Anthropological epistemologies

Anthropological

Nicholas Thomas

epistemologies

In his classic 1973 essay, ‘Thick description: toward an interpretive theory of culture’, Clifford Geertz declared that the analysis of culture - with which he equated anthropology - was ‘not an experimental science in search of law but an interpretive one in search of meaning’ (1973, p. 5). This was to deploy one of the polarities that had haunted and still haunts the discipline. To a greater degree, perhaps, than any other, anthropology has straddled the divide between the social sciences and the humanities, and has been stretched uneasily between a broad- ly positivistic explanatory approach to social and cultural phenomena and an empathetic exploration of communication and significance. It may be hard to imagine a synthesis of ‘experimental’ and ‘interpretive’ science, but neither these terms nor the ‘laws’ and ‘meanings’ they sought respectively to reveal

disembedding the philosophy of the discipline from theoretical reappraisals and from ethnography: few monographs are devoid of reflection upon the making of anthropological knowledge, and there are fewer theoretical treatises on the subject that are devoid of pri- mary ethnography. This essay is consistent with this bias of the discipline towards its own practical grounding; it avoids the ‘philosophy of social science’ genre and instead addresses

Nicholas Thomas is Director of the Centre for Cross-Cultural Research at the Australian National University, Canberra, ACT 0200, Australia. He has written extensively on the relations between history and anthropology, exchange theory, material culture, colonialism, and art, especially in the Pacific. His most recent book is In Oceania: Visions, Artefacts, Histories (1997).

appear in the same way today. An article on anthropological epistem-

ologies could, of course, rehearse arguments and paradigm shifts that apply to the social sciences generally, or to knowledge as a whole. The conflicting philosophies of Popperian scep- tical rationalism, phenomenological hermen- eutics, deconstruction, and the new realism of Roy Bhaskar and others all have their expressions and echoes in various strands of anthropology. Yet anthropologists have resisted

I

epistemological problems peculiar to anthropological knowledge, that arise to a significant extent from the discipline’s basis in fieldwork. I am not concerned so much with the issue of how this grounding has always been pivotal, but rather the ways in which fieldwork and its contexts have changed. This is not to presume that anthropology is no more than ethnography; one of the discipline’s strengths in recent decades

has been its capacity to incorporate historical research and to extend itself to commentary on literature and art. However, anthropologists tend to work in relation to the local encounter of fieldwork even when they are doing something different, as they write ethnographically about history and literature. For better or worse, the practice of fieldwork and the making of ethnography are central to the discipline. These practices are changing, creating new promises and new risks.

ISSl 153/1997 0 UNESCO 1997. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 IJF, UK and 350 Main Street. Malden. MA 02148. USA.

334 Nicholas Thomas

Stances

In reflecting on the consequences of fieldwork, Geertz’ essay provides a precedent. As is well known, he focused, not on a formal definition of the discipline or its theories, but on what its practitioners did, namely ethnography. For him this meant thick description, the interpretive inscription of social discourse, primarily in its interpersonal and local rather than its insti- tutional and global expressions. This was an appealing portrait of an analytical style, with a bias towards localized knowledge, that remains attractive today for many anthropologists, including those who would not count themselves among Geertz’ followers (although there is now an increasing preoccupation with the ‘local knowledge’ of larger phenomena, such as nations and transnational forms). Yet, as a characterization of ethnography, it seems now to stop rather short. Ethnography is not just thick description (which, as Geertz acknowledged, also characterizes the novel); it refers both to fieldwork and to writing, to a practice and a genre, and both have ramifications for anthropological epistemology.

These consequences, moreover, have received new twists since Geertz’ text. The business of fieldwork almost necessarily impli- cates ethnographers in social milieux remote from their own. Sites of research become second homes, more or less partially and with more or less awkwardness. The intensity and the romance of field research are not topics I am concerned with here; the point is rather that this intimacy generally prompted ethnographers to adopt an affirmative attitude towards the people studied, and even to write accounts of their culture that were to some degree complicit in dominant local understandings. This complicit relation has in fact been consistently advo- cated by the idea that the anthropologist should ‘adopt the native point of view’. This has been a powerful tenet since Malinowski, but has always been a rather facile one too. If it is obvious that certain forms of methodological relativism are indispensable, and that no serious inquiry can be engaged in unless there is a certain degree of common ground and respect for local understandings, it must also be clear that there are profound tensions between the aspiration to grasp and share an indigenous

‘point of view’ and the incorporation of that perspective into an analytical or theoretical discourse defined by Euro-American social science.

