This strategy of defining itself by contrast to adjacent and antecedent discourses limits ethnography’s ability to explain or examine itself as a kind of writing. To the extent that it legitimates itself by opposition to other kinds of writing, ethnography blinds itself to the fact that its own discursive practices were often inherited from these other genres and are still shared with them today. At times one still hears expressed as an ideal for ethnography a neutral, tropeless discourse that would render other realities "exactly as they are," not filtered through our own values and interpretive schema. For the most part, however, that wild goose is no longer being chased, and it is possible to suggest that ethnographic writing is as trope-governed as any other discursive formation. This recognition is obviously fundamental for those who are interested in changing or enriching ethnographic writing or simply in increasing the discipline’s self-understanding. In this essay I propose to examine how some tropes of ethnographic writing are deployed and how they derive from earlier discursive traditions. In this essay, I propose to focus on the vexed but important relationship between personal narrative and impersonal description in ethnographic writing and to look at some of the history of this discursive configuration, notably its history in travel writing.

A recent controversy in the American Anthropologist underscores the difficulties ethnography has had in establishing its relations to adjacent discourses. The controversy surrounded Florinda Donner’s Shabono: A True Adventure in the Remote and Magical Heart of the South American Jungle (1982). The book is a personal account of the experience of a graduate student in anthropology who, while doing fieldwork in Venezuela, is chosen by members of a remote group of Yanomamo to live with them and learn their lifeways. The book has been a great success. On the cover of the paperback Carlos Castaneda hails it as "at once art, magic and superb social science"; a Queens College anthropologist calls it "a rare and beautiful book.. . [that] illuminates the world of the Yanomamo Indians [and] conveys a sense of the mystery and power still to be found in ritual"; Newsweek praises it for going "way beyond anthropological questions and categories into the far reaches of a fascinating alien culture."

Controversy about Shabono broke out when the September 1985issue of the American Anthropologist published a comment accusing Donner of plagiarism and fraud. "Frankly," says Rebecca B. DeHolmes, "I find it hard to believe that Donner spent any length of time with the Yanomamo" (DeHolmes 1983: 665). Donner’s ethnographic data, she suggests, were "rather expertly borrowed from other sources and assembled in a kind of melange of fact and fantasy for which Castaneda is so famous" (ibid.). DeHolmes’s most serious accusation is that much of Donner’s borrowing was outright plagiarism from another extraordinary book, a contemporary captivity narrative called Yanoama: The Narrative of a White Girl Kidnapped by Amazonan indians, which appeared in Italian in 1965 and in English translation in 1969. This book, whose authenticity no one has questioned, presents the life story of a Brazilian, Helena Valero, who lived from childhood to adulthood with a group of Yanomamo who kept her following an attack on her family.’

DeHolmes’s accusation of plagiarism is supported by a series of sample passages from the two books and a sizeable list of what she calls "parallel accounts of the same events, plus similar or identical time sequences." The anthropologists with whom I spoke had found this evidence immediately convincing. They had agreed with DeHolmes that Donner’s whole book must be a fabrication and that Donner had probably never lived with the Yanomamo.

I was surprised both by how quickly people rushed to this extreme conclusion and by how schematic the terms in which the issue was discussed were. Donner’s book was either true or false, meaning, apparently, that she had either lived with the Yanomamo or had not, and nothing more was at issue. I was also suspicious, especially at the eagerness to settle the matter quickly and at Donner’s expense, despite the fact that many anthropologists had apparently read and appreciated Donner’s book, and found it believable. Shabono, it looked to me, was being "killed by science," in Malinowski’s words, and without much in the way of a trial. The case obviously threatened some delicate disciplinary boundaries. Most pointedly, it brought to the surface the anguished and messy tangle of contradictions and uncertainties surrounding the interrelations of personal experience, personal narrative, scientism, and professionalism in ethnographic writing. By way of explanation, let me push the example a little further.

To an outsider, one of the most interesting puzzles in the singularly impoverished debate on Shabono was that the book’s factual accuracy did not seem to be at issue. DeHolmes’s meticulous scrutiny had produced only a single ethnographic error in 300 pages (a reference to running between rows of manioc). Implicitly it was accepted that given a certain quantity of secondary material, one in fact could construct a convincing, vivid, ethnographically accurate account of life in another culture without personal experience in the field. Why, I wondered, would ethnographers be so willing to concede such a thing? If it were so, exactly what sort of public threat would Donner’s (alleged) deception be?

What was at issue was not ethnographic accuracy, but a set of problematic links between ethnographic authority, personal experience, scientism, and originality of expression. If Donner really did live with the Yanomamo, why would her text so resemble Valero’s? But by the standards of ethnography, the opposite question also arises: How could her account not resemble Valero’s? The allegedly plagiarized passages DeHolmes cites are indeed very similar to Valero’s text, though never exact repetitions. The first five cases listed include: (a) Nabrushi club fight between men, (b) women’s fishing techniques, (c) girl’s coming-of-age confinement and subsequent presentation to group as a woman, (d) preparation of curare and testing on a monkey, (e) invitation to a feast (DeHolmes 1983:665). Of these, one describes a generalized practice or custom (fishing techniques), the other four are rituals—events anthropologists have always specialized in treating as codifiable, repeatable forms, rather than unique events. If Donner’s account of these rituals had not coincided in detail with Valero’s, fraud by anthropology’s own lights would have been certain. But what was being argued was the exact opposite.

In the end, for DeHolmes, the authority of the ethnographic text is directly constituted by the writer’s personal experience, in turn attested by originality of expression: "If Donner’s Shabono is to be called ‘superb social science,’ as Castaneda claims, it must be shown that the ethnographic data on which she bases her story were actually gathered personally by her while living among the Yanomama and not rewritten from previously published works."

By contrast, for Debra Picchi, who reviewed the book in the same journal, Shabono’s failure to be science arises rather from its "narcissistic focus" on Donner’s "personal growth in the field." "To confine anthropology to the personal experiences of specific anthropologists is to deny its status as a social science" and "renders the discipline trivial and inconsequential" (Picchi 1983:674). Donner fails to display what for Ricchi is the distinguishing characteristic of anthropology’s project, namely a "commitment to the documentation of relationships between behavioral variables on a cross-cultural basis." For Picchi, the idea of borrowing and rewriting is not a problem. Since Donner rejected formal field methods, destroying. her notebooks early in the game, "one assumes," Picchi says, "that the standard anthropological information included in the book is the result of the reconstruction from memory or research of the now extensive literature on the Yanomamo Indians" (ibid.). Picchi takes the book to be a bona fide "ethnography of the Yanomamo Indians ... based on 12 months of fieldwork" and recommends it to teachers of introductory anthropology.

