writing for culture

:ULWLQJIRU&XOWXUH:K\D6XFFHVVIXO&RQFHSW6KRXOG1RW%H'LVFDUGHG$XWKRUV&KULVWRSKb%UXPDQQ6RXUFH&XUUHQW$QWKURSRORJ\9RO1R66SHFLDO,VVXH&XOWXUH$6HFRQG&KDQFH")HEUXDU\SS663XEOLVKHGE\7KH8QLYHUVLW\RI&KLFDJR3UHVVRQEHKDOIRI:HQQHU*UHQ)RXQGDWLRQIRU$QWKURSRORJLFDO5HVHDUFK6WDEOH85/http://www.jstor.org/stable/10.1086/200058 .$FFHVVHG

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press and Wenner-Gren Foundation for Anthropological Research are collaboratingwith JSTOR to digitize, preserve and extend access to Current Anthropology.

http://www.jstor.org

This content downloaded from 128.118.88.48 on Mon, 16 Jun 2014 09:58:37 AMAll use subject to JSTOR Terms and Conditions

Current Anthropology Volume 40, Supplement, February 1999 1999 by The Wenner-Gren Foundation for Anthropological Research. All rights reserved 0011-3204/99/40supp-0003$3.00

There are times when we still need to be able to speakholistically of Japanese or Trobriand or Moroccan cul-ture in the confidence that we are designating some-Writing for Culturething real and differentially coherent.

james clifford, The Predicament of Culture

Why a Successful ConceptFor a long time, defining cultural/social anthropologyShould Not Be Discarded1 as the study of the cultural dimension of humans wouldhave raised few objections among the disciplines prac-titioners. Now the place of culture within that defini-tion is considerably less certain. Within the past decadeby Christoph Brumannor so there has developed what Sahlins calls the fash-ionable idea that there is nothing usefully called a cul-ture (1994:386), and one prominent voice even advo-cates writing against culture (Abu-Lughod 1991),

In the past decade, the idea that speaking of a culture inevitably giving a name to a whole writing against culturesuggests an inordinate degree of boundedness, homogeneity, co- movement in Fernandezs observation (1994:161). Al-herence, and stability has gained considerable support, and somethough the scepticism over the culture concept has itscultural/social anthropologists have even called for abandoning

the concept. It is argued here, however, that the unwelcome con- origins in deconstructionist and poststructuralistnotations are not inherent in the concept but associated with cer- thought, anthropologists sympathizing with it cometain usages that have been less standardized than these critics as- from an amazing range of theoretical positions thatsume. The root of the confusion is the distribution of learned

reaches far beyond that specific vantage point. It will beroutines across individuals: while these routines are never per-fectly shared, they are not randomly distributed. Therefore, cul- worth while to document this disciplinary discourse atture should be retained as a convenient term for designating the some length before contrasting it with standard anthro-clusters of common concepts, emotions, and practices that arise pological formulations of culture. It turns out that whatwhen people interact regularly. Furthermore, outside anthropol- is being addressed by the critics is certain usages of theogy and academia the word is gaining popularity and increasingly

culture concept rather than the concept itself, and I ar-understood in a roughly anthropological way. Retaining the con-cept while clarifying that culture is not reproduced unproblemati- gue that it is possibleand not very difficultto disen-cally, has its limits in the individual and the universal, and is tangle the concept from such misapplications and tonot synonymous with ethnicity and identity will preserve the find historical precedents for this in anthropology. In acommon ground the concept has created within the discipline.

next step I will address what I consider to be the rootMoreover, it will simplify communicating anthropological ideasto the general public and thus challenging mistaken assump- of the confusion, namely, the fact that the sharing oftions. learned traits among humans is never perfect, and how

this can be dealt with. Finally, I will present pragmaticchristoph brumann has been a lecturer at the Institute of Eth- reasons for retaining culture and also cultures: thenology of the University of Cologne (50923 Koln, Germany) concept has been successful, and other scientific disci-since 1992 and is currently conducting ethnographic fieldwork plines as well as the general public increasingly employon preservation-related conflicts and urban identity in Kyoto, Ja-

it in a way we should not be entirely unhappy about.pan ([email protected]). Born in 1962, he receivedhis Dr. phil. in 1997 from the University of Cologne. Among his Some of these uses are certainly problematic, but re-publications are Strong Leaders: The Charismatic Founders of taining the concept and the common ground it has cre-Japanese Utopian Communities, in Leaders and Leadership in ated will bring us into a better position to challengeJapan, edited by Ian Neary (Richmond: Japan Library, 1996), Die

them.Kunst des Teilens: Eine vergleichende Studie zu den Uberle-bensbedingungen kommunitarer Gruppen (Hamburg: Lit, 1998),and Materialistic Culture: The Uses of Money in Tokyo Gift Ex-changes, in Consumption and Material Culture in Contempo-rary Japan, edited by John Clammer and Michael Ashkenazi (Lon- The Critiquedon: Kegan Paul International, in press). The present paper wassubmitted 13 i 98 and accepted 3 ii 98.

The major concern of the sceptical discourse on cultureis that the concept suggests boundedness, homogeneity,coherence, stability, and structure whereas social real-ity is characterized by variability, inconsistencies, con-flict, change, and individual agency:

1. In writing this article, I have benefited from discussions of previ-The noun culture appears to privilege the sort ofous versions with Hartmut Lang, Thomas Widlok, and particularly

Thomas Schweizer, from the remarks of participants in a presenta- sharing, agreeing, and bounding that fly in the facetion to Schweizers research seminar, and from the valuable com- of the facts of unequal knowledge and the differen-ments of the editor and two anonymous referees. I am grateful for tial prestige of lifestyles, and to discourage attentionall this advice, although responsibility for any shortcomings of my

to the worldviews and agency of those who are mar-argumentswhich did not find favour with all those mentionedrests with me alone. ginalized or dominated. [Appadurai 1996:12]

S1


S2 current anthropology Volume 40, Supplement, February 1999

The classic vision of unique cultural patterns . . . specificity comes to be and is reproduced. [Friedman1994:207]emphasizes shared patterns at the expense of pro-

cesses of change and internal inconsistencies, con- Despite its anti-essentialist intent . . . the cultureflicts, and contradictions. . . . From the classic per- concept retains some of the tendencies to freeze dif-spective, cultural borderlands appear to be annoying ference possessed by concepts like race. [Abu-exceptions rather than central areas for inquiry. . . . Lughod 1991:144]The broad rule of thumb under classic norms . . .seems to have been that if its moving it isnt cul- As a result, the differences between the anthropolo-tural. [Rosaldo 1993[1989]: 2728, 209] gist and the people under study are exaggerated, and the

latter are placed in a subordinate position. This in-Culture . . . orders phenomena in ways that privi- creases the distance between the two parties to the eth-lege the coherent, balanced, and authentic aspects nographic encounter while enhancing the anthropolo-of shared life. . . . Culture is enduring, traditional, gists privileged position as the expert and translatorstructural (rather than contingent, syncretic, histori- or even the very creatorof such utter strangeness:cal). Culture is a process of ordering, not of disrup-tion. [Clifford 1988:232, 235] Culture operates in anthropological discourse to

enforce separations that invariably carry a sense ofThe most-dangerously misleading quality of the no- hierarchy . . . it could be . . . argued that culture istion of culture is that it literally flattens out the ex- important to anthropology because the anthropologi-tremely varied ways in which the production of cal distinction between self and other rests on it.meaning occurs in the contested field of social exis- Culture is the essential tool for making other. As atence. [Friedman 1994:207] professional discourse that elaborates on the mean-

ing of culture to account for, explain, and under-Applied in this way, culturea mere anthropologi-stand cultural difference, anthropology also helpscal abstraction (Borofsky 1994b:245)is transformedconstruct, produce, and maintain it. Anthropologi-into a thing, an essence, or even a living being or some-cal discourse gives cultural difference (and the sepa-thing developing like a living being:ration between groups of people it implies) the airof the self-evident. . . . It would be worth thinkingA culture had a history, but it was the kind of his-about the implications of the high stakes anthropol-tory coral reefs have: the cumulated accretion ofogy has in sustaining and perpetuating a belief inminute deposits, essentially unknowable, and irrele-the existence of cultures that are identifiable as dis-vant to the shapes they form . . . our conception ofcrete, different, and separate from our own. [Abu-culture almost irresistibly leads us into reificationLughod 1991:13738, 143, 146]and essentialism. [Keesing 1994:301, 302]

In effect, the concept of culture operates as a dis-Much of the problem with the noun form [of cul-tancing device, setting up a radical disjunction be-ture] has to do with its implication that culture istween ourselves, rational observers of the humansome kind of object, thing, or substance, whethercondition, and those other people, enmeshed inphysical or metaphysical. [Appadurai 1996:12]their traditional patterns of belief and practice,

