Includes bibliographical references. / The research reported in this thesis involves the measurement of stable carbon isotope ratios in human bone collagen as a means of reconstructing prehistoric diets. The sample population includes 67 skeletons of hunter-gatherers, pastoralists and agriculturalists from the Holocene of the Eastern Cape Province, South Africa. The aims of the thesis include the testing, through direct quantitative measurements, of the validity of archaeological conclusions about prehistoric human behaviour in the Eastern Cape. Secondly, the usefulness and applicability of the 13c tracer technique is demonstrated in what is arguable the most complex situation an archaeologist is likely to encounter. The natural environment included c3 and c4 plants, browsing and grazing ungulates, and a marine component - all subject to environmental change over the period under study - while the cultural environment included three different subsistence systems plus transition stages between them. A third, or subsidiary goal, was to test whether burial practices can be correlated with subsistence economies in this situation - that is, whether ritual and dietary behaviour formed part of some larger cultural whole such as "pastoralists" - in order to be able to assign individuals to socio-economic groups on the basis of burial pattern. The results of the laboratory analysis realize these goals with varying degrees of success and with important consequences for the archaeologist.

Includes bibliographical references. / When Europeans first began to move into the central portion of the interior of southern Africa in the late eighteenth century, they encountered Bushmen hunter-gatherers who were already occupying the land. The impact which these Europeans had on one geographical region, the Seacow River valley in the north-east Karoo region, and on the hunter-gatherers living there, form the basic theme of this thesis. The aim is to place the Seacow River valley and the events that transpired in that region between 1770 and 1900 in a documented context, in order to use this context to make sense of the archaeological data pertaining to this Post-Contact period. The expansion of European farmers into the Seacow River valley was gradual, and through the eighteenth and nineteenth century this gradual expansion affected not only the large herds of game grazing on the Karroid vegetation, but the vegetation itself. A wide spectrum of documentary sources shed light on the nature and timing of the extermination of valley game animals and provided clues to the effect which the European presence had on the environment. Bushmen hunter-gatherers living in the region responded to the Europeans in various ways. Although some Bushmen may have moved ahead of the colonial frontier, or were destroyed by it, many Bushmen survived by adopting different roles in colonial society. The degree in which valley Bushmen retained their identity, subsistence patterns and material culture was largely dependent on their relation to the colonial presence in the form of farms, mission stations and towns. Some Bushmen lived independently from these centres, while others were related to these centres in different ways. The documentation of this differing response to the colonial presence not only provides a context for the accumulation of Post-Contact archaeological deposits, but allows some light to be shed on the context of individual archaeological sites.

Much of our current understanding of prehistoric human behavioural patterns during the Stone Age, is derived particularly from a robust set of chronological and technological sequences from caves and rock shelters, with some focus on open-air sites. The information gained from shelters cannot be ignored or downplayed, however, they offer a spatially and temporally limited view of prehistoric lifeways. The aim of this thesis is to provide an understanding of landscape use during the Stone Ages along the Bos River in the Tankwa Karoo, Northern Cape. Surveys were carried out around the Bos River, with the intention of mapping out and analysing all the surface stone artefacts. Analysing at the scale of the individual artefact, particularly temporally iconic artefacts, permits the landscape, although geologically and ecologically variable, to be viewed as a continuous space. The benefit of this approach allows for all artefacts across all types of settings to be analysed, providing a spatially subjective distribution of artefacts across the landscape. The evidence described in this thesis demonstrates an episodic occupation of the Tankwa Karoo during periods of increased resources, particularly the availability of food. The Bos River is a low-energy river that receives little rain and does not facilitate the formation of large rounded cobbles and boulders, explaining the lack of an occupation during the Earlier Stone Age (ESA), whereas an expedient organisation of locally sourced raw materials for stone tools characterise the Middle Stone Age (MSA) and Later Stone Age (LSA) periods in the Tankwa Karoo.

Bibliography: pages 138-149. / This dissertation is the result of a field-project conducted in the lower Thukela Basin over a twenty-month period in 1984 and 1985. The dissertation sets out to document a regional survey of Early Iron Age sites in a part of the Lower Thukela Basin and report on the excavations and analysis of material from two sites, Mamba and Wosi. On the basis of the archaeological evidence iron smelting practices are discussed and the nature of first millennium mixed-farming, valley bottom settlements reported on. The ceramic finds are described and compared with other known samples from this period and are chronologically placed in the light of recent classificatory suggestions. Some arguments are submitted as to the changing nature of the archaeological record through time. In response to appeals made at the 1985 conference of the Southern African Association of Archaeologists (Hall 1985a; Lewis-Williams 1985) I have, subsequent to the conducting of the field-work, attempted to inform myself more widely of prevailing applications of social theory in the interpretation of the southern African pre-colonial historical record. The latter part of this dissertation is my attempt at such an application and further, an attempt at testing specific theories and models against the field-data collected. Unless otherwise acknowledged, all of the field-work reported on is my own. I remain indebted however to those who have gone before and provided the necessary frameworks on which much of my interpretation is based.

This dissertation examines the identity of the Xhosa communities that settled in the frontier zone of the Northern Cape during the first half of the 19th century. It does this through the archaeology, and ethnographic and historical accounts. The concept of a baseline Nguni identity in the Eastern Cape is examined with an emphasis on settlement, mobility and cultural interaction. The historical background and a brief history of the of the Xhosa in the Northern Cape will be detailed, focusing on the Pramberg community. The archaeology of three Xhosa sites in the Pramberg will be described and analysed, and then contrasted and compared with the ethnographic and historical evidence. The result of this comparison is a discussion of the identity change and continuity of the Pramberg Xhosa in the context of the cultural milieu of the frontier and the appropriation of land by the expanding Cape colony.