Why are Donner’s fairly explicit claims NOT to be writing a work of anthropology or social science irrelevant to this discussion? Though it is not written in the standard idiom of ethnographic description, what places her personal narrative within anthropology’s purview? Once there, why is there such confusion about how it should be evaluated?

For some reason, it all made me think of a teenager hanging around outside a strip joint who gets dragged inside in order to be turned over to the police and kicked out again—that’ll teach him. So "disciplining" is often done, if not at strip joints, then certainly in academies. What Donner clearly did was write an infuriatingly ambiguous book, which may or may not be "true," is and is not ethnography, is and is not autobiography, does and does not claim professional and academic authority, is and is not based on fieldwork, and so on. An ungrateful apprentice can do no worse than this. For if Florinda Donner did fabricate much of her story (as she may have), she has disgraced the profession by lying, and lying so well no one could tell. If she did not fabricate her story, she scored one of the anthropological scoops of the century. For her experience, as she recounts it, is in many respects an ethnographer’s dream. She is invited by the group to study their lifeways; instead of sessions in which she interviews them, it is they who sit down to teach her. She is spared the anguish and guilt of paying her way by distributing Western goods; the group has chosen to stay so remote that it is a near first encounter. To realize this ethnographer’s dream, and then to refuse to convert it into the currency of the discipline that made it all possible, this is indeed a monumental betrayal.

I have dwelt on the case of Shabono because it illustrates some of the confusion and ambiguity that personal narrative, not having been killed by science, raises in the "discursive space" of ethnography. Personal narratives like Donner’s are not unknown within academic anthropology. Indeed, personal accounts of field experiences are a recognizable anthropological subgenre, but they always come accompanied—usually preceded—by a formal ethnography, the book Donner has not (yet?) written. One thinks of such paired books as David Maybury-Lewis’s Savage and the Innocent and Akwe-Shavante Society; Jean-Paul Dumont’s Under the Rainbow and The Headman and I; Napoleon Chagnon’s Yanomamo: The Fierce People and Studying the Yanomamo; Paul Rabinow’s Symbolic Domination and Reflections on Fieldwork in Morocco. Earlier examples include the writings of Clyde Kluckhohn and Roy-Franklin Barton; this personal subgenre is also the conventional textual space into which Malinowski’s diaries were published.

Of these pairs of books, the formal ethnography is the one that counts as professional capital and as an authoritative representation; the personal narratives are often deemed self-indulgent, trivial, or heretical in other ways. But despite such "disciplining," they have kept appearing, kept being read, and above all kept being taught within the borders of the discipline, for what one must assume are powerful reasons.

Even in the absence of a separate autobiographical volume, personal narrative is a conventional component of ethnographies. It turns up almost invariably in introductions or first chapters, where opening narratives commonly recount the writer’s arrival at the field site, for instance, the initial reception by the inhabitants, the slow, agonizing process of learning the language and overcoming rejection, the anguish and loss at leaving. Though they exist only on the margins of the formal ethnographic description, these conventional opening narratives are not trivial. They play the crucial role of anchoring that description in the intense and authority-giving personal experience of fieldwork. Symbolically and ideologically rich, they often turn out to be the most memorable segments of an ethnographic work—nobody forgets the frustration-ridden introduction to Evans-Pritchards’s The Nuer. Always they are responsible for setting up the initial positionings of the subjects of the ethnographic text: the ethnographer, the native, and the reader.

I find it quite significant that this kind of personal narrative, in the form both of books and opening anecdotes, has in fact not been "killed by science," that it persists as a conventional form in ethnographic writing. This fact is particularly noteworthy given the multiple pressures on ethnography that militate against narrative ("mere anecdote") and devalue it as a vehicle of usable knowledge. Against these pressures operates an urgent sense that ethnographic description is somehow not enough on its own. The fact that personal narrative has this conventional place in ethnographic discourse suggests why Shabono, regardless of its author’s own wishes, does fall within anthropology’s purview and must be reckoned with. The fact that personal narrative is marginal and stigmatized explains why a book like Shabono has to be recognized only to be rejected.

I think it is fairly dear that personal narrative persists alongside objectifying description in ethnographic writing because it mediates a contradiction within the discipline between personal and scientific authority, a contradiction that has become especially acute since the advent of fieldwork as a methodological norm. James Clifford speaks of it as "the discipline’s impossible attempt to fuse objective and subjective practices" (see p. 109 below). Fieldwork produces a kind of authority that is anchored to a large extent in subjective, sensuous experience. One experiences the indigenous environment and lifeways for oneself, sees with one’s own eyes, even plays some roles, albeit contrived ones, in the daily life of the community. But the professional text to result from such an encounter is supposed to conform to the norms of a scientific discourse whose authority resides in the absolute effacement of the speaking and experiencing subject.

In terms of its own metaphors, the scientific position of speech is that of an observer fixed on the edge of a space, looking in and/or down upon what is other. Subjective experience, on the other hand, is spoken from a moving position already within or down in the middle of things, looking and being looked at, talking and being talked at. To convert fieldwork, via field notes, into formal ethnography requires a tremendously difficult shift from the latter discursive position (face to face with the other) to the former. Much must be left behind in the process. Johannes Fabian characterizes the temporal aspect of this contradiction when he speaks of "an aporetic split between recognition of coevalness in some ethnographic research and denial of coevalness in most anthropological theorizing and writing" (Fabian 1983 : 36). In other words, the famous "ethnographic present" locates the other in a time order different from that of the speaking subject; field research on the other hand locates both self and other in the same temporal order. There are strong reasons why field ethnographers so often lament that their ethnographic writings leave out or hopelessly impoverish some of the most important knowledge they have achieved, including the self-knowledge. For the lay person, such as myself, the main evidence of a problem is the simple fact that ethnographic writing tends to be surprisingly boring. How, one asks constantly, could such interesting people doing such interesting things produce such dull books? What did they have to do to themselves?

