dialogue-an temporary research proj'ects that raise et Ica l e .~~~~~~:o a~ur work and our profession than it is at . ~Iready. with fewer references to "infonnants" and "subjects" and more reference to "collaborators" and "participants.. natural1y..~. Code of Ethics of the American Anthropological Association. wo. physical dimensions of anthropological research. (Final ""i£: same manner. rather than hierarchical and relatively nonparticipatory.:'!r~ Report of the CommIssion.:'.n ~:s':::~o. and in conformity with recent developments in sociocultural research.a stu . t . understanding that the development of knOWledge can lead to change which may be positive or negative for some people.~~:~~~n~el~~ge~e~~~~ao:e~~~~~::~~g. "Anthropological researchers have primary ethical obligations to the people. where competing but legitimate ethical claims can arise.arch. Second. .. 1997). omplex SOCial envlronmen s.. ental agencies or one 0 a :.. and models of research that are collaborative. f the fundamental importance of ethics I Developmg awareness 0 ." When the spirit of infonned consent is implemented. responsibility to the truth.:~i~gSs~~~. and probability of goods to be gained and harms to be avoided is the task of the informed and ethically conscious researcher.~~u:nr~:::r~ho. research.': research proposa . These might include ethical responsibility to the people studied in their complex relations with one another and to their state and other communities and institutions that impact their lives. This extends. a public or private employee.els of the Unis that have anthropology across national borders at the suggesllon of ~e er h versity of Amsterdam.' . With these complexities in mind. . species. and responsibility to the client or funder of research.. irrespective of discipline or sub-: discipline. This meeting acknowledged the maJor: ange duct of taken place worldwide in the practice of anthropology and In the con . Avoidingdece. the mom~nt. "? 0. This may be reflected in a detectable change in terminology of social research. the utility of informed consent may need some explanation and justification.~. data. .methods. ever-shrinking global context. . lanedsema~~~ i:~~~da~~y~~~. .

Adopted by the councll. Social Scientists' Ethical Responsibility to Superordmates.M. M J995 Informed Consent in Applied Research: A Comment. A. Human Organization 53(1):]-9. . Tucson: University of Arizona. Galliher. 1982.. Handbook on Ethical Issues in ~n:roP~/~~. Anthropology Newsletter (April): I7-18.. Englewood Cliffs. a New Era. .. Informed Consent in Anthropological R. Bulmer. monthly column. Ethical Dilemmas (this is a good way to keep current with ethical discourse in the United States). ': I pu ' bl' I· No 23 Washington DC' American Anthropologica SSOCIB 10 . Martin. Ch. Boas. Cambridge' Cambridge University Press. . 1994.esearch: We Are Not Exempt. Federai Register 59(96):26116-26119. Government PnntlOg Ice. in collaboration with Nancy . Problems 27(3):298-308. . March I. . 1979. . 1996. p. New York: Columbia University Press. MA' Crossroads Press. " S Chambers. Report to the American Anthropological Association. Fieldwork Ethics in Policy-Oriented Reseatch. Universal Values and Professional Codes of Ethics. · ts 1990 Part 1028 Code of Federol Regulotions. The most fee Code of Ethics is reprinted here as an Appendix by permission of the Ameri Anthropological Association. . 1980. 1973. mants" as more egalitarian ideals of the relationship between researcher and researched evolving. L I I ~ and Regulations. Deloria. Appell.. Naroll. Organization 54(3):330-331. Developingthe New Code of Ethics. Brinton. Barbara Frankel. ClInton. Final Report ofthe Commission to Review the AAA Statements on Ethics. Fluehr-Lobban. Talal. SOcIal Levy. lanet. 1995. L. Problems 27(3):265-271. ed. Franz. Unger. . Reprinted in Eth:CS an: t~e p.. March 1998. NY: Baywood Publishing. ~ell L L L L Fluehr-Lobban. eds.G. 8rc reprinted in my edited volum Ethics and the Profession ofAnthropology: Dialogue for a New Era (1991). Pinkard..S.~~~ ~j" 0/ Anthropology: Dialogue for a New ~ra. and Kathleen Gibson. Ruth R. 1963. "participants" or "collaborators" is gradually replacing "subjects" and "infor. ~es Ethics of the Society for Applied Anthropology. NI: Prentlce-Hall. London: IthacaPress. London: Alien and Unwin. American Anthropologist 83:626-628. Human Wax.:~~~c:nAnthropological Association. e. Beauchamp. 1973. 1986. Barnes. The Nation 109:279.Anthropology Newsleller . (April): 17-18. p. SOCIal MESA Newsletter 16:1.May1971 (as amended throug ovem er . Rhoda K. Human OrgamzQlIOn SUtement on . \lcauch'amp. Jane! Levy. Joan and Sue. . Carolyn. May 19. Philadelphia: University of Pennsylvania Press. . Anthropology Newsletter (April): 13-1 6. ed. C. REFERENCES ecuon rrot d . L. AAA Newsletter (October):3(}-31.EIIen Jacobs. 1987. Fluehr~Lobban.113'137. Anthropology and the Colonial Encounter. In Representations: Social Construcil'oni ol6eitl~ der. 167-185. C 'I fthe . History and Theory of Informed Consent. 1981. .. Historically. 1980. Human Organization 34: 197~~~~: Code of Ethics of the American Anthropological Association. Murray Wax.l. Erve. (-1-90) e . Our New Research Society: Some Warnings to SOCial SCientists. J.:~icanAnthroPOIOgicaIAssociation. . 1980. . . . Asad. New York: Oxford Umverslty ~ress. 19~3. Proposed Statement on Professional and Ethical Responsibilities. research participants have been referred to as "subjects" in biornedi and psychological research. . through the 1990 AAA revision." ( . Ethics and the Profession ofAnthropology: Dialoguefor. May 1994. 1994. . 5.1978. L L t philadelphia: University of PennsylvaOla Press. N. Who Should Know What: Social Science. and Ronald Cohen. Tern L. Privacy andEthics. Pp. . Erve. WORKS RELEVANT TO ETHICS AND CULTURAL ANTHROPOLOGY MA Newsletter. and T. urray. and Terry P. John F. A Faden. Dialogue with Guatemalan Indian Women: Critical Perspectives on Constructing Collaborative Research.Eth/coIDilemmasmAnt rop oglca nq . Adopted b~~e OU~CI 1~76) . Jordan. Chambers. 1989.~ Bernard. A. . Vine. Business and ProfeSSional Ethics 10(4):57-68. 3. 1919. . Amityville. Other members of the Commission to Review the AAA Statements on Ethics inclu lames Peacock.. Ethics 193 L I 192 FLUEHR-LOBBAN NOTES L 1. : . Final Draft. J991 a. Draft AAA Code of Ethics. ed. Trend. All of these codes.. Fluehr-Lobban. . 1976. 1994. Raoul. P: Kmg. Carolyn Fluehr-Lob an. Carelyn. ]996. Ethics ond Public Policy: An Introduction to Ethics. Professional Ethics and Anthropology: Tensions Betwee Its Academic and Applied Branches. h 01 . Ann. Fieldwork and the Law. Not for further reproduction. G. M. pUbiished in Anthropology Newsletter 39(3): I(}-II. The Anthropologist as Hired Hand. Anthropologists-sociologists have used the term ·'informants. A Handbook of Method in Cultural Anthropology. ed'. . pecl8 lca Ion. However. Human Organization " 42:367. t on Ethics: Principles of Professional Responsibility. Anthropology : Newsletter (February): I. Social Research Ethics. Review of the AAA Code of Ethics. om 1028 1 Washington DC: U. I1 of Human Sub1ec . . 2. and M.~e~. eds. ed. P oblems of Anthropological Research and Ethics. Carolyn. 1983.of I "~~. 1%7. Lykes. Social Problems 27:330-341. I l uiry· A Cose Book. Waltham.. 'C I . 1991 b. 1996. Carolyn. Correspondence: Scientists as Spies.

Spradley. London: Athlone..~:.~{or all aspects of humankind-archaeological. Federal Regulations: Ethical Issues and Social Research.i~a~~~e3 I ~~. .f lUidelines for making ethical choices in the conduct of their anthropological work. Pert~.-.. as well . Sage Pub Hcations. Freedom and Responsibility in Research: The Spring dale Case lluman rgamzatlon 17:1-2.. publishing. Ethical Dilemmas: Notes from Outside the Field A th Newsletter 34(7):5--Q. .4. 1 Rynkiewich. 1994.. 'The principles and guidelines in this Code provide the anthropologist with tools to engage D developing and maintaining an ethical framework for all anthropological work. and sociocultural. Susan. Cassell. r br' ou er an andon Johnson. 1994. n rapo ogy ewsletter 1 wasAhbur~. JosePh" 1971. On Appendix " Code of Ethics of the American Anthropological Association E~hics " and Anthropology. and the environment. the generation and utilization of knowledge should be achieved in an ethical manner. Qualitative Research Methods aks. the American Anthropological Association(AAA) starts from the position that generating and appropriately utiliZing knowledge (Le. ~'thropological researchers. Ethics and A Ih I . Wh~te. Ahmed. Re~~~~~~:~~I~avidson. .<. L. . 1995. An important part of that mission ~'4i:'X. Wiolliam ~. Is to help educate AAA members about ethical obligations and challenges involved in the l~I". 'yf::yAnthropology has roots in the natural and social sciences and in the humanities. ed. 1978. L 32604. ' erating an additional level of ethical considerations. 1 I 1 la a field of such complex involvements and obligations. 1?58. Margaret. Anthropological Advocacy in the Hopi-Navajo Land Dispute merrcan Anthropologist 91:738-743. Rebecca Lyon. and application. " ousan SI CipFlinary journal (formerly Bus(ness and Professional Ethics). Pels. and the students of the University of South Carolina's Eth' Anthropology class. The Ethics o[Collecting Cultural Property· Wh . CA. Anthropologists have moral Obligations as members of other groups.. past and present.y:. . Wax.1. . They also have obligations to the scholarly discipline. and James P. teaching. Ethics for the Future. and that for moral and practical reasons. Englewood Cliffs. such as the family. other species. is a worthy goal. The mission of American Anthropological Association is to advance all aspects of IrIlhropological research and to foster dissemination of anthropological knowledge through PUblications. biological. Dse Culture? Whose Property? Albuquerque: University of New Mexico Press. Cris. The dPoOlitics and Ethics ofFieldwork. '. religion. Lessons for the Field-Ethics in Fieldwork A th I N 35(5).. and J. Furthermore. AAA News/etter (SePtember):~~~ Jorgensen. Alasdalr. the profession. ameSVl e. public education. which includes the study . Messenger. National Codes of Ethics and European Anthropology: A Can for ~ooperatton and Exchange. Drlemmas lrJ Fieldwork. s e evance to t e Skomal. each with its own moral rules or codes of ethics. . dissemination. Shore. Pu~c:. . Practicing Developmental Anthro"%m. Anthropologtcal Research. and Gretel Pelto. 0 e enve from oglca le Id Prachce. 1993.The purpose of this Code is to provide AAA members and other interested persons with I 1 l l .. developing programs. The Future o[Anthropology' It R I h . NJ:. linguistic. The American Anthropological Association (AAA) does pot adjudicate claims for unethical behavior. European Association 0/ Social Anthropology Newsl it . Boulder. and to the human species. eds. 1979. Ed. Daedalus 98:361-386. teachers and practitioners are members of many different munities. 1969. Cambridge' Cambr'd OIverslty Press. 1 gc. M. and A. pro~~s~::IIE5toh1jC7s'Ga ~ulti~II' . and community. Current Anthropology 12:321-3i Preamble Maclntyre. fdworkers may develop close relationships with persons or animals with whom they work. . ed. Ethics and SOCial Science Research.'M. and utilization of anthropological knowledge. ~Anthropology is a multi disciplinary field of science and scholarship. Research with Human Beings' A M dID .ward C. eds. and the need to make choices among apparently incompatible values will "lrise. B Id dL Westvlcw Press. Phyllis Mauch. and informing policy) o( the peoples of the world.'. Westview Press. Anthropologists are responsible for grappling with such difficulties and struggling to 'fCSOlve them in ways compatible with the principles stated here. Peter. 1989. e er peltuo. 1982. conflicts. . . The purpose of this Code I$to(oster discussion and education. . ""p~. eds. Contemporary World. Wilcomb E. .194 FLUEHR-LOBBAN Ethics 195 Green.. ranging in '. teaching. to the wider society culture. that the generation of anthropological knowledge is a dynamic process using many different and ever-evolving approaches. it is inevitable that misunder~ '~dings. Michael A. .eptember):9-1O. Because elt. . n ropolo Mead. n ropo agy.erahon. . 1976. As the principal organization representing the breadth of anthropology. 'approach from basic to applied research and to scholarly interpretation.. New York: WHey. 1989. . .. 1984. d Anth ropoI ' I F' . 1995.

In. facts an~ issues' on which those choices are based. a condition to enga~e in research. Anthropological :~ea~chers working with animals must do everything in their powe~ to enSure th~t the research does not harm the safety." or "proprietary").ake c1. the pr~cess should be initiated in the project design and. yet also be alert to proper demands of good citizenshIp or host-guest relations. addition.of persons being studied. potential impacts. and source(s) of support for research projects with funders. .e. Researchers are responsible for identifying and complying with the various informed consent code~. understanding that the development of knowledge can lead to change which may be positive or negative for the people or animals worked with or studied . Anthropological researchers should obtain in advance the informed consent . the AAA Code of Ethics provides a framework. and historical records . Anthropological researchers who have developed close and en~~rin~ relatio~shiPS (i. T~e individual anthropologist must be willing to make carefully considered ethical chOIces and be prepared to make clear the assumptions.L L L L L L L I 196 FLUEHR-LOBBAN Ethics 197 anthropologists can find themselves in complex situations and subject to more than One code of ethics. 4. does not necessarily imply or require a particular written or signed form. persons studied or providing information. and can lead to decisions not to undertake or to discontinue a research project when the pri~ary obligation conflicts S. that is relevant. These obligations can supersede the goal of seeking new knOWledge." "basic. and ~. 1 Anthropological researchers must determine in adva~ce wheth. anonymity may be compromised or recognitIOn fall to materlahze. Anthropological researchers have primary ethical obligations to the people. such as those owed to sponsors or clients. and with relevant parties affected by the ~esearch. Informed consent. psychological well-being or survival of the animals or species with which they work. dignity. It is the quality of the consent. ~ " No code or set of guidelines can anticipate unique circumstances or direct actions in spec~fic situat~ons.' ought periodically to receive training on current research activities and ethical issues. t L L L L I. or private enterprises.." "pure. These guidelines therefore address general contexts priorities and relationships which should be considered in ethical decision making i~ anthropological work. species. Persons using the Code as a guideline for making ethical choices or for teaching are encouraged to seek out illustrative examples and appropriate case studies to enrich their knpwledge base.te their best efforts. providing information. Anthropologists have a duty to be informed about ethical codes relating to their work. A. and make every effort to comply with those wishes. Research In both proposing and carrying out research. or perform other professional activities..or with hosts must adhere to the obligations of openness and mformed consent. or othenvise identified as having interests w~lch might be Impacte~ by the research. laws and regulations affecting their projects.. detachment. • To consult actively with the affected individuals or group{~). Ill. depending on circumstances. Similar principles hold for ~nt~ro~ological ~es~a. public msutuuons. and . Researchers must present to t~elr re~earch participants the possible impacts ofth~ choices. • To work for the long-term conservation of the archaeological. departments offering anthropology degrees should include and require ethical' training in their curriculums. ~esear~hers must expect to utilize the results of their work in an appropriate fashion and dlssemmate the results through appropriate and timely activities. Research fulfilling these expectations is ethical. Active contribution and leadershipin seeking to shape public or private sector actions and policies may be as ethically justifiable as inaction. continue through implementation by way ofdialogue and negotiation with those studied. for making decisions. owning or con!roIli~gacces~to material being studied. covenantal relationships)with either indivldual persons ~rovldlOg mformatlon.ear that de~pl. with other responsibilities. "applied. These ethical obligations include: • To avoid harm or wrong. . not an ironclad formula. or noncooperation. With the g~al of establishing a working relationship that can be benefiCial to all parties involved 2 Anthropological researchers must do everything in their power to e~sure that their ~earch does not harm the safety. not the format.e~ their hosts/ providers of information wish to remain anonymous or receIve recogmtIo n. fossJl. and materials they study and to the people with whom they work. • To respect the well-being of humans and nonhuman pnm~tes . anthropological researchers must be open about the purpose(s). Responsibility to people and animals with whom anthropological researchers work and whose lives and cultures they study.:he~s employ:d or otherwise affiliated with nonanthropologicat mstltutlOns. or privacy of the people With whom t~ey r rk conduct research. while carefully and respectfully negotiating the limits of the relationship. It is understood that the degree and breadth ofmformed consent reqUired will depend on the nature of the project and may be affected by requireme~ts of other codes laws and ethics of the country or community in which the research IS pursued. colleagues. for the purposes of this code. regardless of the source of funding (public or private) or purpose (Le. Furth~r it i~ understood that the informed consent process is dynamic and continuous. Anthropological researchers should be alert to the danger of compromising anthropological e~~ics as.

Anthropological researchers should seriousl co s· access to their data and other research material/fi n Ider all reasonable requests for also make every effort to insure . Teachers/mentors should conduct their programs in ways that preclude discrimina· tion on ~he basis of sex. in both proposing and carrying out research. and reliable in communicating evaluations. That is.. marital status. ':'~i1e j#sponsibility to students and trainees "10:- B.als.or 0 structthe scientific/schDlarlyresearch 1. ngs 0 t e scientific and h I I 5. animals or cultur l b ' y . Cif obhgabon to reciprocate with ~. At the same time.position of advocacy.. Teachers'/mentors' duties include continually striving to improve their teaching! training techniques. A~thropological researchers should utilize the results . .~l~~~C~~ ~ateri. national origin. and helping studentsftraineeswhen they seek professional placement. encourage them to reflect upon this and other codes. or pur~osesofresearch. :o~. age. at sue Information is well underst d. k lts of their research apprDpriately d . give appropriate credit for coauthorship to students/trainees. {situations). They must avoid sexual liaisons with students/trainees for whose education and professional training they are in any way responsible. of their work 10 an appropriate fashion. . anthropologists must be open with funders. I i 3.. and must make section raising and responding to research proposal. " CISlon roa el'S and oth OlOg. groups. preservation of theu fieldwork data for use by l l C. religion. and make clear the limits of anthropol p.They should posterity.Ethics 199 198 FLUEHR-LOBBAN "__ 2. ~hich. Teachers/mentors should publicly acknowledge student/trainee assistance in research and preparation of their work. e ' la eVidencecDnduct: fals'f they I should . anthropologists may gain personal! from th . on may cause people with whom they work or V. . Anthropological researchers must ex claims and conflicts in advance whe g . encourage dialogue with colleagues on ethical issues. This is an individual decision. A n la et leal issues should be part of every /\"t adhering to ethical and legal codes governing relations between teachers/mentors and "While -"stUdents/trainees at their educational institutions or as members of wider organizations. Teachers/mentors should impress upon students/trainees the ethical challenges involved in every phase of anthropological work. I y. In so their statements but also mu~t consd o~ y responsi~le for the factual content of . Anthropologists may choose to move beyond disseminating research results to a '. Teachers/mentors should beware ofthe exploitation and serious conflicts of interest which may result if they engage in sexual relations with students/trainees. or att empttopreventreportingofmiscondu t b ' Df others. of scholarship and f ' Ity for the Integnty and reputation of sUbj~ct to the general moral ruies of~c~~~~~:~T:US. being fair. Application \. mdlvlduals. " race. their discipline.~t? encounter e~htca~ dilemmas at every pno~:ei-a~mgh~roPOsalS ~Ith efforts to Identify potential ethical and as projects proceed. sexual orientation. and whenever possible d" Issemtnate their findi t h se 0 ar y community. 5. . they must not exploit their debt to the societies in'which th:Y O~o. 4. like other l l ~thropologicalteaChers l " teaChers/mentors.. they must be truthful' they are not I ' er nonanthropologists. and recognize g1ca alert to possible harm their informatt expertise. and responsibly be candid about their qualifications and m~tr~~: h~ses upon. being available and responsive to student/trainee interests. encourage publication of worthy student/trainee papers. conscieotiouSIysup· ervising." social class. disability. y can tD preserve opportunities for 4.lc~1 or political biases. They should make clear the e . encouraging. .. assisting students/trainees in securing research support. their reports stand. coun· se li ng students/traineesrea! isticallyregarding career opportunities. and compensate studentsltrainees justly for their participation in all professional activities. Anthropological researchers should do all the future fieldworkers to follow them to the field. .disCiplin (for example. ~nthropological researchers should make the resu . l l l . 1. The same ethical guidelines apply to all anthropological work. anthropological researchers are bncate ' n sch~larly not deceive or knowingly misrepresent (i . when teaching involves close contact with students/trainees in field e . should be particularly sensitive to the ways such codes apply in their '. pror~rlY contextualized. c . they must be colleagues. ey must 0 everything in th . and supporting students'/trainees' studies. elT work. 2. and discourage participation in ethically questionable projects. Responsibility to the public 1. hut not an ethical responsibility. ethnic background. ~ I er carelully the SOCial a d I" I' o t e Inlormation they disseminate Th d n po IlIca Implications f h th h . 3. Anthropological researchers bear responsibil' . or other criteria irrelevant to academic performance. available to sponsors students de . should follow are: 2. I p . political convictions. Among the widely recognized precepts which anthropological teachers. !hey should recognize people studied in appropriate ways. ell' power to insure Oo utilized. P aglanze). Responsibility to scholarship and science stage of their work. prompt..

l . 900 1996 Principles of ArchaeologIcal. Second Street. Anthropologists may be involved in many types of work. Yolanda Moses. 300 South Jefferson Avenue.. Proactive contribution and leadership in shaping public or private sector actions and policies may be as ethically justifiable as inaction. DC 20418). and sQurce(s) of support for ~ work. 2.ameranthassn. the follOWing individuals' participated in the Commission meetings: philosopher Bernard Gert. f th Commission was pu IS e n ) D afts 'rbe Final Report 0 e bite (http'//www.dlJc1ISsed at the May within a reasonable time. . \996 annual meetings of the of the AAA Code of Ethics. Applied anthropologists. Kalhleen Gibson. Anthropology an e Vtll. Since anthropologists are members ofa variety of groups and subject to a variety of ethical codes. ers n d Code of Ethical Conduct. anthropologists bear the same responsibility to be open and candid about their skills and intentions. pose choices for which anthropologists individually and collectively bear ethical responsibility. (American Board of Behavior 41:183-186. Felix Moos. I VI. The Comm1Ss~~~. facts and issues on which those choices are based. National Association for the Practice of Anthropology 1988 Ethical Guidelines for Practitioners. depending on circumstances. In situations in which anthropological knowledge is applied. MO 65806). Applied anthropologists must intend and expect to utilize the results of thtir work appropriately (i. The Commission members were James Peacock Society for American Archaeology .e. Forensic Examiners. (ReVIse (Chair). persons hired to pursue anthropological research or apply anthropological knowledge should be honest about their qualifications.. Guidelines for the Use of Animals in Research. The Commission L L . Rather. CarolynFluehr-Lobban. Shirley Fiske. like any human actions. VII. publication. Other Relevant Codes of Ethics following list of other Codes of Ethics may be useful to anthropological researchers. frequently affecting jndi~ viduals and groups with diverse and sometimes conflicting interests. they must review the purposes ofprospective employers.• long u g . C d t· Research. . This statement does not dictate choice or propose sanctions.y the work about the purpose(s). taking into consideration the employer's past activities and future goals.96s and 1996 annual meeting edition of the 10r the Code were published 10 t e p web site. leachers and practitioners: Animal Behavior Society 1991. teaching. the second draft was briefly Am~:~~~:n:~~:o~~st ~raft 10 O~en n the Code were held at the 1995 and 0 hearmgs ical Association. detachment. 2d ed. .L L Ethics 201 200 FL UEHR-LOBBAN " colleagues.. EthICS. American Scientist 1992 Sigma Xi Statement on the 80:73-76. DC: 1995 On Being a Scientist: ResponsIble . and aims. . Sigma Xi Use of Animals in Research.• S 't 411 Springtield. American Journal of Ar~h e I ~ I Institute of America.org. it is designed to promote discussion and provide general guidelines for ethically responsible decisions. a e from the codes of et ICS O • d th Society for American Archaeology. Washmgton. -I L. David Freyer. good citizenship and guest status. . bl ' h d I the September 1995 edition of the S:me1l1bership. Washington. potential impacts. and members of the American Sociological . In all dealings with employers. 1983 Professional and Ethical ResponSibilities. r . and monitor the effects of their work on all persons affected.L L I This Code was drafted by the Commission to Review the AAA Statements on Ethics during the period January 1995-March 1997. ' iation Committee on EthiCS. Archaeological Institute of America 1991 Code of Ethics. National Academy Press (2121 ConStitutiOn Avenue. and comments were soli~ited from ~he .'AalhroPOIOgy Newsletter and on ~e ~ ~. and Niel Tashima. as any anthropologist. upp ernen replace the earlier Code of Ethics). &cussed at the Novem er b 1996 meeting of the MA Section Assembly. and relevant parties atfected-. 67~ 1994 Code of Professional Standards.:'Anthropology Newslette: ~d the ~d all comments from the membership m formulalmg . The individual anthropologist must make carefully considered ethical choices and be prepared to make clear the assumptions. and Murray Wax.was '. 3. Acknowledgments L . Prior to making any professional commitments. DC 20002 3 Society for Applied Anthropology . In addition. NE. program and policy deveJopmeJli) . UI e a 010 95:285. National Academy of Sciences . (Arc a~o O~I a ts and expands but does not Commonwealth Ave. American Board of ForenSIC EXaInlO . they should be especially careful not to promise or imply acceptance of conditions contrary to professional ethics or competing commitments. h' f the National Association for the PractIce 0 ~. . Animal . should be alert to the danger of compromising anthropological ethics as a condition for engaging in research or practice. choices must sometimes be made not only between the varied obligations presented in this code but also between those of this code and those incurred in other statuses or roles. Boston. capabilities..O~~:~ommission gratefully acknowledge the use o~ som~ '1Ile final draft m February . Suite 12. d 1983 ) . . oCle y _ 557). or noncooperation. persons studied or providing information. anthropologists Cathleen Crain. teaching. Epilogue Anthropological research. They should also be alert to proper demands of hospitality. lanetLevy.on uc In NW Washington. (S' t for American Archaeology.' ieited comments fr~~9~J1:t:Section Assembly meeting. BarbaraFrankel. and application. MA 02215-1401. In working for governmental agencies or private businesses.

This interdisciplinary space is one valued in feminist research in general. . Standards of Research p r . 1 Introduction Anthropology as a field lends itself to feminist methods. (Society of Professional Archaeologists. 1 1 1 'I I 1 1 1 l l . ox 911. there is a historical tendency in anthropology to focus on the local. Forthcommg Umted Nations Declaration on Rights of I d' . the everyday. At its best. Third. p~r ~man~oe and Institutional Standardi 73146-0911). it becomes the witness for human practices that resonate with but are not mere reflexes of. interviews. or the marginal. At its worst.202 FLUEHR-LOBBAN CHRISTlNE WARD GAlLEY • Society of Professional Archaeologists 1976 Code of Ethics. Perhaps it is this tendency that COntributed to the formative influence that anthropologists in the early 1970s had in the development of what bave become international or transnational feminisms. while remaining skeptical of unified theories based on nationalist assumptions or images of human nature rooted in particular societies. Oklahoma City. e tmlnatJon of All Forms of Discriminati~ J987 Feminist Methods Uni~ed Nat. because of the core concern with cultural dynamics and shifting terrains of meaning. the processes of giobal capitalism or earlier colonialism. at least in its North American version. worlds of meaning in the making. permitting those who wish to seek patterns to do so. the methods of feminist anthropology_its use of fieldwork. First. 0 United Nations 1948 Un~versal ~eclaration of Human Rights. 203 . surveys. the creation of meaning among those who are not powerful in the reductive sense of wealth or political control or even social prestige. Six ~ 1 1983 United Nations Convention on th El' . .. the subordinated. anthropology can ignore global structures of meaning or transform peoples into exotic objects for consumption by more powerful audiences. Taken narrowly as techniques of investigation.ions Co?vention on the Rights of the Child n Igenous Peoples. it flourishes on the borderland between the social sciences and humanities. and quantitative data Collection-aren't distinct from those of other critical methods in anthropology. Second. Anthropology can value the local. Against Women. anthropology is comparative and historical in scope.

Depending on what questions were asked. what we can assume. Leacock once did a ground survey of fur-trapping lines among the Innu of Labrador and used these data in a political-economic analysis.arate made historical trans orma Ion a Th nly "evolution" was one of chronology. (2) in ences of people and the meanmgs that ~eople at. Private Property.~y There is no uwoman"--essentla lze .:~~f. and how resistance and whatever emancipatory practices may exist are grounded in everyday practice. Feminists disagree whether to our ~nderstandmg of SOCial and ~u tura ~~t there is agreement that gender is gender IS always constructed as a hIerarchy. dynamics of social hierarchies. and various types of surveys.IS~: Marxist tradition.a who e 's em stitules "women" is complicated by concerns dimension of analYSIS. but what co. . .. the United States and in other countrIes. Her work aS. shows a unity of concern and a diversity of techniques of inqUiry.:~rious 7"Y disservice differ in how we conceptualize power and w ~t er we s t~ctures or as more "located" in current parlance-in social relallons and s textual analysis. r L :L Leacock's overarching concern-linking a wide range of ethnohistorical and ethnographic accounts and analyses of social and cultural theory-was with the herWith texts. . Her research rellected the various terrains in which these struggles ensued: serving Innu activists in Labrador Who were trying to stop NATO overflights that disturbed animal life..L L L L L L L L 204 GAlLEY Feminist Methods Even the coupling of theory and method found in this paper. a keen sense careful reading in the social sciences. Leacock and Nash 1977). city. archival research. epistemological and ontological debates within feminism-what we want to know. One 'c' amorphously associated with conceptual schemas a~d la~~u~g~i~:r~~~e:' feminist us see these as dialectically related. Despite suc~. Ime e . and inist and antiraclst. an principle that shapes the reproduction of culture and. At the suggestion of Gene Weltfish.ur that: (I) power t ~b~. but she did not flinch from criticizing tho~e for whom she advocated.I~ and historical research.ssively defined by others-in not destiny Leacock. Leacock used participant observation. cla~s'i. direct observation.an d rat:e ~ ss and p.und~d~nr~~17~::a. gro. let us ground ourselves in one feminist anthropologist's work. her international anthropology colleagu~s. I . . What '" ofdynamics and structure.d er is a key dimension of The bedrock assumption in femmlst der or that subfa sume it without in. We shall see that there is no one feminist method in anthropology (or in any other discipline). Instead. Feminist anthropology. always does she sltua e e l . inclUding gender hierarchies. lion of cultural practices. r e Some of . to see how feminism infonns her range of writing and activism. What is distinctive how theory and method are related: The bridges between the two are ontology (the set ofassumptions that underlies one's research design) and epistemology (how one knows what one thinks one knows. the basis for knowledge claims). then. generatIOn. generally associated WIt I eren la s h inherently shapes the kinds Feminists in anthropology share a concern t at power. stretching from 1954-1985. while typical of feminist methods in anthropology. Feminist Ontology . we can consider the distinctive features of feminist methodologies in anthropology. WI e \ .tl~. '.:~~~~ose experiences. Leacock collaborated in producing anthologies on women and colonization (Etienne and Leacock 1980) and women's work viewed through place and time (Leacock and Safa 1986). lbat were shaped by human agency. EJearior Leacock's work. .~dle'~st. sO~la ac.fem­ undergraduate~dg . pointing out the limitations of structuralist formUlations regarding gender (Leacock 1977a. 1983). and the methodological and epistemological problems associated with assertions of universal male dominance (Leacock 1981. . can be found in other forms of critical anthropology: Theory informs method and method shapes theory. with kin roles. h d'ffi r I ocial power or authonty. 6 inist in the sense that gender is always a cenlra beyond. Her feminist claims were mechanical evolu~IOOIsts WI~. t . Correlatl~ely: st.the example of Leacoc k'm ml'nd . Her work is clearly and unapologetically situated as advocacy. n i bles that together produce ~te societies. Leacock posed questions tackling different aspects of these unifying concerns.s?r~.. and what ends our produced knowledge serves-unite the diverse research efforts> 205 is the eo le who facilitate the research. m 0: inevitabili~n(:::cs~~~I~ 9~~~arate set of audiences to address: her an~ The Feminism of Eleanor Leacock Before delVing into the general questions of ontology and epistemology in feminist anthropology. is distinguished by the ontological and epistemological dimensions of method. r~search'IS ~ h at gen. only in her autobiography in the nested contexts of race. country. . conseque y. interviewing. ties and cultures But we of understanding we develop of our own and o~he~ sOCle e it as grou~ded-or human societies and cultures.vestigatio? into " \I~ ~~ics.~ eore IC: the Iiv~d experihape anthropologists probably conc. k' (S'lverblatt 1988) I think was instead have criticized as evolutlOna~~ m mtea~ock'sdialectic~lsense of causation disjunctures and contingencies a tailoring of writing :0 disi. pUblishing one of the first accounts of racism and resistance in innercity education. r ethnicity as well as historical context. eo <: gets a sense of her respect for 'd p an~ n~ ~ua~:~~~S~tructural fact~a~:~testudents. gender is one of several mte~sec~n(~)va:~der is a major organizing g ntl the transformadIfferences in power among sets of people. birthing a generation of feminist anthropologists through her introduction to Engels's The Origin of the Family. and the State (1972).

feminist anthropology makes research and the researcher-no marter how elaborately theorized the research or how seemingly abstract or microlevel ihe concems of the researcher herselfaccountable to a range of audiences.. for anyone calling themselves a feminist anthropologist to investigate the intricacies of hoW gender hierarchies operate in society X while expecting a pregnant or new mother colleague to carry on as if the l \ I I ! Feminist researchers argue that research should . Such self-reflection a . y m he same way as th . of positivist science. Researchers ha~e an obligati~n to let our audiences :no~e::e~lves~r scholarlytra.amewor~s that shape their t"ti qu. Sometimes.a~lltallvetechUlques-thatsome­ how these are inherently masculinist 0 1992). convey those to audiences. lfic mqulry. The feminist rubric that "the personal is political" disallows such compartmentalization. IS . epIstemologIcal claims of other tradit" ( .Gender ideologies and hierarchies are cross-cut by other hierarchies and ideologies. . Impossible' I' (see f ' . Listening to the words and observing the actions of people whose everyday lives . In the larger fra e p.ditions. an so on) This e .r~cls~s. e or 0 not make betw .): What is the context ti I~~S st~cturahst. for example. Gender is distinct from sex differences.' but also a way of living in the world. It would not be acceptable. 1987).sed at supp~rt the status quo. are e ontological and feminist. The knowledge we produce re ardi c IVlty. e or standpomt Harding Because we are cultured beings .. or local communities. ' and doesn't claim to encompass all pos ibl' P ~spectlvels necessarily partial cultural serting. Indeed.The problem.assumptions t~at remain extent that researchers can work to be aware of ptlons. own assumpltons and can the audience situate and evaluate claims w'th' nd I~tellectual forthrightness helps I checking for bias.' VIews a society. is setting priorities. Feminist anthropologists also assume that gender varies culturally 'across time periods and soCIal strata. . our own . are riven by conflicting stresses of race. the claims of .. middle-class women pr . . how Acknowledging that inquiry includi~ a:ci:n~'co~clus~ons we draw. whether mutually constraining or synergistic. and class helps us dismantle " widely held notions that there is an essential nature to women or men (see Stack 'l\i:cpIY implicated in and by the research.Feminisr Merhods 206 GAlLEY 207 Feminist anthropology is distinguished a I inquiry.g rende~ configuratIOns and women's everyday lives (see Lamphere et al 'l~~)m~y. The activist aspect usually stops at arenas deemed private. mte 0 emmlst anthropology. to the their Bias can be aVOIded . gender. for instance. always. hes m reflexivity. . l I .. deeply committed to the goals positivism are held up to scrutl'n . hbe intersection of these hierarchies. Feminist research as dl'st' t O m onn our own social engage. sexual orientation.ec t"IVlty . Reinharz as bias by making our assumptions and e i t nelutr~ ' and ~ur goal may be to reduce Some feminist anthropologists are als~ s~~o. techniques can be viewe~ an I eml~lSt (see. but more commonl th' s through whIch the research woman or of some category of wo~e e~Viewpoint is from the perspective of fessional women battered wo d n ( or example. one of the hallmarks of femmlst lIIeory in the late twentieth century is a mistrust of the assertion that there is a :eoncept of "woman" apart from or independent of national and cultural agendas. ' caIrns 0 obJecllvlty gird a set of research approaches and polici th can situate themselves within aCknowle. uman agen~y as a matter of struggles in whatever contexts shouldg . class. etc. feminist methodology is "not just a reconceptualizationof a way of 'doing research." Other critical methodologies don't insist on such efforts. helping shape the policies of one's professional organizations. particularly family and personal relationships of one's colleagues. P0.makes . clearly defining one's research agenda to one's infonnants.. our research 15 designed and conducted d h ~en our assumptIOns.~d I l 1 1997). Bias differs implicit or that are not even recognize~a:~:::lY . race. . oglcal claims clear to our audiences.ststructuralist. m thIS framework and how does practice? ' ow oes thIS understanding further feminist ~:nl rCfle~ive methOdology. then. ti ment. men.. although gender ~tegories may be isomorphic with the categories of sex difference in particular $ocieties. produces different e~periences of gender and other dimensions of identity-age. universities. gender in general is t:e by the vantage point taken . calls for action and engagement are at the public-domain level-addressing national or intemational policies. thinking should not translate into a ~je~ti:~~:ultura. A biased resea Y wever.:.l fr. religion-that may operate in a society. The audiences to whom the researcher is accountable include not only colleagues in the social sciences and the humanities or informed lay readers. takes place in a social world and is conducted by pe~ple h . m e lrorn research 0 theorellcal understandings with practical at"· h n women. Marxistexpertise established and for whom? ~ha~si: c al~~? Who IS the audience? How is the author claim it? Most notably' h d trut. emmIst ontology thus produces a l l . More than other critical methodologies. Typically. We claIm. for feminist anthro I ~n a ~Ider field of argument. The call for integrating feminism into one's relations in daily life also distinguishes feminism from other critical methodologies. seeks to link course.Abandonmg claims to objectivit ho y e thl~k we ca~ claIm what we from Situating oneself. ob. It openly advocates a decentered ~onauthoritarian approach to all human relationships and to thinking about those relationships. doesn t mean blas. one that sees the researcher's own social engagement as h Another assumption now made by feminist anthropologists is that gender is iocially constructed. s e perspectives on a particular society 0 It also is not politically neutral-it is a stanc . gu~de our own efforts. The connecting thread in this myriad ofaudiences is the call to engage in helping reduce gender inequities wherever these occur. As Kaufrnan (1996:165) writes. in critically . ethnicity. . Indeed. Crossexamining the links that we mak d po oglstS. but also the subjects of the research and the researcher's family and community members.

. and. . ch can be conducted. is necessary in this 'pIlase. development of evidence..It isn'! sUfficient and sexuality m socIety Y whit reproducing gender stereotypes or ra isn't enough to teach about gende~ di'. more explicitly feminist research. and (5) assuming that women did the same things as men. whetherwomen were universaJlysubordinated. whether sex differences were the same as gender differences between people labeled "women" and those labeled "men" in various societies. e a wIder latitude k Th ere IS . The next step was to design research that would correct these distortions. l~g rem mIst Critique of re ers remam commItted to scientific inquiry 'I ~ Beginnings: How Does Focusing on Women Change Anthropological Inquiry? While always useful as a research strate fe" ..• . structuralists. teac her ' s gender stereotyping in thes cJassroo b e. '. usually within the same theoretical tradition. While not read as feminist at the time.. which stimulated cross-cultural analyses of gender roles and ofwomen's authority patterns through time and place (see Jacobs [197IJ.gnoTIng student complaints about sexual harassment by one' d h' . . (2) assuming that whatever men did was more important that what women did.c~. . epartmental colleagu . on the other hand. data analysis. It h. among others. questions emerged regarding how one L establishes what"women's status" is. ow were women's s f ' . ignored or merely forgotten in the academy.as been called the "add ofanalysis-<lr even including them-d ~ h at doe~ pJacmg women at the center .emmIst research that m k . most shared a feminist anthropology was uncoverin . 'Golde 1970. Lamphere 1974. Schlegel [1977]).~su~s are outlined. and so on. practical and theore' I' epistemological dimensions of fe '. Rosaldo and Lamphere [l974J.:~ph?bia i. 1988). but there are types of ' .. Rapp [ReiterJ I 975)'i ::sr:~rch was ~onducted (see Rosaldo and women and mix" period (Harding 1987) Wh rospect. . Itum to the mmlS me odologles m anthropology. t th tlCa .mnlSl e.emlmsm there· . d' . Alongside the anthologies came a series of theoretical and ethnographic studies. th ar perlO s m fern mist . by reconsidering how marriage was defined. sCience.. such as Leacock's commentaries (I 972) and Friedl's (1975) and Sanday's (1981) comparative research. this early work would of inspire later. different degrees depths and 'a try. orth 0 d oxy. ecauseethlcsarepart Now that these ethical. challenged the ways that anthropologists had defined marriage.or what IS considered feminist resear h th d es prov. raphers should be reexamined and given the value they warrant (see Lutie 1966. le b~rst ~tep m ~he. eSe debat 'd .da worklo.::. [ReiterJ Rapp [1975J. .L L L L L L L L 208 GAlLEY Feminist Methods new state of being had no effect on her eve . Paulme [197IJ. Feminist Epis!emologies How do feminist researchers know what th kn ? edge claims made? What are data in femini~~ ow. of observed reality to evidence? Wh t . the conclusions that can be drawn? What difference does h'make to have research done by women? ~ In this form of feminist research.• Appreciating the efforts made by pioneers gives us a better sense of why feminist anthropology came to be shaped the way it is. How did tHe""'.t an. important than what women did? H Y assumptIOns that what men did Was more . or S TUggIng off a terms of attention time. On what basis are knowl•.. India. The basic questions most research designs posed were: "What does this topic look like when women are included?" and "How do portrayals of this society change (Harding 1991). and conclusions. 209 7. d e wentle century Th b IZe as: gynocentrism. These approaches can be seen in early anthologies. '.'ms.. and kinship studies were where references to women were clustered. no c an etractors might th'm. Kathleen Gough's (1959) research on the Nayars of Kerala. the proliferation of .sm or gynocentrism 1970--1980.efits from it in-· adoption ofsuch responsibility is part f . g. " h I sc 0 arshlp m the second half of th t . · an d value Ignored or denied in the w ources 0 authorIty. being attentive to indications of masculinist bias in assumptions made. even as many feminist resea h . . . o 0 t e questIOns we can imagine. of the methodology (Jaggar 1991). They represent· • " l i Cc S 0 an overarch' c. f mtersectlOn analysis.' ese can e character. Researchers corrected the biases inherent in: (I) selecting only male informants. Anthropological literature in and before the 1950s presented marriage as the central institution ofkinship. are de bates regardIng each of thes' Th I m . y investigate the relationship between ende. the researcher first reviews existing literature Go the topic. Whateverthe disagreement~a~s~n 0 fe f~m. by integrating women's work and sources of authority through kin connections into the analysis. research design. It hlle . The creative disco: ecause yo~r son ben. Retrieval earlier studies.ad. standpoint theo and' . Gacs et al.fort or tensIOn that results trom the 0 one s state of bein b . (3) analytically demeaning the contributions of women. was most prevalent in the earliest h gy. roughly an underlying commitment to positivism Th g fi mmlsts ~n thIS period.. as Eleanor Leacock's (I 954) research on the Innu of Labrador did. following major theoretical orientations in the discipline among functionalists. (4) discounting research topics that would be associated with women in the researcher's own culture.thropology. Approaches to feminist anthropological research were diverse. So we can expect that reminist critiques of anthropology would begin.n one's own milieu. Contributions by earlier generations of women anthropologists or ethnogi. process of creating g central questions ofanthropology embod ma e la~ m the dISCIpline. By the mid-1970s. e Issues.mpmc. research? What IS the relationship a conslltutes proof! W·th· . and Marxists. dIgnity. why women would have more or less power in different types of societies. the ways ' when women are the focus or are included in the analysis?" Based on these early investigations. on the one hand.

it is still unknown what women ssumptlons and SOCIal contribute in terms of I their research on gender saW a flurry of rejections.9 . women's \iv~S b ve cultural and social value.Feminist Methods 210 GAllEY 211 -." us ma lume 1987) ocusmg on what the women are doin " is sl'lI ' ' g I' . observing and talking with women a w 11 g. I?n t at t IS first spate even If the terms used then weren't tgh ex and gender ISn t accurate: some did ' e same as the d' I' ' researchers. In those arenas . It generally remains a starting point for ethnographIC . about whether gender hierarchies existed in societies not encompassed by states. (Lamphere 1. These issues became painfully obvious when the first round of tenure decisions of newly included women scholars publishing l l l l! On a practical level.s ~sa. such as Krige and:.en: tnIst anthropology didn't distin uish s ' . A more radical departure from the early studies entails questioning the "naturalness" of seX differences. I a sources senes of ofhlStoncal a th' and.search. As researchers were uncovering dynamics in other societies that valued women's work.~tadm:n (Slrathem 1972). and that .ri e's ographles that dId deal with women challenged assumptions about the presumgd Th~1 Realm of a Ram Queen (1943). later feminist scholars . conceptuah~edand ~. and whether state l l i \ l . there were those who under-' gendering changed historically and in . on the one hand. and international political and economic processes. on the other hand. among other subfields Tb p g~. The sharing of research results in the emerging cohort of feminist anthropologists led to national campaigns to redress male bias. Friedl's (1975).S I SIn ow developme t . 1978).I a VIable research strategy in arenas of social research that retal'n . Ime see Stack 1974.~yo 1980). mVlsl e see. economic an. Edholm et a!. including orchestrated action to affect university policies and funding priorities in private and public foundations. for example. and sometImes to Ditton 1980). The first two sets of issues exposed key areas of male bias in anthropology and related disciplines: Research by or about women was not comparably valued.pa ICU ar ywhen claims F .:s3 m hmng and. what constitutes sex difference has changed with new technologies and. national. term gender was used only rarely at the I' ( to racIal stratification. stood that sex differences were not th ar y perIOd. . indeed. noted ' " nca 1men so the contenl' h h' of . researchers came to realize how similar dynamics operated or did not exist in their own milieus. . Van Alien l productive and reproductive activities and how this work and. a person's perception of what "counts" as a natural difference is culturally shaped. con· to women at various stages in their life ar h u only avaIlable authoritative. Many would point out the limitations I'nhe er ~o es as sex roles. and Sack's (Brodkin [Sacks] 1979) comparative studies suggested that women in less-stratified settings enjoyed greater social authority. While at first blush this may seem absurd. ~nh eSlgnmg research that required provided insights into cultural dyna~i. Here the social engagement relationship of theory and practice came into play. .. but that. identified gend I IS mctlOns used now. research on women within anthropolo wa . Yet even in thi: eX It er:nces as unmediated by cultural. Other women o~slY wIves a~d mothers and husbands and their being situated locally as categoS~ 0 ars. and ways in which vl'tal c' t ebna~ were some women could become .e. or how women's contributions were structurally eclipsed in more hierarchical settings.. sometimes tegional or religious affiliation) will prevent "women" from becoming a monolithic category. whether a division of labor by gender meant gender hierarchy (Leacock 1972..Including other dimensions of social hierarchy (class. . Quinn 1977. some arguing that the split between sex and gender is itself a Cartesian dichotomy. ongomg women's movement saw im ortant cha es . Leacock and Nash ' The effort to retrieve earlier submer ed ethn " ' in important ways. Leacock's (1972). however. ' a mascu mIst set of a ' re atlOnS. although focusing on gendering will often change initial impressions and . Brodkin [Sacks] 1976. Within the academ thes:" own condmons (Omvedt 1977. on n utlOns made by w socta recognItion and value or were rendered' "bl ( omen were accorded 1972). and (3) how the social category of the researcher shapes the relationship of researcher and those studied. Rapp ~~78/ exammed m a senes of review essays 77 By focusmg on what women were doio d cl . Alongside a large number of individual researchers' case studies. researchers women from Important social decision¥makin g : en a~sumed ?efo~e to exclude temporary case studies researchers identl'fied d Iverse . ' e concomItant push fa ffi ' . of kinship studies. lrmahve action by the tenuring and promotion of women (~anjek 19. Bell and. to a lesser extent.:~oen as gender (Oboler 1980). however.8 ' Most ofthe early studies (1970-1980) a sex differences or uncritically accepted essu:f~d that gender differences paralleled perception of naturalness. see also Lamphere 1993). d pohtlcal anthropology (Vincent . how to assess women's status (Ortner 1974. ren t 10 such analyses rt' I 1 were be109 made about "women's stat "(A d' . Brown 1975). (2) how the audience and discipline values tesearch and researchers. Researchers differed.. for example. n was how projects were implemented (Leacock 1977b' Ok local women organizing to ameliorate th ' . I ' . 1977. Lovedu females could be simultane Ie para el of sex and gender: Prominent fathers. the argument is not that there are not physiological distinctions between females and males related to reproduction. Most feminist researchers now appreciate the importance of analyzing how gendering occurs in a specific setting and how it might be changing in relation to local. conducting this kind of corrective research raised three other sets of issues for feminist anthropologists: (I) how funding for field research is granted and consequently bow research topics are valued. reflectmg on field experiences.hro ol:ppr~~ches I~d to a profound reevaluation 1990). as well as in telation to life cycle and kin dynam ics. whether pUblic/private realms existed in all societies (Lamphere 1974. Debates ensued regarding: variations in the meaning of marriage (Etienne 1980). even if the 1977). the effort led to im ortant ch ' local ethnic groups (Stack 1974) h'ft ' hP anges m state policies toward . l l . ethnicity or race.