Includes bibliographical references. / Evidence for iron smelting, agriculture, elaborate pottery styles and increased sedentism appears abruptly in areas previously inhabited by hunter-gatherers and herders during the Early Iron Age (EIA) of southern Africa from around 250CE. Ceramic evidence connects these (cultural) populations to the second millennium Iron Age sites in eastern Botswana. This material culture differs from second millennium Late Iron Age (LIA) sites in South Africa which are attributed to migrations from east Africa and are connected, via the material culture, to modern Sotho-Tswana and Nguni speakers. Although the material culture of this period is well-studied, there is a gap in correlating Iron Age biological identity with the established cultural identity. Here I present an analysis of metric and nonmetric dental and cranial variation to better understand biological relationships among these samples. Specimens from the LIA, EIA and Eastern Botswana are compared with each other, and to specimens from Iron Age Zambian sites, modern Bantu-speakers and a historic Ndebele site from the midnineteenth century. This research indicates few differences between the EIA and LIA groups, although surprisingly a sample from eastern Botswana is more similar to the LIA group than the EIA group. The Iron Age samples are significantly different from the modern sample, while the historic sample lies intermediate to the Iron Age and modern samples, indicating that Iron Age peoples had a pattern of dental and cranio-mandibular variation that differs from what is seen in modern (admixed?) descendants. This research has important implications for our understanding of the sub-Saharan African dental complex, showing population differences within this complex (between Khoesan and Iron Age peoples) as well as variation over time (between Iron Age peoples and modern Bantuspeakers). This indicates that, while farmers within the Iron Age of southernmost Africa are generally homogenous, there are important differences between populations in sub-Saharan Africa that reflect complex and differing histories.

In the last centuries before incorporation into the Cape Colony, the Riet and Orange River areas of the Northern Cape, South Africa were inhabited by communities of hunter-gatherers and herders whose life ways are little understood. These people were primarily of Khoesan descent, but their large stone-built stock pens attest to the presence of substantial herds of livestock, very likely for trade. This region was too dry for agriculture, although we know that there were links with Tswana-speaking agricultural communities to the north, because of the presence of characteristic styles of copper artefacts in Riet River graves. This was a frontier region at a turbulent time in South African history, so one of the questions about these societies is the extent to which they were homogeneous or heterogeneous - were many outsiders incorporated into these communities? What was the relative importance of herding compared with hunting in the local economy? Did connections with farming communities extend to the trading of cereal foods?

Bibliography: pages 85-96. / Heavy stable strontium isotope ratios (⁸⁷Sr/⁸⁶Sr) may be used to characterize the habitat choices of early hominids. The basis for this approach is that different geological substrata may have a characteristic range of ⁸⁷Sr/⁸⁶Sr ratios in the available soil strontium pool. This range may be represented in the plants growing on these various geologies and thus into the bones of animals feeding in these areas. Through an examination of strontium isotope ratios for isotope and soils from 10 different geologies in the Sterkfontein Valley and immediate vicinity, it was possible to obtain some tentative indication of the possible range in ⁸⁷Sr/⁸⁶Sr ratios in an area most likely to cover favoured foraging habitats of early hominids, such as Australopithecus robustus. On the basis of isotopic data from extensive dolomite sampling, it is clear that there is a wide range of variation in strontium isotope ratios for dolomite substrata around Swartkrans. This may be as a result of topography, water action and atmospheric pollution. However, in spite of such variation, it is possible to distinguish isotopically different habitats, such as riparian from slopes. Such variation will complicate the interpretation of early hominid home ranges, but does provide alternative avenues for behavioural interpretations and also emphasises the necessity for the collection of background isotopic data.

This thesis examines the use of stable carbon isotopes as a means of reconstructing the palaeoenvironment of the Caledon River Valley of the eastern Free State, South Africa, and western Lesotho. In doing so, this work draws upon previous studies that have shown that the distinct distribution and δ¹³C values of C₃ and C₄ grasses are influenced by seasonality of rainfall and growth season temperatures. In general, C₃ grasses dominate in areas where conditions are cool/moist during the growth season, while C₄ grasses characterize those that are warm/arid. The isotopic composition of the grasses of an area, and thus climatic and environmental data, is passed along the trophic levels, through dietary intake by grazers, and decomposition into soil sediments. By measuring the ¹³C/¹²C ratios of carbon extracted from the calcified tissues of grazers and soil organic matter recovered from within an archaeological context, a palaeoenvironmental sequence has been reconstructed for the study area for the last 13 500 years. Results have shown that although C₄ grasses have dominated, the presence of C₃ grasses, at various times during this period, suggest that growth season temperatures fluctuated temporally and spatially.

Includes abstract. / Includes bibliographical references (p. 177-199). / In this research I investigate Class 3 and Class 4 (Doornspruit) homesteads in the Bankeveld in order to establish their temporal and spatial distribution. Although the research area does fall within a cultural matrix historically dominated by Sotho/Tswana speakers, the cultural homogeneity portrayed in the oral texts is breaking down. Of direct relevance to this work is the argument put forth by Huffman that at least four movements of Nguni-speakers took place across the Vaal River from present-day KwaZulu-Natal during the Late Iron Age. These are broadly relevant to this research and specifically the most recent movement of Mzilikazi’s Khumalo. It is around this issue of Nguni identity and their homesteads that this research contributes.