Personal narrative mediates this contradiction between the engagement cailed for in fieldwork and the self-effacement called for in formal ethnographic description, or at least mitigates some of its anguish, by inserting into the ethnographic text the authority of the personal experience out of which the ethnography is made. It thus recuperates at least a few shreds of what was exorcised in the conversion from the face-to-face field encounter to objectified science. That is why such narratives have not been killed by science, and why they are worth looking at, especially to people interested in countering the tendency toward alienation and dehumanization in much conventional ethnographic description.

The practice of combining personal narrative and objectifled description is hardly the invention of modern ethnography, however. It has a long history in those kinds of writing from which ethnography has traditionally distinguished itself. By the early sixteenth century in Europe, it was conventional for travel accounts to consist of a combination of first-person narration, recounting one’s trip, and description of the flora and fauna of regions passed through and the manners and customs of the inhabitants. These two discourses were quite dearly distinguished in travel books, narrative predominating over description. The descriptive portions were sometimes seen as dumping grounds for the "surplus data" that could not be fitted into the narrative.

To give a representative example, a book called The Captivity of Hans Stade of Hesse in A.D. 1547-1555among the Wild Tribes of Eastern Brazil achieved a wide readership in the sixteenth and seventeenth centuries. Stade’s account was divided into two such parts, the first some 100 pages long, recounting his captivity among the Tupi Nambas, and the second some 50 pages long, giving a "veritable and short account of all the by me experienced manners and customs of the Tuppin-Imbas, whose prisoner I was" (Stade 1874:117). In this latter section, Stade’s descriptive agenda has much in common with that of modern ethnography, induding chapters on "what their dwellings are like," "how they make fire," "the places wherein they sieep," "how skilful they are in shooting wild animals and fish with arrows," "how they cook their food," "what kind of regimen and order they have in government and laws," "what they believe in," "how many wives each of them has, and how he manages them," "how they are betrothed," "how they make their beverages wherewith they drink themselves drunk, and how they order their drinking," and so forth. Moreover Stade’s descriptions resemble those of modern ethnography in their specificity, their search for neutrality and evenhandedness, their linkage of social and material orders, as in this description of a house, where spatial organization is seen as determined by social relations:

They prefer erecting their dwellings in spots where they are not far from wood and water, nor from game and fish. After they have destroyed all in one district, they migrate to other places; and when they want to build their huts, a chief among them assembles a party of men and women (some forty couples), or as many as he can get, and these live together as friends and relations.

They build a kind of hut, which is about fourteen feet wide, and perhaps a hundred and fifty feet long, according to their number. The tenements are about two fathoms high, and round at the top like a vaulted cellar; they thatch them thickly with palm leaves, so that it may not rain therein, and the hut is all open inside. No one has his specially-prepared chamber; each couple, man and woman, has a space of twelve feet on one side; whilst on the other, in the same manner, lives another pair. Thus their huts are full, and each couple has its own fire. The chief of the huts has also his lodging within the dwelling. They all have commonly three entrances, one on each side, and one in the middle; these are low, so that they must stoop when they go in and out. Few of their villages have more than seven huts. 125)

I use the example of Hans Stade deliberately to underscore the point that this discursive configuration I am talking about is the product neither of an erudite tradition nor of the rise of modern science, despite its similarities with contemporary ethnography. Hans Stade was a ship’s gunner with little formal education; his book was a very popular one, and it predates the rise of "natural history" in the eighteenth century.

In some cases, in travel writing, the descriptive discourse is found enmeshed in the narrative, as illustrated by this excerpt from Mungo Park’s Travels in the Interior Districts of Africa (1799):

We stopped a little at a village called Dangali; and in the evening arrived at Dalli. We saw upon the road two large herds of camels feeding. When the Moors turn their camels to feed, they tie up one of their fore legs, to prevent their straying.. . . The people were dancing before the Booty’s house. But when they were informed that a white man was come into the town, they left off dancing and came to the place where I lodged, walking in regular order, two and two, with the music before them. They play upon a sort of flute; but instead of blowing into a hole in the side, they blow obliquely over the end. (Park 1860:46)

Though interwoven, particularized narrative and generalized description remain distinguishable here, and shifts from one to the other are clear, the most conspicuous signs being, of course, the shift from past tense to present tense and from specific persons to tribal labels. (As I shall be illustrating below, this is the configuration that turns up in the work of Malinowski and Raymond Firth.)

In its various guises the narration-description duality has remained remarkably stable in travel writing right down to the present, as has the conventional ordering—narration first, description second; or narration superordinate, description subordinate. By the late nineteenth century, however, the two modes often had about equal weight in travel books, and it was common for a trip to result in two separate volumes, such as Mary Kingsley’s masterpiece Travels in West Africa (1897) and West African Studies (1899). Richard Burton’s The Lake Regions of Central Africa (1868) alternates chapters of narration with chapters describing the "geography and ethnology" of each region passed through.

Modern ethnography obviously lies in direct continuity with this tradition, despite the disciplinary boundary by which it separates itself off from travel writing. Ethnographic writing as a rule subordinates narrative to description, but personal narrative is still conventionally found, either in the separate personal volumes or in vestigial form at the beginning of the book, setting the stage for what follows.

It is no surprise, then, to find that the opening narratives in ethnographies display dear continuities with travel writing. For instance, Firth in We, the Tikopia (1956) introduces himself via the classic Polynesian arrival scene. This scene became a commonplace in the literature of the South Sea explorations of Cook, Bougainville and others in the 1760s and 7os. It is a memorable passage, to which Clifford Geertz has also recently turned his attention (Geertz 1983c).

In the cool of the early morning, just before sunrise, the bow of the Southern Cross headed towards the eastern horizon, on which a tiny dark blue outline was faintly visible. Slowly it grew into a rugged mountain mass, standing up sheer from the ocean .... . In an hour or so we were dose inshore, and could see canoes coming round from the south, outside the reef, on which the tide was low. The outrigger-fitted craft drew near, the men in them bare to the waist, girdled with bark-cloth, large fans stuck in the backs of their belts, tortoise-shell rings or rolls of leaf in the ear-lobes and nose, bearded, and with long hair flowing loosely over their shoulders. Some plied the rough heavy paddles, some had finely plaited pandanus-leaf mats resting on the thwarts beside them, some had large clubs or spears in their hands. The ship anchored on a short cable in the open bay off the coral reef. Almost before the chain was down the natives began to scramble aboard, coming over the side by any means that offered, shouting fiercely to each other and to us in a tongue of which not a word was understood by the Mota-speaking folk of the mission vessel. I wondered how such turbulent human material could ever be induced to submit to scientific study....

