In her book, Karin Barber sets the stage for a meeting between ‘new criticism’ and ‘social anthropology’ in Africa. She describes this interdisciplinary encounter in terms of her own intellectual growth, but she was not alone. Literature and anthropology edged toward each other, mediated by linguistics, for a good part of the second half of the twentieth century. Barber's Anthropology of Texts is a direct descendant of the American Dell Hymes's essay on the ‘Ethnography of Speaking’, which, in its own turn, was a synthesis of formalism, structuralism, and linguistic anthropology. For many years anthropology maintained an ambivalent attitude toward indigenous texts. Recognizing their import, researchers considered them a ‘mirror of culture’ that reflected social structure, religion, world view, or history, but hardly noticed their literary qualities. Literary research also failed to perceive their poetics, relegating them to the category of ‘primitive literature’, not worthy of any critical idiom. The winds of change have blown for some time. Barber consolidates them by discussing key concepts in literary theory, folklore and linguistic ethnography, such as genre, text, personhood, audience, and public and private writing, devoting an erudite chapter to each.

Literary genre is a basic concept that has been available to literary critical discourse since Aristotle's Poetics. Over the centuries it has acquired multiple meanings and uses, and functioned in the articulation of a number of literary theories. Acknowledging this rich history, Barber draws upon Bakhtin's approach and the more recent studies in folklore and anthropology, and considers genres ‘as a localized way of organizing speech activity’ (p. 37). Like any other cultural categorization, the organization of speaking is predicated upon social interaction and hierarchy, as well as on religious beliefs and rituals. Astutely, Barber proposes that genres are not isolated from each other in society. Rather, it is their systemic relations that make them significant.

For a comparative scrutiny of a single genre in several African countries, Barber selects a genre the existence of which in Africa was questioned only forty years ago, namely the epic. Subsequent research has confirmed its performance in a broad swatch, stretching from West Africa's Mande-speaking area through east Nigeria and Cameroon, and through central Africa up to the western shore of Lake Victoria – an area now known as the ‘epic belt’. The occurrence of epic performances in this area is correlated with political systems. The epic bard, a necessary role for epic performance, validates the political authority of the ruler, narrating the heroic feats of his ancestors. He does so with the aid of a musical instrument, upon which either he or his musicians play. Barber is aware of the fact that the epic themes and their function, as well as their geographic distribution in Africa, are tentative. The so-called ‘epic belt’ is as much a function of our ignorance as of our knowledge; the bard's political function, as analysed in the scholarship, follows the griot model – but other models, only scantly reported, occur as well.

In Barber's anthropology, ‘text’ is both a metaphor and a thing. In oral societies there are no written texts by definition. Yet, there are recurrent narratives, repeated epics and songs, quizzed riddles and spoken proverbs that are interspersed in conversations and speeches. Those are stored and retrieved from the cultural memory, not any archive or library. Because of their relative stability and availability, these tales, songs and other forms have been conceptualized as texts, or have been entextualized. They lack the verbal stability that script and print have, yet they are remembered, performed and uttered in appropriate contexts, combining stability and creativity on verbal, thematic, contextual and performative levels. Barber, incorporating concepts from current scholarship in folklore and anthropology, and contributing her own ideas, proposes repetition, interpretation and audience response as ways of constituting ‘texts’ in oral cultures. These are indicators of the cultural perceptions of particular discourses as part of the corpus of the cultural texts of an oral society, analogous to texts in literate society. The proposition that interpretation constitutes a discourse as ‘text’, is insightful and applicable to literate societies as well.

While so far the encounter between ‘new criticism’ and ‘social anthropology’ has been mutually intellectually rewarding, it comes to a clash in the analysis of ‘personhood’ as textually constituted in oral poetry, particularly in praise poetry. This is a genre with which Barber is very familiar. Through her superb study of the Yoruba oríkì Footnote ^* she has gained profound knowledge of the society and its texts. But this very local knowledge has made her intolerant of intruding analytical concepts and terms as means of interpretation. She dismisses semiotics, one of the recent incarnation of ‘new criticism’, in favour of cultural experience as a way toward gaining understanding.

The Anthropology of Texts offers sets of concepts, terms and methods, refreshing the study of folklore and oral literature in Africa and elsewhere. It is a book that every cultural anthropologist and every folklorist ought to read.