Eighteenth and nineteenth century Karoo history has received relatively little archaeological attention. This is despite the fact that the area underwent a major economic transformation in the nineteenth century. The pastoralists of the eighteenth and early nineteenth century Roggeveld predominantly practiced small-scale subsistence farming. This changed rapidly with the introduction of merino wool sheep in the region in the mid 1800s. The region went from small-scale subsistence pastoralism to commercial and global scale Merino wool production, which fully integrated the Roggeveld into the economy of the colony. This dissertation looks at this sequence through the identification and mapping of the infrastructure of pastoralism using Google Earth and begins the process of linking this survey data with refined research on the ground by excavating at one of the sites located on Google Earth.

Includes abstract. / Includes bibliographical references.

Includes abstract. / Includes bibliographical references (leaves 117-131). / This thesis presents the digital amalgamation of eleven collections spanning six decades of archaeological research in the south-western Cape of South Africa.

This thesis focuses on Klipfonteinrand Rock Shelter in the eastern Cederberg, during the Late Pleistocene period (22.3 - 13.4 cal kyr BP), and specifically the materials analysed after the completion of excavations at the site. The study aims, firstly, to unpack changes in the composition of the archaeological assemblage of Klipfonteinrand during Marine Isotope Stage 2 (MIS 2), by analysing variation in the abundance and composition of lithics, ostrich eggshell, marine shell, charcoal, ochre, bone and roof spall from the site; and, secondly, to situate these changes in the context of changes in palaeoenvironments and patterns in the use of nearby and more distant archaeological sites. Results display a hornfels-dominant lithic assemblage, with a distinctive silcrete-rich Robberg-like industry during the period 22.3 - 16.3 cal kyr BP, categorised by micro-bladelets and bipolar cores. Ostrich eggshell fragments undergo variation in thickness in relation to variation in environmental and climatic changes, and also contain a handful of decorated fragments that mirror a sample from Boomplaas during a similar time period. Donax serra dominates the marine shell assemblage, which peaks in number between 16.6 and 15.9 cal kyr BP, at the time of rapid sea level rise across the subcontinent. Links are drawn between Klipfonteinrand and sites further afield such as Elands Bay Cave, based on the movement of hornfels from the interior to the coast and marine shell from the coast to the interior at the time of sea level rise. Roof spall is smallest and most abundant during colder periods, and charcoal and bone are most abundant then too. Red ochre is the most frequently occurring colour, with ochre having been ground most frequently in the younger levels between 14.4 - 13.4 cal kyr BP. These results are drawn together and Klipfonteinrand Rock Shelter is contextualised in a larger framework of MIS 2 archaeology and palaeoenvironments, on a local, regional and inter-regional scale. The various materials recovered from the excavation show specific diachronic patterns and suggestions are made about lithic technologies, craft and design, and complex landscape use of Late Pleistocene hunter-gatherers. The overarching purpose of this is to attempt to gain a better understanding of human behaviour during the environmentally unstable time period presented.

Includes bibliographical references (leaves 107-113). / This thesis investigates whether or not there is a material cultural expression of the economic (and by inference, social) difference seen in the stable isotope values of human skeletons from Robberg/Plettenberg Bay and Matjes River Rock Shelter between 4500 and 2000 B.P. After 2000 B.P. the introduction of pastoral subsistence in the area changed existing modes of production in such a way as to alter Ձ15N values. The two major excavated archaeological sites in this region are Nelson Bay Cave and Matjes River Rock Shelter, which lie only 15 kilometres apart. For this thesis, previously published descriptions of the artefact assemblages from these two sites were studied, and selected categories of artefacts were re-examined. More original work was necessary on the Matjes River collection, due to the poor quality of previous reports. The thesis focuses on the Wilton and pre-ceramic post-Wilton. In general, the same types of artefacts were found at both sites, but a number of types that were common in Layer C (i.e. in the Wilton) at Matjes River were not a feature of the Wilton levels at Nelson Bay Cave, although they became common in the post-Wilton. Backed scrapers were much more common in the Wilton levels of Matjes River than in any levels at Nelson Bay Cave, and chalcedony was more strongly favoured. Stone sinkers and perforated turtle carapace were present at Nelson Bay but were very rare at Matjes River. Several of the differences noted are not readily explained in terms of different functions or activities at the two sites. The contrast in the proportions of backed scrapers is best understood in terms of different traditions of artefact manufacture. Similarly, differences between the two sites in highly visible decorative items such as shell pendants are likely related to the negotiation of group or personal identity. These differences are consistent with a territorial separation between the groups that occupied the two sites, as postulated on the basis of the isotopic evidence.

