Moving aside a dense epistemological screen

Since the 1820s the history of the KwaZulu-Natal region in the era before a European colonial presence began to be felt has very largely been written round the emergence of a series of kingdoms, culminating in the rise of the Zulu kingdom in the early nineteenth century. This polity was, and still is, typically positioned as the most “developed” example of the region’s independent “tribes,” or in later discourse, chiefdoms, kingdoms and states, on the eve of colonialism. As the most documented and discussed of these polities, it has long been treated as the example which illuminates by extrapolation the nature of less well-documented earlier political formations.

In the historiography, the rise of great African kingdoms received particular attention from the era of decolonization onwards. From the late 1960s Africanist writers portrayed the emergence of the Zulu kingdom as a positive achievement of uniting different polities over a large territory under the rule of Shaka.Footnote ¹ In the 1970s and 1980s Marxist scholars conceptualized the Zulu kingdom, and what they saw as a number of precursor polities, notably Swazi, Mthethwa, and Ndwandwe, as states whose rulers extracted a social surplus from subjugated groups.Footnote ² The “mfecane debates” of the late 1980s and 1990s saw the emergence of a challenge to the long-established notion that the upheavals and migrations that affected much of southeastern Africa in the 1820s and 1830s were generated by the rapid rise of a militarized Zulu kingdom under Shaka. But the debates largely left in place the idea that the rise of the Zulu kingdom formed the central theme of the history of the KwaZulu-Natal region before colonial times.Footnote ³

The upshot of two centuries of writing, from a variety of perspectives, about the history of the region before colonialism, with the rise of the Zulu kingdom as the apogee, has been to conceive of polities of the time as increasingly centralized, circumscribed entities, settled in demarcated lands, with distinct historical identities under historically legitimated ruling houses. These notions, for all their internal variations across time, are well-established tropes.

A corollary of the focus on kingdoms, or states, understood in this way, has been the conceptualizing of many of the people who were not encompassed in these entities as “refugees.” The notion that Shaka and other rulers of the time depopulated vast swathes of territory remained unquestioned in the literature until relatively recently in spite of scholarly arguments to the effect that these ideas were to a large extent rooted in the “myth of the empty land” propounded by European settlers.Footnote ⁴ The mfecane debates served to broaden discussion by attributing the widespread population movements of the time to the impact of European colonial expansion on African societies, but left in place the notion that groups which moved were by definition refugees “driven out” by stronger powers. But as Mathieu Rey has pointed out in his genealogy of the political concept “refugee,” the term acquired a specific meaning after the Napoleonic wars, shaped by the European order of the time and its understanding of states with sovereign territorial authority over a people with a unified national identity. This view was informed by an understanding of a body politic as a fixed entitity, and the term refugee was firmly differentiated from the idea of migrant.Footnote ⁵ In the light of this perspective, uncritical use of the term “refugees” to designate many of the groups that featured in the history of the KwaZulu-Natal region in the late eighteenth and early nineteenth centuries requires attention.

The notion that polities of the region in the period under study were more or less bounded units has gone hand in hand with the idea that each had a discrete “tribal” history and chiefly genealogy. This idea in turn is based on a particular way of reading the source materials. Writings on the history of the KwaZulu-Natal region in the era before colonialism are based ultimately on oral accounts that have been put into written words in a variety of forms by a wide range of authors under a wide range of circumstances. In general, these oral accounts have been seen, and are still generally seen, as “oral traditions,” that is, as accounts relayed across time in ways that preserve a core of information about the past. Variations in detail between accounts of the same events are often attributed to lapses of memory, faulty transmission and idealizations on the part of successive narrators. Efforts to fill in the gaps and smooth out inconsistencies have long been at the center of the methodology of using these accounts as sources of evidence on the past.Footnote ⁶

In short, inherited epistemological thinking about the polities of this region as more or less bounded units was inextricably linked to ideas about each of these units as having a discrete tribal history and chiefly genealogy, expressed in oral traditions handed on over time, ideally faithfully, but in practice subject to poor relay. These mutally reinforcing ideas about “states” and “refugees” on the one hand, and “oral traditions” on the other hand, in effect form a dense epistemological screen which has to be moved aside if we are to see the pre-Zulu period more clearly.

In response to History in Africa’s call for papers that look at how source materials, theories, and methods help us understand the complexities of motion, this article offers, in the first place, an overview discussion of the extent of the evidence for various forms of mobility in the eighteenth and early nineteenth centuries in what is today the KwaZulu-Natal region. While scholars have long recognized that there is considerable evidence for mobility, the full extent of this is far greater than we might think. In the article we marshal the big picture regarding mobility, explore the attendant political flexibility that this entailed, and discuss the underlying factors. In the second place, the article points to the implications of people regularly moving and often shifting their political allegiances from one ruling house to another. It further examines the understandings of sovereignty and identity that went hand in hand with this mobility and flexibility.