If a tension of this kind has long been implicit, it can only have been accentuated over recent decades. Anthropologists formerly presumed that the peoples they studied - whether they were European peasants or Pacific island- ers - would not be among the readers of their published ethnographies. Professional scholarship is no longer contained in this way, but tends to reach diverse audiences and to be used by them. Not only anthropologists, but also a literate fraction of the people studied, will read one’s work. It will also, in all likelihood, reach some in government in the nation researched; indeed it is a condition of many research per- mits that publications are supplied to various institutions and departments, perhaps only to be filed away, but sometimes to receive surprising attention. In the country one writes from, also, those in foreign affairs and official multi- culturalism routinely use anthropological knowledge. Insofar as anthropological writing is drawn into these fields, and even into ‘area studies’ such as Asian or Middle Eastern studies, it will be used in a way that is at odds with its anthropological reading, more for what it adds to knowledge of a place than for reflection on a theory or an issue. Under these circumstances, the question of how, and to what effect, a particular ethnographic account col- ludes with or subverts local perceptions is not an abstract epistemological issue, but something subject to open contention.

The issue of the ethnographer’s stance has become more acute in the wake of an overall politicization of social scientific and cultural knowledge. I suggest that this trend has unpro- ductively exaggerated the political significance of scholarly work, but that it nevertheless points to a specifically epistemological issue that was not important to Geertz, at least in 1973. Even at the time his essay was published, anthropology had been accused of endorsing and tacitly or actively supporting colonialism, and Marxist analyses were gathering a following. Although that particular perspective was much diluted by the late 1970s - the more determin- istic arguments being abandoned by their earlier proponents, while more general tenets became widely embraced - there was a shift towards

0 UNESCO 1997

Anthropological epistemologies 335

a sense that social knowledge was inevitably political, and indeed ought to be political critique. The understanding of knowledge as a project connected with, and justified by, efforts to reform or transform society was bolstered by the growth of feminist anthropology with its overt commitments, no doubt a desirable reac- tion against the bald assertion that social science could and should be value-free.

Yet - despite what I have said concerning the diverse audiences for ethnographic writing - this inflated sense of the role of knowledge in the world was unrealistic, and curiously so, given that the scholars concerned were supposed to be experts in the significance of beliefs and ideologies in social life, among other matters. Cultural studies, which became increasingly visible as a new discipline in the 1980s and early 1990s, to some degree in competition with anthropology, made the megalomanic pretensions of politicized scholarship and theory evident. The flavour of activist urgency that came to pervade a surprising range of writing, particularly around questions of culture and identity, surely reflects a wildly disproportionate sense of the efficacy of theory as a genre, and of what are often specialized if not arcane publications. The shifting economics of knowledge mean that no scholar today can be a Tom Paine, or is even likely to be a Margaret Mead, even if anthropological works are often appropriated locally in significant and unexpected ways. We need to define an intermediate point of view, which does not attempt to recover the pretensions of value-free neutrality, but acknowledges that research and writing take place in domains that may have important connections for cultural policy, but are generally at some remove from the most consequential theatres of political action and transformation. This may mean aban- doning both the liberal notion that intellectuals provide some kind of conscience for societies as wholes (which implicitly denigrated those in other spheres of social analysis and reflection such as the media, bureaucracy, and so on) and radical vanguardism alike. What is needed in their place is a more localized sense of the place of the anthropologist as a commentator and critic.

The issue was inadequately discussed as recently as the mid-1980s. In Anthropology as Cultural Critique (1986), Marcus and Fisher

reiterate the long-established idea that the distinctiveness of another culture questions received ideas at home: the foreign relativizes the familiar. Though apt enough as a gloss on one dimension of anthropological reflection, and on the critical logic of major recent works such as Geertz’ Negara (1980) and Strathern’s Gender of the Gift (1988), this does not prepare us for an unavoidable division in the ethnographer’s voice. Because the people studied have ceased merely to be scholarly objects, and have become partially incorporated within an expanded field of discussion, the anthropologist’s text may be increasingly drawn in two directions: on the one hand towards a global (in fact typically a Euro-American) professional discussion, that privileges the discipline’s questions, and the elevated register of ‘theory’, and on the other towards audiences within the nation if not the locality studied. I am not, of course, asserting that this trend has proceeded in an even and pervasive way. For those working in some regions, or those disinclined to respond to local circumstances, little may have changed. But I suspect that many anthropologists find that their own situations, and the contexts of their texts, epitomize the interplay of the global and local that has become theoretically fashion- able. The global-local relation may, in fact, be theorized in a partial way precisely because cosmopolitan scholars model the flows on at least slightly romanticized versions of their own travelling lives. Scholarship may be geographi- cally dispersed, but cannot count as universal in relation to local particulars.

