The difaqane wars around the area of the later Transvaal/Orange Free State are usually assumed to have begun in about 1822, after invasions from the east across the Drakensberg. But general restlessness and instability began much earlier - as Ngcongco indicates above - with a rising tide of violence from the later eighteenth century that cannot be clearly distinguished from the beginnings of the so-called difaqane in about 1822.

This essay looks at ways in which historians can use the findings of oral traditions and archaeology to reconstruct ‘pre-difaqane’ and ‘proto-difaqane’ periods of increasingly general crisis from the seventeenth up to the early nineteenth century. Understanding of these ‘pre-’ and ‘proto-’ periods should not only increase our understanding of the so-called difaqane/mfecane, but should also help to push back the frontiers of history - the continuous story of identifiable people - into the misty realms of prehistory.

Available oral traditions of the seventeenth and eighteenth centuries were mostly collected from Sotho/Tswana/Pedi-speaking elders, by European missionaries or officials and African intellectuals, before 1940. They tell of the growth of royal lineages, the spread of small states and small migrations of people. Mid-twentiethcentury historians, drawing on the perspectives of anthropology and nuclear physics, saw these traditions as evidence of structural processes of ‘fission’ and ‘fusion’ at work in precolonial societies. Later twentieth-century historians, drawing on the perspectives of archaeologists and environmentalists, have emphasised the importance of climatic changes in determining political-economic developments.

Lepalong is a Tswana name, recounted in oral records collected by Paul-Lambert Breutz, for a large natural cavern system in the vicinity of modern Carletonville and Potchefstroom. Lepalong was an extreme choice for a home, made, so the oral records say, by displaced Kwena people who had fled southwards away from Mzilikazi. This cavern was occupied between 1827 and 1836, and the record of that occupation is preserved in the substantial remains of what must have been a complete underground village. The value of this archaeological site is that it provides a record of social history and evidence for one kind of extreme strategic response to the strife of that period which, in the shifting matrix of what might be called the ‘difaqane’ historiography, provides a concrete expression of what life was like.

Julian Cobbing's critique of settler ‘alibi’ historiography and liberal interpretations of the difaqane locates causality for this strife and turmoil away from purely internal African agency and exposes the potential role of imperial Europe and its slaving agents. Central to his critique is a demythologising of the role of the Zulu kingdom as the prime catalyst and epicentre from which all disruption and turmoil ultimately emanated. It may be that Cobbing's analysis has swung the historiographic pendulum rather violently, and detailed regional investigations may find his general hypotheses out of step with the evidence. It is the aim of this essay to introduce archaeological material into the debate about the difaqane as an additional source from which a more detailed history for the period may be constructed. Cobbing's analyses extend the range of possibilities for assessing the archaeological evidence as it currently exists and as more comes to hand.

It seems that the present archaeological data contribute to a history of the period of upheaval labelled the difaqane in several ways. Firstly, and in the case of the western trans-Vaal, it helps untangle multiple causes by pointing out potential relationships between changes within Sotho/Tswana settlements during the late eighteenth century and an expanding northern Cape frontier as well as other tensions.

Julian Cobbing presents not one, but three arguments in his challenge to accepted versions of the mfecane:

The first argument has been widely accepted. The other two have provoked spirited controversy. Essays in Part One of this book take up some of the issues raised by the second and third of Cobbing's propositions.

Settler History and the Mfecane

Cobbing regards the mfecane as the masterpiece of settler history, and in so doing aligns himself with a long-standing tendency of South African scholarship to emphasise the political uses of history. Afrikaner nationalists have at various times advertised their intention to find a past that serves their cause. English-speaking historians such as the influential amateurs George McCall Theal and George Cory made clear their own intentions to write the sort of history that would ‘heal’ divisions between Boer and Briton. In various ways ‘liberal’ historians of the 1940s, 1950s and 1960s consciously wrote history ‘against apartheid’. Those scholars were in turn attacked by radicals for deliberately concealing the functional relationship between segregation and capitalism.

Against this background, Cobbing's assumption of a linkage between what he regards as mfecane theorists and settler interests amounts to a normal reflex action for a South African historian. That does not mean, however, that it deserves to be accepted without examination. It is a proposition that can be tested according to normal canons of historical proof.