This in turn foregrounds the question of how people at the time thought about and strategized in relation to movement as well as political and economic changes, opportunities and threats. For the most part, scholars have relied on ethnographic studies from the early twentieth century in their attempts to think about how people in the eras immediately before colonialism understood rulership and identity. However, most of these studies were made at a time when so-called tribes had been firmly contained within bounded tribal areas under colonially recognized chiefships, and when ready mobility and political flexibility was no longer possible. When we turn to other sources, most notably historical accounts based on statements made by African commentators, the extent of the evidence for the movement of people and things in the earlier periods becomes clear. These accounts often indicate how people thought about mobility and change, and how they navigated change in their discourses.

The key theoretical move of the paper thus involves moving aside determining theories about primordial tribes and, more latterly, about settled state formation and the closely correlated notion of relayed oral traditions as bodies of material relevant to the history of a particular tribe or state. Instead, we focus on the ways in which ubiquitous mobility and ongoing political changes were constantly being navigated in dynamic oral discourse and deliberation. More specifically we argue that the past was continuously mobilized and reassessed in light of such changes, most notably for newly relevant historical connections and possibilities. We elaborate our methodology for drawing on notes of spoken accounts and on other writings about the past mostly by African commentators that were current 100 years and more after the period that we are concerned with as sources of evidence on political discourses and practices in the eighteenth and early nineteenth centuries. In many cases, the evidentiary detail turns on the giving of close attention to the form, content, and conditions of production of any one account. It also depends on the historical meanings of specific local language terms and discourses in such an account, the nuances of which are often effaced in translation. Our purpose in this article is not to rehearse the methodological work involved in unpacking this kind of detail in all the various accounts of particular events in the past, nor the many ways in which events came to be reassessed over a long period of time. That kind of detailed discussion in relation to events in the histories of a range of pre-Zulu polities and the early Zulu polity is variously available in certain of our earlier works.Footnote ⁷ More recently we have reworked this methodology and brought it into sharper focus in a number of publications and in the manuscript of a contracted book.Footnote ⁸ In this article we draw attention to the patterns which this approach brings to the fore, most notably how it highlights and makes sense of a variety of forms of mobility, long present in the available evidence, but not readily interpretable in terms of theories of either primordial tribes or state formation.

The central substantive argument of the article is that the multiple forms of mobility and the extent of political flexibility which characterized the region in the eighteenth and early nineteenth centuries meant that people needed always to mobilize connections with others. To do that, they continually reassessed the past in the light of contemporary changes and opportunities, and with a view to having open options in the event of further changes and movements in the future. The makers of oral accounts that invoked the past, the article suggests, seldom sought to produce seamless and straightforward narratives which sought to relay a timeless truth about the past or else instrumentally to support a current claim with a definitive historical argument. Rather, they typically engaged in reassessing the past in the light of present circumstances, while remaining alert to the possibility of future changes requiring further reassessment, and at the same time assiduously safeguarding ancestral well-being. Crucially, oral accounts functioned as banks of possible historical connections.

The article further shows that oral accounts of this kind, together with their written versions, point to ideas of sovereignty as being linked to the success of rulers in providing for shifting bodies of adherents, notably in the face of challenges and changes. Sovereignty was constantly being renegotiated, to the extent that a person might in the course of a lifetime give allegiance to several rulers from entirely different ruling houses, and live in as many different political formations in different parts of the region and beyond. Political identity was thus situational and contingent, even as ancestral connections were carefully nurtured and honored over time.

Signs of mobility and political dynamism

In the very first sentence of the article (written sometime in the period 1911–1913) that forms the opening chapter of the collection, A History of the Zulu and Neighbouring Tribes, missionary historian Alfred Bryant wrote, “The map of South Africa during the past 400 years has been quite kaleidoscopic in its ever varying arrangement of its Native tribes.”Footnote ⁹ Much of his later and widely influential synthesis of published works and oral narratives, Olden Times in Zululand and Natal (1929), was about this ever-varying arrangement. It opened with an account of the migrations of “Nguni” groups from the northwest into the KwaZulu-Natal region after about 1500. This was followed by bewilderingly complex accounts of the histories of numerous small “clans” (Bryant’s term) which, in his rendering, moved about the landscape, and split and resplit into offshoot groups, “before the present state of stabilization was reached in recent times.”Footnote ¹⁰

While historians today often recognize that movement in the past took place, they have generally paid little attention to the evidence, and its implications, for the mobility of groups that abounds in Bryant’s work, informed though it was by detailed oral accounts. Only recently has Norman Etherington drawn attention to how Bryant’s text in Olden Times “describes people in constant motion: political, temporal and geographical,” and how his work “opens a window on a longue durée of mobility and fluidity in social and political formations stretching back over many centuries.”Footnote ¹¹