Yet the divided locations of anthropological writing do have profound implications. The exoticism which structures much classic anthropological argument loses salience if the argument itself has ‘exotic’ circulation. And the question of the presence of the ‘native point of view’ in a particular text ceases to be a literary flourish on the part of a Malinowski, an ‘I was there’ gesture, but becomes an assertion that may be readily tested by ‘native’ readers who see their point of view misrepresented or appropriated. The idea that anthropology produces a ‘cultural critique’ of relations and mores ‘at home’ leaves us unprepared for the question of its commentary on the relations and mores actu- ally studied. Does the discipline simply attempt to represent these, in some sense on ‘their own

0 UNESCO 1997

336 Nicholas Thomas

Fijian policemen. Sam WaagenaarlRapho

terms’? Or are they equally to be subjected to the politically deliberate scrutiny of Western social science? The rhetorical strategy moreover too often leaves the home-point of our society unanalysed; it is no more than a stereotypic ‘West’. As anthropological discourse circulates to a greater extent than hitherto among the communities classically studied, and turns its vision upon communities at home, the paradigm of an us-them juxtaposition seems increasingly inappropriate.

Culture-ma king

Thus far I have suggested, all too generally, that what I have called the expanded audience for anthropological work creates a novel problem for the anthropological voice. The problem is perhaps most poignantly exemplified by a line of research that flourished during the 1980s

and early 1990s, that is perhaps now almost exhausted. I refer to the literature on the subject of the invention of tradition and identity.

A global trend of signal importance has been the elaboration of explicit constructions of local custom and identity. Though related to earlier ideas of local folklore, national distinctiveness, ethnicity, and the like, and thus not wholly unprecedented as a cultural phenom- enon, the objectification of culture at national, regional, and local levels has become singularly powerful over the last twenty years. Everywhere from the margins of Britain and Eastern Europe to Oceania and the Amazon, peoples have become conspicuously oriented towards the rhetorical elaboration of their identity, often towards cultural affirmation, autonomy, or sepa- ratism. No doubt these projects of identity are more heterogeneous than they appear, but the vocabulary employed is often that of a popu- larized anthropology: although all people’s cul-

0 UNESCO 1W7


tures are different, they seem to be becoming the same to the extent that they are concerned to affirm their different cultures.

My purpose here is not to analyse these dynamics, which have been discussed extensively in many theoretical essays and case studies, but to raise the problem of what the pro- cesses mean for the anthropological analyst, who confronts what we presume to call a ‘folk’ version of an anthropological concept - or rather, the anthropological concept - of culture. It has been shown, in many cases, that ethnographic research has been wittingly or unwit- tingly complicit in the codification of reified local ‘cultures’ of this kind. Old ethnographies are frequently mined for customs by culture- makers; publications may be upheld as author- ized versions of particular cultures. More subtly, the process of ethnographic inquiry frequently brings a new level of explication to ideas and behaviours.

I am not aiming here to recapitulate the effort to debunk traditions via the imputation that they are inauthentic concoctions, to be juxtaposed with ‘real’ cultures that were simply ‘lived’ in an unselfconscious manner. The point is rather that the anthropologist may, in the course of obtaining knowledge from an ‘informant’, mediate the activity of that ‘informant’ in facilitating a process of cultural explication. Even if this process takes place quite indepen- dently of anthropological complicity, the ethnographer can encounter a situation in which the paradigmatic object of analysis (not necessarily the object of his or her particular project), that is a ‘culture’ constituted out of certain dispositions, practices, rites, texts, and so on, is already present in indigenous articulations. Hence the work of systematization that anthropologists might once have had to do appears redundant, and ethnography appears less as an exogenous project than as a kind of repetition or transcription, not only of what informants already know, but of the form in which they know it.