Culture . . . consists in transforming difference into whom we profess to observe and study. [Ingoldessence. Culture . . . generates an essentialization of 1993:212]the world . . . [Friedman 1994:206, 207]

In global terms the culturalization of the world isabout how a certain group of professionals locatedA powerful structure of feeling continues to see cul-at central positions identify the larger world and or-ture, wherever it is found, as a coherent body thatder it according to a central scheme of things. [Fried-lives and dies. . . . It changes and develops like a liv-man 1994:208]ing organism. [Clifford 1988:235]

The essentialism of our discourse is not only inher-This brings the concept of culture uncomfortablyent in our conceptualizations of culture, but it re-close to ideas such as race that originally it did a greatflects as well our vested disciplinary interests indeal to transcend:characterizing exotic otherness. [Keesing 1994:303]

Viewed as a physical substance, culture begins to Proceeding from the diagnosis to the cure, a numbersmack of any variety of biologisms, including race, of writers suggest that a simple grammatical shift mightwhich we have certainly outgrown as scientific cate- help:gories. [Appadurai 1996:12]A view of the cultural (I avoid culture deliber-Where difference can be attributed to demarcated ately here, to avoid reification as best I can) . . .populations we have culture or cultures. From here [Keesing 1994:309]it is easy enough to convert difference into essence,

race, text, paradigm, code, structure, without ever I find myself frequently troubled by the word cul-ture as a noun but centrally attached to the adjecti-needing to examine the actual process by which


brumann Writing for Culture S3

val form of the word, that is, cultural. . . . If cul- agree on very many other issues.2 I am convinced thatmost readers of this article could easily furnish similarture as a noun seems to carry associations with

some sort of substance in ways that appear to con- references from equally diverse sources. A profounddoubt about the validity of the culture concept, justifiedceal more than they reveal, cultural the adjective

moves one into a realm of differences, contrasts, in terms of the many misleading associations it is pre-sumed to carry, has undoubtedly become an importantand comparisons that is more helpful. [Appadurai

1996:12] trope in current anthropological discourse.

Nationalists were themselves using what lookedvery like anthropological arguments about culture. Historical and Optimal Usage. . One possible escape from this dilemma mightbe to abandon talk of different cultures alto- There is no denying that anthropologists in their ethno-gether, because of its taint of essentialism, but to re- graphic and theoretical work have committed the afore-tain some use of the adjectival cultural. [Barnard mentioned sins in abundance, but I am not convincedand Spencer 1996a:142] that they have done so because of the culture concept.

To demonstrate this, I will turn to anthropologicalFollowing Keesing . . . I use the term culturaldefinitions of culture, since the conception of that termrather than culture. The adjectival form down-ought to be most clearly expressed there. Modern text-plays culture as some innate essence, as some liv-books define culture as follows:ing, material thing. [Borofsky 1994b:245]

A culture is the total socially acquired life-way orReformulating culture: return to the verb. [Friedmanlife-style of a group of people. It consists of the pat-1994:206; the verb is not given]terned, repetitive ways of thinking, feeling, and act-

Further, despite Moores belief that even if one ing that are characteristic of the members of a par-wanted to, it would be impossible to trash the culture ticular society or segment of a society. [Harris 1975:concept because it is so deeply rooted in the history of 144]ideas and in the discipline of anthropology (1994:373),

Culture . . . refers . . . to learned, accumulated expe-some writers go so far as to envision an anthropologyrience. A culture . . . refers to those socially trans-without it, albeit not in very strong terms:mitted patterns for behavior characteristic of a par-

It may be true that the culture concept has served ticular social group. [Keesing 1981:68]its time. [Clifford 1988:274]

Culture is the socially transmitted knowledge andbehavior shared by some group of people. [PeoplesWe need to be fully conscious of the varying bound-and Bailey 1994:23]aries, not so much of a culture but of cultural prac-

tices. A recognition of these features may make usHere and in other textbook definitions, no mentionwary of simplistic notions of cultural homogene-

is made of boundaries, universal sharing, immunity toity. . . . It may indeed make us wary of . . . even us-change, or cultures being a thing, an essence, or a livinging the term cultural altogether. [Goody 1994:255]being. Since the negative tendencies identified by the

In its application, the concept of culture fragments culture sceptics are ascribed to a classic perspective,the experiential continuity of being-in-the-world, iso- however, one might expect them to be more present inlating people both from the non-human environ- older formulations. Here are some well-known exam-ment (now conceived as nature) and from one an- ples:other. . . . Would it not be preferable to move in the

Culture, or civilization, . . . is that complex wholeopposite direction, to recover that foundational conti-which includes knowledge, belief, art, law, morals,nuity, and from that basis to challenge the hegem-custom, and any other capabilities and habits ac-ony of an alienating discourse? If so, then the con-quired by man as a member of society. [Tylor 1871:cept of culture, as a key term of that discourse, will1; Kroeber and Kluckhohn 1952:81]have to go. [Ingold 1993:230]Culture embraces all the manifestations of socialPerhaps we would do best if we stopped privileging habits of a community, the reactions of the individ-the representation of culture, and instead focused ual as affected by the habits of the group in whichon the level of events, acts, people, and processes. he lives, and the products of human activities as de-[Barth 1994:358] termined by these habits. [Boas 1930:79; Kroeberand Kluckhohn 1952:82]Perhaps anthropologists should consider strategies

for writing against culture. [Abu-Lughod 1991:147] The culture of any society consists of the sum totalof ideas, conditioned emotional responses, and pat-In assembling the above collage I do not want to sug-

gest that each of the quoted writers supports each of theideas expressed. Nonetheless, there is a surprising de- 2. For a more sustained and systematic analysis of the presumed

defects of the culture concept, see Brightman (1995).gree of common ground among scholars who would not



terns of habitual behavior which the members of fiable elements and use a noun followed by of and anenumeration of the elements to define it, as in thethat society have acquired through instruction or

imitation and which they share to a greater or less above integral whole of or sum total of. Clearly,this additive notion of culture is only one way of con-degree. [Linton 1936:288; Kroeber and Kluckhohn

1952:82] ceiving it. But while the nouns used for this purposetend toward either the abstract (e.g., summation,Culture means the whole complex of traditional be- set, system, class, organization) or the con-havior which has been developed by the human race crete (e.g., mass, pattern, body, total equip-and is successively learned by each generation. A ment), even the latter ones are almost invariably em-culture is less precise. It can mean the forms of tra- ployed in a clearly metaphorical way (which, it shouldditional behavior which are characteristic of a given be added, is hardly unconventional even with a wordsociety, or of a group of societies, or of a certain such as bodywho has ever touched a body of evi-race, or of a certain area, or of a certain period of dence?) Some of the older formulations are indeed sus-time. [Mead 1937:17; Kroeber and Kluckhohn 1952: pect of conceptual animism, as when Kroeber and90] Kluckhohn themselves muse on how the fate of a cul-ture depends on the fate of the society which bears itExcept for the occasional use of an outmoded word(Kroeber and Kluckhohn 1952:16566), or when Rich-(such as race or, arguably, civilization) and for maleard Thurnwald defines culture as the totality of usagesbias, these definitions do not deviate fundamentallyand adjustments which relate to family, political forma-from the modern ones. Incidentally, most anthropologi-tion, economy, labor, morality, custom, law, and wayscal textbooks and encyclopedias I consulted quote Ty-of thought. These are bound to the life of the social enti-lors formulation, invariably with extensive commentsties in which they are practiced and perish with these(Harris 1975:144; Keesing 1981:68; Kottak 1982:6; Peo-. . . (Thurnwald 1950:104 as translated in Kroeber andples and Bailey 1994:21) but often without giving anyKluckhohn 1952:84). Apart from these exceptions, how-alternative definition (Barnard and Spencer 1996a:137;ever, the most reifying and essentializing definitions inGoodenough 1996:291; Seymour-Smith 1986:67). Thisthe Kroeber and Kluckhohn collection do not comemakes me wonder (pace Brightman 1995:527) whetherfrom social/cultural anthropologists.4 And when Lesliethere really is a significant gap between what modernWhite characterizes culture as an elaborate mecha-and classic social/cultural anthropologists take to benism . . . in the struggle for existence or survival (1949:the core meaning of the word culture; rather, they363; Kroeber and Kluckhohn 1952:137), he puts thisseem to have different theories about the same thing.into a context in which it is obvious that no more thanAnd what applies to modern definitions of culture ap-a metaphor is implied.plies to the older ones as well: in the above quotationsAt the same time, however, one also comes acrossas well as in the other anthropological definitions in

formulations such as the following:Kroeber and Kluckhohns famous collection (1952),there is none which explicitly denies that a culture has We can observe the acts of behaviour of . . . individ-clear boundaries, is homogeneous, does not change, or uals, including . . . their acts of speech, and the ma-is a thing or an organism. I find it significant, however, terial products of past actions. We do not observe athat none of them unambiguously says so either, leav- culture, since that word denotes, not any con-ing these aspects open for investigation instead. One crete reality, but an abstraction . . . [Radcliffe-Brownmight argue that many of the definitions postulate dis- 1940:2; Kroeber and Kluckhohn 1952:253]crete cultures by attributing a culture to a specificgroup,society, or area, but none of them says Culture is essentially a construct that describes the

total body of belief, behavior, knowledge, sanctions,that these units are always clearly bounded or that theymust be so to have a culture attributed to them. Most values, and goals that mark the way of life of any

people. That is, though a culture may be treated bydefinitions are also mute on the evenness of distribu-tion required for delimiting a culture. The few that the student as capable of objective description, in

the final analysis it comprises the things that peo-mention it, however, speak of a greater or less degreeof sharing (Linton above) or even of every individuals ple have, the things they do, and what they think.

