Slayed Author lives to purge imperial guilt

There was a time when anthropologists worked cheerfully at the level of tribal consensus and wrote unproblematically of the Bongo-Bongo "thinking", "feeling" and "believing" that being located on an active faultline dividing the individual from the communal, ethnographic knowledge had few living people and those few tended to be justified as either properly representative of a larger whole or specimens in some particular and excusable study of deviance. It is only recently that the individual self has come to be seen as something that has to be created and sustained in ways that may show cross-cultural variation and so become problematic.

African Voices, African Lives sets out to document the relationship between the anthropologist, Pat Caplan, and the (Tanzanian) Mafia islander, Mohammed, who is one of those semi-invisible figures known in the traditional monographs as "native informants". It is a relationship that has lasted some 30 years and has seen major transformations in both enquirer and informant as they have moved through life events - wed, procreated, divorced - and different fashions of ethnographic episteme. In many ways, neither they nor their relationship remain the same even through the process of composition of the book and the Mohammed of the end is certainly not the Mohammed of the beginning. The result, it must be confessed, is part of the "All-Mankind-Smiles-Nobly-Through-The-Tears" school of anthropology, rooting human experience in the blows of the life cycle, the growing alienation of age from the changes in contemporary life and the inevitable disappointments of parenthood. Universality is ultimately based on human sympathy and life is really one long disappointment.

The work is topped and tailed with references to the recent theoretical literature on (auto)-biography where western loss of epistemological nerve is normally puffed up into a penny philosophy of hegemony. The publisher's blurb entices that we are to have a privileged look behind the screen and - before our very eyes - naughtily see anthropological knowledge being constructed. We are urged to consider the magic question "Can informants also be ethnographers?" In fact, most of this is a protective dust cloud thrown up to show us that Caplan has done her homework and read hard books and that the unaffected tone of this one is to be understood as the result of art not mere artlessness.

For the core of the book is a perfectly conventional and unusually interesting, low-level account of spirit cults and their role in the attribution of affliction in this part of East Africa, examining how they may interact in complex explanations with each other, western medicine and Islam and how some are favoured over others by particular people and in particular circumstances. (Mohammed, I must explain in a Caplan-like intervention, has developed a late taste for spirit cults.) The analysis is excellently clear and insightful while not ignoring difficult details that frustrate easy generalisation. This book deserves to become a major source for scholars working on this area.

The work as a whole, like so many others in anthropology, has been put together from a combination of informant's diary, tape-recordings and the memories and records of focused and unfocused interactions of the ethnographer, though the actual process of choice, creation, transmutation - pace the publisher's blurb - remains mysterious. The price paid for such free-style ethnography is a slow, meandering pace and lots and lots of quotes. The promised benefit of polyvocality remains elusive. The reader's powerful sense of Mohammed's character is of the same dangerously illusory and manipulative sort as that formed in the course of an ethnographic film. Most of the interesting information comes from Caplan, rather than Mohammed and his family, in the form of inserted cultural and historical explanation and, naturally, there has been a huge amount of editing to satisfy the demands of a book rather than an experience. The voice, then, remains squarely that of the anthropologist and it is far from clear that Mohammed would wish it any other way. He has earlier declined the offer of co-authorship or any other kind of more active participation. The question about informants and the creation of ethnography goes, then, largely by the board. Why, after all, keep an anthropologist and bark yourself? Postmodernism may have tried to slay The Author, but publishers, for the moment, continue to pay her. Mohammed gets half the royalties. Perhaps that is the real and practical way for anthropologists to purge authorial authority of its agonising postimperial guilt rather than writing more books about it. I wonder how many others will race to embrace Caplan's excellent solution.

His real name is not Mohammed, by the way.

Nigel Barley is assistant keeper, Museum of Mankind.