Forager variability and transitions to food production in secondary settings

The present work tests a hypothesis that ca. 8,000-2,000 years ago in East Africa, a society existed that systematically differed in terms of mobility, economy and social organization from both immediate- and delayed-return foraging models as traditionally defined by anthropologists. If true, this would suggest that there was greater variation in prehistoric foraging adaptations than is seen today. It could also address the question of why some African foragers adopted food production, while others did not. Methods used include excavation, dating and faunal analyses of four sites in the Lake Eyasi basin of northern Tanzania, and faunal analyses and dating of two sites in the Lake Victoria basin of western Kenya. These two areas share a ceramic tradition called Kansyore, an immediate predecessor to the Pastoral Neolithic (PN). Kansyore ceramics were recovered at half the sites studied, while the others contained PN and later ceramics. The data from these sites are compared with published archaeological and ethnographic data to interpret the economic systems represented. This study confirms that in the Lake Victoria basin, Kansyore foragers represent a moderately delayed-return foraging system that is distinct both from other prehistoric foraging traditions and from modern African foragers, making Kansyore sites unique in sub-Saharan Africa. This may explain why domestic animals are found in some late Kansyore sites, since delayed-return foragers would be more inclined to adopt herding. By contrast, data from the Eyasi sites do not suggest a delayed-return foraging economy, but do give mixed signals for foraging and herding economies. This mixture is typical of the region and indicates a patchy, gradual transition from foraging to food production. This work contributes to East African prehistory in three specific ways: it represents the only large-scale zooarchaeological study of Kansyore sites, which have been hitherto neglected; it provides archaeological data and dates for the poorly-known Eyasi basin; and improves understanding of the chronology and tempo of the spread of herding. At a broader level, this study suggests that anthropological models of forager socio-economic organization can be flawed when applied to archaeological data, since there was likely far greater forager variation in the past than is seen today.