economic dominati~n. however. researc hers had to abandon more .t . The differential effects ofrace. . .naIve assumptions that "being a woman" alone gave a privileged lens thro~gh w~l. Correlatively.~ ~st eV~~Pactivism 1980s-Early J990s: Problematizing Gender.- L L I . ian d embedded in other kinds of sociocultural hierarchies-and ~hlS compleXIty req~ r~ careful investigation The fonns of social hierarchy that mterwove and colhd~ . 1987a. Gailey 1980."t theory. spective of the volume. religious conversion) as eadh mvolved .Ie~arc Ica. a script that might not be isomorphic with the categories of sex difference. An anthology by Whitehead and Ortner (1981) soughtto discuss the relationship between sexuality and gender structs Some of the contributors approached the issue of sexualtty m a c. gender was assumed to be mherent y .' . Ortner 1981).. state "analyz~ng ethnic and class fonnation. re a e s that articulated disparate structures of racial. Gayle Rubm s em: y (1975) led the way by challenging the "naturalness" of sex. te how these colonial agendas were sometimes ~' that jnvesuga . gendered power dynamics.ch to view power relations in a setting.. G~?~enn~ was ':. presenting kin networks and housing arrangements as fraught with internal. L- point in time. legitimating the study of sexuality that long had bee? I.. changing set of beltefs th. 'to tn~reargument was that gender was a historical. Moreover. Mullings 1986. f':' colonlzmg ·t ~a~~int b~storical gen~e~ docu:entm~ le. ". sexual orientation on the lives of women and men made feminist anthropologists move toward analyzing settings as "intersections" of social hierarchies that produced distinctive perceptions of the world (see. fonnrtatltonimpose new notions of gender or to ordain aspects of local gen edr. 1987b. The case stu les agents. The works were mostly empirically grounded-that is.system s . based on field or archival research. where agency mattered.lmkag~ 0 sexuality with gender systems anticipated the move among many femmlst an~ pologists in the 1980s toward Foucault's version of poststructuralism (Yan~glSa 0 and Delaney 1995). the 1980s saw the emergence of a series of studies that complicated the category "women" with both historical and comparative variation and th~t began to examine the gendering of analytical categories commonplace in anthropology (MacConnack and Strathem 1980 [1987]).:am for scholarly inquiry. for example. Some were also positivist. of gays a'nd lesbians within the United States opened a new. racism..te. The volume. Morgen 1989). class. of structuraI ' to the types of women d s examining the relationshIp oppresSIon activities and activism. but depended also on women's and men's actions. d profound (sometimes successful) local reSIstance. ~ pr?~e~n colonialism. .. This realization helped move the discourse of feminist anthropology to discussions of gender as a configuration of cultural expectations of people. The influence of the flood of field studies and theoretical writings led researchers to question whether it was accurate to talk about women as if they were a homogeneous group in stratified societies. notions 0 propr cl a lso . Etienne and Leacock's Women and Colonization (1980) was a series of case studies . Some were empiricist-presuming a tangible. measurable real world existing independently of interpretation.ender systems and calling attention to the heterosexual bIas m the anthro study of kinship. researchers who had not already done so were moving away from discussing "women" or assuming that writers could present women as having a modal experience even in one culture.at were I t ~ to political and economic changes but not reducible to them.L L 212 GAlLEY Feminist Methods 213 L LI L L L fonnation involved the historical development of consistent gender hierarchy (Rapp 1977. . and contemporary resistance to the imposition of hlera:c the uneven development of gender hi:rarchies even under condItions 0 capt a d I ment (Nash 1985' see also Galley 1987b). This led to feminist anthropologISts cntlcal~y . In keeping with the ~verall st~~~ralts~~er. particularly the assumption that difference is ~ooted m coni ceptual schemas where the tenns may shift as part of a configuration o~ cUlt~ra practices. Complicating Women's Identities and Oppressions By the early 1980s. The 1980s also saw demands for inclusion and diversification within feminism by women of color. which brought new concerns on both practice and research fronts to feminist anthropology. as cross-purposes an often ·"f. including the ways in which the issues are framed by those conSldere as experts or powerful.s . ·th few exceptions to heterosexual and clinical studies.u tu contex. at least as a "ghetto" that most departments should have represented in some manner. I bBecause of the examination of feminist movements that these prevIOus y s~ . feminist anthropologists concerned with peasant communities eroded notions that "the household" was a monolithic unit vis-a-vis outside power relations. ethnicity.'. This growing compleXIty moved ~ost 0 e terrain of debate from positivism toward what became known as standpo. that included historical change. .. merged or ignored VOlces deman ded. and re IglOUS orn'. . 'dimensions of colonization (conquest.on ~n ~IOgiCal hete~ose~ualtty ~s ~ulturallc~~i rr I '. The evolutionism that marked much of the earlier work was replaced by researchers focusing on configurations of political economy and sociocultural dynamics that shaped diverse sources of oppression and authority within societies at any given W~ntr~:':. Feminist anthropology in the 1970s problematized the study of women In a number of dimensions and legitimated the line of inquiry. Silverblatt 1978.'mlte~. although taking different fonns in different circumstances. also creat~d arenas where a~~n Was possible. producing multiple identities for women and men. or age~cy explored later in Stoler 1991). aiming at general social laws and the achievement of a gradually perfected truth. Women ' s I'd entl't'les were more complex--cross·cut . nationality and. ethnic. f iety .effo "fit" s the o . Others analyzed the transfonnation of gender through imposed political and economic processes and resistance to those processes as a vehicle for understanding the range of feminist practices in the world. It also introduced ways of.

ms. I ' 'solt encouraged wlthm t e SOCl . she carried with her the· consciousness of her own oppression within her home society.e tex.: pose particular proble. and for shaping the research agenda. witnessing.'s Nisa: The Life and Words of a !Kung Woman (1981). ethnicity. rendering the subjects of research into objects. h 'al 'diJpeos100S. It is a logl~~ oU~~:I:ms. and so on. for being heard."~o:: ~ifferent audience and the consequences of that .some . produced. of course. but as many pointed out. 1975). panicipant observation.Cl culture and gender assump~ reader the role played by the researcher s c ass. Even Ihough she is privileged in being able to conduct the research at all (national privilege). ' tlOns. A day challenges cultural relativism where hVIOI:~~:u~:ra::~ertise of feminist anthroand can facilitate recovery from number of researchers h~ve argued t~at t e p pology can help stop VIOlence agamst women S d 1990 1995 violation (Hoff 1990).ebY recovering the entire research same critical plane as the overt su Jec ma I I I process for scrutiny in the res~lts of ~~searcI~ . "'nshapingtheresearc . SItuate produced and a hlstonca Y s. t women is concerned. Thus. d este been sugg y.o e. ~ ac I:V: that knowledge is socially no t for insisting that the audience has a ng h to d t J'o'ln the evaluation process 'th powering t e rea er 0 ..: necessary in feminist research.S. Bell et al. g t db the encounter between people who knowledge.r II hiftin terrain of contentious meanings. ofw'a'trsasn~o:~:.e:~::'.~~ :2knoWledging for the audience/ She elaborates that standpotnt ental s le~~ . 1993). . and religion of outsiders might make them better witnesses than someone whose gender or national identity was the same (see Whitehead and Conaway 1986. I deed for U. was criticized for its inherent evasion of the question of commitment and its reproduction of llathering. I' t' "-e or two Im .. bl willing to recogmze these ot er 'illem [rom the researcher wh.'" d' ensions t ere IS pr . lema e gem l I field research: advocacy. ItS Pth 'n uirer himlherselfbe placed on the standpoint as feminist analySIS thabt. because on some level she is one of them. The problem is. I . A number ofstrategies were developed for addressing the problem of power in anthropological . there would be a kinship that might even provide a privileged lens for analysis (see. . ' . h b. .m what is called "standpoint shared meanings unlik. " h 'dimensIOns cou e .. . uans atlon. It is sometimes seen as a vehlc e. by making the author accountable for th. made the pretense of "shared fate" in feminist fieldwork seem another form of imperial arrogance (Narayan 1995). One of the m J . Harding (1987) characterized theory" and a resolutIOn of . h~s ?~co~:opolO:yin the mid-1980s The question of accountabilityente~s mto . ThIS concern Wl em wides read in the 1990s.ects of research possibly a lena mg -.e researcher mIght co h ' l'vl'leged access or understanding. nsider that because there IS a ma c ·'.. the researcher in the marginalizedcategory develops a kind oflidouble consciousness"that pennits insight into the kinds of power relations that might be oppressing the people whom she studies. c . The position raised two questions: On what basis do we make claims to which audiences? How do we reduce the chasm of power between anthropologist and informants/subjects? This argument was controversial in the 1970s and remains so today. and so were the people among whom the researcher lived and observed. behefs. women. :.ely. emt~ a.~ow to integrate one. . mtslS~ter et~e. " t hI! with informants in .ntradiction between feminist and aniculates in an import~nt way With an Im::~~e~ates continuing to the present practice and cultural relaltvlSm. This position was consciously parallel to that of cultural nationalism within the black power movement in the United Slates: The researcher from a disempowered category has an edge in conducting research on other disempowered people." that is.emlms s. Testimonies. Demands from oppressed groups-primarily women of color-within industrial capitalist countries and the postcolonial world for inclusion. and Transnational Questions of Power It a series of clearly interested arguments-has Creating "situated knowledges f k' g clear that knowledge is socially d b th s as a means 0 ma m . as c:~i~f:. which categories matter more for the research in question. conducting research on those more powerful. for example. it was thought that because the researcher was in a marginalized category. Id b more Important to t e su . on the other.214 GAlLEY Feminist Methods 215 -. s know e ~l life history/ethnography that investigated the f • matter of conten:IO~. voyeurism by the more powerful. The nature of one central technique of field research.o unta ela~: d~~erences within a shared ethnicity . based on the meanmgs crea e d y tandin that social hierarchies make are different in various ways. Studies of gang and ~cqua:ntanc'~. b~t .. Some called for research to be done only by women from the same culture. 1992).. the class. l l l l l . n. depend~ o~ un ers th fr. wife battering (Counts et al.. rhacdee's'lgn pr~cess and results. The . an e aVlOr 1 • ert the hidden authontattve researcher in this stance can't remalO :h~ a~ony~~usie~PauthOrial responsibility and voice. for developing ongoing patterns of sharing and collaboration with informants versus focusing on fieldwork as a place and the relationship with informants as a lens to explore the conundrums of the complex identities of the researcher in her own society. . Rohrlich-Leavitt et at.. But other . db h .~. Acknowled~mg po~~ e of them into the research design and process IS . the subjects? In early studies.~:u\il':::i::(Gru:nbau~ 1996). too. They might relate to her better. Some proposed a kind of resolution to this by having the researcher's configuration of social categories be somewhat aligned with those of the subjects. What emerged from this ongning controversy are different feminist strategies.. sciences. race. on the one hand. how does the social category and position of the researcher influence her or his relationship with those who make the study possible. I: l l l I \ The 1990s: Translation. Standpoint Approaches The early feminist approaches raised another set of methodological issue< namely.

people's identities and affiliations are being constructed from experiences of rnigration. .~OI:t~es~arch~rs within more women and national-level activists The power d u ~I e. Weston's (1997) research in lesbian and gay communities in the United States and Lewin's (1993) work on the implications of lesbian mothering for American kinship have stressed the need for negotiating and developing consensual strategies for communication between the ethnographer and her infonnants where both are aware of the implications of the research for an entire community. ass.:~::.~~re. such as battered women or breast cancer activists. ~~::i~~a:~st::. The qu~stion debates defended the practice But 10 I t ' .of Some kind of was not whether female genital mutilation was t rnatlOn~ohdan~.an~ asylum or addressing th. legal-judicial.YS at rocked feminist countries had denounced the practices and c~lle~%la'th u. activIsts from other countries such I' ore appropnate for gr. or religious hierarchies. raCial.~mm~ ea er~hJP.o)~tri~s.d e SOCial versial example. r eIT a an onment before 1980. It also has given rise to an offshoot. . class.. Feminist anthropologists engaged in field research where the anthropologist's identity overlaps those of her infonnants in key ways have emphasized that shared dimensions of identity are no guarantor of either access or resolution of questions of expertise or power.h::e~p~n:i~~~~f~. For example. without either overwhelming the voices and perspectives of local people or irritating the readers with self-indulgent fonns of reflection? The strategies taken to address the question of feminist solidarity versus intersecting hierarchies that create power differences depend largely on the political stance of the researchers in relation to the women who make the research possible (Brodkin [Sacks] 1988).o wlthm the activists' own countries' and (3) Wh h omen on a dally baSIS local feminist interventio h '. collaborative the planned project is and how it articulates with work being done by local feminist activists and organizations (Stephen 1997). particularly where these women come from diverse backgrounds (Connell 1997. n. at appens to the effectiveness of intellectuals from fonnern~rwpr:~:uetrccellve~ lleadership is too closely identified with o OOla powers? The ensuing discussions have revealed the 0 relativism: The notion of h p wer ynamlcs mherent in cultural women and men in socie~ie~~i~~~n~ous tUlture does not ~t the r~alities lived by r.in b 'L L L 'L These African activist scholars and p bl" h I h from the United States-both whit ud I~I e~ t workers have called the activists images of Africa and for focusing ~:~ on :~ -:-t. Calls for arguing from a stated position are efforts to make inquiry accountable. born of the efforts to make standpoints clear and to produce an anthropology that forefronts the unequal negotiating process inherent in field research. highlights the need to not revictimize as well as the vagaries of participant observation where empathy.-murde~: ~n~C~~tft~~~~~:e o~nwtheir own cou?tries t. a condition oftransnationalism has been created.:~sc~~~7~~~~~ fro~ allvlsm were the health and safety of marginalized e l ' 0 re_ issues involved.strategy necessar~ly mimic masculinist models o~-sbo~i:~ :~:. El Saadawi fi~:~i:U~~~~b~~at~~~:~:Sr~~. . for instance. The concern with standpoint approaches in feminist anthropology continues. medical. (See. emmlst and infibulation are routine respond:~ t~O~i~ou~:n:s where ~emale circumcision circles across the globe. Field research involving vulnerable populations.~el~:~~~:~~ But t?e power to the kmd of research we undertake Let' . lies lor activism might be m . deployment of data. Many of these recent studies situate the anthropologist more clearly and involve collaboration with the subjects of the research in a more forefronted way. power. s consl er an early and contro- 217 the on female circumcision. and legal outcomes of women seekin d vi?l~nt pahrtners call into question anthropology's h~t:~i. ~:~: was a~ outcry a~ong several anthropologists from clitoride~~encan co~ntnes c~lIi~g for an international ban I L L L Afrhrica. As the new global division of labor that draws on women and men differently and intensifies conditions leading to international migration (Nash and FernandezKelley 1983).S. living "in between" as well as in two or more countries (and gender . mVlOlat~? Whose interests are .dan an: other relevant feminists from other parts of the Id . ey eman t at the researcher examine how d '. and confidentiality are so clearly implicated in women's safety and well-being (Hoff 1990. ethOlc. Investigators to local national feminisms are intricate cO~Plicated hi' yna~lcs ~n the range of interY Ines 0 national status. One positive result of the debates has been the proliferation of research on the role of U. oler~ble no one mvolved m the What happens to girls-ran~ing fro: ac IVlsts mSlsted that others consider: (I) Prosti~ution-ifthe practice is legally bans~::~it~arassme~t .to sexual ~ssault and estabhsh respectability' (2) What ven < " out provldmg alternallve ways to . Sanday 1995) or in addressing the multiple needs of women who have been violated.~n~? Does a top-down At the same lime the deh t h h 11 ' . Connell 1997). Feminists from Egypt ~~ I~ w. Richie and Kanuha 1997). the power problem remains: Some knowledges are better situated for inclusion in global debates and policy than others. g:lnded .L L L L I L 216 GAlLEY Feminisr Methods 1982 [1993]). On an unprecedented .~I~:. Anglin 1997. and calor (see BOlles 1996) Tb' d d h . war became mvolved (for example.. While clearly articulated standpoints make evaluation easier and place debates on a straightforward footing. How do we situate ourselves as ethnographers. scale. Thiam 1986 [1978]). also underscore the relationship of th~ ~:S:a:. even as the author remains accountable to both community and wider audiences for her arguments. and social institutions in facilitating violence against women (for example. Whose version of the cultuYr c. Walley [1997] for a ant apologists and feminists in gene I f i ' .o t~Sk drar r~inforcing colonial "global feminism" than on showing needed in. e IS prtVI ege when researcher h a partIcular practice is "traditional" and th fi " s argue t at served by recourse to legal rather than c~m. powerful countries t~ examin: .

.2/8 GAlLEY Feminist Methods 2/9 systems). the edifice. a reference to become knowledge. a n · .at :~~ seen as a feminist agenda and ad~pt strategies of empowerment t at e lOe l l the subjects of inquiry. our sense of our own phySIOlogical ma e. research in this vein risks audience alienation. American anthropology's canonical "four fields" are dis~~~:d and d~fended arenas where addressing questions about hu~ans. into betrayal (Alarcon 1994). Behar and Gordon 1995). has made concerns about feminist practice in field research even more important.~~e:~~:~:C~I~~::t:~: ::~:i. The risk here is collapsing the certainly problematic anthropological encounter into a kind of psychological memoir or. of the text as it is to make any other claims (Behar 1993.erent . shaped. In the testimonial of Maria Teresa Tula (1994). l . these twm aspects make the research interdisciplinary as well. f . ! ---:] . The degree to which reflexivity characterizes a feminist text varies. Yet feminist anthropology has developed unevenly in the subfields ofanthropolog~. or a main. and tapping cultural resources from a variety of sources.d b r eaplta IS fr ' both parties' efforts and bringing the collectivewlS om to ea ons sharing le. but all seek to make the act ofwriting and filtering experience-ofcreating knowledge claims-apparent to the reader. done in association with Lynn Stephen. a travelogue. more skeptical of outside expertise. r t processes state ynamlCS. sparked. for example. ~ with holds that this is due to both social exclusion and hermetiC c~nce~ . participants come to d~n~ w:. Stephen argues that collaborative encountets are necessarily flawed in a global setting where power accrues to the researcher. Writing about experience without these intervening concerns is memoir.nt:· ~~:~ would therefore develop data and communicate it in different ways todre a e 0 Hent people. The techniques used depend largely on the quesf~mlOlSt a~th : Wh~ can be developed as data in the research context. While this is not a problem for memoir as a genre. As Mascia-Lees et al.ss born. Indeed.di:~~~USaf~:I:~S~r:. The multiplicity and often amorphous swirl of perceptions. class.p. or informed by issues in their own society andlor the concerns of the people with whom they work.1 research.. and our sense 0 f our pas .:s:sa~. I have the subfields. to a degree at least. linked but power-laden spaces (see Grewal and Caplan 1994). In the interest of redressing the imbalance of power that originally makes translation of one people's experiences for an audience in another culture. Politically. but that this is no eXcuse to withdraw from feminist practice. Some go so far as to write about experience as if it and the production of knowledge were the same. and Linguistic Subfields Today fat instance. for instance. Irst emmlS anthology in ethnology was published 20 years before the first one 10 archae~IO~y (Oero and Conkey 1991) and the uneven development of the concept of gen. at worst. pologists tackling the thorny issue of how not to impose a form of feminism embedded in a particular configuration of gender. for exa. meanings. the research conducte~ has to ~ake sense l -. 'fc rtions of women to men m a fact that reflects in large measure the dI .~rt to anal ical category and (2) the purpose 10 contrlbutlOg to an overa e . In another volume 00 issues 10 e::te::~:~ci~l movements in Latin America. about the society sending the researcher and the researcher's own concerns as it does about the people who are and facilitating roles. furthering the agendasofthe informants and a~~:a~ f their own experiences her pnvtleged lens onto g the subjects sense 0 d . h· th Feminist anthropology both merges some ofthose terrains and operates Wit 10 em.p:opo . and migration have meant that informants are better informed.~~C:i·qUes of investigation can be used in AS.ehn~ ~udiences as well-social scientists or undergraduate students. ty informing both collaborators' practices. Transnationalism . Most feminist anthropologists would argue that the problems they research are. literacy. but it is not given. If feminist ethnographers are are. In addressing these issues. the. For those most committed to the approach.~~~~~':n:nity. the researcher assumes that the entire research process-from design to audience-reveals as much. d the like Tbis kind of collaboration involves .i . object of research. a problem. the field memoir can create a new form of silencing. dis:'ntle gendered forms of oppression and exploitation.~~::0~g~995) been later to develop than those within ethnology (Wyhe 1991 )'. ~~. u iogful-our symboling capacity. WyHe argues. She argues that feminist ethnographers can have a distin~:ivelY i..yp~~e~~~~s~:7c~nd:~~e~ll~c~~~e. ISCuS~~ o~o ic. t . and cultural tradition have begun to explore what methodology means in these new transnational. the task of the author is as much to unmask and lay bare the artificiality.. the form of writing and sense of audience shape the knowledge produced (Wolf 1992). Feminist anthr().t et:uinfluence policy. (1989) point out. Impact on Archaeology. PhysicallBiologica1 . What marks the project as feminis~ is (1) t~e focus on gen er~: af. and 8fe not [: . er m ' h i ' d' ~ rt with the notion 0 f these initial articles reflects some of the sc 0 ars lScom 0 . it is a problem for ethnography (see Wagner-Martin 1994). and emotions accompanying a significant event or process must be related to an audience. The shift in communications technology. movements often develop out of concerns for the well-be 109 0 wor~IO. sociocultural structures and untverses 0 meamng. Stephen (1997) argu~s tha~ on w~men-c . Flannery and Marcus 1994). the researcher uses the words of the informant as a foil for her own exploration of self. In these approaches. . ! -'--"'~'-- •• l l . t empiricism (for example.ill1essin g .1S seenkas ou. if not more. according to the Latin American saying. . Also relevant 1I0nshPose di:~ces for whom the research is intended. Reflexive approaches are one response to the problems raised (see Enslin 1994). that femmlSt methods m a. invisibly and silently ensconced far from the researcher's home. Generally. race. Some feminist ethnographies done in the name of reflexivity convert the reseatcher into the main. . Experiential knowledge is possible. ~.

however. she argued. But despite efforts to integrate gender and feminist epistemologies in the anthropology curriculum (Morgen 1989.ons tha~ gender IS denved from sex differences. The exclusion has not stopped the methodologicai and theoretical debates in feminism and feminist anthropology (see Folbre 1993). which stressed the centrality of gathenn~ and mother:mfant bonding (Slocum 1975. ~his moves the argument beyond the gender division of labor. howev~r.der. that is. a c~ncern that has dominated the subfield since its inception. although that 's ~Iso I.a~d ~rder!ng social relationships (Bonvillain 1993. warfare. men (s~e :-Vright 1996). Ewick 1994). Brumfiel~s research (1996) shows how gender can be central to the analysis of state dynamICs. justice can hardly be done to the exciting work on gender and language in this small space. 1991) and how gender bias enters classrooms through language use in naming. born of colonial encounters. and cooperation were ~ountered by u~oman the gatherer" ones. I~ th~ archaeology of state formation that attention to gender is best limited to the dlV. and so on. Collier et a!' 1993). and distancing (for example. ThIS IS not to say that gender doesn't matter in these arenas only that women ar more than producers and vehicles of political alliance among 'the "real" movers an~ sh~ers. food sharing. Tannen 1994). Hrdy 1981). Wellhouse and Yin 1997). Pica et al. a gender dIVIsIon of labor. Brumfiel challenges the prevailing ideol~:.lsl. Usmg stylIShc variations from central and more marginalized ~reas of the empire. Most of the analyses of how science has objectified women's bodies and biological processes and how women's diverse experiences can lead us to revise such konograph~c visions ofphY. Usin standard field techmques and a theoretical orientation informed by debates regard' g ?ender from a number of disciplines. such as kinship. are slower to e~er?e ~n phySIcal anthropology circles (see Zihlman 1985). new studies explore the ways that gender.siology. only emerged during Homo erectus times (Leibowitz 1983).. Feminist Methods in Action By the 1990s. But even without these L L . whenever gender is deployed as a lens (Collier and Yanagisako 1987. In the following decades. to mark a. but that th~orizin~ g~nder can help us develop new data (see conk:y and WlllIarns 1991). Brumfiel constructs an argument that local material culture can lIterally embody through gendered symbols. During the 1970s widely accepted "man the hunter" arguments that stressed the centrality of purport~dIy allmale hunting groups in the development of language. rather than simply women. claiming. and how gender provides a novel way of understanding the dynamics within ethnic and working-class communities (Lamphere 1987.mportant.oluti?n (for example. L L L L L L I. Only Lelb~WItz questIOned why ferilinists would assume that any gender division of labor e~lsted from the outset of humanity and offered a revisionist interpretation o~th~ fossil and matenal cUI~u~e. Frankhn 1993a. new ethological studies have refined or challenged these VI~WS of human e. assumptI. Related research focuses on how in linguistically diverse postcoloniaJ settings. Zihlman and Tanner 1978). Exciting research today is going on into the ways that gender affects classroom discourse and the marginalization of girls and women (Tannen 1996). and class are mutually constructive (Brodkin 1996). The 1990s have seen feminist anthropologists reexamining central anthropological concepts. Others have analyzed gender dynamics in primary and secondary language acquisition (for example. femmlst lIngUIsts investigated gendered languages of specific workplaces and their implications for how technology. Other arenas include the gendering of nationalist discourses or the deployment of gendered languages in the shaping or dissolution of community in transnational settings (see Urciuoli 1995).on of labor and the succession disputes or alliance strategies of ruling elites.nd provide the contours of cultural struggle. specialization. It was one of the most important contributions during the more discIplinary. even ensconced. "corrective" phase of feminist anthropology. But there are excellent examples of ho gendermg archaeology can provide vital new insights into sociocultural dynami w cs (see Wright 1996). Feminist linguistics has focused from the outset on analyzing how various languages are gendered and how that gendering. Critiques of basic. In blOloglc~1 anthropology.. race. In the 1980s. Cohn 1988). enter into societal discourse (for example. canonical texts in anthropology-including the so-called new ethnography that discusses the power dynamics of fieldwork and representing other people through writing (for example Clifford and Marcus I986)--continue to ignore or give short shrift to the influences of feminist anthropologists (Lutz 1990). remains that proposed an initial age-based technical dIVISIOn of labor. in turn. Since Iinguistks is itself a discipline. Similarly. women's access to learning national languages may be blocked by educational structures and local practices..L 220 GAlLEY Feminist Methods 22/ gender. resistance to a dominant state ideology of g~nder. are gendered in ways that may hamper women's actions in postcolonial and decolonizationefforts. Ginsburg and Rapp 1995). feminist anthropology had become a legitimate. Brumfiel's work also shows that archaeologIcal data need not lImIt the extent to which sex can be analytically separat d from g~n. shapes ways of thinking about the social world . pIOneering examinations in the I970s by Lancaster (I?73) and Lelbowltz (1983 [1975]) of male bias in primate studies and in hmoan ongms work produced novel theorizations that changed the "story" ofhuman origins p~rmanently. Feminist linguistic anthropologists also have been active in documenting the ways that creoles and pidgins. L. Brodkin 1993). have come from cultural anthropologIsts (GottlIeb 1988. given the matgmalIzatlOn of women whose work does not fit into the dominant paradigms in the field (see Lancaster [1991] for an extensive review of these issues). Haraway's Primate ~is!ons was pathbreaking in its expose! ofthe gendering ofprimatology and human ongms research (Haraway 1989). Martm 1992.

new questions are being discussed. Research on gender and power proceeds apace in this decade. ' The work is unusual in how that agency is portrayed. The research has clearly drawn Implications for health care. but with new attention to ways that ot~er vect~rs of social power affect the research process and new strategies for redressmg the Imbalance and producing participatory research (see Jarrett 1994. the study addresses the power relations of anthropologists and informants struc- communities encouraged reexamination of what kinship means and how kmshIp IS forged. and the question of reflexivity m socIal SCIence mqUlry. For example. t:-v0 other are~as are of gro"':tng iJnportance in feminist anthropology today: the poltllcs of sexualtty and ItS rel~t. sperm donation. In doing so. female-to-maletranssexual operations in the United States make a strong claim that maleness in the United States is conceived of as having a penis and more testosterone than estrogen production. In a series of articles Susser examines. is made more complicated (Franklin 1993a). Wang et al. Weston 1991). because of the intervention oftechnologies permitting transsexuality. how dire poverty shapes women's lives as mothers wives sexual partners and intermittent workers (Susser 1991. Franklin 1993b. Today. the interdisciplinarity of feminist anthropology marks it as a model for the de~ates r~gardi~g t~xtuality. This arena of inquiry pushes the frontiers of feminist anthropology past the dichotomies of earlier eras to a more fully social constructionist appreciation of both sex and gender. Indeed. and transnationally (di Leonardo 1993).lOn­ ship to gender and other aspec:s of social hierarchy and the cultural and poltllcaleconomic aspects of reproductlOn. with interviewers whose racial and ethnic backgrounds would provide useful differences in how people interviewed might deploy information-the perception of relative power between interviewer and interviewee. partiCUlarly where these adoptions tend to favor one gender or the other. The research being conducted on sexuality and culture demands that we reconsider how closely or obliquely sexual orientation and practices are linked with gender systems and identities. yet also structurally and therefore personally vulnerable. but is not necessarily related to ejaculation or sperm production. ln~er­ nationally. Women are both active on behalfofthemselves and their children. on the one hand. 1996). race. on the one hand. Works such as Riley's an~ Weston's differ~nt treatments of American kinship when focused on lesbian famlltes (Rlley 1988. these studies ask what shapes commitment or orientation of persons to gender identities.ate whether people involved are also transgendered. social welfare. the writer as expert. . on the other (Franklin 1988). Let us examine some actual studies to see how the methods work in practice and how theoretical inquiry and practice can be mutually instructive. Strathem 1992. we can see from the dIscussion above that there is little in the concerns of the "new ethnography" that hasn't been discussed over the past 20 years in feminist anthropology's methodological debates. creates new arenas of complicated gender and racial as well as class and national identities (Gailey 1998) and call on anthropologists to develop new models of kin formation (Modell 1994). 1996. draws on the technologies invented and deployed in metropolitan centers of late capitalism to reexamine what is meant by sex differences. and familial kin roles. Perhaps bec~u~e women-and women of color in particular-in the industrial capitalist societiesthat produce most anthropologists are barraged with naturalizing and reductive images of their being. 1995. The research Itself was team constructed. Susse. and what are presented usually as natural sex differences.hough difficult to conduct. and G~nzalez 1992). similarly. Alt. and Lewin's (1993) argument about reproduction and parenling in lesbi~n ~he l l l l "I workers. This work on reproduction in the new world order articulates with research on cultural variations in how human status and social personhood accrue . More fundamentally. the expansion of transnational and transracial adoption in many industrialized capitalist countries. in-vitro fertilization. Research on the gendering of reproduction. and a team of other anthropologists on homeless families in Ne~ York City in the 1990s shows keen attention to intersections of identities in the development of research strategies and in outreach work. through narratives from homeless women "talking to" social By by diversifying the team and making policy arguments at key public events city without agonizing about the personal identity of the researchers. they often are the only outsiders with any sustained connection to these families and because they simultaneously hold power over them but also can get resources no one else can (Susser 1998). The class. 1993. feminist anthropologists have been more critical and earlier in their criticism of the objectifying and reductive tendencies in social research. 1997). given the transience of the interviewees. For example. and domestic violence services. multiple mtervlews were conducted. In :"dem with this research into the relationship of gender to the polttlca~ and economic aspects of oppression in late capitalism within the United States. Such technologies challenge the naturalness of sex difference and make apparent that what is received as nature is actually a cultural encoding ofphysicality. egg transplantation. Researchers are called on to evalu. Men and children are considered actors as well because the work seeks to present a comprehensive view of gender relations across generations and in relation t? the agencies and gatekeepers the people must engage to survive. and the relationship of theory and practice. Ida Susser's ongoing work with Denise Oliver Maritza Williams. on the other. the problematic between the social construction of gender. and even cloning demand that we evaluate what kinship categories mean in this new social setting (Ginsburg and Rapp 1991.222 GAILEV Feminist Methods 223 e~citing new ventures onto classic anthropological terrains. The multiple predicaments faced by these women and their families are placed in a political-economic and historic context of ~ ra~idl~ restructuring New Y ~rk City (1996).. in an effort to call attention to and change policies directed at women. research that can help readers approach l l l l l I l l l . Correlatively. and gender issues surrounding surrogacy. Thus. to fetuses and infants (Morgan 1993). Social workers were included because as Susser explains.

Bell. 1976. Anglin. 1993. Local variations in what constitutes gendering the relative importance of gender in social structures and cultural gen . Culture. Minneapolis: University of Minnesota Press. Norma. we need . to engage. and other social positioning. London: Routledge. eds. and Ethnography. REFERENCES L L I Alarcon. The global impact of industrial capitalism and the upheavals that accompany its normal operations call on us to listen. 1979.L L L 224 GAlLEY Feminist Methods ~:. Work. we must understand how these important aspects of social differentiation (if not always hierarchy) shape the lives oflhe people who make our studies possible (see Morgen 1989). Men. Behar. and Home in Kingston. but the process of Othering can be held up to scrutiny. Feminist anthropology can't escape the inherent production of Others so long as difference is reproduced through the normal operations of our political economies and attendant ideological practices. 1997. Karen. Law: -The Old and the New: Aboriginal Women in Central Australia Speak Out. 110-133. . present' practices. Traditora: A Paradigmatic Figure of Chicana Feminism. Women WritlngCullure. Moreover. CT: Greenwood Press. Social Science and Medicine 44(9):1403-1415. Gendered Fields: Women. Conclusion A range of feminist anthropologists are operationalizing the often-invoked need to understand how class and race. class. Ruth. or silenced because of their gender. By keeping linkage of theory and practice vigorous. Female Husbands: Gender and Sex in an African Society. The view offeminist methodologies presented in this handbook is in part a response to those efforts. State Bias and Women's Status. andCommunication: The Meaning ofMessages. L L I '-- ACKNOWLEDGMENTS I am grateful to Karen Brodkin and H. Russell Bernard for very helpful criticisms and suggestions regarding the paper. to see how shifting and intersecting social and cultural identities affect the questions we can imagine and pose and how we arc perceived in our research settings and among our audiences. 1987. Diane. Ruth. a warning and a claim for appreciating the insights and diversity of approaches and theories that informed feminist practices in anthropology from the outset. In Scattered Hegemonies. HaITison 1991). there will be a tendency. 1994. Sislersand Wives: The Past and Future o[Sexual Equality. This concern with "intersections" cannot be dismissed as a genuflection in the direction of political correctness at home. Nancy. Such an anthropolog~­ b speaking from our various and shifting standpoints. We need to guard against viewing our own scholarship and engagement as inherently feminist simply"' because we define ourselves as feminist or are seen that way. 78(3):565-569. American Anthropologist I L. Translated Woman. The vitality of feminist methodologies lies in the linkage of practice and social engagement with theorizing and collaborating with others to produce useful know ledges. and Deborah Oordon. Karen. Lynn. London: Zed. for if we make claims about ourselves through the inevitably partial knowing gleaned from others. Pp. and Pam Ditton. I . how gendering shapes people "s perceptions 0 f posS!'bl e past s. Behar. One of the consequences of its becoming institutionalized as a specialization is the defense of boundaries signaled by the recent efforts to produce canonical feminist texts (Behar and Oordon 1995) and to inscribe canonical histories (Viswesrayan 1997). NJ: Prentice Hall. Mary. Sister Jamaica: A Study of Women. As the specialization becomes entrenched. and an anonymous reviewer for their comments on an earlier draft. West port. and to convey to our own communities and other audiences. Thomas Patterson. Englewood Cliffs. eds. Brodkin [Sacks]. Pat Caplan. Berkeley: University of California Press. Brodkin [Sacks]. Amadiume. A. Language. Working from the Inside Out: Implications of Breast Cancer Activism for Biomedical Policies and Practices. the dering of struggles throughout the world. Boston: Beacon Press. toward methodological orthodoxy and toward theorizing estranged from our own social involvements. I also thank Mary Anglin. MD: University Press of America. however. Canberra: Aboriginal History. Feminist anthropologists. 1993. race. Diane. 1993. and Wazir Jahan Karim. Bonvillain. Bolles. 1995. and futures matter more now than ever before if we are not to be consumed and fragmented beyond all ability to construct meaningful lives. ethnicity.. or nationality shape the experience of gender and our strategies for transforming oppressive relations and structures within our own societies and elsewhere (Brodkin [Sacks] 1988.1996. one that side-steps the polemics. by engaging in research that demands that they pro~ lematize the participation end of participant observation and develop more explicitly collaborative studies. Lanham. 1980. exciting research prospects can be envisioned-ways of helping understand and project voices of those who have been disenfranchised. which we can resist. and s. Bell. 225 L L debates over reproductive rights in a different way. and mostly ?y standmg iongside and opening audiences for previously silenced or muted vOlces--offers ~uch to the development of feminist agendas and practices throughout the world. marginalized. are at the forefront of these new studies. Male Daughters. Ifi. Traddutora. Inderpal Grewal and Pat Caplan.

Pp. Rita Wright. and Gender. Karen. New York: Praeger. CA: Mayfield. Essentialism. Clifford. Golde. Pp. Berkeley: University of California Press. 1988. 1988. Special Issue: Gender. and Jacqueline Campbell. 1993b. Alma. Gender and Society 2(1):92-108. 102-139. Ginsburg. Aisha Khan. Class. In Women and Colonization. Franklin. Carol. 1997. adelphia: University of Pennsylvania Press. l l l l l. 1994. Original Narratives: The Political Economy of Gender in Archaeology. 1991. Gacs. 1980.226 GAlLEY Feminist Methods 227 'I I Brodkin [Sacks]. 294--322. Berkeley: University of California Press. Mona Etienne and Eleanor B. Jane Collier and Sylvia Yanagisako. 1987: Toward a Unified Analysis of Gender and Kinship. In Gender and Archaeology. The Hidden Face o[Eve: Women in the Arab World. 1995. Sarah. Joan. Environmental Values 6(4):427ff. Making Kinship in the Wake of History: Gendered Violence in U. Gera. 1993a. Brumfiel. Mona. Conkey. 1970. Cultural Anthropology 9:537-568. Sarah. and Rayna Rapp. New York: Praeger. 14-50. Beyond Writing: Feminist Practice and the Limitations of Ethnography. 1975. Journal of Homosexuality 3(4):27ff. Introduction: Women and Anthropology. eds. In Gender at the Crossroads of Knowledge. Pp. and Rayna Rapp. Peggy. 1986. Conceptualising Women. Judith K 1975. Friedl. I987b. Women. Cambridge: Basil Blackwell. CT: Greenwood Press. William Havitand and Robert Gordon. Franktin. 1-23. and Kate Young. 1993. Patricia. Leacock. Dorothy. Science as Culture 17:522-542. Kinship to Kingship: Gender Hierarchy and State Formation in the Tongan Islands. 1994. Edholm. Understanding Victimization and Agency: Considerations of Race. forthcoming. Judith K. 1997. 84-99. Brown. l l l . Brodkin [Sacks]. Counts. Ideological Domination. New York: Holt. Chicago: Ald!ne. 235-251. Folbre. Toward a Unified Theory of Class. Gaitey. Ewick. 1 l I Ginsburg. Austin: University of Texas Press. Race and Gender in the Construction of Class. Figurines and the Aztec State: Testing the Effectiveness of . InAnalyzing Gender. Kent. New York: Monthly Review. ed. Race. In Women and Colonization. Anthropological Quarterly 66(2):76-81. Michelle Rosaldo. I di Leonardo. Women in the Field: Anthropological Experiences. Phil. Olivia Harris. Karen. Engendering Archaeology. \988. The Politics of Reproduction. Collier. 1998. CA: Sage Publications. 1992. Coho. l 'I . Berkeley: University of California Press. Rayna Reiter [Rapp]. In Women and Colonization. eds. and Eleanor B. flannery. Christine Ward. Mona Etienne and Eleanor B. 1996. and Sarah Williams. eds. eds. 1978. 1993. Women Anthropologists. and Ruth Weinberg. franklin. Franklin. and Margaret Conkey. Brodkin [Sacks). Westport. In Women on War. Mona. El Saadawi. ed. 1988. 1988. History of political Economy 25(1): 167ff. eds. Writing Culture: The Poetics and Politics of Ethnography. Ernestine. 1996. American Ethnalagist 16(3):534--550. and George Marcus. Cloth and Colonization: The Transformation of Production-Distribution Relations among the Baule (Ivory Coast). Postmodem Procreation. o[Women. 214--238. 1987a. Sarah. 1991. Jane. Felicity. Etienne. Beth Hess and Myra Ferree. 1980. Reproductive Futures: Recent Literature and Current Feminist Debates on Reproductive Technologies. and Sylvia Yanagisako. ed. Evolutionary Perspectives on Gender Hierarchy. 1980. 1993. Karen. and the Law. Gottlieb. Integrating Feminist Epistemologies in Undergraduate Research Methods. Pp.oquois Women: An Ethnohistoric Note. Mountain View. Feminist Studies 14(3):545-561. eds. Christine Ward. In Talking about People: Readings In Contemporary Anthropology. Is There a Family? New Anthropological Views. Which Essentialism? Some Implications of Reproductive and Genetic Techno-Science. Gailey. 1994. Margaret. What a Difference Poiitical Economy Makes: Feminist Anthropology in the Postmodern Era. Collier. eds. Dolly: A New Fonn of Transgenic Breedwealth. Micaela di Leonardo. Elizabeth. and Gender. Leacock. Women and Men: An Anthropologist's View. eds. On the Perils of "Politically Correct" Archeology. Annual Review in Anthropoiogy 20:311-343. In Towardan Anthropology . Current Anthropology 35:441--442. Men. I. 1980. Christine Ward. 143-166. Gailey. and Sylvia Yanagisako. Patricia. Sanctions and Sanctuaries: Cultural Perspectives on the Beating of Wives. Sex and Death and the Rational World of Defense Intellectuals. Micaela. 1993. Euro-Ethnic Working-Class Women's Community Culture. Ensiin. lames. Violence. Leacock. Boulder: Westview. eds. New York: Praeger. cd. Jerrie Mclntyre. Mona Etienne and Eleanor B. Sarah. Connell. Pp. Elizabeth. Faye. New York: Simon and Schuster. Stanford: Stanford University Press. Blood Magic: The Anthropology o[Menstruation. Brown. Conceptual Problems. pp. ed.S. Thousand Oaks. How Does She Know? Feminist Theories of Gender Bias. Faye. Science and Society 60(4):471-478. Political and Legal Anthropology Review 20(2):115-143. Putting Down Sisters and Wives: Tongan Women and Colonization. In Gender and Kinship. Gaitey. Older Child Adoption. 1991. Nawal. Critique o[ Anthropology 3(9-10):103-114. Daniela Gioseffi. Leacock. Berkeley: University of California Press. London: Zed. Pp. Pp. Christine Ward. Etienne. 151-158. Nancy. Frontiers 14(1):lff. Uta. Pp. and Joyce Marcus. Rinehart and Winston. 32-67. Identities: Global Studies in Culture and Power 5. Conceiving the New World Order. eds. Jane. Pp.

Hoff. ' . and Marilyn Strathem. Lewin. Pp. DC: Association of Black Anthropologists/American Anthropological Association. From Working Daughters to Working Mothers: Women in a New England Industrial Community. Dialectical Anthropology 7(4):263-284. 123-147. Boston: Beacon Press. 1973. Judith. The Changing Family and Levi·Strauss. and June Nash. Cooperation. Lesbian Mothers: Accounts of Gender in America. MacCormack. 1991. Leacock. African-American Women. ed. 1977. Strategies. 1974. Decolonizing Anthropology. Prospects. In Women. Stanford: Stanford University Press.L L L L 228 GAlLEY Feminist Methods 229 Gough. In Feminist Ethics. 1971. New York: New York Academy of Sciences. Sandra. In Woman's Nature. 7-67. eds. 1991. Helen Ragone. Hall. In Praise of the Achieving Female Monkey. 411-422. Louise. Eleanor B. . Feminist Ethics: Projects. The Realm ofa Rain Queen. and Anthropological Facts and Fictions. Is There a Feminist Method? In Feminism and Methodology. and Colleen Ballerina Cohen. Pp. The Nayars and Ihe Definition of Marriage. L Leibowitz. cd. and the State. Lamphere. I" L Leacock. Frances. Faye. and Patricia Zavella. Rethinking. NI: Prentice Hall. Ellen. The Woman That Never Evolved. Berkeley: Uniyersity of California Press. 97-112. Sandra Harding. LaIIlphere. 1977. Lancaster. 1972. Reprinted in 1993 in Gender in Cross-Cultural Perspective. Leacock. ed. . Pp. Problems. Eleanor B. Eleanor B. Robin. 1977a. Englewood Cliffs. Krige. Donoa. eds. Eleanor B. 1954. Kaufman. Conceptual and Historical Problems. Emily.Lees. Modell. 1990. and Society. C. Myths of Male Dominance. Signs 2(3):612-627. Kinship with Strangers: Adoption and Interpretations of Kinship in American Culture. SeaUle: University of Washington Press. Marian Lowe and Ruth Hubbard. In Battered Women as Survivors. BretteIl and C. Louise. MA: Bergin and Garvey. 1996. Nancy. Yeorbook of Physical Anthropology 43:1-11. 1993. Leacock. Kathleen. Washington. Lancaster. Signs L Lamphere. Urbana: University of Illinois Press. Alison. Women in Perspective. 1987. ed. and Gender. 1994. Englewood Cliffs. Annals of the New York Academy of Sciences 285: Issues in Cross·Cultural Research. Jaggar. lthaca: Comell University Press. 1983 [1975]. Women. Bloomington: Indiana University Press. Postmodernist Turn in Anthropology: Cautions from a Feminist Perspective. Louise. Pp. I ! L Hrdy. New York: Monthly Review. 1994. D. [1980]1987. The Woman in the Body. Reflecting. eds. Carol. eds. Eteanor B. Debra Renee. Leacock. lane. eds. The Movement Against Clitoridectomy and Infibulation' in Sudan: Public Health Policy and the Women's Movement. and I. Catherine. Sarah Blaffer. In Gender in Cross-Cultural Perspective. Anthropological Institute 89:23-34. Eleanor B. 1989. New Gmenbaum. Caroline BrelleIl and Carolyn Sargent. Introduction. Leonore Adler. Leacock. Harding. Cambridge: Cambridge University Press. Harrison. Patricia Sharpe. South Hadley.. Claudia Card. Whose Science? Whose Knowledge? Ithaca: Cornell University Press. Grewal. Pp. ScatteredHegemonies. NI: Pr~ntice 15(1):7-33. 1-14. Elmsford. Pp. Leacock. 1991. 29-83. In Pioneers of American Anthropology: The Uses ofBiography. 1981. Sandra. Leacock. Louise. Origins of the Sexual Division of Labor. 1993. Sargent. Journal ofthe ROYaJ. 1983. 1959. eds. Ideologies ofSex: Archetypes and Stereotypes. 1966. New York: Routledge. New York: Routledge. Minneapotis: University of Minnesota Press.tL~. The. and Pat Caplan. Pp. Jarrett. Krige. Primate Visions: Gender. Harding. 1981. Mascia. Cambridge: Harvard University Press. NY: Pergamon Press. Rewriling: Teaching Feminist Methodology. Situated Lives. Review Essay: Anthropology. Jacobs. The Review of Education/Pedagogy/Cultural Studies 18(2):165-174. Eleanor B. LUrie. 1991. ed. Lutz. 618-645. 1997. Nalure. Development. 1992. 1987. 1990. New York: International Publishers. Michelle Rosaldo and Louise Lamphere. lane. . EIIen. Lamphere. and Nature in -the World ofModern Science. The Montagnais "Hunting Territory" and the Fur Trade. American Ethnologist 17(4):611-627. and Helen Safa. Race. Washington. Psychology Todoy 7(4):3(}-36ff. Culture. Inderpal. Haraway. Louise.. Eleanor B. Pp. Women's Work. ·Sue-Ellen. Lee Ann. 1977b. 1994. 243-255. Theoretical and Methodological Issues. Pp. Ithaca: Camel! University Press. 67-76. The Domestic Sphere ofWomen and the Public World ofMen: The Strengths and Limitations of an Anthropological Dichotomy. Medica/Anthropology Newsletter J3(2). The Erasure of Women's Wriling in Sociocultural Anthropology. Martin. eds. lune Helm. Living Poor: Family Life Among Single Parent. Eleanor B. and Conflict among Women in Domestic Groups. In Frederick Engels: The Origins of the Family. Social Problems 41(1):30--49. Women in Early American Anthropology. Private Property. 1986. Lawrence: University Press of Kansas. I L Lamphere. DC: American Anthropological Association Memoirs 78. Interpreting the Origins of Gender Inequality. 1943. 1989. Eileen lenson. or Whatever Happened to Fathers? Social Research 44(2):235-259. A Feminist and Evolutionary Biologist Takes a Look at Women. 1982 [1993J. Lila. London: Oxford University Press. Latin American Perspectives 4(1-2):8-17. ed. Culture. York: Routledge. Leacock. eds.