I arrived among the Kwaio announcing my intention to record their customs . . . Since [the political movement] Maasina Rule (1946-53), they had themselves, in interminable meetings with millenarian overtones, sought . . . to codify their customary law . . . the political goal . . . was to create the equivalent of colonial legal statutes . . . As a professional chronicler of icustomi . . . I could

be enlisted in their cause both to write kastom and secure its legitimation. As long as I collected genealogies, recorded stories of ancestors, explored the structures of kinship, feasting, and exchange, and recorded ancestrally policed taboos . . . my work and the expectations of traditionalist (male) leaders meshed closely . . . Indeed, their politically motivated commitment to (the impossible task of) codifying customary law and my theoretically motivated commitment to (the impossible task of) writing a icultural grammar? in the manner of Goodenough, Con- klin and Frake doubtless, in retrospect, entailed a good deal of mutual co-optation. (Keesing 1985, pp. 28-29)

In this case, there is indeed a deep col- lusion between the anthropological account and the ‘native point of view’. But in response to these codifications and affirmations, anthropologists like Keesing shifted their ground, to engage the construction of culture itself as an analytical object (Keesing 1989). If this was for a time a fertile step - there was, at least, a proliferation of studies of cultural inventions and codifications - it can also be seen as a poignant one. Just as ‘natives’ had ceased to be objects of anthropological analysis, and became co-objectifiers, or co-interpreters, of their own cultures, anthropology stepped away from the prospect of co-authorship. This was a consequence of a grand disinvestment in bounded and homogeneous notions of culture, that seemed a necessary move, in theoretical terms, yet led almost to a further instance of the ‘denial of coevalness’ with which anthropology is charged in Fabian’s Time and the Other (1983), and certainly to a reaffirmation of the privilege and authority of scholarly know ledge.

If the critical stance towards the construction of culture therefore distances the anthropological account from ‘the native point of view’, it perhaps does so in a way that points to the contradictions in that notion. Although sometimes evoked, almost as a self-evident value, it is striking that the ‘native point of view’ has almost no counterpart in any other scholarly discipline. Neither psychologists nor sociol- ogists are generally concerned to present their objects of knowledge in terms faithful to particular human understandings of those objects; the aim may precisely be other than that, where an effort is made to expose common nonsense, artifice, and mystification, and this is true or thought to be true, in varying ways in philosophy, literary criticism, economics, art, and

0 UNESCO 1997

338 Nicholas Thomas

other fields. The nearest counterpart is perhaps in historical research, in the sense that the scholar may aim to capture values and understandings of a period, rather than assess events in a purely retrospective or ‘presentist’ fashion. But the powerful notion that another culture ought to be presented on its own terms, in some unde- fined sense, is morally rather than intellectually compelling. It is a consequence of the Maussian logic through which ethnographers understand the profound indebtedness incurred towards one’s hosts in the field. However those people themselves understand the relationship, our sense is that there is no way their support and their patience can be reciprocated, yet we nevertheless feel the need to attempt to do so through the register of writing: our writings are sometimes morally framed as efforts to validate or help those others, yet surely more typically help ourselves (cf. Fabian 1991, p. 264).

The anthropological project will thus generally be at least Janus-faced, towards ‘home’ and its intellectual traditions and disciplinary questions, as well as towards the presumed second home, into which one has generally invited oneself. The anthropologist may embrace a ‘native point of view’ at one moment but will generally turn a cold shoulder towards it at another. That is perhaps the wrong way to put it, because it could only be dishonest for scholars to pretend that they were without intellectual baggage, that where they - that is we - come from does not compel us, in most instances, to see questions in ways that cannot be reconciled with local constructions. Or, if we do so, we subsume our voices to a local, extra-scholarly debate, and drift out of a professional anthropological discourse.

If it was always absurd to imply that an anthropological account might faithfully reflect local self-understandings, anthropological writing is nevertheless being locally received in terms of that aspiration. Work in the ‘construction of culture’ genre is being vigorously critic- ized by local intellectuals, in effect for failing to collude with the ‘native point of view’, for insisting on, and perhaps too zealously overstat- ing the point that cultures are remade in and for the present. The arguments of Keesing and others have been contested by a Hawaiian scholar (Trask, 1991); while, perhaps with more justification, F. Allan Hanson’s account (1989)

of the ‘making of Maori culture’, which was widely reported and excerpted in American and New Zealand newspapers, was angrily rejected by Maori scholars and Maori activists (see discussion in Thomas, in press). A more generous attitude towards indigenous reaffirmations might acknowledge that these are themselves efforts of cultural interpretation and reinterpretation, perhaps not radically unlike the anthropological project. These particular contentions may be among the most obvious challenges to anthropology from indigenous knowledge, but they are not the only ones, and perhaps not the most fundamental ones.

