3 FIELDWORK AND ITS INTERPRETATION Theory without data is empty, but data without theory are blind. — C. Wright Mills FIELDWORK Anthropology distinguishes itself from the other social sciences through the great emphasis placed on ethnographic fieldwork as the most important source of new knowledge about society and culture. A field study may last for a few months , a year, or even two years or more, and it aims at developing as intimate an understanding as possible of the phenomena investigated. Many anthropologists return to their field throughout their career, to deepen their understanding further or to record change. Although there are differences in field methods between different anthropological schools, it is generally agreed that the anthropologist ought to stay in the field long enough for his or her presence to be considered more or less 'natural' by the permanent residents, the informants, although he or she will always to some extent remain a stranger. Many anthropologists involuntarily take on the role of the clown when in the field. They may speak strangely with a flawed grammar; they ask surprising and sometimes tactless questions, and tend to break many rules regarding how things ought to be done. Such a role can be an excellent starting-point for fieldwork, even if it is rarely chosen: through discovering how the locals react to one's own behaviour, one obtains an early hint about their way of thinking. We are all perceived more or less as clowns in unfamiliar surroundings; there are so many rules of conduct in any society that one will necessarily break some of them when one tries to talce part in social life in an alien society. In Britain, for example, it is considered uncultured to wear white socks with a dark suit; still, it happens that people who are not fully conversant with the local dress code do so. In the field, anthropologists have been known to commit much more serious mistakes than this. A different, and sometimes more problematic, role that can be assumed by the anthropologist in the field, is the expert role. Many fieldworkers are treated with great deference and respect by their hosts, are spoken to in extremely polite ways and so on, and can thus run the risk of never seeing aspects of society which the locals are ashamed of showing to high-ranking strangers. 24 Fieldwork and its Interpretation 25 No matter which role one takes on in the field - most ethnographers are probably partly expert, partly clown, at least in the early stages - fieldwork is extremely demanding, both in professional and in human terms. The tidy, systematic and well-rounded texts written by anthropologists are more often than not the end-product of long periods in the field characterised by boredom, illness, personal privations, disappointments and frustration: few anthropologists can state squarely that their fieldwork was a continuously exciting journey of exploration, full of pleasant experiences. In a foreign setting, one will usually master the language and the codes for behaviour poorly at the beginning, and one will feel helpless in many situations. Besides, one runs the risk of encountering suspicion and hostility, and it can be profoundly unpleasant for the body to have to cope with an unfamiliar climate, strange food and a different hygienic standard than one is accustomed to. Last but not least, it can be very trying for people with a middle-class 'Western' background (which is that of most anthropologists) to adapt to societies where being alone is considered a pitiful or pathological condition. Plainly put, in many village settings one is never left alone. This problem does not usually exist for the growing number of anthropologists who carry out their fieldwork in modern urban settings. In their case, the problem may be the opposite: in societies where people have TV sets and cars, and where time is considered a scarce resource, an ethnographer may quickly discover that his or her presence creates neither excitement nor curiosity among the natives, and that continuous immersion in local life is difficult. Urban fieldwork tends to be more discontinuous than village fieldwork, and often depends on more formal methods, such as the structured interview. Even fieldwork in 'exotic' settings should not be unduly romanticised. Among some Native American peoples in North America, a new profession has emerged in recent years: that of the professional ethnographic informant. Some cultural specialists may thus charge handsome fees for spending their time explaining the intricacies of myths and customs to visiting ethnographers. IN THE FIELD A principal requirement in fieldwork nonetheless consists of trying to take part in local life as much as possible. Anthropologists also use a variety of specified, formal techniques for the collection of data (see, for example, Pelto and Pelto 1970; Ellen 1984). Depending on the kind of fieldwork one is engaged in, structured interviews, statistical sampling and other techniques may be required to varying degrees. Most anthropologists depend on a combination of formal techniques and unstructured participant observation in their fieldwork. Participant observation refers to the informal field methods which form the basis for most fieldwork, whether or not it is supplemented with other 2 6 Small Places, Large Issues techniques. The aim of this method is to enter as deeply as possible into the social and cultural field one researches; in practice one becomes, as Evans-Pritchard remarks (1983 [1937], p. 243), a 'doubly marginal' person, in a sense suspended between one's own society and the society under investigation. During participant observation, one tries to immerse oneself in the life of the locals and tries not to be noticed, so that they can carry on with their own lives as usual. In this regard, the issue of hidden versus open