Claire. Silverblat~ Irene.. Karen. Peggy Reeves. Pp. Paula Rothenberg. Shostak. Pulling Train. 186-213. eds. and Society. Berkeley: University ofCalifornia Press. In Women Writing Culture. Carol. Women ofTroplcolAfrlco. Ortner. Pierette Hondagneu-Sotelo. Rcinharz. Signs 4(3):497-513. Toward on Anthropology of Women. 157-210. Women and Change in Latin America. New York: Harper and Row. 33-48. Patterson. Jtapp. Mountain View. and Violence. Narayan. Okeyo. Lyon. Leith. Boston: Allyn and Bacon.. See Brodkin. In Toward an Anlhropology of Women. Ruth Behar and Deborah Gardon. L10yd Holliday. Anthropological Studies on the Status of Women. 28-38. In Toward an Anthropology of Women. pp. and Society. Barbara Syke~ and Elizabeth Weatherford. 1995. 1990. and Maria Patricia Fernandez-Kelley. Stanford: Stanford University Press. Albany: State University of New York Press. Rosaldo and Louise Lamphere. Pp. Uneven Development: Class. Rayna. 1977. 1971. Women. In Toward an Anthropology of Women. Culture. Naomi. Ortner. Slocum. Denise. The Position of Women in the Major Departments of Anthropology. New York: Monthly Review. 1991. 36-50. Rayna. Sexual Stratification: A Cross-Cultural View. In Race. Women and Men and the International Division of Labor. Annual Review in Anthropology 6: 181-225. RHey.. 1988. and Gender in the United States. Gender and Sexuality in Hierarchical Societies. eds. Pp. ed. When Does Life Begin? A Cross-Cultural Perspective on the Personhood of Fetuses and Young Children. The Traffic in Women. In Women and Colonization.ed. Sacks. Sanjek. Annual Reviews in Anthropology 17:427-460. DC: American Anthropological Association. Definitional. 41-56. 1974. MA: Bergin and Oarvey. I. William Haviland and Robert Gardon. 1967-1976. Rayna [Reiter]. Ornvedt. 1977. 1975. Alice. Peggy Reeves. In Through the Prism of Difference. and Action Considerations. Pp. Gayle. 1-20. Rosaldo. Sandra Morgen. Leacock and Helen Safa. .230 GAlLEY Feminist Methods 231 Morgan.. Critique of Anthropology 3(9-10):5-24. Mona Etienne and E1eanor B. New York: Random HouselVintage. Race. Sanday. 1989. Pp. 1986. Rubin. Beth. Rapp. 1985. Cambridge: Cambridge University Press. 1980. Female Power and Male Dominance: On the Origins of Inequality. Stack. Feminist Issues (Fall): 75-94. Peggy Reeves. ed. Pp. Sanday. A Woman Scorned: Acquaintance Rape on Trial. Nash. Karen. Sherry. Pp. The Hague and Paris: Mouton. 1997. New York: Monthly Review. MorgeD. Banered Women of Color in Public Heallh Care Systems: Racism. Toward a Social History ofArchaeology in the United States. 1977. Aboriginal Woman: Male and Female Perspectives. eds. eds. Berkeley: University of California Press. Rayna 1 . New York: Monthly Review. New York: New York University Press. Cambridge: Cambridge University Press. See Rapp. M. 1981. New York: St. June. and Michael Messner. Marjorie. 1975. Rayna Reiter (Rapp). ed. In Ta/king about People: Readings in Contemporary Anthropology. Martin's Press. MA: Bergin and Garvey. 1981. ed. Rohrlich Leavin. Rayna Reiter (Rapp). ed. Ruby. 110-126. Thomas. Fraternity Gang Rape. 1974. Pp. 1993 [1989]. 1 Silverbiall. Saliy. ed. Peggy Reeves. South Hadley. New York: Praeger. 1980. and Louise Lamphere. Achola Pala. Pp. eds. Washington. Is the Female Husband a Man? Woman/Woman Marriage among the Nandi of Kenya. Regina Smith. Is Female to Male as Nature Is to Culture? In Woman. 1978. 1996. Pp. Nisa: The Life and Words ofa !Kung Woman. Leacock. Pp. Woman the Gatherer: Male Bias in Anthropology. Language Learning Through Interaction: What Role Does Gender Play? Studies in Second Language Acquisition 13(3): 343-361. 1974. I Reiter (Rapp). New York: Monthly Review. Richie. In Women's Work. Teresa. Maxine Baea Zinn. Rapp. 1992. All Our Kin. Rayna[Reiter]. New York: Columbia University Press. ed. 1988. New York: Doubleday. Women in States. Sanday. Daughters of the Lakes and Rivers: Colonization and the Land Rights of Luo Women. 1983. and Nora Lewis. Quinn. CA: Mayfield. Kirin. Introduction. Ethnology I9( I):69-88. 373-393. I I l 1 . 121-129. Rayna [Reiter]. Feminist Studies 4(3):37-61. 1978. 1975. Sanday. eds. eds. June. 1995. 1981. Oboler. In Sexual Meanings: The Cultural Construction ofGender and Sexuality. American Anthropologist 80:894-904. Class... Stanford: Stanford University Press. GaB Huizer and Bruce Mannheim. South Hadley. Schlegel. Roger. and Gender in the United States before 1900. 359-409. Sexism. 1977. Eleanor B. Review Essay: Anthropology. I . Shulamit. Participant Observation. New York: Oxford University Press. 396-402. Pp. Reiler. ed. i I I Paulme. 1978. The Search for Origins: Unraveling the Threads of Gender Hierarchy. and Valli Kanuha. American Kinship: A Lesbian Account. ed. In Gender and Anthropology: Critical Reviews for Research and Teaching. Pica. Michelle. Gai!. 1975. Andean Women in the Inca Empire. Irene. 67-88. Feminist Methods in Social Research. eds. Sherry. Sandra.. 1995. Nash. eds. I Mullings. Culture.ln The Politics ofAnthropology: From Colonialism and Sexism Toward a View from Below. On the Participant Study of Women's Movements: Methodological. ed. Harriet Whitehead and Sherry Ortner. Fort Worth: Harcourt Brace.

and Zenong Yin. Wylie. New Guinea. 207-225. 1991. :. Women and Language 20(2):35ff. Yanagisako. Bonnie. Naturalizing power. Sylvia. [987--91. loan Gero and Margaret Conkey. Austin: University of Texas Press. Annual Review in Anthropology 26:59[-621. and Mlchael Messner. In Through the Prism ofDifference. [996. Pp. Gender Theory and the Archaeo[ogica[ Record: Why There Is No Archaeology of Gender. and Carol ri)aney. Race. New York: Roudedge. Chinese Village Women as ~~~ual Anthropologists: A Participhtory Approach to Reaching Policymakers. 79-85. Pierette Hondagneu_' Sotelo. and Sherry Ortner. Tula. [993. Reproducing the Future: Essays on Anthropology and the New . Adrienne. Alison. Sex and Gender in Cross-Cultural Fieldwork. [da. Pierefte Hondagneu-Sotelo. Weston. and Lynn Stephen. Adrienne. Tannen. Van Alien. Gathering and the Hominid Adaptation. Sex. Awa. Hagen. Susser. . Different Visions: Gender. trans. Gender in Research on Language: Researching Gender Patterns in Classroom Discourse. and Ethnographic Responsibility. Mt. Gathering Stories for Hunting Human Nature. The Construction of Poverty and Homelessness in U. Feminist Ethnographies. Telling Women's Lives: The New Biography. unlit. J991. Tesol Quarterly 30(2):341 ff. New York: Columbia University Press. Whitehead. [986. Walley. Carol. and Xiuang Vue Ping. In The Dual City. Mary Ann Surris.tivist from El Salvador.logy 14: 307-232. Gays. Lyon. Tannen. and Morality in Colonial Asia. Kinship. 163-194. In Enger. Zihlman. Postmodernism. Urciuoli. and M. Women and Social Movements in Latin America: Power from Below. J995. Austin: University of Texas Press. Deborah.cds. 1997. Annual Review in Anthropology 25:411-435. Ida. Tucson: University of ACizona Press. Hear My Testimony: Maria Teresa Tula. Anthropology. Tony. Pp. and Gender Relations in a Global City: New York. and Videotape: The Prevention of AIDS in a New York City Shelter for Homeless Men. Ida. Gender and Discourse. New York: Oxford University Press. Visweswaran. Christine. Strathern. Cambridge: Ba . eds. Urbana and Champaign: University of Illinois Press. . London: Pluto Press. Carnal Knowledge and Imperial Power: Gender. and Mary ElIen Conaway. Philadelphia: University ofPennsy[vania Press. Deborah. loan. cd. The Separation of Mothers and Children. Annual Reviews in Anthropology 24: 525-547. Social Science and Medicine 42:1391-1400. Searching for "Voices": Feminism. 51-102. Maxine Baea Zinn. Identities: Global Studies in Culture and Power 5. 1981. 1972. Strathern. Fieldwork in GI/and Lesbian Communities. Inequality. and the Debate Over Female Genital Mutilation. Berkeley: University of California Press. Nl: Rutgers Uni¥lrsity Press. 31-56. ering Archaeology. Feminist Studies 11:364-377. ed. London: Seminar Press. ~ 232 GAlLEY Feminist Methods 233 Stack. Pp. Gonzalez. Pp. 1998. Judith. In Gender at the Crossroads ofKnow/edge. 1996. Gender and Archaeology. 1978. Cities. forthcoming. 1'95. Karyn. J996. Marilyn. "Sitting on a Man": Colonialism and the Lost Political Institutiods of Igbo Women. Susser. Mollenkopf ' and M. Pp. A Thrice-Told Tale: Feminism. B1ackwell. eds. and Nancy Tanner. Violence. Reproductive Technologies. 51-57. Rita. Stanford: Stanford University Press.~ L L L . and Michael Messner. I? Through the Prism of Difference. Critique of Anthropology 3(3):267-285. Susser. 1996. L L L L t I Stephen. Susser. Language and Borders. eds. Clitoridectomy and Infibulation. eds. J. Families We Choose. Wellhouse. Staler. Harriet. Ka• . Human Rights A. New York: Routledge. and Moral Reasoning. 1994.. Ida. 1992. 1994. . Culture. Creating Family Forms: The Exclusion of Men and Teenage Boys front Families in the New York City Shelter System. 199~ Politics and Anthropology. Ida. 1986-1996. Maxine BacaZinn. L I I L Susser. Marjorie. In Speak Out. In Female Hierarchies. 1992. Women in Between: Female Roles in a Male World. Wolf. Kath. eds. Wagner-Martin. 1972. eds. 1985. Zihlman. Lionel Ti~~ and Heather Fowler. New Brunswick. 1997. Boston: Allyn and Bacon. New York: Russell Sage Foundation. Ann. Cambridge: Cambridge University Press. 57-87. Kath. Weston. eds. [991. Castells. 1997. Different Voices. Canadian Journal of African Studies 6: 165-18J.L L .S. [99 I. [997. Pp. Maria Teresa. Micaela di Leonardo. Black Sisters: Feminism and Oppression in Black Africa. Thiam. Vincent. Sexual Meanings: The Cultural Construction of Gender and Sexuality. Wright. Marilyn. Chicago: Beresford. Whitehead. Boston: AlIyn and Bacon. Cultural Anthropology J2:405-438. 1986 {l978]. 1997. Lesbians. "Peter Pan Isn't a Girt's Part": An Investigation of Bias in a Kindergarten Classroom. 1997. Dorothy Blair. 1992. Pp.· . W Caroline. Se/[. Medical Anthropo. 1994. Drugs.

and new jonmals such as Public Culture (since 1988) and Identities (since 1994) are devoted. Others argue that it presents a major challenge to rethink radically some central assumptions in anthropological and other social science research traditions. mainstream journals likewise pay frequent attention to them. Some are wary of it on 235 -. Older. to issues and phenomena relating to them. this is an area where different research orientations and objectives overlap. Moreover.ULF HANNERZ • Seven Transnational Research l l l. although I emphasize the considerable methodological continuity between it and other kinds of social and cultural anthropology. and engagement with global and the transnational issues often becomes linked with general theoretical issues of modernity and poslrnodernity. In this chapter. I also map the main genres of transnational research. Some suspect that it's all a fad that (they hope) will soon disappear or that will take a more modest place in the wide array of possible research interests. Then I present some practical and methodological concerns that may become particularly important in transnational anthropology. I Transnationality and globalization have become increasingly prominent in anthropology since tbe late 1980s. I lI - l . concentrating on contributions by anthropologists. They are recurrent foci at professional meetings. I "i i 1 'l l l Transnational Anthropology Past and Present Reactions to the recent wave of transnational interests in anthropology have varied. I comment briefly on conceptions of globalization and transnationality and their place in the development of anthropology. to a large degree.

comment that this is "reinventing anthropology"-in the same sense as reinventing the wheel. \ . . but fo\ m~:~::':~ey frequently shade into certain genres as being partIcularly Im~ortan." or modernization in the decades around tnidcentury. rm transnational is thus meant to draw attention to t~e '. 2 Long-distance social and cultural interconnectedness may.~~~. these aren't all the same. kinship groups. with long memories. indeed. Different dlSclplmes may an ..'SlOt~i d is also more a matter of cultural implications: e .Vis-a-vis ?ne another. in an attempt to restrict the use of the latter term to contexts where states appear as corporate actors Some Genres of Transnational Anthropology b d 'b d as transnational in one sense or Anthropological research that may e eS~I. It is now widely recognized that globalization need not mean global homogenization.hey actually refer to may be states-pohtIII1bigui ty mheren ~ 0 f . Others.in . However." "social change.ic idea of comparative aspect of transnational researc. 1 d' ' . is again apparent here. . tral\Snational is expected to encompass lvers ?nth. so~ial movements. trends that can proceed only if cultural islands are broken up and rendered compatible with the flow of ideas. commodities and labor through them" (Shryock 1996:30). have intensified and broadened its scope during the twentieth century. and industrial enterprise is hardly fair.' Without reviewing the varIety of eXls~mg rese 'erve as examples.undarte:: s~c~hat:~ What . If we prefer to limit the use of the term "globalization" to processes with a literally world-encompassing or at least transcontinental scope.h. will turn them into something altogether new. with "resistance" to alien· influence and domination as a key idea. national bo. we ma d . of expanding international markets and the delocalization of industrial production. on a s~alIa 1. One tendency is to contrast transnational and international. The tendenCYn~iOn) is a nation-sta. yet the recognition that the world is not a mosaic of bounded entities. existing in isolation from one another. but not everything that is part of it may contribute very much to a usable past. t~~: ~. commercial. We could construct an intellectual history oftransnational anthropol· ogy. but in some ways it'~ as old as humanity itself. may draw attention to ~ cecta.autolntomdosuStoUnri~~~t~t: s~~~ i:~~. the global ecumene of connections and cultural flows may be a very old thing.~. often m the framework of the go . and programma!ic natur~. e r d' ity of meanings and meaningful forms .n~~~~~~ ~~ well as place. they argue. .a~c~1 work we can identify conceptual. for that matter. and products did not appear all of a sudden in the twentieth century. consisting of :th?lc. its characteristics need to be specified with reference to time as studies are· n?t by definition implicit comparisons are often mvolved. it has rather been a variety of critical anthropology. The suggestion that the anthropology of globalization and transnationality in its recent form has been merely an acquiescent scholarly annex to expansive Western financial. ommunities defined on the basis of common ical units-rather t an na lOnS--<lt conflate these two and to assume that every culture and history.op:~~:." It is equally true that globalization may have varied through history.se kinds of ambiguities in their own ways. and only ignorance: or cultivated intellectual amnesia.. This is one reason why one recurrent theme in anthropology and related fields has been the opposition and interplay between "the global" and "the local. Consequen y. also involves ongoing interconnectedne~s. one commentator proposes that it may be "the ideological partner. More often than not.' This also implies that transnational events. due in large part to new technologies oftransportation and communication. Obviously. structures. an interest in such phenom~na ~nc u es corn arison is a prominent or national adaptations and contextu~hn:~on~:~ :~e clas. What ~ ally distributed in a manner that complex of phenomena can be seen as transn~ 10~ction exchange or mobility. have been anthropological concerns for a long time. ethmc ~.~ional is something tha~ cros~es.::o~ in L L L L "culture contact. m . :~:~iO~n :~~:~er. in form and intensity. f extending across many state boun anes. then the term "transnational" need not always be equally inclusive. The te kinds of actors-individuals.' dients of varying strength. On e 0 e. Some comments on the range of conditions and phenomena it refers to may be useful. now or in other periods. ~velryl Yfind entities like "transnational nations. " I bal mosaic" and perhaps analyzing local life.e. to carve out its own niche in the academic marketplace. . dIaSPOra~-~~:~~. tn~e a con~ern with c~ntrasting local Frequently. or even its last few decades. The global and transnational. "long-dIStance. The transnatlona en Th . ther-is clearly not present. poups. ~ t been of a theoretical. arch I will comment on each other. perhaps combined with a new generation's anxiety . firms. In some ways.~: ~~~~~i:S betw~en politically defined. is a common denominator. a ." tl and paradoxlca y. or with aCCUlturation.Transnational Research 237 L L 236 HANNERZ L L L I L political grounds.' state (and.s~~:~r. processes. Global or transnational interconnectedness. We can recognize transational and global themes in anthropology's past in the preoccupation with diffusion early in the twentieth century. f ·ties and relationships that transcend '~'.et~~e~h~a. anthropology-that units are mdependent 0 one ano ~::s:~r~nal ca~~~a~:::. A large part 0 It as. in trans- In . is rather too complex and multifaceted for any simple moral or intellectual stance of being for or against it.• pawing mvolvement of other . another takes many forms. each of these genres here and point to a few studIes that can s Communities Open to the World h I we were intent on portraying and [n an older tradition ~f ant ropo ogy.

concentration of concrete research activities even as we atch for foreign social. word reSIstance IS used to refer to so many practices that it risks lOSing some of Its potency.nd the P.6 Studies m popular culture have made important contributions to such underst~ndings of transnational cultural process. though.). In the working perspective toward the interrelations of "the local and the global. music is one example.238 HANNERZ Transnational Research 239 ?rie~ted tow~rd '~he ethnographic presenl. and the places are somewhat peripheral within the global order.. technologies. This theme suggests that as previously separate cultural Ite~s (Ideas. or at least tntranatlonal. system :heory. even exemplary. There is. also a related tendency to use border and borderland concepts more metaphorically. but the idea comes up often !hat hotels. Terms such as "hybri~ity." In that context. e Departing somewhat from this stance. that are intensely involved in mobility and in !he encounters of varied kinds of mobile people and where this is quite central to their organizational characteristics. Robert Redfield's A Village that Chose Progress (1950) or Margaret Mead s New Livesfor Old (1956) come to mind as classic examples. political. marginal. or cultural influence~. aesthetIC forms. and Sulayman Khalaf(l992) has discussed Arabian Gulf societies as sites with a lot of transiency. An interest in world's fairs. zinc buckets. l l l l l l l -1 Border Studies Research on border communities and border areas might be seen as special cases of communities open to the world-there is a concern with local life in a single place. as Robert Alvarez (1995) has observed in a recent review. Gupta and Ferguson 1997. already mentioned. sedentary. Waterman's (1990) ethnography of NIgerian jilj'. and similar institutions deserve more attention. tlOn. com~Unttles open t~ the world"-a fairly continuous concern of anth~opolnglsts gOIng back to the times of acculturation and modernization studies at ml~century. new inventions and syntheses of various kinds are generated. This has often been a response to the tendency. economic.lSregar? the presence of the district officer and the missionary and the avaIlability of Imported matches. Under the influence of recent !heorizing about modernity and postmodernity. to assume that globalization equals global cultural homogenizati?n. centerIng ?~ domInation and confiict between core and periphery. There can still be a single local. we may have been· Inclined. and radio batteries in th vIllage market.. business and leisure weave together various circulating populations with kinds of 'locals'. Eric Wolfs magisterial overview Euro a. and so ?o) c~me IOta contact through greater transnational interconnectedness. however. These communities also tend to be nodes within transnational social and cultural processes. Trans localities Most studies of communities open to the world deal with places where most people are. There are sites. it does not quite bnng the study or long-distance interconnections into the recent past and the presenl.n~lSts that the study of local life cannot ignore the development of the gl?bal political economy over the past 500 years or so. . recently exemplified in several disciplines.' Anot~er recurrent theme can be described as "cultural creativity. No doubt a great de~1 o~ rich a?~ illuminating ethnography has been generated here. after all. On~ theme in t~e e:hnography of communities open to the world has been that of reSIstance: WhIle Influences reaching in from the outside appear massively ~owerfu!." suggesting that "many such locations create complex conditions for the production and reproduction ofIocality. site of such studies in anthropology and other human sciences. Appadurai 1995. Olwig and Hastrup 1997). I have tried to outline the characteristics of world cities and their cultural processes in related terms (Hannerz 1996:127ff. of course." too. or sheer mlSunderstandmg so as to blunt their effectiveness.Research o~ the openness of communities to the world can obviously remain qUIte close to tlll~e-teste~ anthropological fie Id practices. organizational forms. the variable constructions ofIocality through social and cultural action have come under more active scrutiny. satIre." a~d "~reoliz~tion" have Come to the fore in these writings. Hannerz 1996:25ff. to ." There is little ethnography of translocalities yet.. but with its particular quality offacing in at least two different directions and being shaped by differences between the respective national contexts. a\though the. work. as have !he discipline's traditional conceptions of "the field" (see. !here has been a tendency to take both categories somewhat for granted.e0ple Without History (1982).d. Arjun Appadurai (1995:216) has described them as "translocalities. there are accounts of what we m d e'b" " ay SCrl e as . coming closer to ethnographic conce::::' SImIlarly l. or interstitial conditions.local people r~spond throu." not necessanly totally separate from that of resistance. evasion. More recentl such studies have also ?rawn inspi~ati~n from Immanuel Wallerstein 's(1974) worlr~ . where glob~ltz~tlOn IS Viewed In tenns of a continued production of new diversity. may have had a start in anthropology with Burton Benedict's (1982) historical study of the San Francisco Exposition of 1915. l l I v: l .. allowing "the local" to take on a quality of primordiality. for example. largely In the terms of political economy. The border between Mexico and the United States-with its highly diverse cross-border linkages and contrasts between countries in different stages of development-has been the most prominent. in which ties of marriage. subversive interpreta.gh confrontation. airports. to refer to culturally and socially liminal.

this has become rather more of a standard operating procedure. 1992b) and Alhwa Ong (1992). Anthropologists have devoted some attention to the activities and the impact of transnational (or multinational) corporations. and exchanges of consumer goods. James L. The work of Roger Rouse (1992) on what he terms a "transnational migrant circuit" including. North Africans to France. I suspect that it's not only a matter of anthropologists demanding more of themselves and each other. but more often now. there may be more coming and going. she says. ~ I L I L migration chain. also of a more metaphorical nature. 1994). These certainly make up a large proportion of the entire migrant population. Transnalional Corporations and Occupations Much of the growth of transnational structures and relationships in this century has taken place in the realm of work. often by labor migration and exile. the people at home and t/Jose abroad (perhaps somewhat problematic categories) thus come to form a single coherent. Koreans. in another early publication.is hardly a sh~ly defined conceptual boundary between diasporas and transmIgrant cOmmUnIties.I. drawing on his understanding of Central African mining companies.) exemplify current diaspora ethnography with their account of ritual life among Iranians in North America. dispersed ethnic or national communities are being continuously generated anew. . it is regarded as both desirable and practicable to do such research at both ends. Michael Fischer and Mehdi Abedi (l990:253ff. and so on. "within a few yards of my office and a ten-minute subway ride from my apartment. activities and patterns of life encompass both their host and home societies" and whose "lives cut across national boundaries" as "transmigrants. migrants and those who remained home can stay in rather close touch through return trips and visits from home. and state agencies (not least those regulating immigration). among others." They draw on their own research experiences in Southeast Asia. although spatially dispersed. A prominent aspect is the concern with the embeddedness of the relationships among the transmigrants themselves in a wider setting. but that it's also related to the fact that transnational migration has somewhat changed character. increasingly. also illuminates this tendency. sending and receiving.. social field. or may have been. (1992:1) refer to those migrants "whose networks. the municipio of Aguililla in Michoacan. It may be. discussed the corporations in general tenus and reported on some exploratory fieldwork at a site. and interactions and exchanges of various kinds. for example. there has been one attempt at historical and theoretical overview (Hunt 1987) as well as more intense ethnography. however. Indians. internally very hetero" geneous. not to mention via telephone calls. media. In the twentieth century. and the United States to formulate a more general research perspective (see also Basch et a!. Increasingly. In Between Two Cultures (1977). In later decades. the Caribbean. Liisa Malkki (1995) has recently reviewed refugee studies as a field of growing anthropological concern. Turks to Germany. nit is impossible to gain a true picture of immigration as a process without investigating the people and their families on both sides. ~ L .' A large proportion of work on transnational migrants centers on working-class migrants." Since then. migration has increasingly become transnational: Pakistanis and West Indians to England. Ong 1987). and more West Indians to the United States. Schiller et a!. but there is likewise Anthropological studies of migration in the 1950s and 1960s were mostly concerned with intranational (or intracolonial) mobility-in large part with African. have paid some ethnographic attention to such groups. in the migrants' own country of origin and in Britain. On the other hand. 240 HANNERZ Transnational Research Migration 241 L L L I I ! ~ i " L . James Clifford (1994) has recently explored the diaspora concept and its place in the contemporary landscape of ideas. and Redwood City in California's Silicon Valley. Yet. category of migrants in the late twentieth-century world are refugeesoften with a very different relationship to their areas of origin than that of the typical transmigrant. There was certainly always some return migration and circular migration. such as those of Jews and Armenians. involved mostly in industrial or service occupations. faxes. 1977) was one of the earliest commentators within anthropology. But we may think of the typical djaspora as one where generations have already lived away from the old homeland and where the relationship to that homeland is somehow problematic. Vet another large. many involving corporations' Third World manUfacturing sites (see. inaccessible. mostly at the receiving end. between the various outposts in other countries. in particular. class structure. of immigrant communities and their involvement with the surrounding society. . Mexico. according to his own experience (in studying a Hong Kong Chinese lineage). June Nash (1979). Watson described it as a unique feature of the book that "all contributors have had field experience at both ends of tile considerable transnationaI mobility among highly skilled." Since then. Diasporas There are some very old diasporas. for instance. In considerable part. Johanna Lessinger (1992a. Alvin Wolfe (1963. higher-income people involving "brain drain" as we~1 as considerable capital movements. There . the studies of migrants have been matters of single-site research. Asian. involving." He suggested that. Mexicans. where there is particular reason to attend closely to the institutionalization of the refugee category through the activities of national and international agencies. especially as the former term has grown in popularity and come into expanding use. or Latin American villagers becoming townspeople.

Yet another obviously transnational occupational grouping for which an ethnography is available is that of media foreign correspondents. we glimpse the author monitoring the news groups "Soc. thIS remams the d?mlOa~t relatlonal form.tiOn­ ships. On the one h?~d. and even those engaged in "anthropology at home" tend to be continuously and consciously involved with colleagues in other countries. the computer company. l l i . Another Swedish anthropologist. homogeneous standards and the relative marginality of language in instruction and performance. Nelson Graburn (1983) offered a general spective soon thereafter. and at the Swedish branch office in Stockholm. there . people cross boundaries without carrylOg the" bodies along. OrvarUfgren's (forthcoming) study of European chartertounsm in the Mediterranean area offers a view of the latter type of leISure travel. the communities-open-to-the-world genre.J j 242 HANNERZ It is. to some extent. which. doesn't necessarily deal with large corporations. She was able to follow the interpenetration of local life with longdistance communication and visits and to see how the California-inspired. Mark Pedelty (1995) concentrated his study on war correspondents in El Salvador in the early 1990s. I Media As media technology diversifies.wert: and Frederick Errington's (1991) study of the Chamb~i of Papua New Gumea IS ~n example of an ethnography.. While the anthropological exploration of cybe~P?ce has har~ly more than begun there is a quickly growing interest in the pOSSIbIlitIes of new lOformation techn~logy. there is also a globalization of leisure: Th. deliberately constructed corporate culture of the company's management was modified by employees in offices elsewhere. By so doing. anthropology of tourism has been an expanding fidd for some lIme ~nd much of . somewhat in line with Tourism If there is a globalization of work. on the organizational culture of Apple. The ongoing debate about outsider and insider anthropologies-for one enlightening contribution. and the journal Annals of ~ourlSm Research has been . Going on tour or training in other countries is part of the round of work.d IS of life and looking for more "authentic" experiences. teaching awareness of cultural differences and thus sharing some of the ideas and concerns of anthropologists. is often made up of dancers of many nationalities). .Islam" on the Internet and worrying somewhat about the ethICS of being an anonymous reader of other people's messages.Religion. In an essay by Davld ~dwards (1994) on the nature of "fieldwork" on (rather than "in. it is unlikely that thIS Will not become a significant genre in transnational studies (unless it is simply absorbed as an aspect of other genres). conversely. On the other. It's also about occupations. fac~-to-facerela.--.l . In many kinds of transnational anthropology. see Kirin Narayan (1993}-may be viewed against this background. occupational communities. Deborah Ge. The volume edIted by Valene SmIth (1977) was an early contribution to this genre.Transnational Research 2. Arturo Escobar (1994) and Gustavo Lins Ribeiro (19~7) suggest general visions of an anthropology of cyberculture.. i "" Cyberspace Classic ethnography involves people in physically pres. however. too. the dividing line between cyberspace and media studies as types of transnational anthropology may become rather blurred. and occupational life-styles that are smaller in scale and more individualized but involve contacts and mobility across national boundaries-experiences that have become common for at least some practitioners. The dancers' careers often take them through companies in several countries (so that a company..CuI. ~~rimportant outlet for anthropological writings on tounsm. of course. by the presence of explicit. m which tourism figures importantly. We can l l l ~. --. toward the end of the war there. I have also considered this group. tourism of individuals or small groups."nt. somewhat in a commumlIes-open-to-the-world style." perhaps) Afghamstan. and a transnational circulation of videos has become important for instruction as well as for promotion (Wulff 1998). it isolates itself from the host society . Finally. there IS ways . One study that involves fieldwork in several places is Christina Garsten's (1994). Increasingly. at the European head office in Paris.. Helena Wulff. i -. A contrast is often made between two kinds of tourists.Afghamstan': and "Soc. clearly. creates its own mstltutlons an e. d' t mass tourism. has recently been studying the transnational occupational culture of ballet dancers. mostly on the basis of biographical and autobiographical accounts (Hannerz 1996:112ff. anthropologists themselves make up a transnational occupational grouping. With a somewhat different range of interests. the contrast between the global and the local enters the picture through differences in styles and working conditions. The transnational anthropology of work.i~ some ways and m ot~er ways it shapes the latter to fit the requirements of the VISItors and those of the mdu~try that handles them. l l l -. Since that technology has such a great cap?city to transcend physical distances and national boundaries. Yet here.).nvlronmen s. A large proportion of them does much of its work abroad. Garsten did researchat the company headquarters in Silicon Valley. Tommy Dahlen (1997) has described the emergent profession of interculturalists: trainers and consultants working mostly in the marketplace.t involves travel between countries. The transnational integration of the ballet world is aided. possible to do such studies in a single site. though. --------. seeking to adapt to local faCIlitIes an. at least by way of reading matter.

This. war. and mterests can be better grasped. and may lead to further action.logy. takes place." however. such study is mostly part of that anthropology of commodities and consumption that has developed quickly since the 1980s. family roles. dramatic shifts in interpretations may L L OCcur along the way-as. and consumers. obs~rvations ~f forei?n tourists." is distorted and thinned out.' Its bases and its development to date have been extensively and illuminatingly reviewed by Debra Spitulnik (1993). T~ansnalIonal media studies need not be limited to studies of how particular media products of one country are received by audiences elsewhere. intensity. Is an example. for example. as centeringmore often on those communication technologies that tend to involve more asymmetrical relationships between special_ ized producers/senders and relatively large audiences. adapted to local contexts.. may be another example of cultural creativity generated by transnational interconnectedoess' o'f hybridity. then. ~nter. local religious debate.) As commodities move over distances and across national boundaries. But the point here is that consuming is itself a kind of producing-interpretation is an active engagement with media content. mentioned above. they are carried through a more or less intricate network of relationships between producers. turned IDto another commodity. relationships." operated. In a competitive market situation. the point has recently been made wIth reference to film and television programs. At present. Yet media anthropology is undoubtedly here to stay and to grow. There is some possibility of a transnational. responses could go either way. Anthropologists may prefer to integrate an understanding of media use into a more wide-ranging ethnographic study where ~he interactions between media experiences and other experiences. around the last turn of the century.'es outside anthropology. A major research interest would also be to learn what these categories of people think of the objects themselves. Laura Bohannan's (1966) well-known essay. anthropologists may at lImes be skeplIcal WIth regard to what passes for ethnography In SOme research contexts. The amount. although Pedelty's study of war correspondents in El Salvador. and their paths of travel. and suffering for our imagined worlds._----245 L L L L L L I 244 HANNERZ Transnational Research think of "mediastudies. but a local production. such a production doesn't necessarily lose out to imported counterparts.v villagers in midcentury central Nigeria responded to Hamlet: is surely a progeDltor of later tra~snational media studies. the bal~ce between lived experience and imagination may have shifted in many human hves in recent times.a stud. b Australian Aborigines and other ethnic minority groups-see Ginsburg (1991):ay be considered as part of transnational anthropo.. in stories of cargo cults in Oceania. "electronic empathy" here. It's not to be taken for granted in the way it frequently has been. Fantasy turns into a major form of social practice as an incre. a sense that those far-away strangers can be understood by our own emotional resour~es. Most often. Another VIew. The assumption that media are "mass media" may be becoming obsolete in several ways. and the activities of these various actors may themselves become one focus of study. In transnational studies. for exam~Je. this becomes part of a critical response to the scenario of global cultural homogenization: eve~ if the same programming becomes available and perhaps popular everywhere. And ethnographic methods have recently grown increasingly important at the consuming end. although they often are. The study by Israeli media researchers Tamar Liebes and Elih~ Katz (l9~0). print.) has suggested that espeCially via the media. She studied how young Muslim women in a town on the Swahili coast of East Africa vie~ Bombay-produced Indian movies on video and integrate such viewing and thelt own related commentary with their educational experience. Although Appadural may be thinking primarily of alternatives that could seem attractive. Minou Fuglesang (1994) used such an approach. too. and breadth of actual observation of audiences and their daily rounds is sometimes very limited. the media may als~ make available scenes ofdisaster. Appadurai (1991: I98ff. Probably. L L L intennediaries.asing variety of alternative ways ofliving are vividly presented.est in t~e actu~1 audience handling of media has made ethnography a popular ~otlOn m medl. Because these movements in space and through varied contexts are often far from transparent to participants. The engagement with the production end of transnational media structures has so far been limited in anthropology.r0adcast on a daily b~is as "infotainment. depending on the wider context of audience interpretations. and life-career prospects. June Nash (1993) has edited a volume of studies that concentrate mostly on the production .. however. was what diffusion ism was in large part about. . . 'o A number of studies of this type deal with art objects-or with objects that Sooner or later come to be understood and evaluated in esthetic tenus. offers striking evidence of this. their origins. Here. t Commodities Studying the transnational mobility of material things is perhaps the oldest genre of transnational anthropology-this. of the. Often technologies and media genres diffuse transnationally. b. To repeat.va~ing interpretations of the TV series Dallas by groups of different natIOnal ongm m Israel. media ethnography has developed mostly at the receiving end of the message flow. . especially in some earlier styles of mass communication research. and electronic media. "Shakespeare in the Bush" about the way Ti.' (Commodities are not always material. after all.. once such pictures are effectively distributed throughout the world. of course. The interest in "indigenous media. _--_. but the tendency to have separate bodies of research for the producing and consuming sides of mediated relationships continues to be strong. suggested by Arthur and Joan Kleinman (1996). is that suffering. It cannot be taken for granted that it is understood in precisely the same way everywhere. The argument is relevant across a range of written. or creolization.

is (in most cases) a network of sites. For instance. When network analysis drew most attention in anthropology in the 1960s and 1970s. there has been a tendency to speak of "multilocale" or "multisited" fieldwork. through marketplaces in towns and cities. In a formulation close enough to a social anthropological tradition. The unit. attention to media almost always plays some part. How. why) in media anthropology to study CNN or the Bombay movie industry. while Steven Sampson (1994) ~ffe~s a ghmpse of how the newly rich in eastern Europe acquire understandings of hfe m affluence from the West. at least as areas of further work and debate. somehow deterritorialized. both at its production and consumption ends? Are all potential sites implied by a research topic accessible?Given that a diaspora or an occupational community may be spread over not two or three.e drawn from this overview of important genres of mformant work. it showed tendencies toward conceptual and methodological involution that limited its attractiveness. to gaIleries in New York. indeed. the sense of problem. Doing Transnational Studies transnaltona~ anthr~pology IS that the latter entails considerable internal diversity. or Jonathan Friedman's (1994: 105) remar~ ?n Coca Cola aod the Congolese. Karen Tranberg Hansen (1994) h~ mvesllgated the sal~ a~d identity-constructing consumption of European and Amencan used clothmg m towns of Zambia. perhaps. and elsewhere. for example. a vlIlage. network analysis may again offer ways of thinking about how units of study are constituted and about the social spaces occupied by varied actors (see Hannerz 1992b). There may be no more or less bounded territory here. and not least in transnational Contexts. available to acceptably comfortable ethnographic surveillance. South Africa. more varied and complicated than in the selection of a single site. And. . Yet a relational point of view toward the cohesion of translocal. one that can yield the most desirable combined ethnography? In one way.246 HANNERZ Transnational Research 247 side. and consequently William O'Barr (1989) has interviewed the man in charge of shaping the idea of Coke in Brazil ~alay.•t can hardly be characterized by anyone set of approaches that ~ould dlStmgulSh It from other anthropology. the formulation multisited ethnography somewhat obscures an important fact: The research may need to be not merely multilocal but also translocal. Units of Study . We can expect that in such cases the topic. in what was described above as communities open to the world. (Of course.IIy. But in transnational and other translocal studies. transnational units seems to reqUire a more varied vocabulary of kinds of connections. Serious effort must thus be devoted to an adequate conceptualization and description of the translocallinkages and the interconnections between these and the localized social traffic. be spatiaIly dispersed. the transnational corporation perhaps at headquarters and at branch sites." The practice of ethnography may have to be distributed over several places. and we may argue that such a sense has not been particularly strong in much of recent transnational research. Nancy Rosenberger (1992) discusses how Western home styl~s are presented in Japanese magazines. Such a research perspective could also include some sensitivity to the range of kinds of relationships that occur in modem societies. Would it be desirable (and if so. face-to-face interactions. 1I Other studies cast light on the global passages of foodstuffs-as does Sidne Mintz's (1985) historical study ofsugar. how does one decide on a manageable selection. as acceptance of commodities often coheres into larger patterns. I l I l 1 I I l l II l . smaller-scale structures-Gemeinschaftrather than GeseUschaft. Observation (more or less participant) One co~clusion that can b. among Middle American artisans who supply crafts to the world market C. Migrants are to be studied at home and abroad. But there are choices to be confronted here and constraints to be taken into account. to delineate the field often appears less obvious in transnational studies than m ~lasSlcallocal ethn~graphy. In discussing this.hristopher Steiner (1994) has foIlowed "African art in transit" all the waY-fro~ vlIlages of the Ivory C~ast. The anthropological tradition is strongest in the study of more intimate relationships.) How do transnational ethnographers conceive of their units of study? The units may. At th~ same time. town. Method~l~glca. or ne. will often suggest fairly strongly how a field may be demarcated and where research should be carried out. (Although they can now rarely be disregarded in local or national research contexts either.ghborhood with a reasonably settled population. These may mvolve some characteristic anxieties and significant challenges that need to be identified. there can be exceptions in this respect. li~ ~istoriesJ surveys. and through the hands of West AfrIcan transnational traders. but dozens or hundreds of locations in the world.'ia. Advertising is a conspicuous part of commodity Imagery that operates transnationaIly as weIl. and parts of one's ethnography may have to be between these sites. textual analyses-all the usual ingredient~ ?f the a~th~o~ologlSt s craft-are as relevant to transnational studies as elsewhere m t~e dlsclplme. It would be helpful here to have a developed sense of structure (and of the room for maneuver within it).) : et a series of methodolo~ical issues seems particularly prominent in transnatIOnal research. since transnational units of study are less often fuIly camed by face-to-face contacts. and this points to a significant aspect of what is involved. some tools and techniques may take on a relatIvely g~eat~r centrahty.

however. etc. Sharon Traweek Cultural Competence Ethnographers' unease about the risk of overextending themselves across too many sites in too little time is likely to be further increased if they feel that they also have to cope with more cultural variation. at considerable distances from omnnother..). although the inhabiltmls pass through. are involved in a study. and perhaps there is no sense of loss of cumulativity and repetition at all. as sketched above. Even as such research is extended in space.). One could study transnational families and kin groups. surveillance relationships. deep ethnography either. on the other hand. Marcus (1995: 101) concludes that so . census data and credit card records may allow one party to keep track of another more or less unobtrusively. As transnational fields are constituted. telephone. there are certainly occasions Where the ethnographer's competences face more serious challenges. lt will perhaps not be possible to follow the shifts between seasons. as. to learn through the everyday redundancies of social life. They are translocalities. but a combination of sites can include some where the ethnographer is not alone in being more or less footloose. Finally. are one-sided. _. it's increasingly true in contemporary anthropology. Some such sites are perhaps durable. Direct relationships are those of physical copresence: "primary" and "secondary" relationships. such ways of keeping in touch on the part of the researcher will not necessarily appear as strikingly deviant. according to a familiar vocabulary." (The former are seen as linking whole persons. Irade fairs.. Insisting on carrying out an entirely local study in a site strongly marked by translocal and Iransnational connections would surely not result in satisfactorily complete. They may have other interaction patterns among themselves.. In some cases. "tertiary" ones are mediated by technology andlor large-scale organization.or multisited. Unless we limit anthropological work to autonomous local units (of which there are not so many these days).. with their own internal patterns of long-distance cohesion. A number of points must be considered here. What Gusterson (1997) has described as "polymorphous engagement" may well be the natural adaptation to these fields. the notion of a "complete ethnography" now seem's rather dubious. and practices across borders. In other instances. Is it possible to meet normal ethnographic standards and to have the expected sense of deep involvement if (assuming that only the normal amount of research time is at one's disposal) one has to scatter one's attention over many sites? (1988:3) notes that the most active researchers in the field make up a community of no more than 800 or 1. as may be the case with language skills.. In any case. "Quaternary" relationships. whether single. equally fully in every site. This may be true in transnational fields as well. We can be aware of these as at least potentially relationships between persons. It's also important to recognize that transnational research is not by definition research on large-scale units. a whole series of different bilingual combinations can be involved. 1992) contrasts "direct" and "indirect" social relationships and distinguishes two types in each of these. the definition of the research problem may imply more focused attention on some phenomena than on others. the anthropologist is a transient in a community where most members are life-time residents. Here. for example. studies of transmigrants are among those that have recently taken such linkages increasingly into account (see Hannerz 1996:9Iff. L L L This is certainly a serious question that has to be confronted (see Marcus 1995:99ff. Yet. and in a study of high-energy physicists.. we need to experiment with the distribution of attention. it sometimes involves relatively limited populations. conventions. of which one party is not entirely aware..\ . In the latter. the latter enactors of specific roles. although usually they are mediated by markets or bureaucracies. and are nn doubt embedded differently in a wider social context. Some anthropologists focus on ethnic or national groups distributed over several countries..000 members worldwide. that fieldwork goes on in some ways. or at least a different set of preparations? For several kinds of transnational study this may not be altogether necessary. Does each additional site imply starting more or less anew in acquiring knowledge of another culture. and passages across borders between states.. In an ethnic diaspora. We may at times have to accept some tradeoff between depth and breadth. and no single ethnographer can be expected to master all the languages involved. Information technology plays a large part here. however. Whether in local or translocal research. also involves certain practical problems. When transnational research deals with the fact that states. As indicated above. Others deal with specialized occupational communities that share knowledge._-------------Transnational Research L I L. but in sheer numbers they are comparable to a largish village. 248 HANNERZ . orientations. 249 L L L L L L L sociologist Craig Calhoun (1991. they may themselves be short-term phenomenasports events. an ethnography that disregards tertiary and quaternary relationships may seem quite incomplete. Dispersion Versus Intensity A study that is spread out over a number of sites. the ethnographer may even be around for as long as the site lasts.) Ofthe two kinds ofindirect relationships. for example. the relationship between the temporalities of the researcher and the field also works out differently in Iransnational and multisited studies than it does in classical ethnography. and even a one-year stay is short compared to the involvement of these natives. even as the ethnographers have absented themselves physically from their fields. Many anthropologists maintain more or less continuous contact with one or some handful of informants by way of letter. or e-mail.

). 6 I'Creolization" seems to be the term that has drawn more conceptual and theoretical com~ent in anthropology. In adiscussion of keywords in transnational anthropology. One additional solution should be identified. for example. thus decreasing the risks of thinness and superficiality.'ett'es and cities Now with studies of the transnallonal and the into peasan . In transNOTES national research. see Verdery (1994). ofmobllt . t the rather concrete genre classification here. follow the thing. Appadurai's (1990) distinction between ethnoscapes. . there's a particular reason to consider the alternative of collaborative work. story. an sup~o:e b the Swedish Resesearch Council for the Humanities and SOCial. . Alfred cO::be~(~945:9) argued 'more than half a century ago. One consequence of this has been an increasing frequency of adversary relationships between expatriate anthropologists and those who do their "anthropology at home" in the same place. ProspectIRetrospect It is hardly time yet to offer a more conclusive view of transnational anthropology and its methods. Certainly such a team might be recruited in its entirety in one country or even in a single academic institution. and the Department of Ethnology. I. GalinaLindquist. we will have to accept that ethnographies will be more partial and in some ways more superficial. see. I th ology may also But with transnational research. finanscapes. see Ortner (1995). some have perhaps not been clearly formulated or even identified. for example. Bengt. more new ethnographic genres may emerge and old ones may be This chapter i~ based on studies carried out within the framework of ~he "National and Transnational Cultural Processes" project based at the Depar. Ronald Stade. Hannerz( 1987. I WIS ~o a~knOwledge the valuable exchanges about questions of research practice tha~ I have had m that context with several colleagues in Stockholm: VIf BjOrklund. Ennque Ro nguez Larreta. with shared intellectual concerns so that confrontations over ethnographic authority and informal ethnographic property claims are less likely to occur than in locally or nationally focused projects. be able to spend more time. I global. Fabian (1978:317). follow the metaphor. . it later continued its journey ~f exploratIOn p p t soc. for example. There could be problems of scholarly compatibility within such teams. For one sociologist's attempt to periodize globalization.ultur~ of all Kr 'ty" If at that time the discipline was still mostly preoccupted wtth ~he h~:~:lof ba~ds. 4. Perhaps when studies are conducted in sites and combinations of sites that involve more cultural and linguistic diversity. Hasse Huss. see Robertson (1992:58-59). There are now anthropologists in and from nearly all countries. and ideoscapes as dimensions of global cultural flow ~se~1 als~ Appadurai 1991) may also be seen to relate. mostly English-speaking settings. Ethnography has tended to be a lone~wolf enterprise (with teams of spouses as the occasional exception). medlascapes. Tommy Oahlen. techooscapes. Drummond (1980). drawing on the complementary backgrounds and competences of several scholars. and Margolis (1994:21 ff. and Hel~na W~lff. in fact. recom b me . As new cohorts of anthropologists formulate their projects. must be "the c. and villages.. and Friedman (1994:208ff. h h While not all multlslted researc h he notes the strategically sltuated. Vet it would then be desirable for researchers to acknowledge the constraints on their work.).). tribes. Marcus (1995). has suggested SIX strategies: follow the people.' . Unlverslty. or alleg~ry. For a critique along such lines. most multisited field studies have. . It is also possible that due to personal background or interests. Garrison and Weiss (1987). On the latter. although perhaps somewhat problematlca y. l l l l l J l l l l l l l . Apart from these strategl~s.. Minou Fuglesang. At least some projects oftransnational research could instead involve more cooperative research between scholars in different locations. 5. however. in one partlcll ar way.250 HANNERZ Transnational Research 251 far. Colen (t990). indeed all the research time available. SCiences. an rop . Grabum (1984:402ff. d seem to be forever more research passl 1 Itles. Perhaps we must also conclude that whether carried out in one place or many. y follow the life or biography' and follow the conflict. and 00 "resistance and accommodation:' Marcus (1989. researchers are better equipped to function as scholarly cosmopolitans. Each member of a small team of scholars might.). in one of the sites of a multisite study. 'b'l' . there cou I . some · d And as the kinds of relationships that cross boundaries proliferate. in a variety of contexts. Christina Gar~ten. follow the plot. Ihave tned to Identify some of the ingredients of such a history (Hannerz 1997). Some problems have not been solved. 1992a:261 ff" 1996:65ff. is necessarily transnational. But it seems to be in the nature of transnational research that collaborations between anthropologists from different countries would be another alternative.. 2.ofLu?d. This could also be a way of dealing with other issues identified above.~nk Bor~s~rom. Britt·Marte Thuren. ethnography is the art of the possible and that it may be better to have some of it than none at all. For an illuminating discussion of resistance conceptions in recent anthropology. in a review of "multisited" ethnography..t of SOCial An~hrOPOIOgYd Stockholm University. there is a partial overlap between this claSSification an~ the one which follows. Garcla Canclini (1995) offers a discussion of hybridity in relation to modernity. been carried out in monolingual.tmen. t resting point The ultimate natural unit for anthropologIsts. 7. SlOgle-slte et nograp y. see also. but there are networks of transnational collegial friendship and trust that could be a fundamental resource here. 1992). . it may face the final test of its tools. 3. . Such collaborative work is now facilitated by new information and communication technology (even in the write-up phase).