This perspective is completely at odds with the common view – encapsulated in the European notion of “tribe” – that people in the times before colonialism lived in bounded and culturally homogeneous groups under the rule of authoritative hereditary chiefs in more or less fixed territories. At an empirical level, even a cursory reading of Bryant’s texts reveals numerous cases where groups of various sizes and composition, and individuals of a variety of standings, moved from one locale to another for a variety of reasons and with a range of effects. The same point can be made about, among others, works like missionary David-Frédéric Ellenberger’s History of the Basuto, the first major book on the subject, published in 1912; printer and journalist Magema Fuze’s Abantu Abamnyama Lapa Bavela Ngakona, the first full-length history book in isiZulu, published in 1922; and the ways in which, in the period 1897–1922, African commentators discussed the past with Natal colonial official and assiduous recorder of oral histories, James Stuart.Footnote ¹²

These works are replete with accounts in which sections of political formations split off to avoid what they experienced as oppression at the hands of rulers, or after disputed successions, with the aim of establishing their own autonomy, or of seeking protection from other rulers. Other passages tell of how homesteads moved to new locations because of drought, or outbreaks of human or livestock disease, or to find more productive agricultural and grazing lands, or to take advantage of new opportunities, or after quarrelling with neighbors, or to avoid accusations of ukuthakatha (malign thaumaturgy), or to avoid pollution. Senior political figures or wealthy homestead heads sent sons out to establish their own homesteads. Individuals and groups were constantly criss-crossing the landscape for a range of purposes – to visit friends and relatives, in search of adventure and opportunity, to negotiate marriages, to procure materials for domestic manufactures, and to carry provisions to family members called up in service of rulers. Parties of men traveled long distances to hunt or raid cattle or make war. Strategically negotiated marriages saw the ongoing movement of women in webs of interconnected natal and married contexts. Specialists in a variety of fields traveled widely – herbalists, diviners, metal-workers, officials, hunters, traders in medicines and snuff and in prestige goods like brass, copper, ivory, and exotic beads. As did porters, guides, envoys, spies, scouts, amabutho (units of men organized by rulers) on missions, and even ancestors being introduced to new homes through the practice of ukubuyisa. Footnote ¹³ All this kind of movement depended on the affordances of local topography in relation to look-out points, defensible redoubts, effective hunting and transit routes, fordable river crossings, mountain passes, and river breaches as conduits for the local movement of people, things, and ideas – among them new crops, expertise, knowledge, information, news, and intelligence – connecting political formations in the region with one another and with developments in the wider world. Mobility was underwritten by a variety of practices designed both to control movement, and to protect and support people who were moving.Footnote ¹⁴ Optimal routes often involved the use of obligatory passages that forced traders and travelers to converge at particular points – passes, traverses, crossings, shelters – that were subject to intensive monitoring and strategic controls. While traveling involved expectations of etiquette and of hospitality, it was also a source of uncertainties and dangers.Footnote ¹⁵ To guard against them, people practiced a range of rituals and made use of local connections and expertise. Movement was central in people’s social and political life.

The plethora of evidence about political and individual movement is complemented by evidence on the movement of stuff. Ivory, brass, iron goods, cattle, beads, finery, herbal remedies, potent substances, provisions, and raided grain, among other things, were all in motion. The movement of people and goods involved also the transmission of technologies, knowledge, expertise, songs, stories, fresh ideas, tidings, and news. Movement of people and the circulation of goods and information were at once both agents and results of processes of change and signals of dynamic political developments.

The ways in which these developments played out in the period under study was to a greater or lesser extent shaped by the particular features of the environment in which people – numbering perhaps 200 000 in an area of 100 000 square kilometers – made a living, an environment which, itself, had been shaped in certain ways by centuries of human activities.Footnote ¹⁶ The subtropical climate was warm enough, and rainfall, nearly all of it falling in the summer months, usually adequate for the seasonal growing of cereal crops in the form of sorghum, millet, and maize, and for keeping cattle and goats. Pronounced variations in the topography meant that resources of cultivable soil and of adequate grazing were unevenly distributed across the landscape. Annual variations in the length of the rainy season and in the amount of rainfall in any one area meant that the productivity of the land might vary considerably from one year to the next. Periodic droughts were not uncommon. These features put something of a premium on people’s expectations that they would be able to move as and when they thought necessary, both in the short term and in the long term, in order to have access to cultivable land and adequate grazing. These expectations lay at the heart of the political order that they held to.