Christopher Saunders sets out to do precisely that, and finds no evidence that a continuous chain of settler apologists stretches from Theal to John Omer-Cooper. Without at all denying that historians write in an ideological context, he points to marked contrasts in interpretations of the mfecane. He draws a sharp contrast between Theal and Cory, who traced all violence to Shaka, and William Macmillan, who strongly suspected that the slave trade to Mozambique had much to do with disruptions in what became the Zululand region, the highveld and even the eastern Cape frontier.

If Julian Cobbing is right in his contention that the root causes of the mfecane lie not in the Zulu kingdom but in disruptive forces emanating from Mozambique and the Cape, then rethinking the mfecane means rethinking the Great Trek. One of the oddest circumstances in historical writing about South Africa is that those contemporaneous phenomena, each of which has been called ‘the central event in South African history’, have been treated as isolated occurrences. According to the dictates of a peculiar historiographical apartheid, the only recognised linkage is the supposition that the mfecane cleared the highveld of people at the very moment the Voortrekkers decided to go and live there. This essay offers some revisionist propositions about the 1830s developed from a bird's-eye view of the historiographical landscape. The word ‘revision’ is used in its original sense. No new archival research findings which change our picture of the past are reported here. Instead, some familiar and obvious sources are re-examined with a view to changing standard versions of history. In particular, an attempt is made to explain the remarkable persistence of certain narrative structures in accounts of the Great Trek and mfecane written by historians working in different periods and informed by dramatically different ideologies.

When Cobbing finds the same story repeated in different eras, he suspects historians of complicity in a lie which serves the interests of dominant groups in South African society. Such explanations, whether cast in terms of interest group theory or structuralist theory, have much to be said for them but are less than totally satisfying because they have flourished not just at home but also abroad. Why should foreign scholarship dance to the favourite tunes of South African politicians, miners and farmers? What could move John Omer-Cooper in Nigeria or Kent Rasmussen in California to serve among the legions of ‘settler history’? Are there not other possible reasons for the persistence of certain story-lines?

The small and possibly smug world of precolonial southern African historians was rudely shattered in 1983 by the appearance of a brilliant polemic by Julian Cobbing entitled The Case Against the Mfecane'. In a series of increasingly provocative papers and lectures, Cobbing challenged the existing conventional wisdom, first put forward by John Omer-Cooper in 1966, that the Mfecane was a ‘revolutionary process of change from a single centre’, namely the rise of the Zulu kingdom. Omer-Cooper was much concerned to stress the Afrocentrism of his perspective. The Mfecane, he maintained, gave the lie to the view ‘that African societies had no record of autonomous development … [The Mfecane] was essentially a process of social, political and military change, internal to African society and taking place with explosive rapidity.’

Cobbing not only rejected Omer-Cooper's definition of the Mfecane, but he seriously questioned whether the event had taken place at all. He argued that the Mfecane was not an internally self-generated revolution occurring within northern Nguni-speaking society, but a convenient alibi, whereby succeeding generations of white historians rationalised and legitimated white seizure of black lands. The wars and dispossessions of the early nineteenth century were set in motion not by the rise of the Zulu state but by two chains of violence emanating from colonial aggression. One chain of violence, in the east, was initiated by Portuguese slave-trading from Delagoa Bay, which impacted directly on northen Nguni-speakers. Another chain of violence, in the west, was prompted by the labour needs of the Cape colony and augmented by the raiding activities of its armed and mounted surrogates, the Griqua.

It is not an exaggeration to say that the entire debate on the nature of southern African precolonial societies has been paralysed by Cobbing's intervention, and will continue to remain so until the crisis created by ‘the Cobbing hypothesis’ has been resolved. Fortunately, one of his articles has been published in an historical journal, ‘The Mfecane as Alibi: Thoughts on Dithakong and Mbolompo’, thereby making it possible for serious historical debate to commence. I will confine myself to those aspects of The Mfecane as Alibi’ which relate to my own area of specialisation, namely what became the eastern Cape/Transkei region. Much of what follows may seem negative and petty.

The term Mfecane has come to be widely accepted as a name for the process of political change and the accompanying wars and migrations which began in the area between the Thukela River and Delagoa Bay during the later part of the eighteenth century. Inter alia, this process resulted in the emergence of the Swazi and Zulu kingdoms, the founding of the Gaza kingdom in southern Mozambique, and the migrations and state-building activities of the Ndebele and of the Maseko and Zwangendaba's Ngoni.

