This article examines the cultural dimensions of financial credit and debt, placing these against a deeper and broader background of entrustments and obligations. A standard response of the largest international aid agencies to African rural poverty has been to set up programmes to lend money and other resources to rural people without understanding what the borrowers already owe to other creditors and claimants, or how strong these competing claims are. The history of credit programmes has been a history of dismal failures and of disappointments for borrowers and lender alike, particularly where land mortgages have been involved. Intensive field research reveals that Luo farmers in Kenya, like other East Africans, already have a broad assortment of borrowings and lendings of their own, some far more meaningful to them than loans from banks, co-operatives, or marketing boards will ever be. Some are only partly economic in nature; some involve sacred trusts or important political contacts. Land, labour, animals, money, and humans themselves are all objects of entrustment and obligation among kin, neighbours, or other familiars. Farmers channel resources from socially distant institutions into uses that are often locally more meaningful than those their lenders intend; and they may not be at liberty to convert them back into liquid forms for repayment. Requirements of land title collateral misfit a cultural context where attachments to land, and to ancestral graves on it, symbolise an individual's or family's social identity. More broadly, the credit strategy of development aid needs rethinking. Rather than continuing to enmesh rural Africans in debts and uncertainties, those who purport to help reduce poverty in rural Africa should shift their strategy from lending to encouraging saving and investment, or to promoting other kinds of locally rooted initiatives not financial in nature.

A case study from western Kenya is used to explore the links between labour migration, rural economic decline and changes in key domestic relationships. Twentieth-century transformations in the regional political economy, together with processes of differentiation, have been closely bound up with changes, and continuities, in relationships within households, and in the ideologies which justify them. A central concept in the analysis is that of divisions of labour, which covers the division of tasks, divisions of spheres of responsibility and authority and contributions to the reproduction of the household. Changes in all these have shaped, and have been shaped, by the trajectory of economic decline in the region. Changing divisions of labour have been slow, piecemeal, non-uniform and non-linear. They have been the subject of intense conflicts within households which have centred on questions of access to and control over resources and in which, as well as power relations, ideas about rights and responsibilities have been crucially important.

Tree cultivation and management are a common form of land use in high-potential areas of Kenya. While some of these practices are related to economic considerations, such as markets and prices for specific tree products, others were derived from or developed in parallel with customary practices. This article traces the origins of contemporary demarcation practices in Kikuyu areas of Kenya, involving the planting of trees in hedges and windrows, from their customary antecedents. Customary law prescribed clear mechanisms for demarcating land to which rights of use had been acquired. These mechanisms, characterised principally by the planting of particular trees on the boundaries of land holdings, were given limited recognition by the colonial administration, and were subsequently incorporated (without any clear awareness of their customary role) in the contemporary body of land law which emerged as a result of the land reforms of the early 1960s. Land reforms tended to obscure customary distinctions between rights of control to trees and rights of use and access, by equating rights of control with rights of ownership. The result has been that rights of use and access, which had been guaranteed to the landless under customary law, were, for the most part, eliminated.

In Ghana, as well as in other parts of Africa, pentecostal Churches have recently become extremely popular. Within these Churches reference is made frequently to the devil, who is associated with the non-Christian gods and ghosts as well as Western luxury goods. Present Ghanaian popular culture reveals a striking obsession with images of the devil and of evil. By analysing stories told and published in Ghanaian ‘born again’ circles about money received through a contract with the devil or one of his agents, the author attempts to understand (1) what evil is denounced in these movements by means of the devil, and (2) how, with the help of the notion of the devil, ‘born again’ Christians think about poverty and wealth. It is argued that collective fantasies around the devil have to be understood against the background of difficult socio-economic conditions. These stories entail both a critique of the capitalist economy in the name of the pre-capitalist ideal of mutual family assistance (although a much more limited critique than Taussig suggested in his The Devil and Commodity Fetishism) and an opportunity to fantasise about things people cannot afford but nevertheless desire.

Analyses of ancestor-related practices were a crucial component of structural-functional models of social organisations in Africa. In the 1970s the theories of lineage and segmentary social organisation upon which these studies depended were called into question. Ethnographic and historical research indicated a preponderance of migratory activities and a high degree of discontinuity and reorganisation in kin relations. As a result most anthropologists turned from lineage-based functional theory to historical models of social organisation. With the waning of lineage theory, studies of ancestors in Africa became a marginal issue for most scholars of African societies. Ancestor-related practices, however, continued to be important in the lives of many African people. On the basis of data from Ohafia, Nigeria, the article suggests that the structural-functional model of social structure and the historical model of social dynamics both have parallels in indigenous representations of the ancestral past. In academic discourse these two models are taken as 'schools of thought' which pro-pound incompatible explanatory arguments. However, these apparently contradic-tory representations unite as an irreducible whole in the lived experience of the people of Ohafia. It is suggested that this indigenous paradigm of knowledge about the past provides valuable insights, not only into how we might productively theorise the social, but also for how we evaluate the contributions of our own intellectual ancestors.

In one small village in western Sudan local political struggles over power and resources are enmeshed in discursive struggles over representations of gender, ethnicity, class and community. Analysis of two specific conflicts illustrates this point. In one conflict over control of a village grain co-operative some villagers sought to exclude women, West African immigrants and the poor from participating in political decision-making. In a second conflict over a roadside market these same villagers, empowered by the divisive rhetoric and policies of the National Islamic Front regime, again mobilised dominant representations of class, gender and ethnicity in an attempt to prevent marginal groups from gaining economic advantage. In both cases, villagers translated national discourses in different ways, refracting ideology to meet their specific local interests. These conflicts and fractured discourses epitomise the process of civil breakdown that has marked the tenure of the NIF regime.