While the nature and extent of movements that took place in the wider KwaZulu-Natal region before colonial times was substantial, it was not without constraints. Prime among them would have been the extended requirements of the seasonal agricultural cycle: preparing land, planting, weeding, harvesting, storage. Factors inhibiting movement to an unfamiliar place included the advantages of having intimate knowledge of a particular area and its affordances. In grassland areas where people had no option but to engage in the labour-intensive practice of building livestock structures in stone, they would probably have inhabited settlement sites longer than did people who had ready access to timber for building stockades. Successful iron-smelting required producers to work for extended periods near sources of ore and suitable wood for fuel.Footnote ¹⁷ Rulers sometimes made efforts to hold back the cattle of people who sought to move away, as was the case in the time of the Zulu kingdom.Footnote ¹⁸

While mobility was in part the result of local contingencies, it was also shaped by many of the factors which have long been cited as the causes of state formation and centralized settlement. Since the 1960s scholars have debated the relative importance of factors that are thought to have generated the political changes which led to the emergence of “states” in the KwaZulu-Natal region in the late eighteenth and early nineteenth centuries: rising population pressure, environmental degradation, long- and short-term fluctuations in climatic conditions, and the impact of external trade. While there is little evidence for the significance of the first two factors, the argument for political effects of the growth of international trade at Maputo bay in what is now southern Mozambique from the mid-eighteenth century onward, and of a period of severe drought in much of southeastern Africa in the early nineteenth century, particularly in the later 1810s and early 1820s is substantial.Footnote ¹⁹ Punctuations in the long-distance trade and competition over control of the trade, as well as in cycles of drought and good rainfall, have thus long been recognized as distinct factors driving regional political changes and movement just as much as they underpinned the establishment of any one significant political formation.

Overall, our perspective chimes with arguments made by scholars who have long recognized that one of the main checks on the arbitrary exercise of power by rulers in African societies before colonial times was the possibility that disaffected adherents would “vote with their feet,” that is, move away to give their allegiance to another leader. In the 1980s, Igor Kopytoff argued, on the basis of half a century of previous Africanist scholarship, that “[c]ontrary to a previously widespread stereotype of Sub-Saharan Africa as a continent mired in timeless immobility, its history has emerged to be one of ceaseless flux among populations that … are relatively recent occupants of their present habitat,” with movements of many kinds and dimensions having taken place.Footnote ²⁰ Similarly, Paul Landau discusses how the history of African societies on the southern African highveld before colonialism was characterized by ongoing movements of people and ideas over long distances.Footnote ²¹ Fred Morton writes that the early history of Setswana-speaking groups “is one of movement and dislocation” and of “the origin and emergence of resourceful, interdependent, quarrelling communities thrown together.”Footnote ²²

For their part, archaeologists, who for long have focused primarily on evidence for settlement, have also given some attention to evidence for mobility.Footnote ²³ Until recently, though, they have done little to theorize it.Footnote ²⁴ Recent work by archaeologist Rachel King points to how the historical disciplines typically read settlement as a sign of political order and success, and mobility as a sign of political disorder and weakness. Questioning this assumption, King develops an understanding of mobility and raiding in the mountainous areas of what are now southern Lesotho and the neighboring Eastern Cape as meaningful sociopolitical strategies undergirded by flexible sociopolitical institutions.Footnote ²⁵ Historian Peter Delius and archaeologists Tim Maggs and Alex Schoeman comment on “the high levels of movement in pre-colonial society” more generally, whether on the part of individuals, families or wider groups.Footnote ²⁶ The picture of mobility in the KwaZulu-Natal region, as in other parts of southern Africa, was no doubt one of considerable diversity in time and space. But in our view the historical evidence on the period that we are concerned with points inescapably to the notion that a degree of mobility was a defining feature of social and political life in the region.

A shifting political mosaic

Discussion of politics beyond a concern with the factors accounting for the rise of what has been seen as a series of increasingly stable centralized states that superseded one another has been cast mostly in relation to three broad themes, all strongly influenced by the forms of “settled thinking” that we discuss above. First, a “billiard ball” treatment of conflicts between states, in which they spent a lot of time fighting and cannoning off one another. Second, the occurrence of succession disputes, with writers paying relatively little attention to the logic of the political order underlying the disputes. Third, the exercise of power from the top down, with very little about the agency of subordinate leaders and of ordinary people.

In our respective earlier work, and in our more recent joint work, we have attended to the wider regional political mosaic in studies that engage with the histories of Ndwandwe, Hlubi, Mthethwa, Qwabe, Thuli, and Cele, and with aspects of identity-making that movements of people from one area to another brought about.Footnote ²⁷ That work shows that polities formed and re-formed in relatively short order across the eighteenth and early nineteenth centuries, and that in the process many groups shifted location. In the present article we further draw out the implications of these studies. The formation of polities before the colonial era, we suggest, had to do centrally with the processes in which ruling houses and potential allies, and established groups and newcomers, constantly negotiated new political arrangements. Our thesis is that, rather than forming through processes in which bounded territorial units coalesced into successively bigger units, the political formations of the period were shaped by the ongoing mobility of people in the landscape, and defined in terms of the particular groups that recognized the authority of particular rulers at any one time.

This leads us now to use the term “political formation” in preference to terms like “chiefdom” or “state,” and to track the politics of the movements of people more often than not simply described as refugees. This demands, we suggest, a new understanding of the nature of political authority in these relatively fluid formations, together with a critique of latter-day notions of “chief,” “chiefship,” “chiefdom,” and “state,” and of their application to earlier periods.

