The importance of oral traditions in the reconstruction of the history of non-literate peoples has virtually ceased to be a matter for debate and is now generally acknowledged. Indeed, within the last few years, historical research, based on such traditions has made possible histories of many societies in Africa. But in spite of this general acceptance, the diversity of oral traditions has not been so fully recognized as to make possible the analysis of each type as historical data. In this regard, the history of the Yoruba provides a good example. Their culture has accumulated around it a rich variety of oral traditions whose study has made significant contributions towards the understanding of their past. For the earlier period of their history, their historians have had to rely mainly on oral traditions; even for the latter period, in spite of the existence of written documents, oral traditions have still proved very useful in giving a balanced view of events. The tendency, however, has been for the historians of the Yoruba people to regard oral traditions as no more than personal recollections and generalized historical knowledge. In the main, they have failed to give cognizance to oral literature; experience elsewhere, however, has shown that a meaningful utilization of oral traditions in historical reconstruction cannot afford to neglect this third category, which is for historians of non-literate societies what literature is for the cultural and social historians of literate societies. Moreover, the Yoruba themselves treat some forms of oral literature as quasi-historical records.

At the level of belief Igala religion is rather unsystematic and comprises belief in a number of different spirits whose mutual interrelationship is not very clear. A Western observer would tend to rank the various spirits in terms of their order of magnitude or degrees of power. At the top of the hierarchy would come Ojọ, the supreme being and creator. Another spirit with wide powers in Igala belief is Anẹ, the spirit of the earth. Then there are various oracle spirits known generically as ẹbọ, with the function of giving protection against witchcraft. Their congregation may be large or small according to the success of the spirit in unmasking witches by striking them with illness and making them confess to their wrongdoing. Operating within a more restricted range are the ancestor spirits who are the guardians of well-being in particular clans, lineages, or lineage segments. Within the complex of ancestor worship falls the dyadic relationship between an individual and his personal ọjọ or guardian spirit. The guardian has responsibility for the destiny of its ward. Finally there are fetishes, ode, which are made and worshipped by individuals, and which are believed to ensure success in daily life by affording protection against the misfortunes that witchcraft and sorcery can bring.

Les Azza appartiennent au vaste groupe culturel formé par les Teda et les Daza, qu'on appelle communément Toubou, population islamisée dispersée au Nord du Tchad sur un territoire d'environ deux millions de km2 et que se partagent les deux Républiques, celle du Niger et celle du Tchad.

Artisans castés, les Azza dépendaient autrefois des Daza: émancipés depuis quelques années, ils adoptent petit à petit le mode de vie de leurs anciens maîtres, éleveurs de bovidés dans cette zone sahélo-soudanaise où ils côtoient d'autres populations pastorales: Arabes Choa, Arabes Ouled Sliman, Peuls.

This is an essay about strangers in a West African society. It asks how ethnicity constrains role selection and role performance. It asks why members of bounded ethnic groups occupy specific professions and exploit specific ecological niches in a multi-ethnic society. Field information from south-eastern Niger reveals the environmental and cultural elements which together generate a distribution of peoples into single ethnic and multi-ethnic occupations. My concern is with the processes of boundary maintenance among ethnic groups rather than with their distinct cultural contents. Barth has pointed out (1969) that in multi-ethnic interaction, ethnicity itself is a critical factor in status ascription. This is a simple and unassailable observation. The behavioural character of the ethnic unit must be demonstrated; it cannot merely be assumed a priori.

The observation that migrants in most of Africa's towns maintain close ties with their home communities, initially made with respect to labour migrants, has been extended to include all kinds of migrants—seasonal, semi-permanent, and permanent (Van Velsen, 1960: 265–78; Gugler, 1971: 405; Caldwell, 1969: 140; Mitchell, 1959: 12–47).

It has been stated that a migrant who spends most of his working life in the town still regards his community of origin as his real home, owes social allegiance to it, ‘and reduces rural-urban migration to a mechanical process involving forward and backward movement between the town and country’ (Imoagene, 1972: 4).

This paper investigates the nature and the role of such links between urban migrants and their various places of origin–both rural and urban–in Nigeria. The study is approached from a micro-level by examining the pattern of socio-economic relationships between individual migrants and their families at their various places of origin rather than using aggregate data.

Recently renewed interest in the history and social structures of the Mande-speaking peoples of the West Sudan has brought to attention a problem of both ethnographical and theoretical importance. It is the problem of understanding the structure and functions of clanship in Mande societies, while at the same time interpreting the meanings of clan names, rituals, myths, totemic usages, and inter-clan relationships. In approaching the first aspect of the problem one is faced with questions of definition which have not been satisfactorily resolved; for example, are the nyamakala groups properly defined as ‘castes’, and are the dyamu (or diamu) to be described as ‘clans’ or as ‘patronymic groups’.