In his Economic History of West Africa (1973) Hopkins points out that relatively little attention has been paid to the history of food production by contrast with export crops, even though it has been clear since early research on African food systems (e.g., Johnston 1958) that patterns of production have been changing. The determinants of shifts in land use and crop rotations are complex but two major factors have been suggested: population pressure on land resources, and the relative prices of different crops. The population pressure argument tends to assume that subsistence is maintained, so that any change in the relationship of population to food land requires shifts in farming practice to allow the maintenance of the same level of living (Boserup 1965). The price argument tends to assume that the agriculture system is penetrated by the market principle, so that farmers' decisions to maintain subsistence production patterns depend on projections about the prices of the cash crops available for sale and the food items needed to purchase (Chibnik 1978). From work on African farming systems comes a modification which suggests that the management of both these constraints depends to some extent on the broader social and economic context in which decisions are made. In particular it has been suggested that the position of women farmers in both indigenous social organisation and national economies is different from men's; they work under different constraints in their farming and have different opportunities for alternative employment (Boserup 1970; Meillassoux 1975). If the sexual division of labour is an important aspect of farming, men's and women's differential access to resources might be expected to have an independent effect on cropping patterns.

It has only been in recent years that anthropologists have begun to investigate women as actors with resources important in the political field and to broaden the concept of political system to include the activites of women (see Collier, 1974 and Nelson, 1974 for another statement of the problem). The analysis of the political processes of the nineteenth century Kikuyu (Kenya) presented here lays the framework for the inclusion of women's subsistence activities and domestic decision making with the field of public or political relations.

As this important Nuer myth suggests, men's control of the disruptive forces around them is limited and problematic. In part this limitation stems from relations of mutual dependence which continually restrict the power and authority of men in Nuer society. In speaking of their dependence upon cattle, Nuer say: ‘“They will be finished together with mankind”, for men will all die on account of cattle and they and cattle will cease together’ (1940:49). Similarly, the dependence of men upon the reproductive and nurturing powers of women assures the latter an important source of control: mutual dependence implies a mutual independence. Women have an exclusive realm of activity and hence an exclusive domain ofxontrol and influence. Moreover, they can subvert political alliances and aggravate divisions within the male hierarchy through the manipulation of their children's loyalties. In this essay, I examine these bonds of mutual dependence uniting men and women in Nuer society so as to outline some of the sources, characteristics and limitations of male domination and authority.

Only recently have precolonial African historical studies broken away from their fixation on dynastic histories and trade. In reassessing the dynastic traditions, historians of the African past have increasingly drawn on oral traditions from non-royal sources. In particular they have placed a great deal of emphasis on ‘clan traditions’ and the clan group has consequently served as an important building block in reconstructing the African past. But in Rwandan studies this has often been undertaken in a rather uncritical manner; in fact, historians may have been building on a foundation which is less enduring than they realize. For their new construct is often based on the assumption of ‘primordial’ clan units reaching far back into the past, antedating the dynastic political units of which they came to form a part. It is usually argued that these enduring clan units could absorb individuals from other clans or experience geographical expansion through migration; but the basic structure of clan identities is seen as unchanging.

Orthodox theories of social change emphasise the importance of formal training in the structural transformation of traditional systems within less developed countries (Peil 1973; Dore 1978). It is assumed that through investment in human capital economically-backward societies will achieve modernisation and growth, and that a sequence of changes in the social and occupational structure following the typical Western model will occur. A close association between educational and occupational attainment is postulated which forms the basis for a new system of social stratification defined by educational qualification and income level (Foster 1977). Intervening between the academic achievement of individuals and their entry into the job market an important additional spatial variable is included in the dual sector model (Lewis 1954; Forde 1968). Given that rural-urban dichotomy exists between traditional and modern sectors the model suggests that considerable selective movement from countryside to town is an additional necessary condition for economic growth. Finally it is assumed that the resultant spatial separation of the educated from their traditional environment coupled by secure employment in town consolidates the process of acculturation into a Western value system (Foley 1977).