The Mfecane as a Process of African State-building

In The Zulu Aftermath the Mfecane was presented as a positive process of political change in the direction of enlarged state power through improved military organisation and increased centralisation, and the expansion of the scale of political organisation through the rapid incorporation and assimilation of members of previously separate political communities. It was characterised as an essentially African revolution, the achievement of African leaders employing and modifying traditional institutions, values, weapons and tactics, rather than borrowing external models or techniques, in building their new kingdoms. It thus illustrated the capacity of African leaders for creative statecraft and the adaptability of traditional African institutions to new purposes. It demonstrated that the nineteenth-century history of Africa could not be adequately discussed simply in terms of external, predominantly European, forces acting on essentially passive African societies, but that it must be seen as a much more complex pattern involving reciprocal interaction between European and African initiatives.

The process of political change in the area inhabited by northern Nguni-speakers was described as beginning in the latter half of the eighteenth century in response to an intensification of intercommunity conflicts probably arising from heightened competition for grazing and garden land consequent upon population growth beyond what could be comfortably sustained within the area by traditional food production practices. The process involved an initial phase, dominated by Dingiswayo's Mthethwa and Zwide's Ndwandwe, in which the conversion of the corporate initiate-mate system to a means of military organisation and a device facilitating the incorporation of young males from other communities was implemented. The potential of this development, however, was not completely realised, in that the amabutho only assembled for fairly brief periods of active campaigning.

In a series of papers and articles beginning in 1983, Julian Cobbing has offered a radical, and often provocative, critique of the mfecane as the pivotal concept of the history of southern Africa in the nineteenth century. He asks vigorous new questions about everything from the identity of the ‘Fingoes’ in the south and the ‘Mantatee hordes’ on the highveld, to the extent of the slave trade around Delagoa Bay. Cobbing's work has stimulated a host of graduate studies on these topics, and has prompted a number of established students of the period to reassess aspects of their earlier work. The sheer scope of the critique is, however, also the source of its greatest weakness. In particular, Cobbing may be criticised for misusing evidence and employing imprecise periodisation.

Nowhere are these criticisms more pertinent than in relation to a central element of Cobbing's thesis, namely, his view of ‘Shaka-the-monster’ as a European invention to mask illegal labour procurement activities and land occupation. In this essay, I focus on Cobbing's reconstruction of the making of the Shaka myth. My purpose is to disentangle the elaborate weave of Cobbing's powerful insights and implausible conspiracy theories. I suggest that while Cobbing's critique is extremely valuable, especially in the way that it forces historians to question many of the assumptions with which they have for too long been extremely comfortable, he fails fundamentally to come to grips with the full complexity of his primary target, past historical myth-making processes.

Cobbing identifies four key elements in the notion of the mfecane as most commonly espoused: firstly, ‘a self-generated internal revolution’ within northern Nguni-speaking societies which culminated in the 1820s in the regionally dominant Zulu power led by a savage despot, Shaka; secondly, attacks by the Zulu on neighbouring chiefdoms which forced the latter to flee their land and which, in turn, displaced other chiefdoms still further afield; thirdly, a ‘cataclysmic period of black-on-black destruction’ (including cannibalism) leading to the depopulation of the interior of South Africa; with all of this culminating, fourthly, in the restoration of security with the advent of the Europeans. Cobbing's observation that this explanation of the depeopling of much of the interior and for the arrangement of the African inhabitants of southern Africa in a surrounding ‘horseshoe’ serves to legitimate white occupation of the land and the ideology of separate development is not new.

In 1991 a colloquium was held at the University of the Witwatersrand, Johannesburg, South Africa, entitled The Mfecane Aftermath: Towards a New Paradigm'. The title was a play on the name of John Omer-Cooper's 1966 publication, The Zulu Aftermath: A Nineteenth-Century Revolution in Bantu Africa. Omer-Cooper, following George McCall Theal's writings some seventy years earlier, argued that the emergence of the powerful Zulu kingdom in the early decades of the nineteenth century caused massive upheaval among neighbouring chiefdoms. This, in turn, set in motion a ripple effect of dislocation and disruption that extended through much of southern Africa, namely, the ‘Mfecane’. Omer-Cooper's study celebrated both the revolutionary changes within the Zulu kingdom which lay at the root of the tumult, as well as African responses to it - the innovations and achievements which underlay the period of migrancy and increased state-building throughout the sub-continent. This, then, was what Omer-Cooper focused on as the ‘aftermath’ of the Zulu.