Stockholm: Almqvist and Wikselllnternational. Fabian. Man 15: 352-374. Annual Review 0/ Anthropology 24:447-470. Bohannan. In At Work in Homes: Household Workers in World Perspective. Annals of Tourism Research /0:9-33. The Production of Locality. and Frederick K. Rober! R. Border: The Making of an Anthropology of Borderlands. Intro~uction: Commodities and the Politics of Value.. Cullural Anthropology 9:302-338. Indigenous Media: Faustian Contract or Global Village? Cultural Anthropology 6:92-112. Reading. eds. Public Culture 5:465-{j04. . 1992. 1990. Cultural Anthropology 9:345-360. 1992:315ff. 1995.. London: Sage Publications. James. New York: Center for Migration Studies. 1994. Pp. Nelson H. ed. I~ Recapturing Anthropology. Notural History 7575(7):28-33. In Caribbean Life in New York City. 33. Craig. and the New World Order. and Stratified Reproduction in New York. Stockholm: Almqvist and Wikselllnternational. Lee. Shellee. 1991. 1994. 1966. 89-118. See. The Anthropology of Tourism. Graburn. 1990. In Social Theory for a Changing SocIety. Appadurai's (1986:41) discussion of "knowledge and commodities" is central h L L L L L L L L r~latIon~hlps nee~ not always be chw:acterized by physical copresence. 1991. J~. H. Pp. Michael M. 1982. 10. 190-210. Pp. Richard G. Annals of Tourism Research l/:393-419. Debating Muslims. Basch. Diasporas. Co1en. 1994. 1995). Madison: University of Wisconsin Press. Santa Fe: School of Amencan Research Press. London: Routledge. Edwards. University Press. Nelson H. Among the lnterculturalists. Public y Culture 2(2):1-24. 9. the present context. Jr. Daniel Miller. Arjun. Afghanistan. 1991. 1991. 1987. ed. Cambrid e: Cambrid e g g Appadurai.6. Wt' es View PresslRussell Sage Foundation. See. for example. In Worlds Apart. 3-{j3. 1995. Pierre Bourdleu and James S . Mineu. Smelser. Miller (1987) and Friedman (1995). Berkeley: S~olar Press. Sutton and Eisa M. 169-189. Gewertz. for exam I Abu-Lughod (1993. and Cristina Szanton Blanc. and this IS certamly of great importance in transnational contexts.deos) allo~ ~uch re~ationships to be carried on at least partly outside the face-ta-face situatIOn. 205-236. Current Anthropology 35:211-223. and Social Integration. 1993. Christina.tnformation Technology. and Carol 1. it should perhaps he noted that primary ·and secondary REFERENCES Abu-Lughod. 1978. Chicago: University of Chicago Press. Appadurai. Fischer. Pp. Hybrid Cultures. C~aig. Veils and Videos. More interpersonal kmds of mformatlon t~chnology (rangmg from letter writing and telephone calls to e-mail and exchanges of ho~e vl. Pp. 11 I ere. Laura. the State. Transnational Research 253 L L I L 8. 1980. Garsten. England: Harwood. Altered Contexts. Burton. . pp. ArJun Appadurai. 235-254. Appadurai. aod Mehdi Abedi. 204-225. Welcome to Cyberia: Notes on the Anthropology ofCyberculture. Lila. Nestor.. Disjuncture and Difference in the Global Cultural Econom . Jonathao. The Cullural Continuum: A Theory of Intersystems. Garcia Canclini. Roger Sanjek and Shellee Colen. In The Social Life of Thmgs. InSocial Change andModernity. Daniel Miller. HansHaferkamp and Neil J. Vivian. David B. 1984. H. Cullural Identity and Global Process. Global Ethnoscapes: Notes and Queries for a Transnational Anthropology. Arturo. Africa 48:315-334. Errington. Constance R. PA: Gordon and Breach. Berkeley: University of California Press. Linda. London: Routledge. Johannes. 1994.\~ I ~ ~ 252 HANNERZ i.S. 1990. eds. IndirectRelationshipsand ImaginedCommunities: Large-Scale Social Inte~ratlo~ and the T~ansformation of Everyday Life. 32. The InfrastructureofModernity: Indirect Social Relationships. Weiss. 1995. Appad~rai. Stockholm Studies in Social Anthropology. Nina Glick Schiller. Popular Culture in Africa: Findings and Conjectures. Apple World. Pp 95. Das. Garrison. Dahlen. 1995) and Das (1995). 1995. Graburn.) and Marcus (1989:24ff. Marcus and Fischer (1986:90ff. The Mexicao-U. Drummond. Washington. Deborah B. Nations Unbound. Clifford. Calhoun. For some prominent examples apart from the work mentioned here. 'Housekeeping' for the Green Card: West Indian Household Workers. ed. Arjun. Friedman. The Anthropology o/Worlds Fairs. Veena.. Twisted Histories.. Escobar. 1995. Tommy.. 198. Benedict. Lila. for example. 191-210. ed.121 . DC: American Anthropological Association. Bo uId er. Fuglesang. Stockholm Studies in Social Anthropology. Chaney. Stockholm: Almqvist and Wiksell International. ~yers (1991) and Morphy (1995) have discussed the Journey of Australian Abongmal art IOta metropolitan art galleries. pp. J. ~rjun. The Evolution of Tourist Arts. Ethnography. Fox. Cambridge: Cambridge University Press. In Counterworks. 1995. Langhome. 1994. Richard Fardon. Pp.. eds. see. 1997. Jonathan. . Calhoun. 1983. Ginsburg. Arjun. Alvarez. Coleman" eds . ed. 38. The Objects of Soap Opera: Egyptiao Television and Ihe Cultural PolItICS of ModernIty. 1994. Friedman. Faye. Shakespeare in the Bush. Consumption and Identity. 12. Stockholm Studies in Social Anthropology. ed. Dominican Family Networks and United States Immigration Policy: A Case Study. 1994. pe. ed. On Soap Opera: What Kind of Anthropological Object Is It? In Worlds Apart. Abu-L~ghOd. n a some~hat simi~a~ vein. Screening Politics in a World of Nations. In. London: " RouUedge.

Olwig. Margaret. London: Sage publications. Pp. The Hague and Paris: Mouton. Vol. 1945. eds. Pp. fronteiras. Adam Kuper. Arthur.. 106-125. Princeton: Princeton Umverstty Press. 1991 . 309-330. Pp. Tobin. Lessinger. . The Dismay of Images: Cultural Appropriations of Suffering in Our Times. New York: Columbia University Press. Ghck Schiller. Representing Culture: The Production of Discourse(s) for Abortgmal Myers. . VIr. 1992. Frances Abrahamer Rothstein and Michael L. Drtner. Johanna. Comparative Studies in Society and History 37:173-193. Hansen. The Ancient Oikoumene as an Historic Culture Aggregate. Johanna. William M. Anthropology as Cultural Critique. Vol ' 645 ' Nma Pacific Rim. and Present' Forms. Berkeley: University of California Press. L~fgren. Limits to Cultural Accumulation: Chinese Capitalists on the ~meri~an . Hannerz. Margolis. Multinational Corporations: Their Origin. 1995. Roland. Md' d Eth~ographies of Late Twentieth-Century Modernity WorldWide. Georg . Tamar. 1995. CT: lAI Press. Joseph J. 1997. Public Culture 2:1-19.--l . Critique of Anthropoiogy 9(3):7-30.1 I Lo~don. 1992. Dealing with Used Clothing: Salaula and the Construction of Identity. O'Barr. e4s. Material Culture and Mass Consumption. Vol. Albany: State University of New York Press. 1992. 1987.Transnatianal Research 255 254 HANNERZ Past Present and Emergent Identities: Requirements for Marcus George E . K trln. Anthropological Locations. Hannerz. 1989. Marcus. Gulf Societies and the Image of the Unlimited Good. Pubiic Cuilure 6:503-523. 62-82. Chicago: University of Chtcago Press. Aboriginal Art in a Global Context. eds. 1996. New York: Morrow. Images of the West: Home Style in Japanese Magaz~nes. Alfred L. 317-375. Akhil. 1992a. A Village that Chose Progress. Gerrit Huizer and Bruce Mannheim. In Research in Economic Anthropology. London: Routledge. The Export ofMeaning. Pp. Dialectical Anthropoiogy 17:53-84. eds.. Orva'. · t 'Sidney W. . . l l l Dng. Barry L. Hannerz. Journal of the Royai Anthropoiogicai 1nstitute 75:9-20. IM'U Morphy. Liisa H. Nash. Development. and James Ferguson. Redfield. 1987. Refugees and Exile: From "Refugee Studies" to the National Order of Things. 1985. 1990. Pp. ulr. . 1996. New Haven: Yale UniverSIty Press. Khalaf. . e E 1995 Ethnography in/ofthe World System: The Emergence ofMultl~ Marcus. Howard.. . Pp. 125-143. . a . Gupta. Annual Review of Anthropology 24:95nd Michael M J Fischer. 1989. Gustavo. Hugh. Liebes. Oxford. G eorge E . Albany: State Umverslty of New York Press. cd. Investing or Going Home? A Transnational Strategy among Indian Immigrants in the United States. Pp. Nina Glick Schiller. Imagining the Whole: Ethnography's Contemporary Efforts to Situate Itself. Ong Aihwa. Pp. Isaac. Nash. George E.In Re-made 'in Japan. F re d . Sweetness and Power. Anthropology Newsletter 38(5):8(}-78. Kleinman. 1997. Transnational Connections. Karen Tranberg. Little Brazil. 0 e~m~1 an Identity. The Global Ecumene as a Network of Networks. In Search of the Virtual-Imagined Transnational Community. Pedelty. Lessinger. Annual Review of Anthropology 24:495-523. Malkki. Culturai Anthropology 6:26-62. Making Hoiidays. The Airbrushing of Culture: An Insider Looks at Global Advertising. Blac we . Forthcoming. and Kirsten Hastrup. . Scott Lash and Jonathan Friedman. 95:671-686. Sited Ethnography. New York: Viking Press. The World in Creolisation. ed. on M ' "Ion. Studying Up Revisited. 1994. . eds. Daedalus 125(1): 1-23. 1979. Routledge: l l ~ I . Ulf. 1992. Sulayman N. hlbridos: Palavras-chave da antropologia transnaciona!. eds. Sherry. 1997. Lins Ribeiro. June.. June. Gusterson. Edwin S. . and Cristina Blanc-Szanton. 1997. London: Routledge. l l . In Towards a Transnational Perspective on Migration. Acrylic Paintings. In. In Anthropology and the Global Factory. . In Towards a Transnational Perspective 19ra . The Appeal of Experience. and Cristina Blanc-Szanton. New York: OXford University Press. 1995. UIf. Robertson. ·· 1993 How Native is a "Native" Anthropologist? American Anthropologist Narayan. Lnndon: Routledge. 1987. 645. ed.' . . l l I Hannerz. 211-239. 1995. . Aihwa. Mana (Rio de Janeiro) 3:7-39. Kroeber. Ulf. l . Anthropology of the Multinational Corporation. Pp. 34-56. 53-80. Berkeley: University of California Press. eds. Oxford: Blackwell. Slim. Annals of the New York Academy of Sciences. 1986. 1992a. Nonresident-Indian Investment and India's Drive for Industrial Modernization. Giobaiization. Spirits ofResistance and Capitalist Discipline. 8. Fluxos. New York. Linda Basch. In Conceptualizing Society. Linda Basch. ' 117 . War Stories. Mark. Africa 57:546-559. Resistance and the Problem of Ethnographic Refusal. Marcus. Chicago: University of Chicago Press. Hannerz. . Cultural Complexity. 1950. Greenwich. 1993. Dame I er. New York: Bergin and Garvey.. M IOZ. 1956. In The Politics of Anthropology. New York: Annals of the New York Academy of Sciences. 1994. In Worlds Apart. Crafts in the World Market. 1987. Miller Daniel. 1992b. Maxine L. 1992. and Elihu Katz. Political and Legal Anthropology Review 20(1):114-119. Robert. ed. and Joan Kleinman. . 421-446. 1992b. New Lives for Old. Rosenberger Nancy. Hunt. 1997. New York: Routtedge. Mead. Karen Fog.. Siting Culture. ed.

Shryock. Spitulnik. 1988. lames L. and Cristina Blanc-Szantoo. and Cristina Blanc-Szantan. Steiner. 1977. Cambridge: Harvard University Press. Katherine. 1994. Oxford: Blackwell. Social Problems 3:153-164. Wolfe. 645. Watson. 1998. Anthropological Journal on European Cultures 3:7-30. Anthropology and Mass Media. Nina Glick SchiJIer. Sampson. Smith. Debra. Oxford: Berg. Valene. New York: Annals of the New York Academy of Sciences. Making Sense of Settlement: Class Transformation. and Transnationalism among Mexican Migrants in the United States. Helena. The Modern World-System. Linda Basch. Nina Glick. 1974. ..~. Chicago: University of Chicago Press. New York Annals of the New York Academy of Sciences. Current Anthropology 18:615--620. Andrew. cd. 1977. Wolfe. New York: Academic Press. African Art in Transit. Annual Review o[Anthropology 2]:293-315. 1992. The Supranational Organization of Production: An Evolutionary Perspective. L L . Culture Without Money: Eastern Europe's Nouveaux Riches. Traweek. Immanuel. Towards a Transnational Perspective on Migration. Christopher A. 25-52. ed. Waterman. 1996. Cultural Struggle. Val. Between Two Cultures. Wolf. 1977. L Rouse. In Towards a Transnational Perspective on Migration. 1990. 1993. Tribes and the Print Trade: Notes from the Margins of Literate Culture in Jordan.. Berkeley: University ofCalifomia Press. Sharon. Hosts and Guests. Eric. Wallerstein. 1994. Beamtimes and Lifetimes.. Alvin W. 1982. Verdery. Roger. Social Text 38:1-19. American Anthropologist 98:26-40. Cambridge: Cambridge University Press. Linda Basch. Europe and the People Without History. eds. PART II ACQUIRING INFORMAnON i I- I. Christopher B. Alvin W. L L I Wulff. Val. Schiller. 645. Juju. 1992. 1963. Money Without Culture. Steven L. The African Mineral Industry: Evolution ofa Supranational Level of Integration. eds.I L 256 HANNERZ L L L 1. . 1994. Beyond the Nation in Eastern Europe. Pp. Oxford: Blackwell. Ballet across Borders.

For us. DEWALT with CORAL B. l l . there is no single agreedon definition for what constitutes participant observation.. Spradley (\980) used the term Uparticipant observation" to refer to the general approach of fieldwork in ethnographic research. structured observation. the method of participant observation includes the explicit use in behavioral analysis and recording of the information gained from participating and observing. while much of what we call fieldwork includes participating and observing the people and communities with whom we are working. We argue below that the method of participant observation requires a particular approach to recording observations (in field notes). all humans are participants and observers in all of their everyday interactions. II l 259 . and the use of questionnaires and fonnal elicitation techniques..KATHLEEN M. participant observation is one among a number of methods that are used in anthropological fieldwork. Despite this. Van Maanen (1988).! I Participant observation is accepted almost universally as the central and defining method of research in cultural anthropology. ~!_. participant observation subsumes the bulk of what we call fieldwork.' We take a narrower view here. but few individuals actually engage in the systematic use of this information for social scientific purposes. and Agar (1996). DEWALT BILLlE R. WAYLAND Eight l 1 I Participant Observation l 1 I Introduction . For writers such as Spradley (1980). and Agar (1996) used it as a cover term for all of the observation and fonnal and infonnal interviewing in which anthropologists engage.. and that the information the ethnographer gains through participation is as critical to social scientific analysis as more formal research techniques like interviewing. That is. We take this position because.

and events of the people being studied as ofthe means ofteaming the explicit and tacit aspects oftheir culture. But we can use the experiences of other ethnograp ers 0 . mitigated by donations of tobacco. (1922:7-8)' Malinowski's approach was distinguished from earlier forms offieldwork in that '. altering it by my very approach. ex?lic!t descriptions of participahnt . who lived for more than 4 years in the neighborhoods in which he worked.. presenting itselfto me more or less as it does ' to the natives. as well as of theirs. The participating observer seeks op~~r· "" . Tacit aspects of culture largely rem outside our awareness or consciousness. described his approach to research in a more contemporary context in similar terms: data ake tinder the observation of the field worker must somehow fit. Sanjek notes.. when someone stands too close to us or touches us in a way that see too familiar. While anthropologists had done ethnographic fieldwork before him.te r:~nio:-~:: . pp.. Malinow (1922. on the street and in crack houses observing dealers and ~. he al$o listened" (I 990b:2 I I). ssible-the lat ers an •• . they ceased to be interested or alarmed.ate y reco d time with and carry out activities with members of communl. It is the feeling of discomfort we have.partlcular theor~tlcal per~ ". . interactions. Perhaps more " . people starting on their errands. and t e c:n et within which they work.an e. example.bs:l~a~~~. or groups of men and women busy at some manufacturing tasks. of toilet. form the atmosphere ofmy daily Jife.derstanding is also bemg d'.. I could see the ~sPent hundredsI o~ant~~corded their conversations and life histories.lintervlewe.. I began to take part. however..' Participant observation is a way to collect data in' a relative unstructured manner in naturalistic settings by ethnographers who observe andA take part in the common and uncommon activities of the people being studied. Bourgois. t bservation was )tif collecting information (Holy 1984).~r an ?'ular ionse . as work presses. we believe that t~cit un..lIes · tl~S t~~~:~ he is working. h d' stepfathers of the crack dealers featured in these pages.0 pa . their theoretical approach. And. ...:d. (Leach 1957:120). partlclpan 0 ':linked with functionalist theory.. k 1990b) or that his theoretical perspectIve mfluenced hIS metho spectlve a n J e . In fact. at its beginnmg. rin extents each of these ethnographers practiced the met~od by I. While linked historically with functionalist theory. or the people well advanced in their working. to wake up every morning to a new day.~vmg }'o dlffe g't taking part in usual and unusual activities.nlatiinn is certainly not tied to it any longer. disallows selectIve leammg about a peop e· P . as always happens with a newcomer to every savage community. Writing more than 70 years later. a necessary evil or nuisance. jokes. the met 0 .an mm Y h \ .. in a way. and ·the c?mmum y'm ared with interviewing) while consciously observmg and · •versl~g (as ~~ngPwhat was observed. for they get up and begin their labors early or late. a level ofknowledge people can communi" about with relative ease" (Spradley 1980:7).Id of the functionalist theoretical perspectIve that assumed that t h ~~~~I~:~:i~:~~s~~~e:~~~:I~S::. It must be remembered that the natives saw me constantly every day.~other of Malinowski's major cont"b~l1ons to anthrop~!ogy ~:St~t:1 . attending parties ~d intim. s ICC I I oints out· "Participant observation . day according to the hour or also the season. dinnerstoNewYearsEvecelebrations. an . "hangmg o~t. (S . Quarrels. Because enculturation takes place ~t the same lime >It ~b~~~d to avoid). 1 could see intimate details of family life. family scenes. ~ seems common across theories. po . sometimes dramatic but always significant. Their field research IS affected by ~ com~ xl mix of their personal characteristics.. Leach 1957). m ~a. 1935) is usually credited with developing"something novel" (Stocking 1983 Sanjek 1990b}-an approach to fieldwork that gradually became known as method of participant observation. it included an emphasis on everyday interactions and observations rather than on using directed inquiries into specific behaviors. and-wen SP f"cnded.a~h' to fieldwork resulted in the development of a .. "As he observed. an . "Explicit' ture makes up part ofwhat we know. to look forward to the important or festive events. and I ceased to be a disturbing element in the tribal life which I was to study.'a~IO:. rituals. events usually trivial. taking of meals. or made self~conscious by my presence. :r one aspect of people's traditions" (p.~~ ~: understanding that is not easIly artIculated or recorded. Malinowski's discussion of his approach is stil the fundamental description of the method of participant observation: Soon after I had established myself in Omarkana Trobriand Islands.'. 44). mothers. toget .. da'l b' is many different types 0 f Adjusting to a new culture provides on a I y as . cooking. I would get out from under my mosquito net. th. they finished by regarding me as a part and parcel of their life.. following others (Fortes 1957. but t (1995:13) L L L arrangements for the day's work. Achieving understanding of people and their behaviors i~posSI ~. grandmothers. d ethnographer closer to un erstan mg 'bl A P' h' (1992) . iC~~~~~~ ~sit~ their families. as they knew that I would thrust my nose intocverything. Participant observatIOn h~S. 260 DEWALT I DEWALT • Participant Observation 26/ L L L L L Here. even where a well~mannered native would not dream of intruding. . h d' articipant The approach that ethnographers use in their field re~earc an m p leX observation is highly individualistic. are the expectations that: (1) We..? h from observation (keeping in mind that the o~server ~ecomes a part 0 i: : t~: being observed)' (2) Being actively engaged m the Itves of peop~e br gd (3) d' the participants' point of vIew. As I went on my morning walk through the village. h d f rtici ant {:. d . to take personal interest in the gossip and the developments ofthe vlllage occurrences. to find around me the village life beginning to stir. glvmg ouses lovers siblings. on an experiences that prevent anthropologists from concentratmgtoo aSSIduously Y . participant observation is a method in which an observer takes part in daily activities. Whether one feels that M~mowskl s ~rtlC '8 r. in the village life.. ~a t nh. been the catalyst for the development of a number of theoretical perspecllves.

participating allows the ethnographer to "know" in a unique way because the observer becomes a participant in what is observed. ca~e~ories to focus only on the aspect of participation. the ethnographer as researcher and writer must be a "vulnerable observer. NonpartlclpatlOn IS when cultural knOWledge is acquired by watching teleVISIO?." Bemard (1995) dlstl. at its root... Big Men made speeches. He reports: "Yanomama nights were an event. (1980:58-62). in the past. lea~ ~ore system~tic~l1y how to craft our own individual approaches. But just how far do you let that other culture enmesh you?" (1996:5).lcal field research.mformants orto participate in only certain of the everyday activities ofthe commumty." Barbara Tedlock has argued that exploring the dynamic tension between Participation and observation is critically important. '. it also highlights what we believe to be the creative tension between the goal of documented observation and the critical goal of understanding the situated observer.:. we go hack to our desks.ngUls~ed participant observation from both pure observation and from pure partlc.It is a strain to try to sympathize with others and at t!'e ~ame time. a paradox (see Tonkin 1984). however. While the focus on a term that is. Examples of this include ethnographers who are l.i and study jazz musicians or anthropologists who become hobos or cab drivers for . While not all of us are ready to adopt the path of vulnerable observation. It wasn't as if the community just went to sleep. I was 'lip. They essentially "commute" to the field to questIOn . She noted that. . shaman took drugs and . as Behar argues. No. At the same time. Iben woke up the next moming. all without regard to the others who were sleeping.. Adler and Adler 1994).~dentified what he described as a continuum in the "degree of ~a~lclpatlOn.oderate participation IS when the ethnographer is present at the scene of the actIOn b. . Eventually I got used to this. Geertz (1995) commented about the process of learning through participant observation in this way: "You don't exactly penetrate another culture. eV. a Yanomama night was like another day. derived from living in the shapono was superior. Our pu~ose is to ~how how ~e can leaI1l from and build on the expellences of others m order to Improve thiS most essential of anthropological ." Behar drew on this metaphor to note: "Yes. a muc~ more system~tic examination of this method may be the best means of unpro~l~g anthropolog. Much information can be ~cqu1fed In thl~ ~ay. We beheve that these two dimensions require separation so that we ha.des for behavior. will live in their own house or perha. our attempts to remain observers of actions and behaviors maintains a certain distance between us and the people we want to "know. Behar noted that participant observation is an oxymoron. But. stllve for sCientific objectivity (1953:69).arts . observation re~U1res d~tachment. for example. methods. babies cried. In this chapter. ~rst issue relates to the degree of "participation" and of observatIOn that are utlhzed. Participation and Observation Recent ~ears have seen a mU~h more self-reflexive examination of participant obseZ:3tlon among anthropologIsts.r''''l~. l l l l i I. Like the ". talks about his decision to move into the shapono (large. All of things went on" (Good 1991:67).banted. S'p~adl~y .r.ive participation" is when the ethnographer actually engages in almost everything that other people are doing as a means of trying to learn the cultural ·\. although his categories confound degree of parttclpatlo~ With the .'. I'd spend eleven hours in my hammock at night to get seven or eight hours of actual sleep" (Good 1991:68-69).lpatlon. circular communal houses) of the Yanomama as an important step in learning about them. to remove the rese:rcher from. but NOT "go native. of the group that is being studied. indeed. modifi~~ his. seeks. because it's part of what he or she brings to the relationship. ." ready to include all ofher pain and wounds in research and writing. to the maximum extent possible. I'd lie in the hammock for an hour trying to get back to sleep among all the i nighttime noises of the shapono. re~dln~ newspapers. < • 1 . Pure partIcipatIon has been described as "going native" and "becoming the phenomenon" (see Jorgenson 1989:62). that first pight and every night afterward. for example. or only occasionally interacts. At first..7~ though no active interaction with people is reqUired.'" When the grant runs out.. as used by some sociologists and psychologists (see Tonkln 1984."~'\ Finally. too.degree of an ethnographer's emotional Involvement in the commumty.' Good has no doubt that the insight .' M. no:.Ps ~ven in a larger community.ut ~oesn't actively partic~pate or interact. a time (see Riemer 1977).Vanomama.. Many anthropologists. men laid plans for a hunt. or reading diaries or novels. t~e ~ctlOns and behaviors so that they are unable to tnfluenc. Because.. it is our View that.I l l l l l l l . Participant observation is a paradox because the ethnographer seeks to understand the native's Viewpoint. paradoxical can be seen as adding to the mystification of the work of ethnography. with people In It.sleeping together. turn. Pure observation. and it is to issues raised in these examinations to :7h1Ch we.when ~e noted that: "Participation implies emotional involvement. Other researchers have focused more on the dimension of emotional involvement in participant observation. Benjamin Paul anticipated some of the current debates . as the mascuJinist image would have it. in "complete participation.e them.' "Act. as well as themselves. You put yourself in its way and it bodies forth and enmeshes you. we examine the' method of participant observatIOn and how it has been used by a variety of ethno~raphers.262 DEWALT I DEWALT Participant Observation 26j parttClp~tobservatlOn IS one ofthe key techniques anthropologists use for leaI1l ing s~bstantlal amounts about the people they study. . In a house in which 75 people were . ." the ethnographer is or becomes a member "'1. Good. The. this was difficult for Good: "When something got me up.

. Robert Desjarlai. What do~s attempting to participate in the events and lives around one mean to ' data coll~ctlO~ and analysis? Living with. because everyone still knows we aren't Mexicans). new styles of behavior. Coy (1989) argued that the apprenticeship experience results in "ways of knowing" and "learning to see" that are distinct from less participatory approaches. He argues. It ~s a tacit understanding that informs both the form of research.of loc~1 values: pat. feeHo ". Observing the behavior of others around us and participating in our society led to our knowledge of correct and incorrect behavior. Fro: ' her perspectIve. This embodiment of tacit cultural form also informs interpretation of meaning. it enhances We ~elieve tha~.ime.26). . emotions. comes as the result of sharing the lives of people over a significant amount of time. he found it necessary to leam how to move and to experIe?c: hIs body . I L L L L tramlng as a shaman. experiencing the body in this manner (including the residual Intennmghng effect it had on how I stepped through a village climbed a h'I1 • approached others) influenced my understanding of Yolmo expe'riences.d"" ga~e. these interactions shaped his "understan~mg .ultural k. while difficult to convey in words. (DesJarlOlS 1992. . that these ways ofknowing are connected to the physical performance of the duties required in the role being examined. T?~t is: b~. it hin~ed :~ . It is embodied in the way we walk. students can b~tter appreciate the cIrcumstances. Now that we have considered these definitional issues concerning participation. the recording of informat' d IOn. observation.:ords. As our ability to communicate through language improved. it's one with which all of us have had experience. move and talk (imperfecUy translated. and the like. and my loose assemblage of bent knees and jointed bones contributed to the springboard technology that gradually brought some force and case to my shamanic "shaking. DesJarlals argues that his body incorporated the meanihgs a. the forms that we !hrou~h t. moving. Narrative ethnography combines the approach~s of wrItmg a sta?dard m~nograph about the people being studied (the Other) wIth a~ ethnograp~lc memOlr centering on the anthropologist (the Self) (Tedloc~ 1~91. It shapes how we interact with others and. however. we learned a great deal by observing parents. As infants and toddlers. like Desjarlais. of course. the The process by which this might take place. By using the body in different ways. We note that the timbre of our voices changes in Spanish to approximate that of Temascalcingo voices and that we are much more animated in our speech and bodily gestures. tlclpated m the pracllce of everyday life.69).a greater understandmg of the images he experienced in trances as part of his observe.patlOn problem and with their emotional involvement. par. Part of what we know about life in rural Mexico is tacit. ways of being and moving through space that 1 did not previously have access to. and others. (i 992) trained to become an appre?t. laughing with the people tha~ one IS trymg to understand provides a sense of the self and the Other that isn't easIly put mto . I stumbled on (but never fully assimilated) practices distinct from my own. so as to maintain their reputations as professional ethnographers. Participant observation is thus both a data collectIOn and an analytIc tool. To do so.terns of actions.alIty of th~ data obtained dUring fieldwork. redlock (1991 :71) argues that successful formal and informal apprenticeships are ways of undergoing intensive enculturation. It enhan~es the qu. He notes that as he gained c. Most obviously. to anticipate and understand responses..irrespective o~ the degree of involvement or of participation. caught the meaning of jokes." an ap. it allows us to understand nonverbal communication. This isn't a mystical procedure. What IS most valuable in the kinds of accounts that she is advocatmg IS that they go a long way toward demystit)dng the process of doing :thnography. these accounts should be part of the data ofanthrop I She argues that we should engage in the "observation of participation. Second.L L L L L L L L 264 DEWALT I DEWALT Participant Observation 265 many ethnographers wrote personalistic accounts of their field research und pseudonyms. relatives. we would like fo turn to a discussion of how one actually does participant observation.as a Yolmo. Doing Participant Observation Children learn their own culture through participant observation.nowledge. this learning process was enhanced. working with." (1992:27) Why Participant Observation Is Important ~racllce of partlclpa?t observatIOn provides two main advantages to research. Touching head to heart merged thinking and feeling (two acts unsegregated in Yolmo society).r~::h that she terms ~~rrallve ethnography. a sense of the body as a vessel dynamically compact led me to see Yolmo forms as vital plenums of organ and icon. the speclfictechUlques ofdata collection. I~amed how to sip tea. He argues that much of the leaming regarding peoples lIves IS tacIt and at the level of the body.ce s~aman. and reactions they are likely to experIence when they begin their own field research. First. Desjarlais is one ofmany ethnographers who have apprenticed themselves in the field in order to gain new perspectives. more fundamentally.The field worker who doesn't try to experience the world of the observed through participant observation will find it much harder to critically examine research assumptions and beliefs and themselves (see Clifford 1997:91). it shapes how we interpret what we the qU~IIty of the mterp~etatlOn of data. an the subsequent interpretation of materials collected In studying Yolmo healing. ways of being. examining how other anthropologists have dealt with the degree of partlcl.

In terms of . returning to those awkward first attempts to learn a culture-but with some advantages and disadvantages. . and bodily gestures. We should be open to being surprised and. BourgolS (1995: 19-20) makes a similar point .h fu . l l l l l l l 1 These general attitudes and elements are guidelines for those who are beginn~g to use participant observation as one of the tools in conductmg ethnograp tC research. While we should be sensitive about intruding into situations where we are not wanted or welcome. personal characteristicsto the enterprise. ituationsand plorations in the field are bound to confront on~ wit con SlOg s ~~nt1icting pressures. suggested that we cou~d sttlllearn a lot by doing a census of the different kinds of stalls in the open~atr Sunday market.lme ~ 4. Almost all people love to tell their story and to share their experiences with those who take an interest in them. of doing anthropological field research-has for too long been shrouded in mystery. These tasks served th~ purpose of attuning us to nonverbal cues. ThfiiS IS a Skl~\ tch:t ~:n t~ e~e~~c~n through practice. the method of doing participant observationindeed. and feelings. roug 6.. If we listen we will learn much more qUlckly. Aside from being a somewhat barbaric approach to teaching. thoughts. As one learns more. The feeling of "culture shock" (Bock 1970) when you are confronted with a wide variety of new behaviors and stimuli is something almost all ethnographers have experienced. As ethnographerS. a good field researcher must react to the goings-on with sensitivity and discretion.' b h d 5. and having read many accounts of hundreds of other experiences. The insights gained from this method often can and should be used later on to be verified and substantiated through more structured techniques. and by making charts of where people sat in meetings. advantages. we reject this attitude because we believe that good ethnographers are not born that way. Villa Rojas (1979). It's important to be a careful observer. while the activities in which the ethnographer is taking part may be extraordinarily exotic or mundane. 3. if an ethnographer shows genuine interest in learning more about behaviors. Temascalcingo. Both of us have often felt vel)' intimidated and anxious about what we were supposed to do on many occasions. appropriate facial. participant observatlo. they also beglO to ee much more comfortable and confident. . those who can't will discover that during their first field experience and presumably choose another profession..n pu~s the researcher into situations in which he or she is acquiring infonnatlon m an open-ended fashion. by . It is normal to feel awkward and unsure when observing and participating ina new situation. 1. An additional advantage is that we should have profited from the experiences of other ethnographers. \) Th h It's important to be a good listener (Leach 1957. Being ethnographers." The implication here is that some individuals are constitutionally suited to become ethnographers. biases. so that some errors are almost ine~itabte. We remember our Ifst expenen . writing about his and hIS collaborators field I l ____~. we are not just interested in how we can "get by" in the culture we are learning. t' e As Whyte has emphasized: "It is important to recognize t at pa lenc . l I Establishing RapportlBecoming a Participant as Well as an Observer The establishment of "rapport" is often talked about as both an essential e~e?.' d Everyone will make mistakes. most reasonable.Participant Observation 266 DEWALT I DEWALT 267 are to use to express or hide emotions.en: in using participant observation as a tool as well as th~ goal of partl. 1. . Unfortunately. language we rapidly acquire a substantial ~oun\ofinformatlon m ashort time. Having been involved in a variety of field research projects during the past three decades. That is. we also want to develop a more systematic understanding of that other culture-an understanding that we can analyze and interpret for our colleagues and students. Sanjek 1990b:21 . and all the other tacit and explicit aspects of our culture. 2. got us out and about in the community ~~ere we began interacting with people. Many anthropologists have sent students to the field to "see if they can survive. we also expect more of our participant observation..clpan observation. Those individuals who develop or use these attributes can gene. in a sense. That LS. Pelto. he or she will be a welcome guest at most activities. One of our mentors. The ethnographer has to approach participating and observing any particular situation with an open mind and a nonjudgmental attitude. These are also disadvantages because we bring a series of preconceptions. Pertti J.rally exp~: to achieve sufficient rapport with people to make partlclpant observation a use ethnographic too I. science research that uses structured techniques in order to test ~ypotheses (Picchi caUs this selective learning [1992: 144]). we believe that the following are the basic elements and attitudes that are required to do participant observation. but mostofthese can be overcome With t.to learningt~h: ~:::~~~~~~oTsh~:i~~ perhaps the strength of participant observation as a me o. Mexicowhen we only had arudimentary knowl~dge ofSpaOlsh. we are adults and have gone through the process of learning a culture and have acquired a lifetime of experience living within a culture. we are. hand. sensitive and intelligent individuals can be trained and educated to be good participant observers and field researchers. by making maps of the houses m the Villages. . and taught us that a lot can be learne Just by being careful observers.--l . and developed but these are probably the times in which we were learning the most ~boutfth~ people and places we were studying. counting the numbers of people at the market with bar~ feet a~d the number wearing shoes or boots. . but few errors are serious enough to abort a project" (1984: 11).

L l L L We ran down the main village street. Suddenly. In some situations in which the . our prevIous questions about sensitive subjects like witchcraft were answered in great detail rather than tossed off and evaded.. simply by continuing to live with and interact with a group of people. Their breakthrough came as a result of a police raid on an illegal cockfight they were o~servl~g. the only basis on which really reliable infonnation can be obtained" (p. characteristics. K.'· Many field workers find that they can point to a single event or moment when the groundwork for the development of true rapport and participation in the setting was established (for example. seemg. : . what they were doing m the VIllage. Sterk 1996:89.y been through this sort of thing before. whipped out a table.en moments later the police arrived. sat down. We found in our work in Temascalcingo that the quality of the infonnatlOn we were receiving improved after our return from a three-week trip to the United States to renew our visas and take a break from fieldwork.. While these dramatic examples are more vivid. exp~ctations of commun. In a number or these cases. and the kmds of infonnation that one wants to obtain. In our own thinking. "Our close contact with local people· has always led to excellent rapport. they chose to run away with the rest of the villagers when the cockfight was raided. when infonnants participate in providing infonnation for "the book" or the study. He relates that it resulted in a subtle change in his relationship with the Gypsies. Whyte and Whyte 1984. agricultural problems. and we aH! without any explicit commUOlcat~on whatsoever. Komblurn risked violence to stand with the people with whom he was working. and experience of the elhnograp~er. In these situations. that the thing to do was run too. (p. The fact that we had visited our own culture and families. DeWalt 1983). AI~ough th~ Geertzes could have stood their ground and presented the polIce wIth their credentIals and pennissions. people seemed to be unsure of our interest in them and their lives. We have often.cloth. Clifford and HIilary Geertz acted lIke Balinese villagers when they could have acted like privileged foreigners. Nader 1986. a ~ble. Clifford Geertz (1973) elegantly describes the event ~hat allowed him and his wife to begin to establish rapport in the Balinese village m whIch they worked. It was truly an anthropologist's nightmare. three chairs. participated in rapid appraisals (Kumar 1993) in which the objective is to obtain a quick impres~ion of fanning techniques. L 268 DEWALT IDEWALT Participant Observation 269 research in the Mayan region of Mexico says. '.nothmg ahead of us but rice fields. The. for example. the ethnographer has to achieve "instant rapport" that is sufficient to put infonnants at ease to answer the questions being asked.. 'we were unable to obtain much infonnation at all on topics s~ch as witchcraft or traditional curing practices until we had spent over six months m the field (see discussion below on ethnographer bias).. In a moment of crisis. L L L i L however. he became one of them. away from where we were living. Until that point. the Geertzes were enthuSiastically incorporated into the community. but had returned to renew our stay in Mexico.ater warmth and affection than they had previous to our break. The Geertzes had been in the village for about a month during which time the villagers treated them as though they were not there. "When in Rome. open country and a v~ry ~Igh volcano. An even more dramatic account of a single event that lead to the establIshment of rapport is provided by Komblum (1996). About halfway down a~other fU~itive ducked suddenly into a compound-his own It turn~d out-and we. ." my wife and Idecided. Dewalt 1979. This discussion of the process of developing rapport and coming to be accepted in a community begs the question of how long it takes to achieve. is an elusive one. who was called on to stand with :'his" gypsy family when they were attacked by Serbians in their camp outside Pans. at least to some extent-that is. pure and simple. and what pennlSslOns they had. We returned to Puerto de las Piedras against the. hIS Wife. he noted. or Significant kinds of illnesses m a community. I. some people would move ~way when they approached. . After that. Whyte 1996). He was then viewed with respect rather than with disdain. demonstrated our commitment to the people in this Mexican community. were rarely greeted. we have often used a definition for which we can no longer find the citation. a lot depends on the ability. people seemed to look right through them. rapport generally is established slowly. northward. The definition of what constitutes rappor!. commenced to sip tea and sought to compose ourselves. In addition.'ty members who had assumed we would not. when they demonstrated a more-than-passing commitment to a com~u. the Geertzes' adopted host was able to ~rovlde . On the other hand in our initial fieldwork in Temascalcingo (H. In Geertz's words: On the established anthropological principle. the breakthrough in rapport Was achieved when the anthropologists showed that their relationship with the community was important and serious. who had apparentl. and when the researcher approaches the interaction in a respectful and thoughtful way that allows the infonnant to tell his or her story. 59). 415) ""?. 113). followed him. Stack 1996. In this fonnulation. In reality. She (1986) wrote: "Rapport. people said "Que milagro" (what a miracle) and greeted us with gre. As the three of us came tumbling into the courtyard. when both the "infonnant" and the researcher come to the point when each is committed to help the other achieve his or her goal. consists of establishing lines of communication between the anthropologist and his infonnants in order for the fonner to collect data that then allows him to understand the culture under study" (p. Everywhere we went. .a lengthy and accurat~ d~scription of who they were. . rapport is achieved when the participants come to share the same goals. the Geertzes bad been in this compound all afternoon sipping tea and knew nothing ~bo~t the cockfight. Nader (1986) suggested a more one-sided view. The bewildered police left. and three cups of tea. nity. the circumstances and characteristics of the group being studied. only slightly less instantaneously than everyone else.

must be a tremendous sense of responsibility in it. o Experience and the literature suggest that there are severa~ important pomts abo~t fi Id notes and their relationship to the participant observation method. Bernard 1994). texts. and informal interview carried out on a day-by-day basis by the resea~c~er. (2) records of prolonged activities and ceremonies (in which interview is not feasible)..--1 . However. The admonttlon m Notes and Queries that "It is unwise to trust to memory.Iose the fine detail of observations and conversation all too quickly. and. Field notes are at le~st ~ne more step removed from objective observation than the nonobjective observatlo~ Itself and are a construction of the ethnographer and part of the process of analysl~. Until recently. Most of this space is devoted to examples of the level of detail (high) that they see as desirable in recording field notes.between the researcher and study participants in a manner that allows for reflexlvlry at the end of the process. !here.5 pages to a discussion of "descriptive notes" as one of four essential types of documentation. The sixth edition of Notes and Queries (Seligman 1951). Sanjek (1990c) addresses issues about field notes in more detail. a set of chronological. plans and diagrams. Sanjek (1990a) reviews the historical changes in the nature of participant bservation and field notes. According to Clifford (1990). and to be able to review the growmg relationshIp . and genealogical and census data. in practice.nowskl recorded both observations and reflections on their fieldwork expenence m field notes and personal diaries.. and analyzing field notes. ThIS mherent contradicti~n embodied in field notes is part of the continuing discussions about the na~e of anthropological inquiry and of ethnography. events ~nd :ontexts by the researcher. situations of violence or deprivation. he writes: "Ethnography cannot. Speakmg of des~np~lOn (thick or otherwise). Pelto (1970) and Pelto and Pelto (1978) provide approximately two pages of discussion of field notes in each edition of their book on Anthropological Research. it will be virtuallY impossible to re. extreme isolation) very suspicious. As one of Jackson's (1990) respondents said about field notes: "Each anthropologist knows it is a dialectic. and casual interviews that are the primary materials of participant observation. but several important issues about field notes should be addressed here. pure recording-cannot be sustamed.) Notes and Queries suggests three kinds of notes. overheard conversations.. you create it together. The second point is that field notes are simultaneously data and analySIS. It has become relatively standard in ethnographic inquiries to think of a minimum of a year of fieldwork as necessary to gain sufficient insight from participant observation. and (3) following the practice of contemporary ethnographers. (The other three are: maps. Sanjek (l990b) notes that pioneering researchers hke Malinowski and Mead knew this. mostly anthropologists. debates over the last two decades have made this point even more salient. By thIS we mean that they should be the careful record of observation. time allocation. constructed by the researcher. Malinowski 1967]). . . s arching for things they did not understand. or on which they felt they had i~complete information (what Agar refers to as "breakdowns" [1986]).c noted.heir variations will remain indelibly etched in the researchers mmd for all tIme. etc. p~re inscription"-that is. but writing field notes is virtually the only way for researchers to record the observation of day-to_ day events and behavior. The infonnant creates it. s. Sanjek 1990c. 14).ents d. to a series of questions about their relationships with their . but this is only a general guideline. that is. Jean Jackson (1990) summarizes the responses of a sample of 70 field workers. Space limitations make this impossible. The first IS ~e t bservations are not data unless they are recorded in some fashion for further ~al °sis. I l l l l I' I . then it didn't happen" (at least so far as being data for analysis). (but that is distinct from the personal diary which a number of ethnographers keep [for example. matntam a l A Note on Notes A whole chapter should be devoted to discussing strategies for writing. If the researcher's daily reactions to events and contexts are not recorded.o as t~ direct the flow of subsequent conversations or interviews. At the same time field notes are a product. In this -volume. Field notes are the primary method of capturing data from participant observation. it may never be possible to achieve a substantial amount of rapport. They continually read and reread field notes. A useful maxim that we have always· used in training students is that: "If you didn't write it down in your field notes. We believe that few a~t~rop?logl~ts really ever believed that their observations were unbIased. Mead an~ Mah. people being studied are for some reason (for example. field notes. More books about field notes are coming out all the time (for example. Participant observation is an iterative process. or that ehmmatmg bIas was even possible or desirable in research. harsh exploitation byoutsiders. as we hav. conversation. part of hat occurs is the development of a tacit understanding of meanmgs. devotes only 1. one version" (p. input and output ofenergy. Researchers can audio. a sense of pohtlcal history:. managing. which the committee called a journal. relatively little had been written about the nature of field notes and how anthropologists record observation. Even though it seems after a few months in the field that ~ommon ev. the vi~W of the fi~ld note as. which are still relevant today: (I) records of events observed and information given (in which the researcher takes time to interview or converse with participants as events take place).. daily notes.or videotape more formal interviews and events to record words and behaviors for later analysis and record more formally the results of response to formal elicitation. .270 DEWALT I DEWALT Participant Observation 271 .c0nstruc~ the ~evelopment of understanding. 1I0tes should be wntten as soon as possible" (Seligman 1951:45) is still relevant today. :emory is unfortunately more fleeting and less trustworthy than that '!'e . --. for example.