We slipped overboard on to the coral rock and began to wade ashore hand in hand with our hosts, like children at a party, exchanging smiles in lieu of anything more intelligible or tangible at the moment. We were surrounded by crowds of naked chattering youngsters... . At last the long wade ended, we climbed up the steeply shelving beach, crossed the soft, dry sand strewn with the brown needles of the Casuarina trees—a home-like touch; it was like a pine avenue—and were led to an old chief clad with great dignity in a white coat and a loin-cloth, who awaited us on his stool under a large shady tree. (Firth 1936: 1-2

Firth reproduces in a remarkably straightforward way a utopian scene of first contact that acquired mythic status in the eighteenth century, and continues with us today in the popular mythology of the South Sea paradise (alias Club Méditerrané/Fantasy Island). Far from being taken for a suspicious alien, the European visitor is welcomed like a messiah by a trusting populace ready to do his or her bidding. For comparison, consider Louis de Bougainville’s arrival in Tahiti in 1767:

We run with all sails set towards the land, standing to windward of this hay, when we perceived a pariagua coming from the offing, and standing for the land, and making use of her sail and paddles. She passed athwart us, and joined a number of others, which sailed ahead of us, from all parts of the island. One of them went before all the rest; it was manned by twelve naked men, who presented us with branches of bananas; and their demonstrations signified that this was their olive branch. We answered them with all the signs of friendship we could imagine; and they then came along side of our ship; and one of them, remarkable for his prodigious growth of hair, which stood like bristles divergent on his head, offered us, together with his branch of peace, a little pig, and a cluster of bananas....

The two ships were soon surrounded with more than an hundred periaguas of different sizes, all which had outriggers. They were laden with cocoa-nuts, bananas, and other fruits of the country. The exchange of these fruits, which were delicious to us, was made very honestly for all sorts of trifles. (Bougainville 1967:213)

Bougainville has a lot more trouble than Firth in getting his ship anchored, but once he does, the same drama resumes:

When we were moored, I went on shore with several officers, to survey the watering-place. An immense crowd of men and women received us there, and could not be tired with looking at us; the boldest came to touch us; they even pushed aside our clothes with their hands, in order to see whether we were made exactly like them: none of them wore any arms, not so much as a stick. They sufficiently expressed their joy at our arrival. The chief of this district conducted and introduced us into his house, in which we found five or six women, and a venerable old man. (ibid.: 220)

The similarities between the two scenes are obvious. There are some interesting differences too. Bougainville’s version of the scene uses one trope Firth does not reproduce, the sentimental commonplace whereby the natives try to undress the foreigners to determine their humanity and, symbolically, level the difference between them. Firth stays dressed, like the king he is to meet. Bougainville also carefully mentions the material relationship that is immediately established between Europeans and natives, an exchange whose spontaneous equality he stresses. In Firth this opening exchange is present, but dematerialized, an exchange of "smiles, in lieu of anything more intelligible or tangible at the moment" and leaving unclear what his material relation to these people is going to be. At the same time, Firth demystifies the egalitarianism of the conventional vision with his note of irony about how all this human material "could be made to submit to scientific study." Indeed, his irony here lightly marks the royal-arrival trope as a trope, and as part of a language of conquest. The fact that his own project is also an assertion of power is tacitly acknowledged. Firth’s opening presentation of self suggests the story of Tikopia’s becoming "his little island," as Malinowski was later to call it (Firth 1937:1).

Malinowski in Argonauts introduces a quite different self-image, also from the annals of travel writing. His "brief outline of an Ethnographer’s tribulations" opens with the now famous line: "Imagine yourself suddenly set down surrounded by your gear, alone on a tropical beach close to a native village while the launch or dinghy which has brought you sails away out of sight" (Malinowski 1961:4). This is unmistakably the image of an old-fashioned castaway. That it turns up here is especially apt, since it corresponds to Malinowski’s own situation at the time. An Austrian citizen living in Australia, he had been sent to sit out the war in the Trobriand Islands rather than risk reprisals or deportation.

There are other reasons why the image of the castaway is an evocative and utopian self-image for the ethnographer. For as the figures of Helena Valero, Florinda Donner, and Hans Stade suggest, castaways and captives in many ways realize the ideal of the participant-observer. The authority of the ethnographer over the "mere traveler" rests chiefly on the idea that the traveler just passes through, whereas the ethnographer lives with the group under study. But of course this is what captives and castaways often do too, living in another culture in every capacity from prince to slave, learning indigenous languages and lifeways with a proficiency any ethnographer would envy, and often producing accounts that are indeed full, rich, and accurate by ethnography’s own standards. At the same time, the experience of captivity resonates a lot with aspects of the experience of fieldwork— the sense of dependency, lack of control, the vulnerability to being either isolated completely or never left alone.

Then again, the image of the castaway mystifies the ethnographer’s situation in ways mentioned earlier. Castaway and ethnographer differ fundamentally in their material relation to the indigenous group. Castaways take up a place within the indigenous social and economic organization; that is how they survive—if, indeed, they do survive, for captivity is a rather higher-risk business than fieldwork. Anthropologists customarily establish a relationship of exchange with the group based on Western commodities. That is how they survive and try to make their relations with informants nonexploitive. But of course this strategy is enormously contradictory, for it makes anthropologists constant contributors to what they themselves see as the destruction of their object of study. The status of the captive or castaway, by contrast, is innocent, and one can see why it would be a compelling image to the contradiction-ridden ethnographer (though actually going native, as captives often did, is taboo for anthropologists).

