A New Approach to the Archaeology of Livestock Herding in the Kalahari, Southern Africa
Lindholm, Karl-Johan, Antiquity
Archaeologists in southern Africa disagree on how to identify livestock herders in the archaeological record: some stress that pastoralists produce sites with distinctive archaeological signatures, others that the identification of a pastoral package should be based on testable hypotheses about what pastoralism represents economically, socially and ecologically (Kinahan 1991). My approach tests a hypothesis, namely that herders would need access to water in the dry season, and I propose pastoral land use can be reconstructed by mapping artificial wells.
The study area is situated in the Kalahari of eastern Namibia (Figure 1). The region contains areas with thick sand layers which, in combination with a dry climate, result in high evaporation rates and paucity of surface water. For this reason the area was considered from colonial times onwards as generally unsuitable for livestock herding before deep-reaching boreholes were introduced in the latter half of the twentieth century. As a consequence, the archaeological record of pastoralists was perceived either as non-existent, or received little enquiry. However, a recent re-introduction of pastoralism helped to explain the long survival of the Ju/'hoansi, a Khoesan speaking people, who were previously considered as providing 'some of the best data we have on a full-time foraging ecology unaffected by the presence of cattle pastoralism' (Lee 1979: 73). During the second half of the twentieth century, their economy and egalitarian social structures attracted considerable attention from hunter-gatherer researchers.
[FIGURE 1 OMITTED]
In the 1980s, archaeological finds of early first-millennium AD pottery in Mahopa and cattle bone in/Xai/xai dated to the eighth century questioned the absence of herders in the past (Wilmsen 1988). Seen in relation to archaeological studies undertaken elsewhere in the region, the findings suggested that pastoral economies may have been established in this part of the Kalahari before the historical period (Denbow & Wilmsen 1986). For this reason, many of the basic assumptions (Lee & DeVote 1970) that had motivated the long term ethnographical research on the Ju/'hoansi (Wilmsen 1989a) were questioned. Sadr (1997) pointed to the paucity of published archaeological evidence to back up some of the claims made by Denbow and Wilmsen (1986). The single maxilla found in /Xai/xai is so far the only find of domestic livestock remains that has been made in the Nyae Nyae Dobe area. In addition, the radiocarbon date associated with the bone is disputed (Yellen & Brooks 1988; Sadr 1997). Wilmsen (2003) has responded by presenting further historiographical information and pointed out a series of inconsistencies in their attacks on his study.
One of the central controversies, and not only in the Kalahari debate, is how pastoralism should be identified in the archaeological record. Conventionally, livestock remains have been considered the prime indicator for pastoralism and pottery is often taken as proxy evidence (Kinahan 1991). In the Kalahari, the early agropastoral settlements at Tsodilo Hills contained large amounts of bone from domesticated fauna and are for this reason generally accepted (Mitchell 2002). In the marginal Kalahari grasslands to the south-west there is a general absence of archaeological sites with faunal remains of cattle, sheep and goat. For this reason, finds of pottery and metals in the area have not been interpreted as proxy data for the presence of herders with livestock, but as exotics indicating incursion by pastoralists and agropastoralists living elsewhere (Sadr 1997; Smith & Lee 1997). The model is to a large extent verified by the opinion that livestock herders 'had little interest' in the region owing to the lack of water and the presence of non-palatable grasses not preferred by cattle (Smith 2000: 69).
It is apparent that the premise for the model is that the activity of livestock herding is necessarily represented in the archaeological record by the bones of domesticated livestock. …