Sedentarization is often characterized by increased cultivation, which can result in either a partial or a total transformation from a livestock to a farm-based economy (Barth, 1961; Baxter, 1975; Nelson, 1973; Salzman, 1980). There are many possible permutations along this continuum from livestock to agriculture, and categories based on a sharp dichotomy between them are inadequate; even sub-units within a single ethnic group may show divergent patterns. Recent evidence reveals considerable intra-group variation in pastoral responses to agricultural opportunities (Alverson, 1979; Behnke, 1981; Haaland, 1977; Kjaerby, 1979; O'Leary, 1980). These differences are often in line with the generalization that large cattle herds are more closely associated with pastoralism than with agriculture.

The early years of the 1970s saw two important milestones for the political economy and historiography of Kenya. The ILO report of 1972 was a major critique of Kenya's development efforts since independence that identified growing problems of poverty and inequality which were the result of policies that in many other respects achieved desirable objectives (International Labour Office, 1972). The ILO diagnosis and proposals had some direct influence on government policies, notably in the Third and Fourth Development Plans, for 1974–78 and 1979–83 respectively, and several but by no means all or even the majority of the proposals were acted upon, e.g. in policies towards agriculture (Clayton, 1978). Equally important, however, was its indirect effect on creating a climate of opinion that emphasized a ‘basic needs’ approach (Ghai et al., 1979) and ‘redistribution through growth’ (Chenery et al., 1974) and brought the attention of the public, the government and of scholars to the problems of rural and urban poverty in Kenya at a time when they seemed to be growing alarmingly.

In June 1978, following a series of arguments in her marital home, a young Temne woman called Isatu travelled from her home in Makeni (a town which is the capital of Sierra Leone's Northern Province) to the village of Petbana Masimbo three miles away in order to consult a diviner about the cause. The diviner she consulted was Pa Koroma, a well-patronized Muslim diviner who was also her father's brother. Pa Koroma happened to be conducting a session of the dramatic and relatively expensive α η-yina Musa mediumship method of divination, which is commissioned by a wealthy client but which more minor clients such as Isatu can also participate in. He put a female medium, completely covered by a protective white cloth and gazing into a white bowl of water containing Islamic medicine (mǝ-nαsi), into a trance in which she was able to communicate with αη-yina Musa, the fearful ‘Muslim’ spirit of divination. Around Pa Koroma's central parlour, in which the divination took place, were seated members of the village, some of whom had come to ask questions, others simply to watch and, where appropriate, to interject comments on what was being said. The medium, an elderly woman called Mami Yeno, was a sister-in-law of Pa Koroma from Makaper, a village about two miles from Petbana. Each person who had come to ask about a problem addressed their question to Pa Koroma, who repeated it to Mami Yeno. Mami Yeno then silently conveyed the question to αη-yina Musa and, after a pause, returned the reply to Pa Koroma, the latter repeating and often elaborating upon her utterances to the audience.

Anthropologists are beginning to investigate the sociological dimension of dreams, at least of those purporting to be channels of religious revelation. Two recent examples of this tendency are Charsley (1973) and Curley (1983), each of whom researched dream phenomena in an African independent church. The sociological approach is a firm rejection of earlier and some current attempts to extract patterns of culture and personality from the subjective reconstruction of the dream-as-dreamed, a purely private event. In contrast, the sociological emphasis isolates as its primary datum the dream-as-told to others, an essentially social act, which leads to an analysis of dreams as social assets, which can be manipulated to advantage or disadvantage. Consequently, the method of analysis demands that as much attention be directed to the act of recounting as to the content of the narration, although in practice this seems to mean that content is subordinated to purpose and can be related to social arrangements only indirectly through purposive action. I shall endeavour to correct this latter impression in subsequent argument. The second methodological requirement is to connect narration to social action, social relations and social organization, and to demonstrate the tactical use of dreams in social encounters.

The relationship between land tenure forms and the twentieth-century transformation of agriculture throughout much of Africa from a strong subsistence to a market orientation – and from food crop production to predominantly export crop production – is a topic of great importance both to Africanists and to policy-makers. From a scholarly perspective, contemporary Africa presents an unusual opportunity to study dramatic changes in land use and especially in land tenure, similar to changes which took place in other parts of the world as commercial farming developed. From a policy-maker's point of view, too, the current shift in land tenure has major implications for possible government measures affecting security of tenure and land distribution – and, in turn, patterns of agricultural investment, productivity and output.