Search

We conducted experiments to compare the micromorphological signatures of modern burnt sedge and grass bedding to laminated layers of carbonized material and phytoliths in Middle Stone Age deposits at the shelter, Sibudu. The experiments were designed to clarify the formation processes associated with the laminated layers and to investigate whether these previously identified layers of bedding were deliberately burned or not. The results indicate that the laminated layers were most likely produced by human activity related to the construction, maintenance and burning of bedding. Furthermore, our experiments demonstrate that large volumes of vegetal material could have produced the relatively thin, archaeological deposits of burnt bedding.

We show that carbonized fruits and seeds recovered from Middle Stone Age deposits in rock shelters are likely to have been carbonized as part of post-depositional processes. We buried indigenous South African fruits, nuts and seeds at pre-determined depths and distances from the centers of experimental fires. The cold ashes of the hearths and the sand surrounding them were subsequently excavated, using standard archaeological techniques and dry-screening. The fruiting structures from the oxidizing part of the fire were burnt to ashes, whereas those buried in sand under anoxic conditions survived in varying forms. Those buried 5 cm below the center of the fire were carbonized; those buried 10 cm below the center of the fire were dehydrated; and those 5 cm and 10 cm below the surface at the outer edge of the fire were unaffected. Size, moisture or oil content of the original fruit or seed did not appear to influence whether or not carbonization took place. Temperatures recorded 5 cm below the experimental fires suggest that the carbonization occurred at or before a maximum temperature of 328 °C, and also at lower maximum temperatures (152 °C) that were maintained for long periods. Even when the quantities of a particular wood are controlled, open fires may produce variable underground temperatures and the temperatures below ancient hearths would have been equally variable. We suggest that Cyperaceae (sedge) nutlets, the most numerous fruiting structures in the Sibudu Middle Stone Age archaeobotanical assemblage, occur in the shelter as a result of human activity, subsequent burial, and accidental carbonization when hearths were built directly above the buried nutlets.