observation has been discussed in the anthropological community. Generally, it is agreed that it would be unethical not to inform one's hosts what one's mission consists of. The people explored must have the right to refuse to be subjected to anthropological analyses; in the case of hidden observation, they are deprived of this possibility. There are many ways of doing fieldwork, and it is impossible to provide a clear recipe for how to carry it out. For one thing, the anthropologist him- or herself is the most important 'scientific instrument' used, investing a great deal of his or her own personality in the process (see Hastrup 1995). Another source of variation is the greatly differing settings and topics investigated by anthropologists. The methods must be tailored to fit the requirements of the subject, but it is difficult to be more specific. Evans-Pritchard once recalled his first attempts to learn about fieldwork in the early 1920s (1983, pp. 239-54). He had asked a number of renowned anthropologists how to go about doing it and had received various answers. First, he asked the famous Finnish ethnologist Westermarck, who said, 'Don't converse with an informant for more than twenty minutes because if you aren't bored by that time he will be.' Evans-Pritchard comments: 'Very good advice, if somewhat inadequate.' Alfred Haddon said 'that it was really all quite simple; one should always behave as a gentleman'. Evans-Pritcharďs teacher, Charles Seligman, 'told me to take ten grams of quinine every night and to keep off women'. Finally, Malinowski himself told the novice 'not to be a bloody fool'. Evans-Pritchard himself emphasises, later in the same account, that facts are themselves meaningless; in other words, 'one must know precisely what one wants to know' and then fashion a suitable methodology from the available techniques. There is - alas - no simple recipe for fieldwork. Many anthropological accounts of the process of inquiry, and especially fieldwork, are probably strongly idealised. The expression 'participant observation', a vaguely defined research technique, may serve as a convenient blanket term to conceal both ethical and methodological shortcomings in the actual research process. Many ethnographers probably develop a profoundly ambivalent, sometimes even antagonistic, attitude towards the people they study. When Malinowski's private diaries were published more than twenty years after his death (Malinowski 1967), they spurred a long and heated debate, particularly in the United States. Malinowski, who was and is regarded as an outstanding ethnographer, turned out to have a less than flattering view of his hosts in the Trobriand Islands. His project consisted of understanding the islanders in their own Fieldwork and its Interpretation 27 terms, yet he personally regarded them as unwashed savages and frequently, he reports, had to force himself to leave his hut in the morning for fieldwork. The question which has been raised in this context is whether it is possible to carry out good fieldwork among people one has this little respect for. The answer is obviously yes; and as long as one does not molest one's hosts in the field, there can be no rules against negative attitudes on the part of the anthropologist. At the end of the day, the value of participant observation lies in the quality of the empirical data one has collected, not in the number of close friends one has acquired in the field. Common problems in fieldwork can be limited knowledge of the field language, gender bias (see Chapter 9) or the fact that one's main informants fail to be representative of the society as a whole. Concerning the latter point, it may be true that anthropologists have tended to pay too much attention to the elite of the community (although elites in complex societies are understudied, partly because access to them is difficult). Frequently chiefs, teachers and other untypical individuals are most efficient in offering their services to a visiting anthropologist, and the anthropologist may also unwittingly be attracted to this kind of people because they resemble him- or herself. Gerald Berreman (1962) once wrote a confessional piece on his own field experiences in North India. He depended on an interpreter, and only after he had been working, rather unsuccessfully, for a while did it become evident that the interpreter was a significant source of distortion in his fieldwork; not because he lied or was inadequate as an interpreter, but because of his position in the caste hierarchy. People did not talk as openly to him as they would to someone like Berreman himself, who was caste-neutral by virtue of being foreign, or to a local with a different caste membership. Fieldwork does not have to be either capital-intensive or labour-intensive: as a research process it is cheap, since the only scientific instruments involved are the fieldworkers themselves and possibly a few assistants. However, and this is perhaps the main point about fieldwork as a scientific method, it is time-intensive. Ideally, one should stay in the field long enough to be able to see the world as the locals see it. Even if this may be impossible, among other reasons because one cannot entirely get rid of one's own cultural background, it can be a worthwhile aim to pursue. The strength of the anthropologist's knowledge can thus be said to lie in his or her mastery of both the local culture and a different culture (his or her own), and of tools of analysis, which makes it possible to give an analytical, comparative account of both. The strength of ethnographic field method can also be its weakness: it is demanding, and rewarding, partly because the ethnographer invests not only professional skills in it, but also interpersonal skills. The ethnographer draws on his or her entire personality to a greater extent