It's sometimes comforting and helpful in assuaging guilt on days when we just need to be out of the scene to say to ourselves: "Well. We would like to ~ L I \ L ! L Ethical Concerns Of all t?e me. and to make sure that we will put it "in the book. in the modem world. Bourgois (1995. it was clear to participants L ! L being interviewed by structured interview schedules or when their interviews are being recorded in some way. Participants in his study even joked about "how the book was coming" (1995:27). the responsibility "to do no harm to our informants" (AAA 1997) is great.thods usually applied in the field by anthropologists. Who may even be anxious to see the finished work. while drinking a beer in the bar after a day's work. Although anthropologists once deposited copies of their field notes in Iibtaries for other researchers to consult.. L L L 272 DEWALT I DEWALT Participant Observation 273 L L I L. Whyte (Whyte and Whyte 1984) reminds us that much of our work can be put to unintended use. With respect to the impact of note taking during events. it is for just this rea~on that Agar (1996) is critical of time spent in recording field notes. Therefore. It can maintain such a relationship only to what is produced in field notes.." But it is more often the case that the informants "forget" that casual interactions may form part of the data to be used in analysis. hanging out. 68). no~es in fr~nt . 1996). work. if informants were always consciously aware of our activities as ethnographers. suggesting that what they had to say was not important enough to record. but "think a bit first" (1994:95). . participating in events. either of "scratch notes" (Sanjek 1990c) and "jottings" (Bemard 1994) or longer more transcriptual notes. In fact. Field workers rarely recite their informed consent scr~pt during afternoon conversations carried out while swinging in hammocks. and the goals of research.. leisure activities. We want them to forget. On the other hand. he went home or to the bathroom in order to record events. Community companions will probably not be aware that the field worker will faithfully record an account of these events as soon as possible. at least part of the time. for a time at least that we are outsiders. and that this will form a data core for analysis. participant observatIOn raises the greatest number of ethical questions. Many peoples do not object to it." And as it is almost always true. We regard this as the strength of our method. Whyte (Whyte and Whyte 1984) went to relatively great lengths to avoid taking The field workeris traveling alongside community members. openly taped both interviews and conversations as he hung out with crack dealers in El Barrio in East Hadem.of partic!pants in his research in ComerviJIe. the activities carried out during participant observation are less clearly so: . as Punch (1994) points out. Punch recommends that we carry out research with this powerful method. it is the perfect excuse. they note that Some people may become suspicious when the ethnographer takes notes. However. most understand that the information resulting from these activities will be used in the research carried out by the field worker. quoting Ditton (1977). Whether public note taking. Notes and Queries on Anthropology suggests: "The investigator must sense the native attitude to note taking in public. What are our responsibilities underthese circumstances? We have often taken the position that providing anonymity for communities and individuals is sufficient to protect our informants."e so comfortable with us as community participants that they will share insights and information that only insiders would know. We have all been Impressed by the degree to which our infonnants will suspend a conversation or intemction to remind us that the topic about which they are speaking is important. However. It's important to emphasize that ethical questions surround not only the published information we provide but also relate to our field notes. The first is the ethical question about taking notes publicly. Others found that participants were insulted when notes were not taken. To gain accurate maps of mteractlOn m a men s club. Doing so. participantobservation is inevitably unethical "by virtue of being Interactionally deceitful" (Ditton 1977:10)." Jackson (1990) reports that a number ofthe ethnographers she Interviewed found that taking field notes in front of participants was uncomfortable and objectifying. Even if fle~d workers make it clear that they will "write a book" or report on their experiences"nforma~ts may not realize that what they share as "gossip" during informal conversations w. . It is the ethnographer's responsibility not only to think a bit first. on the other hand. It is by its nature deceptive. When individuals are I L comment on two issues. We want to develop sufficient rapport and to have them beco. and when he openly taped interviews and conversations on the street. However. which he sees as more profitably spent in interviewing. I . A final note on field notes: Participant observation can be a very stressful experience (we often call this culture shock). an experience lOany of us have had with participants." (p. it is increasingly difficult to even suggest that the identities of communities will remain hidden.'1l form part of this report. The Participants in Bourgois' s crack research knew he was writing a book. today I MUST stay in and catch up On my field notes. and a few people who are "otherwise friendly may never tolerate the practice. or while in bed with a lover. based on careful consideration of the ethical dimensions of the impact of information on those who provide it. but to make conscious decisions on what to report and what to decline to report.. . While this is probably good advice. has an impact on the flow of participant observation is a question that has been answered in a number of different ways by researchers. simply regarding it as one of the European's unaccountable habits" (Seligman 1951:45). That is. reinforces for participants that what is being done is research. constant descriptive relationship to cnltural phenomena. the information we acquire would be less rich. this practice is unfortunately being lost.

are almost always transIent (Punch 1994). Not only is there the potential that field notes can be subpoenaed' ~overnmental funding organizations such as NSF and NIH are becoming increas: mgly concerned with research integrity and falsification of data in funded research. unul the bIOlogIcal grandchildren were available to take over. However. Judging was to focus on originality and eloquence of individuals who were given a topic on which to speak.the group they are studyIDg. When the regional senior picnic was planned for a recreation area near Central County. Realizing the implied commitment to the community in the director's words. for example: . In fact. AnthropologISts need to be aware of the implications of relationships and obligations that they incur in the field. As part of research with older adults in rural settings in Kentucky. would be available for inspection by revIew boar~s should allegat~ons of a breach of research integrity be alleged." Kathleen's heart sank because she realized that the project was nearing its end and that she was not prepared to maintain a long-term relationship with the center and participants after the project was over.. Much ofthe afternoon was spent in competitions among senior citizens. With computenzaUon.1 l -: 1 l 1 l. Many anthropologists enter into fictive kinship relationships in order to find a place in a community. . Universities are developing policies concerning research integrity of nonfunded and funded res~arch in which data. baking contests.274 DEWALT I DEWALT Participant Observation 275 l what was on the record and off. field notes have always constituted a gray area. was "adopted" into a .:"ithout names. the Central County Senior Center Director introduced Kathleen and her co-researcher by saying: "I would like to introduce two people to whom we have become very close over the past few months. and ate the meals provided by the centers. inclu?ing field notes. Another important ethica. at personal expense (loss of precIOus field time). When It became necessary to assume the responsibilities of a true granddaughte~and care f~r an elde~ly "gra~dfat~er" she did it. because there was a lot of out-migration. For us. Over the years. though. Here they chatted with the program participants. Although protesting that they might be biased toward the contestant from Central County. Assuming this kind of relationship has a number of implications and ~dds a ~eries of responsibilities for the anthropologists that should be carefully conSlder~d before the identity is accepted. we have developed strong personal relationships in different places and with different people. taking notes or taping events openly alleviates some of the concern that participants lose awareness that they are participating in research. however. Even when the relationship is not intimate or familial. While many informants become true friends such friendships are difficu.' issue is about maintaining relalionships developed in th~ field.. anthropologISts have gone to jail to protect the identities of their ~nformants. Two field researchers we knew joined ritual dance groups in a Mexican village in which they were doing research._------l . family as a daughter and granddaughter. the researchers were invited to attend and arrived with the participants from the Central County Senior Center. One of the contests was an extemporaneous speaking contest. They ate box lunches. i l l l 'l I I .:" that the researchers were studying food and nutrition problems in the commumtIes. Catherine Lutz (1988) became a "daughter" to fit into Ifaluk society. she began to develop a plan to put more distance between herself and the residents of the center. They are demanding that primary data be available for review by others. they agreed. While most internal review boards for th. there were no obviously unbiased potential judges.1t to maintain when the anthropologist is thousands or miles away. etc. participated in the auction for such goodies as fried apple pies and home-canned vegetables and pickles. Both ethnographers danced for two years during which they were in the field. On occaSion. I l _ . rode the transportation vans.. Most of the people at the picnic were participants or staff of individual centers. Becky Ross. During the introduction of jUdges. with ." use of human subjects require that all questionnaires and transcribed IDtervle~ data be stor. . but were unahle to fulfill their commitments for the full terms of their appointments. some concern with the protection ofthe identity of informants ID field notes can help alleviate these problems. While we have tried to retum as often as possible to communities we have studied to let individuals know we still think of them. they sometimes suggested that he turn on his r~corder when he turned it off. this has not always been possible. the community was used to individuals who cou~d not get back in time for key festivals in which dancing was required. However.. . we must recognize that this transience may not be the expectation of those individuals who become our informants. The message is that while the field worker is almost always transient. it becomes relatively easy to assign code names or numbers to ~~icipants and use these from the outset or use the global search and replace faclhUes of word processing programs to expurgate real names from field nOles. These relatIOnships. Ethnographers actlv~ly try to develo. one of the places Kathleen and her collaborators "hung out" to get a better understanding of the problems facing older adults was the Senior Citizens' Centers.e. it may present ethical dilemmas."d . It ~as pOSSible for them to send money to the community to fulfill their commitments ID those years when they could not get to the field to dance but this could have caused substantial problems. The research integrity policy statement for the University of Pittsbur~.now suggests that real names be expurgated from field notes agamst the pOSSlblhty that they will be requested by others. whose story forms one of the ethical case studies published by the AAA (Cassell and Ja~obs 1997). Good intentions to correspond have rarely been POssible to maintain. . Fortunately for them. She discussed the benefits and costs to her research of this relationship.p close relationships and to identify . The second issue has to do with preserving anonymity for participants identified ID field notes. The commitments into which they entered were for several years. including walking races. All the participants and the program staff kn. Kathleen and one of her collaborators were asked to be judges.

I'd come to study them.. This kind of thing went on. During hi§' research WIth the Yanomama of South America. drug cultures. Bdl faced a very dIfficult SItuatIOn In getting rid of his neW-found friend without insulting him. one outcome of popular justice. I'll help you find out about the culture of those fucking Indians. I was an anthropologist. There are also often occasions during which the e~hno~rapher faces a difficult decision about whether or not to inlerv~ne . he did intervene in a similar sit~ uation (1991:104-105). or high-risk sexual practices.<:" SItuatIOn. Good ascertained that the teenagers were trying to drag the woman off to rape her while the old women were trying to protect her. he came across a situation in which' a group of teenage boys and three older women were engaged in a tug of war. we dedicate a special section to it. She. It was expected behavior. subject to violence or damaged in some other way. pl." Because this is one of the most important questions with which most ethnographers eventually have to contend. chances were she'd be raped.in.r '- L . Tomorrow we'll go up in the hills to talk t~ them. not a policeman. 1996). Limits to Participation? There are some dramatic cases of the need to establish limits to. patiently explaining that we were not there to study only "Indians" and that we wo~ld feel ~uch better about using our own methods for getting people to talk WIth u~. (1991:102-103) whether to engage in romantic and/or sexual involvements with members of corn. If a woman left her village and showed up somewhere else unattached. Gender Issues in Participant Observation One of the important contributions of theoretical discussions over the past two decades has been the axiomatic acceptanceofthe ethnographer as a gendered. 1had no doubt I could scare these kids away. It's increasingly difficult to justify not intervening. ~enneth Good gave a particularly wrenching example of this. They were half-afraid of me anyway. became quite involved with the drug dealers with whom he was working.family status-will detennine how we interact with and report on the people we ate studying. . the establishment of our own limits to participation depends greatly on our own background and the circumstances of the people we study. I thought. Obvious examples include situations in which eth. although he abhorred the violence and other activities in which they engaged. However. BIll decI~ed that. She used this as an argument for working to try to help the tin miners she was studying in Bolivia fight for their rights. On a less dramatic level.stol at the ready. A woman whom he had befriended earlier was in the middle. We may appeal to "cultural relativism" or to the role of "objective observer" 10 avoid intervention in situations like those faced by Good. Nash concluded that the world should not be seen as simply a laboratory in which we carry out our observations but rather a community in which we are "coparticipants with our infonnants" (1976:164). there were probably better venues .t~an the cantma for fi~~mg out what was going on in the community. She describes how she chose to intervene in the punishment of several young boys caught stealing in a South African village. even though her research was. There are also many accounts of ethnographers being confronted with I stood there. 276 DEWALT I DEWALT Participant Observation 277 L L L L L L I . dangerous to the personal health of the ethnographer. try to inject my own standards of morality? I hadn't come down here to change these peopte or because I thought I'd love everything they did.clpate or not to participate in the life of the people bemg studIed IS not easy to deCIde. the man showed up at our house. clngo. knew it. sex. To a large extent. my heart pounding. threatening yell. BlII decided the town cantina would be a go~d place to fin? out . Bill's explanation that we were there to study the local culture led his new companion to pull a very large pistol out of his belt and to state: "The Indians around here only understand one thing and that's this. participation because engaging in these behaviors may be illegal. . or both. class. When we began research there. religion.Decldl~g h?w much to partl. and if I picked up a stick and gave a good loud. On the olher hand. in the future. Scheper-Hughes (1996) argues even more strongly that the role of the ethnographer includes activism. . Castaileda 1972)." The next morning (not very early). to assu~e t~e r?le of research assistant to the anthropologist. In the following sections. prisons. when we see the people with whom we are working being exploited. we discuss some of the most important ways in which the personal characteristics of the ethnographer affect participant obser· 'vation-and how these personal characteristics often results in limits to Participation. Our L L L L I L personal characteristics as individuals-our ethnic identity. and . munities they study (see below). L" A final ethical issue relates to what we might call "limits to participation. they'd scaner like the wind. in part.as g~ing on in the community. Every ethnographer sooner or later faces dilemmas like these that become difficult ethical issues (see Rynkiewich and Spradley [1976] for a useful compilation). for example. She intervened to take an accused boy to the hospital to save his life after his punishment at the hands of villagers. What was I supposed to do. All was going well untl! one of the IncreasIngly Inebriated patrons asked Bill what we were doing there. there are experiences like those of Bill in Temascal. Bourgois (1995.:hat w. A month later. they knew it. raced.. He described his dilemma as the young boys succeeded in pulling her off into the bushes: Good decided to do nothing but wrote that this was a turning point in his integration into the community. I wasn't supposed to take sides and make value judgments and direct their behavior. nographers study shamanistic use of hallucinogens or other drugs (for example.

I should use the tag mawesh (sweetheart) when addressing someone. (A contrasting case however is that of the discrep~ncies in the rep~rts of Mead [1923] and Freema~ [1983] ~onceming the sexual hves of Samoan girls. If one is to be successful as a participant observer. 1997). 1990). who focused more of her work On exchanges involving women.) Catherine Lutz (1988) has written about the experience of being a woman in Ifaluk. a male. claims that his data about premarital sexual behavior are more detailed and accurate than Mead's because he spoke Samoan fluently and she did not. for example. husbands and wives may spend very little time together. The gender of the ethnographer has an impact on several areas of the ethnographic enterpris. however. Warren 1988). had antIcIpated that she would be able to achieve the "genderless" or "generalized gender" status that a number of women ethnographers reported in other settings (see Fluehr-Lobban 1986. Just as men are often barred from situations in which they can know the intimate worlds ofwomen." of the island's men was called. and Tarnelakar would from then on refer to me before others as his daughter. T~e reports of ethnographers. She noted that the lives of men and women on lfaluk are sharply divided. individual. m g~?eral. Men and women. she found that she could not achieve this.278 DEWALT IDEWALT Participant Observation classed. however. Murphy and Murphy (1974) showed a view of Mundurucu society that was almost unique for its time in the way it placed in counterpoint the perspectives of men and women. A week later. She wrote: 279 On the first evening. etc. although she notes that her original entrance into the· community was eased because she was accompanied by a male col1eague. but that world is highly gendered. have successfully stepped outside the prescribed roles for women within a particular cultural setting. or other economic activities in which both men and women might work. he anxiously asked. a toi. For example. Seeing me. Tamalkar [her fictive "rather"] and my "mother. women make naturally better field workers because they are more sens~tlve ~nd open than are men (Nader 1986. Al1en (J 988). suggest that women may find it easier to gam access to some aspects of men's lives than male ethnographers find it to gain ~ccess to the worlds of women (Nader 1986). in which women contribute strongly to the economy through control of agricultural production. however. race. Freeman. Having the perspective andlor assistance of a member of the opposite sex can often be quite important. In research in rural Kentucky. As direct requests are rarely refused." Ilefagomar. Several classic ethnographic debates are at least partly the result o~ the differen~ vantage point of the ethnographer. she reports months of stymied research in which community members essential1y shunned her because she couldn'l keep her temper and act like an Utkuhiksalingrniut woman. Lederman 1986. Other researchers have argued that. I was to consider myself their daughter. A number of women. The Murphys could do this because they had simultaneous access to different events and to different informants during the same events. Lutz recorded the event that finally convinced her to abandon the hope that she could create a role outside of the Ifaluk system of expectations. I should say siro (respect or excuse me) when passing a group of seated people. on hearing this. the view of economIc exchange m the Trobriand Islands that Malinowski (1922) presented is enlarged and enhanced by the work of Weiner (1988). Women have been involved with research on agricultural production. '"Where are you gOing?" and looked both uncomfortable and displeased when he heard I was interested in observing the meeting. A quite important influence relates to the experiences of the ethnographerdunng the field research. Jean Briggs (1986) was never able to adopt the role of kapluna (white) daughter to the satisfaction of her fictive Utkuhiksalingmiut "family. rather than as a neutral research tool. l l gardens and birth houses. Tamalekar emphasized. (1988:36--37) Lulz expected to be able to choose Ihe role that she would adopt in Ifaluk and anticipated that she would "allow" (1988:33) herself to be socialized in arenas in which she was interested. Tamelakar was already there. or '"mass meeting. but was required to conform to the gender expectations ofthe community around her. Warren 1988). and I spent the great majority of my time with women in cook huts. With this and subsequent encounters. .ifferential acc~ss to the lives of women has resulted in generations of predommantly male-biased ethnography. This is not to say that women do not have high social status in many domains of life on Ifaluk. Being a man or woman may be the most significant social fact concerning an individual and obviously should have an impact on participant observation. After a number of months of in-depth intervieWing with samples of key informants in each of two counties. but waved me to sit off to the side by his relatives. but the spheres of men and women are different. Some feminist and ethmc wflters argue that true rapport and accurate portrayal of the voice of the ~articipant ~an only be achieved by researchers who come close to matching the mformants m gender.e. women ethnographers are sometimes barred from important parts of the worlds of men. I should crouch down rather than remain standing if others were sitting." As a result. D. was able to study the tasks of both men and women in her Peruvian research. Some of the most successful and fascinating fieldwork is done by teams of men and women. which has often paid little heed to the lives and concerns of Women. In fact. Jackson 1986. I should not go into the island store if there were more than two men inside. however. and Kathleen DeWait were investigating the nutritional strategies of older adults (Quandt et al. such all-male occasions soon lost their interest for me. I said that I would like to see it and was brought over to the meeting site by a middlepaged man from the village. they had heard virtual1y l l l J l l l l l l l l . they said. that is nol always possible. Lutz. Beverly Morris. Sara Quandt. gave me a first elementary primer on what 1 should and should not do. Women in the field have often been harassed and have become victims of violence in ways different from men (Warren 1988). he did not respond. and class (hooks 1989. and. It is a matrilineal society.

Although his initial agreement to this arrangement was made aimost casually in a conversation with a village headman. ") Over the next hour or so. resulted in rich data on those" aspects of life (B. we would not claim that working as a couple has given us any "special". Mr. following Newton (1993). Our own experIence has been that haVing a man and woman involved" fieldwork at t~e s~e time has provided a more balanced view of communitY hfe. and their eventual marriage and moving the United States. ~'1ean Gearing (1995). DeWalt 1979. I rest between field worker and informant didn't exist... and of the interaction of households and families than' we w. access to different settings. he looked at UqUlllas. or at least our ~ sexuality (Kulick 1995:3).into the co~munities and people we have studied. It IS much more hkely that collaboration by males and females in field research can occur."Good's account is less about sex and sexuality. the sexuality of the field worker hasn't ". An engaging and personal account." (p. Kulick 1995. the problems of drunks and how they were dealt wIth m the commuDlty. ched to Yarima during several years of returning to South America. "'Kulick and Willson (1995) find the taboo on discussing sex and the sexuality of '~ ethnographers somewhat curious. but felt would have been '0 much discussed by ethnographers. became his "girlfriend" and eventually arried him. N. IlIat silence about sexuality has served the purpose of "fortifying male hetero. . The presence of another man gave Mr. For this reason. the eventual consummation of their relation'p after she began menstruating. which was viewed as appropriate by the community. The book also includes observations from Yarima's perspective 'Good 1991). He includes among these: the supposed objectiVity of the observer (Dwyer 1982). and bodies of ~owledge. and more general cultural taboos about discussing sex. She had no idea which.h he had be~n a great source of information about use of wild foodS and rec~pes fo~ cookmg g~e he had never spoken of drinking or moonshine produc~lOn. Any ethnog. B talked abo. that sexuality should not make a difference in the objective recording and analysis of the customs and habits of other people.. but at the same tIme. she learned in graduate school llbecause it was never mentioned-that erotic inappropriate to discuss with women.adorIan socIologIst who had expressed an interest in visiting the Kentuc field sItes.. community v~lue~ concerning alc~hol use. Fortunately. .in the field part of the methodolo~ical 'theoretical trammg of many anthropologISts. either in their monographs or methodOgical notes. We base this on not only our experIenc~ m ~orkl~g WIth one another in a number of projects. and ethnohistory. the r~sear~h team traveled to Central County with Jorge Uquill an Ecu.wn spec181 perspective to the field.' rapher br~ngs theIr o. Some well-known exceptions to the ong silence about sex and sexuality in the field include references in Malinowski's " 'es (1967). both through a shifting in her relationships with others "and with her husband as an informant. the general disdain (until recently) in the discipline for personal narratives Q'ratt 1986).J" much less data on issues like kinship... K. would be inappropriate. B the opportUDlty to talk about ISsues he found interesting..lnsight into Vincentian life. however. Single ethnographers with characterIstIcsthat dIffer fr~m our Own would not have discovered some things that w. but opened up the opportunity to gain significant '. people. T' L L L 280 DEWALT I DEWALT Participant Observation 28/ L nothing about alcohol use or production of moonshine. and saId Are you a drinking man?" (Beverly whipped out the tap recorder and switched it on. Even carefully word and rather oblique interviewing about alcohol. . insight . and provided a number of stories about moonshi~e in Central Co~nty.field. He describes . men and women have been about equally represe~te? among stude~ts entering cultural anthropology programs.. Kulick (1995) reviews some ofthe factors implicated in the lack of discussion 'of sex and sexuality in fieldwork. One of the informants they visited was Mr. . but describes his relationship with . but also on' other proJects m whIch we have worked with larger teams involving men and women. One day. a . and Bdl s mterests m economIc and agricultural issues.Xuality by keeping above the bounds of critical inquiry and of silencing women d gays" (1995:4). Lewin and Leap 1996a). sexuality.ould ~ave ha? if >. not only increased ber acceptance in the community. Cesara's (1982) reflections on her fieldwork experience among the Lenda. how he dealt with his rage and jealousy after 'arima was raped by another Yanomama. we ha. c:ouldn't be mentioned.e noted. Good makes no claims that this !ationship enhanced or hindered his understanding of the Yanomama. He first agreed to become betrothed to Yarima when she was 'sS than 12 years old. there are other behaviors that our own perspective ' dl~ ~o~ allow u~ to see. and from one of the very few women to speak f this. symbolism.e had worked alone. a natural storytelle who had spoken at length about life of the poor during the past 60 years Althoug. '? L. of key relatIOnshIps. 213). even m the chmate of reflexivity. describes how she became attracted to her "best formant" on the island of St.s'increasing emotional involvement. She argues persuasively that her romantic relationship with a :Vincentian. Tumbull's 1986) mention of his Mbuti lover. B. On the other hand. 'yamomama woman. At the same time. as ethnographers haven't hesitated to discuss '0: sexuality of the people they have studied. was met with 11 denials. Rabinow's (1977) discussion of an affair in Morocco. I L Up Close and Personal: Sex in the Field As several re~ent writ~rs have noted (Caplan 1993. J: "tn the introduction to one of several recent volumes on sex and sexuality in the L.o~ was a discussion of sex . however. Vincent. Kulick suggests as well. K~thleen's interests in medical and nutritional topics. Good became more . Wlthm.:ew mmutes of entering his home on this day. for many decades. Esther Newton (1996) has written ~. Men and women have. DeWalt 1983).

sexual relalionships in the field raise two issues that are quite Im~o~ant to reVIew here. summanze. Leap 1996. and relatively powerless. As several writers have noted. What is the for all researchers.282 DEWALT I DEWALT Participant Observation In recent years. and gender put most ::'ethnographers in a more powerful position than the citizens of the communities in '.. IS. The answer to this question. ~or example. 250).nts not elicited by foreigners" as data (p.ethnographer is a white. heterosexual male from a developed country? Despite the Ffeeling of many ethnographers (especially new researchers) that they are dazed. especially women. Altork 1995.1S tIme to ask: "What are the implications of the anthropologist as a sexually cogmzant knower?" (p. Ibis has resulted in a number of.d t~at "conclusions based on sexual partlcl~at~on are dlst?rt~d by ~on~smg the mtlmacies possible with strangers with natlve s everyday mUmate Itves (p. then. However Murray argues tha~ "~aving .. Several of our . (Caplan 1993.admlts to havmg thought about the urepresentativeness" of his "sample" at one po.'.lOn for ha. . Discussions of sexuality.however. Kulick a~d Willson 1995.. especially when the .. 242). Gender differences may not be at issue for gay and lesbian (researchers.:::-female graduate students have found themselves in awkward situations because of lbe perception among some Latin American men that all U. etc.:. c'confused. he suspects that even behaviorunder these Circumstances IS adjusted to fit what the participants believe the researchers want to kno.context of intimate relalionships. . It seems apparent that much greater discussion is required concerning the elbics of sexuality as a research tool or as a component of . to what extent does sexual activity place researchers.n~er . Further. to homosexualtty m Central America only in retrospect.(19.odological approach.1996) has written several thoughtful essays on the use of info~­ matl?n ga~ned ~unng sexual activity as data in research. 'J'What are the implications for the use of information gained during sexual l -. ~f contributors to his and several other volumes. Bolton . A second trend is the increase in research on gay and lesbian" commumtles by gay a?d lesbian ethnographers. at risk for sexual assault? Both women and men may be at risk for sexual assault I l i i 'l l l .Which they work.. 6). It would seem that this potenlial would be magnified in the . . Kathleen included. ~olton 1995. . have experi- l l l l I Eve~ this cooclusl.sex with the nati~es is not a royal road to insight about alien sexu~h~les. ~ccoun~s of the expenence of being a "native ethnographer. We have developed elaborate guidelines for ensuring that the people we study are given a full explanation of the purposes of the research. having sexual relat~onshlps wIth other ~ay ~en in Guatemala became fieldwork on eliciting terms rela~mg. participant observation raises a number "potential for sexual exploitation of research participants. ethnographers 'who discuss these issues argue that intimate relationships provide access to infonna'iion Ibat might not have been available otherwise (Gearing 1995. KUIi:1995. Lewm an~ ~eap 1996). provided by a number enced reluctance and even hostility from potential women informants as a result of their expectation that "V. Lewin and Leap 1996b: Ne'.dld have intimate relationships with men (including married individuals) in the Community.To. Sev::~ ' on re f1" rary. . As Kulick CI995) puts It. 1996. S.95.1s conducting research. the.• emp haSIS eXlvlty wh' IC h suggests that the ethnographer is situated sexually well as wit~ respect to. Lewin and Leap 1996a). class. For Murray. he IS ~onceme. It. although he . Should we be ~developing informed consent scripts that can be whispered at an appropriate moment? This is not an idle question.2?0).meth. is that there are a number of l factors have contributed to this mcreased attention. they ended up not having sex.?' . several volumes of essays have been published that deal m 'h d 'Irect Iy Wit th e 'Issues of sex and sexuality in the field (Whitehead and Cona ore' 1986. -t'' 1: Intimate relationships. these problems were the results of previous experience with researchers who . He pref~rs "native doc~m.. the matter of informed consent. • >~f The second issue has to do with participation. Gearing 1995. gender."PArticipation in sexual relationships may be important to acceptance in a community 'and the development of rapport (Tumbull 1986). which is becoming more important ')iresearch involving human subjects. class. Bell et al. l l i . discusses not only Ibe ~ays m whIch hiS homosexuahty mfluenced hIS examination of gay communities m the "years of the plague" [AIDS] but also how his sexual activity became "data" ~urra~ (1991. both the effect of being accepted as a "native" and the information they gain as a result of sexual activity (Cesara 1982. Increasedattention to reflexIvity suggests that sexuality is a key characteristic of the observer apart from. -.. there is' mcreasmg acceptanceof the discussions concerning sexual relationships in the field' 283 . One has to do with observation.(although obviously these two cannot be separated).: Finally. sexual relationships on the experiences of subsequent researchers in doing ethnography? Several women researchers we know. but class and ethnic differences are still likely to be important. differences in race. Murray 1996." 'l~ part. His primary '"~encounters as "data"? As we noted earlier. women are "loose. ". principally because it's not always clear when the ethnographer ~. Coming from a more textually oriented t~eoreticaland . More commonly. 1996). and we stress that we should always ask permission when making sound ior photographic recordings. women" were out after their husbands. rather than attraction. The first is the contempo lniplications and recognizing the observer as a sexually situated observer is lInportant both in the writing of ethnography and the reading of the ethnography. . Murray 1996).Another disturbing question that arises for us is: What will be the impact of . ge. particIpant observaUon with interviews and other research materials. I~ part.!on ~996).mt. In the one instance in which he reports that he went with a man out of curiOSIty. raise several ethical questions. I?93.~eral et~nographershave written about the impact on their research of a fuller particIpatory mvolvement in the community itself. (I996.doing field research.S. race. J mottvat." Finally.S. ':of ethical questions. become part of the' process ~f:eflexlvlty.ving sex with Guatemalans was not to recruit informants.n.In this vein. the result of his juxtaposing "experience " near data from. however. becomes extremely relevant.

. During their first field .' married for more than a year without having a child and without signs of Kathleen '~. noting at the same time that this probably is a significant underreporting of rape. Lee 1995). In the literature. The risks-ethical. learned how Hutterite children engage in very different behaviors when they are in . Klass and Solomon Klass 1987). it can also teach '~i. G. And children can help gather data that is inaccessible to adults. " may be viewed as strange. although the results of these encounters has further reinforced our contention that sex and fieldwork is not a' good combination. many anthropologists bring their families. as a community member. strategy that became much less effective When he became sexually active. it was a concern for many people that we had been t. fie. including their children. Many Sudanese doubted . we have always strongly advised students that sexual relationships with informants or other individuals in communities in which. (ethnographers are. She had been using him as a foil against other men. but several cases are personally known to us.(Flueh~­ t'Lobban and Lobban 1986. Others have. the anthropologist.thropologists have written about their experienceswfili" 0' chIldren m the field.~receive a constant stream of advice about child care from friends and neighbors. Children can help ease the loneliness and isolation characteristic of fieldwork in foreign cultures. but. can serve as male protectors. Huntington (1987) reports that her nine-year-old daughter was an invaluable source of information about Hutterite children's informal culture. to the field with them. we were working should be avoided.ldwor~.. the presence of children accompanying a researcher can signilY his or her adult status. the villagers with whom she lived devoted a lot of time teaching .. even in their' In most cases. :)i· . When informants are enculturating children. or invent. Our own' advice to graduate students has often been ignored. In many cultures a childless adult. (Nichter and Nichter 1987). of his woman companion. absence. or the object of pity.iexperience in the Sudan. E. they also present a number of challenges to anthropologists in the field. especially a married childless adult. Like many other anthropologists. Renate Femandez (1987) learned that children sleeping alone in their own room was viewed as a type of social deprivation in rural Spain. L I Participating and Parenting: Children and Field Research" Fieldwork is traditionally portrayed as a solitary endeavor. . dangerous.kthat they were truly married. a number of themes emerge regarding the effect of children on participant observation. Many anthropologists report that their children had a positive impact on participant observation. During Diane Michalski Turner's (1987) fieldwork in Fiji. after all. personal. Mimi and Mark Nichter ·i1::' reported that they gained valuable Insights into rural Indian ideas about child \1t" development from villagers' comments made about their son's "constitution" ~~. An interesting byproduct ofthe discussion of L L L sex in the field provided by Murray (1996) is his suggestion that his beginning sexual activity with other men in Guatemala may have compromised the position.".· L L L 284 DEWALT / DEWALT Participant Observation 285 in the field. Bemard (1995) notes this advice as common for beginning anthropologists. Mirni and Mark Nicht~r '(Nichter and Nichter 1987) believed th~t the p~esence of ~heir young s~n ~ade It easier for villagers to relate to them dUrIng their research m a rural IndIan vIllage. Bringing children into the field can also open up new areas of information to .Bourgois (1995) describes how his son's cerebral palsy was diagnosed in a clinic . rolling his walker over trash and crack vials in the streets of El Harrio helped establish ~:Bourgois '~ultures. During oUr first field t experience in Temascalcingo. In El Harrio. a number 'Of an. suggest that the risks. the people who are being studied are' able io relate more easily to an anthropologist living with his or her family.9aL. a.her two-year-old daughter to become Fijian. As part of the trend to demystilY anthropolo&). Michalski Turner was able to learn not only how one becomes Fijian. they are also teaching the anthropologist about their culture. The presence of children in the field shapes the research experience in a number of distinct ways. Bringing children to a field site can lead to increased rapport with the research community. 1987) reported that people had trouble acceptmg theIr status as a married couple because they had no children. Howell (1990) reports that 7% of a sample of women anthropologists reported rape or attempted rape in the field. While an overabundance of friendly advice can be exasperating. in reality.> op ':j :'1. Most researchers who bring young children in the field often ':. being pregnant (see also. burly husbands and boyfriends.' are not always great and a "blanket prohibition" is not only impossible:. By watching how villagers interacted with her daughter. anthropologists about the culture they are working in. but their return to the field ten years later with their daughter Josina reassured their friends and acquaintances. unexamined cultural biases and assumptions. However. WhIle each fieldwork experience is unique. Carolyn Fluehr-Lobban and Richard Lobban . but also about Western/Fijian power relationships. Rape ofth~ researcher in the field is not often talked about. Some women attempt to assume a "sexless" identity to help protect against assault. Huntington . though. • ~. and that his son's ability to negotiate the neighborhood. A solitary anthropologist showing up in a remote area 'L L L L to live alone for a period of a year or more may seem extremely bizarre in many Also. but the risk for women is higher (Warren 1988. who. The narratives of researchers who' have developed intimate relationships in the field. and to the research enterprise-have always seemed too high to us. human) but perhaps not even desirable. Anthropologists can also learn about a culture from the way people react to their children. This advice can also challenge the anthropologist's . Eva Moreno (1995) discusses how a combination of ambivalence and inattention to sexual cues resulted in her rape by a male research assistant. As a result.

As our children moved into their teens and began to have obligations and WIShes of their own. Most anthropologists who amve at a field SIte for the first time with their children seem to experience more. In the most striking finding. Joan Cassell (1987) relates how she frequently missed nighttime events because she felt compelled to stay ?ome . Scheper-Hughes 1987). The relatively small amount of formal examination of ethnographer bias in anthropology is evidence that these issues merit much more attention than they have previously received. unruly children can also disrupt meetings and mtervlews. Mead [1923] and Freeman [1983) on Samoa. problem~ than veteran field workers (Cassell 1987. Bringing. we could more easily afford child care in Mexico or Honduras than we could in the. This acceptanceof the reliability ofdata contrasts markedly with the controversies embroiling anthropology and other social science disciplines concerning the interpretation or theory built with the data. however.J m our careers when we were relatively poor graduate students or assistant professors. l l . adapt more readily and experience less severe culture shock (FluehrLobban and Lobban 1987. states. Until recently. Nichter and Nichter 1987). United St~tes. During these years. and the field worker's family situation suggest. Everyone had denied that there were any such beliefs in the community. Redfield [1930) and Lewis [1951] concerning Tepoztllln in Mexico. one of our key informants began regaling us with a recounting of l 1 i l --. i I . especially those solely responsible for child care. of course. it became more difficult for us to take them to the field. depends on having reliable data. sexuality. early I worker's time in the field with his or her children (Fluehr-Lobban and Lobban 1987. it's quite apparent that these personal attributes can substantially affect participant observation in field research. Reflecting on her research in Jamaica. In addition. Whether children help or hinder participant observation depends on a number of factors. who brought three children to rural Jamaica to conduct postdoctoral research with her. Older children seem to have a more diffi.Itself also shapes the experience with the children. has its disadvantages.i' " J. Before our brief hiatus. Building theory.J. Perhaps the biggest disadvantage of bringing children to the field is the amount of time that researchers devote to child care (and hence lose to the fieldwork). ethnographers surely differ in terms of their abilities and qualifications. 1970). SomeAnthropologists report that the responsibilities of child care forced them to miss' out on certain opportunities. Above we referred to leaving the field for three weeks after our first six months in the field. I Ethnographer Bias Because participant observation is perhaps the quintessential qualitative method. The pioneering work on ethnographer bias was done by Raoul Naroll (1962. we began to schedule our time in the field separately so that one of us stayed in the United States with the children while the ' other was engaged in doing field research. Many anthropologists. however. It seems that there is a significant effect depending on whether the anthropol. who became concerned that his cross-cultural research results may have been affected by systematic errors in ethnographic reporting. Very young children.286 DEWALT f DEWALT Participant Observation front of adults and when they are among other children participating in their own culture. ?~r personal exp:riences with children in the field have been generally quite poslllve.cult ~im~ adapting to new and foreign cnltures (Scheper-Hughes 1987). We also found that. Bringing children into the field.ogist is retumi~g to a site or arriving for the first time. Hugh-Jones 1987).with her two children. particularly because these topics were so relevant to KathJeen's medical anthropological research. report that the presence of children severely curtailed the amount oftime they could devote to fieldwork. As the above discussions of the effects of gender. children to a field Site where they already speak the language is easier on both the Children and the parent than introducing them to a culture where they are only able to communicate with their family (Cassell 1987. Melanie Dreher. During fieldwork among an indigenous tribe in the northwest Amazon Christine and Stephen Hugh-Jones (Hugh-Jones 1987) had to devise altematin~ fieldwork schedules so that one of them would always be available to supervise their two children. Naroll (1962:88-89) found that the incidence of witchcraft reported in particular societies was related to the amount of time the ethnographer spent in the field. Michalski Turner 1987). we had asked people many times about magical and witchcraft beliefs. however. Almost the very day of our return. the question of reliability is critically important. Postmodemist writers particularly emphasize that the observer and his or her circumstances and biases cannot be separated from the accounts that they write. it has been rare for the accuracy of field reports to be questioned. He s~owed that ethnographers who spent more than a year in the field were sigmficantly more likely to report the presence of witchcraft beliefs among the societies they studied than those who spent shorter amounts of time in the field. Th: field sItuation . Our two fall-haired children were an instant magnet everywhere we have traveled in Latin America and opened many doors for us. The age of the chIldren also shapes the field experience. while needing more care."l suspect it took me twice the time to accomplish half the work that I would have normally accomplished" (Oreher 1987: 165). Our own research in Temascalcingoprovided a striking personal confirmation of Naroll's finding. This is so despite an increasing number of controversies coming to light in which the data collected by different anthropologists who have worked in the same area differ substantially (for example. I I I l 1I . Young. The presence'" of another parent to share child care responsibilities certainly facilitates a field l l -. so it's lamentable that so little attention has been given to the issues of reliability and validity of the information collected. and Benedict [1934] and Bamouw [1963] on the Zuni).

although we still occasionally hear similar sentiments from some of our colleagues in anthropology. We believe that this is a form of intellectual elitism. and others both as a method of dying other cultures as well as a personal means for coming to terms with their reactions to their research. And. we' improve Our observational skills en w...am(p.or example lan u b'l' trammg.' . and hang-ups to the field with '\' however. L L L 'viduals we study should be a beginning point. . Abram Kardiner. laid the groundwork for much of the theoretical development autoch.. the increasing ..argues . ') I utilizing more formal methods of d~ta all IS . by one of the parties that witchcraft w b' . pertence . appa~ent. 288 L i DEWALT IDEWALT Participant Observation a conflict that had occurred during our absence Th '. . Although tied . the reliance on .anthropological research since the beginning of the twentieth century and has been distinguishing characteristic for anthropology compared with other social sciences. their training and ex .': observation. They then went about the business of trying to l!bnstruct social scientific explanations of peoples' behavior through ethnography. should attempt to make thes:~~a:::e VIew ?t~er cUltu~es. ~n cross-cultural research since N~rojl':re~~ has b:come increasingly common y Wor . .s eY a.explicitlyai~( ethnography" (Tedlock 1991'69) Th " r .)... psychoanalysis was commonly used by anthropologists like DuBois.. is a continuing demystification ofthe process of fieldwork and ethnographic writing.ntatlOn.~t mcluded accusatio remaining months in the field wI'tch raft as emg used agamst them. quality of c. not the final product of ography. Our reporting. identify alternative strategies and begin to craft their personal approaches to participant observation '. e pomt IS often made th t" b' tOO " I pOSSIble m the study of human beh ' Wh'l a. reactions.8 . and perhaps theirtheoretical orie ' ' .gender. all of us bring biases r d' .or a mIx of meth d ' h' observatIOn is just one of the tools that a th I' 0 s m w Ich participant of the people they study.. wlthm a smgle narratIve . don't accept the corollary that is O~VIO. aroll ~1~62) and We can atte make a very large difference in thePkl'nd f gfi gd~d m partIcIpant observation doe 0 m mgs that may ber rt d . . The movement into more reflexive ethnographic writing has resulted in a quan~ tum increase in the number of accounts of the fieldwork experience presented by ethnographers from a number of different theoretical approaches.~ un I ely to report the negative (1961 :381). may e a natural sub·. Making explicit the process of participant observation allows the reader to better understand the information presented by the ethnographer. ethnic' .. The work of Naroll and others who followed his lead are suggestive ' 'of how informative such studies can be. They quoted Levi-Strauss . while not atheoretical (no method is) is so closely tied to a relatively unchanging theoretical core in anthropology as to provide the basis for a wide range of theoretical development around that core. One among the people they study. 0 Ias m reporting b t • rejectIOn and its importance in personality d v I .. Another example of the Importance of ethnographer bias come fi Their work focused on the e/liects f b~ r?m the work of Rohner et al. Rohner ~t al n an o~: ne~altve personality traits at thIs seems to be linked to a "bias of romanticism" among anthro I' reporte other than just participant observati:"o ~tsts. age. ~ at therefore we should not strive to human behavior. . Indeed. . ~owe~er. a convinced opponent of trad'f I . and other factors affect their 'Observations. sexual orientation. We are convmced that the willing level of confidence and comfort with u Th ernes ~Ilected a breakthrough in their the length of time that a person s en~~ en u~ as . thonous to anthropology. (1973) .e.: e e. who observed that "At h than d hIves. and we cannot completely esca~:t: ISposltlOn. .comes respectful of even the most. bur perspective is that we should go beyond the individual postmodem musings that ire too common in contemporary anthropology to more systematical/yexarnine how 'lbe anthropologist's race. parental acceptance e striking finding of these analyses was th t ehopment m chIldren and adults. eontro IImg for sources of ethno ra h b' epo e . but no Sooner has he in focus' conservative practices" Th. Narrative ethnography (Tedlock 1991) and personal accounts of field experience also provide the opportunity for new researchers to begin to anticipate problems. makes clear..~~~ ification efforts report more parental re 'ecti: et .!~articipant Beyond the Reflexivity Frontier ·. may use these in judging Our work Wh t ' aS expltclt as pOSSIble so that others. The approach to training in ethnography common a generation ago held that each new ethnographer should go out and reinvent anthropology methodologically and sink or swim on their own ability to do so." I ~ we can agree WIth this position.s. it may be essential to anthropology. It may be constitutive-that is. . L L L ' " The "quality" of participant observation will characteristics ofethnographers (for ex I vary dependmg on the personal affiliation). The result." toward presenting both "the Self and Oth.. . unle:~ethnographers Use methods aspects of their subjects' personaliti. field notes. etc. suggests that the method. DUring th • J C was a common them f WIth peopie who had denied its existence befor ~ 0 our conversatio ness of people to talk with us about such th e. ' . sexual preferences.ltm m ant~ro'pology. n ropo oglsts use to find out the behavior 289 L. gender. Understanding ou orIsearc d or expl~natory theories concerning rse Yes an Our reactions to field research and the observation as a technique of fieldwork has been a hallmark of '.'s a. reportmg. which also include the incorporation of the ethnographer's thoughts and .a ou. That the descriptions of the research enterprise provided by Malinowski in 1922 and Bourgois (among many others) in 1995 can appear so similar.: a society different from his Own th 'I~na busage. is that by. participant observation and the recording of chronologically oriented descriptive . to functionalist theoretical approaches early in the century.a so. we believe. e conll. Ilty. . Ruth Benedict. As mterpretlveanthropology " us. ame e ant Topologist b ' verslve.og~ap?cers W?O l~se multiple ver:-." early in the fieldwork experience. 0 ~ec 'Vlty' IS not.g age a. we may improve the qual'tyco edctlon ~n cOlljunctlon with participant' I an conSIstency of our ' M uc h ofthe recent trend in postmodemist w " g.

The aim sho. our responsibility as ethnographers is to try to ensure that the people we study are not harmed by our personal involvements with them and are not negatively affected by the information we collect and write about them. One hypothesis that was very common in the literature was that as people switched from semisubsistence crops to cash crops their nutritional status became Worse. Systematic study of the effects of biases. We accepllhat none of us can become completely objective measuring devices. We need to be aware of who we are. l l l l l l l I I d'·': ~ I " . Finally. while we should be aware of our own identities and how these may affect our field research.. Certainly. we should continue to work toward scientific observations of people and their cultures.. Our experience has been that. That is. or people who don't look or act much differently from ourselves-is an essential tool. Rather than using the latter as justification for givin~ up on making participant observation and other anthropological methods mBre"" verifiable and reliable. When people switched from semisubsistence crops to cash cropping. ~. some of which are specified in her work. participant observation raises many important ethical issues for ethnographers. to draw important policy conclusions. are also involved and must be studied (DeWalt 1993). Both of us have engaged in comparative work in which we have utilized the work of other anthropologists and other social scientists. There is no substitute for gaining tacit and implicit knowledge of cultural behavior than living among people and sharing their lives. there is the problem ofmaintain~ ing the anonymity of the people whom we study. There are no easy answers to these ethical issues. we acknowledge that every one who chooses to use this method will bring his or her own biases. An example of Kathleen's work is her evaluation of research on the relationship between cash cropping and human nutrition. We can. these problems require discussion in methods classes. In both cases. In a similar vein. we believe it's more productive to focus on the generalizations that can be derived from such data. The objective of ethnography should not be to learn more about ourselves as individuals (although that will happen). Conclusions Participant observation is the hallmark of anthropological methods. It was possible. In this chapter. she was able to show that there is not a simple relationship between these two phenomena. what is heartening to us is that much previous anthropological and other social scientific research can be quite useful in building generalizations.: l l . reliable data concerning human behavior. but we see these as complementary to the use of participant observation as a means of collecting verifiable. despite differences in theoretical perspectives. and personal characteristics on the research enterprise is a valid social scientific endeavor and requires further development. Approaches that emphasize the "observation of participation" are quite useful and important (especially for training budding anthropologists). as a starting point. the hroad-brush observations of indi~idual researchers concerning human behavior are relatively consistenl. however. Summarizing studies from around the world. should they establish sexual relationships with informants. we see reflexivity as a beginning point rather than as an end to ethnography. we have focused on the most important of these and have tried to convey what we have learned from our own personal experiences and those of other anthropologists. 1996). but to learn more about others. Our own view is that. and other personal factors.uld then be to improve our methodological skills to work toward bUlldmg generalizations that are I even stronger. an exotic subculture of our own society. In addition. Once we have done that.290 DEWALT I DEWALT Participant Observation 291 I number ofmonographs and textbooks addressing issues in ethnographic methods and the number of formal courses in methods available in our universities suggest that we have gone beyond that time. gender. often for their oWn protection. understand our biases as much as we can. like the processes of doing participant observation and fieldwork. there has not been a necessary improvement or decline in human nutrition. The active engagement of an ethnographer in the lives of the people being studied-whether they are a group from halfway around the world. We believe that the practice of participant observation has been one of the catalysts for theoretical development in anthropology. These issues include the problem of establishing "limits to participation"-should ethnographers engage in illegal behaviors. predilections. If we look at the fine detail or if we look for consonance in theoretical conclusions we will find many differences. The primaty message is that. and should they take up the causes held dear by the people we study? Another ethical issue on which we touched was finding the proper mix between participation and observation. use participant observation in conjunction with other methods to serve anthropology as a scientific pursuit. ethnicity. From a personal point of view. on the basis of comparative analysis of the ethnographic materials. places. he and his collaborators were able to find common patterns in the data presented by individuals working at many different times. and to understand and interpret our interactions with the people we study. Other factors. we can strive to detennine whether there are regularities in human behavior. and with different theoretical perspectives. predilections. and personal characteristics with him or her and will face a number of challenges and choices.:. Bill has looked at the literature on agrarian reform communities in Mexico (DeWalt and Rees 1994) and on development in indigenous communities in Latin America (Roper et al.

