According to their ethnographers, the Bena of the Rivers in colonial Tanganyika emerged ‘from the shadows into the light of day’ when they undertook a great migration from the forested hills of the Iringa highlands to the treeless floodplain of the Kilombero valley in the late 1800's. The phrase had two meanings – on one level, it described the physical relocation of the Bena from the shadows of the forest to the open floodplain. On another level, ‘light of day’ was a metaphor for the remembered cohesion of a distinct Bena tribal identity, distinguishable from the ‘shadows’ of prior ethnic obscurity. The Culwicks considered the chief or mtema of this period, Ndaliwali, to be the founding ancestor of a new political lineage.

The defining moment for the emergence of the Bena of the Rivers was the Battle of Mgodamtitu in 1874, when neighboring Hehe attacked Bena settlements in the foothills and forced the Bena permanently out of the highlands and into the valley. Once they had taken up their new residence, the Bena absorbed the existing settlements of the western end of the valley into their kingdom. They did this initially ‘under the guise of protector’, wrote the Culwicks, but their goal was political control of the region. By 1890, they wrote, most of the Ndamba as well as several smaller valley groups had become subjects of the Bena chief.

The Battle of Mgodamtitu caused a ‘sudden and wholesale change’, according to the Culwicks, because Bena were forced to give up cattle keeping and take up rice cultivation; to learn the ways of the river and skillful use of canoes. They had previously looked down upon ‘rice and fish-eaters’, but after their migration they found that ‘there is but one crop really worth growing, and that is rice’.

This essay is about a conceptual category of historical and environmental knowledge and about how a particular group of Africans use that category to understand and debate change. It is, in effect, an exercise in translation. In the middle and upper Sangha basin forests of the Central African Republic (C.A.R.) and Cameroon, Mpiemu speakers have articulated a broad category, doli, through which they express, debate and make claims of truth about the past and present. Glossing doli as ‘history’ does little justice to the richly complex dimensions of this category, for doli encompasses a multitude of relationships to the past. It can refer to a distant unchanging past, as well as to the knowledge, beliefs and practices associated with that past. Mpiemu people hold up the knowledge, beliefs and practices as an idealized framework to guide their behavior toward one another and their uses of fields, forests, rivers and streams. But doli can also describe and frame the accumulated experiences – identifiable events, people and places – of elderly people. In all of these expressions about the past, Mpiemu use idioms linking persons and their environments, those of cords and vines and of mobility (wandering) and stasis (sitting), to articulate doli's central aim of ‘leaving a person behind’. Tracing doli's different meanings, genres and aims can illuminate how the category has changed over the twentieth century, how Mpiemu have interpreted environmental interventions in the Sangha basin, and why they have engaged in conflicts over their entitlement to valued forest resources. Hence, it offers insights into why people use natural resources as they do and provides an alternative to exclusively materialist explanations for conflicts over resource use.

The days are approaching when you will hear new things. Soon you will hear that an Arab is not an Arab, a Shirazi is not a Shirazi, and an African is not an African. You will hear this, and you will be told that you are all Zanzibaris.

Al-Falaq, paper of the Zanzibar Arab Association, 1946

Nani awezaye kumnyoosha Binaadamu pindi alitiwa kibyongo na Mungu? [Who can straighten out mankind, whom God has made a hunchback?]

Afrika Kwetu, paper of the Zanzibar African Association, 1952

STUDENTS of the history of African ethnicity will recognize in the second of these quotes a Swahili version of the aphorism from Kant with which Leroy Vail introduces his important volume on the creation of tribalism in modern Africa: ‘Out of the crooked timber of humanity nothing straight will ever be fashioned’. True to their spirit, both aphorisms can be interpreted in contradictory ways. Vail intended to point up a theme that runs throughout his volume: that ethnic categories are rarely as clear-cut as ethnic nationalists would have them seem, and that, where such clarity exists, it is only momentary, having arisen out of complex and messy historical processes. But the editors of Afrika Kwetu, the weekly newspaper of the Zanzibar African Association that published the Swahili proverb in 1952, drew less subtle implications. The immediate context was a plea against religious chauvinism, part of the newspaper's allegation that Zanzibar's Arab nationalist leaders had violated the Islamic injunction to respect other religions ‘of the book’. But the clear implication – elsewhere made explicit – was that God and nature had fashioned humankind into irreducibly separate races and nations, and that it was foolish and even blasphemous to mix or combine them, as (they charged) their Arab rivals were seeking to do.

RAPID population growth is commonly depicted as one of the greatest problems facing modern Africa. For decades, the tendency of birth rates to exceed mortality rates has prompted predictions of land shortage, resource depletion and mass starvation. Underlying causes of high fertility are hypothesized to have been an unusually high demand for human agricultural labour, ‘traditional religious pronatalism’ and a ‘horror of barrenness’, while in some areas the later colonial period saw a shortening of the durations of post-partum sexual abstinence and lactation. Mortality decline from the 1920s is commonly linked to the establishment of cash crop economies, networks of roads and railways, and the diffusion of western medicine, maternity facilities, missionary activity and primary education. Yet the empirical evidence supporting this model of population growth is contradictory. Areas such as Buhaya, Buganda and Bunyoro should have experienced rapid demographic expansion by natural increase in the colonial period according to dominant theories but instead experts in the early decades of this century feared the extinction of the Haya, Ganda and Nyoro. This paper will attempt to explain why population decline among the Nyoro was more severe than anywhere else in colonial Uganda, and probably East Africa.

THE recruitment of African labour at poor rates of pay and under primitive conditions of work was characteristic of the operation of colonial capitalism in Africa during the nineteenth and twentieth centuries. The implications of these conditions have been generalized very widely in the historiography of colonial Kenya. Where capital was centred upon extractive industries or upon settler agriculture (as in Kenya), historians have found much evidence to indicate that colonial states (and the metropolitan government) readily colluded with capital in providing the legal framework within which labour could be recruited and maintained in adequate numbers and at low cost to the employer. The state itself was the largest employer of labour throughout British colonial Africa and shared an interest in encouraging Africans into the labour market. Criticisms of labour conditions prevailing in any colony were thus likely to be interpreted as criticisms of the state itself.