Firth and Malinowski, then, both invoke well-established images from travel literature in their opening view of the ethnographer. The bodies of their ethnographies are also similar, at least in their deployment of narrative and descriptive discourse. Both writers move freely and fluidly between the two, introducing anecdote constantly to illustrate or elaborate on the ethnographic generalizations, in a way somewhat reminiscent of the text from Mungo Park cited above (p. 35). Firth is full of complex passages like this one, where his ethnographic generalizations, his eyewitness anecdotes, and his personal irony interweave:

Relatives by marriage do occasionally get in sly digs at each other without absolutely transgressing the bounds of good manners. Pa Ranifuri told me with great glee of how the Ariki Taumako spoke to him of his classificatory son-rn-law Pa-Panisi as "Matua i te sosipani,, —sosipani being the native pronunciation of saucepan, of which sooty vessel this man was as far as I recollect the only possessor in the island. As a dark-skinned foreigner he was slightly sneered at (behind his back) by the Tikopia. Scolding, 1 was told, though not permitted by convention directly, may take place at a distance. (Firth 1936:274)

What Firth and Malinowski (his teacher) seem to be after is a kind of summa, a highly textured, totalizing picture anchored in themselves, where "self" is understood not as a monolithic scientist-observer, but as a multifaceted entity who participates, observes, and writes from multiple, constantly shifting positions. Such are the reflective capacities of this versatile, larger-than-life subject that it can absorb and transmit the richness of a whole culture.2 In this subject is also anchored the heady optimism both Firth and Malinowski convey about the ethnographic enterprise. There is a significant irony in the fact that the speaking subject in the work of these founders of "scientific ethnography," as Malinowski called it, is anything but the self-effaced, passive subject of scientific discourse.

The richly perceptive, but terribly unsystematic, textual being that is the ethnographer in Malinowski and Firth contrasts sharply with the frustrated and depressed figure that appears in some subsequent classic ethnographies, such as those of Evans-Pritchard and Maybury-Lewis. If Firth shows up as a benevolent eighteenth-century scientist-king, Evans-Pritchard comes on stage in the later guise of the gruff Victorian explorer-adventurer who exposes himself to all sorts of dangers and discomforts in the name of a higher (national) mission. In his famed opening description of field conditions in The Nuer (1940), Evans-Pritchard joins a century-long line of African travelers who lose their supplies and cannot control their bearers. His initial self-representation reads thus:

I arrived in Nuerland early in 1930. Stormy weather prevented my luggage from joining me at Marseilles. and owing to errors, for which I was not responsible, my food stores were not forwarded from Malakal and my Zande servants were not instructed to meet me. I proceeded to Nuerland (Leek country) with my tent, some equipment, and a few stores bought at Malakal, and two servants, an Atwot and a Bellanda, picked up hastily at the same place.

When I landed at Voahuang on the Bahr el Ghazal the catholic missionaries there showed me much kindness. I waited for nine days on the river bank for the carriers I had been promised. By the tenth day only four of them had arrived.. . . On the following morning I set out for the neighbouring village of Pakur, where my carriers dropped tent and stores in the centre of a treeless plain, near some homesteads, and refused to bear them to the shade about half a mile further. Next day was devoted to erecting my tent and trying to persuade the Nuer... to remove my abode to the vicinity of shade and water, which they refused to do. (Evans-Pritchard 1940: 5)

Such episodes are a commonplace of the African travel and exploration writers of the last century. Here for instance is the eternally cranky Sir Richard Burton, whose expedition recounted in The Lake Regions of Central Africa (1868) is given much the same inauspicious beginning:

We were delayed ten days off Wale Point by various preliminaries to departure. Said bin Salim, a half-caste Arab of Zanzibar, who, sorely against his will, was ordered by the prince to act as Ras Kafilah, or caravan-guide, bad, after ceaseless and fruitless prayers for delay, preceded us about a fortnight, for the purpose of collecting porters.. . . He had crossed over, on the 1st of June, to the mainland, and had hired a gang of porters, who, however, hearing that their employer was a Muzungu, "white man", at once dispersed, forgetting to return their hire. About one hundred and seventy men were required; only thirty-six were procurable... . It was necessary to leave behind, till a full gang of porters could be engaged, the greater part of the ammunition, the iron boat which had proved so useful on the coasting voyage to Mombasah, and the reserve supply of cloth, wire and beads, valued at 359 dollars. The Hindus promised faithfully to forward these articles. ... Nearly eleven months, however, elapsed before they appeared. (Burton 1961: 12

And so it goes, over and over, till Burton’s narrative, like Evans-Pritchard’s introduction, reads like one long, frustrating master-servant feud. Evans-Pritchard’s choice of this particular self-representation is not haphazard. Clifford Geertz’s recent analysis of an old military memoir of Evans-Pritchard’s (Geertz 1983a) connects Evans-Pritchard directly with the tradition of African colonial exploration and writing.

With respect to discursive conventions, Evans-Pritchard must also be thought of as producing a hugely degraded version of the utopian arrival scene exemplified in Bougainville and Firth. This is first contact in a fallen world where European colonialism is a given and native and white man approach each other with joyless suspicion. MayburyLewis gives a similarly degraded version of the scene in Akwe-Shavante Society. He does find help with his luggage, but the chief who awaits him is not recognizably regal, the opening exchange of goods not reciprocal:

A number of Shavante from the village had come to the airstrip when our plane landed and helped to carry our baggage to the post. They set it down at the feet of an elderly man, who we discovered was the chief of the village. He dearly expected us to open the trunks then and there and distribute their contents. (Maybury-Lewis 1967: xxiii)

The problem here, it turns out, is contamination from outside. The Shavante have been spoiled by Brazilian army officers who fly in to see them as a curiosity and bring "elaborate gifts." Like Evans-Pritchard (his teacher), Maybury-Lewis complains at length about his informants’ hostility and uncooperativeness, their refusal to talk to him in private, their refusal to leave him alone, his problems with the language, and so on.

In both these anti-utopian instances, the opening narrative is given by way of explaining the limitations on the ethnographer’s abilityto carry out his scientific mission. Paradoxically enough, the conditions of fieldwork are expressed as an impediment to the task of doing fieldwork, rather than as part of what is to be accounted for in fieldwork. The contrast with the tone of Firth and Malinowski is obvious. Evans-Pritchard and Maybury-Lewis are the heirs of the scientific, professional ethnography Malinowski invented. The scientific ideal seems to press on them acutely, calling for codified field methodology. professional detachment, a systematic write-up. Whatever about the other culture impedes these tasks is an ethnographic obstacle, as well as an ethnographic fact. Both writers complain, for instance, about the impossibility of having private conversations with informants, as if private conversation, once baptized as a field method, ought to be culturally possible everywhere. As methodology gets increasingly codified, the clash between "objective and subjective practices" becomes increasingly acute.