Cambridge: Cambridge University Press. In the former case. and observation serv as source of questions about which to interview. 4. MichaeL 1986. :'. New York: Houghton Mifflin. VoL is participation in the context around us that allows us to gain insight into the tacit. Bolton. ~L 2d ed. Introduction. Men and ~'" I ' . Urbana: University of tIlinois Press. Peggy Golde. Code· of Ethics of the American i Anthropological Association.htm. We may find it difficult to articulate what it is that makes us feel uncomfortable because these aspects of cultural knowledge remain outside of our general consciousness. Coming Home: Confessions of a Gay Ethnographer in the Years of the Plague. Spealcing ofEthnography. '(Altork. 1996. 1994. 2. Patricia. Malinowski further cautions against living in compounds apart from the people ubder investigation like other Olwhite men" do and insists on the need to live in the community. Diane Bell. The Hold Life Has: Coca and Cultural Identity In an Andean Community. but allowing these to emerge out of the flow of everyday conversation. Walnut Creek. 107-139. 1988. however. he sees participant observation as providing the context for the rest of the enterprise. 377-392. In Gendered Fields: Women. http://www. In Out in the Field: Reflections of Lesbian and Gay Anthrcr pologists. " 9. 1996. Thousand Oaks.' Anthropology that Breaks Your Heart. Philippe.'\ Oaks. Bernard. Pp. Culture Shock: A Reader in Modern Cultural Anthropology.~ 6. Don Kulick and Marga. Nader also quotes Robin Fox as commenting after a trip to the Southwest: "There were all the anthropologists. In Search of Respect: Selling Crack in El Harrio. In Handbook of Qual. Caplan. . Victor. eds. . Ralph. andErotic Subjectivity InAnthropological Fieldwork. CA: Sage Publications. Pp. This brings to mind a well-known joke in anthropological circles about a "postmodem· anthropologist" Postmodern anthropologists are very concerned about documenting their. Washington. 8. Can we now talk about me for a while!" 10.t Willson. 1970." Building rapport often involves not directly addressing certain issues or asking pointed questions. 1934. The tourists were asking the Indians all the questions that the anthropologists wanted to ask. 1995. and Wazir Jahan Karim. Observational Techniques. Patterns o/Culture. There is no doubt for us that greater understanding comes from living in the community. It '~gar. 11. Boston: . Pat Caplan. the native becomes restless and says to the ethnographer: "Excuse me. Berkeley: University of California Press. eds. In Taboo: Sex. Pp. 19-27. 1994. I ~.Behar. However. eds. Bolton. 1995.ldentity. Homewood. 147-170.Briggs. IL: Dorsey. Don Kulick and Margaret Willson. Ralph.. 1963. MichaeL 1996. Bourgois. London: Routledge. :" 5. but we've talked enough about you. Thousand :. Tricks. 19-44. Ethnography. 1993. In Women in the Field: Anthropological Experiences. Jean. The Vulnerable Observer." As she points out. Experience. /~'. . in each case the individual continued to publish ethnographic accounts (1991 :70). pp. CA: Sage Publications. Finally. and Peter Adler.. Kate.Bamouw. and Erotic SubjectiVity in Anthropological Fieldwork. 3.orglethcode. of communities. own personal responses to what the people they study are telling them.~' itative Research. Pat. 1986.ry: American Anthropological Association (AAA). Philip K. . eds. 'Agar. Gendered Fields: Women. Postmodern anthropologists may base their analyses on "texts" that can be written or spoken materials. During this research project. New L L York: Knopf. London: Routledge. 140-167. 12. DC: Smithsonian Institution Pres!l' . Confronting Anthropology. Men and Ethnography.can Anthropologist 98(2):249-265. . Philippe. Tedlock discusses a number of cases of anthropologists who are candidates for having': "gone native. 1997. Walking the Fire Line: The Erotic Dimension of the Fieldwork. Identity. The ProfeSSional Stranger: An l'I!ormallntroduction to Ethnography.'''''e'"''' L I '- 292 DEWALT I DEWALT Panicipant Observation 293 NOTES REFERENCES L L L I l o_ 1. Amer. This ethnographer is conducting an interview with a "native" that stretches on for many hours. Ruth. Qualitative Research Methods Series. Most anthropologists. In terms of ethics. London: Routledge. Norman Denzin and Yvonna Lincoln. no face-to-face interaction with informants is required. however. Friends and Lovers: Erotic Encounters in the Field. For Agar. CA: AltaMira Press. Pp. H.. it is important' that ethnographers always explain how recordings will be used and ask permission before taping ' or videotaping any interview or conversation. .:. would argue that the range of behaviors that are susceptible to pure observation or can be understood through pure observation is small. > 2.Allen. Culture and Personality. London: Routledge. WaYland. In Taboo: Sex.t Although Good saw the violence in their culture. 'Adler. ~. This section was prepared with the assistance of Coral B. Pat Caplan. 1995. Eilen Lewin and William Leap.ameranthassn. Ruth. Beacon Press. 2d ed.: Bell Diane. the ethnographers had asked permission and re~orded many of their conversations with informants. 2d ed. Russell. Katherine. Education and Inner-City Apartheid. Pp. and there were all the tourists. 1996. he also saw a substantial amount or" harmony and group cohesion. Good reports that it was this experience that contributed to his increasing divergence from his then-adviser's portrayal of the Yanomama as "the fierce people" (Chagnon 1983). Research Methods in Anthropology: Qualitative and Quantitative Approaches. but didn't because they were afraid of ruining rapport (Nader 1986:113). 7. and Wazir Jahan Karim. eds. the interview is more important than the participation. j ~ Benedict. San Diego: Academic Press. L L Bourgois. We have carried out fieldwork in an essentially "commuting" situation and as residents. Kapluna Daughter. 1993. ed.

Gender.ichae~.Jean.htm Castaneda. 53-71. R. Joan CasseIl. Dreher. Clifford. Joan Cassell. \57-\88. 1957. Lnndon: MacMillan. '·cll. eds. In Children in the Field. Introduction. Pp. Special Publication of the AAA #26. and Martha W. Gender and Methodologyj : the Sudan. Carolyn. Conaway.. Philadelphia: Tern Umverslty Press. 1987. New York: Simon and Schuster. Part-Time Crime. 1984. In Children in the Field. and Sheila Solnmon Klass. '''. 121-147. L. Identity. . Ladislav. eds. Kathleen M. Pp. Participant Observation. Human Organization 43:16-34. E. Routes: Travel and Translation in the Late Twentieth Cent : Cambridge: Harvard University Press. Yearning: Race. ' Firth. In . nsen. g. In Children in the Field."can EJido: A Study in Economic Adaptalfo' New York and Cambridge: Cambrtdge University Press. D.fieldwork. In Apprenticeship: From Theory to Method and B~ Agam. Jam~s. 1987. "an. 47-70. Into the Heart: One Man's Pursuit 0/ Love and /. Carolyn. Rineb and W 105tOO. Thousand Oaks. 1990. London: Rourledge. Joan. Don Kulick and 'ar8aret WilIson. eds. and Richard Lobban. 1987. F. Joan Cassell. Philade!phia: Temple University Press. In Children and Anthropological Research. loan Cassel!. In Children in the Field. Conaway. Clifford. Pp. Pp. Melanie. Pp. 27~3. E. Thinking Black. 1986. Clifford Geertz. Handbook on Ethical Issues in Anthrop%gy~ special publication of the American Anthropological Association. Jean. Clifford. J. Ann Arbor: UMI Research Press. an~ S~e Ellen Jacabs. G. M. '\cs. 1993. Washington. ed. .>alca. ed. Whitehead and M. Philadelphia: Temple University Press.'~ DeWalt. ~' Cesara. 1-26. . 1Jlhropological Myth. J gb-Jones. IL:' Waveland Press.S. pp. 1989. Morton. Jean. '-' - COY. ' DeWalt. 1983. 19~7. Boston: South End. On Trying to Be an Amazon. Rees. Fluehr-Lobban. 1987. Roger Sanjek. Social Science and Medicine 36:1407-1416. In Ethnographic '. E. In Man and Culture: An Evaluation ofthe Work ofBranislaw Malinowski. New Yo" Academic Press. 1984. with David Chanoff. 1990. Pp. Billie R 1979. Meyer. 3rd cd. They're Not My Shoes!": Fieldwork in the Blue MOuntains Jan. Fernandez. ed. and Cultural Politics.zation ~·n a Mex. Philadelphia: Temple Universily Press. ed.-Mexican Studies. Philadelphia: Temple University Press. In Self Sex and Gender in Cross-Cultural Fieldwork. L. 1990. In Taboo: Su. Joan CasseH. SurviVing Fieldwork. ed. One Anthropologist. llhaca: Comel! ~niversity Press. 1973. The End ofthe Agrarian Reform In Mexico Past Lessons. Birthing in the Bush. ~ bell. Christine. Billie R. Pp. New York: Simo~"" Schuster. Pp. 1983. Derek. Sex and Gender in Cross-Cultural . Body a~d Emotio~' Th~ Aesthetics o/Illness and Healing in t Nepal HJmalayas. 182-193. Drink from the Nile. Pp. Robert 1992. Barbara Butler and Diane t1ichalski Turner. Urbana: University "~f lIIinois Press. Mlchael Coy. . 149-171.search: A Guide to General Conduct. Man and Culture: An Evaluallon oJ the Work of Bronislow Ma/inowski. In Fieldnotes: The Makings ofAnthropolollY. Deep Play: Notes on the Balinese Cockfight In The Interpretation of • Cultures: Selected Essays. . New York: Plenum Press. Margaret Mead in Samoa: The Making and Unmaking of an . DeWalt. Pp. CasseIl. Theory Methodology and the Research Process. ' Dwyer. r&. 1957. 412-435. 1994. W. Four Decades. Urbana: University of l1Iinois Press. Pp. Yanomamo: The Fierce People. 1995. Children and Parents in the Field: Reciprocal Interests.294 DEWALT I DEWALT Participant Observation 295 .'Fieldnotes: The Makings ofAnthropology. Philadelphia: The University of Pennsylvania Press. 1987. 263-274. Talking Back: Thinking Feminist. Boston: South End. ~n. The Deveiopment of Rapport Through the Negotiation of Gender in 'fjeldwork among Pnlice. . Pp. Order Rules the World: Our Children in the Communal Society of 'Ille HUllerites. New :V~rk: Harper Torchbooks. ed. ly. 1982.Joan. Ma~da. DC: American thropological Association. Whitehead and M. Albany: State University of New York Press. Nutritional Strategies and Agricultural Change in a Mexlc. bell. 185-215. Children in the Amazon. Fear and Loving in the Westlndies: Research from the Heart.orglsp23. Reflections ofa Woman AnJhropologist: No Hiding Place. New York: Basic Books. Raymond.owledge among the Yanomama. 1987. eds. Nancy. l l I 'cd. ed. Pp. ed. 1982. T. San Diego: Center for U." Community.ameranthassn. New York: Harper Torchbooks. 13-34. Notes on (Field)notes. 1989. 1977. Renate. M?dern. CA: Sage Publications.' Dcsjarlais. Raymond Firth. Three Children in Rural Jamaica. Future Prospects. DeWalt. In Self. New York: Halt. Journey to IXIlan: The Lessons o/Don Juan. T. _Lobban. \86-218. and Erotic Subjectivity in Anthropological Fieldwork. 1972. tlntinglon. Prospect Heighls. ed. In Children in the eld. A Report ofthe Advisory Panel on Health and Softey in Fieldwork. London: Academic Press. Nutrition and the Commercialization of Agriculture: Ten Years Later. ed. 1997. 1990. Moroccan Dialogues: Anthropology in Queslion. Philadelphia: Temple University Press. "I Am a Fieldnote": Fieldnotes as a Symbol of Professional Identity. ! I 1 I l 1 i!:kson. ~apoleon. 1 1 -. Roger SanJek.~ Jennifer. Number http://www. 1989. In Children in the Field. 3-33. Families. 'f Clifford. 1983. "Oh No. 1995: After the Fact: Two Countries. Pp. Cambridge: Harvard University Press. Joan Cassell. ~bridge: Harvard University Press. Kenneth. 1986. ~ames. Kathleen M. Bllen. Malinowski and the Study of Kinship. 1987. and Richard Lobban. Pp. Ithaca: Camell University Press. ' Chagnon. Ditton. Kevin. cd. ed. 237-255. 1991. x-xv. Carlos. Cassel1'.







Participant Observation


Kornblum, William. 1996. Introduction. In In the Field: Readings on the Field Research Experience, 2d ed. Carolyn D. Smith and William Komblum, eds. Pp. 1-7. Westport, Praeger. ,


Kulick, Don. 1995. Introduction. In Taboo: Sex,/dentity, andErotic Subjectivity inAn/hropological Fieldwork. Don Kulick and Margaret Willson, eds. Pp. 1-28. London: RoutIedg~
Kulick, Don, and Margaret Willson, eds. 1995. Taboo: Sex. Identity. and Ero1l'c Subjectivity in Anthropological Fieldwork. London: Routledge.

l.turpby, Yolanda, and Robert F. Murphy. 1974. Women afthe Forest. New York: Columbia University Press. .'Murray, Stephen O. 1991. Sleeping with the Natives as a Source of Data. Saciety of Gay and ~)i;' Lesbian Anthropologists Newsletter 13:49-51.

t Murray, Stephen O.


Kumar, Krishna. 1993. Rapid Appraisal Methods. Washington, DC: World Bank.

1996. Male Homosexuality in Guatemala: Possible Insights and Certain Confusions from Sleeping with the Natives. In Out in the Field: Reflections ofLesbian and Gay Anthrapologists. Ellen Lewiu and WilIiam Leap, eds. Pp. 236-261. Urbana: University of Illinois Press.



Leach, Edmund. 1957. The Epistemological Background to Malinowski's Empiricism. In Man' and Culture: An Evaluation ofthe Work a/Branislow Ma/inowski. Raymond Firth, ed. Pp.
119-137. New York: Harper Torchbooks. . Leap, William. 1996. Studying Gay Englisb: How I Got Here from There. In Out in the Field: Reflections ofLesbian and Gay Anthropologists. ElJen Lewin and William Leap' eds. Pp. 128-146. Urbana: University of Illinois Press. ' Lederman, R. 1986. The Return of Redwoman: Fieldwork in Highland New Guinea In Women in the Field: Anthropological Experiences, 2d ed. Peggy Golde, ed. Pp. 359-387. Berkeley: University of California Press. Lee, Raymond. 1995. Dangerous Fieldwork. Qualitative Research Methods Series, Vol. 34. Thousand Oaks, CA: Sage Publications. Levi-Strauss, Claude. 1961. Tristes tropiques. New York: Criterion Books. Lewin, ElIen, and William Leap. I996a. Introduction. In Out in the Field,'· Reflections of Leshianand Gay Anthropologists. ElIen Lewin and William Leap, eds. Pp. i-28. Urbana: University of Illinois Press.

Nader, Laura. 1986. From Anguish to Exultation. In Women in the Field: Anthropological Experiences, 2d ed. Peggy Golde, ed. Pp. 97-116. Berkeley: University nf California Press. Naral!, RaouJ. 1962. Data Quality Control: A New Research Technique. New York: Free Press. Naral!, Raout. 1970. Data Quality Control in Cross-Cultural Surveys. In A Handbook of Method in Cultural Anthropology. Raoul Narolt and Ronald Cohen, eds. Pp. 927-945. Garden City, NY: Natural History Press. Nash, June. 1976. Ethnology in a Revolutionary Setting. In Ethics and Anthropology: Dilemmas in Fieldwork. Michael A. Rynkiewich and James P. Spradley, eds. Pp. 141l-166. New York: Johu Wiley. Newton, Esther. 1993. My Best Informant's Dress: The Erotic Equation in Fieldwork. Cultural Anthrapology 8(1):3-23. Newton, Esther. 1996. My Best Informant's Dress: The Erotic Equation in Fieldwork. In Out in the Field: Reflections ofLesbian and Gay Anthropologists. ElIen Lewin and William Leap, eds. Pp. 212-235. Urbana: University of Illinois Press. Nichter, Mimi, and Mark Nichter. 1987. A Tale of Simeon: Reflections on Raising a Child while Conducting Fieldwork in Rural South India. In Children in the Field: Anthropological Experiences. Joan Cassetl, ed. Pp. 65-90. Philadelphia: Temple University Press. Paul, Benjamin. 1953. Interview Techniques and Field Relationships. In Anthropology Today. A. L. Kroeber, ed. Pp. 430-451. Chicago: University of Chicago Press. PeUo, Pertti 1. 1970. Anthropological Research: The Structure of1nquiry. New York: Harper and Row. PeUo, Pertti J., and Gretel H. PeUo. 1978. Anthropological Research: The Structure of Inquiry, 2d ed. New York: Harper and Row. Picchi, Debra S. 1992. Lessons in Introductory Anthropology from the Bakairi Indians. In The Naked Anthropologist: Tales from Around the World. Philip R. DeVita, cd. Pp. 144-155. Belmon~ CA: Wadsworth Publishing.


Lewin, Ellen, and William Leap, eds. 1996b. Out in the Field: Reflections ofLesbian and Gay Anthropologists. Urbana: University of Illinois Press. ' . ~ Lewis, Oscar. 1951. Life in a Mexican Village: Tepoztltin Restudied. Urbana: University of Illinois Press. . Lutz, Katherine. 1988. Unnatural Emotions: Everyday Sentiments in a Micronesian Atoll. Their Challenges to Western Theory. Chicago: University of Chicago Press. Malinowski, Bronislaw. 1922 [1961]. Argonauts ofthe Western Pacific. New York: Dutton. Malinowski, Bronsilaw. 1935 [1978]. Coral Gardens and Their Magic. New York: Dover. Malinowski, Bronislaw. 1967. A Diary in the Strict Sense ofthe Word. New York: Harcourt, Brace and World. Mead, Margaret. 1923. Coming ofAge in Samoa. New York: Wittiam MOIT()\V'~'" :':'';;';~;-:'' Michalski Turner, Diane. 1987. What Happened When My Daughter Became Fijian. In Children andAnthropological Research. Barbara Butler and Diane Michalski Turner, eds. Pp. 97-114. New York: Plenum Press. Moreno, Eva. 1995. Rape in the Field: Reflections from a Survivor. In Taboo: Sex, Identity, and Erotic Subjectivity in Anthropological Fieldwork. Don Kulick and Margaret Wilson, eds. Pp. 219-250. London: Routledge.


Pratt, Mary Louise. 1986. Fieldwork in Common Places. In Writing Culture: The Poetics and
Politics afEthnography. James Clifford and George Marcus, eds. Pp. 27-50. Berkeley: University of California Press.

Participant Observation
298 DEWALT I DEWALT ":( Punch, Maurice. 1994. Politics and Ethics in Qualitative Research. In Handbook of Qu ilative Research. Norman Denzin and Yvonna Lincoln, eds. Pp. 83-98. Thousand 0 CA: Sage Publications.


\Tcd;~I<E::rgence of Narrative Ethnography. Journal of Anthropological Research 47(1).

rbara. 1991. From Participant Observation to the Observation of participation:

Quandt, Sara, MaraZ. Vitolins, Kathleen M. DeWalt, and Gun RODs. 1997. Meal PaUe of Older Adults in Rural Communities: Life Course Analysis and Implications for Under. nutrition. Journal of Applied Gerontology J6(2): 152-171. Rabinow, Paul. 1977. Reflections on Fieldwork in Morocco. Berkeley: California Press.
Redfield, Robert. 1930. Tepoztldn. Chicago: University of Chicago Press.

'. 69-94. hAG 'd t <' k' Elizabeth. 1984. Participant Observation. In Ethnographic R~searc : e K Ton In, I C d R F Ellen ed Pp. 216-223. London: AcademIC Press.


Genera ,on uc. . . " . ., . . b U Colin. 1986. Sex and Gender: The Role of SubJecttvlty In Field Research. In Se~j. ,;i:,Turn u a~d Gender in Cross· Cultural Research. T. L. Whitehead and M. E. Conaway, e s. s a I. ' P Pp. 17-27. -Urbana: University of I I1000S ress. . . . . Maanen, John. 1988. Tales ofthe Field: On Writing Ethnography. Chicago. UOIverslty




Riemer, Jeffrey W. 1977. Varieties of Opportunistic Research. Urban LIfe and Culture
5:467-478. Rohner, Ronald, Billie R. DeWalt, and Robert C. Ness. 1973. Ethnographer Bias in CrossCultural Research: An Empirical Study. Behavior Science Notes 8:275-317. Raper, J. Mantgamery, John Frechiane, and Billie R. DeWalt. 1996. Indigenous People and Development in Latin America: A Literature Survey and Recommendations. Latin American Monograph and Document Series, No. 12. World Bank and Center for Latin American Studies, Pittsburgh: University of Pittsburgh. Rynkiewich, Micbael A., and lames P. Spradley. 1976. Ethics and Anthropology: Dilemmas in Fieldwork. New York: John Witey. Sanjek, Roger. 1990a. The Secret Life of Fieldnotes. In Fieldnotes: The Makings ofAnthropology. Roger Sanjek, ed. Pp. 187-270. \thaca: Comell University Press. Sanjek, Roger. 1990b. A Vocabulary for Fieldnotes. In Fieldnotes: The Makings of Anthropology. Roger Sanjek, ed. Pp. 92-121. Ithaca: Cornell University Press. Sanjek, Roger, ed. 1990c. Fieldnotes: The Makings of Anthropology. \thaca: Comell University Press. Scheper-Hughes, Nancy. 1987. A Child's Diary in the Strict Sense of the Term: Managing Culture Shocked Children in the Field. In Children in the Field. loan Cassell, ed. pp. 217-236. Philadelphia: Temple University Press. Seligman, Brenda Z. 1951. Notes and Queries on Anthropology, 6th ed. London: Routledge . & Kegan Paul. Spradley. James P. 1980. Participant Observation. New York: Halt, Rinehart and Winstan. Stack, CaroL 1996. Doing Research in the Flats. In In the Field: Readings on the Field Research Experience, 2d ed. Carolyn D. Smith and William Kornblum, eds. Pp. 21-25. Westport, CT: Praeger. Sterk, Claire. 1996. Prostitution, Drug Use and AIDS. In In the Field: Readings on the Field Research Experience. 2d ed. Carolyn D. Smith and William Kornblum, eds. Pp. 87-95. Weslpor\, CT: Praeger. Stocking, George. 1983. The Ethnographer's Magic: Fieldwork in BritishAnthropology from Tylor to Malinowski. In Observers Observed: Essays on Ethnographic Fieldwork. George Stocking. ed. Pp. 70-120. Madison: University of Wisconsin Press.

Vi1~e~:~~~h~~f~~~~~II~::r~~e~1:r~;0::: ~:~:~,~~~~o;~~~:~~~I:~~~~~-~~~:O:I:~~
Robert Kemper eds. Pp. 45-64. New Yorl<: AcademiC Press. 88 Gender Issues in Field Research. Thousand Oal<s, CA: Sage Warren, Caro1 A.B , 19 . Publications. . h rt Annette B. 1988. The TrobriandersofPapua New Guinea. NewYorl<: Holt, Rine a and Winston. Whitehead, T. L., and M. E. Conaway, eds. 1986. Self. Sex and Gender in Crass-Cultural

of Chicago Press.




Research. Urbana: University of Ulinois Press. Wh te WiHiam Foote. 1996. Doing Research in Cornerville. In ~n ~he Field: Readings;n ;he'Field Research Experience, 2d ed. Carolyn D. Smith and Wtlham Kornblum, eds. p.
73-85. Westport, CT: Praeger. Whyte, WiUiam Foote, and Katherine King Whyte. 1984. Learningfrom the Field: A Guide from Experience. Beverly Hills, CA: Sage publicatiOns. . Willson Margaret. 1995. Afterward: Perspective and Difference: Sexualization, the F~el~ and ;he Ethnographer. In Taboo: Sex. Identity. and Erotic Subjectivity in :nt~;pOlt~'Ca Fieldwork. Don Kulick and Margaret Willson, eds. Pp. 251-275. Lon on. out e ge.

l l

l l






Direct Systematic Observation of Behavior


., havior-what people do-is a fundamental dimension of cnltural diversity. ether we think of behavior as customs, habits, practices, Hfeways, or activity " ems, an important task of all ethnographic inquiry is to illuminate the patterns otaction and interaction of the people we study. The central thesis of this chapter ~ihat direct systematic observation is our best approach to developing trustworthy .ICcounts of people's behavior. As such, it deserves a more prominent place in the ,dlUtographic toolkit than most anthropologists seem to appreciate. :F Systematic observation is structured by explicit rules about who we observe, .when and where we observe them, what we observe, and how we record our '(observations. These entail selection among options, each with associated tradeoffs. Here we guide the newcomer to systematic observation through the maze of choices ,from project conceptualization to the sampling and recording of behavior patterns.


Why Do Direct Systematic Observation?
';,We distinguish three broad methods of ethnographic field research: interview, , participant observation, and direct systematic observation. Interview research, an 'eclectic category that includes a large and diverse array of specific methods, relies entirely on research subjects as sources of ethnographic knowledge. Participant ~bservation places the ethnographer at the scene, where a combination of direct



Direct Systematic Observation of Behavior


observation and interview provide the evidence from which rich ethnographic accounts may be constructed. By direct systematic observation, we mean those

ethnographic methods that-in contrast to interviews-rely primarily on the: researcher's first-hand observations and that-in contrast to participant observation -are seriously attentive to problems of sampling and measurement. Ratherthan study behavior for its own sake, anthropologists commonly undertake behavioraldescription as one component in more holistic ethnographic investigations. that include people's attitudes, discourse, and organization. In these holistic efforts, interviewing and participant observation-the traditional mainstays of ethnography -are often the methods used for describing behavior as well as for describing what people think, how they are organized, and so on. We endorse multiple methods and recognize the complementarity between them. We argue here, however, in the strongest terms that interviews and participant observation are, by themselves, ' inadequate to the task of constructing trustworthy accounts of activity patterns. There is an irreducible need for other, more rigorous observational methods.

Interviews Versus Direct Systematic Observation
The methodological distinction between interviews and direct systematic observation is related (but not identical) to the emic/etic distinction in the sense that to obtain emic data-which requires an interpretation by a research subject-it is usually

necessary to conduct some sort of interview, whereas the data of direct observation
.are usually classified as etic. What confounds this neat dichotomy is that interview data may also be classified as etic to the degree that our research subjects are regarded as reporters of events they have witnessed first hand (Harris 1990:53). The general distinction between interview and direct observation, therefore, is most properly that between interview methods that emphasize meaning, interpretation, and subjective experience as against research methods (including some interviews) that emphasize accurate reporting of observed scenes and activities-ideally, without interpretation. This last caveat is fundamental and raises methodological problems that are too often swept under the rug by researchers unwilling to face the painful implications. .. . ... The fact is that hum~ns (including both trained field workers and untrained research subjects) are surprisingly incapable of accurately describing scenes they have observed with their own eyes (and ears and other senses). Abundant evidence shows that when research subjects are asked to report on their own behavior, and these reports are compared to researchers' records of the subjects' behavior .based· on direct observation, the research subjects' accounts of their own behavfor'ak'i';" substantially "wrong"-that is, they show errors of from 50%-80% when cOlllparGcl '.,' to the observational data (Bemard et a1. 1986:388; Engle and Lumpkin 1991). This means, to put it bluntly, that anthropologists )'Yho rely uncritically on their research

b' cts £or descriptions of behavior are more likely to be wrong than right. Certainly, this is a methodological issue worthy of the most senous attention. . ':'~ Where does the problem lie? It is most serious for re.se~rchers who tr~at theIr 'j' fonnants' descriptions of the past as accurate descnpttons of behavJOr. For "Ill pie for years the University of California asked faculty to report annually on exam, . . b d' hl'gs t "" their use of time for the previous year, the mformatlon to e use 1n e pm e 8 ublic poliCY on such issues as how much time professors were expected to sp~nd ~ teaching and supervising students as opposed to research, conferences, and outSide :nsulting. Routinely, professors reported their average work week t~ be more ~han 60 hours with abundant time devoted to teaching. The question skeptlcs are entitled to ask i;, "Are these self-reports to be believed?" Or, more gener~cally, are ou~ research subjects the "videotape-Iike creatures with near-perfect retneval systems that our research methods sometimes assume (D' Andrade 1974;124)? The methodological issue is clarified by the distinction between short- and longterm memory (D' Andrade 1995:42-44). Roughly speaking, direc~ly obse,:"e~ events are first stored in short-term memory, which appears to be relatively obJectl~e ~nd accurate about recording what actually transpired, but which has a very hmlte.d storage capacity. Quickly, therefore, knowledge stored in short:term ~emory IS transferred elsewhere in the brain-to long-term memory. HoW qUickly thiS happens is debated, but it is generally thought to take a matter of seconds: . Long-term memory has the advantage over short-term memory IU ~hatlts storage capacity is far greater, but it has a serious defect fro~ the ~tandpolUt of resea~ch methods: The correlation between records made Immediately on observatIOn (depending on short-term memory) and records based on later recall (long-term memory) is extremely low-generally around r = .25 (~hweder 19.8~). What happens to information in long~term memory? The answer IS not surpnsmg to cultural anthropologists: The direct observations that were store~ in short-te~ memory are systematically distorted to conform to cultural expectatIOns. In particular:


. .

l l l

l l

What people remember as going together are the kinds of be~avior they judge to be similar. Humans show a systematic distortion in their memories. They falsely recall "what goes with what" based on "what is like what.:' This e~fect has b~en demonstrated across a wide range of kinds of materials, not Just behav\or frequenCies. . . . Overall, these results throw doubt on a broad class of retrospectively based research data. .. . The results [however) are much happier for the. study ?f the organizati~n ~f culture . ... Cognitively shared salient features are an mteresttng part of a society s


culture. (D' Andrade \995:84) We can expect, then, that when anyone is asked to report on his or he.r o~ behavior in the past, or on the observed behavior of others, systemalic dlstottions-selectiverememberingand forgetting-will shape the memory so stron.gly as to make it largely irrelevant as a description of the detailed behavior in questIOn.

l l




Direct Systematic Observation of Behavior


, ,


The silver lining of this cloud is that, if in fact people's cultural models of behaviot more or less accurately reflect average behavior in their communities, then their: long-term memories will not be far off as reports of what people usually dO (Romney et aI. 1986). What the reports of UC professors tell us, therefore, is most likely som: combina~i?nof their implicit, shared vi~ws of how professors typically behave, WIth the addItion of some more-or-Iess delIberate distortion based on their perceived self-interest: After all, professors know qUite well that the activity questionnaires play a role in the political process of university funding. The compiled· data almost certainly exaggerate the length of the faculty work week and likely also· the proportions of time spent in teaching. They are, of course, entirely useless as records of such potentially interesting behaviors as socializing in the hallways or staring out the window. Anthropologists and other behavioral scientists should be more distressed by this problem than they appear to be. Although there are certainly times during ethnographic research when long-term memory must be relied on for behavioral descriptions, field workers should refrain from doing so casually.Although direct observation is sometimes difficult or awkward, infonnant recall--or even the ; researcher'swrite~up offieldnotes at day's end-is no SUbstitute. Recall data, based on long~term memory, are about cultural pattern (the informant's or the;
researcher's), not about observed behavior.

tbropological modes share a view of which activities ate important and worth :ention and which can safely be ignored.

Comparison of Anthropological ActiVity Descriptions with Actual Time Allocation ¥. of Behavior Descriptions
...c, Activity
Textbooks (0 .. 2.50) Ethnographi« (0 .. 250)
An/el'. Amh. (0 .. 2.50) Overall
(0 .. 7.50)

Global Adult Time Allocation

, :: politicaVreligious .; Ibod production
eating, rest, sleep

c:omrnercial car.
making/fixing things

3&.4 22.0 \4.0 12.0 9.2 4.4

36.4 \&.4 1&.4 9.6 11.2 6.0

34.0 26.4 12.& 10.0 &.& &.0

36.3 22.3

15.\ 10.5
9.7 6.1

2.& 11.0 62.0 &.& 12.0 3.4


Participant Observation Versus Direct Systematic Observation
The problem of erroneous descriptions arising from cultural distortion also weakens participant observation as a method of behavioral research. One problem is that anthropologists don't naturally or automatically gather representative data on behavior. On the contrary, like all people, they are driven by their interests and

L 1 L

cultural models to privilege certain behaviors over others as "relevant" or Uinteresting." To demonstrate this, Sackett (1996) examined the spectrum of activities described in a number of anthropological accounts drawn from textbooks eth' nographies, and five years of articles from the American Anthropologist. H: took a random sample of descriptions of behavior found in these publications, which resulted in 750 activity descriptions divided equally from these three kinds of " anthropological writings. He then compared these activity descriptions with the actual frequency of those activities in people's behavior as reported in time allocation studies from around the world (Table I). Two features of Table I are especially noteworthy. First, the relative conspicuousness of activities reported is quite similar across the three modes of anthropological writing (for example, the frequency of discussion of political and religious behaviors hovers around 36%). This suggests that writers in the various

Second, there is no consistent relationship between the conspicuousness of a behavior in our professional discourse and the amount of time people around the world actually spend engaged in that behavior. In fact, political and religious practice-mainly, public ceremonials like rituals and raIIies-commands the greatest proportion ofanthropological attention (36%) yet actually occupies the least amount of people's time «3%). Self-absorbed activities like eating, sleeping, and relaxing occupy by far the greatest proportion of people's lives (over 60% of their time), yet constitute only 15% of published descriptions. Sackett's research also reveals a strong gender bias in activity descriptions (Table 2). Far more anthropological attention is being paid, at least in professional writings, to activities cross-culturally associated with men rather than women. Men's activities are eight times more likely to be described in these writings than are women's activities, although the actual frequencies with which these activities are performed by the men and women of the world are not greatly different (12.3% for predominantly women's activities, 18.6% for predominantly men's). Tables 1 and 2 provide evidence of strong shared biases in how anthropologists describe behavior. Our point here does not concern whether or not these biases areappropriate, only that they exist and carry important implications for behavioral measurement. As long as participant observation alone stands as the main method of behavioral description, we can expect these biases to shape the way observations are recorded during fieldwork and later reported in scholarly publications. The same kinds of biases that shape 1000g-term memory also direct the field worker's attention· and,colleagues' and editors' judgments regarding what is deserving of description, analysis, and publication. The methodological solution to this problem is to take the researcher's attentiott off auto-pilot, so to speak, by introducing rigorous procedures for sampling and



Direct Systematic Observation of Behavior



Comparison of Anthropological Attention to Women's and Men's Activities
% of Behavior Descriptions

division of labor

(n" 250)

(n ... 250)

Amer. AlIIh. (n .. 250)

(n '" 150)

Global adult time allocatiOlll"

,} texts in which research is done, but tack of explicit attention to definitions ~es it impossible to distinguish real cultural diversity from differences in ethIll. graphers' arbitrary and idiosyncratic conventions. Although the solutions to these ~ f I ., .difficulties come from scientific methodology, the broad goal, 0 va ue to SCientists -liid humanists alike, is verisimilitude-a recognizable similarity to the actual lives of the people we study.


Activities perfonned predominantly by women Activities showing little cross-cultural gender bias Activities perfonned predominantly by men
SOURCE: Sackett 1996.






The Chal1enge of Behavior Measurement

A commitment to include behavioral measurement in fieldwork immediately brings
29.6 34.4 32.8 32.3 69.\ 63.9 56.8 60.8 60.4 18.6

recording behavior-in effect requmng field workers to observe and report behaviors they might otherwise neglect. Such neglect is virtually inevitable unless' addressed. Overtime, field workers become so familiar with their subjects' behavior that they begin to stop noticing quotidian commonplaces. Field workers fall into comfortable field routines that make some scenes and locations much more likely to be observed than others; and they find it easier to remember cases that confirm. their own understandings of what is going on and to forget the negative cases that defy their understandings. Hence, there is much in anthropological method-indeed, in human nature-to make it highly improbable that routine descriptions of behavior based on participant observation research will validly describe actual behavior. For example, we can't make sense of an ethnographer's report that child care is predominantly women's work unless we are told precisely what child care is: Does it include producing the food the child eats, or having socially recognized responsibility for its well-being, or is it limited to directly handling the child? And what is the quantitative basis for - - - - -'the- report: Is it the relative. frequency of care, total time 'spent; the amounFof' physical exertion, or some combination of all of these? The current state of anthropological fieldwork is such that ethnographers are generally inconsistent in the ways they describe activities, differing suhstantially in their definitions of such basic categories as labor, housework, and child care, It is' now a commonplace in activity studies that how much "work" people do depen~s sensitively on just which activities we consider work. Even a small shift in defim,.,· tions can lead to radically different conclusions (Johnson 1975). Some differences between ethnographies are inevitable given the difference in social and cultural

to a dilemma: Measuring the ongoing stream of behavior in its natural detail, complexity, and context is so daunting as to be a practical impossibility. Behavior ':-somebody doing something-refers in phenomenological terms to ohservable changes in location, posture, expression, and vocalization. Simply describing a 'subject's location requires at least six pieces of information (latitude, longitude, \0 altitude, compass azimuth, orientation to the horizon, and time). Posture is far more tJ'~complicated: Given the human body's 206 bones and their articulation at joints, we would have to recreate 218 joint angles precisely in order to reproduce a subject's posture at any given instant (Alexander 1992). Add to this a minimum of 58 facial muscle groups involved in expression, and we would have to record a total of 282 pieces of information (6 location measures, plus 218 degrees ofpostural freedom, plus 58 facial muscle groups) simply to describe someone's behavior at a moment

l l l


in time.
And these 282 pieces of information allow us to dip only once into the stream of behavior. To animate our description with the verisimilitude of a motion picture (which itself is only a sampling of the stream of behavior), we would have to update all these variables 24 times each second. At this frame speed, we would have to make 406,080 measurements per minute of observation, amounting to over 24 million measurements to describe just one hour of spontaneous behavior. And we haven't even addressed the problem of describing a subject's speech or the setting in which the behavior is happening. Thi~ dilemma is real for every behavioral researcher and has long attracted serious methodological discussion (for example, Chapple and Arensberg 1940; Harris 1964). A large part of the solution is beyond the scope of this Chapter: In ~e process of research design, clarifying the purposes of the research allows the 'eld "" Worker to select and distill from all the possible ways of describing behavior those 2#0" that most efficiently answer the key questions of the research project (Martin and ;jf~; Bateson 1993; Bemard 1994). Hence, flexing the shoulder, extending the elbow, ':~i~ pr~nating the hand, and flexing the fingers around a piece of fruit become t; . .. ~. grasping an orange," and that, in turn, becomes part of an act of food procurement, ',J,yi food preparation, gift exchange, theft, or eating <as the case may be). Any of these












Direct Systematic Observation of Behavior





higher-order categories of activity mayor may not be of interest to any given research project. Once we have a clear sense of the subject and goals of the research, we need to address several questions that have profound implications for how we conduct our behavioral srudy: Is observation an appropriate technique? How does systematic . observation integrate with other activities? What information do we need to collect?

Is Observation Appropriate?

'L 'L

Whiting and Whiting (1973) have argued that behavior observation is more expensive in field and analysis time than other ethnographic activities and should only be used when other techniques are less effective. To justify this investment, first and foremost the study must be about behavior. That is, activity descriptions must be able to help answer the key questions posed by the research proposal. When those questions ask what activities are like, who performs them, and in what contex~s .~ey oc~ur,. o~servation is .appropriate. Questions about the consequences of

actiVItIes for mdlvldual well-bemg, alteration of the environment, genetic fitness,
and so on, can also be answered by observations carefully supported by other eth. nographic evidence. . Practical concerns limit the appropriateness of behavioral observations. Will research subjects tolerate the presence of the observer? Will they act in the observer's presence as they would if the observer were not there? Are some kinds of observation less acceptable to subjects than others? Can the behaviors of interest be observed without violating subjects' sense of decency and privacy? Is the observer likely to wi~ess illegal or stigmatizing behaviors and, if so, are the conseque.nces of reportmg or n?t reporting them morally and ethically acceptable? Expenenced field workers WIll agree, however, that these practical concerns arise in an~ participa~t observ~tion research. Perhaps the methodological rigor of system,at,c observatIOn h.,ghhghts the problems, for the option of quietly putting away one s notebook, tummg away, or leavmg the scene is available to the participant obse,:er mo:e so t.han.to the systematic observer. And, given the serious problems

be present at particular times and places, when we might prefer to pursue interesting events elsewhere in the community-to interview an unexpectedly cooperative informant or to escape the community in private contemplation. Furthermore, . systematic observation is sometimes a less comfortable role for the researcher, owing to its requirements of detachment and objectivity. It creates a distance, in contrast to the essentially friendly, murually attentive, and empathetic relationship fostered by interviews and participant observation. But it's best to recognize and embrace the synergy between systematic observation and both interviewing and participant observation. On the one hand, systematic observation keeps participant observation honest by explicitly confronting the implicit biases of long-term memory, variations in contemporary research fashions, and inconsistent definitions. On the other hand, it addresses the potential biases and ambiguities of an interview, helping identify conscious and unconscious deceptions and distortions. Reciprocally, both interview and participant observation increase our confidence in the validity of the results of systematic observation by helping us understand the consequences of behavior for subjects' well-being, motivations, and emic conceptions of activity.

What Information Do We Need to Collect?
Actors, Actions, and Settings
We find it helpful to distinguish three broad categories of variables common in activity srudies: actors (including both the focal subject(s) and the social others with whom she or he interacts), actions (the behaviors we want to study, whether specific


acts and activities, the content of speech, or the consequences of behavior) and
sellings (including the location of action, details of the physical space in which the

assocIated With usmg mformant recall to answer behavioral questions,


must be

prepared to make the investment in direct systematic observation where it is the only appropriate method for answering key research questions.

actors act and interact, and the "props" which they manipulate and use, such as furniture, implements, foods, etc.). For ethnographic research, the most common and most useful focus is on actors, a sample of whom are selected for observation such that their actions and the settings in which they actare the content of the observations. More rarely, it is an action (for example, political speeches) or a selling (for example, a public speaking space) that is the focus.
Level of Behavior Measurement Different research questions require different levels of behavior measurement. Since higher levels of measurement usually cost more (in various ways), we should be clear about the minimal data quality needed to answer a question before settling on a suitable method (Martin and Bateson 1993). In general, actionS can be

How Does Systematic Observation Integrate with Other Ethnographic Activities?
Participant observation, interview, and systematic observation compete for our Iimite.d field tim.. an~ att..ntion. The inflexible scheduling often required of systemallc observatIOn h,ghhghts the conflict: Most sampling strategies require us to

measured as nominal variables or as quantitative variables.



Direct Systematic Observation of Behavior I. Nominal variables. Some issues can be resolved with nominal-level data-that is, as presence or absence of particular behaviors. We might, for example, ask whether aparticular behavior is in the repertoire of the population we study (for example, does anyone in the community fish, or make pottery, or observe Hindu
food restrictions?). Alternatively, we may want to.~now whether a certain ~eha~ior occurred during a given time interval (for .mple, does anyone fish


Temporal Resolution The focus of behavioral observation may be on short episodes (events) or long "ones (states), each with different methodological implications. An ~xample of ~n ,5event is a schoolyard fight: A glare is met by a challenge, promptmg a shove In iresponse, erupting in a brief skinnish that is quickly stopped by schoolyard tmonitors; within seconds the event is over. The interesting feature of behavioral events tends to be their frequency: How / often do schoolyard fights occur and are they frequent in relation to other , behaviors on the schoolyard or elsewhere? Since events last such a short time, they require a deliberate sampling strategy, such as continuous monitoring for 'periods of time. .. . By contrast, the daily commute to work has the quahlIes of a behavloral state: ; It is extended in duration, often rather repetitive, and possibly monotonous. The salient features of states are usually their total duration over some period of time ·and the duration of each occurrence. For example, the duration of the daily commute between home and workplace, and the total time spent commuting as a part of the working day are much more revealing about how we spend our lives than the fact that we do it twice a day. Because states aTe of longer duration than events, they are easier to encounter during research and can be observed either by continuous monitoring, or by sampling the behavior stream at discrete moments, a strategy called instantaneous


wmter, or make pottery on Sunday?). Such questiQns


relatively easy to

answer and may even be reliably addressed through infonnant recall. The chief disadvantage of nominal data is that they are poor indicators of

how much of a behavior people do or of a behavior's relative importance in their lives. For example, two subjects could have fished on the sam'e day, but the actual time they spent fishing and the number offish they caught CQuld have differed greatly. 2. Quantitative variables. Some questions require data Qn the actual amount Qf be~aviQr. Measuringt~e actual amount of behavior performed increases our cap. BClly to evaluate the Importance of the behavior. Behavior researchers commonly distinguish four quantitative measures of behavior: F~eque.ncy .refers to the number of Occurrences of a behavior during a glven time Interval, most conveniently expressed,~s a rate (instances per unit time). Duration is the length of time for which a single occurrence ofa behavior lasts, measured in time units. Total duration is the total amount of time spent performing the activity during a particular interval of time, numerically equivalent to the frequency of the behavior multiplied by the average duration. It is the measure most commonly used in time allocation studies, expressed either in time units (for example, hours per day) or as a percentage of time. Intensity refers to the pace at which the behavior is performed. It may be measured by a local rate such as elemental acts per unit of time (ax swings per minute) or rate of production (meters of cloth woven per day), or by a more generally comparative measure like energy expenditure (Montoye et al. 1996).


Descriptive Resolution We have discussed the overwhelming amount of detail possible in behavioral observation. A simple activity like grasping an orange could potentially require thousands of points of measurement. The more detailed the resolution of behavioral study, the more costly the research in time, technology, and, probably, the good will ofthe research subjects. Detailed description of the discrete acts that go into making up the generalized sequences we call activities, therefore, must be justified by the theoretical and practical goals of the research project. In our experience, even those observers who collect data on specific acts tend, in the long run, to abstract and generalize these into descriptions of activities anyway. For example, the act of "walking down the trail While carrying a rifle" is detailed and specific, but the most useful fonn of description usually ends up being "hunting deer." Why not save a step and record the activity "hunting deer" in the first place? If we collect too much detail in the field, we waste field time and then require additional time later-probably after the field site is long behind Us and far away-trying to infer meaningful activity from data sheets full of 'incremental acts.

l l


Behavior Flow . In addi~ion to m~asuring the presence or amount of a behavior, we might also be mterested m behavlor flow. The flow of behavior is the sequence of behaviors from one activity to another. We may be interested in the flow of behavior in the short tenn, such as the component actions in weaving a basket, or in longer tenn sequences like the daily round or annual cycle. . Measuring the amount and flow of activities increases the power of our data but , It gener~lly entails. ~ore fieldwo.rk and gr~ater risk of influencin!l tjle pehavior.of.;.t. ,our subjects (reactivity). Measurmg behavlOr also places greater constraints on the:" kinds of sampling and recording methods we use. Below, we examine further ihe '!' tradeoffs among alternative methods.





" the ideal is to make observations . . the investigator's language facility. Paohsso and 5ackett 1988). shoppmg. the temptation is great to reduce the scope of the research by selecting specific categories of individuals for focus. sex. While attractive as a . to work in distant cities or go on trek. The convention in ethnographic research is to study a community for at least a full year. "dawnto-dusk" studies are often a practical necessity. Field workers must impose a distance rule . research focuses on individuals. When systematic observations are limited to daytime. In our ' experience. some other technique such as participant observation or interview should be used to establish in a general way what goes on at night. those who study only children wish they' had included adults. such as the day-range of common activities. In a small community. or ': ~siting relatives. urpose of the trip was stated to be. but they are problematic because they systematically overestimate time use in some activities and underestimate it in others (Gross 1984. as is often done in time allocation studies of neighborhoods or villages.e. The size of the targeted social universe may vary greatly. but such research is rare because it is intrusive and :making night visits is often culturally prohibited and dangerous. Kiiski. activities. Sampling for .~takes to travel to a site and back in a single day (Carlstein 1982). Step 2: Choosing the Units to Observe Having chosen a sample for observation. Scaglion 1986). where seasonal variations can be demonstrated not to be very important (perhaps in some cities or with respect to key activities like infant care). Researchers who study only the women in a community . " t ' d'Icate strategy in such cases is to define a category 0 f activity o. . social class. we urge researchers to resist this temptation. It is possible to add information to this ~ode by spelling out w~at the . primarily. L L i . the research budget-it is still the case that the best research population is one that is most appropriate for answering the key research questions identified in the research proposal. and Liikkanan 1981) and multinational (Szalai 1972) communities. Step I: Establishing the Limits of the Study We first need to establish the sampling universe. Targeting People L L L Most commonly. most behavior studies concern some combination of three variables: actors. regret the absence of information on men. This is true when subjects leave ." The only realistic option IS. it may be possible to include all Individuals for observation. from a single Yanomami family (Lhermillier and Lhermillier 1983) to entire national (Niemi. the envelope of people. such as workmg for wages. samples of individuals may be drawn for observation. limited by the time systematic behavioral observation requires several steps: 'that subjects are temporarily resident elsewhere (see.1n observations. . happens that individuals in that sample leave the field site and are beyond the limits' to which the researcher can travel to observe them. we must decide what to target for observation. A common It . Arguments for shorter time frames can be made in some cases. cod ed"A way. influence the selection of a population for study-such as access to the population. for exampl. But these data are not comparable to the data of direct observation and should be treated separately in later analyses. and so on. The best sample in such caseS is always a random sample. L L Social Boundaries Although it is typical in ethnographic research for many practical issues to . and settings. It is far more satisfYing to develop sampling methods that include the whole range of kinds of people in ' research populations. those who study only the Hindus of some village wish they had data on neighboring Muslims for comparison. it leads to no end of lamentation once the research is over. it often . because the most important "context" for individuals' activities is the activities of the others around them. we focus on one of these and observe how the others vary in relation to it. throughout the 24-hour cycle. way of conserving resources.-m~ntioned "god's-eye view. to observe the variations of the seasons and the full annual cycle: This rule applies as well to behavioral observation.L L L L 3/2 JOHNSON / SACKElT Direct Systematic Observation of Behavior 313 Sampling All field workers interested in fair and accurate descriptions of what people do have wished at one time or another to be a fly on the wall or an omnipresent (and invi~i~le) obse. a compromise ID the fonn of sampling. Usually. " In large-scale studies (and even some smaller ones). who should be selected for their representativeness.:er-gifted with the oft. of course. using criteria like age. Temporal Boundaries The major issues of temporal sampling concern how long a period of study is required to get a good sample of ordinary behavior and how to handle the problem of nighttime activity. Regarding the daily "observation window. places and times from which we will select entities for observation. Geographical Boundaries Once a research population has been identified and a sample drawn.. or ethnicity. In a larger community. As we have said.i . Therefore.