In a series of papers presented at seminars starting in 1983, Rhodes historian Julian Cobbing mounted a campaign for the ‘jettisoning’ of the concept of the mfecane. Cobbing argued that the idea of a ‘Zulu explosion’ which set in motion the mfecane was a settler myth which conveniently obscured the disruptions of local societies caused by the labour needs of the Cape colonists and the demands of the Delagoa Bay-based slave trade. Despite their preparation for prestigious forums such as the seminar of the Centre for African Studies, University of Cape Town, and the African Studies Institute seminar at the University of the Witwatersrand, Cobbing's early papers elicited little response from historians. Then, in 1987, Cobbing presented yet another version of his case at the University of Durban- Westville, ‘The myth of the mfecane’. Shortly after that, in 1988 and 1989, an eminent historian of Natal and the Zulu kingdom, John Wright, collaborated with Cobbing in joint presentations to seminars at the University of the Witwatersrand and the University of Cape Town, entitled The Mfecane: Beginning the Inquest'. Attention to the debate shown by Natal-based interests and historians of the Zulu kingdom was stimulated in the later 1980s both by an efflorescence of new scholarship on the early Zulu kingdom and by contemporary political developments within Natal and KwaZulu.

The title of Julian Cobbing's article, The Mfecane as Alibi', suggested a key part of his argument: that the view of the Mfecane held by previous historians was not only a false one, but had diverted attention from something else of greater importance. What Cobbing called in his article ‘mfecane theory’ - the idea of a great upheaval caused by the Zulu - was, he claimed, ‘integral to a white settler, “Liberal” history', and had been refined and elaborated in the 1960s. For Cobbing, settler historians, liberal historians and apartheid apologists were essentially all of one mind in believing that the Mfecane began with, or was caused by (if they confined that term to the consequences of the rise of the Zulu state), ‘a self-generated internal revolution’ in what became known as Zululand. That these ‘mfecane theorists’ all treated the Mfecane as something separate from the colonial history of South Africa was, he argued, no accident. There was, he claimed, an ideological purpose at work, in particular a concern to justify the racially unequal division of the land. What these historians had failed to point to, as the major process at work in the early nineteenth century, as later, was the violence of colonialism.

In what follows I discuss some of the earlier historical writing on the Mfecane to demonstrate that Cobbing has presented too stereotypical a view of what previous historians - hardly ‘theorists’ in any usual definition of that word - said about the Mfecane. He did not allow for the way in which views had changed over time, nor did he discuss previous work in its historical context. Historians from Theal to the liberal Africanists of the 1960s all to a greater or lesser extent saw the Zulu as playing a major role in the Mfecane, but they by no means all agreed either on its causes or its consequences. Nor did they all treat the Mfecane as something entirely separate from the history of colonisation, or ignore the violence associated with colonialism. That the liberal historians failed to give more weight to colonial violence was not because, as Cobbing suggested, they supported or were linked to segregation and apartheid, but for other reasons. And it is pointed out that the main ideas in Cobbing's anti-Mfecane critique are also to be found elsewhere.

The intention of this essay is to establish that the period from about 1750 to 1820 was a time of conflict on the western highveld and in the adjacent southern Kalahari characterised by warfare and raiding. It should be noted that the notion of the outbreak of ‘Tswana wars’ in the late eighteenth century is not new - it was mentioned by Neil Parsons in his New History of Southern Africa (1982). This essay constitutes a fresh look at the evidence on which some of Parsons’ s conclusions were based - most notably, the ethnological surveys of Paul-Lambert Breutz. It also draws on my own research into the Hurutshe, and on theses by Leonard Ngcongco and Gary Okihiro on the Ngwaketse and Kwena. All of these works owe much to the pioneering work of Isaac Schapera, though his findings have not been exempt from critical evaluation. I attempt to reconcile and periodise oral traditions collected by Breutz, and to flesh out some economic aspects of the conflict only suggested in the works of Schapera and Ngcongco. I also refer to Legassick's innovative work on the emergence of the Sotho/Tswana.