than any other scientist. For this reason, many emerge from the field exhausted, but with a material of extraordinary richness and depth. At the same time, this degree of personal involvement has important ethical implications. Are friendships 2 8 Small Places, Large Issues and other confidential relationships developed in the field 'real' or 'fake'; what are the moral obligations of the ethnographer vis-ä-vis the informants? The AAA (American Anthropological Association) and other professional bodies, as well as many university departments, have developed ethics codes for the protection of informants, who may both disapprove of the 'anthro-pologification' of their personal concerns and of their 'close friends from abroad' who suddenly vanish, never to return. A different set of problems concerns professional bias caused by personal biography, which may lead ethnographers to see only those parts of social reality that make sense in terms of their earlier experiences. On the other hand, existential involvement in one's own research can also improve the quality of the work (see Okely and Calloway 1992). In many cases, fieldwork is as profoundly personal as it is professional, and most anthropologists probably feel a lifelong attachment to their first field site. A topic rarely talked about, but probably not uncommon, concerns sexual relations between anthropologist and informant. An edited volume devoted to this topic (Kulick and Wilson 1995) is entitled, characteristically, Taboo. THEORY AND DATA The relationship between theory and empirical material, or data, is fundamental in all empirical science, including anthropology. No science can rely on theory alone (it then becomes pure mathematics or philosophy), just as it cannot rely on pure facts: in that case, it would be unable to tell us anything interesting. To put it differently, research has an inductive and a deductive dimension. Induction consists of going out there, 'watching and wondering', collecting information about what people say and do. Deduction consists of attempts to account for facts by means of a general hypothesis or theory. Suppose I were to explore the hypothesis that the rank of women in society is proportional to their contribution to the economy (see Chapter 9). Working deductively, I would develop an argument showing why this made sense. In the actual research process, however, I would have to shift to an inductive mode, exploring the relationship between the position of women and the economy in a number of existing societies. As soon as I came across one or several societies where there was no apparent relationship between the contribution to the economy and the relative rank of women, I would have to modify my initial hypothesis. We may envisage the search for this kind of general insight as a zigzag movement between the observation of fact and theoretical reasoning, where new facts modify the theory and (modified) theory accounts for the facts. Each time one shifts from theory to description of empirical process and back, one's insight has become a little bit more accurate. If one were to reproduce everything one's informants said and to describe everything they did, one would be unable to falsify, or for that matter justify, Fieldwork and its Interpretation 29 specific hypotheses. One would virtually drown in details without being able to present patterns and regularities. The description of society would be as complex and ambiguous as society itself and therefore superfluous. The anthropological project consists, to a great extent, of imposing ordering patterns and regularities on to the observed material, and we depend on our own theoretical abstractions in order to do so. The challenge lies in saying something significant about culture and social life through these abstractions. The choice of an precise topic for investigation, therefore, is an important part of the preparation for fieldwork. One should, at the very least, know if one is interested in, say, resource management or child-raising before embarking on fieldwork. Otherwise one will end up knowing too little about everything rather than knowing enough about something. Godfrey Lienhardt (1985), borrowing an analogy from Geertz, has compared the relationship between theory and ethnography to an elephant-and-rabbit stew. What is required, says Lienhardt, is one elephant of ethnography and one rabbit of theory. The art, as he sees it, consists of bringing out the flavour of the rabbit. ANTHROPOLOGY AT HOME Anthropology has traditionally distinguished itself from sociology through (1) the great emphasis placed on participant observation and fieldwork, and (2) through studying chiefly non-industrialised societies. Sociology has concentrated on understanding, criticising and managing modern societies, whereas the historical task of anthropology has been to account for the variations and similarities in human existence and, to some extent, to rescue disappearing peoples from oblivion by recording their way of life in writing. For a number of reasons, fieldwork in the anthropologist's own society or a neighbouring one has become much more common since the early 19 60s. First, our discipline today faces a number of new challenges because of historical changes in the world, including the virtual disappearance of 'the tribal world'. It has become impossible to posit sharp distinctions between 'us' (moderns) and 'them' (primitives), not least because modernisation and 'development' have contributed to shrinking spatial distances and have blurred boundaries between cultures which formerly seemed relatively clear. What is 'home' and what is 'abroad' is no longer always clear. Second, the analyses of tribal societies have inspired researchers to use similar analytical models when dealing with their own society, and have also provided a useful basis for