In Evans-Pritchard’s case, the difficulties translate into a rigid separation between personal narrative (his long and vivid introduction) and impersonal description (the rest of his account). Gone is the constantly shifting position of speech of Firth or Malinowski, and gone is the sense of authority, the possibility of reliable totalizing. Evans-Pritchard strives for a totalizing picture of the Nuer, centered on cattle, but feels he must be emphatic, both at the beginning and the end of his book, about the limitations on his capacities and his achievement. Maybury-Lewis’s response, on the other hand, is to try to reaffirm the authority and legitimacy of personal narrative, and to make it explicit as a subgenre. He publishes his personal account first (The Savage and the Innocent 1965), fills his ethnographic book with personal narrative, and suggests that "it is time we abandoned the mystique which surrounds field-work and made it conventional to describe in some detail the circumstances of data-collecting" (Maybury-Lewis 1967: xx).

Each of these opening narrative self-portraits (Malinowski, Firth, Evans-Pritchard, Maybury-Lewis) comes straight out of the tropology of travel writing. Intriguingly, they often come out of the specific tradition of writing on the region in which the ethnographer is working (Central Africa, the South Pacific). At the same time, each expresses in an only slightly mediated fashion some specifics of the particular situation in which each ethnographer finds himself. They are emblematic self-portraits, which operate as a prelude to, and commentary on, what follows. In this respect they are not trivial, for one of their tasks is to position the reader with respect to the formal description. As often as not in modern ethnographies such passages undertake to problematize the reader’s position, as they do in Evans-Pritchard and Maybury-Lewis. The ethnographer’s trials in working to know another people now become the reader’s trials in making sense of the text.

In all the cases I have discussed, much is mystified in the ethnographer’s self-portrait. Much is ironized, indirectly questioned, but never named—notably the sheer inexplicability and unjustifiabiity of the ethnographer’s presence from the standpoint of the other. Evans-Pritchard’s dogged misreading of a conversation with an informant in his preface could provide no clearer example. Equally mystified is the larger agenda of European expansion in which the ethnographer, regardless of his or her own attitudes to it, is caught up, and that determines the ethnographer’s own material relationship to the group under study. This relationship is one of the great silences in the midst of ethnographic description itself. It is the silence that shapes the traditional ethnographic project of trying to describe the culture as it was before Western intervention.

I have been making a loose generational argument in this essay, and propose to end up with one final example from a yet more recent generation of ethnographers. Marjorie Shostak’s Nisa: The Life and Words of a !Kung Woman (1981) is widely regarded as one of the more successful recent experiments in rehumanizing ethnographic writing. The arrival trope is only one of many conventions of ethnographic writing that get remolded in her text. The introduction to Nisa contains two versions of the conventional arrival scene. Here is the first:

Walking into a traditional !Kung village, a visitor would be struck by how fragile it seemed beneath the expanse of sky and how unobtrusively it stood amid the tall grass and sparse tree growth of the surrounding bush. ... A visitor who arrived in the middle of the cold season—June and July—and just at sunrise would see mounds of blankets and animal skins in front of the huts. covering people still asleep beside their fires. Those who had already awakened would be stoking the coals, rebuilding the fire, and warming themselves in the chilly morning ..... . A visitor on another morning, in the hot, dry months of October and November, would find people moving about, even at dawn, up early to do a few hours of gathering or hunting before the midday heat would force them to rest in the thickest shade. (Shostak 1981 :7—8)

As in Firth, this hypothetical arrival is at dawn—new day, new place. And, in a slightly different way, it presents an ethnographic utopia: here is a traditional society doing its traditional thing, oblivious to the alien observing presence. Unlike the other arrival scenes considered above, this one is hypothetical and normalized. It represents what "a visitor (not an anthropologist) would experience if The status of this fantasy is made apparent several pages later when Shostak recounts at great length her own arrival experience, the. night she meets Nisa:

By the time we arrived at Gausha, from Coshi, where we had our main camp, it had long since been dark. We drove the Land Rover past one ! Kung village and stopped at a deserted village site farther down the road. The full moon, high in the sky, appeared small and gave off a cold light.... Kxoma and Tuma, two ! Kung men traveling with us, suggested we make our camp at this site where Richard Lee and Nancy Howell, other anthropologists, had set their camp four years earlier. Living where someone had lived before was right, they said: it connected you to the past. The slender stick shell of Richard and Nancy’s hut was still there. It stood out in the moonlight, a bizarre skeleton set apart from the surrounding bush.. . a traditional ! Kung frame. The grass had long since been taken and used in Nisa’s village. As it stood, the hut offered no protection from the weather, nor any privacy. (Shostak 1981 : 23)

To begin with, this is a nocturnal arrival, relatively rare in both travel writing and ethnography. For the anthropologists, this is the land of the dead—the moon is small and cold, and inexplicably they take themselves past the village of the living, to a deserted one, and set themselves up in the skeleton of a hut. Shostak’s own symbolism contrasts sharply with the !Kung’s understanding of what they are doing. For them, the link with the past is a haven, a comfort. Shostak and her companion are haunted by the ghosts of their anthropological predecessors.

There is no spontaneous native welcome in Shostak’s account, but unlike both Firth and Evans-Pritchard, these anthropologists do not want or need one: "It was too late for visiting. They knew us by then; we would still be there in the morning." They unpack by themselves ("I thought of the time Nancy had found a puff adder in her sleeping bag"). What is biting the dust here, of course, is even the pretense or fantasy of a first encounter. Far from being the first European on the scene, Shostak is even a long way from being the first anthropologist. And this point gets woefully dramatized when the indigenous welcomers, themselves unwelcome, belatedly appear after the luggage work is done and the party is ready for bed: "Nisa wore an old blanket loosely draped over the remnants of a faded, flower-print dress, sizes too big. Bo was wearing a pair of shorts, even the patches of which were worn through in places." There they are, in European clothes they have obviously been wearing for a long time. And instead of a song of welcome to Shostak and her partner, what Nisa sings is a praise song for their predecessors. "Richard and Nancy! I really liked them! They liked us too—they gave us beautiful presents and took us everywhere with them. Bo and I worked hard for them.... Oh how I wish they were here!"

