Subtle and continuous adaptation: Understanding Middle Stone Age technological adaptations through Marine Isotope Stages 6 & 5 in the Doring River Catchment, South Africa

This thesis investigates Middle Stone Age (Early MSA; ^~315-80 thousand years ago [ka]) lithic technology at three rockshelter sites in the Doring River Catchment, Western Cape, South Africa, dated to Marine Isotope Stage (MIS) 6-5. I define this study period as ‘Early MSA’ in the broadest sense – as characterised across southern Africa by the general use of local rocks, a spatio-temporal mosaic of technological adaptations, and the absence of ‘precocious’ technological and behaviour traits that emerge from MIS 4 onwards. In lieu of such novelties, the study of Early MSA has languished in comparison with the study of later periods. There is growing evidence for technological innovations in the Early MSA, namely in late MIS 5, though such instances are often short-lived, temporally discontinuous, and spatially patchy. The Early MSA’s lithic technological adaptations, often characterised as ‘generic’ and stagnant owing to their protean nature, increasingly demonstrate a heterogeneity that existing culture historic frameworks do not easily capture.

The prevailing culture historic ‘technocomplex’ scheme used in southern Africa, initially developed by Thomas Volman (1981, 1984) and subsequently updated by Lombard and Colleagues (2022, 2012), features limited explicit engagement with the drivers underlying technological change. Furthermore, specific technological changes are rarely linked to underlying processes. Consequently, while the approach deals well with clear package of change such as those that occur in the Still Bay and Howiesons Poort, it struggles to accommodate the subtle variability that characterises the Early MSA.

The technocomplex scheme has been delineated into three successive technocomplexes based on changes in a small number of artefact types and attributes, namely consistent use of locally derived raw material, retouched flake types of notches, denticulates, and scrapers, and changes in flake reduction methods (e.g., blades, Levallois, laminar, and so on). Scrapers and denticulates in the European Middle Palaeolithic have been shown to be poor fossile directeur, or type fossils, as their reduction acts on a continuum with their discard state reflecting continued retouch through reuse. Developed by Harold Dibble (1987, 1995) and colleagues (e.g., Clarkson, 2005, Hiscock and Attenbrow, 2003, Hiscock and Clarkson, 2007, Holdaway, et al., 1996, Lin and Marreiros, 2021), the reduction model for retouched flakes has been tested in various contexts but not for the Early MSA. The first proposition that this thesis tests is whether the retouched flakes of the Early MSA fit with strict typological characterisations, reduction mediated by retouch form, or a pattern of expedient, on-going retouch. The three rockshelters provided ^~9,000 lithic artefacts across seven stratigraphic units that were analysed through an attribute-based methodology. The morphology of scrapers, notches, and denticulates is shown to be the result of on-going, flexible, and often ad hoc patterns of retouch that were more likely responsive to imminent needs than to mental templates or culturally defined production systems.

The classification tool of typology for the Early MSA has been shown to be ineffective, while there is increasing evidence that the technocomplexes are time transgressive and spatially varied. This means that a critical assessment of technocomplexes for the Early MSA is required. The rockshelter Klipfonteinrand 1 formed part of Volman’s technocomplex scheme, contributing to the definition of the MSA 2b (^~100-80 ka) unit that was applied across southern Africa. Klipfonteinrand 1 has largely fallen out of discussions of Early MSA technology owing to concerns over mixing. Renewed excavations at the site from 2012-2013 provide a new and better resolved assemblage with which to work. The renewed excavations showed the site to be less mixed than previously thought and older, with two Early MSA stratigraphic units - GGLBS dated to ^~85 ka and PBS at ^~156 ka. When tested against the defining characteristics of MSA 2b, Klipfonteinrand 1 was shown to have a low affinity to the technocomplex scheme it helped develop. While each unit contained many relevant features of their respective technocomplexes, they did not contain all of them, nor were the features exclusive to their technocomplexes. A review of coeval assemblages within the Winter Rainfall Zone and the adjacent Year-Round Rainfall Zone found greater inter-site than intra-site differences, with few sites sharing all or many technological features.

From these results, the thesis turns towards models of technological organisation to illuminate the drivers of variability through the Early MSA. Technological organisation models, generally discussed in terms of mobility, procurement, provisioning, and reduction strategies, have proved to be practical tools for understanding behavioural choices in various settings. Mobility is a prime adaptive measure for hunter gatherer populations in response to changes in environmental and social conditions. Raw material procurement can provide information on scales of mobility and provisioning insights in how groups ensured they had tool making stone available. While these strategies have been used to explain the relatively short-lived Still Bay and Howiesons Poort, the preceding Early MSA has only seen rare use. Persistent features of the Early MSA such as the use of locally derived raw material, subtle technological changes, and expedient patterns of artefact production and reduction have previously limited their use. Technological organisation is the focus of the latter half of the thesis.

Technological organisation models will be tested at Mertenhof Rockshelter, across three distinct stratigraphic units dated between ^~112-98 ka. The site exhibits some of the highest proportions of non-local raw materials in the Early MSA, offering an opportunity to test models of mobility, procurement, and provisioning of raw materials. The sequence displays a series of subtle technological shifts over a relatively brief period of ^~14,000 years. These subtle changes represent a series of tactical responses within technological organisation strategies at a similar timescale of the Still Bay and Howiesons Poort but without the associated technological turnover.

Following the results from Mertenhof and the importance that mobility played in its technological signatures, the final part of the thesis looks at what may be driving these changes in mobility and site occupation across the Doring River Catchment. As the catchment is situated in a semi-arid region at the ecotone of the Fynbos and Succulent Karoo biomes and receives most of its water during the winter months with limited retention during summer, it presents an opportunity to test the influence of water availability on mobility, occupational intensity, and technological organisation though MIS 6-5. Using historical surface water data from Water Observations from Space and known raw material sources across the catchment, the final chapter finds water availability across the catchment and in the immediate vicinity of the rockshelters influenced their occupation and occupational intensity, and mobility. Technological organisation responded to changes in water availability through quicker movements during periods, greater sedentarism at others, and changes in tactical responses in core and retouched flake reduction.

In conclusion, this thesis critically evaluates the existing frameworks of the Early MSA and tests models of technological organisation, finding that models of technological organisation are effective tools in deciphering the heterogeneity in lithic technology. Rather than a stagnant and protean period for Homo sapiens, the results from this thesis indicate the hunter gatherers in MIS 6-5 applied a flexible and adaptive technological repertoire and were adaptively sensitive to the changing climatic and environmental conditions present. Ultimately, there is more to the Early MSA than meets the eye, and its lithic technology should be foregrounded in understanding our species' evolutionary history.