comparison. It is easier to see what is unique to our own society when we have intimate knowledge of other societies than it would otherwise. Third, there are today many researchers competing for scarce research funds, and far from everybody is able to raise funding for long-term fieldwork in a remote place. In addition, a number of governments in the Third World 30 Small Places, Large Issues are sceptical of anthropologists. Anthropology is no longer a 'science of the tribal' or of the 'non-industrial world', but it remains a truly global science which may just as well study Internet use in Trinidad (Miller and Slater 2000) as Hindu nationalism (Van der Veer 1994), sacrifice in eastern Indonesia (Howell 1996) or ethnic complexity in Britain (Baumann 1996). One argument sometimes used against fleldwork in one's own kind of society is that the overall aim of the discipline is to account for cultural variation in the world. It therefore seems reasonable that one should study people who seem culturally remote. Another argument is that it is necessary to use our own society as an implicit basis for comparison, something which vanishes when we study our neighbours. On the other hand, thorough field studies of 'modern' societies have revealed that they are far more heterogeneous, in terms of culture and social organisation, than is generally assumed. Also, the distinction between 'self and 'other' is no longer unproblematic. A German ethnographer may, in important respects, have more in common with middle-class, urban Kenyans than with neo-Nazi skinheads from his or her own hometown. A general argument in favour of anthropological research 'at home' is that the most fundamental questions we ask about culture, society and so on are equally relevant anywhere in the world. One of the grand old men of social anthropology, Sir Raymond Firth, expressed his own view in a lecture given on the future of anthropology in 1989: 'Since we can explore the anthropological problems anywhere, we might as well go to places where it is comfortable to spend some time' (Firth 1989). For his own part, Firth has carried out much of his fieldwork in Tikopia, a tropical island in the Pacific (but, it must be conceded, he has also done fieldwork in England). As a matter of fact, today's anthropology encompasses the whole world, including the areas which anthropologists call home. Fieldwork at home, like anywhere in the world, depends on the anthropologist's professional skills. In a familiar or semi-familiar setting, one has the advantage of mastering the language and cultural conventions better than in a culturally distant place, but one also tends to take too much for granted. This problem is sometimes called 'homeblindness', and it can, at least to a great extent, be overcome through proper training. The comparative, detailed study of cultural variation which forms the core of the education of an anthropologist enables us to study societies we believe to be familiar with roughly the same methods and analytical apparatus we would apply to distant societies. INTERPRETATION AND ANALYSIS When modern anthropology took shape in the early years of the twentieth century, there were still large areas of the world that had hardly been visited by Europeans, much less been subjected to systematic exploration. When Boas studied the Kwakiutl and neighbouring peoples on the west coast of Fieldwork and its Interpretation 31 North America, when Malinowski lived in the Trobriand Islands, when Bateson visited the Iatmul in New Guinea and when Evans-Pritchard went to live among the Azande, they could only to a limited extent prepare themselves through reading earlier studies of the same people. They knew relatively little about the sites they were going to. The world has changed quite dramatically since then. There now always exist studies from the region one visits as ethnographer, although it may still be possible to come across smaller groups, for example in the Amazon basin or New Guinea, which have not been studied anthropologically. Frequently, one can learn the language before leaving home, and there normally exists a large specialist literature devoted to the region, which is it usually necessary to study before embarking on fieldwork. The classic anthropological monograph, or the most common type of -anthropological book from the 1920s to the 1950s, dealt with a people's most important institutions, usually taking detailed village fieldwork as its point of departure. It frequently aimed at a comprehensive overview of 'the way of life' of a people, describing the interrelationship between religion, politics, the economy, kinship and so on. For a number of reasons, this model is no longer common. One obvious reason is that most anthropological studies now take place in complex large-scale settings and are not limited to villages. Another reason is the growing specialisation within the discipline, which has turned many professionals into highly specialised sub-disciplinary experts focusing, say, on medical systems, socialisation, public rituals or political rhetoric in particular societies. A third, related reason is the fact that a wealth of general ethnography has already been carried out on most regions, so it is frequently unnecessary to begin from scratch. When Annette Weiner left for fieldwork in the Trobriand Islands in the 1970s, therefore, she did not study every aspect of the Melanesian island society in detail: Malinowski and others had already done so, and she could relate her own work to lacunae or controversial points in earlier analyses of the Trobrian-ders (Weiner 1988). Her studies deal chiefly with the role of women and gender relations, and they engage in a critical dialogue with earlier work in the area (see Chapters 9 and 12). In addition, social theory is now being produced from within many of the