Dynamic sovereignty: Inkosi yinkosi ngabantu

If people moved or expected to move relatively often, the prevailing political philosophy about rulership and sovereignty needed to accommodate this feature of their lives. The picture of changing political relationships across time means that sometimes within a few generations, people could shift from acknowledging one line of rulers to another, and indeed, within a line, first an incumbent and subsequently a successful rival. This raises questions about what enabled people to identify with one ruler and that ruler’s adherents at one point in time, and relatively soon thereafter with another ruler and, importantly, in many cases, quite different co-adherents. We suggest that the evidence points to an understanding of sovereignty shaped not by immutable lines of authority and related to a delimited entity but by dynamic and flexible politics and pervasive mobility, in terms of which rulers were accepted as rulers because of their abilities to act both effectively and potently (more on this below) on behalf of adherents.Footnote ²⁸ This point is summed up in the much-cited regional adage, “Inkosi yinkosi ngabantu,” a ruler is a ruler by dint of the people. Of course, this political notion, like ideas about political power the world over, was an ideal that was realized in varying degrees in various situations. But as an ideal it is revealing about how the political order was understood.

Where imported European thinking about rulers was rooted in ideas of kingship as inherited by, and even god-given to, a defined people – what we might think of as a descending theory of government – rulership in the settings that we are concerned with appears to have involved aspects of an ascending theory of government based on the terms on which people agreed to it. On the one hand, this meant that rivals could seize rulerships and, by dint of their success, justify their positions of leadership. The acclaimed Mthethwa ruler, Dingiswayo kaJobe (who reigned from sometime in the late eighteenth century to about 1818), offers a case in point. Significantly, when Dingiswayo unseated and killed his father’s designated successor, Mawewe, there was no attempt to suggest that Dingiswayo was in fact the rightful heir. Rather his forcible assumption of the rulership was accepted and remembered for the benefits it brought Mthethwa. Accounts by African commentators draw attention to the markers of potency demonstrated in how he gained the rulership.Footnote ²⁹ By potency we mean the expression, through a charismatic personal aura, of the ability of a ruler to take power and of his capacity, as perceived by his adherents, to intercede with the ancestors and to demonstrate the necessary skills to ensure people’s security, freedom from misfortune, success in harvests, victory in warfare, and so on. The ethnographic literature relies on the idea that a ruler was seen as empowered to act for his adherents primarily by dint of his descent from a particular line of ancestors. By contrast, in our view a ruler’s potency lay not only in an ancestral relationship shared with certain of his adherents but in the power of his ubukhosi – that is, his ability to act as a sovereign ruler – to bring benefits not only to his kin but to his adherents in general. A ruler’s capacity to intercede with the ancestors of the ruling house was certainly seen as vital, but so was his ability to create conditions conducive to the well-being of various groups of his adherents with their own ancestral relations.

An ascending theory of government also meant that people could give their support to one ruler at a particular point in time, and to another later in time, in one political formation or by moving elsewhere. At the very heart of relations between people and rulers lay the practice of ukukhonza. Dictionaries gloss the word as meaning to serve, to subject oneself to; in political terms it can be seen as referring to the acts in which leaders of subordinate groups, as well as individuals, publicly acknowledged allegiance to a particular ruler, with the expectation of receiving protection and access to resources in return.Footnote ³⁰ In her incisive study of relations between people and rulers in the Table Mountain area near Pietermaritzburg from the eighteenth century to the 1990s, Jill Kelly sees ukukhonza as a long-established cultural inheritance that “enabled chiefly subjects to define belonging in a chiefdom and demand accountability, land and – especially important in times of war – security.”Footnote ³¹ As her work cogently shows, the precise workings of ukukhonza as a practice varied widely according to historical circumstances, but it was important enough to rulers and people alike for it to remain central in political life through a long series of often violent and disruptive changes.

Identity in motion

The threads of evidence on people and things in motion, on changing allegiances, and instances of ukukhonza, occur everywhere across a variety of recorded narratives. Insofar as colonial recorders picked up on them, they were regarded as confusing and set aside (by researchers like Stuart), or harnessed to the theories of migration being used (by writers like Fuze and Bryant) to account for origins and the populating of the region. By way of contrast, the approach that we are putting forward suggests that it is productive to see these threads as attesting to identities as dynamic, both at the level of political formations qua political formations, and as tied into changing ideas about the relative status and positioning of groups within political formations. This approach builds on recognition of a way of thinking, shared widely across the region, of distinctions articulated in local African categories between original inhabitants, incomers, and latecomers.Footnote ³² Rather than seeing this as something of a timeless traditional three-tiered system, in the way that some structuralist analysis does, the focus on mobility highlights the way in which the terminology involved itself labeled people situationally in time and space, and not as having fixed identities. It focuses attention on the dynamic ways in which ideas about identity played out in particular historical settings and how they were mobilized under changing circumstances, often in new places.