It is an awful scene, a real return of the repressed. These others are fallen, corrupted not only as non-Europeans, but specifically as ethnographic informants. Bo and Nisa arrive praising the guilty relationship of exchange based on Western commodities, the point at which the anthropologist preserver-of-the-culture is the interventionist corrupter-of-the-culture. It is the naive informant turned hustler. In short, this arrival scene contemplates, reluctantly and in the dark of night, the aftermath of the ethnographic episode: the indigenous people who, specifically through the ethnographic contact, have acquired a vested interest in westernization and a concrete day-to-day link with the larger structures of exploitation. It is like a bad dream, and Shostak’s reaction, as she describes it, is to try to pretend it is not happening and wait for a more standard arrival scene in the morning, complete with the ritual exchange of goods: "That was when we would give [tobacco] to the rest of the !Kung with whom we planned to work" (Shostak 1981 : 25).

After long resistance, Shostak ultimately capitulates to this degraded anthropological world, consents to be hustled/seduced by Nisa, and then and only then finds a fieldwork relationship that is in fact enormously productive for ethnographic purposes. For Nisa is advertising a genuine talent for story telling and for reflecting on her own culture and experience.

Out of that encounter, Shostak produces a text that, in a way very different from Malinowski and Firth, again seeks to recognize the authority of personal narrative within the ethnographic description. Shostak’s own story is concentrated at the beginning and end of the book. In between, Nisa’s narrative predominates, edited into a life story by Shostak, who introduces each section with a few pages of ethnographic generalizing and commentary. So this book tries to mediate the contradiction between objectifying ethnographic representation, in which all natives are equivalent and equal, and the subjective experience of fieldwork, in which all informants are not equivalent or equal. What is sought here as an alternative is not, however, a totality or a synthesis anchored in the multifaceted subjectivity of the fleldworker/traveler/scientist, but something less unified and more polyphonous.3

A utopian element persists, however, for the polyphony Shostak creates is strikingly harmonious. One finds little evidence of struggle among the various voices, despite Shostak’s awareness of the intolerable contradictoriness of her position. Current Western conceptions of female solidarity and intimacy seem to be part of what produces this cross-cultural harmony: she and Nisa are bound together in ways that perhaps transcend culture. It is to Shostak’s credit that the last lines of the book challenge that harmony. Nisa’s final words in the text are: "My niece . . . my niece.. . you are someone who truly thinks about me." Shostak’s are: "I will always think of her and I hope she will think of me, as a distant sister." Each assigns the other a different honorary kinship.

In the darkness that shrouds Shostak’s arrival scene, Western readers will recognize the symbolism of guilt. In part there is guilt linked to the particularities of Shostak’s situation as the last in a series of anthropologists working among the !Kung. And in part it is guilt that comes down to her from much farther in the past. For as with the other examples I have discussed, Shostak’s text, and those of her fellow participants in the Harvard Kalahari Project, displays direct continuities with a long tradition of writing about the !Kung—three centuries of it in fact. It is a nonprofessional "lay" tradition from which the Harvard anthropologists energetically dissociate themselves (most explicitly by replacing the earlier European name for the group, the Bushmen). What that tradition documents is a long and violent history of persecution, enslavement, and extermination. The contemporary !Kung are the survivors of that history; the contemporary anthropologists the heirs to its guilt. A few details will clarify this point.

It is at the end of the eighteenth century that the Bushmen start turning up in European accounts as objects of (a) ethnographic interest and (b) pathos and guilt. Not surprisingly, this was also the point at which the Bushmen definitively lost their struggle against European encroachment on their land and lifeways. For the previous century, they had existed in European writings as hordes of wild, bloodthirsty marauders fiercely resisting the advancing colonists, raiding their farms at night, turning loose or stealing cattle, and sometimes murdering colonists or laborers. Given a free hand by colonial authorities, the settlers embarked on a war of extermination. Commandos descended on Bushman kraals, often at night, killing men and either killing or enslaving women and children. Gradually the settlers won this war, so that by the 1790s "the Bushmen were still numerous along the interior mountain range, but in other parts of the colony there were hardly any left" (Theal 1897:1, 198-201).4

It is at this point that the discourse on the Bushmen changes. Late eighteenth-century travelers, like the Swede Anders Sparrman (Voyage to the Cape of Good Hope 1785) and Englishman John Barrow (Travels in the Interior of Southern Africa 1801), vociferously deplore the brutality of the colonists and the injustice of the extermination campaign. The same writers construct a new ethnographic portrait of the Bushmen. No longer seen as militant warriors or bloodthirsty marauders, they acquire the characteristics that the powerful commonly find in those they have subjugated: meekness, innocence, passivity, indolence coupled with physical strength and stamina, cheerfulness, absence of greed or indeed desires of any kind, internal egalitarianism, a penchant for living in the present, inability to take initiatives on their own behalf. Thus Sparrman describes the Bushmen as "free from many wants and desires, that torment the rest of mankind," "detesting all manner of labour," yet easily induced into slavery by a little meat and tobacco (Sparrman 1975: 198—20 1). Barrow finds they are "mild and manageable in the highest degree, and by gentle usage may be moulded into any shape." "In his disposition," says Barrow, "[the Bushman] is lively and cheerful; in his person active. His talents are far above mediocrity." Their constitutions are "much stronger and their lives of longer duration, than those of the Hottentots"; "universal equality prevails in his horde... they take no thought for the morrow. They have no sort of management or economy with regard to provisions" (Barrow 1801 : 287).

Presented as objectified ethnographic description of the (eternal) Bushman and his (natural) disposition, this is a portrait of a conquered people, simultaneously acknowledging the innocence and pathos of their condition, evaluating their potential as a labor pool, and legitimating their domination on the grounds that they do not know how to manage themselves.

Caught between celebrating and deploring, historicizing and naturalizing the Bushmen’s condition, Barrow expresses the guilt and anguish of his position in a nocturnal arrival scene that has numerous points in common with Shostak’s, despite being written 180 years earlier. In Barrow’s account, too, the all-important personal contact with the other is achieved only through a kind of descent into hell. Though his motives are entirely benevolent, the only way Barrow can make contact with the terrified Bushmen is by hiring a group of Boer farmers to help him ambush a group at night. As with Shostak, the result is a nightmare of contradiction, one of the few episodes of personal narrative in Barrow’s book. Barrow describes the raiding party descending from the hillsides down onto the sleeping camp: "Our ears were stunned with a horrid scream like the war-hoop of savages; the shrieking of women and the cries of children proceeded from every side" (Barrow 1801 : 272). Despite Barrow’s instructions, his Boer guides begin shooting down the fleeing people; Barrow’s protests are ignored. "‘Good God!’ [the Boer farmer] exclaimed, ‘Have you not seen a shower of arrows falling among us?’ I certainly had seen neither arrows or people, but had heard enough to pierce the hardest heart" (272). Later Barrow remarks in shame that "nothing could be more unwarrantable. . . than the attack made by our party upon the kraal" (291).