societies studied by anthropologists. Sociology, anthropology and other theoretically informed texts are being written by the grandchildren of Radcliffe-Brown's and Kroeber's informants. This implies that today's anthropologist may engage in a dialogue with the society explored to a much greater extent than her predecessors did. It also implies that anthropological studies may affect local communities directly. If someone today writes a monograph on, say, a South African neighbourhood, the book will necessarily influence South African society; it will be read by some of the 'natives', and thus becomes part of the social reality of the informants. This situation has created obvious ethical problems. It would not have occurred to Malinowski or Bateson that their books on Melanesia might have a direct 3 2 Small Places, Large Issues influence on the societies in question (although Malinowski's work has for a long time been well known in the Trobriand Islands): they could write freely without taking such issues into account. This is no longer possible, at least if one's work is published in English or French. THE ETHNOGRAPHIC PRESENT AND THE PAST Anthropological texts are usually written in the present tense. Many of the most influential monographs were nevertheless written half a century ago or more, and in virtually every case the societies they deal with have changed radically since the original fieldwork took place. Frequently, moreover, fleldwork was carried out during an unusual, not a 'typical', historical period. For example, classic African anthropology was developed during the last phase of French and British colonialism, namely between the First World War and 1960. It has since been pointed out that this period was atypical because it was a time of political stability - which has been absent both before and afterwards in large parts of Africa. Social anthropology has never tried to replace history. Anthropological analysis has traditionally been focused on social and cultural interrelationships at a particular point in time and, until recently, rarely emphasised the historical processes which have led up to the present. In the British, American and French traditions, the aim has usually been to account for the workings of a particular society or culture, not to try to explain how it emerged. Boas, Radcliffe-Brown and Malinowski were all critical of the rather speculative forms of cultural history which preceded modern anthropology. However, an appropriate response to the bad cultural history exemplified in much early anthropological writing would not be to discredit history as such, but rather to improve one's historical accuracy - as witnessed at an early stage in the work of a fourth 'founding father', namely Mauss, whose essay on gift exchange draws extensively on well-documented historical material from Norse, Indian and Roman society. Although the historical dimension has become more important in recent anthropological work, particularly since the 1980s, a majority of anthropological studies could probably still be described as synchronic 'snapshots'. We may use the term 'the ethnographic present' to characterise the literary tense involved. In a certain sense, it can be irrelevant when the Nuer, the Trobriand islanders or the Swat Pathans enact a certain culture or form of social organisation as described in a monograph. The importance of the studies of these peoples does not lie primarily in their historical or genealogical explanatory power, but rather in their contribution to our understanding of differences and similarities of social life in the widest sense. They contribute to our comparative knowledge of forms of human life. As Kirsten Hastrup argues, the ethnographic present does not imply that timelessness is a feature of other societies, 'but we do stress that ethnographic knowledge transcends Fieldwork and its Interpretation 33 the empirical' (1992, p. 128) in that it deepens our understanding of the human condition in general. There is also a clear methodological advantage involved in the synchronic study of social life. Anthropology may be described as the process whereby one wades into a river and explores it as it flows by, whereas historians are forced to study the dry riverbed. One cannot engage in participant observation of the past. On the other hand, many anthropologists have followed the lead of Kroeber and Evans-Pritchard in stressing the importance of knowing the history of a society and its contribution to the present. This can be especially rewarding - some would say absolutely necessary - in studies dealing with societies with a written history. Further, the connections between different societies, which are often crucial for the understanding of each society, can only be properly investigated historically (see Wolf 1982). It would be impossible to understand, for example, the Industrial Revolution in England properly without prior knowledge of the slave trade and the cotton plantations in the United States. In sum, the ethnographic present and the historical dimension should not be seen as mutually exclusive. The critics of diffusionism, who correctly pointed out the importance of studying societies and cultures as more or less integrated systems, tended to overemphasise the relative isolation and unchanging character of societies. To the extent that historical sources are available, they doubtless make important contributions to the contemporary understanding of single societies. WRITING AND READING ETHNOGRAPHY Although fieldwork remains the most important method of generating new knowledge in anthropology, the transmission of knowledge within the professional community generally takes place through the writing and reading of texts. Geertz claims, in a study of the writings of several prominent practitioners, that the most characteristic