Search Results

Edwin N Wilmsen and James R Denbow


Tora Nju is the local name for a collapsed stone walled enclosure situated approximately 20km from Sowa Spit, 200 m south of the Mosetse River, and 7 km east of the present strandline of Sowa Pan. The site that takes its name from this ruin includes several midden areas containing pottery, stone tools, and faunal remains along with house structures and grain bins. Excavations were carried out in parts of all these site components. The middens contained a moderately rich suite of materials including sherds, glass and shell beads, metal, and animal bones. The enclosure, however, yielded very little. Consequently, we concentrate here first on the middens before turning to the enclosure. Typical Khami vessel forms predominate throughout the midden stratigraphy; a few midden sherds are comparable with Lose wares in part contemporary with Khami ceramics. A possible earlier Leopard’s Kopje presence is also indicated. Glass beads characteristic of Khami Indo-Pacific series were also recovered from all midden levels. Three charcoal samples yielded contradictory radiocarbon dates for the middens, and we have no direct means for dating the enclosure. We evaluate evidence for a takeover of Sowa salt production by the Khami state sometime in the early 15th century. Finally, we examine historical records and incorporate current linguistic and dna studies of Khoisan and Bantu speakers to illuminate the social history of the Tora Nju region.

Edwin N. Wilmsen, David Killick, Dana Drake Rosenstein, Phenyo C. Thebe and James R. Denbow

Over the last 30 years Wilmsen and Denbow have recovered and studied pottery from 28 sites in Botswana dated between ca cal AD 200 and AD 1885. Some sherds in several of these assemblages appear, on stylistic evidence, to have been made in other sub-regions of Botswana than where they were found. These inferences are confirmed in this paper by use of an independent archaeometric technique, optical petrography. We are able to demonstrate the transport of pots from the Okavango Delta to Bosutswe in the eastern hardveld, some 400–600 km distant, as early as cal AD 900–1100, and of others over equal distances to the Tsodilo Hills probably before that time. We are also able to demonstrate several shorter itineraries at contemporary and later times in the Tsodilo-Delta-Chobe region as well as in the hardveld. Furthermore, we demonstrate that clays were transported from geological deposits to sites where pots were made from them. We consider some implications of these findings for a deeper appreciation of the movement of peoples and goods at several time periods of the past and present as well as further implications for understanding the participation of the region in the Indian Ocean trade during the 8th–10th centuries.

Edwin N. Wilmsen, Anne Griffiths, David Killick and Phenyo Thebe


Current potters in Manaledi village in the Tswapong Hills of Botswana aver that they and their ancestors for five generations have made pottery exclusively with clay from nearby sources. We begin with an examination of Manaledi and its clay mine to uncover current dialectics between village, landscape, clay, potters, and ancestors. Archaeological sherds found around the village and clay sources document occupation by makers of Early Iron Age (ca. AD 500-750), Middle Iron Age (ca. AD 750-1050), Late Iron Age (ca. AD 1420-1800), and 18th-20th century wares related to current Manaledi pottery. The proximity of archaeological deposits, clay sources, and village made it possible to conduct simultaneously what might otherwise be considered three separate projects. As a consequence, we are able to document that Manaledi clays have been used to make pottery for some 1500 years and to consider long-standing constraints on potting this implies.