Amazon grace

I blame Levi-Strauss. It was his Tristes Tropiques that established the genre of personal accounts of fieldwork and legitimated the desires of anthropologists who wished to open their hearts onto the page. Trouble was, his magisterial travelogue, first published in 1955, was not followed by a succession of equally dazzling reports on time spent in the bush but by a mostly dismal series of mediocre narratives of life among different peoples. Instead of enlightening the nature and process of fieldwork itself, most tended merely to display the unfulfilled literary pretensions of their authors. A bit of "colourful" description, a few anecdotes, and some rather heavy chest-beating seems to sum up most.

Philippe Descola bucks this depressing trend of the third rate by producing a first-class book that bears comparison with that of his old supervisor, Levi-Strauss. The Spears of Twilight is pellucid, exactly written and Descola has much to say. It is sure to become a modern classic.

Astonishingly, Descola is the first anthropologist to write a book that focuses methodically and unswervingly on the way he carried out his inquiries in the field, in his case among the Achuar, one of the Jivaro tribes in the Ecuadorian Amazon. Writing in the ethnographic present, he introduces the reader to the state of ignorance the fledgling fieldworker faces on first entering the village after slogging it through the jungle. He then shows how, in a painstakingly and frustratingly slow manner, he and his wife gathered information that they eventually assembled into meaningful patterns.

At times his book reads like a good whodunit, with the anthropologist as plodding gumshoe diligently collecting data until the clues hidden among them emerge, which, once put together, provide an answer to the question at hand, whether that be identifying the murderer in a tribal vendetta or the most satisfactory, rounded way to interpret shamanic trance. Again, like a seasoned detective writer, he details the way he and his wife followed up false leads. His aim here is not to maintain a created suspense but to allow readers to appreciate the process of trial and error dedicated field-workers are forced to pursue if they want to understand what is going on around them.

Conventional ethnographies present societies in a tightly organised manner, as though the local way of life worked according to some invisible clockwork and as though their ethnographers worked according to an equally organised routine. Seeking to dispel this chronotypical image of how he and his peers go about their business, Descola stresses the aleatory nature of field-work: the way the locals, not the outsider, set the pace of investigation, the large role chance constantly plays in the process of research. For fieldwork is above all the study of social relationships by the forging of social relationships. An anthropologist is ultimately, almost totally reliant on others, on what they choose to tell their guest, and when. Descola's account illustrates the point well.

Descola does not want his readers to gawp at the people whose lives he describes. He is not writing a tale of exotic others for the sake of titillating chair-bound westerners. He wants to purvey some of the beauty of Achuar society and - a classical anthropological device - to show that their seemingly unusual customs are not so strange after all. The first way he does this is to place traditional practices into their indigenous and regional contexts. Thus he tries to make the infamous Jivaro custom of shrinking the heads of enemies appear less sinister by showing how it exemplifies in a particular way elements within a politics of otherness common to a diversity of Amazonian groups practising headhunting or ritual cannibalism. But, as he admits, this is not enough, as it serves only to dilute "a particular strangeness in a wider, more encompassing strangeness". Thus the need for a second way to make the apparently odd reassuringly familiar: by taking Jivaro ideas to a higher level of abstraction and then comparing that with their closest equivalent in western societies. The strategy is a well-established one, already employed to great effect by Montaigne, Montesquieu, Swift and a long train of illustrious others. Descola's intention, when, for instance, juxtaposing Jivaro and western forms of the treatment of the dead, is not just to make the strange familiar but to remind his readers how relative their own cultural beliefs are. In this sense, and in this way, he wants anthropology to resume the culturally critical role it once had and should never have allowed to fall into abeyance. This is engaged anthropology, albeit in a mild form.

Like a good ethnographer, Descola complements his lengthy text with a set of photos, their poor focus and lack of composition only adding to the air of authenticity they lend to the text: "The Achuar, a tribe so remote no professional photographer has yet got to them."

Unlike almost all ethnographers, Descola includes a set of commissioned engravings of scenes from Jivaro life. By choosing this style of illustration, more suited to Victorian travelogues than to modern ethnographies, he wants to emphasise visually the sense of artifice in representing cultures that conventional anthropological writing downplays. These penned pictures are there to create a tension and to remind the reader that among the Jivaro all is not what Descola's words might make it seem. Aware of the persuasive power of his own literary style, he does not want his audience to think that the Jivaro can, in the end, be completely contained within his elegant phrases.

Descola's is a complex book, written with grace and fluidity. It is a welcome sign that an anthropology book as good as this has not been put out by an academic press but by an astute commercial publisher prepared to promote it. The Spears of Twilight is a splendid achievement. It deserves the widest readership.

Jeremy MacClancy is senior lecturer in anthropology, Oxford Brookes University.