But if Cobbing succeeds in establishing the outlines of an alternative grand hypothesis, his attempt to put empirical flesh on its bones is in many respects badly flawed. Too often he seeks to make his case through a process of assertion rather than of argumentation; too often he makes sweeping judgements on the flimsiest of evidence; too often he overbalances in arriving at conclusions. For instance, in drawing attention to the development of the Delagoa Bay slave trade, he has made an original and important contribution to the debate on African 'state-formation' in the region. But in my opinion he puts too much weight on the trade as a factor for political change in the Delagoa Bay hinterland in the 1810s and early 1820s. Concomitantly, he does not give enough attention to the expansion of the ivory trade which had begun in the 1760s and 1770s, and which has long been seen by a number of historians as one of the causes of the intensified conflicts which were beginning in the region several decades before the 1810s. The size of the slave trade in the 1810s is not the issue here: the point is that by presenting the impact of this trade as suddenly originating widespread and dramatic political change, rather than as feeding into ongoing processes of change, he distorts its probable historical role. Africans emerge as victims of the European presence rather than as actors capable of shaping their own responses to it.
Cobbing injects another important element into the debate by reviving and carrying further the arguments put forward by Martin Legassick more than twenty years ago about the role of Griqua, Kora and other raiders on the Cape northern and north-eastern frontiers in destabilising the highveld region in the early nineteenth century.3 He also forces us to look in a new light at the historical significance of the European settler demand for African labour in the eastern Cape. But in arguing that missionaries like Robert Moffat and officials like Sir Richard Bourke were involved in conspiracies to raid slaves he is straining the evidence to the point where he risks undermining his whole thesis. The expeditions which led to the massacres at Dithakong ( 1823) and Mbholompo ( 1828) which feature in the title of his article may very well have provided the opportunity for some of the victorious parties to seize slaves from among the defeated. It is perfectly likely that this was all along the aim, if an unstated one, of some of the participating groups, Griqua in the one case, colonial settlers in the other. But to say this is very different from implying that the expeditions were from the start deliberately planned as slave-raids by government officials and missionaries. Conspiracy theories of this kind are usually bad history in that they are prone to lump together into a single, undifferentiated category of actors groups with widely differing aims and interests, and to see the outcome of events as preordained in their causes.
The same point can be made with regard to Cobbing's treatment of the way in which white writers since the 1820s and 1830s have presented the upheavals of the period as the product of Zulu rather than European expansionism. I would agree that, as used by historians, the concept of the 'Zulu wars', or mfecane, has over the years often functioned as an 'alibi' by serving to deflect attention away from the role played by Europeans in stimulating the conflicts of the early nineteenth century, and to justify European land-grabbing later in the century. But this is not the same thing as implying, as Cobbing does, that white historians deliberately invented the concept as an alibi in the mid-nineteenth century, and have continued consciously to use it as one ever since. As Carolyn Hamilton argues in detail in her essay in this volume, this kind of reductionist argument distorts the processes in which the history of the 'Zulu wars' has been constructed.
The Natal Region: Reconsiderations
Soon after the publication of Cobbing's seminal if lopsided article, my own research into the precolonial history of the Natal region reached the point where I was able to enter the emerging debate on the mfecane with an article which in certain respects lent support to his critique.4
In reviewing the secondary literature on the subject, I argued that the entrenched view that Zulu armies had 'devastated' and 'depopulated' Natal south of the Thukela River in the 1820s was based not so much on empirical evidence as on the uncritical repetition by generations of historians of a stereotype whose origins date back to the writings of European traders and settlers in the 1820s and 1830s. Hamilton has rightly criticised this article for taking little account of the role played over time in the development of this view by African intellectuals, both literate and non-literate.5 But this criticism does not alter my conclusion that there is little evidence to show that the Zulu played the exclusively destructive role usually ascribed to them in what I have called the 'prototype of all other regional mfecanes'.6
I followed up with an article in which I examined the presentation of the history of the region south of the Thukela in what has long been the standard source on the history of the Natal–Zulu kingdom region before the advent of Europeans, A. T. Bryant's well-known Olden Times in Zululand and Natal.7 Most later writers have usually assumed that much of this work is based on oral traditions collected by the author. My investigation showed that his account of the upheavals of the early nineteenth century south of the Thukela was in fact based on an uncritical reading of the works of earlier writers, and could no longer be regarded as reliable.8 The idea, which Bryant developed in some detail, that in the late 1810s and early 1820s Natal south of the Thukela had been devastated by four successive waves of refugees fleeing from the Zulu, and then by a series of Zulu invasions, was shown to have very little foundation in the sources which he had used.
In a third contribution, 'Political Transformations in the Thukela–Mzimkhulu Region of Natal in the Late Eighteenth and Early Nineteenth Centuries' (pp. 163–81 in this volume), I present in outline a new interpretation of the existing evidence on the political history of Natal south of the Thukela during this period. My main findings can be summarised under four heads: first, that discernible political changes in the region began half a century before the emergence of the Zulu kingdom; second, that though the region as a whole experienced considerable political upheaval in the 1820s, it was not subject to wholesale devastation or depopulation by the Zulu or anyone else; third, that the Zulu kingdom under Shaka did not have the military or the political capacity to establish direct rule over the region south of the Thukela; fourth, that there were other important agents besides the Zulu in bringing about the complex series of changes that took place in the region in the 1820s. The concept of the mfecane finds no empirical support in the evidence on the history of the territories south of the Thukela: in effect it dissolves away.
Cobbing's 1988 article has provided a departure point for three further contributions, in the form of the chapters in this book written respectively by Elizabeth Eldredge, Carolyn Hamilton and Jeff Peires, which bear directly on the concerns of the present essay.9 For her part, Eldredge accepts the essence of Cobbing's argument that the Zulu were not responsible for most of the conflicts of the 1820s and 1830s. She concurs that the slave trade across the Cape frontier was an important cause of conflict on the highveld. But she rejects Cobbing's arguments about the involvement of missionaries in the slave trade in the interior. She regards as a retrograde step his depiction of African societies as simply 'reacting' to the European presence rather than as having aims and objects of their own: as she says, twenty years of Africanist historical scholarship in southern Africa is thereby disregarded. And, in a reappraisal of the evidence on slaving at Delagoa Bay, she argues that the trade there was of no significance before the early 1820s: Cobbing's thesis that intense conflict was set in motion by the expansion of the slave trade in the later 1810s is therefore without foundation.
As indicated above, my own opinion is that Cobbing puts too much explanatory weight on the slave trade as a factor in the conflicts of the pre-1820 period, and there seems little doubt that he has