A Green Place, a Good Place: Agrarian Change, Gender and Social Identity in the Great Lakes Region to the Fifteenth Century. By David Lee Schoenbrun. Portsmouth, NH: Heinemann: Oxford: James Currey, 1998. Pp. xiv+301. £40 (ISBN 0-325-00041-7); £15.95, paperback (ISBN 0-325-00040-9).

An African Classical Age: Eastern and Southern Africa in World History, 1000 B.C. to A.D. 400. By Christopher Ehret. Charlottesville: University Press of Virginia; Oxford: James Currey, 1998. Pp. xvii+354. £35 (ISBN 0-8139-1814-6).

Recently several historical reconstructions based on linguistic evidence and dealing with ancient times have been published in African history. In 1998 alone there are the two books reviewed here as well as a major work by Gerda Rossel. Linguistic sources contribute much to the recovery of aspects of the past, which would otherwise remain out of reach, and the standard methodologies of historical linguistics are well known to readers of this journal. Yet in practice many historians remain all too often disconcerted by such studies because they have great difficulty in evaluating them: i.e. in linking assertions made to the evidence provided and so to establish the credibility of such statements. This is not just because many historians are unfamiliar with linguistic evidence but because all the evidence necessary for evaluation is usually not available in the work studied, and often enough authors do not clearly indicate where it can be found. Indeed sometimes it is not available at all. In such cases one has to take the statements made by the authors on faith: one believes the author or not. That is clearly unacceptable. For is it not a fundamental rule in history writing that assertions must be substantiated and hence evidence must be cited or provided? Any work without substantiation cannot be considered to be a work of history at all.