The essay then examines the causes of the violence and offers a discussion of the processes of state-formation among these communities. Finally, this analysis is related to the re-evaluation of the difaqane initiated by Julian Cobbing. The intention is not to contest his wide-ranging critique, but rather to stress that conflict and violence engulfed regions in southern Africa other than the Natal/Zulu kingdom area where the origins of the difaqane are said to have been.

The peoples whose history is recorded here were the inhabitants of the Tswana chiefdoms (the Kgatla, Fokeng and Kwena) located in the present-day Rustenburg district and, directly to their west, the Hurutshe in the Marico, and the Ngwaketse and Kwena in the Kanye and Molepolole regions of southern Botswana. Reference will also be made to the Rolong and Tlhaping whose activities impinged on the affairs of the Hurutshe and Ngwaketse.

This book comprises a revised selection of the papers delivered at the colloquium The Mfecane Aftermath: Towards a New Paradigm’, held at the University of the Witwatersrand in September 1991. A full list of the papers presented at the colloquium is reproduced at the end of the volume.

The book has been divided into three sections. Each of the sections is preceded by a specially commissioned contextualising essay which offers an overview of the section, together with discussion of the main areas of debate among the contributors to the section. The contextualising essays are designed to serve as introductions to the dense arguments which follow, while their review of debated evidence gives non-specialists the benefit of the evaluation of a scholar well versed in that particular area.

Part One takes stock of the major historiographical and methodological issues. Part Two is concerned largely with the history of the eastern coastal region in the late eighteenth and early nineteenth centuries, and deals with some of the major debates over sources and their interpretation. Part Three examines events in the interior of southern Africa in the same period. Readers will detect greater methodological refinement in the use of sources in the essays in Part Two than in Part Three, as well as more detail and a better established chronology of political events. This unevenness is a consequence of the ‘Zulu aftermath’. Interest in the Zulu kingdom resulted in a long-standing concern with the history of the KwaZulu-Natal region, whereas the myth of the depopulated interior underlay the neglect, until recently, of the precolonial history of the highveld. Scholars like John Wright and Jeff Peires, working on the coastal areas, are able to draw on substantial bodies of recorded oral traditions, about which considerable background material exists, whereas the oral history of the interior has been relatively little attended to. The essays in Part Three prepare the ground for studies of events and close analysis of the pertinent texts of the kind already begun for the coastal regions.

Together, the various contributions to the book provide one of the first detailed overviews of the major developments of the later precolonial period in southern Africa.

During the 1820s the entire region of southern Africa was affected directly or indirectly by tremendous demographic upheaval and revolutionary social and political change. The period was marked by massive migrations, sporadic raids and battles, and frequent periods of privation and famine for many people in the region. This ‘mfecane’ has been explained in many ways by historians, but never adequately. The sociopolitical changes and associated demographic turmoil and violence of the early nineteenth century in southern Africa were the result of a complex interaction between factors governed by the physical environment and local patterns of economic and political organisation.

Increasing inequalities within and between societies coupled with a series of environmental crises at the beginning of the nineteenth century transformed long-standing competition over natural resources and trade in south-eastern Africa into violent struggles for dominance and survival. Trade at Delagoa Bay, involving primarily the export of ivory, had allowed some Africans to accumulate wealth and consolidate power, leading to political amalgamation at the expense of the weak. Increasing political inequality between chiefdoms and increasing socioeconomic inequality within societies made weaker people, with fewer entitlements to food, more vulnerable to famine in times of drought-induced food scarcity. When the area was struck by severe droughts and other ecological problems in the first two decades of the nineteenth century, competition became keen over fertile, well-watered land. Those who had already consolidated their power prevailed over weaker groups in the open contests that emerged. The weak found themselves incorporated into the lower echelons of stronger societies, either conquered involuntarily or submitting voluntarily for the sake of survival. The stage was set for the emergence of various strong leaders, some, like Shaka, who ruled with terror and others, like Moshoeshoe, who won the voluntary devotion of their followers.

The resulting period of state formation, often involving initial stages of conquest by force, was prolonged into two decades of violence because of the activities of Europeans encroaching from both east and west. In the east, slave traders found a new supply of ready victims in the area around Delagoa Bay and fostered continued violence to meet their demands for slaves.