In this essay I question what I suggest is a common-sense practitioners’ notion of the discipline, as an intellectual field constituted on the one hand by general theories and problems and on the other by localized studies. I have already suggested that the intellectual entailments of fieldwork are conducive to a highly differen- tiated audience for anthropological work, that is to say to a divided if not a confused voice, and to contradictory expectations manifest in the controversies cited, among other contexts.

Another sense in which the seeming comp- lementarity of universal theory and ethnography is misleading arises from the marginalization of the regional as a frame for anthropological discussion (redressed in an important, but neg- lected collection; Fardon, 1990). The defining significance of regions as theatres of research activity and intellectual debate is known inti- mately to anthropological practice, but typically denied by anthropological epistemology. All anthropologists in fact work, not only within a specific locality and the discipline as a whole, but also, if to a varying extent, within intra- disciplinary and cross-disciplinary area studies milieux, Mediterraneanist, Sinological, South Asian, or whatever.

Institutionalized regional specialisms are important in obvious ways for the sociology of the profession, but they are also significant for defining its forms of knowledge. Many osten- sibly central theoretical issues are not really global anthropological issues at all, but rather problems that arise from the encounter between

0 UNESCO 1997.


particular strands of the discipline and particular societies, that are then sometimes exported elsewhere, with translation problems internal to anthropology that are hardly less significant than those at the point of initial translation into the anthropological idiom. Anthropological texts are formed, not by a pure encounter between a theoretical language and an unmediated experi- ence of local fieldwork, but through regional traditions of anthropological scholarship. In some cases these have long histories originating in travel writing or colonial scholarship; in others the imprint of particular, eminent professional theorists can be lasting. The stamp of India and Dumont on the theory of hierarchy might be instanced; or the earlier specificity of the lineage debate in British studies of Africa; honour and shame in the Mediterranean; evolutionism in Polynesia; peasants in Latin America, and so on.

The regional theoretical vocabularies - academic equivalents of Melanesian pidgins - often entail their own hegemonies, and could be seen to inhibit genuinely global dialogues by focus- ing scholars’ interests on relatively parochial problems. To make this adjudication, however, is perhaps only to reintroduce a false univer- salism, or pretentious intellectual cosmopoli- tanism, that anthropologists of all people might be expected to be suspicious of. If ‘area studies’ debates indeed tend to be introverted and anti- theoretical, they may also be theoretically marked by engagement with the site of research, and thus reflect a more genuine compromise between a Euro-American discipline and a theatre of field research. At the moment when the global pretensions of cultural studies are becoming more evidently exhausted, the interplay between area studies and wider disciplines may suddenly provide something anthropology needs. Regions should not, however, be received simply as natural geographic entities that frame research and professional discussion: their histories and particular implications need to be traced more self-consciously. In any event, the metaphor I introduced earlier breaks down: anthropologists have not only one foot in their field site and the other in their discipline, but a third within a regionalist sub-profession, and probably others elsewhere.

This is perhaps not an ‘epistemological’ issue of the conventional kind, if we imagine

epistemology as a kind of meta-knowledge. But I would resist the idea that meta-knowledge constitutes more abstract knowledge, or something like the philosophy of social science, in anthropology. Reflection on the conditions of anthropological knowledge can incorporate debate about the project of translation, questions of scale, the status of models, and issues at that level of generality, but it must also address the determining contexts within which understandings are arrived at and fashioned into public products - into lectures and other teaching performances as well as publications, films, and reports. If this wider sphere of knowledge- making is acknowledged, the significance of regional traditions for better or worse must be addressed.