This approach builds on an argument that we have made in a previous study to the effect that the notion of shifting layerings of groups does not match neatly with notions of “ethnic” identity-making rooted in colonial-era notions about “tribal” identities.Footnote ³³ It is possible to track how at particular moments, diverse people, using discourses of movement, came to identify themselves situationally and in a variety of ways as belonging to a particular political formation at a particular time. It is also possible to track how, within emerging formations, ruling elites and their allies, again using discourses of movement, established social and political distinctions between themselves and categories of subordinate groups.

To make these kinds of points about identities being situational and tied to a politics of mobility is not to argue they were invented. In the first instance, people were concerned to give proper attention to matters relating to respected and honored ancestral relations, some dim and distant and inevitably open to interpretation, others immediate, pressing, and incontrovertible.Footnote ³⁴ This involved the active management of ancestral relations, notably when people changed their allegiance from one ruler to another, and set limits on how the past was reassessed. Other limits related to how newly emergent identities were understood to be rooted in historical grounds.

Stories of arrival from elsewhere, for instance, were backed by narratives of movement anchored in landscape references. As an example, we take accounts of the origins of a named political formation, Ndwandwe, which emerged and flourished in the late eighteenth and early nineteenth century in an area centered on Magudu and the Nongoma plateau in what is today KwaZulu-Natal.Footnote ³⁵ These accounts, given by descendants of the ruling house, anchored the earlier history of the forebears of what became the Ndwandwe ruling house in movements from a distant place somewhere south of Maputo bay. They make clear that these movements took place in a shadowy, pre-Ndwandwe past, and that the origins, or umdabuko, of Ndwandwe as a distinct political formation lay in the settlement at Magudu and the subsequent expansion of the area under the ruling house’s authority. The thrust of these accounts was quite different from that of an early white writer like Bryant, who read the evidence as indicating that the origins of the Ndwandwe kingdom lay in the Maputo bay area.Footnote ³⁶ Where Bryant assumed that the origin of Ndwandwe came before movement to Magudu, for the descendants of the ruling house, movement culminated in the origin of the Ndwandwe political formation at a particular point in time and space.

Similarly, in an especially vivid instance of terminology labelling people situationally in time and space, the incoming forebears of what became the ruling house of the Swati kingdom to the north of Ndwandwe came to be known as the original people, the bomdzabuko, while those whom they found in situ were known – and differentiated from the bomdzabuko aristocracy – as the people found ahead, emakhandzambili. Footnote ³⁷

In another instance, the ruler of the new Zulu kingdom, Shaka kaSenzangakhona (reigned c. 1816–1828) identified assertively as Ntungwa, an identity associated with groups which were said to have come into the area from the north “by means of a grain basket.” This was a broad claim, shared by groups in an otherwise diverse emerging elite, to an identity that distinguished them from others in the new Zulu political formation.Footnote ³⁸ In Shaka’s reign the Zulu leadership sought to establish a clear social distinction between the Zulu ruling house and its close allies, on the one hand, and the generality of the groups over which it exercised authority, on the other. Identification as “Zulu” was reserved strictly for members of the ruling house.Footnote ³⁹

Because fluidity, flexibility, and mobility, including the forming and reforming of political identities, were, so to speak, in the DNA of regional political dynamics alongside carefully maintained ancestral connections, leaders of large and hybrid political formations needed to develop as well as they could means of holding such formations together. This involved, among other things, defensive alliances, strategic connections, and marriage networks; controls over communication, intelligence, expertise, and knowledge; trade and resource monopolies and distribution networks; and centralized coercive capacities for both cattle raiding and for defence. It also entailed active engagement in identity-making and remaking, involving the mobilization of ideas of commonality and difference. As power shifted, the newly powerful – new rulers, ruling houses, and their allies – had to wield and think about the nature of power, and about how to mix and sometimes weld disparate people into a composite political formation. Many of these concerns were shared, whether in support or resistance, by those who were subjected, or subjected themselves, to new powers.

Extensive evidence concerning political mobility and flexibility suggests that under certain circumstances, people could be more amenable or more vulnerable to political centralization, and that in other circumstances they were less so. Our approach suggests that for as long as movement elsewhere was possible, it remained a valued potential option. Potential was a factor that shaped the political practice and discursive activity of both rulers and their adherents. It depended on having options in one’s connections, and also on maintaining communication with others in order to have knowledge of possible opportunities for establishing advantageous connections. The implication is that identities could not be narrowly fixed if people wanted to be open to potential relationships, knowing that they might need to move if such opportunities opened up. Theorizing this kind of political potential provides an opportunity for breaking out of a genealogy of political theory shaped by a focus on the differently connected and constrained Eurasian world of this same period.

