to recognize stone artifacts gave this educated elite knowledge of the past from a time before written history began, and, in some sense, through their knowledge they controlled that unwritten precolonial past.

The Zimbabwe-type structures apart, there are few monumental archaeological features in the South African landscape that draw attention and demand interpretation. Instead, numerous shell middens were encountered along the coast in positions far above the reach of the sea, and yet it took many years of debate to resolve whether these piles of shell were natural or owing to the activities of people. The shell midden controversy (Goodwin 1935) can be traced back to the exchange between two early travelers—the Englishman John Barrow and the Swedish naturalist Henry Lichtenstein—at the beginning of the nineteenth century. Neither seem to have been aware of Van Riebeeck’s journal, which was written at the time when the Dutch East India Company founded a refreshment station at the Cape in 1652, in which Van Riebeeck referred to people he called “Strandlopers” who were harvesting shellfish and other marine resources for food.

In the mid-nineteenth century, the same debate was taken up in the pages of the Cape Monthly Magazine, but by this time it was known that the heaps contained not only shellfish but also animal bones and even human burials and that the piles were food remains discarded at living sites. There are still numerous shell midden deposits along the coast, but they are under threat from the continuing development of resort towns.

In the last decades of the nineteenth century, there were attempts by Dunn, Gooch, Stow, and others to write archaeologies of South Africa. Stow’s book, The Native Races of South Africa (1905), published more than twenty years after it was written, is a good example of the state of the knowledge, and in its time, it was a magnificent piece of scholarship. The book takes the view that until recent times, the San (bushmen) had been stone toolmakers and that they represented a people long resident in southern Africa. Other peoples were seen as being later immigrants. At the time Stow was writing, there were folk memories of stone toolmaking groups of San (Kannemeyer 1890), and recent genetic research (Soodyall and Jenkins 1992) has amply confirmed the antiquity of the San genotype. Shrouded in mystery were the people who may have preceded “the bushmen.”

A feature of the book is a map that is a classic example of the migration model. The map is a mass of arrows showing waves of migrations from eastern Africa of San painter and sculptor “tribes,” who were responsible for the rock art; of Khoekhoe (or Khoikhoi, known as Hottentots to white settlers) herders; and of different siNtu- (Bantu-)speaking agricultural groups migrating into southern Africa. By assuming that the social, political, and linguistic divisions had gone unchanged, Stow projected the ethnic diversity that was understood in his own times into the past. That projection is too simplistic as individuals and communities interact and social groupings are and were in a state of continuous change or transformation. However, the importance of Stow’s work cannot be underrated. His was a first attempt to explain the peopling of southern Africa. Migration models, although not as extreme as Stow’s, have remained a popular form of explanation in South African archaeology, particularly in the study of the “spread” of early farming communities.

Stow is also remembered as a pioneer recorder of rock art. He was not the first to make field copies, as several of the early travelers had reported and made copies of rock paintings and engravings. Barrow (1801), for example, on a visit to the Graaff-Reinet area in the early nineteenth century, commented that the paintings were so fresh that they must have been made very recently. Rock art became known as “bushman paintings” and stone artifacts as “bushman implements.”

Without any reliable estimates of the time scale of precolonial history or knowledge of “prebushman” peoples, archaeology could not develop. Many of the finds of stone artifacts occurred on the surface, not deeply buried, which was taken to indicate that they were relatively recent and that most, if not all, artifacts were of the same age. Comparisons could be made with the deeply buried and undoubtedly ancient