This scene that disrupts Barrow’s highly impersonal account is an explicit reversal not only of the utopian arrival scene a Ia Bougainville, but also of the traditional image of the Bushman horde descending on European ranches. Its symbolism and position in Barrow’s text are similar to those in Shostak’s, and the two share the same discursive history. (Interestingly, Shostak’s predecessor and colleague Richard Lee also uses a nocturnal arrival scene to introduce his third visit to the I!Kung [Lee 1979:xvii].)

But it is not only arrival scenes that contemporary anthropologists share with earlier writers on the Bushmen. Their ethnographic descriptions likewise reproduce the discursive legacy, even as they openly repudiate it. One is struck by the extent to which twentieth century anthropological and journalistic writers have continued to celebrate and naturalize in the Bushmen many of the same characteristics singled out by Barrow, Sparrman, and the rest. Cheerfulness, humor, egalitarianism, nonviolence, disinterest in material goods, longevity, and stamina are all underscored with admiration and affection in both journalistic writings like those of Laurens van der Post (The Lost World of the Kalahari 1958) and ethnographic work from Elizabeth Marshall Thomas’s The Harmless People (1959) to the writers of the Harvard Kalahari Project (1963—70). What turns up throughout this literature is the same blazing contradiction between a tendency on the one hand to historicize the !Kung as survivor-victims of European imperialism, and a tendency on the other to naturalize and objectify them as primal beings virtually untouched by history. (ln both cases, they stand doomed to extinction.) Detailed discussion of the extensive contemporary literature on the !Kung/Bushmen is impossible here, but Shostak’s own text can exemplify the ambivalence I am talking about.

By introducing Nisa to us clad in a dress and selling her talents on the anthropological free market, Shostak repudiates the image of the pure primitive so often associated with the !Kung. Yet it is ultimately that image of the primitive that motivates Shostak’s inquiry. For her, as for the Harvard Kalahari group as a whole, the !Kung are of interest as evidence concerning our human and prehuman ancestors. Shostak hopes !Kung women will be able to "clarify some of the issues raised by the American women’s movement," especially because, she says,

their culture, unlike ours, was not being continuously disrupted by social and political factions telling them first that women were one way, then another. Although the !Kung were experiencing cultural change, it was still quite recent and subtle and had thus far left their traditional value system mostly intact. A study revealing what !Kung women’s lives were like today might reflect what their lives had been like for generations, possibly even for thousands of years. (Shostak 1981: 6)

"Recent" and "subtle" are not the adjectives that come to mind when one ponders the grim history of the Bushman conquest. This is a history of which Shostak and her colleagues seem at times deeply aware, at times totally oblivious. Repeatedly, Richard Lee and others warn that the !Kung are not to be treated as "living fossils" or "missing links" (Lee 1979: xvii), that their colonial past and changing present must be given full recognition to avoid dehumanization and distortion. Yet the inquiry the group proposes is explicitly an evolutionary one (initiated by primatologists) in which the !Kung are important as "evidence which will help in understanding human history" (S. Washburn in Lee and Devore 1976: xv) as examples of an ecological adaptation "that was until ten thousand years ago, a human universal" (Lee 1979: i). Focusing heavily on physical and biological issues like diet, physiology, use of time, settlement patterns, spacing of births, use of food resources, disease, aging, and so on, this literature naturalizes current ! Kung lifeways with a vengeance. The researchers’ sincere desire to be sensitive to the !Kung’s situation in present historical circumstances is simply incompatible with their project of viewing the !Kung as a complex adaptation to the ecology of the Kalahari desert, and an example of how our ancestors lived. The use of primarily quantitative methods (producing tables like "Average number of child-caring acts by a subject per child per hour of observation" [P. Draper in Lee and Devore 1976:214]) intensifies the reification.

An outsider, on reading the history of European contact with the !Kung/Bushmen, inevitably questions this image of them as representatives of hunting-gathering life as it was lived 10,000 years ago. Is it not worth even asking the question whether 300 years of warfare and persecution at the hands of white settlers (to say nothing of the competition with indigenous pastoralists) have had an impact on the life-ways, the consciousness, the social organization, even the physiology of the group undergoing these traumas? Did the long-term practice of massacring men and enslaving women have no impact on "what women’s lives were like" or how women saw themselves? What picture of the !Kung would one draw if instead of defining them as survivors of the stone age and a delicate and complex adaptation to the Kalahari desert, one looked at them as survivors of capitalist expansion, and a delicate and complex adaptation to three centuries of violence and intimidation? There are times when such a perspective does seem to be implied, ever-so-indirectly, as in Shostak’s characterization of a !Kung village as looking "fragile" and "unobtrusive" (see quotation on p.43 above).

To make sense of the conflicting concerns of the Harvard group, one must locate them on the one hand in the context of the American counterculture of the 1 1960s many of whose social ideals seem to realize themselves in the !Kung, and on the other hand, in the context of the expansion of the biological, "hard science" sector of anthropology that has made Harvard the center for sociobiology in the 1980s. At the same time, one must also recognize their continuities with the discursive history coming down from Sparrman, Barrow, and the rest, a discursive history they sometimes wish to "kill by science."

As I have been arguing throughout this paper, anthropologists stand to gain from looking at themselves as writing inside as well as outside the discursive traditions that precede them; inside as well as outside the histories of contact on which they follow. Such a perspective is particularly valuable for people who would like to change or enrich the discursive repertoire of ethnographic writing—especially that "impossible attempt to fuse objective and subjective practices." Surely a first step toward such change is to recognize that one’s tropes are neither natural nor, in many cases, native to the discipline. Then it becomes possible, if one wishes, to liberate oneself from them, not by doing away with tropes (which is not possible) but by appropriating and inventing new ones (which is).