The ‘moving frontier’ and the transition to food production in Kenya

Abstract Since at least the 1960s, models of the transition to food production in East Africa have tended to be driven by various hypotheses derived from historical linguistics concerning the spread of various language families across the continent through the process of population migration. In more recent years, archaeologists, historians and historical linguists have offered various refinements to the initial models. This has certainly encouraged more nuanced reading of the different sources of evidence. However, such accounts rarely consider in any detail the possible variable rates of adoption or the different social conditions under which this occurred. Consequently, despite awareness of the likely complexity of the processes and the growing range of critiques, this critical transition in the history of East Africa is still widely thought of as having universally entailed the co-appearance and/or adoption of particular ‘packages’, namely cereal cultivation, iron smelting and proto-Bantu languages in the case of early farming communities, and domestic livestock and proto-Cushitic or Southern Nilotic languages in the case of early herders. In the former case, this transition is also often thought to be signalled in the archaeological record by a single diagnostic marker—the presence of Urewe and other Early Iron Age pottery styles. Building on the pioneering work of John Alexander in the application of ‘frontier theory’ to account for variable rates and mechanisms for the transition to agriculture, this paper offers a summary of the current state of knowledge regarding the transition to food production in parts of western and central Kenya.