[Herskovits 1948:154; Kroeber and Kluckhohn 1952:being a representative of at least one subculture (Sapir1949:51516).3 84]The majority of the definitions in Kroeber and Kluck- A culture is invariably an artificial unit segregatedhohns volume see culture as a set consisting of identi- for purposes of expediency. . . . There is only one

natural unit for the ethnologistthe culture of all3. Every individual is . . . in a very real sense, a representative of atleast one subculture which may be abstracted from the generalizedculture of the group of which he is a member. Frequently, if not 4. These speak of culture as a stream (Blumenthal 1937:12; Ford

1949:38; Kroeber and Kluckhohn 1952:130, 132), an embodimenttypically, he is a representative of more than one subculture, andthe degree to which the socialized behavior of any given individual (LaPiere 1949:68; Kroeber and Kluckhohn 1952:112), or a self-gen-

erating . . . pattern-creating order (Panunzio 1939:106; Kroeber andcan be identified with or abstracted from the typical or generalizedculture of a single group varies enormously from person to person Kluckhohn 1952:106). Even here, however, I assume metaphorical

intentions.(Sapir 1949:51516; Kroeber and Kluckhohn 1952:247)



humanity at all periods in all places . . . [Robert Lo- talistic conception (Brightman 1995:51213; Kuper1994:54041).6wie, as quoted in Kroeber and Kluckhohn 1952:165]Especially for exaggerated assumptions of bounded-

It is difficult to attribute essentialism, reification, or or- ness and homogeneity, however, I believe that responsi-ganicism to these statements or to similar ones by Mur- bility cannot be simply deflected onto particular theo-dock or Sapir.5 It rather seems that, at least on a general retical approaches. Rather, a number of rarely discussedlevel, a good number of typical representatives of the but powerful assumptions implicit in traditions of eth-classic perspective were no less aware of these dan- nographic writing, traditions that are much older thangers than are todays culture sceptics. Here I agree with the discipline of anthropology, must also be accused.Brightman that an expendable straw culture is . . . be- These assumptions include the existence of a mosaic ofing retrospectively devised (1995:528) by the critics as territorially bounded, discrete cultures of which thea selectiveand itself rather essentialistconstruct world supposedly consists; the irrelevance of intra-andthat excludes those disciplinary traditions that are more interindividual variation, the timelessness of the cul-in line with current theoretical concerns (pp. 52728). tures under study (which either have no history or haveWhere, then, are the unwelcome aspects associated acquired one only by coming into contact with colonial-

with the culture concept presumed to have originated? ism), and the superiority of precontact cultures as an ob-Kuper identifies Boas as a main culprit, taking him to ject of investigation. In much classic ethnographic us-task for importing and bequeathing to his students a no- age, a culture was simply understood as synonymoustion that was heavily influenced by Herders idea of the with what formerly had been called a people, and theVolksgeistthe spirit of a people presumed to be inher- units so designated were taken as the natural, internallyent in all of its material and mental creations (Kuper undifferentiated, and unproblematic reference units for1994:539). Fox, however, emphasizes that Boas himself description just as they had beenand continued towas not consistent and that his followers were divided bein most pre- and nonanthropological ethnography.about the coherence of cultures, some of them (includ- As a consequence, many portraits of Japanese or Tro-ing Kroeber, Benedict, and Mead) holding to a highly in- briand or Moroccan culture are indeed marred by thetegrated notion while others (notably Lowie and Radin) shortcomings deplored by the critics. And when Mali-spoke of shreds and patches (Fox 1991:102) or concen- nowski meritoriously reminds his readers of the natu-trated their research on diversity and individuals (1991: ral, impulsive code of conduct, the evasions, the com-1016; see also Brightman 1995:53034). Overall, it ap- promises and non-legal usages of the individualpears to me that the former perspective gained weight savage/Trobriander (1976[1926]:120), he is somewhatwith the synchronic turn in anthropology, the re- like a statistician who gives the average, says that thereplacement of an earlier diachronic orientation (in evolu- is variance, but does not care to calculate the standardtionism, diffusionism, historical particularism, or Kul- deviation. No doubt it would be mistaken to search forturkreislehre) with a focus on the analysis of cultural a full-fledged theory of praxis in the work of Malinow-systems at a fixed point in time (as in the culture-and- ski, Lowie, or Radin (see also Brightman 1995:540).personality school, structural-functionalism, structur- Yet still, at least in their more general and theoreticalalism, and, later on, culture-as-text interpretivism). In writings, there was a clear awareness of the constructedthe latter approaches there was certainly a strong incli- nature of the culture concept among a good number ofnation to see more cultural coherence than actually ex- representatives of the classic tradition, and one ofisted. This was further exacerbated in American anthro- them even elevated allowance for variation to the sta-pology by Parsonss influential segmentation of culture tus of a central problem (Firth 1951:478). Hence, if aand society as separate fields of study, a theoretical de- disciplinary precedent is needed, anyone seeking to re-cision that discouraged what interest there was in the tain a nonreified culture concept as an expedient ab-social differentiation of culture and supported a men- straction (see below) can find it here. Definitions, as I

have tried to show, have been open in this regard any-way, and therefore I propose to hold apart the historical5. According to Murdock (1937:xi, Kroeber and Kluckhohn 1952:

251), culture is merely an abstraction from observed likeliness in usage of the conceptwhat it has been taken to meanthe behavior of individuals organized in groups. Sapir (1949:515 by many in the pastand its optimal usagewhat it6; Kroeber and Kluckhohn 1952:24748) argues that the true locus could mean if used to its best intents, so to speak.. . . of . . . processes which, when abstracted into a totality, consti-

Sometimes a scientific concept can no longer be sal-tute culture is not in a theoretical community of human beingsknown as society, for the term society is itself a cultural constructwhich is employed by individuals who stand in significant rela-tions to each other in order to help them in the interpretation of 6. By contrast, leading proponents of British social anthropology

such as Fortes, Nadel, and Firth continued to view society and cul-certain aspects of their behavior. The true locus of culture is in theinteractions of specific individuals and, on the subjective side, in ture as complementary concepts of which the significant ele-

ments phase into one another in such a way that they cannot bethe world of meanings which each one of these individuals mayunconsciously abstract for himself from his participation in these adequately studied in isolation (Firth 1951:483). According to Mur-

dock, however, British social anthropologists rather one-sidedly ne-interactions. . . . It is impossible to think of any cultural patternor set of cultural patterns which can, in the literal sense of the glected the analysis of culture for that of society, so much so that

he suggested repatriating his trans-Atlantic colleagues into sociol-word, be referred to society as such. . . . The concept of culture, asit is handled by the cultural anthropologists, is necessarily some- ogy, where they would constitute a specialized subfield (1951:471

72).thing of a statistical fiction.



vaged; for example, that of race has been proved to lematic way that, for example, a cat or a bicycle is.Rather, the term refers to an abstract aggregate, namely,be empirically unfounded, has been abused enormously,

and in subtle ways keeps getting in the way even of the prolonged copresence of a set of certain individualitems, and thus is employed not too differently fromsome physical anthropologists who use modern, non-

racist methods to assess human biodiversity (Keita and other nouns such as forest, crowd, or city. Inidentifying a culture, we have to abstract such a set ofKittles 1997). I am not convinced, however, that past

and present misapplications of the culture concept are items from observed instances of thought and behavior,selecting that which occurs repeatedly rather than thatof a comparable degree and warrant similar avoidance.which is singular. This is a mental operation that is notin principle different from, say, identifying a style in in-dividual works of art, and the same capabilities of mem-Culture as an Abstractionorizing previously perceived instances and ignoring mi-nor differences for the sake of commonalities areDiscussing the culture concept, one has to distinguish

between culture (or Culture) in a general and required of anyone who undertakes it. Since in the em-pirical world no two things are completely identical,culture/s in a specific sense (in the same way as Mead

did; see above). The former meaning refers to the gen- the result of any such operation is always contestable,and therefore one can no more prove the existence oferal potential of human individuals to share certain not

genetically inherited routines of thinking, feeling, and Japanese culture than prove that of the Gothic style.Cultures can have no natural boundaries but onlyacting with other individuals with whom they are in so-

cial contact and/or to the products of that potential. It those that people (anthropologists as well as others) givethem, and delimiting a certain set of elements as a cul-is not very clear-cut and mentioned only in few defini-

tions; besides, it seems to be derived from the second ture can therefore be only more or less persuasive, neverultimately true. Nonetheless, we may consider it ex-meaning, on which most of the definitions concentrate.