activity of anthropologists is writing, and he therefore calls attention to the way in which such texts are produced (Geertz 1988). Far from being neutral and objective descriptions and analyses of customs and cultural systems, anthropological writings are shaped by each author's biography, literary style and rhetoric, as well as by the historical period in which they were written (such as colonialism) and, of course, by the character of the fieldwork. These aspects of the production of anthropological knowledge have often been understated or dismissed as irrelevant to the end-product, the monograph or academic article. The approach to ethnography exemplified in Geertz's book may seem to undermine the authority of these works as sources of knowledge about society and culture, and to reduce them to mere literature. This was not Geertz's intention although it may have been an unintended side-effect: he argued for more professional reading, offering, through his examples, 34 Small Places, Large Issues contexts of interpretation enabling the reader to appreciate the full significance of ethnographic texts. For example, he shows how Firth establishes his ethnographic authority by starting his famous 1927 monograph on the Tikopia with a lengthy descriptive literary passage 'marshaled with Dickensian exuberance and Conradian fatality' (Geertz 1988, p. 13) to communicate his familiarity with the society he then goes on to describe in strict sociological terms for the following 500 pages. This observation does not mean that Firth's work is invalid, but that a different anthropologist would have written a different book - in other words, that the anthropologist as author is situated in the field and in his or her own text. Several anthropologists (or meta-anthropologists) have in recent years followed Geertz's lead in applying techniques from literary criticism to ethnographic writings (see Clifford and Marcus 1986; Manganaro 1990; James et al. 1997) and have suggested, for example, that Malinowski's style was indebted to Joseph Conrad, that Evans-Pritchard's representation of Africans was tainted by his colonial background and attitudes, that Ruth Benedict's study of the Japanese, The Chrysanthemum and the Sword (Benedict 1974 [1946]), had to be understood in the context of the Second World War and not as a research monograph, and so on. A possible implication of this focus on narrative strategies and implicit agendas is that anthropological studies tend to be persuasive rather than convincing; that they evoke more than they describe; that they are shaped by the author's personal bias, not by the studied society; and that they create a 'suspension of disbelief in the reader not so much because of the data presented but because of the author's style and rhetoric. The recent interest in the writing of ethnographic texts does not necessarily lead to a conclusion this radical. Geertz's aim was not to dismiss ethnography as fiction, but rather 'that we shall learn to read with a more percipient eye' (1988, p. 24). Some understanding of literary techniques and the importance of personal and historical contexts in the production of knowledge can in the long run only lead to more accurate comprehension than a naive reading would. Such insights into the social conditions of the production of knowledge led, in the 1980s and 1990s, particularly in the United States, to a proliferation of 'experimental ethnographies' that tried to solve the problem of representing others in novel ways (see Clifford and Marcus 1986; Marcus and Fischer 1986), as well as 'postcolonial ethnographies' that are either written by members of the formerly colonised peoples or by foreign anthropologists arguing against previous understandings of the areas in question, which they see as informed by a colonial way of thinking or at least by exoticism and stereotypes (see, for example, Guha andSpivak 1988; Gupta and Ferguson 1997). THE PROBLEM OF TRANSLATION In the earlier discussion of ethnocentrism, I remarked that we would not arrive at a satisfactory understanding of a society if we were to evaluate its Fieldwork and its Interpretation 35 achievements in relation to the standards and values of our own society. It therefore becomes a central challenge to connect our analytical concepts to the alien social and cultural world we deal with. The related problems arising from this project are frequently spoken of as the problem of translation. How can we translate an alien way of experiencing the world into our own mode of thought; how can we be certain that we do not misinterpret or distort the society when we try to describe it in our own terms? And how can we be entirely certain that we understand the alien society and culture at all, imbued as we are with our own cultural background, concepts and values? These interrelated problems are fundamental to anthropology, and they are dealt with extensively in Chapter 15. For now, we will restrict ourselves to an outline of some dimensions of the problem. Within the discipline of anthropology, it is necessary to use abstract terms such as kinship, social organisation, social control, religion and so on. These terms are necessary for the discipline to be comparative in its scope: how could it be possible to compare, say, the kinship system of the Trobrianders with that of the Yanomamö if we did not have a general concept of kinship? However, the abstract, technical terms used by anthropologists exist only rarely in the societies we study: they form part of our world, not theirs. How, then, can we justify accounting for an alien society in terms which are demonstrably not its own, if the aim of anthropology is to understand societies and cultures from within? There are several possible solutions to this seeming paradox. A first step could be to distinguish between description and