Search

The archaeological record of Africa provides the earliest evidence for the emergence of the complex symbolic and technological behaviours that characterize Homo sapiens1–7. The coastal setting of many archaeological sites of the Late Pleistocene epoch, and the abundant shellfish remains recovered from them, has led to a dominant narrative in which modern human origins in southern Africa are intrinsically tied to the coast and marine resources8–12, and behavioural innovations in the interior lag behind. However, stratified Late Pleistocene sites with good preservation and robust chronologies are rare in the interior of southern Africa, and the coastal hypothesis therefore remains untested. Here we show that early human innovations that are similar to those dated to around 105 thousand years ago (ka) in coastal southern Africa existed at around the same time among humans who lived over 600 km inland. We report evidence for the intentional collection of non-utilitarian objects (calcite crystals) and ostrich eggshell from excavations of a stratified rockshelter deposit in the southern Kalahari Basin, which we date by optically stimulated luminescence to around 105 ka. Uranium–thorium dating of relict tufa deposits indicates sporadic periods of substantial volumes of fresh, flowing water; the oldest of these episodes is dated to between 110 and 100 ka and is coeval with the archaeological deposit. Our results suggest that behavioural innovations among humans in the interior of southern Africa did not lag behind those of populations near the coast, and that these innovations may have developed within a wet savannah environment. Models that tie the emergence of behavioural innovations to the exploitation of coastal resources by our species may therefore require revision. Human populations in the southern Africa interior were collecting non-utilitarian objects at around 105,000 years ago, suggesting that the development of this innovative behaviour did not depend on exploiting coastal resources.

Over the last few decades, investigations of the southern African Late Pleistocene archeological record have transformed our understanding of Homo sapiens origins and evolution. However, the intensity of research on coastal and near-coastal records has surpassed that in the deep interior. One reason for the geographic bias is the rarity of well-preserved, datable rockshelter sites outside the more mountainous and karstic regions of South Africa. Here, we report stratified Middle Stone Age and Later Stone Age deposits at Ga-Mohana Hill North Rockshelter (GHN), near Kuruman, in the Northern Cape, South Africa. Contrary to a previous report, we show that the archeological deposits are in good context with minimal disturbance based on a detailed fabric analysis of plotted finds. Three dating methods, single-grain optically stimulated luminescence, radiocarbon, and uranium-thorium, indicate that the artifact-bearing deposits date to ~ 105 ka, ~ 31 ka, and ~ 15 ka. Thus, Ga-Mohana Hill North Rockshelter provides an important opportunity to investigate early human adaptation in this semiarid region of southern Africa.

The archaeological record of Africa provides the earliest evidence for the emergence of the complex symbolic and technological behaviours that characterize Homo sapiens

Early archaeological evidence for symbolically-mediated behaviour, which is our ability to create and share coded information between and within groups, comes from the African Middle Stone Age. Nassarius kraussianus shell beads, discovered in the Late Pleistocene, Still Bay archaeological deposits at Blombos Cave, Western Cape, South Africa, are some of the worlds earliest personal ornaments and their discovery significantly pushed back the origins of complex human symbolling. Further analyses of these beads led to the hypothesis that stringing arrangements at Blombos Cave changed through time, with important implications for the development and maintenance of social norms and style in early human populations. This hypothesis was supported by qualitative comparisons of archaeological and experimental wear distributions. Here, we present the results of a quantitative approach, applying a modified edge damage distribution method and statistical modelling to published diagrams ( Vanhaeren et al. 2013 , Journal of Human Evolution 64, 500–517) of wear on N. kraussianus shell beads. Our results support the original findings that different beading arrangements result in different wear distributions, and that the wear distributions on Blombos Cave beads exhibit temporal variability. However, our results vary with respect to which stringing arrangements best match the archaeological samples. Furthermore, we conclude that a combination of multiple processes may best explain the archaeological wear distributions, a finding more congruent with a long and complicated life history of curated objects like beads. These findings add to a growing record of early human social behaviours, and contribute methodologically to use-wear analyses of personal ornaments recovered from the archaeological record.