Here a culture is the set of specific learned routines pedient to go on using the concept in the same way thatwe go on speaking of art styles, forests, crowds, or cities;(and/or their material and immaterial products) that are

characteristic of a delineated group of people; some- and wemay do so in spite of the disagreement that oftenarises over whether these terms really apply to the spe-times these people are tacitly or explicitly included.

The existence of any such culture presupposes that of cific body of art works, concentration of trees, gatheringof people, or settlement that is so designated or whereother sets of routines shared by other groups of people,

thus constituting different cultures. The debate in fact precisely their boundaries are located in a given case.The core of the problem of identifying cultures can befocuses almost exclusively on this second meaning, and

I will concentrate on it accordingly. It is the act of iden- illustrated with the three diagrams in figure 1. In these,capital letters stand for individuals and numbers fortifying discrete cultures that is held to be empirically

unfounded, theoretically misleading, and morally objec- identifiable ways of thinking, feeling, and acting.8 In thetop diagram, there is perfect sharing among individualstionable by the concepts critics.7

Of course, cultures are always constructed, but they A through F regarding features 1 through 6 and amongindividuals G through L regarding features 7 throughare so not only because of being written (Clifford and

Marcus 1986) within the confines of sociohistorically 12. Identifying cultures is not difficult here: Features 1through 6 represent one culture, features 7 through 12constituted tropes and discourses but also in a more

profound sense. A cultureas the above quotation from another. Since there is perfect discreteness between thetwo groups of features as well as between the twoLowie reminds usis not simply there in the unprob-groups of individuals carrying them, this partition rep-resents the only possible way of distinguishing cultures.

7. One might object that culture is more than just the sum total In contrast, features in the middle diagram are ran-of certain features that co-occur with a certain frequency in a group domly distributed across individuals, and it is impossi-of people. This latter, additive view will perhaps appear entirelyble to make out cultures in the same unproblematicmistaken to those who see culture as a process rather than as a

static distribution of traits or who follow Bourdieu (1977 [1972]) in way or perhaps in any convincing way.assuming that the loose structures of habitus will guide peoples The problems start with a situation such as that ofeveryday improvisations but never determine them in a strict the bottom diagram. This distribution is far from ran-sense. Neither of these ideas, however, can manage without de-

dom, yet no discrete blocks can be discerned either. Onetermining distributions of features across people as a methodologi-cal starting point. Otherwise, there would be no recognition of pro- possible partition would place features 1 through 6 incessesdefined as that which causes the presence of certain one culture and features 8 through 12 in a second. Eachfeatures at one point of time and their absence at anotherin the culture, however, would then contain features that arefirst place and no way of discovering that some specific habitus is sometimes associated with features of the other. More-at work leading to creative variations on a common (i.e., tempo-rally and interindividually stable) theme in actual behavior. Wher-ever we locate culture and however dynamic we consider it, there 8. For the sake of this argument and also for the criticisms raised,

it does not matter if any of the latter are excluded or if institutionswill be no way around determining how given features are distrib-uted over a given number of persons at a given time and how this or artifacts are added. Moreover, any observable feature can be in-

cluded in such a matrix, including emic categories, ideal as wellcompares with other times and/or other persons if we are to iden-tify a culture, a cultural process, or a habitus guiding individuals as observed behavior, norms and values, and peoples cultural self-

perception and self-categorization.improvisations.



are no such clusters of habits and that the distribution1 2 3 4 5 6 7 8 9 10 11 12of cognitive, emotive, and behavioral routines amongA x x x x x x . . . . . .humans is as in the middle diagram. However, theyB x x x x x x . . . . . .seem to fear that by identifying cultures when con-C x x x x x x . . . . . .fronted with a distribution like that of the bottom dia-D x x x x x x . . . . . .gram anthropologists will invariably be misunderstoodE x x x x x x . . . . . .as implying a distribution like that of the top diagram.F x x x x x x . . . . . .Ceasing to speak of cultures, however, also entails aG . . . . . . x x x x x xcost, namely, being understood as saying that featuresH . . . . . . x x x x x xare distributed randomly, as in the middle diagram. II . . . . . . x x x x x xdoubt very much that this kind of misunderstanding isJ . . . . . . x x x x x xpreferable, since it is not borne out by the results of an-K . . . . . . x x x x x xthropological research. Moreover, it flies in the face ofL . . . . . . x x x x x xthe experience of the billions of amateur anthropolo-gists who inhabit the world, who in their everyday lives1 2 3 4 5 6 7 8 9 10 11 12continue to identify commonalities in the thought andA x x x . x . x . . x . xbehavior of different individuals and attribute these toB . . . x . . x . x . x xtheir belonging to the same family, kin group, gender,C . . x x . x x . x . . .age-group, neighborhood, class, profession, organiza-D x . x . . x x . x x . .tion, ethnic group, region, nation, etc. Of course,E . x . . x x . x . x x .they do so in an on-and-off, often semiconscious wayF x . x x x . . x . x . .thattrue to its commonsensical naturecares lessG . x . x . x . x . x x .about oversimplifications, contradictions, and incom-H x . x . x . x x . x . xpleteness than anthropologists do and often explainsI . x . x . x x . x . x .difference incorrectly, for example, in terms of geneticJ . x . x x . . x . x x .or quasi-genetic transmission. But many of theseK x . . x x . . x . x . xamateur anthropologists would be puzzled indeed ifL x x . x . . x x x . x .we tried to persuade them that what until recently wewould have advised them to call a culture (instead of,1 2 3 4 5 6 7 8 9 10 11 12for example, the way we/they do it) does not reallyA x x x x x x . . . . . .exist.B . x x x x x . . . . . .Just as there is no way of deciding whether a glass isC x . x . x x . . . . . .

half-full or half-empty, there is no ultimate solution toD x x x x x x . x . . . xthe dilemma of being misunderstood as implying eitherE x x x x . x x . . . . .perfect boundedness and homogeneity (when speakingF x x x x x . x . . . . .of cultures) or perfect randomness of distribution (whenG . . . . . . x x . x x xdenying the existence of cultures). Confronted with thisH . . . . . . x x x x x xdilemma, I propose that we go on using the concept ofI . . . . . x . . x x x xculture, including the plural form, because of its practi-J . . . . . . . x x x x xcal advantages. We should do so in a responsible way,K . . . . x . . . x x . xattentive to the specific audience and also to the prob-L . x . . . . . . . x x xlem of communicative economy.9 There are many situ-

Fig. 1. Three hypothetical distributions of features ations in which Japanese culture is a convenientacross individuals. shorthand for designating something like that which

many or most Japanese irrespective of gender, class, andother differences regularly think, feel, and do by virtue

over, feature 7 does not readily group with either of the of having been in continuous social contact with othertwo cultures, and individual D may be seen as partici- Japanese. And I am confident that at least among con-pating in both. temporary anthropologists the first phrase will very of-

ten be understood as equivalent to the second. After all,anthropology did not discover intrasocietal variation

Incomplete Sharing and the Identification of only yesterday. While many classic studies of small-scale, out-of-the way societies certainly do not showCulturesany awareness of it, peasant studies, explorations ofgreat and little traditions and of center-periphery rela-No distribution of learned routines among real people

will ever be much clearer than that in the bottom dia-gram, and consequently there will always be more than

9. By the latter I mean that, with space and time always scarce,one way to cut out cultures from the fuzzy-edged clus- simplification is inevitable, and therefore it is wise to decide cir-ters of habits that we observe. I suspect that most of the cumstantially rather than in principle how much of it is permissi-

ble without distorting ones argument.culture sceptics do not really want to imply that there



tions, research on gender, and the ethnographic study of ter off not even trying, and for this purpose, a word todesignate the clusters will be useful.complex societies and cities have been with anthropol-

ogy for quite some time now and have frequently occu- Let me now turn to the way in which one culturesceptic arrives at the conclusion that positing a cul-pied themselves explicitly with such variation or at

least acknowledged its existence.10 Consequently, the ture is something we should avoid. Borofsky, doingethnographic fieldwork on Pukapuka, a small Polyne-danger of being misunderstood by fellow anthropolo-

gists when speaking of a culture is, I think, much sian atoll, learned that the islanders were all well ac-quainted with a certain tale of Wutu which accordingsmaller than the critics claim.