analysis. The descriptive aspect of an anthropological account is usually close to the native conceptualisation of the world, and a major challenge lies in translating native concepts into the anthropologist's working language. When describing a social and cultural life-world, the anthropologist will often resort to direct quotations from informants, to give an account of the world as it appears from within. The analysis, on the other hand, will try to connect the society, at a theoretical level, to other societies by describing it in the comparative terms of anthropology. It will, in other words, describe the society with concepts which do not exist in the society itself. What kind of 'kinship system' do the people have? How do they resolve 'conflicts'? What is the 'division of labour between men and women'? What is the role of 'religion and rituals'? How are 'power relations' structured? The suggested distinction between description and analysis, although helpful, is not absolute. Even the description is necessarily shaped by the anthropologist's selection of facts and own interpretations, and he or she can never become a native. In addition, it would be plainly impossible to include everything in a description, even if one's aim is to provide a comprehensive account of a society and one has a thousand pages at one's disposal. 36 Small Places, Large Issues EMIC AND ETIC Let us look a little more closely at the relationship between the view from within and the view from the outside. Ethnographic description lies closer, in a sense, to the world as experienced by the informants than the analysis does, as the latter may ultimately aim at general statements about culture and society. This level - life as experienced and described by the members of a society themselves - is sometimes spoken of as the 'emic' level. Its counterpart, the analytical descriptions or explanations of the researcher, is the 'etic' level. The emic-etic dichotomy was introduced into anthropology by Marvin Harris (1964,1979), but it was first developed by the linguist Kenneth Pike (see Headland et al. 1990), who derived the terms from the linguistic distinction between phonetics and phonemics, referring to the objective relationship between sounds and the meaning of sounds, respectively. The 'native's point of view' is emic, whereas the analytical perspective of the anthropologist is etic. However, even if the anthropologist aims to reproduce reality the way it is perceived by the informants, there are three reasons why the result may never be an emic description. First, we must usually translate between two different languages, and the translation is different from the original. Second, we use a written medium to reproduce oral statements, and the meaning of utterances changes when they are transformed into writing. Third, the anthropologist can never become identical with the people he or she writes about. The only truly emic descriptions possible in anthropology are therefore accounts written by natives in their vernacular. It is a common assumption that emic perspectives are wrong, whereas the etic perspectives are correct. This is an unfortunate way of framing the issue. The point is not whether the 'natives' or the 'scientists' are correct, but rather that social scientists have specialised interests and that the kinds of insight they aim at are frequently not identical with the interests of their hosts in the field. There are a number of equally correct ways of describing a cultural and social system; one's choice must depend on one's interests. A further misunderstanding regarding the emic-etic distinction amounts to the idea that emic notions are 'concrete' whereas etic notions are 'abstract'. This may be the case, but it is not inherently true. As Geertz (1983) reminds us, many of the studied peoples use highly abstract, or 'experience-distant', concepts such as 'God', 'witchcraft', 'mortgage interest' or 'karma'. However, a necessary condition for anthropological research to be meaningful at all is that the researcher knows something essential that the native does not know. He or she must have the ability to connect a local reality to a comparative conceptual apparatus, enabling a particular society to shed light on other societies and contributing to the growth of our total body of knowledge about social and cultural variation. There are 'strong' Fieldwork and its Interpretation 11 and 'weak' programmes in this regard: some anthropologists plainly see themselves as kind of natural scientists in search of general theories and laws of culture and societies, while others are more strongly concerned with elucidating dimensions of a single society in great detail, doubting the validity of theories with a highly general scope. What all of them have in common is, nevertheless, an interest in the similarities and differences of forms of human existence, and confidence in the ability of anthropology to say something meaningful about it. ANTHROPOLOGY AS POLITICS Unlike university subjects like law and medicine, anthropology does not lead to a particular profession, and there are anthropologists practising in a growing number of professions. Only a minority teach and do research at universities. Many are involved in development cooperation, and an even larger number work in public administration. There are also anthropologists working in publishing, in private enterprises, in hospitals, in the media and so on. In other words the discipline may be useful in a variety of professional practices. Here we focus on the production of anthropological knowledge in the most powerful milieux of the profession, namely where the authoritative texts - monographs and articles - are written and evaluated, where appointments are decided and where reading lists are made; in short, the arenas where the discipline is being