I say 'it seems': a feature of Eldredge's otherwise measured and carefully constructed argument that I would query is her retention without explicit justification of a number of conventional 'mfecane' datings. Do we have enough evidence to say that conflicts in the Phongolo–Thukela region broke out precisely in 1817, or that the Hlubi attacked the Tlokwa precisely in 1822, or that the Hlubi and Ngwane fought a decisive battle precisely in 1825? Much of the chronology of the conflicts of the 1810s and 1820s still needs to be properly researched: until this is done, and the results published, writers on the period should treat many accepted datings with circumspection.
Eldredge's essay is not simply a reaction to Cobbing's article of 1988. She also has some important things to say about the connection between environmental and demographic forces and political conflict in the late eighteenth and early nineteenth centuries. Although it is not always clear from her argument whether she sees these forces as primary causes of 'the dramatic sociopolitical changes of the early nineteenth century' (p. 150), or simply as contributory to them, her points about the differences in the ways in which different social categories in African societies were affected by drought take current debates an important step forward. The useful references which she cites on the occurrence of drought in south-east Africa in the early nineteenth century remind us that a thoroughly researched study of this topic is still badly needed.
Like much of Eldredge's contribution, Hamilton's essay is also a 'reconsideration' of aspects of Cobbing's thesis. While expressly stating that she is not making a case for the mfecane, she takes Cobbing vigorously to task for the reductionism of his argument that the roots of modern mfecane theory are to be found in depictions of Shaka as a tyrant and monster produced by British traders at Port Natal in the 1820s. She puts forward two main sets of criticism. In the first place, she argues in detail that Cobbing has failed to look at the evidence which indicates that before 1829 the traders were not generally concerned to represent Shaka in pejorative terms. Different groups among them had different views of him which changed over time. In the second place, she argues that the unfavourable image of Shaka which was developed by European writers from the 1830s onward was in fact taken up from images originally produced by some of his own subjects. In the same way as Cobbing allows Africans no autonomous role as actors in the upheavals of the early nineteenth century, so he permits them no role in constructing the history of the period. The processes involved in the making of the historical images of Shaka, and of the concept of the mfecane itself, have been much more complex than Cobbing makes out.
Some elements of the empirical argument which Hamilton puts forward about the aims and activities of the British traders at Port Natal in the 1820s can be queried. It could be argued, for instance, that the traders were rather more concerned to engineer official British intervention in the Port Natal region than she allows for. And the nature of Shaka's policies towards the British in the Cape remains something of a puzzle. But her fine-grained article serves as a sharp reminder to historians who are seeking to move beyond the mfecane of the importance of grounding their arguments firmly in the available evidence. She also points to the need to develop a much more nuanced understanding of the relationship between history and ideology than is to be found in Cobbing's argument. This she directs particularly at students of the ways in which the concept of the mfecane has come to be so widely accepted.
The Career of Matiwane
Misuse of sources is also the main theme of the criticisms which Jeff Peires levels against Cobbing. He focuses specifically on Cobbing's handling of the career of Matiwane kaMasumpa, chief of the Ngwane and, according to mfecane historiography, one of the central figures in the history of the southern highveld in the 1820s. The conventional view is that the Ngwane were driven out of the Natal region by Shaka in the early 1820s, and spent several years raiding and fighting on the highveld. After being attacked either by the Zulu or the Ndebele, they moved away south and south-east into the country of the Thembu. In 1828, in the battle of Mbholompo, they were defeated and broken up by a combined army of Thembu, Gcaleka and Mpondo assisted by a British force from the Cape frontier. This latter had been sent out to repel a threatened Zulu invasion, and ended up attacking the Ngwane in the belief that they were Zulu.
Cobbing argues by contrast that the Ngwane were driven out of the Natal region either directly or indirectly by slave raiders from Delagoa Bay. On the highveld they fell victim to attacks by Griqua raiding for slaves and cattle, and were driven across the Drakensberg into Thembu territory. The main aggressors at Mbholompo were the British, whose prime aim was to secure labour for the eastern Cape settlers.
In a detailed response, Peires maintains that there is little in Cobbing's interpretation of the history of Matiwane and the Ngwane which can stand up to close scrutiny. He accuses Cobbing of concealing the existence of five African accounts which depict a quite different scenario. These sources indicate clearly, in Peires's view, that the Ngwane left the highveld not because they were attacked by the Griqua but because, after defeats at the hands of the Zulu, the Sotho and the Hlubi, and after a rebellion by his followers against his authority, Matiwane felt that the only means of reasserting his leadership was to make 'a brand-new start in a brand-new country' (p. 221). In his assessment of the events leading up to Mbholompo, Peires rejects Cobbing's argument that the British expedition was sent out specifically to acquire labourers. In his view the established explanation of the battle is correct. The capture of Ngwane refugees by the British was a by-product of the encounter, not its main purpose.
In sum, Peires concludes, Cobbing's account of Matiwane's history is based on misrepresentation of the evidence. He feels that the same is likely to be true of what he sees as the other two bases of Cobbing's hypothesis: that the Delagoa Bay slave trade was of central importance in stimulating intra-African conflict in the 1810s, and that the battle of Dithakong in 1823 was the culmination of a slaving expedition largely planned by officials and missionaries. The new paradigm which Cobbing is putting forward has no evidence to support it; the old paradigm of the mfecane should be retained until more convincing reasons for discarding it have been put forward.
I have indicated above my own opinion that Cobbing's portrayal of the British expedition which culminated in the battle at Mbholompo as an officially sanctioned labour-raid is intrinsically unlikely; on this score I agree with the thrust of Peires's argument. In fairness it should be pointed out that since the publication of his 1988 article Cobbing has modified his argument to take account of the differences of interest which existed in the 1820s between the colonial administration in Cape Town and the eastern Cape settlers and military.11 This is something that Peires needs to make more of: he is correct to insist that by this time British capitalism 'had long outgrown the smash-and-grab phase of primitive accumulation' (p. 227), but the same can hardly be said of embryonic settler capitalism on a remote African frontier. 'The capture of refugees by Colonel Somerset was a by-product of the battle rather than its cause,' Peires argues of Mbholompo. Yes, if Somerset is seen as representing the interests of the Cape administration, but a decided no if he is seen as representing the interests of the eastern Cape settlers. Peires himself states that it was Somerset's practice (as it was no doubt the practice of many other frontier officials) 'to entice Boer volunteers to military service with easy pickings in cattle and child servants' (p. 236). Mbholompo was not the product of an official labour-raid, but it certainly seems to have been the product of an officially sanctioned 'rescue' operation which, like many other official operations on colonial frontiers everywhere, was hijacked by ambitious soldiers and labour-raiding settlers.
As for the flight of the Ngwane from the highveld, we will not be in a position to understand this more clearly until we have a much more substantial account of highveld politics in the troubled times of the 1820s than either Cobbing or Peires has given us. I find Cobbing's