Moreover, when there is enough time and space, to most of them dealt with a man who cleverly escapespersecution by a couple of anthropophagous ghosts.nothing prohibits us from representing the arbitrariness

and internal variation of such cultures as faithfully as However, individuals renditions of the tale varied con-siderably, and even having it told repeatedly by thepossible or resorting to formal methods of analysis for

delimiting cultures instead of trusting our intuition same person could produce different versions. Thesewould deviate again from what the very same personoras is commonly done when delimiting ethnic cul-

turesthe judgment of the people we investigate. One presented when telling the tale in a group. Accordingly,there was no single content element of the plot that wascould also specify a minimal numerical level of consen-

sus required for a culture and then search for maximal included in every rendition of the tale, and even thoseelements that reached a 67% consensus in any gendersets of features that fit this requirement (standard sta-

tistical procedures such as cluster analysis offer them- or age-group were few. Only focusing on the 67% con-sensus of those individuals considered experts on theselves for this task). When describing the two cultures

in the bottom diagram, we may distinguish between topic of tales by their fellow islanders would includeenough elements to produce a version that approachedcore features that are shared universally (feature 3 for

the first, features 10 andarguably12 for the second) the clarity and coherence which none of the individualrenditions failed to display (Borofsky 1994a:33134).or close to universally by the carriers of the culture in

question and others that are less widely and unequivo- Clearly, there is no universal sharing here. Having anumber of other Cook Islandersor readers of this arti-cally distributed and may be seen as less central. Noth-

ing prevents us from introducing temporal variation clerender the tale, however, would result in versionsthat would show hardly any commonalities either withinto the picture: searching for the same features in the

same individuals at other points in time may produce one another or with the Pukapuka renditions. Most peo-ple would very likely reject the task, saying that theydifferent distributions which, however, could again be

expressed in matrices and superimposed on the previ- did not know the story. But among the tales Pukapukaindividuals told, a general family resemblance is diffi-ous ones to introduce a third dimension. One may also

think of replacing the simple dichotomy of presence/ cult to deny. Some elements appeared with greater fre-quency than others, and one may see these as more cul-absence with quantitative values, since people will of-

ten act differently or with varied intensity in repeated tural and the rarer ones as more idiosyncratic orintroduce an arbitrary minimal frequency of occurrenceinstances of the same situation. All this increases com-

plexity, but the distribution will very likely still be above which a specific element is to be considered cul-tural (and mention that limit whenever speaking ofclustered, and we are still not necessarily thrown onto

intuition as the only method for finding such clusters. such cultural elements). Alternatively, one may searchfor those persons showing the highest consensus withThus we are also left with the problem of naming them.

It may be objected that the total matrix we are dealing each other and take their average version as the mostcultural11 orin a kind of analysis that Borofsky doeswith (Lowies only . . . natural unit for the ethnolo-

gist) has 6 billion rowsone for each living individual not considersearch for elements that co-occur with acertain frequency or that even implicate one anothersin the world, not to mention corporate actors that could

also be regarded as culture carriers and dead individ-ualsand that it has an almost infinite number of 11. This is commonly done in consensus analysis and is based on

the assumption that, when asked for their judgment on culturalcolumns, there being hardly any limit to identifiablequestions, experts will agree more often than laypeople. The latterfeatures. On top of that, the matrix changes at abase their answers on chance or improvisation, not on knowledge,tremendous pace. Nevertheless, from all we know and and are therefore less likely to come up with identical answers to

from what social psychologists have found out about a given cultural question. Consensus analysis is a statistical modelfor determining whether there is a common culture behind infor-human striving for conformity (Lang n.d.), we can bemants responses and, if so, how to estimate the culturally cor-sure that it will not show a random distribution but willrect answer to a given question from those responses. The an-be highly patterned. In an analogy with what I have said swers of each informant are weighted in proportion to theabout historical and optimal usages, the fact that we are informants average agreement with the others, a method which

as yet not particularly well equipped to describe and ex- privileges the experts, since they tend to agree with at least somepeople (other experts). It should be noted that fairly small samplesplain this enormous matrix and the clusters thereinsuffice to produce highly reliable estimates (Romney, Weller, anddoes not mean that we never will be or that we are bet-Batchelder 1986:32627). Romney, Batchelder, and Weller (1987)and Romney in an annotated bibliography on his Internet home-page give overviews on the now numerous applications of this10. Thus, there is perhaps less need than Keesing has argued (1994:

303, 3078) to turn to cultural studies as a guide in this regard. method.



presence, making for larger building blocks that can be for specific instances of individual thought and behav-ior. But sometimes communicative economy maysubjected to the above operations instead of the individ-

ual elements. In any case, sequence seems to be unprob- make it expedient to speak of a culture and identifythe constituent units of such a cluster as elements,lematic, since Borofsky offers without comment an ap-

parently standard succession of all content elements features, parts, or traits of that culture. In doingso we at the very least avoid the impression that there(1994a:332). Whatever the approach, it is clear that all

these content and sequential elements and aggregates is no such thing as the tale of Wutu on Pukapuka.In my view, speaking of culture while making it clearwhich occur with significant frequency belong to a rep-

ertoire on which individuals may draw when telling the that universal sharing is not implied does not automati-cally privilege coherence. Just as we may concentratetale, constituting the material for their guided impro-

visations (Bourdieu 1977 [1972]). No Pukapuka indi- on explaining why a glass is half-full as well as why itis half-empty, sharing is as good a theme for anthropo-vidual is unaware of this repertoire, while most outsid-

ers certainly are. logical research as nonsharing, and I wonder how wecan avoid either when attempting to portray and ex-In contrast to Borofsky, who would not venture be-

yond the cultural (see above), I do not consider it plain peoples ways of life realistically. And neitherdoes such an approach preclude temporal variation orproblematic to call this repertoire a part of Pukapuka

culture. Moreover, a description of the tales elements presuppose that the always arbitrary, abstract entitythat we call a culture becomes a thing, an essence, orand their frequencies and likelinesses of co-occurrence

or evenif such can be foundthe identification of a living being. Moreover, defining anthropology as thescience of culture does not mean that culture must belarger clusters that constitute alternative versions of

the tale would in my eyes constitute a faithful ethno- the sole focus of analysis: obviously, we do want toknow what events, acts, people, and processes (seegraphic representation of that specific part of Pukapuka

culture, without confusing anyone about the fact that Barth above) do with culture and what they let culturedo to them. Dropping culture/s, however, will leaveindividuals will disagree with each other and even with

themselves in their ways of making use of that reper- us without a word to name those clusters that, ill-shaped though they may be, are nonetheless out theretoire. Representational techniques such as bell curves

of certain features distributions or identifying center and do play an important role; and it also makes it dif-ficult to define the discipline in short and positiveand periphery within a cultural inventoryor domain,

or schema, or semantic networkmay help us here. words, at least if we do not content ourselves with prac-ticing the fieldwork science.This would perhaps come close to what Keesing seems

to have had in mind when he expressed the hope that As pointed out, there is no ultimate logical reason toretain culture/s (or to abandon it), but there are prag- a culture as a bounded unit would give way to more

complex conceptions of interpenetration, superimposi- matic ones even beyond that of communicative econ-omy. They have to do with the success of the concept,tion, and pastiche (1994:310) and what Appadurai is

looking for when he proposes that we begin to think and it is to them that I will now turn.of the configuration of cultural forms in todays worldas fundamentally fractal, that is, as possessing no Eu-clidean boundaries, structures, or regularities. . . . we Pragmatic Reasons for Retaining Culture/shave to combine a fractal metaphor for the shape of cul-tures (in the plural) with a polythetic account of their The concept of culture has undoubtedly exerted an in-overlaps and resemblances (1996:46). fluence beyond the borders of the discipline (HannerzThe approach just outlined can easily be extended to 1996:30):other domains not only of knowledge but also of ob-served behavior. Everywhere we find sets of certain Suddenly people seem to agree with us anthropolo-learned features that are shared more extensively by gists; culture is everywhere. Immigrants have it,people who interact with each other than between these business corporations have it, young people have it,people and others with whom they do not interact or women have it, even ordinary middle-aged menamong those others. And everywhere we will find that have it, all in their own versions. . . . We see adver-people are aware of this fact, while they are certainly tising where products are extolled for bed culturenot ignorant of individual variation even among those and ice cream culture, and something called thewho have much in common. We should try to describe cultural defense plea is under debate in jurispru-the unevenness of any such differential distribution dence.(Hannerz 1992) as well as we can, and it is clear that asyet the precise extent of interindividual conformity and It is concern for the nations culture that makes the