defined. Anthropology, arising out of a particular kind of social environment, can itself be studied anthropologically. The knowledge offered to students is not developed in a social and cultural vacuum, and theoretical as well as empirical directions are decided through social cooperation, competition, the search for personal prestige and political decisions. Students must, for their part, be socialised (see Chapter 4) into a particular mode of reasoning and style of writing in order to succeed. Anthropology may, in this way, be regarded as a social and cultural field; and like any social phenomenon, it entails power disparities (see Bourdieu 1988 for a critique of academia along these lines). This fact is frequently pointed out, and criticised, by students, who form the least powerful group. In the 19 70s it was thus commonplace for students all over the Western world to form informal 'counterculturaľ groups where they analysed and criticised what they saw as an inherent ideological bias in the subject. Notably, they argued that a feminist perspective was lacking, and that traditional anthropology was incapable of analytically coming to terms with coercion and exploitation. In this period, the dominant professional priorities were challenged. Above all, the critics stressed that anthropology, as a human science, is intimately related to society as such and that it is therefore influenced by power interests in society. Some Third World critics claimed that anthropology was simply an extension of the colonial ideology, trying 38 Small Places, Large Issues to subjugate non-white peoples by incorporating their way of life into a Western body of knowledge. A related issue was the question of how knowledge advances. Are research environments fundamentally self-critical and ruthless in their impartial search for knowledge (Popper 1968 [1959]), or are they rather deeply conservative, since every new idea threatens to challenge their own claim to authority (Kuhn 1962)? The radical students, and other critics of the anthropological practice, did contribute to a transformation of the discipline, even if it remains true -perhaps necessarily - that it is being defined by a professional power elite. What is most remarkable about the kind of autocritique referred to is perhaps the willingness to apply insights from social science to one's own situation. Finally, something must be said about the anthropological production of knowledge as a social process. Although fieldwork is emphasised as the main source of new insights, the production of anthropological understanding mainly takes place at universities and research institutes. The exchange of knowledge between anthropologists occurs at international conferences and through professional journals, doctoral theses and books. Since many anthropologists compete over prestige and power within their professional environment, a rather frenzied rate of publishing can be observed in parts of the international anthropological community. The American maxim 'Publish or perish!' is as valid among anthropologists as in other scientific disciplines. In other words, it may occasionally be the case that the 'love of knowledge' programmatically underlying scientific publishing is not necessarily the main motivation for publishing. Of course, the discipline of anthropology is about understanding social systems and cultural variation, not about the professional careers of individual anthropologists. We should nevertheless keep in mind that this knowledge is not being produced in an ivory tower inhabited by pure vocational spirit; that fallible human beings of flesh and blood are responsible for the advancement of analysis; and that our modes of analysis, claiming universal applicability, should occasionally be applied to ourselves as well asto 'theOthers'. The population explosion which has taken place among anthropologists since the 1960s, perhaps especially evident in anthropological publishing, has made it difficult for any single individual to follow every development within the subject. The number of professional journals and monographs published annually is enormous, and we have also witnessed an increasing specialisation within the discipline. Through the development of a growing number of sub-disciplines, such as regional specialisations, medical anthropology, symbolic anthropology and development anthropology, the subject may seem threatened by fission. In the 1940s it was relatively easy to acquire an overview of the discipline; there were perhaps only 30 or 40 canonical books to relate to. Today the number of relevant studies is enormous, and specialisation is inevitable. However, in this kind of situation it is more important than ever to retain a common core of shared concepts and some Fieldwork and its Interpretation 39 shared knowledge of culture and social systems if the discipline is to remain one. This book aims at presenting - and critically engaging with - the bulk of this common foundation, which makes it possible for anthropologists everywhere to have informed professional exchanges, even if their specialisations are very different. If we lacked such a shared professional language, we would not have a common discipline, nor would we be able to exchange ideas and experiences across the many specialisations and emerging sub-disciplines. SUGGESTIONS FOR FURTHER READING Gerald Berreman: Behind Many Masks: Ethnography and Impression Management in a Himalayan Village, Monograph no. 4. Indianapolis, IN: Society for Applied Anthropology 1962. Kirsten Hastrup: Á Passage to Anthropology: Between Experience and Theory. London: Routledgel995. Don Kulick and Margaret Wilson, eds: Taboo: Sex, Identity and Erotic Subjectivity in Anthropological Fieldwork. London: Routledge 1995. C.W. Watson, ed.: Being There: Fieldwork in Anthropology. London: Pluto Press 1999.