The collection of republished essays entitled Right and Left: Essays in Dual Symbolic Classification, edited and with a substantial introduction by Rodney Needham (the University of Chicago Press, 1973), has been long awaited. All of the essays are of great interest, and all are concerned with the binary mode of symbolic classification, something which appears to be found, in greater or lesser degree, in all human cultures. The work will no doubt be widely reviewed in the appropriate journals. What follows is not an attempt at an over-all review. It is rather a response to, and a critical examination of, some of the arguments Needham advances in his introduction to the book. As such, I hope that it may contribute, however modestly, to our understanding of some of the problems involved in the application of the ‘binary’ mode of structural analysis to ethnographic data. The context is the symbolism of the Banyoro of western Uganda.

This paper is an attempt to establish certain contested facts about Nyoro symbolism. The need to do so results, in brief, from the following circumstances. Eight years ago I published an analysis of the significance of right and left in Nyoro symbolic classification (Needham 1967). John Beattie, who had done field research among the Nyoro, then brought out an article in which he contended that I had seriously misrepresented the nature of Nyoro representations of their own culture and history, ‘even in so far as these can be understood from readily available published sources’. His main concern in his critique, consequently, was ‘to set the Nyoro ethnographic record straight’ (Beattie 1968: 413).

Certain aspects of van Gennep's (1908) analysis of the structure of ritual symbolism have in recent years been elaborated by E. R. Leach in a series of articles dealing with such topics as magic, the symbolism of time as well as initiation rites. This development has followed two related courses. Firstly, Leach has tended to emphasize the critical significance of rites of separation rather than van Gennep's overall schema involving rites of separation, rites of transition and rites of incorporation in effecting changes in the ritual status of persons (Leach 1970). Secondly, attention has been directed instead to the symbolic importance of the residues of such rites of separation (hair, foreskin, etc.) (Leach 1958) and, to a lesser extent, the agent effecting the separation (Leach 1961). The primary purpose of this article, however, is ethnographic rather than theoretical but in so far as the results of this empirical work tend to confirm the emerging analytical schema it is not without general interest and especially as the manner of this confirmation is not quite straightforward. It will be useful first of all to consider further and in general terms the issues indicated in these opening remarks.

The relationship between the living and the dead looms large in Africa, but the study of such interactions is notoriously fraught with pitfalls and perhaps only now are we equipped with a valid conceptual framework to enable us to deal adequately with the problems posed by ancestor worship and by the various spirits of the dead in Africa. Fortes (1965: 16—20; 122—142) first gave us a comparative framework of ancestor worship in Africa but part of his conceptualization has been recently questioned by Kopytoff (1971) who argues that the distinction between elders and ancestors is not as clear cut as our own implicit dichotomy would seem to imply. However, his argument, stressing a continuity between the living and the dead, leaves out the fact that elders, when they die, very often become ancestor spirits and, even if they do not become such, take on new qualities and new powers that are different in nature from those they held when alive (Brain 1973; Sangree 1974).

Among the traditional patrilineal Zulu of South Africa, women are more often associated with mystical experiences than men. While on the one hand, as daughters or sisters, women may be associated with the positive mystical forces as diviners, on the other hand, as mothers or wives, women are often related to the negative polluting mystical forces. It is the logic behind these notions that I want to examine in this article.