nents of southern African Iron Age sites. Their correct identification is crucial to understanding technological processes performed at these sites. This paper presents criteria for distinguishing between iron smelting slags, iron forging slags, copper smelting slags, crucible slags resulting from melting activities, vitrified clay and various biomass materials. Slag identification should entail a combination of morphology, microscopic study, chemical analysis, and assessment of the archaeological context. It is a necessarily specialist activity and superficial classification without materials analysis can be misleading. Archaeologists need to be mindful of both the archaeological opportunities and the potential technical difficulties in the interpretation of slags.
This paper provides the results of a detailed metallurgical analysis of the gold, copper, bronze and iron artifacts and slag recovered from excavations, carried out in 1990 and in 2001–2002, at Bosutswe on the eastern edge of the Kalahari Desert. While we find that the general manufacturing technologies of smelting and metal artifact production did not change greatly over time, and are indeed similar across vast distances of southern Africa, the cultural context of these materials attests to their importance as productive tools and weapons, as well as jewelry and ornamentation that were important in the construction of sumptuary distinction and social status. The important new technology of copper-alloy bronze production makes its appearance at Bosutswe around CE 1300. The quantity of bronze goods recovered indicates that during much of the 2nd quarter of the second millennium CE the occupants of Bosutswe participated in elite networks of inter-regional exchange and luxury consumption that were dominated by the larger regional polities of Mapungubwe, Great Zimbabwe, and Khami. While the occupants of the site were able to express some degree of political and cultural autonomy through their elaboration and use of uniquely styled ceramics, their subordinate position vis-à-vis these more powerful entities was also attested through many of the same mechanisms — the possession of small numbers of imported glass beads and iron, copper and bronze ornaments, and the occasional gold bangle.
The archaeological evidence for iron and especially copper production at Marothodi indicates that output far exceeded local requirements. Preliminary analyses of slag and metal provide insight into the technical processes of this production, while well-resolved spatial data comment upon the social and cultural organization of production. In this paper we attempt to integrate both technical and social aspects of production into the regional historic context with a view to developing ideas about the contextual specificity of surplus metal production from Marothodi early in the 19th century. Generally, Marothodi was occupied in a period of increasingly competitive economic and political relationships between lineages. The evidence from Marothodi indicates that although copper ore quality was poor, and had been largely mined out by previous producers, it was clearly worthwhile to produce a surplus because of regional demands. Importantly, the Tlokwa elites at Marothodi had the regional power to do so. Furthermore, although the location of Marothodi was a compromise between several factors, we suggest that proximity to the copper ore sources was important. Spatial data suggest that political authority did not physically centralize copper production, and that most home-steads were independent producers.