(In press: Not for citation)

Paper invited for Colloquium X the XIII International Union of Prehistoric and Protohistoric Sciences Congress, FORLÍ, ITALY. SEPTEMBER, 1996

Southern Africa in the debate on the origins of modern people


Department of Archaeology,
University of Stellenbosch
Matieland 7602, South Africa.
e-mail: hjd@maties.sun.ac.za


The debates on the emergence of modern people are in fields as diverse as molecular biology, physical anthropology and archaeology. Many have to do with regional questions, like the fate of the Neanderthals in western Europe or how, when and from where of getting to Australia. These parochial concerns are not separate from the larger question of the geologically recent global expansion of modern humans. They are part of the understanding the diaspora of modern people. This paper is concerned with southern African region. It argues that the evidence supports an early Late Pleistocene expansion of morphologically and behaviourally modern people in Sub-Saharan Africa. The emergence of modern humans in Africa would date even earlier in time. The paper emphasizes the importance of evidence for past population histories in regions of Africa.

Southern Africa is one of the better researched regions of Sub-Saharan Africa with a tradition of Palaeolithic research that began in the 1920s. The region was inhabited from the time of the primary distribution of humankind and stone artefacts, of all ages and as virtually permanent markers, are abundant at open stations and rock shelters in the landscape. Some of the stratified sequences, notably Border Cave and Klasies River, have received prominence in the debates on the emergence of modern humans because of the claims made for the age of the fossils and artefacts found at these sites.

Sequences and dating

A.J.H.Goodwin (1958) formulated the culture stratigraphic sequence in a pioneering synthesis of South African archaeology. He avoided using established European terminologies because any direct correlation between the archaeological units, recognized the length of a continent apart, could not be shown. With hindsight this was the proper course to follow but the stage divisions he proposed, the Earlier, Middle and Later Stone Ages, have a broad technological equivalence to the Lower, Middle and Epi-Palaeolthic stages in the classic European sequence.

These stage divisions were formulated on acceptable stratigraphic observations but there was no appreciation of the duration of the divisions. The manufacture of Earlier Stone Age Acheulian bifaces may have ceased by 200 000 or more years ago but reliable dating for the termination of the Acheulian is still lacking. Middle Stone Age occurrences, typologically not the earliest phase, are dated by alternative methods to radiocarbon to more than 100 000 years (Deacon 1992). Even older dates are available for the same stage elsewhere in Sub-Saharan Africa (Mehlman 1991). The most recent Middle Stone Age occurrences fall within the range of radiocarbon dating with the 22 000 year-old site of Strathalan (Opperman & Heydenrych 1990) one of the best dated sites from the end of this stage. The dating of the Later Stone Age is similar to that of the Epi-Palaeolithic of North Africa and the Late Stone Age (a corruption of Later Stone Age) of other parts of Africa.

Advances in dating techniques are making it possible to interpret the Middle and Late Pleistocene archaeological evidence with finer temporal resolution than afforded by the broad stage divisions. Interregional and intercontinental correlations of sites and their evidence are essential for the study of the global expansion of modern people.

Associated human fossils

The Earlier Stone Age is associated with archaic humans, the Middle Stone Age with early modern humans and the Later Stone Age with the immediate ancestors of contemporary populations. In Sub-Saharan Africa continuity rather than population replacement is accepted under both the competing recent-African-origins and multiregional models for the emergence of modern people. Continuity without large scale replacement suggests that in Africa there should be a graded series of fossils exhibiting increasing selection for more modern characters with time. The series is far from complete but in South Africa the Saldanha calvarium, like Kabwe, may represent archaic Homo sapiens, Florisbad a later archaic population, and the Klasies River and Border Cave fossils within the range of modern humans.