French government establish commissions to search forvariation within human groups has received insuffi-cient attention, and therefore we do not have a clear indigenous equivalents of unwanted loanwords, and it

is again in the name of culture that the Chinese and In-theory, for instance, about how much social interactiongives rise to how much culture. We must also face the donesian leaderships reject the claim to universal appli-

cation of the Declaration of Human Rights, declaring itfact that once culture is found to be incompletelyshared it will have that much less explanatory power a product of Western culture unfit for exportation. Ev-



eryday ways of contemporary talk have been heavily in- blocks that are no less incompatible than the old, ideo-logical ones.12 While far from controversial within hisfluenced by our anthropological concept of culture

(Keesing 1994:303). Thus, it is no longer certain that an own discipline (see Axford 1995: 19194), Huntingtonswritings have certainly had a greater influence on theevaluative, elitist view of culture (Goody 1994:254

55) prevails, and it cannot be taken for granted that general public than any contemporary anthropologicalstudy can claim, extending to, for example, the Germanlaypeople will invariably associate the word with the

original meaning, in which it was reserved for improve- president Roman Herzog, who found Huntingtonswork a useful companion when visiting China (Nassment and its results (first of gardens, then of individu-

als, and finally of societies [see Clifford 1988:337; 1996).Huntington is an extreme representative of a moreKroeber and Kluckhohn 1952:15, 44]). Instead, they will

often understand us fairly well, and this is quite re- general figure of thought that is identified as cultur-alism or cultural fundamentalism (Stolcke 1995). Itmarkable, since the word in its anthropological mean-

ing did not enter standard dictionaries of the English posits the existence of a finite number of distinct cul-tural heritages in the world, each tied to a specific placelanguage before the late 1920s (Kroeber and Kluckhohn

1952:63). It is precisely this success that makes anthro- of origin. Since these are taken to be ultimately antago-nistic and incommensurable, they and the individualspologys brainchild difficult to control (Eller 1997:253,

251): associated with them are considered best kept separate,ideally in their respective homelands or, if that fails, in

As many commentators have noticed, the first thing ethnically defined quarters, as is currently being sug-to realize is that anthropology no longer owns the gested by some urban planners in Germany (Die Tages-concept of culture (if it ever did). Virtually all ele- zeitung, November 26, 1997). Stolcke finds that in Eu-ments of societyacross the political spectrum ropean reactionary political discourse the new rhetoric. . . have learned the language of culture. . . . of culture has largely supplanted the older one of race.American society has become culture-conscious, to Culture is a more egalitarian notion, since everyonethe point of a culture cult in civic society. . . . is supposed to have it (although, of course, in distinctCulture and difference have become the dominant variants). But this is still unlike racism, in which someparadigm of the day, and individuals are being en- people are believed to be genetically defective whilecouraged, even driven, to conceive of themselves in othersusually egos groupare not. Cultural funda-these terms. mentalism, therefore, will not serve as ideological but-

tressing for new colonialisms, but for fueling xenopho-Moreover, this trend is by no means restricted to postin- bic tendencies in the Euro-American immigrationdustrial societies or those aspiring to such a position. countries it is already being amply used (Stolcke 1995:On the contrary, Sahlins (1994:378) states that 48).The notion of incommensurable cultures best keptthe cultural self-consciousness developing among

distinct is not restricted to the political right wing, asimperialisms erstwhile victims is one of the moreStolcke emphasizes (1995:6). It can be detected in recentremarkable phenomena of world history in the latepapal encyclicals that introduce the concept of incult-twentieth century. Culturethe word itself, oruration, that is, synthesizing elements of two culturessome local equivalentis on everyones lips. . . .while maintaining the integrity and identity of eachFor centuries they may have hardly noticed it. But(Angrosino 1994:825), on either side of the currenttoday, as the New Guinean said to the anthropolo-multiculturalism debate over educational contents ingist, If we didnt have kastom, we would be justthe United States (Eller 1997:252; Amit-Talai 1995:like white men.140), and among those Greek anthropologists who denyforeigners membership in their association becauseWithin the academy, the culture concept is also gain-

ing popularity. At least in Germany, major feuilletons they consider them not really able to understand Greekkeep announcing the cultural turn in the humanities(e.g., Bachmann-Melik 1996), and the replacement of

12. There are whole passages in Huntingtons article which read asGeisteswissenschaften with Kulturwissenschaften,if copied from anthropological textbooks. Consider the following:centering on a less high-brow notion of culture (Schle- Villages, regions, ethnic groups, nationalities, religious groups, all

sier 1996), has its proponents. Cultural studies has fast have distinct cultures at different levels of cultural heterogeneity.The culture of a village in southern Italy may be different from thatestablished itself in many countries, and its adherentsof a village in northern Italy, but both will share in a common Ital-have moved into a more anthropological direction ofian culture that distinguishes them from German villages. Euro-conceiving culture (Keesing 1994:303), with, for exam- pean communities, in turn, will share cultural features that distin-ple, scholars of high literature descending onto the guish them from Arab or Chinese communities. . . . People have

worldly levels of popular novels, comic strips, soap op- levels of identity: a resident of Romemay define himself with vary-ing degrees of intensity as a Roman, an Italian, a Catholic, a Chris-eras, and advertisements. And after the demise of thetian, a European, a Westerner. . . . People can and do redefine theirtwo-block paradigm in the field of international rela-identities, and, as a result, the composition and boundaries of civi-tions, the Harvard political scientist Samuel Hunting- lizations change (Huntington 1993:2324). Especially the last sen-

ton predicts a clash of civilizations (1993, 1996) in tence, however, is all but forgotten in the further course of Hun-tingtons argument.which cultural differences give rise to multiple new



culture (Kuper 1994:54546). It is also never far from them as far as possible to the interests of specific indi-vidual and corporate agents, thus giving authors (Foxmost contemporary nationalist, ethnic, and fundamen-

talist movements. Here, at last, one finds culture being 1995:27778) to culture. I am fully aware that if thereis one thing that Foucault wanted to discredit it was theused in the way denounced by the culture sceptics,

with, for instance, routine references to questionable idea of individual authorship, but what may be appro-priate for the very large discursive formations he inves-megacultures such as African culture or Western

culture (Eller 1997:252, 253), reductionist conceptions tigated need not be so for all of culture.Secondly, there are limits to culture.14 On the onethat restrict culture to, for example, ritual (Angrosino

1997:828), or vague Volksgeist-like ideas of a mystical hand, culture does not suffocate the idiosyncratic, andindividuals can never be reduced it. To conceive of cul-substance or ethos that suffuses a given culture and the

community of its carriers (Eller 1997:252; Angrosino ture as a toolkit that can be put to manifold uses butwill never do anything of itself, however, is hardly con-1997:827). Whether anthropologists like it or not, it ap-

pears that peopleand not only those with power troversial now for the numerous anthropologists whohave taken up a concern with praxis and the relationwant culture, and they often want it in precisely the

bounded, reified, essentialized, and timeless fashion between structure and agency. More neglected is theother limit of culture, between it and what is commonthat most of us now reject. Moreover, just like other

concepts such as tribe, culture has become a political to all of humankind. Anthropological research on hu-man universals has not flourished recently, to the pointand judicial reality, requiring any attempt to authorize

more deconstructed notions to reckon with consider- that there are no entries or index entries on universalsin two major new encyclopedias (Barnard and Spencerable institutional inertia (see the experience of the ex-

pert witnesses for the Mashpee Wampanoag claim to 1996b, Ingold 1994; but see Brown 1991). Cross-culturalstudies leading to the identification of universals havecultural continuity [Clifford 1988:277346]). In my

view, however, this should not discourage us from de- not fared much better, if their share in major journalsis any evidence. Moreover, research on the expandingconstructing such understandings and developing our

own truths (which does not necessarily mean speaking level of global culture (wearing T-shirts, mourningLady Diana, having heard about global warming, know-for others in any case). For this purpose, I think that

three fundamental insights about culture require spe- ing how to use a thermometer, liking soccer, etc.) thatis socially transmitted but no longer tied to any specificcial emphasis.

First, the social reproduction of culture is always location or group (Hannerz 1996:38) is only just emerg-ing (Brumann n.d.). Yet it is precisely with reference toproblematic and never guaranteed. Maintaining cul-

tural consensus across time and individuals requires (genetically generated as well as acquired) universalsthat we can reject exaggerations of cultural differenceconsiderable effort. This point is almost always side-

stepped by cultural fundamentalists, who seem to pre- and the notion of incommensurability, pointing also tofieldwork experiences in which anthropologists andsuppose stability as the natural condition of cultures

and speak unproblematically of, for example, theusu- their informants frequently develop common under-standings and emotional affinity relatively quickly.ally old but unspecifiedage of a given culture.