Includes bibliographical references. / This dissertation quantifies and compares the mass of the ochre assemblages and the surface area of rock paintings from the same sites based on colour classification from the Later Stone Age assemblages of three rock shelters, De Hangen, Andriesgrond and Diepkloof in the Western Cape, South Africa. This work begins to bridge the gap that exists between the excavated archaeology and the examination of rock paintings which has been primarily focused on the iconography. The colours present in the rock art are quantified using a standardised colour system. A preference is shown for saturated 10R hues, based on the Munsell classification, in the fine line and handprinting rock art traditions, whilst the colonial era paintings may illustrate a more ad hoc approach, an interpretation bolstered by survey analysis. The colours of the ochres from the excavated assemblages do not match the colours seen in the rock paintings, the colours of the ochres being clustered more toward 2.5YR and 5YR Munsell hues. The concept of “ghost ochre” has been proposed in order to describe those ochres, in the colours shown to be preferentially used in the rock art, that have been utilised in their entirety and are accordingly absent from the archaeology. A proposed cause for this is that the primary ochre processing strategy employed by the inhabitants may have been that of pulverisation rather than grinding. Some experimental hearths were conducted to examine possible colour changes in buried ochres, and these samples were analysed by colorimetric means and by X-ray diffraction. Ultimately it has been demonstrated that the relationship between ochre and rock paintings from the same site is more complex than has been previously assumed. The findings in this thesis have implications for future research in rock art studies and analyses of LSA and MSA ochre assemblages.

The farm Blaauwbergsvalley, situated on the Cape west coast just outside Cape Town, has been identified as the place where a field hospital was set up for the casualties from the Battle of Blaauwberg in 1806. The significance of the site, however, extends beyond this specific event and includes a longer sequence of colonial and precolonial occupation. This is because a vlei provided a continuous supply of fresh but brackish water. Documentary evidence is discussed that draws attention to the growing importance of the western Cape Slagtersveld from 1652 as a region for livestock production to supply that VOC and its trading fleet. This contrasts with the Stellenbosch and Franschoek areas that developed more broadly around agriculture. More specifically, while it is likely that Blaauwbergsvalley, was a node in the 18th century development of this livestock landscape, it only formally appears in the documentary record from the late 18th century. The documents suggest that Blaauwbergsvalley, never fully functioned as a livestock farm but that it served as an outspan and a place that served the wider region and the flow of livestock and goods between the Table Valley settlement and the western Cape interior. The documented character of Blaauwbergsvalley is cautiously assessed against the archaeology of one area associated with the vlei. It is suggested that the archaeological evidence supports the transient, outspan function of Blaauwbergsvalley particularly in the period between 1800 and 1837 and that its material signature is not typical of other farms and werfs in the region. This needs to be assessed through future research.

Bibliography: pages 202-231. / Osteological and dental analyses have been widely used to outline a graded response to nutritional and physiological stress in human bone. It is argued that agriculturalists and transitional agro/pastoralists are more stressed than the hunter gatherers who preceded t hem. This is evinced by mortality profiles, mean age at death and the number and extent of stressors observed in the skeleton such as enamel hypoplasiae, porotic hyperostosis and Harris lines. Agriculturalists and agro/pastoralists are thought to be more prone to these stressors as they relied heavily on root crops and cereals for their nutrients. This exposed them to periods of episodic starvation and physical stress. Hunter gatherers in comparison are thought to have subsisted on a relatively healthy diet, offering more and better quality protein and so reducing the incidence of episodic and general stress. An alternative to this diet-dependent hypothesis is suggested by the analysis of forty-six skeletal remains from the nonagricultural, marine-dependent population of Oakhurst from the South coast of southern Africa. Porotic hyperostosis and enamel hypoplasiae are just as common among these marine-dependent people as among transitional agro/pastoralists. These findings imply that both individual development and population growth rates at Oakhurst were interrupted episodically and generally, and that these interruptions were substantially more common than in living and recently extinct hunter gatherers and pastoralists in southern Africa.

Bibliography: leaves 71-78.

Summary in English. / A unique opportunity to study historic burial practices in Cape Town arose in 1994 when construction activities at Cobern Street, Green Point revealed an eighteenth century burial ground. Subsequent salvage excavations unearthed approximately 65 burials and scattered skeletal material (both historical and precolonial) representing a total of 121 individuals. A variety of cultural material was found with the burials. The following is a summary of the excavation activities, and a detailed description of the burial patterns and grave goods unearthed at the site. An attempt is made to construct a cultural identity for the Cobern Street burials, and to determine what, if anything, burial practices have to contribute to our understanding of eighteenth century colonial society. The burial patterns were divided into four analytical categories, covering a spectrum ranging from the Later Stone Age to the end of the eighteenth century. The artefacts are divided into six groups; Later Stone Age artefacts, coffin hardware, burial items, clothing accessories, personal items (excluding clothing residues), and intrusive items. Burial items and artefacts are considered against the spatial layout of the site to determine that Cobern Street was used as used as an informal cemetery by lower class members of Colonial Cape society, primarily during the eighteenth century.