Oxford Handbook of African Archaeology, edited by Peter Mitchell and Paul Lane, Oxford University Press 2013 http://ukcatalogue.oup.com/product/9780199569885.do#.UeqzRW1m9JI Chapter 44: The archaeology of herding in southernmost Africa Karim Sadr Introduction During the last fifty years or more the conventional view has been that livestock and pottery first reached the frontiers of southern Africa with immigrant, Bantu-speaking farmers. The herding way of life then spread southwards with the Khoekhoe-speaking branch of the indigenous ‘Khoesan’ populations (Walker 1983; Elphick 1985; Parkington et al. 1986; Smith, 1990, 1992; Boonzaier et al. 1996). Archaeologically, the Bantu migrations are well documented (Huffman 2007; de Maret, this volume; Mitchell, this volume), but the evidence for Khoekhoen migration is meagre. Indeed, linguistic and chronological evidence refute the conventional view. In many southeastern Bantu languages the words for livestock are of Khoe origin (Ehret 2008), and southern African pottery predates the arrival of the first farmers (Sadr and Sampson 2006). Two alternatives present themselves. Either, the earliest livestock and pottery reached southernmost Africa by a process of diffusion and without the help of any migrating herder-potters (Deacon 1984; Klein 1986; Kinahan 1996a; Sadr 2004, 2008a). Or, a migration of non-Bantu speaking herders brought livestock to the subcontinent before the arrival of Bantu farmers (Ehret 2008; Fauvelle-Aymar 2008; Güldemann 2008; Blench 2009). The remainder of this chapter discusses the archaeological evidence from these two perspectives (Fig. 1). Ceramics The earliest Later Stone Age (LSA) pottery from the western parts of South Africa, northern Namibia, and northern Botswana/western Zimbabwe is all thin walled and well fired (Sadr and Sampson 2006). Vessels are generally small. Before about AD 300, in the South African coastal areas and the Orange River Basin, thin ware was undecorated (Sadr 2008b). To the north, in the Limpopo Basin, the Kalahari and northwestern Namibia, thin ware was decorated with incised lines, comb stamped impressions and an effect referred to as rippling. Spouted vessels were not uncommon here, but absent in the south. A ceramic-free zone separates these early northern decorated thin wares from the southern plain ones, but this gap could reflect insufficient research. Further north, in the upper Zambezi and Congo Basins and along Angola’s coast, the earliest thick wall ed ceramic pots (later clearly associated with Iron Age, farming villages) are known from Benfica in Angola and Situmpa in western Zambia (Clark and Fagan 1965; Dos Santos and Ervedosa 1970). Further south, thick ware first appeared in the late third century AD at the farming village of Matola near Maputo in Mozambique (Morais 1988; Maggs and Whitelaw 1991). Given the much earlier dates of southern thin walled pots and the absence of thin ware in Central and East Africa, there is little reason to think that the first southern African potters were influenced by immigrant farmers from further north.