The Klasies River finds come mainly from two horizons. All the finds reported by Singer & Wymer (1982) relate to the younger 90 000 year old horizon and finds dated to 120 000 years have been reported from the ongoing excavations. These fossils represent a single dimorphic population within the modern range of variation (Rightmire & Deacon 1991, Bräuer et al. 1992). The provenance of the important original Border Cave finds has had to be reconstructed (Beaumont 1980) and although the deposits are of similar age to those at Klasies River the dating of the individual finds is not secure. The Border Cave material is more gracile than that from Klasies River and given selection for gracility in the Late Pleistocene, they may be more recent.

The taphonomy of the fossils from Klasies River (White 1987) suggest dietary and/or ritual cannibalism rather than conscious burial. Apart from a possible burial associated with the Middle Stone Age at Border Cave (BC3), the oldest dated burials from South Africa date to the end-Pleistocene and most are from the Holocene. The associations of these hyper-gracile Holocene individuals, products of the selective regime, are with the Later Stone Age.

The Middle-Late Pleistocene human remains from southern Africa complement finds from elsewhere in Africa. Human fossils with modern morphology (Bräuer 1992) and of similar age to those from Klasies River are known from eastern and northern Africa, including the Levant. The human fossil record is limited because of generally unfavourable conditions for the preservation of bone in Africa and because many areas have been investigated at only a reconnaissance level. The available sample from the continent, however, documents an evolutionary continuity that is more persuasive than for any other part of the world. The debate between the proponents of different models for the emergence of modern humans is about possible discontinuities in the fossil record outside of Africa.

Late pleistocene populations

The population histories of the different regions of Africa are of interest because population size is related to genetic diversity and the contemporary population of Sub-Saharan Africa is the most diverse. The higher diversity may indicate this Middle-Late Pleistocene population was larger. A recent study (Relethford & Harpending 1995) suggests that the African population was some three times larger than any other continental population. If population size differences are taken into account, then higher genetic diversity does not necessarily imply that the African population was ancestral to modern humans, although it is still possible that a significant portion of the ancestral modern population lived in Africa.

Progress in the dating and the correlation of rock shelter sequences in southern Africa provides relevant data on population history. There was widespread occupation of sites in the region in the first half of the Late Pleistocene (130 000-60 000 years). Non-occupation deposits, like rock falls, discontinuities or brief occupation events are recorded in sequences in the interval that follows. The evidence for site occupation increases after 16 000 years but the most significant increase is in the last 5000 years. A complex population history is indicated with the initial Late Pleistocene expansion followed by an extended period of contraction and a very recent re-expansion. A contraction in the early Holocene, when much of the western interior of South Africa was effectively depopulated, is well documented and correlates with a period of aridity (Deacon & Lancaster 1988). Environmental forcing was also the likely cause of Late Pleistocene population fluxes.

A prediction, based on nucleotide divergence theory (Sherry et al. 1994), is that the major episodes of global population expansion occurred in the latter half of the Late Pleistocene, 65 000-30 000 years ago. Collapsing the mtDNA mismatch distribution data into three continental areas, one of which is Africa, makes the scale of geographical resolution coarser than considered here. The large confidence intervals in the estimates of timing of the expansions precludes any conclusion on whether expansion was earlier in Africa. The estimates for expansion are more recent than suggested here for southern Africa.

There is archaeological evidence from other regions for the initiation of the African population expansion in the beginning of the Late Pleistocene. The population history in eastern Africa may have been as complex as in southern Africa and not necessarily identical in all trends. There are Middle Stone Age sequences in Central Tanzania (Mehlman 1991) dating to 130 000 years that support an early expansion. These sequences appear to show continuity in occupation after 60 000 years and any subsequent population contraction may have been less severe than in southern Africa. Equatorial Africa, possibly a marginal habitat even under drier Pleistocene conditions, currently provides little relevant data but data from the Sahara and North Africa (Clark 1992) are again consistent with expansion in the beginning Late Pleistocene followed by contraction in post-Aterian times. Plots of the numbers of archaeological horizons (Deacon & Lancaster 1988) radiocarbon dated to the last 20 000 years, which are proxy data for the population trends, show few records for the period 20 000-12 000 years BP in northern and eastern Africa. In the following period there is a marked increase of dated sites and this increase coincides with pluvial conditions associated with strengthening of monsoon circulation. This recent re-expansion preceded that in southern Africa by some 7 000 years because environmental changes in eastern and southern Africa are out of phase. Habitats throughout the continent appear to have favoured population expansion during the first half of the Late Pleistocene and contraction in the mid-Late Pleistocene was pronounced in the northern and southern regions. All regions show a very recent expansion.

