THE COMPARATIVE ADVANTAGE OF ANTHROPOLOGY

Despite general agreement that there is a variety of materials (film, video, audiotapes, music, sermons, documents) that anthropologists are working with and a variety of methods applied to these materials, the empirical part of anthropology continues to be primarily based on the ethnographic method, which consists of a long stay "in the field," a familiarization with the "way of life" of the people one studies, and thus in many cases a long linguistic preparation to acquaint oneself with the local language. The rather vague term way of life suggests a "holistic" approach in which a society is examined in its entirety. As one knows, it is practically not possible to have such a total vision, unless one fantasizes societies as outside history and outside a larger world of interaction. In the early stages of the development of anthropology, students of small-scale societies may have been tempted to isolate their fieldwork sites, but no student of an Indian or a Chinese village would have been able to sustain such a fantasy (despite the questionable focus on the "village" as a self-contained unit in the 1950s and 1960s). What ethnographers do claim, however, is that their study of everyday life in a small setting allows them to interpret a larger entity (local, regional, national, or even global) and that that knowledge cannot be gained through the deployment of large-scale surveys. The claim here is that through close study of a fragment one is able to comment on a larger whole and that an understanding of the larger whole allows one to interpret the fragment (whatever that particular whole or fragment may be). It is the choice of a particular fragment of social life that determines what its relevant context is or what the larger question is that one wants to address. While this resembles the hermeneutic circle of textual interpretation, one needs to recognize that social life is not a "social text" and certainly not a closed text and that the openness of social change is multidirectional. Similarly, one needs to steer clear of a universalizing approach that first defines some kind of essence, like "ritual" or "prayer," and then studies it comparatively across cultures.

At this point it is important to emphasize that what I am suggesting is not to be misunderstood as a process of generalization from the particular. The purpose is not to come to some general truth but to highlight something that is not general, something specific without any pretense to general truth, but definitely of broader significance. What is general is often banal, and while anthropologists deal with ordinary life, they strive to say something about it that is not banal. This is also the reason why anthropologists, however much they are prepared when going on fieldwork, generally have to change the questions they ask and the general direction of their fieldwork while confronted with the real-life situations in the field. Moreover, these situations are in fact processes and can only be grasped in dynamic terms. They change during and after the fieldwork, and this process implies that the findings are often difficult to replicate in later fieldwork by other fieldworkers. This problem is compounded by the fact that ethnographers do not ask the same questions and thus come up with different pictures of the societies they study. There is a significant difference not only between ethnographers who continue to study a particular society for most of their professional lives (for example, Kenneth Dean on Fujianese society or Chris Fuller on South India) and those ethnographers who spend a fieldwork period in one site after the other (the prime example is Fredrik Barth on the Pathan, on Oman, on New Guinea, on Bali, on Bhutan), but also in the kinds of questions asked. Anthropologists study historically evolving, open configurations, and the anthropologists themselves are part of them.

The emphasis on fieldwork should not be taken as a far-reaching empiricism in which it suffices that the minutiae of social life are recorded in exasperating detail. As Marilyn Strathern wrote in 1998, "if at the end of the twentieth century one were inventing a method of enquiry by which to grasp the complexity of social life, one might wish to invent something like the social anthropologist's ethnographic practice." Such a purpose cannot be sustained by precision of observation and description (extended case study or situational analysis) alone or by claims to producing a representative sample. The perspective I offer is obviously close to that of Clifford Geertz in his famous essay on "thick description," where he argues that the ethnographer seeks to give an interpretation of "what our informants are up to, or think they are up to." Geertz also offers an enticing example of what comparison can offer in his essayistic comparison of Islam in Morocco and Indonesia. What I do not agree with, however, is the Geertzian penchant for generalization, as in his definition of religion as a universal entity (as a "cultural system"), a definition that fails to reflect on the nature of the global Western political expansion that enabled the universalization of this concept as a way to explain and govern practices all over the world, or in what in his work on Indonesia he calls "the general Javanese religious system."

The move from fragment to larger insight is a conceptual and theoretical one and not a form of generalization. It does not come from mere observation but is theory-laden. Theory should here be taken in its original sense of observing and contemplating. This is not theory as generalization, as in "a general theory of action" (Parsons) or a "theory of practice" (Bourdieu). Therefore, I take the concept of "holism" to refer not to a form of generalization or to the ethnographical method per se but to anthropology as a conceptual engagement with a fragment that allows us to interpret another conceptual universe, in which translation plays a central role. Some observational methods in microsociology, urban geography, and actor-oriented political science resemble the ethnographic method but do not share this radical theoretical orientation of anthropology.

Anthropology's starting point is to question the universality of what in modern society is taken to constitute the separate domains of the "economy" of "politics," of "law," and of "religion," as well as the dichotomy between state and society or between the individual and society, or even between "inner feelings" and "outward appearance." In fact, these pervasive dichotomous conceptualizations have a particular history in modern Western societies and languages. The "holistic" perspective of anthropology allows us to...