Networks of connections and communications

Activating potential depended on communication and trade networks, intelligence-gathering, knowledge of diverse areas and their topography and terrain, and logistic capabilities. What emerges from highlighting evidence for movement across space and time is the ability of people to take advantage of the affordances of carefully cultivated and constantly curated networks of cross-cutting connections, whether they were moving of their own volition or being forced to move, or moving information and things. In all cases, networks of relationships were key forms in the management of risk and the exploitation of opportunity.

At the heart of these connections, we would argue, was the strategic management of two fundamental sets of relationships. One had to do with politically purposeful relationships entailed in distributed marriage connections, often built up over generations, and the movement of women between their natal and marriage homes, also continually referenced in the accounts. The other had to do with notional relationships of common origin and past reciprocities, as expressed in complex and shifting discourses about the past and expressed in thickly layered claims, with multiple reference points in the available accounts.

Marriage relationships were key in developing connections between unrelated or distantly related people. Married women continually drew on political support from their natal homes in managing the affairs of their households in the contexts into which they had married. They relied on their natal contexts to underpin their status in their married contexts, and to secure successful accessions of their sons and contest the successions of the sons of other wives. Often diplomats extraordinaire, women also used the affordances of their married contexts to direct benefits back to their natal contexts and vice versa. Under certain circumstances, they and their sons, and even on occasions their husbands, had reason to draw on the married woman’s mother’s and grandmother’s natal contexts. Marriage strategies thus gave the families of both sons and daughters opportunities to secure multiple connections, capable of supporting flexible immediate and future strategies for exploring opportunities, movement, and risk management.

Women did not simply disappear into new married contexts, but actively visited, and were visited by, relatives, and participated in matters relating to both their natal and married contexts. Movements of these kinds facilitated ongoing exchanges of information and, in certain instances, political intelligence-gathering by women and visiting relatives. Similarly, women who moved from their homes to live and work in izigodlo (rulers’ establishments of wives and of unmarried women), where they existed, effectively connected their natal homes to the court, the workings of which they were closely and continuously exposed to. As izigodlo women whose marriages to assigned men were arranged by the ruler, they further connected the court to their married contexts. By sharing information among themselves, married izigodlo women facilitated the lateral movement of information across these contexts.

Considerable work also went into the making and maintenance of other kinds of networks of connection within and between political formations, among them offensive and defensive alliances; tributary relations; and relations involved in setting up new imizi (homestead) for sons. Particularly important in linking people of different political and social standing were relations of ukukhonza (see above) and relations of ukusisa, the practice in which wealthier individuals loaned out cattle to poorer acquaintances with the expectation of receiving their political support. As is well recognized in the literature, it was to the advantage of wealthy owners of cattle to spread their herds by lending out several head at a time to poorer individuals. This practice provided a dispersed form of insurance against outbreaks of cattle disease and the effects of drought; it also enabled owners to conceal the real size of their herds from the eyes of jealous neighbors and rapacious rulers.Footnote ⁴⁰ Furthermore, debts in the form of cattle, whether loaned in ukusisa or in ukulobola (transfer of bridewealth) transactions, created social and political connections which both lender and borrower had an interest in maintaining over time.

The strategic management of notional ideas of common origins, past connections, and previous reciprocities was crucial in face of changes, whether in negotiating a move into a new setting, mitigating the effects of drought or famine, introducing new kinds of crops or meeting new forms of resource demand, or responding to the emergence of aggressively expanding neighbors. Taking up options in relation to these kinds of change required having multiple potential allies as well as access to refuges, military support, loan cattle, and alternative land to settle on, as well as multiple forms of capital in the form of specialist knowledge and expertise that was desired locally and in other places.

The capacity of people seeking for whatever reasons to move, to forge new political relationships, to activate earlier connections as well as existing ones, and to mobilize alternative possibilities at different points in time in response to new conditions depended on ongoing communication and the work of historical memory, and sometimes its refurbishment. The seeking of support, refuge, or new opportunities was often bolstered by the mobilization of historical arguments about past marriages, historical reciprocity, kinship, and indeed many other kinds of connection. In the process, advisors, strategists, logisticians, spies, envoys, custodians of history, praise poets, court and local intellectuals, performers and ritual specialists, as well as ordinary people, thought critically about and discussed changing power relations and possibilities, maneuvered in response to change, assessed available resources and gave expression to their ideas in imaginative ways. The matter of having viable options was no accident but the result of careful strategizing and the maintenance of relationships across time, often rooted in historical obligations, and the preservation of historical memory which required assiduous curation and ongoing work. In short, historical consciousness was located at the heart of political relations, and the production of history was a vital aspect of political praxis.