The surprising result of studies of human variability is that humans are genetically very alike. This is evident from, for example, the sequencing of the complete mtDNA genome of humans and great apes (Horai et al. 1995). The root of the chimpanzee tree is four times deeper than that of humans, showing apes preserve much greater genetic diversity. The implication is that the effective human populations remained low, fewer than 10 000, from the initial dispersal of humankind, or there was a population contraction, a bottleneck, prior to the Late Pleistocene expansion. The difficulty that archaeologists have experienced in piecing together the record between older Middle Pleistocene Acheulian and Late Pleistocene Middle Stone Age sites supports the possibility of a period in the later Middle Pleistocene when populations densities were close to or below the limits of archaeological resolution. Ill defined and poorly dated archaeological "cultures" like the Sangoan in tropical Africa and the Fauresmith in South Africa and possibly Kenya, are presumed to represent this "intermediate" stage. A comprehensive dating programme is needed to clarify the significance of these entities and the nature of events prior to the Late Pleistocene expansion. An early Late Pleistocene expansion of African populations would be consistent with Africa being a donor continent and with movements of people out of the continent. Eurasia may have been a population sink, registering no net gain in populations in this period. In those continental regions, however, expansion was registered in the latter Late Pleistocene. Environmental forcing was the major factor driving changes in distribution and density of populations in Africa and environment constraints may explain the retarded expansion in Eurasia.


Why the Late Pleistocene population expansion happened when it did is an origins-type question like the origins of farming communities or the origins of complex societies. The timing was dictated by environmental factors but as there have been repetitive cyclic environmental changes in the Pleistocene the question remains why this major expansion in global populations happened as recently as the Late Pleistocene. Post hoc accommodative explanations could include the evolution of a new human species conferring adaptive advantages, the emergence of a new means of production or again the emergence of new levels of social complexity and cognitive abilities. The evidence from southern Africa is discussed here.

There are Acheulian as well as Middle and Later Stone Age sites in the same landscape in southern Africa. The Acheulian sites are poorly dated. All would date prior to the Late Pleistocene population expansion and it is possible that the widespread mid-Acheulian occurrences, like those associated with the Vaal River terraces, are of early Middle Pleistocene age and long pre-date the expansion. The Middle Pleistocene Acheulian sites are located in valley bottom situations, around ephemeral lakes, springs and seepage points. The pervasive association with wetland habitats suggests those populations occupied a very narrow ecological niche. This is in contrast to the distribution of Middle and Later Stone Age sites which are found in all landscape positions and not infrequently at the same locations. The narrow Acheulian niche is consistent with wide dispersal of a small population. Population expansion on the scale of the Late Pleistocene would not have been possible without the use of a broad spectrum of resources and social mechanisms permitting a dispersed rather than restricted landscape distribution.

The Late Pleistocene population expansion in Africa is associated with people who were morphologically modern. There have been suggestions (Klein 1992) that modern anatomy does not imply modern behaviour but these arguments are eurocentric in defining modern behaviour in terms of the Upper Palaeolithic. There are conceptual problems in defining what is and what is not modern behaviour. However, the traces of activities that reflect in site formation (Deacon 1992), like the positioning of hearths and the disposal of food waste, are so similar at Middle Stone Age and younger sites that differences in behaviour that may relate to cognition can be discounted. The mid-Acheulian populations can be considered as the outgroup showing non-modern behaviour in contrast to the modern behaviour associated with populations from at least the beginning of the Late Pleistocene.