Reflexivity

The emerging discussion of the distinctiveness of regional anthropologies and national traditions can be seen as part of a trend towards greater self-consciousness among anthropologists, though the trend has generally been regarded suspiciously rather than positively. This is the context in which a highly visible debate of the 1980s may be addressed. The Clifford and Marcus collection, Writing Culture (1986), was seen to introduce, quite suddenly, the question of the literary qualities of anthropology and their relevance for ethnographic authority. The book, like Tala1 Asad’s similarly widely cited collection on Anthropology and the Colonial Encounter (1973), was in fact diverse, yet consistently referred to as though it rep- resented a manifesto for a particular stance. In this case the stance was seen as a new affirmation of subjectivism: ethnographic knowledge was no sure representation of another culture, but first an artifact of dialogue in which the ethnographer’s voice was as significant as that of the native, and second an artifice of tex- tuality, of authenticating devices that conveyed a sense of specificity to place while recapitulat- ing the conventions of travel and travel fictions.

What was perhaps most shocking about this intervention, from the perspective of those anthropologists most attached to the notion that anthropology really did produce and circulate valid accounts of other cultures, was the attitude

~

Q UNESCO 1997

340 Nicholas Thomas

Drawing by children of the Gossanah tribe, on the island of Ouvea, New Caledonia, depicting a clash with the French authorities in 1988. J.F. Marin/Editing

imputed to the authors, which was that ethnographic artifice needed not only to be acknowledged, but might be relished. Rather than con- tinue to pursue the chimera of truth, we could experiment playfully with genres; the ethnographer’s centrality in the making of ethnography, moreover, licensed a confessional style in which the writer would be increasingly visible.

What was disappointing about the flurry of literature on the theme was the tendency to polarize in a hackneyed opposition between a rigorous engagement with the external world and an indulgent preoccupation with text and self (e.g. Spencer, 1989). Although Writing Cul- ture certainly included, and perhaps encouraged, an introspective, literary habit among some scholars whose literary aspirations were more conspicuous than their literary accomplishments, this was really not what the debate needed to be about. Although novelists and painters may write or talk in interesting ways about their own creativity, one does not look to them for

the most critical or revealing account of their own moment in literature or art. Given that anthropology is predicated on the revelatory potential of the unfamiliar, we should always have known that the richest critique of anthropological writing would not be an auto-critique. What might have followed from the observation that ethnographic texts obeyed various conventions and used various devices to evoke a sense of reality and particularity could have been a historical examination of traditions in anthropological writing and anthropological knowledge (a project in fact taken up by Geertz (1988) and Stocking (1987) and others). In its most positive form, such a critique can address not only the most obvious texts such as those of Malinowski, which were in fact pre-adapted for the project by their self-consciously rhetorical and mannered style, but also other genres such as museum bulletins that are now remote from dominant styles but remain important for their accumulations of data that are still drawn upon

0 UNESCO 1997.

Anthropological epistemologies 34 1

(cf. Thomas, 1989). In other words, the questions ought to have licensed not talk about ourselves, which would lead inevitably to disguised self-justifications, but a richer sense of the diversity of anthropological genres, of the strengths and limitations of descriptive modes at varying times.

Totalities

Over the last twenty years there have been basic reorientations in anthropological thought. Of many that might be cited, I suggest that two have particular epistemological ramifications. The first of these concerns the holism that was long fundamental in the social sciences, and the second the significance of language as a dominant metaphor.

The Enlightenment antecedents of professional anthropological discourse tended to deploy grand concepts such as the idea of the form of government. This was not a narrow classification of political institutions, but rather a far more encompassing notion of national esprit, present in laws, religion, temperament, and the arts: it was essentially a generalized culture-concept. It is useful to remember this, simply because explanation in anthropology has proceeded most commonly ever since by relat- ing some particularity to a total entity. The total entity may be understood as a cultural type, an ethos, a social structure, or a political form; the mode of analysis and the key concepts have of course varied a great deal, but at the crudest level this analytical strategy, that could be called a contextualizing one, has been fundamental (Strathem, 1991).

Since the early twentieth century, the most important context-defining entities have been ‘society’ and ‘culture’, which constituted both larger systems and bounded systems. If the assumption of boundedness has long been ques- tioned, most recently by advocates of the significance of globalization, other notions fundamental to a systemic analysis have come to seem less and less satisfactory. The preoccupation with the construction of culture could only implicate anthropological inventions along with those of our informants, while society, most particularly in Strathern’s arguments (1988), was similarly recast as something other

than a field or container somehow naturally inhabited by actors and relations. In her analysis, Melanesians might busily be evoking collec- tivities through events such as ceremonial exchange, rites, and dances, but these were not social systems as much as rhetorical artifacts - insecure evocations of particular occasions, imaged entities, more like nations in Benedict Anderson’s Imagined Communities (1983) than the societies of conventional anthropological and sociological reference.