Moreover, culture has often simply been adopted as a And it is alsopossibly onlyfrom here that legitima-tion for universalist projects such as basic human rightsless controversial word by people whoconsciously or

unconsciouslystill hold to racist ideas of pseudo-ge- can be drawn and their rejection as merely Westernculture denounced. I do not agree that anthropologynetic transmission and its relatedness to phenotype (a

point that is also made by several of the commentators is fundamentally about difference (Eller 1997:251) ifthis is intended to be a programmaticinstead ofon Stolckes aforementioned article). Almost automati-

cally, recognition of the problem of reproduction will merely historicalstatement.Thirdly, culture is not always ethnic culture, and nei-lead to the role of power in achieving cultural consen-

sus. Here it will be necessary to overcome the remnants ther is it always tied to identity. Yet anthropological aswell as lay expositions of culture are frequently prem-of the Parsonian divide and re-sociologize anthropol-

ogynot simply by reciting the Foucault-inspired dis- ised on a presumed synonymy of the three, quite irre-spective of any commitment to cultural fundamental-coursepower mantra and by routinely ascribing dis-

courses, and culture in general, to very large and vague ism. For example, the Encyclopedia of World Cultures(Levinson 199195) lists ethnic cultures, and althoughforces (such as technoscience, colonialism, or the

German imaginary [Linke 1997]13 but rather by tracing thought is being given to including national cultures in

13. This article demonstrates how even the deconstruction of racistideas can sometimes border on cultural fundamentalism. Linke ob- treme utterances of politicians that are quoted (1997:56061) pro-

voked public outcries showing that German political fantasy is,serves an interesting obsession with blood imagery in the Germanpublic imaginary. That the latter is homogeneously distributed at the very least, divided. While I do not want to deny continuities

in the history of German racism, I think that an analysis of publicand clearly bounded and has been handed down unproblematicallyfrom the Nazi period to the present seems, however, to be taken discourses which does nothing to gauge their distribution and in-

fluence over time must remain incomplete and suspect of arbitrari-for granted throughout her analysis. This appears questionable,since the metaphors of floods of immigrantswhich, represent- ness.

14. On this point, Hannerz, borrowing from Redfield, has an in-ing another kind of liquid, are found to be related to blood (1997:56465)are hardly unique to Germany and since many of the ex- sightful discussion (1996:3043).



the Human Relations Area Files (Ember 1997:12) there Still, there is no denying that many ordinary peoplehave grasped at least part of anthropologys message:is still no talk of including, say, academic culture, punk

culture, or gay culture in their clearly trans-ethnic man- culture is there, it is learned, it permeates all of every-day life, it is important, and it is far more responsibleifestations. Not that I would envy anyone who wanted

to undertake such a difficult task, but I still think that for differences among human groups than genes. There-fore, I think that retaining the concept will put us in awe should be careful not to overethnicize anthropology

and pay due attention not only to gender cultures but better strategic position to transmit the other things weknow than we would achieve by denying the existencealso to age (Amit-Talai and Wulff 1995), regional, pro-

fessional, and class cultures, as well as to the global cul- of culture/s.16 Choosing the former strategy, we can tryto establish anthropology as the expert onif no longertural level mentioned above. We should also more

closely analyze the interplay of ethnic cultures with the owner ofculture, whereas opting for the latterplaces us in the difficult position of denying somethingthese other cultures that often do not stop at ethnic or

national boundaries. On the same account, I do not con- about which we rightly claim to be more knowledge-able than others.sider it a wise move to follow Appadurai in restricting

the concept of culture (or the cultural, as he prefers) Staying with culture/s, we could object to Hunting-ton that he is justified in paying attention to the roleto those differences that either express, or set the

groundwork for, the mobilization of group identities of culture but that the drive to power and wealth thatunderlies much of global politics is very likely universal(1996:13) or to agree with Knauft that culture is now

best seen . . . as a shifting and contested process of con- and more often clothes itself in cultural differencesthan is caused by them. We could add that there are in-structing collective identity (1996:44). To do so would

prevent us from showing that not all culture is relevant deed anti-Western or anti-Islamic feelings in the worldbut that currently none of the pannational civiliza-to identity formation and that what collective cultural

identity exists need not be ethnic. I believe that anthro- tions he identifies can count on a degree of internalsolidarity that is in any way comparable to what fre-pologys critical potential with regard to ethnic and na-

tionalist movements and to cultural fundamentalisms quently develops in smaller culturally defined groupssuch as nations or ethnic groups. We could alert him toin general would be seriously hampered as a conse-

quence. After all, it could be a healthy reminder that the fact that almost any of the eight major civiliza-tions he identifies conceals so much cultural diversitywhat people of a given nation really have in common is

often trivial things such as familiarity with certain soap that their analytical value must be doubted and thatglobal communication, migration, and cultural diffu-brands, commercial slogans, or TV stars and not an

ever-present awareness of their common history and sion will certainly not make the picture any clearer inthe future. We could point out that political salienceheritage. Anthropologists should remain capable of

showing people that what they see as their culture is seems to be more important than cultural diversitywhen categories as narrow as Japanese civilizationoften a rather arbitrary selection.

These insights, of course, cannot be allowed to ob- and as broad as African civilization are considered tobe of the same order. We could refer him to anthropo-scure that the reified notion of culture has become a so-

cial fact that itself deservedly receives anthropological logical studies that try to identify wider cultural areaswith less intuitive methods (Burton et al. 1996) and ar-attention. Nor will they rescue us from the dilemma

that the demand for unproblematically reproduced, over- rive at units far from congruent with the civilizationshe proposes. We could sensitize him to the degree tolarge, and ethnicized cultures often comes from pre-

cisely those people we sympathize with and that this which his separating a Confucian civilization from aJapanese civilization disregards important East Asiankind of culture is often deployed or commoditized more

effectively than what we have to offer as an alternative. cultural commonalities (including precisely the influ-ence of neo-Confucianism) and thereby falls prey to aMoreover, anyanthropological or amateuridentifi-

cation and description of ones own or another culture myth of Japanese uniqueness that Japanese and foreign-ers alike have done much to maintain (Dale 1986,is potentially reactive, that is, capable of influencing

that specific culture and the people carrying it when it Miller 1983, Yoshino 1992).17becomes publicly accessible. Consequently, ethno-graphic innocence is a vain hope in an age in which 16. One anonymous reviewer considered the idea to educate themass media proliferation can very quickly turn any general public . . . a strangely sanctimonious view of anthropolo-

gys role in the world. I agree that a scientific discipline as suchstatement about cultural affairs into a political assetoris not obligated to anything, but those of its practitioners who aretarget (see, e.g., the public debate about Hanson 1990 aspaid for teaching or for research, often out of public funds, shoulddocumented inHanson 1991, Linnekin 1991, and Levine feel some responsibility for disseminating truth. I also sympathize

1991, or the controversy about Karakasidou 199715). with pleas to integrate the discipline more centrally within acade-mia, and in public policy debates (Weiner 1995:14). As I see it, allthis leads to one or another form of educating the general public.15. This is an ethnography about Greek Macedonia which in 1995

was rejected for publication by Cambridge University Press not be- 17. I restrict myself to the objections we can raise as experts of cul-ture. There are other, equally serious flaws in Huntingtons modelcause of qualitative deficiencies but for fear of retaliation from

Greek nationalist sources (see the Internet documentation of which, however, need no anthropologist to denounce them, suchas the claim that armed conflicts arise more often at the fault linesevents, opinions, and protests under http://www.h-net.msu-edu/

/sae/threads/CUP). between his civilizations than within themthe inhabitants of



Moreover, we could tell practitioners of cultural stud- continues to be so, especially in the hands of culturalfundamentalists. But, weighing the successes and fail-ies and other disciplines that they are indeed right to

extend their study of culture to the more mundane and ures, I am not convinced that the concept really entailsthe criticized connotations, and I think that it can beeveryday, but we could go on to argue that an internal

analysis of the products of popular culture alone (as, for dissociated from them and used to its best intents.Staying with culturewhile emphasizing its problem-example, in most contributions to Schwichtenberg

1993) remains ungrounded if it is not complemented by atic reproduction, the limitations imposed on it by theindividual and the universal, and its distinctness fromethnographic field research on recipients engagement

with these products and the resulting practices, dis- ethnicity and identitywill enable us to retain thecommon ground it has created within anthropology andcourses, and fantasies, referring them to, for example,

anthropological research on television (Kottak 1990, A. profit from the fact that the general public increasinglyunderstands what we mean when we employ it. Deny-P. Lyons 1993,H. D. Lyons 1993,Mankekar 1993a, Pace