Concluding discussion

The debates on the emergence of modern humans have stimulated interest in the archaeology of the Late Pleistocene, encouraged reassessment of the fossil record and the development of dating techniques applicable to time ranges older than those datable by radiocarbon. Advances in techniques in molecular biology and genetic theory are providing a wholly new and independent perspective on problems associated with the emergence. While the debates appear polarized between two competing models, the recent-African-origin and the multiregional models, these models are more heuristic devices for discussion of the evidence.

This paper argues for a beginning Late Pleistocene expansion, 130 000 years ago, of populations in southern Africa and the African continent as a whole. It shows there were dynamic changes in population size during the Late Pleistocene in Africa and links the cause of these changes to environmental forcing. This population expansion in Africa was of people who were morphologically modern and behaviourally modern. The expansion of populations in Africa preceded that in Eurasia by 50 000 years or more. The history of populations in Eurasia rather than in Africa is the prime concern of the competing recent-Africa-origin and the mutiregional models. However, documenting the emergence of modern humans in Africa is an important challenge.


Beaumont, P. B. 1980. On the age of the Border Cave hominids 1-5. Palaeontologica Africana 23:21-33.

Bräuer, G. 1992. Africa's place in the evolution of Homo sapiens. In: Continuity or replacement: controversies in Homo sapiens evolution (eds. G. Bräuer & F. H. Smith), pp. 83-98. Rotterdam: A. A. Balkema

Bräuer, G., Deacon, H. J. & Zipfel, F. 1992. Comment on the new maxillary finds from Klasies River, South Africa. Journal of Human Evolution 23:419-422.

Clark, J. D. 1992. African and Asian perspectives on the origins of modern humans. Transactions of the Royal Society London B 337:201-215.

Deacon, H. J. 1992. Southern Africa and modern human origins. Philosophicaal Transactions of the Royal Society London B 337:177-183.

Deacon, J. & Lancaster, N. 1988. Late Quaternary environments of southern Africa. Oxford: Clarendon Press.

Goodwin, A. J. H. 1958. Formative years of our prehistoric terminology. South African Archaeological Bulletin 14:25-33.

Horai, S., Hayasaka, K., Kondo, R. & Tsugane, N. 1995. Recent African origin of modern humans revealed by complete sequences of hominoid mitochondrial DNAs. Proceedings of the National Academy of Sciences (USA) 1992:532-536.

Klein, R. G. 1992. The archaeology of human origins. Evolutionary Anthropology 1:5-14.

Mehlman, M. J. 1991. Context for the emergence of modern man in eastern Africa: some new Tanzanian evidence. In: Cultural beginnings: approaches to understanding early hominid lifeways in the African savanna (ed: J. D. Clark) pp. 177-196. Mainz: Römisch-Germanisches Zentral-museum.

Opperman, H. & Heydenrych, B. A 22 000 year-old Middle Stone Age camp site with plant food remains from the North-Eastern Cape. South African Archaeological Bulletin 45:93-99.

Relethford, J. H. & Harpending, H. C. 1995. Ancient differences in population size can mimic a recent African origin of modern humans. Current Anthropology 36:667-674.

Rightmire, G. P. & Deacon, H. J. 1991. Comparative studies of Late Pleistocene human remains from Klasies River Mouth, South Africa. Journal of Human Evolution 20:131-156.

Sherry, S. T., Rogers, A. R., Harpending, H. C., Soodyall, H., Jenkins, T, & Ttoneking, M. 1994. Mismatch distributions of mtDna reveal recent human population expansions. Human Biology 66:761-775.

Singer, R. & Wymer, J. 1982. The Middle Stone Age at Klasies River Mouth in South Africa. Chicago: Chicago University Press

White, T.D. 1987. Cannibalism at Klasies? Sagittarius 2:6-9.

Back to home page
Stellenbosch University home page