If Strathem’s singularly anthropological deconstruction of feminist scholarship and the Durkheimian legacy, particularly through their manifestations in Melanesian ethnography, is a radical project at some remove from much contemporary anthropology, its subversion of the conventional totalizations has much broader relevance. The affirmation of practice, agency, and particular context and circumstance, could indeed be seen as the intellectual zeitgeist of late twentieth-century anthropology. Not only are explicit determinisms, such as those of Marxism, avoided, but so too is a more basic operation of systemic contextualization, that gives analytical value to particular instances through the demonstration that they conform to some cultural or social rule. To be sure, the evocation of generality remains a necessary moment in any ethnographic description, since an account that privileged pure idiosyncrasy or contingency would be biographical or historical rather than anthropological. But the tendency is to no longer treat a systemic or conventional aspect of a particular practice as an end-point of analysis, as an accomplishment that established a regularity. Regularity or convention of whatever kind is instead the ground of innovation, the basis of a transformative action. Hence even structuralism comes to focus on the contexts in which structures are risked and transformed (Sahlins, 1985), and even symbolic anthropology addresses the innovation of meaning (Weiner, 1994).

This is a theoretical sea-change that anthropologists at present see the inside rather than the outside of. We therefore take agency, con- testation, complexity and flux - the dimensions of social life that we understand to have been suppressed by all over-systematizing paradigms, whether they were functionalist, structuralist or Marxist - as hitherto unacknowledged features

0 UNESCO 1997.

342 Nicholas Thomas

that it is self-evidently valuable to discover again and again. But social thought has long tacked back and forth between collectivism and voluntarism, holism and atomism, and systematization and history. I share the sense of many scholars that contemporary approaches in fact do greater justice to the multiple determinations of social life, and the capacities of actors to shape their own circumstances, than the theoretical languages of thirty years ago, but current trends may in retrospect seem to reactively overstate the particular, as preceding generations overemphasized the force of totalities.

Meanings

I have found it useful to refer back to key formulations in Geertz’ influential 1973 essay on thick description, as a measure of where anthropologists have subsequently moved. At the time he wrote, it seemed obvious that what anthropologists, or at least interpretive anthropologists, searched for was ‘meaning’. Analysis discovered significance, communication, coding, and symbolism. Despite the profound differ- ences between Geertz’ anthropology and that of LCvi-Strauss (which Geertz regarded as a quaint pursuit of ageless reason) both presumed that language and a linguistic model were central to social life and its analysis. ‘Thick description’ was a description of discourse, even if Geertz’ famous example concerned not words but gestures, blinks and winks.

In the wake of Pierre Bourdieu’s critique of mechanistic semiotics and affirmation of practice (1977), it was perhaps inevitable that the centrality of language would be challenged. This has not been a unitary theoretical effort, but rather a highly dispersed one, conducted on different fronts in different fields, against textualism and for performance in one context, and against communication and for materiality in another. Studies of embodiment, emotion, material culture, and art have all, in quite different ways, shifted away from what was pre- viously axiomatic: that anything socially consequential or efficacious was perforce meaningful and significant. Even if not understood as mess- age in relation to code, or specifically as a text, a practice or artifact was understood to communicate. Though it would, of course, be

unproductive to deny that language, iconogra- phy, and discourse are tremendously important, it has become increasingly apparent that pres- ences as well as representations, substances as well as significances, doing as well as meaning, are of vital and constitutive importance in most domains addressed by cultural analysis. Surpris- ingly recently, it seemed exciting when theorists in philosophy, literature, and history as well as anthropology pointed to the cultural constitution of the body, as it did when Barthes and others drew attention to the semiotics of consumer goods. Yet subsequent work tends to take us back to the common-sense point that critics had rhetorically distanced themselves from: the body is always more and less than a text, and the values and desires invested in consumer objects depend on their materiality as well as their imputed ‘meaning’.

This may be a theoretical rather than an epistemological issue, but it raises a basic question for future anthropological knowledge. If we have ceased to see gestures such as winks as bearers of implicit linguistic meaning, as practices that stand for something else, as alterna- tives to language that are linguistically explic- able, and instead regard them as distinctive and substantial performances or practices in themselves, we confront yet another translation issue. In their practice, ethnographers have always had to address the problem of adequately rendering foreign idioms in one’s own terms, they now increasingly face the issue of describing and presenting the embodied and the implicit through an explicit analytical language.

