Student Number : 8905619P - PhD thesis - School of Geography, Archaeology and Environmental Studies - Faculty of Humanities / In this thesis I identify and clarify the archaeological signature of rain-control sites in the Shashe-Limpopo Confluence Area (SLCA). I use a landscape-based approach to investigate rain-control in the ideology of SLCA farming communities. I investigate the archaeology of ritual by viewing rain-control as materialised ideology. Using this insight, I examine the material culture and spatial manifestation of rain-control, the transition from ritual to residential sites, and how these transitions articulated with the ritualised landscape. Specifically, I explore the local manifestation of rain-control and its relationship with the ideologies of farming communities in the period leading up to SLCA state formation, between AD 1000 and AD 1250. I also scrutinize the relationship of the Leopard’s Kopje elite with hunter-gatherers and other farming people on the same landscape, as this relationship was partly grounded in ritual and raincontrol. Furthermore, this thesis explores the ideological roots of the Mapungubwe state. The ideology manifest in the location of the Mapungubwe royal residential area germinated during the K2 occupation. In this period rain-control was slowly removed from nature and located in farmer society. The final step in this course was nationalising rain-control and locating it on Mapungubwe hill. Initially, however, rain-controllers resisted this centralisation.

Calabrese (2005) identified two distinct ceramics styles in the Shashe- Limpopo basin at the same time – Leokwe and K2. This is the first record of ethnicity in the Iron Age of southern Africa. With this identification come new avenues for research. How these groups interacted, and their relative status through time is the focus of my research. According to Calabrese, some Leokwe groups maintained a higher, or at least equal status on initial contact with K2, before K2 became the dominant political group. He bases this claim on the identification of what he terms ‘Elite Symbolic Objects’ at sites, such as Castle Rock. Using GIS, it is clear that the locale of sites differ within the landscape. Specifically, locations vary through time on the escarpment and floodplain and their relationship to primary and secondary resources. This variation suggests that access to resources was controlled, and this implication influences ones assessment of the relative status of K2 and Leokwe groups. Additionally, new excavations at Castle Rock call into question the validity of ‘elite symbolic objects’ in determining status.