1993, Wilk 1993) as a source of inspiration. ing the existence of culture and cultures will be difficultto transmit to the many that see them out there, andOf course, if a sufficient number of anthropologists

agrees that the use of the term culture undermines they will very likely turn to others who may then dis-seminate their questionable expertise without serioussuch a strategy and contradicts all our scientific results,

its meaning will eventually converge with this assess- competitors. Any scientific concept is a simplifyingconstruct and has its costs, but once the advantagesment and the term will have to be dropped (Brightman

1995:541). But I am not convinced that this is inevita- have been found to outweigh these costs it should beemployed with a clear conscience.ble, and I regard the resulting speechlessness as too high

a price to pay. We might consider a move similar to thatof the pop star Prince, who lately gained much attentionby renaming himself The Artist Formerly Known asPrince, or TAFKAP for short.18 If only for the more Commentsdifficult pronunciation, however, I doubt very muchthat TCFKAC would become a comparable success.Therefore, I propose that we retain culture the noun lila abu-lughod

Department of Anthropology, New York University,in its singular and plural form and clarify for those non-anthropologists who are willing to listen what the phe- New York, N.Y. 10003-6790, U.S.A. (lila.abu.lughod

@nyu.edu). 2 vii 98nomenon so designated really iswhich, as I have triedto emphasize, requires very clear and definite formula-tions about all the things it is not. Brumann has done us a service in laying out, side by

side, many of the attempts by contemporary anthropol-ogists to question the culture concept. He also remindsus of the complexity and richness of the anthropologicalConclusiontradition with regard to this central concept. There is atendency these days to close out the history of the disci-There is no immanent justification to be drawn from

the empirical world either for using or for discarding the plineto forget that those who came before us reflectedcarefully on a range of theoretical problems and repre-culture concept. Any set of persons who have specific

routines of thinking, feeling, and acting in common will sented a variety of approaches.His intelligent arguments in defense of the cultureinvariably be different with regard to other such rou-

tines, and therefore wherever we find sharing there is concept, however, miss the mark. I think he needsto ask why it is that so many of us might want to ques-also nonsharing. If we agree, however, to imagine the

world in which people dwell as a continuous and un- tion the concept. What is at stake? As he is careful tonote, the anthropologists in his collages come out of dif-bounded landscape, endlessly varied in its features and

contours, yet without seams or breaks (Ingold 1993: ferent theoretical traditions and make their argumentsfor different reasons. I will let others speak for them-226), we will still need a vocabulary for describing its

mountains, plains, rivers, oceans, and islands. The an- selves. My own argument for writing against culturewas developed in the context of trying to think how Ithropological concept of culture offers itself for that

task, all the more so since it has persuaded many people might write an ethnography of a Bedouin community(in which I had worked for many years) that did justiceoutside anthropology of its usefulness. There is no de-

nying that it has often been applied wrongly and that it to the complexity, uncertainty, and contestations of ev-eryday life and to the individuality of its members. Iwas hoping to achieve in my text some fidelity to myRuanda certainly did not need cultural differences of a civiliza-experience of being there. I was also working with ational order to slaughter each other, and the Europeans for centu-

ries did not need them eitheror the specter of an Islamic-Confu- strong sense, following Edward Saids critique of Orien-cian alliance forming against Western civilization, which, to talism, of the ways in which representations of peoplemy mind, is simply unfounded. in other parts of the world, particularly parts of the18. An interim as symbolgraphically expressed by a symbol

world that are viewed with antipathy in the West,and never written outattests to this artists versatility in the fieldof applied semiotics. might reinforceor underminesuch antipathies. I



concluded that the idea of a culture, with its inevita- racist politics. It seems to me that our role is not to useour expertise in culture to correct him (by showingble generalizations and typifications, had become a cen-

tral component of the distancing and othering against that his cultural units are too big or too incommensu-rate, too homogenized or too crude) but to criticize thewhich I wanted to work, even while I recognized that

humans are, in the broadest sense, cultural beings. very notion of setting up groups of people defined byshared cultures as hostile and opposed. If civilizationsI now feel even more strongly that the concept is

problematic, mostly because I do not agree with Bru- are extensions of cultures and cultures depend on cul-ture and we do not question the notion of culture, thenmann that usages can be separated from the concept

itself. I cannot accept the idealism that distinguishes we are not in a position to mount this critique.There are other arguments to be made about the wayhistorical usages from optimal usages. Perhaps this

is a disagreement about language. I favor pragmatics in which the culture concept is unhelpful. These canbest be made by taking up a small, perhaps throwawayover semantics. I do not think that concepts have tran-

scendent or true meanings. Concepts are human cre- point that Brumann makes in defense of culture. Hepoints to the contributions that anthropological studiesations and socially embedded. They are employed only

by real people in real situations. To quote the defini- of television might make to cultural studies. I agree ab-solutely that we must complement the study of thetions, many very subtle, offered by anthropologists of

the past is to be too literal. Such definitions tell us noth- products of popular culture with ethnographic field re-search on recipients engagement with these productsing about the contexts in which they arose or, more im-

portant, the contexts in which the concept is put into and the resulting practices, discourses, and fantasies,but as someone who has been engaged in exactly suchplay and with what impact.

Brumann says that there are pragmatic reasons for field research for almost eight years now I fail to seehow this will vindicate culture. Quite the opposite.staying with the concept; I think that there are prag-

matic reasons for questioning it. He gives us two rea- In a recent article dedicated to Clifford Geertz (surelythe most influential American theorist of culture, inex-sons for reclaiming the concept: communicative econ-

omy or expediency (it is a shorthand for the notion of plicably absent from Brumanns discussion) entitledThe Interpretation of Culture(s) After Television, I re-shared routines which surely we all recognize exist) and

the preservation of our authority as experts to intervene flect on how and why anthropologists should study tele-vision on the basis of some thoughts on Egyptian televi-when nonanthropologists such as cultural fundamen-

talists (a wonderful term!) misuse the concept. In some sion serials place in the lives of some villagers in UpperEgypt. I conclude one section (1997:121) as follows:ways, I grant him the first. We all know in some sort of

rough way that different groups of people share certainthings, ways of thinking and doing. Who could deny it? Taking television seriously forces us to think about

culture not so much as a system of meaning orAnd what other name is there for this? Despite the vir-tues of Brumanns definition of culture as an abstrac- even a way of life but as something whose elements

are produced, censored, paid for, and broadcasttion, though, I doubt that anyone today would be will-ing to undertake the formidable task he proposes of across a nation, even across national boundaries.

The hegemonic or ideologicaland thus power-re-drawing up matrices of shared and nonshared features.The larger problem is that refining and redefining can- latednature of mass-mediated cultural texts in the

service of national, class, or commercial projects isnot solve the problems created by the fact that the cul-ture concept carries historical accretions and takes its undeniable. This, in turn, should lead us to think

about the ways that aspects of what we used tomeaning from the many contexts in which it is and hasbeen invoked. The concept is always contaminated by think of as local culture, such as moral values about

the proper age of marriage or the propriety of wom-the politicized world in which it is used, the same worldin which anthropology was born and flourished. That ens education, are themselves not neutral features

to be interpreted but the sometimes contested re-billions of amateur anthropologists would find usstrange to deny what they know is precisely the sult of other, more local, projects of power that are

worth analyzing.problem.This, I believe, is the crux of our disagreement. Bru-

mann admits that the concept of race had to be aban- I go on to note that ethnographies of television (becauseits cultural texts are produced elsewhere and inserteddoned as scientifically invalid and so subject to devasta-

ting political uses as to be dangerous. Culture, he into local households, communities, and nations) con-firm for us the need to rethink the notion of culture inclaims, is different. But is it? The fact that it is such

a successful and popular concept should be cause for the singular, as a shared set of meanings distinct fromthose held by other communities sometimes calledsuspicion, not self-congratulation. That the concept

lends itself to usages so apparently corrupting of the an- cultures (p. 121). I do not deny the existence of reac-tive processes of cultural assertion, the same ones thatthropological ones as the pernicious theses of Samuel

Huntingtons clash of civilizations is, for me, serious. Brumann, following Sahlins, notes. I see these as politi-cized processes of identity formation in global contextsHuntingtons glorification of Western superiority and

gross simplification and reification of cultures and cul- where culture has already been made (through colonial-ism and neocolonialism) definitional. But I also pointtural difference resonate with popular sentiment and



out that these reactive uses of culture are balanced an enormous growth of what Barrett has memorablycalled no-name anthropology (1996:179): ethno-by many other processes in the world today that unset-

tle the boundaries of cultures. The encounters that tele- graphic writing which avoids the

writing for culture

Documents

jstor terms

jstor archive

culture concept

collaboratingwith jstor

ffhvvhgyour use

terms conditions of

place of culture

amall use subject