The place of theory

Finally, I suggest that the discipline of anthropology must be understood on a different model from the hierarchy of description and analysis, data and theory that is so often presumed in discussion of social and cultural analysis. Whereas engagement with the particular is often understood as a limited activity, circumscribed by philosophical empiricism, and legitimate only as the basis for a higher effort of abstract theorizing, anthropological practice in fact lifts theory from the level of the abstract to the concrete. As, in quite different terms and different contexts, Marilyn Strathern (1988, p. lo),

0 UNESCO 1997.


Michael Herzfeld (1987, pp. 202-205), and others have suggested, descriptive ethnography can be seen as a higher-level, or second-order discourse, that is intelligible only by virtue of its theoretical and analytical grounding. More than any other discipline, anthropology con- stantly reminds its practitioners of the pretensions of our analyses, which may differ from

~ ~~

those of our subjects, but are not obviously privileged or authoritative in relation to them. To acknowledge the formative character of ‘the field’ with respect to anthropological knowledge is not only to prefer practical theory to theoretical practice: it is to realize that one works not with informants, but with co-interpreters.

References

ANDERSON, B., 1993. Imagined Communities. London: Verso.

ASAD, T. (ed.), 1973. Anthropology and the Colonial Encounter. London: Ithaca.

BOURDIEU, P., 1977. Outline of a Theory of Practice. Cambridge: Cambridge University Press.

CLIFFORD, J.; MARCUS, G., 1986. Writing Culture: the Poetics and Politics of Ethnography. Berkeley: University of California Press.

FABIAN, J., 1983. Time and the Other: How Anthropology makes its Object. New York: Columbia University Press.

FABIAN, J., 1991. Time and the Work of Anthropology. Chur and Reading: Harwood Academic Publishers.

FARWN, R., 1990. Localizing Strategies: Regional Traditions of Ethnographic Writing. Washington: Smithsonian Institute Press.

GEERTZ, C., 1973. ‘Thick description: towards an interpretive theory of culture’. In The Interpretation of Cultures. New York: Basic Books.

GEERTZ, C., 1980. Negara: the Theatre-State in Nineteenth-Century

Buli. Princeton: Princeton University Press.

GEERTZ, C., 1988. Works and Lives: The Anthropologist as Author. Cambridge: Polity Press.

HANSON, F.A., 1989. ‘The making of the Maori: cultural invention and its logic’, American Anthropologist, 91, pp. 89Cb902.

HERZFELD, M., 1987. Anthropology Through the Looking Glass. Cambridge: Cambridge University Press.

KEESING, R., 1985. ‘Kwaio women speak: the micropolitics of autobiography in a Solomon Islands Society’, American Anthropologist, 87, pp. 27-39.

KEESING, R., 1989. ‘Creating the Past: Custom and Identity in the Contemporary Pacific’, The Contemporary PaciJic, 1. pp. 19- 42.

MARCUS, G.; FISHER, M., 1986. Anthropology as Cultural Critique. Chicago: University of Chicago Press.

SAHLINS, M., 1985. Islands of History. Chicago: University of Chicago Press.

SPENCER, J., 1989. ‘Anthropology as a kind of Writing’, Man, (ns) 24, pp. 145-64.

STOCKING, G., 1987. History of Anthropology (book series). Madison: University of Wisconsin Press.

STRATHERN, M., 1988. The Gender of the Gi’. Berkeley: University of California Press.

STRATHERN, M., 1991. Partial Connections. Savage, Md.: Rowman and Littlefield.

THOMAS, N., 1989. Out of Time: History and Evolution in Anthropological Discourse. Cambridge: Cambridge University Press.

THOMAS, N., in press. In Oceania: Visions, Artefacts, Histories. Durham: Duke University Press.

TRASK, H.-K., 1991. ‘Natives and Anthropologists: the Colonial Struggle’, The Contemporary PaciJic, 3, pp. 159-67.

WEINER, J., 1994. ‘Myth and metaphor’. In Companion Encyclopaedia of Anthropology: humanity, culture and social life, ed. T. Ingold, London: Routledge.

0 UNESCO 1991

Documents

Anthropological epistemologies