FlInn 988).._ _. systematic observation (Whiting and Whiting 1973 J. For example. Bemard 1994). we record either sequentially or synchronously the behavlOrs of a number of associating individuals (Altmann 1974. and how to sample time within the session (Gross 1984. to the exclusion of others.. we select one subject and " Behavior patterns are usually structured by the time of day. random· interval instantaneous sampling. For example. or setting. Hames 1979. t: Step 3: Scheduling Observations ~havioral research to "A strategy for scheduling observations should be developed at the beginning of ensure that the study period is sampled evenly. Dunbar 1976).1 -:'may be practical to scan an entire community sequen~ial1y by walking a cir~uit .think we see in the data.as a description of the full range of behavior patterns. etc. to eliminate weather as a confounding . shelters (900 minutes).' taneous walking speed in 15 communities ranging in size from 365 people (Psychro. Dufour 1984.. arrow poison (300 minutes). . Scheduling Observing Sessions Targeting Activities or Settings Occasionally. Hill et al. Sampling Strategies How long do we keep the target individual or group under observation? The answer depends on which of four main sampling strategies we choose: continuous monitoring.000 (Brooklyn. New York). In the downtown of each com. house to house (for example. By the same token. researchers generally have the choice of concentrating on a single individualljollows) or tracking the activities of members of a group (scans). In group scans. In/oea/. if the schedule of observations is such that individual X is always observed on Tuesday. and continuons observation (Gross 1984). psychologists Bomstein and Bomstein (1976) examined spon. the research questions call for targeting certain activities. how long the session should last.600. its unit labor .. in contemporary society if behavior .such cases. rather than i distorting it. as the Bomsteins did in the~r research. Under some circumstances (not holistic ethnographic research).'be acceptable to limit observations to a single time or circumstance: For example. the absence ofweekendbehaviors would "Strongly distort our data. and season. It may be tempting to build a~ oPP?~unity samp~e in . factor in their study. like houses. ~ Borenstein and Borenstein (1976) limited their observations of walking speed in all :'i cases to sunny days of moderate temperature. By determining who made each type of artifact. observations were made only on weekdays. for groups and individuals). They found a remarkably strong correlation (r = 0. In the 1970s. l l l l l Continuous Monitoring For the fUllest account we may choose to target a single individual and follow her or him throughout the day (for example. Lee (1979) observed the time his !Kung (Botswana) informants took to make ostrich eggshell canteens (360 minutes). For example. 1985) or for several days in succession (Szalai 1972). cost. activity. best solution is to schedule observing sessions at random. ThiS should be avoided because it will compromise the reliability of the sample as being representative of the population and thwi threaten the study's overall validity.314 JOHNSON / SACKEIT Direct Systematic Observation of Behavior 315 drawn using a lottery or table of random numbers. Continuous monitoring is Bemard's (1994) term. schoolyards.'lJlonth.individua//ollows. Baksh 1989-90). ~~ing people with whom the researcher is already familiar (and. weather patterns.91) between the size of the community and walking speed: The larger the community. even though X was chosen at random for inclusion in the study. women. 45 minutes). avoldmg those who are unfamiliar or with whom the researcher doesn't get along). and the number of such objects maintained by a family. 64 minutes. the observations on X. A common strategy in time allocation studies is household scan sampling.' munity they marked a 50-foot stretch of walkway along a main street and recorded the tim~ passers-by took to cover the distance. With its goal of producing a detailed description monitor her or his behavior over time. since with a large enough : sample the habitual activity rhythms will be fairly captured in the data. it may . the faster people walked. The . day of the week and '! .---_1 . meeting places.>. Unless we use a strategy to ensure repre'~ntative sampling. fixed~interval instantaneous sampling.. Later research established this as a general and robust cross-cultural pattern and even related it to the incidence of heart disease (Levine 1990). these habitual activity rhythms will confound the patterns we . synonymous with focal-animal sampling (Altmann 1974). Lee' could estimate the total time men and women devoted to manufacturing and repair . and. Scans or Follows? Whatever the focus on actor. Betzlg and Turke 1985. all aspects of the observation routine should be randomized as far as possible.. and one·zero sampling. Another possibility is to target settings. " Crete) to over 2. which targets whole households and records the activities of all the members present at the moment ofcontact (Johnson 1975. will not be representative of her or his general activity patterns. its useful lifetime.' each day (men. by ImplIcatlon. In some ethnographic situations l . The main issues are: when to begin each observing session. Continuous scan/focal follow (Hames 1992. continuous recording (Martin and Bateson 1993).

and less precise. making eITors and reduced levels of detail more likely the longer the session goes on.ii. SUbjects were always aware that they were being observed but rarely interacted with the observers.m~nly L L . 13. total dailY time spent in each activity can be estimated. "resting" has been described for 22 of those observations. including the sequences of verbal or nonverbal exchange between subjects that are indispensable in studies of interaction. type of weapon carried.5 hours less per day). Typically. Instead of . child care. The most common method is to note the onset of each activity and the time at which it began. yet little In detallls known about how pregnancy and lactation affect women's behavior. Where appropriate.d. it can provide levels of measurement from nominal lists of behaviors to quantitative behavior frequencies. faithful rendering of sequences.5 hours less per day). such as the 5-minute intervals in Whiting and Whiting (1975). they used I-hour focal follows. "" For these reasons. As with all systematic observation methods. with continuous monitoring. (1988). so she selected focal· subject follows. this L L I L I L L \ ~ work through much of their pregnancy and nursing periods. it thron. location. ~ behavior in the interval is described in narrative fashion. continuous monitoring is most appropriate when behavlor 'measurement is the primary goal of th~ research and the po.specific cultural practices. it imposes heavy demands on observers. : senting a brief frame in a sequence of data points over the observing session. It provides. recording the continuous flow of behavior. resultmg ID 7. tt tS "a useful adjunct to other ethnographic methods as a way to document in great d~tail . In a remote Tamang farming community. continuous monitoring results in the most comprehensive and flexible descriptions. If its chiefadvantages are its detail. In fanning communities.longer intervals.with diverting too much field time away from other important ethn0graphtc actlvttles. Tamang (Nepal). Bailey following the men and Peacock the women. instantaneous and scan sampling (Altmann 1974). Of all the systematic observation strategies. it results in a list of intervals. for example. Efe (Zaire). If we are making one observation per minute for a 60-minute session.te. of data at lower cost and often at greater reliability. responsibility for tending a cooking pot. One of its great attractions is maintaining a record of the sequencw . It is subject to difficulties in determining exactly where activities begin and end (Gross 1984:539). a relationship of considerable theoretical interest in cultural anthropology. By scheduling observations over a full day (either "daylight" or around the clock). say. durations. then we may calculate that 22/60. recording their activities minute-by-minute. 1 Fixed-Interval Instantaneous Sampling This is also known as time-point sampling (Martin and Bateson 1993). Panter-Brick found she could not use spot checks because people dispersed too much during daily work. In . and Peacock. and social interactions (who interacted. there are tradeoffs to be considered . and find that. with each observation repre. icles both brief events and protracted states. and an intuitive storylike record. Among her key findings: pregnant and lactating women work significantly less than others during slack seasons (1. a kind of ethnographic narrative structured operational rules for sampling and description.u~e 'method provides are essential to answenng key research questiOns. wherein every minute on the minute (as indicated by a digital watch with count-down functions) the observer recorded the subject's activity using a verb-object code. liable in their descriptions. Whiting and Whiting (1975) found in their comparative study of child socialization that after only a few minutes even trained observers became increasingly telegraphic. or 37% of that subject's time during that hour was spent resting. but work nearly as much as others during heavy work seasons (under 0. and fixed-interval sampling (Gross 1984). The Efe pygmies are tropical forest foragers who live in close symbiosis with settled Lese horticulturalists. it raises the likelihood of subject reactivity. OtherwIse. fixed-interval sampling divides the :-' observing session into discrete equal-length time intervals and makes observations . Less demanding on observers. since it is difficult to act normally in the presence of a foreigner (or even a fellow community member working as a field assistant) following closely behind making continuous notations on a clipboard. ing of behavior ("flow"). pregnant .5 in summer) she used local assistant pairs to keep two people each day under. each characterized by a particular behavior. when the observer's presence had an obvious impact on their behaviors. a new notation is made. For a whole annual cycle (1982-83) and all daylight hours (11 in winter. this is the most intuitive of the sampling strategies. For one thing.). The resulting record is a list of behavioral events that can be analyzed for duration and sequence.5 to 2. and unte. when their labor input is decisive for their household food supply. what type ofinteraction.L L 316 JOHNSON 1 SACKFrr Direct Systematic Observation of Beluzvior 317 of the subject's stream of behavior over the observation session. : The resulting record is much like a motion picture. problems a~socia. therefore. its costs are nonetheless significant. L narratives of scenes and processes that can provide illustrative anecdotes. The primary use of fixed-interval sampling is to measure the total duration of activities. their posture. they also recorded the type of food being eaten. co~tinual direct observation. and is but one level of abstraction from continuous monitoring. . but intervals from once every 15 : seconds to every 5 minutes are common."'erful data ~at . N. For another.678 hours of observations on 297 woman days. the weight of the load carried. obser~ vations are made once a minute on the minute. and the nearest neighbor within 10 meters. Continuous monitoring is subject to statistical problems that result from the lack of independence of observations. selective. Whenever the subject changes activities. and 19 who were neither. and to ~om~n com. verbal exchanges.. R. Example: Bailey. Example: Panter-Brick (1989). C. a kind of minicontinous monitoring. and even intensities. etc. on the instant marking the transition from one interval to the next. Other strategies are available that produce the same kmds i. with 24 women who were either pregnant or lactating.

but over the research year visits were made on 13S days totaling 3.1. the researchers applied the spot observation technique (Johnson 1975). school. fixed-interval sampling is less oft~n. In contrast to continuous monitoring.t observations. interviews can be used j.' ample time for detailed descriptions of setting. duration. Results \. activities makes the method appealing even in research where behavioral measure· ':1"r . used than ~ther tec~D1ques of systematic observation. B~t we cann?. If all we need are total durations of . Activities were coded into a hierarchical coding scheme based on eleven major activity categories (with subcategories in each): eating.e n denved. If the research subject is absent. etc. such estimates are rehable only If the sampling interval is shorter than the duration of each instance of a gi~~n behav~or.:n ~s spo. they visited all households in a scattered settlement of 13 households at random during daylight hours (6 am to 7 pm) only. if work includes all food production. Target individuals and observation times should be selected at random. Between July 1972 and August 1973. wild food collection. "The brief lime spent in recording activities took only a small fraction of the total field time. Behavior. but this advantage evaporates when individuals separat dunng the observation.xtreme: The entire observing session is reduced to a single instantaneous observatIOn.. This almost seamless integration with other ethnographic The advantages of fixed-interval sampling over continuous monitoring are th f field ~ork~rs experience less fatigue and can collect much more information at eac~ sa~plmg mte~~1 than they could on a continuous observation routine. Context data (location. if the target saw '\ the observer first. manufacture. and wage labor. But also like such a . and unlike either continuous monitoring or fixed-interval sampling. abstracted only by restricting observation to sessions with clearly marked start and end points. more compatible with other ethnographIc acllvltles. t?en r~n~?m-mterval sampling is cheaper. ~although compromises for practical reasons-such as targeting whole households at 'random rather than individuals. garden labor. the field worker is left with . although this should be kept as a separate kind of data to " be distinguished in later analysis. because ofother demands on the researchers' time. be recorded atthis moment. kn~w this unless we already have information : on actIVIty duratIon. a gIven samphng Interval may give stable estimates of fr~quency for ~ome behaviors but not others. spot Example: Johnson. Continuous monitoring records the flow. then continuous monitoring is more tru~~?rthy and has gr:ater face validity.. the field worker can also monitor the activities of mo~e than one person. duratIons and sequences. Wanting a descripti~n of how tropical forest for~ger. thIs IS a hIghly preferred method for certain purposes. In fact. spot observation is only a formalized (and randomized) version of common ethnographic visiting practices. A.) can also .the order of images carries little or no infonnation about the sequence of activities iIlat precede and follow each observation. Machiguenga (Peru).described. tools being used. '!' ". solely from fixed-interval sampling. Ideally. activities. spot checks.to fill in the description. and '". forcing the observer to do a focal-follow on an individua~ or subgroup that remains together. . like lime. ~! and records the activity at the instant the target was first seen (or.:. it is debatable whether ~stlmates of frequency. sUb~e~t reac~lVltyJ and the successive observations within a session are not statIstIcally. or following a circuit through the community-are . visiting. "because travel after dark is hazardous and Visiting at night is not encouraged. When subjects remain In one place. { '~servation is analogous to an album of random photographs: A few might seem scattershot and unrepresentative. In a sense. events versus states). Since each observing session is only a moment." . Johnson (I 988). Techniques for choosing the appropnate sample Interval are discussed by Martin and Bateson (1993:93). the observer finds the target individual at the selected time . 13% for women. . weather. idleness. If the record of fixed-interval sampling is analogous to a motion picture. but a large number of truly random snapshots . Since . the activity she or he was doing just before she or he paused to : greet the observer).horticul~uralistsspe~d t~me. Some of the disadvantages of continuous monitoring remain true for fixedInte. flows seamlesslyfrom moment to moment. ~¥ ':Ii·~ activity frequencIes. Among Ihe findings of the study Was the sensitivity ofconclusions about "work" to differences in definition: If"work" is garden labor. food preparation. and O. If the goal is data on C C l ~7. Fixed-interval observation abstracts further from the behavior flow by sampling the observation session itself at a series of evenly spaced discrete moments. As a rule. j' (Efe women spent III minutes per day ID swiddens as compared to only 40 minutes per day for Efe men). Random-interval sampling takes this process of abstraction to its logical e. women only 7%. Independent.'. . and perhaps more reliable. informal ad libitum observations.rval samp!ing: the continuous presence of the field worker increases the risk of. add food l l l :I ! .' '{ment is secondary to other ethnographic goals. Perhaps because of thes~ potential drawbacks. photo album. and much of their "foraging" activity is spent obtaining honey and meat for trade to Lese villagers.provide an overall summary of how people spend their time. Random-Interval Instantaneous Sampling AI~o kn0v." Households and visiting times were specified in advance using a table of random numbers.318 JOIlNSON I SACKETI Direct Systematic Observation of Behavior 319 was noted and the observation was omitted from the data in the sample. or simply as the time allocation t~chmque. and sequence are entirely trustworthy wh. hygiene. then men work 16% oflhe daytime. the visits brought us into contact with community members who could be interviewed for other purposes" (Johnson 1975:304).behavlOrs ~Iffe~ in their typical duration (for example.'.495 spot observations of individuals. Visits were not made every day. but not wanting to commIt large amounts of field time to continuous momtorlng. the percentage climbs to 34% for men.'. child rearing.I showed these "foragers" ~pending surpr~sing amounts of time tending Lese gardens . Spot observation begins with an observing schedule that specifies the day and lime of each observation and the target individual(s) whose behavior is to be . often acceptable. ethnographic interviews.

but We don't care how much of the activity was performed. 1987). wage labor) and those done in private (sex. sometimeS . ~nother problem with spot checks is that as a practical reality we end up relyin~ .L L L . On the other hand. Cross-cultural psychologists have found evidence that individuals in a variety of cultures integrate successive events over a period of about three seconds to construct a subjectively experienced "present" (Feldhutter et al. the most prominent activity would likely be "greeting anthropologist" (Scaglion 1986). defecation. would allow for a broadly accurate picture of how people spend time (Baksh 1989-90). That is. The person-day diary is a running account of who performed a particular activity each day of the study. i 320 IOHNSON I SACKEIT Direct Systematic Observation 0/ Behavior 321 preparation and manufacture. Not only does the absolute amount of work change.. "Random spot checks are in fact very economical of observer time [Compared to other techniques) in terms of sampling validity there is little question that random spot checks is the method and level of detail of choice in time allocation studies.·:/'. this opens us up to the same problems of memory distortion and biased attention as any other interview data. socializing. Bemard and KiIlworth (1993) have estimated the number of observations needed to achieve varying levels of accuracy depending on how rare or common a givett activity is (see Bernard 1994:325-326). as few as five obServations on each individual. In reality.. if spot observers did consistentiy record the behavior that occurred the instant they first spotted the target. Perhaps we should acknowledge that "instantaneous" behavior sampling actually involves a threesecond observation window during which the observer constructs the observed activity in ways that have not yet been explored.. Researchers have noted the' difficulty of observing an instant of behavior. Since each SUbject can be surveyed quickly. eating. for example. without the benefit of long-term . 46% for men. Martin and Bateson 1993). it allows the researcher to keep track of a large number of subjects each day. It gives a semiquantitative assessment of activity by measuring the number oJ intervals in which the target activity occurs. Spot observations will give a rough picture of time allocation within a small community with a surprisingly small number of observations. insights. Spot checks of readily visible . memory. But the more detailed a description needs to be. The leave-and-return log is a refinement that records the time the subject leaves to perform an activity and the time she or he returns. including person·day records. and even such food production as hunting and gardening. confounding efforts to uSe observational data to understand how environment (context) influences behavior. food preparation. 1990).." ca. hunting or wage labor----occurred during some specified interval of time. if conducted strictly at random. In common with other methods of direct observation.:~" . typical of spot observations. tradeoffs exist here as well. the researcher is still making a choice about which aspects of the stream of behavior to code into "activity" (Altmann 1974. since the method doesn't provide true frequencies and S)'stematically underestimates activity frequencies (since some activites are missed) L L L L namely. add child care and the men's figure remains the same but women's rises to 56%..000. activity-presence recording is most useful when applied to general activity categories of considerable duration. some statistical problems disappear. field workers tend to watch during a brief "observation window" before they come to a judgment about which behavior the subject is performing: eVen here. On the one hand.:11"0 1. in a community of 200. '- L L L L L social activities(Hawkes et a!. including activities at a distanCe (hunting. Several techniques are commonly used in one·zero sampling.:. on mformant reports and recail to describe a significant amount of behavior. . another bias of spot observations concerns nighttime activities (Bernard 1994:325-327). violence). including ritual. But. we collapse context and behavior where ideally they would be separate data points that We could compare later. . broadening exposure to local sCenes and bringing serendipitous . Hence. Furthermore. Scaglion's (1986) study of 24-hour time allocation in an Abelam community (New Guinea) makes clear that many interesting activities occur at night.::lbJe to overestimate '1TF' . but does not tell just how much of the time window was actually Spent in the specified activity as opposed to rest. however. One-Zero or Activity PresenCe Sampling One-zero sampling is useful when all we need to know is whether a particular behavior of interest-say. will be missed or inaccurately estimated by small numbers of spot observations. In practice.:_. it introduces other biases since activities conducted away from the observing eye of the field worker ar~ often certain kinds of activities. The label "frequency" is • mIsleading. Rare activities. and so forth. and household activities (Borgerhoff Mulder and Caro 1985). the rule: Record the activity in progress before the presence of the observer became known to the target. that of individuals not locatable during the spot check. to the degree to which We use context to decide what the person was doing. Because each observation is statistically independent. leave-and-return logs. the more observations must be made. 47% for women. It is analogous to an industrial worker's time card in that it gives a list of the tasks performed and the window of time within which they occurred. has the obvious result of OVerrepresenting daytime activities." The method also improves the ethnographic' enterprise by requiring researchers to be present at places and timeS they might not ordinarily chOOSe. socializing.rl . perhaps as low as 1. such as labor patterns. and the in-out diary. based either on direct observation or interview. On the other hand.lIed a "Hansen frequency" (Martin and Bateson 1993). but also our evaluation of whether men or women work longer hours (see Baksh 1990). Confining the observations to daytime. Gross (1984:540-541) writes.

Recording Strategies Descriptions and Codes At the moment of recording observations. e Example: Hurtado aod Hill (1987). It allows . . . confirming their status as foragers. During one 2S-day field ViSit.leaves us with three options: (I) avoid coding entirely by ke~pmg textual deSCrIptIons and treating our data as a corpus of qualitative narrallves." Coding has its own advantages. gender.. reliability. Minimal observer judgment is required to score behavior and recall' data can be used with perhaps less error than in other forms of' behavioral ' observation. and seasons. position in the family. and perhaps other information like occupation. As part of a general ethical requirement to protect the anonymity of research subjects. and do so more efficiently. . as a rule. 517 for men and 404 for women). T~ls. On the other hand. At one end. typically in a detailed codebook that spells out how codes are . The c~ief advantag~s of a~tiv.e~ th~ subject: the risk of reactivity is minimal.'epres~ntations with a selective.ity-presence recording are that it is cheap in . symbolic . use of \ text descriptions of hehavior allows conceptual confusion to lay buried in ways that . (2) collect textual descriptions to be coded later using categories I 1 "::'. minimizing overlap '>. in some contexts. identifY the subject to a casual outside observer.322 JOHNSON I SACKElT Direct Systematic Observation of Behavior 323 while overestimating durations (since uncounted activities also helped fill the tim . explicit. will have a unique identification code. The subject should be on a census list that contains relevant information like age. (mutually exclusive and exhaustive. and health status. and pethaps the game being sought. I I . This problem can be addressed by using participant observation or systematic behavior observation to detail daily patterns of activity among a subset of individuals that can then be used to interpret person-day records. age-sex ': groups. codes are brief. Researchers have distinguished two sorts of behavioral descriptions: structural "descriptions of an actor's physical movements and functional descriptions of the ~: presumed objectives or consequences of the behavior (Borgerhoff MuIder and Caro :i 1985. It calls for a degree of selectivity: The goal of '''''. observer lime.. Unhke other foragers. They can be as detailed and topically broad as the situation allows and as finely contextualized as necessary. household. and places great demands on the observer. limitations: codes reduce the complexity and ambiguity of .They have a potential richness that can allow reanalysis at a later date or by . Robson 1993).. it also established that the bulk of their food still comes from wild game and roots. real behavior and to that degree represent a loss of information in exchange for order and clarity.s~d a battery of method~. Fieldwork in 1985 u. usually. it is ideally possible to have a complete set of codes to cover ""'::" all activities. what do the data indicate in levels of time and effort? Two individuals may be reported to spend the same number of person-days at work. Codes are ideally explicitly '•. coding is not to capture reality in all its complexity hut to highlight those aspects . Each observation should also ·pinpoint the time of observation on a 24-hour clock Structural and Functional Descriptions :(. too. these codes will not.. '. Structural descriptions document the actor's posture (sitting. days of the week. Just . behavioral descriptions are open-ended textual accounts of behavior. It gives valuable data about labor variation among individuals. we have two approaches open to us: Describe the activities we ohserve or code them. where the observer's presence is highly likely to lead the . nontechmcal. .• !. with discrete codes for discrete activities. at the other end.~':: for descriptions of similar events to vary greatly from one instance to the next. Each subject. i . this study shows that Cuiva mcn produce more food than the women do. Cuiva (Venezuela). The real limitation of the method is the ambiguity of interpreting the results. noting food resources brought back and the co~position offoraging parties (0=921 person-days. li l J The Behavior Record Each observation in a behavior recotd should have a unique identification code fOt record keeping. This can be very valuable. but for one this may represent a few hours of labar whereas for another full days at work. developed once the research is complete. These strengths of coding are t:~ also.:. or (3) code activities at the moment of . 10 actIvItIes hk~ hunt~ng. The method IS also hIghly compatIble wIth other ethnographic activities. different researchers.{ to be applied so that different coders can achieve a high degree of interobserver {.nt. With codes. observation using categories developed prior to or during early phases of the i research. and reliable within the limits of what it seeks' to measure. and preferably unamblg. The recording style is free to adapt as the research progresses i and the goals of the research change or become refined. ?unters t~ modIfY thel~ route: pace. I ! l: l . Because the observer doesn't follow or ~ncou. ethnicity. we consider significant in the context of the research. would be immediately exposed by requiring explicit rules of description. Hames 1992). Scholars had disputed whether Cuiva still lived as foragers or had now become settled agriculturalists. the authors used an m·out notebook" to record departure and return times of all individuals leaving the settlement.to document subsistence ecology. defioed and ohjective. obJectIve.uous meanmg.: Text descriptions have the advantage of being like conventional ethnographic Fdescription and may be used as such during analysis. window).

The specific functional activity classification scheme-which is best worked out in advance of making observations-depends on the nature of the research problem and the habitual activity patterns ofthe study population. Coding To facilitate cross-cultural comparisons of activity patterns. . including mining CM Manufacturing articles for sale CS Shopping. Thus. cultivation of nonfoods or cash crops tfthese are undifferentiated from subsistence production. with sub-categories added that are appropriate to partiCUlar cases. CA Cash cropping. buying and selling. Analogous coding schemes have been used by sociologists working in urban societies since the 1960s (Szalai p H . and soil preparation. Since structural codes are straightforward descriptions 0 observable actions. rather than recording in detail all the component movements of the actor's behavior. tbe observer describes the action using broader functional categories such as food production. etc. at the beginning of fieldwork can be invaluable for testing and refining activity classification schemes. Includes foraging. The goal of functional description is to characterizeleast in a general way-the purposes and likely outcomes of the actor's movements. touchin . agriculture. For example. Table 3 shows a hierarchical coding scheme developed for use across cultures in small-· scale and peasant societies (Johnson and Johnson 1988). with minimal inference on the part of observers. and so forth. "make ther~sultSof different investigators collected for different purposes usable for comparative purposes. traps. clearing an irrigation ditch. and husbandry producmg food for household consu~ptlon. fences. improves the richness of our descriptions of general (functional) activity categories such as "production" and. Gross (1984:542) calls on investigators to "standardize their coding of behavior into a set of broad descriptive categories of behavior. we stress that any standardized coding system can be modified. etc. However. manipulatory movements (carrying. PC Cooking food PG Handling or processing food for storage PH Handling or processing food for consumption in the near future PS Serving or transporting food PX Other food preparation PU Food preparation. irrigation channels. service for money CX Other commercial activity CU Commercial activity. and serving food. nonpaid work on facilities such as roads. storing. the action of repeatedly raising and lowering a hoe while standing may have different outcomes depending on whether it is perfonned while weeding a garden. implements. TABLE 3 Standardized Hierarchical Activity Codes Designed for Cross~Cu/tural Comparisons of Time Use Patterns. Recording both physical actions and their presumed consequences increases confidence that measurement was not biased by preconceived notions of function. MA Making or repairing portable artifacts MC Making or repairing clothing MF Building or repairing immobile facilities MM Acquiring materials for manufactures tv1X Other manufacture MU Manufacture. tidying. cooking. etc. pet care. receiving an object. cleaning clothing. or killing a snake that wandered into the yard. cleaning HM Fetching and managing household water and fuel M "leisure.324 )OHNSON I SACKETI Direct Systematic Observation of Behavior 325 walking.). and bartering CW Wage laboring. etc. Struc descriptions may be conveniently recorded in grammatical data "sentences" "individual X walks rapidly" (subject-verb-adverb) or "individual Y peels potato" (subject-verb-object). The field worker need not choose between structural and functional description: There are good reasons for collecting both simultaneously." he points out. type unknown Manufacture Making and repairing the household. handling. selling labor to others.). This would. A period of infonnal observation. they can be recorded reliably and objectively. striking. type unknown Food preparation Processing. child care. cleaning related to meals. Food production ." With Gross (1984). furnishings. and social interactions (speaking. HH Housekeeping. type unknown Housework Household tidying and cleaning chores. hygiene.). climbing. raising livestock for sale CC Col1ecting wild/natural products for sale." and preserves information on both what the observer saw and how she or he interpreted it (Borgerhoff Mulder and Caro 1985). the same physical act can have very different consequences depending on the context. type unknown c Commercial activities Activities oriented toward the production and exchange of money and trade goods. FA Agriculture Fe Collecting wild plant foods FF Fishing FH Hunting and fowling FL Tending food and draft animals FX Other food production FU Food production. managing household water and fuel. clothing (all for household use). including travel to and from site of food production.

2. Individual Self~involved activities that do not tit well into other categories.·:rreliability. Note: this code is not to be used to distinguish unobserved activities ("time-outs") when the subject is temporarily out of sight. Whichever is chosen must be consistently adhered to and made explicit in publishing the data.?. visiting Other social activity Social activity.326 JOHNSON / SACKEIT Direct Systematic Observation of Behavior 327 HX HU E Eating Other housework Housework. on an extended trip visiling another community). This preserves information about simultaneity but raises the problem of double accounting (inflated n). napping Other individual activity 2. 6. but probably unlikely) UH Away to do housework activity (also unlikely) UE Away 10 eat (unlikely) US Away to do social activity VI Away to do individual activity UX Away to do activity classified as "Other" UU Away for unknown purpose there might be an activity code (cooking) and a social code (child care). although field workers using this activity list undoubtedly will want to subdivide and specify this residual category. Code the different activities as separate dimensions of behavior. Use a structural coding system and describe physical acts. lE IG 11 IN IP IR IS IX Acquiring education or infonnatlon. such as coding food preparation over child care: In this ease. such doubling-up of activities is actually an interesting empirical question: For example. There are two questions that frame this problem: Does a single observer consistently record the same behavior in the same way (intracoder reliability)? Do two observers consistently code the same behavior in the same way (intercoder reliability)? The reliability problem must be addressed early in the research -even prior to it -so that a detailed codebook can be constructed. Codes must be applied in practice and evaluated for usefulness~ coders must practice also and Qb . No secondary codes suggested. In this case. the woman would only be counted as cooking. socialization in a face-Io. Develop combination codes such as cooking/child care to cover the most Common cases of simultaneity. The reliability problem. I. it is often argued that women's activities are constrained by their responsibility for child care. th~ prima?. and ingestion of non foods ES EX EU Suckling Other food consumption Food consumption. The codebook should do more than just list the complete set of codes. political activity (including both spectating and participating) Group recreatlonal activity or public entertainment (participating or spectating) SS SX SU Socializing." SOURCE: Based on Standard Activity Codes. so that their other activities must be compatible. Code both activities. 4. chatting. ritual.refer 10. UF Away to do food~producing activity VC Away to do commercial activity UM Away to do manufacturing activity UP Away to do food preparation (possible. type unknown S. Eating and other meaJ-time activities. type unknown Other Activity that does not fit into any of the above categories. 3. Away from community unobserved This catego. actively tending child SE Acquiring or giving education. applies only to activities that take place outside afthe specified geographical range of the behavlOr study (for example. type unknown Three problems that confront the researcher during coding are simultaneity. Establish priority rules. ' se Child care. Code both activities as a full observation. I I Each of these solutions has its strengths and shortcomings. The simultaneity problem. perhaps along with a functional activity code (BorgerhoffMulder and Caro 1985). I I SO SP SR Care for anolher (nonchild). UCLA Time Allocation Project (see Johnson and Johnson 1988). Secondary co~e~ . activity best describing the purpose of the trip. thus preserving information about simultaneity. S Social Includes a broad range of activities not better classified into the above categories which are distinguished by social exchanges. drinking. This avoids the problem of inflated n but discounts activities performed simultaneously: nursing a baby while cooking only counts half as much as nursing a baby alone. This solves the problem of inflated n but loses information about simultaneity. alone Self~grooming. and context. type unknown Strategies for handling the simultaneity problem: I. What do we record when an actor is performing what could be coded as two different activities? A common example is a woman nursing a baby while tending a cooking pot: is her behavior "child care" or "food preparation"? A small but significant proportion of activities are not mutually exclusive in this sense. . but treat each as a half~observation. I I IU U Individual activity. receiving care Group cerem~ny. giving priority to those activities earlier In the list. It should spell out how the codes will be applied. information. dressing.face context I I . EE Eating. hygiene Idle due to illness Idle "doing nothing" Participating in individuaJ religious observance Participating in individual recreation or entertainment Sleeping. .

Current Anthropology 26:323-336. weather.. Baksh. M. R. C. H. we must remain flexible and ready to change methods to meet contingencies. Cultural Anthropology Methods 1:3. to go 'overboard with context descriptions: in our experience few anthropologists do much with the context data they collectpeople tend to do agriculture in their fields. such as whether the behavior was directly observed or established by hearsay. R C. Ethnology 32:207-215. worship. Much of ethnographic description is intended to tell us how members of a group behave: how they make a living. 1985. I. and H. and P. Zaire. celebrate. 1990. CA: AltaMira Press. O. L L L L Alexander. Bomstein M. We do not take seriously enough the severe limitations that the cultural construction of long-teno memory place on the ability of both research subjects and fieldworkers to develop accurate descriptions of ongoing behavior. Turke. Time Allocation of Efe Pygmy Men and Women oJrhe Iturt Forest. M.. and Ecology._". 2d ed. fairness. and T. 1988. and P. in the sense that they change little or not at all during the research (for example. Peacock. L. M. Current Anthropology 27:382-396. Borgerhoff Mulder. K. Researchers who do so will be able to make the inevitable compromises fieldwork entails while achieving the most accurate and representative descriptions possible. New Haven: Human Relations Area Files. map. Boster. L L L L REFERENCES Understanding of context can also be enhanced by attention to. Time Allocation among the Machiguenga of Camana.i' observe. Johnson. t. As with behavioral descriptions.. resolve conflicts. '. 2:6-7. The Construction of Primary Data in Cultural Anthropology. The Spot Observations Technique in Time Al1ocation Research. Bailey.. I. and technology used. G. A census. H. Werner. A steep learning curve means that a few weeks of practice may be needed in order to stabilize recording or coding and achieve a high degree of intercoder reliability (Robson 1993). Common context variables include date and time. Society. Kiiiworth. Measuring Time Allocation: ObserVation and Intention. verisimilitude-ofethnographic descriptions is of central importance to us. New Haven: Human Relations Area Files. The context problem. Research Methods in Anthropology: Qualitative and Quantttative Approaches. Accurate descriptions of behavior matter because. this is only possible to the degree that our descriptions of behavi are separate from our descriptions of context. political institutions.. Caro. Romney. when. and N. Bernard. from the most theoretically abstract to the most immediately applied concerns./. Vol. Altmann. Yet participant observation typically produces descriptions of behavinr with such error rates.'. It is possible. housework in their homes. Some aspects of context are constant. Other aspects ofcontext are variable and must be recorded on each observation. location. 1985. social interaction. community matter. nur understanding of the causes and consequences of human behavior depends on them. R. A. raise children. Betzig. yet we as ethnographers are astonishingly cavalier about how we develop our descriptions of behavior. 1993. We would ordinarily have little respect for theories or policies based nn data that had error rates ranging from 50%-80%. 3. T. the methodological issues we have addressed in this chapter . But. The Use of Quantitative Observation Techniques in Anthropology. 1982. The best way to do this is to develop skills in a variety of methods of direct systematic observation and an understanding of the rationales behind them. The Human Machine. The accuracy.L 328 IOHNSON I SACKETI Direct Systemartc Observarton of Behavior 329 L L L learn how to increase their consistency. As noted above.. Carlstein. H. 1989-90. H. 3:4-5. M. P. in all the relevant contexts of their lives? These are issues that we should address as far as possible in the design phase of research. Preindustrial Societies. 1976. Time Resources. 1986. L. Once it is accepted that accurate descriptions of the behavior of members of a or L . Bemard. however. D. Sampling inTime Aiiocation Research. completeness-in a word. and (2) provision for' open-ended comments on the circumstances of the observation. R M. and A. Bemard. what people do matters. Baksh. . 1994. effort put into context descriptions should be constrained by the question: What are the data for? f become paramount What does it mean to describe an actiVity? Whom should we '. New York: Columbia University Press. 8. and where? At what level of detail? How can we sample from the ~. 7 in CrossCultural Studies in Time Allocation. R. roads. 8ehavior 48:1-4\. The Observational Study of Behavior. R Ember. The Pace of Life. Kasakoff. etc. London: George Anen and Unwin. Current Anthropology 26:647-650. unless the observations are carefully recorded immediately as IIw hrhnv. 1992. (I) metlwdological details needed to establish the reliability of the observation. Peito. Walnut Creek. Bornstein. and status. or general ethnographic observations. Vol. stream of behavior in a community to generate a fair representation of people of differing age. because research in real communities is always unpredictable to some degree.. Nature 259:557-559. Conclusion In ethnography. We use context both to help us interpret behavior and to look for connections between environment and behavior. and a general ethnographic description often suffice to supply this information. amplification of the data recorded. Vol. gender. mourn. I. A. 3 in Cross-Cultural Studies in Time Allocation. 1974. markets).or occurs.

Paolisso. eds. V. 1 In Cross~Cu/turalStudies in Time AI/ocation. Naroll and R. New Haven: Human Relations Area Files. Hili. Real World Research: A Resource for Social Scientists and PractitionerResearchers. II in Normal Personality Processes: Progress in Experimental Psychology Research.. K. B. Boletin Antropologico 3:39-{)5. 1993. New Haven: Human Relations Area Flies. 2d cd. American Anthropologist 88:313-338. Whiting. I. R. K.. Weller. Szalai. B. Hurtado... Thousand Oaks CASage Publications. 1995. The Development ofCognitive Anthropology. 8. In Emics and Etics: The lnsider/Outsider Debote. K. Lee.. A Problem of Bias in Scan Samplmg. Measuring Behavior: An lntroductory Guide. Shweder.D. 1975. R. g DUfour. M. and A. eds. H.H. and K. 1973. Methods for Observing and Recording Behavior. Robson. American Scientist 78:450-459. and P. Culture as Consensus: A Theory of Culture and Informant Accuracy.W. 'Lhennillier. Human Ecology 13(1):29-47.. Ethnology 14: 301-310. Whiting. ... R. R. physical Activity and Energy Expenditure. Johnson. Men's Time Allocation to Subsistence Work among the Ache of Eastern Paraguay. Romney. Chicago: Aldine. M. S.. Washburn. R. I. and W. Martin. 1990.. 1964. Time AI/ocation omong the Machiguenga ofShimaa. R. and M. 282-315. 1979. Hawkes. Whiting. 1986. In A Handbook oJ Method in Cultural Anthropology. H. 1996. M.G. New York: Random House. L. 1987. Time Allocation in a Machiguenga Community. Ecology 17:205-228. A Comparison of the Efficiencies of the Shotgun and Bow in NeotropicaJ Forest Hunting.. Pp. R. Cambridge: Cambridge University Press..40. How Accurate Are Time Use Estimates?Effects or' Cognltlve Enhancement and Culturai Differences on Recall Accuracy. 1982. Measuring Human Relations: An Introduction to the Study of the Interaction of Individuals. Annual' Reviews in Anthropology 13:5 t 9-558. 48-6 I. Liikkanen. A.:~. Cambridge: Cambrid e Untverslty Press. V. ' . t 992. Hames.n the Sac/a. Applied Cognitive Psychology 6:141-159. Memory and the Assessment of Behavior. Hawkcs. Maher. 1984. New York: Academic Press.'tliemj. 1989. M.I. Scaglion. Pp. Vol.. Turke. 1976. Pp. Hill. Lhennillier. The Nature o/Cultural Things. Oxford: Blackwell. A~ensberg. A. and C. and R.ter··Briick. 1987. E. Pike. The Time and Energy Expenditure of Indigenous Women Horticul_ turaltsts In the Northwest Amazon. . Fact and Artifact in Trait Perception: The Systematic Distortion Hypothesis. K. In Measuremen/.M.. and N. L.C. A. 1991. Behovior 58:78-98. R. Sciences. and I. In Human Reproductive Behavior: A Darwinian P~rspec/ive. M. A. 159-186. Pp. Lumpkin. ed. Time Allocation among the Yukpa afVenezuela. Blalock. A. Eibl-Eibesfeldt. and R. D'Andrade: R. Cohen. D. A. HaITis. Cambridge: Cambridge University Press. C. Motherhood and Subsistence Work: The Tamang of Rural Nepal... diss. l 't . 65-100. K. Time Allocation. Hurtado. K.M. M. C. 189-200.~ S. American Journal o/Physical Anthropology 65:37-46. The Pace of Life. 1979. and O. Batchelder. 1. and A. 1974. Moving in the Beat of Seconds Ethology and Sociobiology /1:511-520. Gross. 1990. Kaplan. Parent-Offspring Interactions in aCaribbean Village. The Importance of Nighttime Observations in Time Allocation Studies. lohnson. The !Kung Son. Vida economica y social de un nucleo familiar Yanomami. L. eds. Bateson. M. Human Ecology 15:163-187. Kaplan. Energy. and I. Emics and Etics Revisited. 1986. B. A. Hurtado. Sackett. Hill. Use ofTime in Finland. T. l l l . Cambridge: Harvard University Press. 1996. New York: Columbia University Press. Children oJ Six Cultures: A Psycho-Cultllral Analysis. Pp. Saris. A. Humon Ecology 7:219-252. Time Allocation: A Tool for the Study of Cultural Behavior. IL: Human Kinetics. Helsinki: Central Statistical Office of Finland. 1972. BorgerhoffMulder.' Smith and 8. Dunbar. R. 1983. 1984. Some Aspects of Research Design and Their Implications in th Observational Study of Behavior. R. eds. Headland. Journal of Anthropological Research 43:239-246. 'Montoy.. Pp. Cambridge: Cambrid8e University Press. 203-235. M. Measuring .M. American Ethnologist 13:537-545. Genetic Psychology Monographs 22:3-147 D·An~rade. and J. 1990. I. ~".D.ol. L. Los Angeles: University of California. 1979. Betzig. Early Dry Season Subsistence Ecology of Cuiva (HIWI) Foragers of Venezuela. The Use ofTime: Daily Activities of Urban and Suburban Populations in Twelve Countries. Flinn.M. Levine. l l l I Feldhutter. Sackelt. B.. Winterhalder. New York: Aldine. Schleidt. Hames: R. Time.. Maher and W. In Evolutionary Ecology and Human Behavior. and P.. Harris. 1993. eds. M. 2 in Cross-Cu/turaIStudies in Time Allocation. ~ 9. Ph. Whiting. M. E. 1988. H.. . R. e Englc. and M. 1975. Johnson.330 IOHNSON I SACKETI Direct Systematic Observation of Behavior 33/ Chappte. R. W. 1988. and J. A. 1985. M. Kemper.. 1988. H. W. and the Indolent Savage: A Cross-Cultural Test of the Primitive Affluence Hypothesis.. Department of Anthropology. A. D. Harris. 1981. ed. The Hague and Paris: Mouton. Kiiski. Champaign. Val.P. H.

is the nature and location of such constructs as self. and action have to do witll social process and with '!he implications of history.aJien to them? What.". cognition in different kinds of communities? How are the . morality. LEVY DOUGLAS W. identity.'i'j'. we will be concerned with certain techniques for investigating '. what as an intrapersonal (psychological) form? When is this traditional "dlchotomy misleading. stability. so that some synthesizing concept of behaviorallocus and Integration might be warranted? What do these various localions of local influences on thought. agency.'in relation to a particular person. . then. sometimes coercive.. and change-and.'.freedom? . HOLLAN L ! ~ Person-Centered Interviewing and Observation L L L I L Introduction . In the interplay between public and private spheres? What. emotion.sometimes enabling interplay with a context that is in some way separale from and c. with the idea of . A number of preliminary cautions and L L L 1:: l L 333 .'. How are community members ·consliluled by their contexts? To what degree and in what way are they at least 'partially autonomous individuals.' In this chapter.. is to be thought of as a communal (sociocultural) form. stabilized..'-': L L L ren ROBERT 1. . in a particular place and . and located .'.these relations of individual and context. engaged in a dynamic.A major problem for contemporary anthropology is to clarify theoretically and "empirically the nature of the relations in various communities-and various kinds "'ofcommunities-betweenindividual members oftbe community and their historical .and current sociocultural and material contexts.(' phenomena relevant to these constructs differentially formed. indeed.

. .. There are still some of them about. as he or she reacts or responds to "reliable" techniques such as those used by "scientific" technicians to assure what they take to be valid and reliable (that is easily replicable) results. determine the issues to be covered in the interviews. market. and errors." as an object of systematic study and observation in him. aware of the ideological and methodological disagreements that some anthropologists will have with the quasi-positivistic or "cookbook" way in which we have necessarily phrased this methodological manual. we hope to elicit behavior that moves beyond roledetermined surface scripts to suggest hidden or latent dimensions ofthe organization of persons and of the sociocultural matrix and their interactions. whIch may well be criticized. One final issue needs comment. and undermined as a result ofthe successful inquiries that they generate. both as output and input. and insofar as anthropologists are working in other kinds of places. therefore. systems.s). We are not only concerned with the nature and interaction of what we may gloss as "relatively private" and "relatively public" fonns in one or another particular community. 2 We are. partially determine private and public forms and their relatIOns. The implication is that the study of individuals is an essential component. altered. l I I To the extent that person-centered interviews engage the interviewee as an "informant. in regard to the aspects of I~fe we. But one can begin. II It' . . the problems sketched here of studying and thinking about individual organization and person-context integrations become significantly transformed and even more complex. . to its sociocultural contexts. and are. not just lIfe hlstort. and a community in Sulawesi in Indonesia-were at the time of the original study quite remote. and subsistence. Rather. colonial or postcolonial relations.~o~ have to make use of social. ap~e."s or autobiographies. these interviews are similar to other types discussed in the social science literature. Our focus is on the relatively neglected anthropological study of individuals' (neglected in comparison with studies of "culture" and "society"). which. the purposes of some comparative personality theory. but they are condu~ted. l .. We also' are concerned with the probability that such local forms are. and proba. 50CI ual This means that none of what follows is to be followed mec haDlca y. and. '!"k it) in some important sense beforehand. They are simply used to get started.. exchange or cash ec~nomie. not just narratives. achievements. to the . In so domg. and make the interview materials intelligible.. But those topics and categories are tentative and may have to be changed as a . It is important to note that these methods are not attempts to study individuals primarily in or within themselves (for. considerably inwardly turned. in some sense. of adequate social theory. psycholo~ical..0 "i I I." that is as a knowledgeable person who can tell the anthropologist-interviewer about culture and behavior in a particular locale.ndin~ Ii~? stories to it). Person-centered interviews must also be augmented by special studies dictated by the investigator's interests and developing sense of significant problems. We believe that the distortions implicit in method-embodying differential power and status and Western ethnocentrism and self-interest-areaspects of the kinds ofat least partially correctable interview distortions that we will discuss and which can be brought to consciousness and in part rectified. results of typological features shared by other communities (for example.. of guides for the mobili~ation and ducation of a preadapted understanding. Person-centered intervieWing and observations are not made up of standard . say. ~:~er be taken as a series of examples. isolation. ~ot every?ne can grasp skillfully (or is interested in) person-centered aspects of soc~al beh~vlor. PartIcular practIces WIll be sketched and related to topics of inquiry and categories of analysis. although almost all studies have been ethnographic and local. they are attempts to clarify the relauons of mdlvlduahty. scale. in some of their dimensions. and this manual is something like a musical score. of course. They are in part such things.result of trying to use them." .334 LEVY / HOLLAN Person-Centered Interviewing and Observation 335 assertions are necessary. This manual of technique has another problem. Tbese were the sorts of places that generated traditional anthropological thought.bly evoluti." .. of systematic comparative interest. These methodologIcal preSCrIptIons are no ~ore mechanical and positivistic than is a musical score for skilled performers. but insofar as traditional communities have changed. Tbe interviewer observes and studies the interviewee as he or she behaves in the interview setting. The three communities where we have worked the most-a remote Tahitian village. density.?nary ~~iIIs of 'al knowing and interpersonal interaction that you brlDg 10 the score. at least-as anthropology began~with considerations arising from smallish and inwardly turned places. We assume that the local forms of individuality and individual-social transactions have to be studied empirically in each particular setting and for each kind of problem (although we will also argue for a concept of "types" of settings for some purposes) and that we cannot thus use some a priori generic model of "anthropological man and woman. you have to already be a "musician. Our main concern here is with interviewing in person-centered anthropological studies and only tangentially with other types of person-centered study. Tbe interviewing and observing discussed here are rather performing arts. we~e principally concerned with at the time. \ . IS '.' But person-centered interviews also engage the interviewee as a "respondent. conservative. nature of embeddedness in larger. but interviewing must be conducted in conjunction with traditional community studies that elucidate context. a city in Nepal's Kathmandu Valley.! . You have to know how to do it (or something very l . The interviews we will be centrally concerned with are not just samples of ~isco~rse-:-notone kind oflocal discourse among others. or to humanize an ethno~raphy br. Much of our dISCUSSIon IS directed at the problems and issues relevant to such kinds of communities.m an attempt to ~ttenuate and disrupt ordinary and conventional patterns of SOCIal dIscourse. complexity. or some issue in general psychology. And as not everyone can go from the score to a satisfactory performance. Tbey are often crudely approximate and are rough startlDg places.or herself.

what do people think?" T~ese oscillations betweenrespondentand informant modes illuminate the spaces." Questions and probes may be relatively focused or ." which is more open than "Tell me about how your mother disciplined you.. Some anthropologists don't work in the native language. of course.:. vague or "open ended" ("What was going on?"). It is not only harder to shape questions for respondent probes than for informant probes (it is harder to be properly vague in a foreign language than properly precise). so that the two types of description ' complement each other. as an expert witness (albeit with a limited and special perspective) about some community procedure. both for social and psychological interpretations." which is more open than "Tell me about your mother._ _ _ _"--. and topics.. of reports and behavlOrs. between the woman-in-herself (either in her own conception. Problems arise when we try to theorize these separations and relations-we cannot in any satisfactory way simply force private into something like "mental" or "personality. _ L : --_. The Interviewee as Infonnant and as Respondent L I iL L L L L L ! L There is a significant difference between asking a Tahitian interviewee something like "Please describe for me exactly how and why supercision (a penis-mutilating rite of passage) is done by Tahitians.". conflicts. L Person-centered interviewing depends.' It is the balanced combination of infonnant and respondent modes of interviewing that is characteristic of person-centered interviews and that distinguishes them from most other types of interviews. . Nepali etc. The: relation between the two sets of answers is directly informative (see Levy 1973: 117-122) and illuminates the force of supercision on males in a way that the informant-based cultural description does not. as is a verbal directive ~ like: "Tell me more about that. ." which is more open than "Tell me about your parents." and public or external-to-the-individual into some contrary teaJinof" "culture" or some related idea.1:..' Person-centere~ interviewing generates a field of often neW phenomena. Where an infonnant is monolingual in some local language. not necessarily in the "' form of a question. .. When informants are bilingual. . One has to know enough of the standard language to realize that the respondent is talking in a peculiar or significant way. or the fower colonial language. on adequate linguistic competence.or herself." (We will return to open and closed interventions.:·. presumably.:.. the national language (pidgin.. The second set of questions treats the interviewee as a respondent. this unpacking of phenomena is not too difficult. The interviewer tries to widen the constraints of the question as much as is relevant to what he or she is interested in at the moment. some private concern or orientation. : Open-ended probes are purposely ambiguous.. Any of these shortcuts will systematically affect and distort the respondent's behaviorin often obscure ways. .. questions. We will consider the mechanics of the interview and note sources of distortion and misinterpretation before turning to aspects of interpretation and the theoretical issues to which they .:. they may work in the areallingua franca. A remark of a young woman informant: "I felt very shy and embarrassed at that time" might be followed by a respondent-type probe: "Tell me more about how you felt" or by the informant-type questions: "What do girls When to Begin Person-Centered Interviewing Several factors detewine when in the course of the fieldwork the person-centered intervIewing should begin: Linguistic Competence usually feel under those circumstances?""How do they usuaHy act?" "If they don't feel or act like that. The tensions and apparent polarities that personcentered interviewing reveal present a challenge in theorization to each worker.. An encouraging grunt may be a probe. "Tell me about your childhood" is more open than "Tell me about the people around you in your childhood family..' How to Conduct Person-Centered Interviews fPerson-centeredinterviews are a mixture of informant and respondent questions and tprobes_ A probe is an intervention to elicit more information. This leaves the respondent a relatively wide choice of responses.) The specific choices and emphases in the contents of responses as well as the form of the responses made by the respondent within the relatively wide constraints of the probe are important material for analysis. that are then subject to interpretation. but the linguistic nuances in the respondent's discourse that convey personal infowation are often meaningful variations of the standard language. trade Malay. ".:.. or in the interviewer's emerging one) and aspects of her perception and understanding of her external context. and transformations. ".. it explores what he or she makes of the procedure._-----" L L L \ 336 LEVY I HOUAN Person-Centered Interviewing and Observation 337 various probes. Notice that if we stay at the descriptive and phenomenal level. as an object of study in him. if any.'. morality.closed" ("Exactly what did he say that made you feel like that?") or relatively . Many anthropologists don't achieve enough fluency in the local language to do person-centered interviewing.. ... the anthropologists may use bilingual assistants..). Different aspects ofidentity.:. Person-centered interviewing moves back and forth between the informant and the respondent modes." and asking him "Can you tell me aboutyoUF supercision?" "What happened leading up to it?" "What happened that day?" "Did it change your life in any way?" "How?" "What did you think and feel about it then?" "What do you think and feel about it now?" The first question uses the interviewee as an informant. coherences.. dictated less by the exact answer to a factual question than by.

course and informant interviewing. Understanding the Culture [n the same way as the interviews will drift from public to private realms (below). relatively long periods of fieldwork are required.) The most helpful way to improve interviewing skills is to study closely the errors and misunderstandings revealed by tape recordings of interviews. lary and phrasing of questions and probes with the help of a patient informant..often than not. insulting 'C to the respondent. it would have .' ~~. It takes considerable general knowledge about a place and its people before we can begin to understand the presence and significance of private variants and transformations of local cultural and social forms. '. therefore. people (unless they have good reason to keep 'silent) are motivated to engage superficially.. but the interviewee soon realizes that something strange is going on. they become. the respondent interviewee must develop . therefore. one doesn't discuss potentially teputation-threateningprivate worlds with someone unless one trusts him or her deeply. when ':. Under modem conditions. hospitality. We gave substantial gifts or reciprocal services to respondents (sometimes. and in a rather deep and comprehensive way. : t it has been our experience in several traditional communities that almost all 'people we asked (representing a cross-section of their communities) were willing 'jo become and remain respondents. They perceive themselves to be taken seriously by someone they may perceive as having high status. compassion for the naive and "confused field worker. Essentially. informants and respondents do become professionallzed. gifts of money to poor respondents). ~.. [n contrast to infonnant interviews. . a long-lasting relationship of mutual obligation is being entered Z: into. there is a sense :that it is prestigious to be chosen as an expert.: ". and curiosity. Many of the language forms used during respondent interviewing are different from many aspects of the language that the anthropologist learns for public dis. so does the understanding of the anthropologist. I 'if' ll': . The field worker must. where face and reputation are essential. We must ask at this point why interviewees engage in interviews at all. and. deeply meaningful to interviewees.:. on the other hand. It has been our experi- "I i I I l' . they express regret at the conclusion of the interview series. with the anthropologist and his fOr her interviews through politeness. Such motivations may operate in the beginning phases of a respondent interview. In the kinds of communities we have worked in. Thus. In traditional societies. We also found that we had assumed some very : long-term obligations to informants and their families.~ been a serious violation of the conventions of exchange to offer money at the start. '\ that is to hire respondents.:~. nor in his or her public role. Early in the anthropologist's stay.338 LEVY I HOLLAN Person-Cemered Interviewing and Observation 339 self-presentation. overt quid-pro-quo offers of AA payments "settling" the interviewer's obligation are disturbing and.. the respondent assumes that an exchange relation is v being set up and that some sort of obligation on the part of the interviewer will ]Tresult. Such interviews may be problematic unless the respondent '~l. including the interviewee's superiors and enemies. [n some kinds of communities. one may hire them. usually "'. which may be successful no matter whether '. may consider helping 'the anthropologist to be part of the responsibility (and validation) of their status. Why should this be? l l l '. More ".'. initiated by the respondent's generosity. High-status interviewees.' I. [n most premodem places. Trust and Repondents' Motivations [n the United States. They are selected over their peers and treated as an expert.:. He or she is not being treated for the most part as a cultural expert.. This is particularly true in relation to the anthropologist who will be' living in the community and who will be in daily contact with many different local people. function as a general ethnologist before turning to person-centered studies. 1 --. t£· All interviews provide some kind of immediate psychological reward for inter.:. So it's often necessary to work out the vocabu. at least. this minimizes community' members' expectations about the interviews. additional motivations for the interview to proceed properly."y lbe reward is intrinsic or extrinsic. i l . and mind are organized around and evoked by the use of each 0' the two (or more) languages the respondent uses. It may be productive and theoretically interesting to wotli with informants in both languages for comparison. at the end of the interview series. as one does a Western expert. and in exchange for their time and expertise. This way if things go wrong it won't compromise essential studies: (It's sometimes a good idea to do these interviews and the linguistic preparation with someone who lives outside of the community. [t is easier to gras(>-Cspecially in small communities where anxieties about public reputation require much self-censorship and self-control or in communities where : litical positions may be full of danger-why people would not want to participate. people are often willing to discuss personal matters freely with experts who are strangers to them and to do so in many public arenas. [t is also very helpful to try some pilot interviews with people who will not be major respondents. 0: $. Much close observation' of and contact with the anthropologist (insofar as that observation and contact leads to trust) is necessary before local people will consent to being relatively frank respondents. often. a master.becomes engaged for the sorts of personal reasons we discuss below. ence in very diverse communities that insofar as respondent-centeredinterviews are :. as in Nepal. The local language is almo certainly most closely related to people's private realms of experience and to the' personal organization. but as an individual. no 'matter how lengthy it has seemed to the interviewer. but it is deeply distorting not to work primarily in the respondent's core language. being conducted properly (in ways we shall try to specify). ~' viewees.people slowly grasp what is going on.

young versus old.' .: life lived in the narrow constraints of a tiny village)." and. it is essential to protect rigorously the reputation (and sometimes life and liberty) ofindividuals. developed historically for the purposes of some model or other of intrapsychic personality description). at least at the level of abstraction we are using here ("identity. modernization. the latter being the case in many traditional Indian families) are deeply suspect as an escape from communal or familial controls. then. on the one hand. The problem is deepened in cases of cross-sexual interviews.'. and.:. local context." Those stories are directly meaningful in themselves.. and construct aspects of one's private world in conditions where this proves safe-away from immediate external communal censorship. In actual interviews. isolated. In many communities. such interviews are possible if the interlocutors have some sort of (and the proper kinds of) fictive kinship relation. (Conversely. any behavior the exposure of which would cause a sig. a woman may be interviewed by a man if she brings a baby or young cbild with her. Clearly. but some categories of people (women in their defining contrasts to men. :t . On other occasions. Obviously. perbaps most people. these are the very communities in which self-exposure is most dangerous. and people with specializedroles-mothers. matter. L L L L Interview Topics r"'nop The subject matter of an interview series consists of discussions of as wide a (t~ possible of the issues in the interviewee's life that seem important to him or her andlor the interviewer. share./. sometimes in the respondent mode. The problems addressed in a caste system are different from those of a small. local leaders. nlficant matters of more or less general human concern. and so are behaviors d~fined as "crime. as being open and interactional aspects of context-person relations. This is particularly true in small. Location of Interviews To maximize private responses. traditional communities (or' those under some sort of rigid social control) where it may be difficult to express. The topics of the personal pole. It is essential to note that our list is selected and biased In certain ways. being an untouchable. friends. comparative. Inquiries into specific topics generate stories. A generalized list cannot deal with such variables. we understand them. articulate. it may be safe to talk about some subject so long as no recordings are made and so long as written notes are coded or disguised in some other manner. In others.. Often.ficant threat to reputallon.) will obviously call for particular topics of inquiry. etc. In these. Although many of these person-centered categories may have intrapsychic Implications (they were." "feeling") are perhaps more general than social variables. or a wife in an arranged marriage.. and in all communities. express. on the other hand. and historical importance (such "cl as. but the form in which they are told also yields a good deal of general information about aspects of personal organization. high status versus low status. or about . "This is the day-by-day life of us untouchables. In some communities. and trustworthy-is of considerable value for many. We will consider the topics first and aspects of how to deal with them in later sections. we would similarly expect such categories as "production" or "exchange"to be so understood for our purposes). in an interaction with someone who is sympathetic nonjudgmental. private. a relationship that is often given to outsiders in traditional kinorganized societies to facilitate social relationships. at the social level. kinship structures. it is essential to interview the respondent as far as possible in isolation from his or her family. through writing some things in a coded language) and must resist any temptation to gossip. although it may be useful to cover most of the topics listed here for comparison with other studies and because they may prove to be valuable at a later time. 340 LEVY I HOLLAN Person-Centered Interviewing and Observation 341 t :1 L L L L I '- It seems that the opportunity to talk about.. for example. the search for a private place for interviewing is a most difficul~ sometimes impossible. But the main reason for our present emphasis is that general aspects of context--economic forms. The categories of the person-pole interacting with these public forms are relatively less familiar to anthropologists. Our list emphasizes the personal pole of person-context Interactions. Person-centered interviews consist. this provides profound responsibilities for the anthropologist. of topics afinquiry that are approacbed and dealt with in certain ways.. and so on-are the familiar subject matter of general anthropological method. Most of these will concern. and person-context relations that cut across many different topics. These accounts of aspects of lives are approached first descriptively. disinterested. or explore private worlds even with friends or family. in large part. L . sig~. and secret conversations (even among friends and/or spouses. Relevant social and cultural patterns vary in different kinds of communities. both poles have to be dealt with. for example. as we have noted. deviants.' Other interviewers in other settings must develop their own lists. and acquaintances. religious forms. There may be some topics that are so sensitive and dangerous that they should not even be asked about and certainly not recorded in ways that could possibly implicate the interviewee~ Political participation in many countries is one such matter. Many topics are relevant to all respondents. relatively egalitarian village. the "forms and encounters of local social. ". There are several reasons for our present emphasis. who must protect the respondent through guarding and disguising field materials (for example.. The presence of others automatically shifts behavior and discourse toward public behavior and socially proper responses. then with an emphasis on the interviewee's personal experience and interpretations-that is to say sometimes in the informant mode.

" liAt what age?. patterns of association. with allies and opponents.. Identity is reflected in the behavior-guiding orientation to the issue of "Who am I?" as it is formulated in different contexts. how people achieve status. l 1. household caretaking and teaching practices." Aspects of Self Quite different from the idea of identity is the local (and individual's) conception about and the actual forms of "self.:other interview materials and from observations of the informant.' depends on whom I am with.." and "Who did you wish not to be like?" One can trace the interviewee's sense of developing consciousness of identity. An important aspect of identity is. with people in the local community. The processes of identity formation can be approached through questions like "Who (an individual or class of people) did you want to be like when you were a child?. class. or different from that?" directly probe aspects of categorization of self and. and other putatively formative and socially differentiating matters.. Direct probes at sexual identity include the repondent's definitions and evaluations of limen" and nwomen" and their lives. One may ask the village headman what a headman does. parents' and other caretaking adults' roles and characteristics. "'"---. importance here are not only specific relationships but generalized ones.. self-definition in relation to these general definitions. 9 Answers such as "I am an aristocrat. In the early phases of the interview series.. This entails definitions of the sexual other (as all identity categories define classes of others). envy or disapproval of aspects of opposite sex roles and characteristics. with people of the same and opposite sex. it is important to explore memories and interpretations of the rites. Ethnic. others. Here we may include such things as patterns of life in the family of origin. l l l . fonnal education. and other components of identity can be approached directly. with equals. Locating Information This cluster of issues locates individuals in the various significant components" of their community's social structure. The concept of self is related to the idea of "person" in that a community's ideas about which members are socially responsible ("persons") implies that they possess competent selves in one or another sense that must be locally investigated. Although ." "In what ways?. and of shifting. We can only suggest briefly what we mean by any particular topic category." "I am a member of such and such a family. Nonself phenomena may be variously explained by respondents and/or their communities as emanating from some natural or supernatural source. It is only at some later point in the interview series that certain items in the locating data may be returned to in respondent mode ("How do you feel about being a headman?" "What about the work makes you anxious?" "What is difficult about the work?" "What is rewarding?" "How do people jUdge and react to what you do?"). both in their current activities and in their earlier life. meanings and attitudes about "heterosexuality" in its defining contrast to "homosexuality. variously sorted as belonging to self or to nonself. Of'.. It guides the inquiry of the first stages ofthe interviews and covers such matters as social roles. of moral responsibility. and with various kinds of outsiders. economic status and activities. etc. produces very informative discussions.. This position is reflected and becomes a behavior-directing force... in his or her "identity."" etc. l l "l ._. but also developmentally in relation to what is taken to be significant past experience. Questions such as "What does it mean to be a Tahitian?" and "In what way are you like. I ." "I am a farmer. . What traits does the interviewee sense in him~or herself to be masculine or feminine? Does he or she wish sometimes in fantasy to change sex? All this overlaps with local and private forms. familial. -~~"'7':C . the direct question "Wbatwould you answer ifl asked 'Who are you?'. and their effects. multiple. alien to some essence of what it is to be a conscious and responsible individual. marital status. . or unified identities by discussing these issues directly-as well as by inferences from other interview materials.." addressed to someone who knows that you already know who he or she is in some sense. and advantages and disadvantages of being male or female. An individual's thoughts and feelings and behaviors are ." "Why?. early economic activities.-." of experientially autonomous action and. Not only is the individual to be located in his or her current life. It is also closely related to and in part distinct from ideas and behaviors about physical sexuality and general orientations about bodies. locating data are gathered principally in informant and conversational modes.. are all obviously of differential importance. in part. of course.. sibling patterns.' Patterns of Identification and Identity Formation The locating data and patterns of relationship are all related to the interviewee's position in a historical and social nexus. friends and enemies. particularly those matters having to do with learning and with the developmental aspects of people's present orientations and behavior. if any." understood as the intersection of (aspects of) an individual and (aspects of) his or her communally provided social roles and social definition.. locally constructed sexual identity.." If: ':'1 am myself. gender.342 LEVY / HOLLAN Person-Centered Interviewing and Observation 343 -. In our experience. I These matters can be discussed directly with the respondent as well as inferred from . children-and whatever else may be socially and culturally significant for placing respondents in the local social network. For further understanding. travel." if self is taken to be that subsystem of a larger mind that is the locus of "I. on the respondent. I." "It . it may be helpful to the reader to see how the categories are: used in the literature we have cited. thUS. In those societies that have rites of passage. with classes of superiors and inferiors.---------~. with relatives and nonrelatives.__. in defining contrast.

"mental disorders" (as defi?ed m local Judgment. poor appetite). usually late in the interview series. and oth~r i.L I I L_ the interviewer's understanding oflocal ideas about selves and actual local struc of self are. As many studies (for example. and one can investigate as well the stability or instability of the .msomma. . excretion. ultimately restored. guilt. to a person's value. p ople's intemal adjustment and external adaptat. sin. Probes can be used to investigate the relation of ideas about self-its separation. Illness. sense of self. to a person's nature and constitution. Related to the conception of the self as the particular focus within the larger mind of moral responsibility is a large set of issues related to aspects of moral eval~' uation and.ally trou'l' behaviors.. Levy 1984) have shown. and the like can be explored directly. local healing and problem-preventmg pracltces. pollution. deal with. repentance. and so on. Emotion L I ~ embarrassment. and the relation of bodies to action. procreation. the strength of the self in relation to internal and external pressures. but refrains from doing? Why? Ideas (and relevant episodes) bearing on what we would gloss as shame. about defense mechanisms. and the relation of those tensions to his or her actions are all important here. ~~al and social fonns we are concernedwith here. for the most part. Why is a particular behavior good or bad? What kinds of feelings and personal consequences are associated with doing good. about SOCta orgamzatlOn.ke E is defended from breakdown and is. cleanliness. Religion and the Supernatural In early psychological anthropology. Such an approach is suppiemented by (and often clarifies) accoun~s of L emotional responses in respondents' stories about other matters. it is possible to approach feeling and emotion directly. and responses about the respondent's own death as well as the deaths of others are rich sources of information about self and identity. he~daches.cate problems (symptoms). and ntuals. certain areas of local culture were thought of as "projections. or antagonism-to the body or to some spiritual entity or to other parts of mind. The OppOSItion between realistic and projective culture is problematic. terminology. m "I response to such things. based indirectly on a range of miscellaneous intem materials. menstruation. particUlarly the ~.ence-and. experience. especially.among these pr~­ sociocultural topics such as eating. Varlo~s henomena (locally interpreted in various ways) md. from where} I Such conceptions are related to the mor~l issue (below) of which tho\lghts and behaviors one (and one's associates) are responsibl for. and Healing 345 . to self-control and attempts to control others that can be approached directly-as well as being implied and illustrated in discussions ofmany topics. These symptoms and heahng and control techmques can be asked b 109 directly and are important clues about local persona1 and 'l . drinking and drugs. blending. rites. Foremost .L 344 LEVY I HOLLAN Person-Centered Interviewing and Observation . justice. doing bad? What kinds of things is the respondent aware of wanting to do. by exploring categories. to ask about respondent's sense of which actions or thoughts are really his or hers. but the exploration of ." then. if poss!bl~. aspects of aggression. m . Stress.." The Body One of the principal foci of private experience is the ways that people in various communities live in and through their bodies. the relation of selfto different contexts. These latter issues. about ~: orientations to time. and by direct observations of respondents' reactions during the interviews. and physical sexuality. ~eat[ng palms. Various miscellaneous matters are relevant here-body-centered . .. ideas of gUilt and punishment.:. to self and identity. All of these can be explored directly. is derived from how it br~aks ential information ahout any systematically organized arena. and if from uth outside. feelings.. behavior and evaluations of the respondent in relation to a range of what the intervieV:er considers (taking account of local definitions) to be the domain of emotions. their sense and evaluations of the functions and form of their own bodies and the bodies of others. . Death Ideas. a~d so on: These are~s of investigation often show clearly the interplay between pubhc doctnnes and pnvate A. the particular resources the self has to control itself. farming or fishing). .. : prevent.de. the relation of selfto the local use of drugs and their effects (and to meditation and mystic experiences in some communities). One can explore respondents' ideas about what constitutes good and bad' behavior. symptoms of anxiety or depression. Included here are such things as indications of the pre~ence of and co?texts ~or i enerating "psychophysiological stress" (for example. L i orientations. which j conventional actions. Morality . and which come from outside the self. congruence. rectification. . say. and the somewhat different issue of which behaviors respondents believe they can or cannot control at some level of thought or action..ions ~nd local ways of a. Tensions between public ideas about moral matters and those of the respon. "" meanings of intimate relationships.ttemptmg . are on the border of moral organization and conception. jective areas were religion and its associated symbols. it is possible. dent. anxIety. vlOl. and relieve the corre~pondmg discomforts. were thought to reveal "deep" psychological orientations. '~ andlor in what seem to be some transculturally recog~l~ed f~rm). and soc.~." These were areas that because of a lack of material constraints (in contrast to the reality-constrained procedures of. su.

. maneuvers and self-protective distortions generated within the interviewer.. by talking about one~elf.espondent's ~iscussio~ of his or her own children is a rich source of mformatIon o~ educalIon. Bntjust how does a friend listen properly and encourage his or her partner's talking. What the interviewer senses are abrupt and often if mysterious alterations in his or her feelings. the uncongenial true ". We may roughly differentiate two overlapping sorts of distortion of 'understanding-<:ognitive and motivated. Ilgnalthemselves.7. in fact.suspects the hstener-fr. or by showing onets cleverness.'~: respondent) and with other kinds of contexts (than the respondent's). Rather.or herself. we are talking 'about systematic distortions in understanding that can profoundly bias an otherwise '\excellent technique..learnmg. obscure. . although they may well be uncertain as to what.~. for example. stirring up defensive J. We will say most about the latter.med friend then may move on to a second stage-': that of tryIng.ences respondents had during their own childhoods (Levy 1996a). its constant . is often a pleasurable and rewarding perception of correctable error. Children . But distortion may be motivated to serve some personal need (including guarding .of the defensive operation. One listens sympathetically and nonjudgmentally. if interviewers are attentive to their own sensations and . ideas of the self.!earning is. . ~ne ~ responSlbIhty m thiS phase ofJistening is (I) to facilitate the friend's commu. it is not only the particular topic areas that make these person-centered " or respondent-oriented anthropological interviews.-. As we have noted.. to be helpful to the respondent. (See.) Fantasy.•. thinking.346 LEVY I HOLLAN Person-Centered Interviewing and Observation 347 ~~e responde~~~s understa~dings and uses of the culturally constituted realm ofth supernatural. These motivated processes lend to i. l I The Need for Self-Monitoring Conducting the Interviews Topic areas suggest what issues the person-centeredinterviews might be directed' toward.1 I l --. and dreams-projective activities generated or. .: During an interview. l I i Interview Distortion '''~. There is a significant .c or literary art. uses. arranged out of cultural pieces by the respondent him. in Jact. Levy's [1990:323-335] discussion of blood sacrifi ' m Bhaktapur and Hollan's [1996] analysis of spirit possession in Toraja. and feeling? What are the interventions he or she uses? Here a discussion . and from self-monitoring 'during the interviews. These person unde~standings. . morality. but we must first consider the basic problem of interview distortion. state of affairs may well be somehow known or intuited..of specific elements of technique will be relevant.. . of being too egoistic. the anthropological terview maintains itself in this first phase. If It moves on too quickly to the advice stage.Correction through feedback from studying the interviews. errors that good technique and skillful theoretical rhetoric. but not how to actually conduct them. responses. the nonfactitiousness. in itself.:interviewer's prior intellectual experience with other kinds of people (than the .':Ks~me deeply and centrally organizing cognitive orientations). the first phases of talking with a friend who has some. That is. . blunting >jI.end of not doing the job rightly.thing im~ortant to discuss.' DIcallon. The interviewer perceives an alteration l l . the whole point of the field study. For our purposes of trying to understand the respondent. 1~' or closing off threatening new understandings. . drawmg the frIend out and trying to help him or her say what is on his or her mind: One doesn't break off the fri~~d's report by changing the subject. and (2) to try to fully understand what he or she is trying to communicate The conve~ation with the conce. by pr~~~tu~ely ?dvlslng or moralizing. the interviewer may be aware that something "psychot logical" is going on about which he or she is unclear-the unclarity being the point {.s of community religion that may be obscure at the level of cultura descnplIon. We now discuss how the interviews should be carried out. Dreams O~r discussion about "~rojective culture" is also relevant to the relatively priva~ creatIOns of fantasy. At least some ':: of this cognitive distortion is correctable through ordinary learning based on what . This corrective . The r. 1 UCognitive distortion" is a misunderstanding of what is going on. of loosely structured interviews depends :'almost entirely on the quality of the anthropologist's interviewing.going on. is going on. Being helpful is not our present concern at least in our immediate role as ethnographers)-although we have argued that 'lIful listening and interviewing seems. they can become aware of signs of the process of distortion while it is ."The validity. Ideas about what children ar: hke and ho:" they should be dealt with illuminate a good deal of the respo~dent s central onentatlOns as well as providing clues to what formative exper. plast. to be h~lpfuI. The General Stance of the Interviewer Perhaps the most basic technical matter in conducting interviews is the general ~tance of the interviewer toward the respondent. exactly. but also how the interviews are " conducted. at least. dIfference here from simple ignorance. In the motivated case. one . etc. and experiences of the supernatural reveal meanings an funct~on. It closely resembles a familiar one m ordinary social interactions..s another nch source of psychocultural information. based on the . nor how to think about them.. We are not concerned here with straightforward unskilled :errors in technique (which can be corrected relatively easily). Creative Art.

to understand the range of a domain and the relative saliency of items in it-"Tell me about how people fish"). However they are produced.. premature re~pond~nt-type m. cloudy-minded.." ~~'::'." and the hke-<leveloped subsequently a I'IV ing?" I • • &'. one may simply' not hear what the respondent was trying to talk about.'. anger. (I) Listening to the recordings will usually make evident what was occurring at the time of altered awareness. the (always embarrassing and painful) awareness of these distorting' responses and maneuvers leads to significant self-corrections and learning." "go on. a brief note of the response and of what seemed to be going on in interview should be made-or later while listening to the tape recording of interview). we begin with locating data-th~ usual . The. or interest and may feel anxio~ L L L L L L bored.." "Describe your ho. Beginning the Interview Series. In our experience in our own interviewing and in teaching interviewing. excited. The interviewer's responses in the interview may be called "interventions. distorting the interview and suppressing the disturbing material. . and experiences. avoid topics of importance that they inappropriately fear are too sensitive.tervtewm would probably be unnecessarily disturbing to the mt~rvIewee. . Sh~fts t~ the respondent mode come gradually over the set of interv~ews as the mt~rvlewer approaches more personal target topics.-.---~_. There are various ways the interviewer may. and thus the understanding of what is going on. the distortions of the interview.:"e thIS purpose. covertly I . one may intervene to change the subject or deflect it. For if the interviewer's first response to uncongenial material is a matter of feeling. to emotionally loaded material that mIght stIr . evaluations. rte ~n' topic.ee III rega d" d " More '. The shift will begin when some mutual trust. are usually evident when the tapes and transcripts are' studied closely." "Where do you live now?. something. In person-centered. IJ Guarding Against Forcing a Wished-for Response . or disapproval in some (often quite subtle) way. Tape recordings (below) are invaluable-for two main reasons. grunts. sa Isue. " ( db 'd' Iik 1 though they will be much too cautious. or one may signal one's own anxiety.:".... In the early phases. Clearly interviewers can be brutal. at least III a preliminary way. Interventions and Open and Closed Probes Open-ended interviews can be used in the early stages of investigating any sociocultural phenomena (for example. or whatever. or infonnation about the infonnant's corner of public culture." and the like-not only indIcate attention. the reaction may be generated again. But interve~tlons-pr~bes.ays of gettmg to .. What was going on just before the disturbance? Was there a pattem to such feelings and alterations? Did they happen with certain kinds of respondents. ee aI t' g him or her to change the subject) close off the d' ISCUSSlon lca y .n.. Thus. inattentive.:-. The early phase allow~ the interviewee and interviewer time to get to know one another. with certain developments in relations with respondents? It is difficult to grasp what is going on while an interview is taking place._ man rd . up ' . and understanding has been built up.:cntrea ID • }infonnative feelings. hnwever. and by signaling back their anxiety to the interviewee.. and uses open-ended interviewing. _-----------LEVY / HOLLAN -". Follow-up questions at ~is st~ge are to ~xpand mfo. deflect' the disturbing interview phenomena." One purpose of interventions is simply to keep the interview going:-grunts and.o.. J know someone-and in an infonnant mode. the next response is liable to be an action. It's necessary at first to begin to map out via the infonnant n:'0de the pubhc world in which the infonnant exists. Introductory Topics :'" The interview series must start with topics and in a manner that seems n?tural j!10 the responden!. A kind of preemptive blandness. pger. L Exaggerating the Fragility of the Interviewee In other sections we comment on the sensitivity of certain topics that have to be recorded with care and sometimes avoided. "uh huh. but such probes quickly lead to more focused. asks questions probmg personal feehngs and ideas. with certain kinds of materials under discussion. for example." "What do you do "How do you spend a day."'.. alertness.. They show the effects of these alterations on the interview process. there is a continuous back-aDd-forth movement between the open· ended probes and more specific questions. The anthropologist asks for locatl~g m. (2) But the tapes have another kind of essential infonnation on self-motivated distortion.ant infonnation.. nods signaling that the interviewer is listening and interested se. treatmg the m!ervlewee as an expert infonnan!. sympathy. • in some systematic way.atlO.-.' . that will serve to strengthen an attractive developing hypothesis or bolster a cherished world view. y this is a matter of an exaggerated fear of the fragility of the respondent I 'mgs 0 f anxi ety • For . L L 348 Person-Centered Interviewing and Observation 349 of his or her state of conce~tration. Inexperienced interviewers sometimes. and m so bemg an y avol 109 '. fact-finding questions.is' called for at these points.use to me. For example. more or less unconsciously. quesllons. 0 . concerns. interview must struggle to understand what happened (either during the interview-in whic case. an 'Insensitive. but are so L L_ It is essential that interviewers be particularly careful to be nondirective as they approach material that they sense they wish to make come out in a certain way. While listening to the tape. which is the world that provlde~ the ~ont~xt g for respondent studies. open-ended interviewing.:"..bom?. and questions about feelings and private mterpretatlons are aVOlde~. Examples mIght be Where were you . irritated. 0 f Cri 't' It . r sadness. WII~out such signals most interviews quickly dry up.

how it felt. but not necessarily. back to something already stated. and often. nor. "What do you think about this?. that is public. especially where interviewees are just checkmg to make sure that anthropologist really meant to he so v a g u e . .. may ask something like "Just exactly what about my childhood do you want to know?" Usually just repeating the question or slightly reph". to encourage the respondent further to elaborate on a particular topic by referring . the respondent will try to deal with the interview in the mode of two-person conversa· tion familiar to him or her. Respondents often pose polite questions to the interviewer such as . infonnation. a dialogue. "Tell me about your mother" is quite open. Respondents. but there are stili significant respondent choices. properly speaking. the interviewee must choose how to respond out of a large permissible universe of responses. etc. "What color hair did she have?" is still narrower and a function of the interviewee's descriptive competence. "A while ago you menlioned X.or herself. I l i -. Within the investigative arena of "mother" (all other areas having been for the moment closed oft). .:sing it will produce a response. on a continuum. and the interviewer must tactfully deflect this into interview form. .J . and the interviewer may want not to bring it up him. ' The interview is not a debate. . As the introduction t~ t~is secti?n n~tes. the interviewer listens and signals ' that he or she is listening. Could you tell me more about what that was like. at least until the interview series is over. Shy or fearful respondents who worry that they have nothing more of interest to say (or fear the interviewer is not hearing them properly) can be encouraged to go on by the interviewer asking." Insofar as the respondent's question was dictated by politeness rather than curiosity. . The openness or closedness of intervention~ is relative. While the discussion of an area is going on. Closed interventions narrow the option$ for response.350 LEVY / HOLLAN Person-Centered Interviewing and Observation 351 placed as to guide the direction of the interview or to encourage the continuati of some topic. "What did she look like?" is narrower. At first. This attentive listening not only helps the interviewer' determine proper interventions. anxious about the unfamiliar mode of questioning or unsure of the intent or linguistic competence of the anthropologist.?" 1. And it enables the interviewer ." or "How do you folks do this?" Such questions can be temporarily deflected by a response like "Let's talk about this after the inter-' view. The ability to listen attentively and accurately is critical to the interviewing process. it also communicates to the respondent that the interviewer is interested in what he or she has to say. ' Interviewees are often at least mildly uncomfortable with very open-ended ques: lions that tend to violate the expectations of normal dialogue. the ideal purpose of an op mterventlOo (open m that It gives the mtervIewee maximum options in his or h response) is to encourage the interviewee to form the response as far as possible' terms of his or her personal organization. the interviewee now acting presumably mostly as an informant. he or she will not bring it up after the interview. shift the response in the direction 0 objective. l' .

a question full of lost t ' soml .' because they are iconic signs (in C. The written notes should ces of communication. The depth. and honesty of respondent r~sponses tends to grow over the course of a series of interviews. . They should be recorded as briefly as possible. :of course.. to grasp." "'. more intImate modes of response. S. . and where. ·s of body states. It thus takes a number of interview~ to discern patterns in the respondent's responses. '-nnrtll1'f !rm('r~r interviews.lWare of linguistic and interviewing mistakes. ironic) 3] on lS mvo/ved in the statement-rage. vocabu:. In our experience. This is particularly problematic for the anthropologist who needs to be at his or her most alert" In our experience. and comments on the . in a sense.. a SlOWing of . on the other hand. depending one one's goals. These mallers should be noted .B. as he or she returns with different perspectives to major. clarified In discussions with local people. for example. grammar. However.. Although audio '. Writing may interfere with "ying attention and it sends signals (often wrong ones) to the respo~dent t~at wbatever he and she is doing at the moment of note taking Is more mterestmg ihan non-noted behaviors. amount of the kinds of information that allow interpretation of respondent. Tape Recording •'. This can skew the interview. In contrast to an informant' responses. a residue of I I of the tongue in which som.. mcompletions. Tape recordings of welkonducted interviews contain an enonnous .Illat respondents generally may find them to be less distracting and obtrusive. if necessary. This can be accomplished comparatively quickly.ymg the mfOrmation. although respondents' limited availability often requires that the .'''} "or 2 mtervlews may be sufficient.. These "DOtes may include: (I) important visuat aspects ofthe respondent's behavior (which. and intervention phrasing can be spoiled. '" a declaration. Most notes about interviews will come from recall immediately after the interview and from the study . S . something between 6 (if goals or time are limited) and 20 or more interviews may be reasonable. that the respondent will become annoyed. The question of the length of individual interviews is simpler-it is a question of how long the participants can pay close altention to each other. the generality of some form among some class of respondentS).i:. :::/ nena. It is best to keep note taking to a minimum. 1~ .'. a strong emphasis. For special purposes (to check..iary. a list of new topics and ~. seen called "paralinguistic" ph . seeme< tIC sense of deity. . notes which can then be expanded in writing after the interview. anthropologist has chosen to cover.':. There are so many variables determining the length of the interview series that it is not possible to give more than very rough numbers..the moment.a respondent do not get used up so quickly.those of. an impression of the : significant behavior of the respondent. respondent. comments on the interactions between inter- ".statement.e. :."< 9. which overlaps in part lid' It ha b Isn .ory. which indicate "th ~ as the kind of utteran~e it is toe ~ re ways that we convey (in allia a verbal . the ideal and culturally an( gomg by too rapidly. that the interviews will become sterile. a~ o~en clarified by their location' e mdlcated by some convention o~ 4. they only begin to become unmformatlvely redundant and undemanding offurther contextualizingafte a large number of responses have been observed. particularly in the early stages of doing 'interviews. These paralingUistic co. wrillen notes should be made. almost any verbal and bodily behavior of the respondeut may prove informative.of tape recordings. .er and Length ofInterviews How many interviews are necessary in the study of a respondent? How la should each interview take? The distinction between informant and respondent is relevant in approach these questions. to react to a set ofquestions an probes. An informant is asked to give systematically a set of verbal descri tions of local phenomena. Note Taking After the Interview As soon as possible after each interview. or abrupt SWitch and dynamics of organization .'n. . . Problems in pronunciation.. These If. it may be necessary to note certam mallers bnefly for mnemomc 'purposes. is asked. The respondent's personal world becomes more and more clarified as he or she comes to trust the interviewer and rapport deepens.: '~\! attend to all the malters that the interviewer thinks might be missed or ohscured in the details of the tape recordings. 1 'ption is "slips of the tongue" \ an Village informant calling' th ead of "he" (which she sUbseque~ s!'e heard it on the tape).L L i L 352 LEVY / HOLLAN Person-Centered Interviewing and Observation 353 Num. gUistic Phenomena eal communication may be noted ~ lakes ofthe interviewer's probes an: le p. uncertain. interviewer interactions and of the respondent's personal world. Tbeyare . Recording and Transcribing the Interviews Note Taking During the Interview 'nlered Interviewing and Obsen L L L L L L L L L L L It is the shifting play of these reactions that provide the materials for a growin understanding of the respondent's private world. Finally. we believe . what seems to be happening in the interview at . and it is often possible in liste : VOice. and tape ptl~ns are of various kinds. the respondent usually regrets having the series discontinued. pnvate concerns. . and as public role behavior begins to shift to private. (2) significant events and . Pitch. and (3) guesses about some ': potentially mum inating and organizing hypothesis. as we have noted. comments on visual behaviors. The most frequendy used recording device for respondent interviews is tape Rcording. as he or she gets some sense of what these ~eculi~r interviews are about. significance. Rcordings obviously cannot capture the visual data of video recordings. should include a list of the topics covered in the interview. it is difficult to be properly alert in a interview of more than an hour or so. The length of the interview series is an important technical tool in itself. would not be picked up by the tape recorder). it will also take certain number of interviews to cover the list of the target topics that th. erba! or phonemic level . for ~xam~le. interviewer's reactions to various parts of the interview.robes and questions are correct )se mterventions that are ignored n the respondent's usual responses ~ar f:~t~/ . . Listening to and transcribing tapes.• along with their context-that is.questions suggested by the interview for later follow-up.'· ~. This is not what happens in welk conducted interviews where understanding deepens. these paralingUistic markers are uniVE .'shifts in the interviewer's internal experience. are excellent ways to improve language competence and to become.-:viewer and respondent. And although only the coher discursive statements of an informant are useful. and. Inexperienced interviewers fear that they will run out of questions. etc. sincere.

Douglas W. Merton. Research Methods in Cultural Anthropology. History. The Ch' ! U I' Thomas B . Emotion. In Selves in Time and Place: Identilies. In Spmts In Culture.. ~alph R: 1967: The Technique and Practice oJ Psychoanalysis. 79-173. The Quest for Mind in Different Times and Different Places. 1994. ":. Interviewing. by Margaret M. Anthropology. Robert 1. Differences in Parental Ideas about Learners and Teaching in Tahiti and Nepal. 213-235.Person-Centered Interviewing and Observation 363 362 LEVY / HOLLAN REFERENCES Iloll • Bernard. Pp. Robert I. 1992. Raymond L. New York: W. Psychological Issues 1(1). CA: Sage Publicatioos. 1956. Honolulu: University of HawaII Press. Denzin. Intimate Communications: Erotlcs and the Study of Culture.. New York: McGraw Hill. 1996b. In The Psychology of Cultural Experience. Mathews and C. New York: Family Servi ~ AsSOCiation of America. 1990. I -. eds. . Approaches. 'iauckhohn.. Interviewing in Social Research. ' ed. Russell. Gilbert.n~ity and the Life Cycle. .)(racket W. owar . erkeley: University of California Press. Annetle Marie. Cobb. Levy. Hollan.. Pre~~ Csordas. Erikson. Pp. 19S0. . San Francisco: Jossey·Bass. New York: l I Bradbum~ No~an M:. eds. Douglas. Qualitative Research Bemard. Raymond L. Yvonna S. Spirits and History: A Theoretical Perspective. Cannell. Robert 1. 1975. Andrew Barnes and Peter Stearns. L. Ashley Montagu.Levy. MD: Rowman and Littlefield. ilolstein. Pp 214-237. 2d ed. Gods.. Herhert H. Douglas W. a)'lllan. New York: Guilford Publications. Grant David. 1986. and Tac/ics rev. Clyde. 1994. 11-28. Thousand Oaks C Sage PubhcatlOns. Chicago: >_ University of Chicago Press. In Social History and Issues in Human (:"onsciousness. Developments in Person-Centered Ethnography. Gubrium. eds. The Dynamics of Interviewing. New York' International Untvers1tles Press. Knowing and Culture. W. Beverly Hills. In Culture Theory: Essays on Mind. 1995. Gottschalk et aI. Learning How To Ask: A Sociolinguistic Appraisal ofthe R / '. Sacred Self A Cultural Phenomenology o·r 'J ansmal. 1996. In press. Erik. CA: Sage Publications. University of Chicago Press... James A. . Walnut Creek. Lincoln. 1957. Mesocosm: The Organization of a Hindu Newar City in Nepal.: r~v.Wiley. 1973. 18G-187. person-Centered Anthropology. eds. Glencoe. 1978. Kahn. From Anxiety /0 Method. Norman K. Self.c Ilea mg. 1996a. Fontana. Methods Series. Levy. . 37. ed. McCracken. C Gorden. eds. Kendall. 1988. H. Erikson. Pp. The Active Interview. H. New York: Columbia University Press. New York: Columbia University Press. <. 1988. Experience and History in Nepal. 1977. Essential Contrasts. Zaki. I' . Robert I.. IL' F E P k Publishers. Levy. In Spirits in Culture and Mind.. 1984. In The Socia/ization of Aggression. . T. Qualitative Research Methods Series. l ~ ! ~ ~ "lI l l l l l . and E1inor P. 1967. WelJenkamp. Robert 1. Marjorie Fisk-e. Garretl. Lanham. Hollan. New York: Humanities Press. 5. Robert"l.e s. Interviewing: Strategy. Thousand Oaks. and Sociology. Gorden. Pp. CA: Sage Publications. LeVine. 1959: Ide. 1L: Dorsey Press. . Frey. eacoc Greeoson. Charles L. New York: W. Robert K. Creative Interviewing. 1968. Mangold and E1inor P. Russdl. Pp. 1975. ll!l. Tahitian Gentleness and Redundant Controls. In The Use o/Personal Documents in History. The Focused Interview: A Manual of Problems and Procedures. Stoller. the Interview in Social Science Research. "::LeVY. Norton : Erikson. Tahitians: Mind and Experience in the Society Islands. 222-235. leanette Mageo and Alan H d d Pp. New York: International UniverSIties Press. . D~ugl~ W. ed. Herdt. Dehra Skinner. eds. eds. Gcorge. Ques/lonnQlre DesIgn. W. Richard A. The Long Interview. I I . CA: Sage Publications. 1994. 1979. 13. New Science Research Council Bulletin 53. Levy. 1982. and Patricia L." of Quolt/atlve Research. H. pp. . Contentment and Suffering: Culture and Experrence m ToraJa. Childhood and Soctety. Pp. In Assessing Cultural Anthropology. Chicago: . Vol. l l l l 1l -. . and Dorothy Holland. 3-40... 1990. Norton. Robert Borowsky. CA: AltaMira Press. 1994. In Handbool. and Robert J. Wellenkamp. ~dr~a. Erik. Improving Interview Method r ~ I Briggs. Zaki. 361-376: Thousand Oaks. Chicago: University of Chicago Press. ' Dcvcreux. Berkeley: University of California Press. Interviewing: The Art of Science. and Emotion. and Seymour Sudman. New York: Cambridge University 0. Vol. and laber F. 1945. Research Methods in Anthropology: Qualitative and Quantitat -:. Cambridge: Cambridge University Press. and William J.• an~ Jane C. Selves in Motion. Rohert I.Douglas W. 1989. -. Hollan. Oxford: Oxford University Press. Techniques. Identity: Youth and Crisis. Force and Persuasion: Leadership in an Amazonian Society. on the Life Cycle. London and New York: Routledge. Its Principles and Methods 3d ed. and James H. Jack D. and Jane C.he Thread oJLife: Toraja Reflections ~:. The Personal Document in Anthropological Science. Robert 1. IL: Free Press. E~ik. 1994. In Parents' Cultural Belief Systems. Jeannette Mageo and Alan Howard. Moore eds New Y rk' . Basle Interviewing Skills ltasca. Robert L. 123-142. In preparation. 0 • Cambridge University Press. HOlllewood.. Thousand Oaks. Levy. Sarah Harkness and Charles Super. London and New York: Routiedge.. 1985. Cultural and Experiential Aspects of Spirit ·Beliefs amoog the ToraJa. 1996. Shweder and Robert A. and Charles F. and Mind. J. New York: New York University Press. AI Pach. Robert 1.

Moral Knowing in a Hindu Sacred City. New York: Houghton Miffiin. help get more details than "WhY/When I would you go to a folk healer?" lented with items from other sources. 1951. I focus here on developing interview materials for studie.) uestions. 1984. W. Carl. the more appropriate are unstructured. n uman. rgamza/lon seare. Learning/ram the Field: A Guide .-. Thousand s. Barbara S. and Irene S. Me~usa 's Hair: An Essay on Personal Symbols andReli ExpeYlence. CA: Sage Publications. Quolitative Interviewing: The Art ofHearing Data.rview purpose. descriplive information may be used to design a detailed study of cultural models. ~pe of information desired. The chapter is organized by inte. In general. the iess that is known about an area. Systematic Fieldwork.mani-based information.. Here. C. Illness Terms L Structured Interviewing and Questionnaire Construction Frequency 5 Disease Term 5 4 4 4 3 3 3 Scarlet Fever VenercaJ Disease Arthritis Migraine L L L C L s Whooping Cough Diphtheria Headache Hepatitis Mental Illness Mononucleosis Rubella Smallpox Strep Throat Stroke Rubin. is conlrasting questions. rmants asked about their distinguishing study health care choices.-------.L L L L 364 LEVY / HOLLAN SUSAN C. open-ended methods..thod alone. Chicago: Umverslty of Chicago Press. Stanley L. 1996.the. or beliefs. M.'--' ' . WELLER • Obeyese~ere.. The First Five Minutes: A Sample of MIcroscopIc IntervIew Analysis. Richardson. Stanley A. A combi!lation of initial descriptive exploration and subsequent systematicinterv'1ew{Ilg prOduces'a study ~fior to one based on eilher me. MarkSchoeptle.C i b' . standardized scales) or to check their validity. The Ethnographic Interview New York' Hol' R' hart d Winston. . . Ithaca: Paul Martineau.'ihat ~~d{recnnl'o. Dnhrenwend. In cross-cultural psychology.sons that might influence a woman to Jngy asked multiple questions of each aspects of each feeding method. I describe different approaches to interviewing and