Dynamic discourse

The evidence for active engagement with the past in order to manage the political changes, flexibility, and mobility which, we suggest, characterized this region in the eighteenth and early nineteenth centuries, lies in what have long been thought of as relayed oral traditions. It comes into sharp relief when we pay close attention to these accounts as instances of dynamic discourse. Evidence of the dynamism of discourse, we suggest, is thickly present in, among others, the many accounts recorded by James Stuart in the late nineteenth and early twentieth centuries, the holdings of the Swaziland Oral History Project assembled in the 1980s, the praise poems of prominent figures of the eighteenth and early nineteenth centuries, and other inherited literary forms such as amahubo and izithakazelo, long-standing local language place names, writings of early generations of African commentators, notably but not only in the early Black press, and the ritual archive of the region. Evidence of the dynamics of discourse can also be discerned in the writings of early white researchers, where the signs of this dynamism are treated as a weakness in the sources. Writers like Bryant and Ellenberger, whose historical works drew heavily on oral accounts, recorded their exasperation at the many ways in which these accounts failed to provide the definitive kind of evidence they were looking for to support their models of large-scale Bantu migrations, bounded units with discrete “tribal” history and authoritative chiefly genealogies.Footnote ⁴¹

For our part, we have for some time been concerned to develop new approaches to reading recorded oral and early written accounts as rooted in an understanding of the dynamics of historical discourse.Footnote ⁴² In the course of the nineteenth century and on into the twentieth century, a wide range of thinkers and commentators engaged with the history of earlier times, in a variety of spoken situations and in writing. Our key point is that, in drawing on notes of conversations and on other writings about the past as sources of evidence on political discourses and practices in the eighteenth and early nineteenth centuries which are the subject of our research, we are drawing on records of discourses and practices that were current 100 years and more after the period that we are concerned with. The people who wrote and spoke about the remote past in the late nineteenth and early twentieth centuries did so in terms of the political conditions, and varied discourses and practices of their own times, only some aspects of which were informed by inherited ideas and practices.Footnote ⁴³ Recognition of these points lies at the heart of our methodological approach. It demands that we deal with changing historical and political discourses and practices in later times in efforts to grapple with how later thinkers conceptualized politics and identities in the remote past. One of the changes was that the kinds of mobility that had long been characteristic of regional political praxis was no longer possible. What remained important in the face of colonialism, however, were inherited practices of political flexibility, techniques for the navigation of change, and the nurturing of options as part of the management of risk and the pursuit of opportunities.

We thus take it as axiomatic that significant aspects of historical accounts were shaped and reshaped over time in situations of ongoing change. Political changes, in particular, drove discursive processes and intellections which drew on the past in a dynamic way to make sense of the present and the possible future. To make this argument is not to deny the existence of stories about the past retold across time, stories that sometimes took the form of heroic epics and fireside tales laden with inherited wisdom outside the roil of high-stakes political rhetoric and argumentation.Footnote ⁴⁴ It is, rather, to recognize and acknowledge the intensity and effort involved in reflection and sustained thought, and its expression in discussion and debate, about the significance of the past in conditions of change, as well as the accumulating effects of this across time. It is to explore how understandings of the past were marshaled in considerations of the key questions of the day, and to offer an intellectual history of the making and reshaping of these understandings, and then to explore what they might tell us about earlier eras.

It involves, among other things, researching the histories of production of the various materials that we use as sources for historical enquiry into eras immediately before colonialism. It requires grappling with their making over time and positioning them in relation to the wider discourses, as well as to the political and intellectual imperatives and the shifting worldviews, that gave them shape. In short, at the heart of our method is a practice of engaging both written accounts and so-called oral traditions as productions of the time of their recording, shaped by the longer histories of the production of ideas of the past on which they drew. In certain cases, it involves engaging with oral accounts of the past as complex and, often, carefully considered intellectual disquisitions, rather than assuming that they are simply relayed and formulaic tales, traditions, and legends. This involves tracking ideas about the past as they travel across oral forms, from oral forms into written ones, and into forms with the oral and written inextricably entangled. We also look at the way in which historical materials have been reshaped over time as they have been reconsidered, recorded, translated, edited, and deployed as evidence.

This approach highlights a point originally conceptualized by Isabel Hofmyer and subsequently elaborated by Carolyn Hamilton that oral historical accounts “live” by their fluidity.Footnote ⁴⁵ Applying this point to historical accounts relevant to the period with which we are concerned not only reveals much about discursive processes and the intellectual work involved in the ongoing production of historical accounts, both in that period and in the times when such accounts were recorded. It also tells us about the nature of the dynamic political changes over time with which these discursive processes and instances of intellectual work were engaged. In short, we have found that paying close attention to the dynamics involved in the production of discourse, as much as to the older pursuit of cores of unchanged material, brings into sharp relief movement in many different forms, notably evidence of the significance of mobility and political flexibility. Such movements, and the imperatives that they generated, perhaps most notably in the navigation of the changes which they involved, prompted ongoing reflection on available historical bases for newly relevant connections and relationships and the taking of new opportunities. In short, movements and the changes which they entailed in the region and the period with which we are concerned were not simply remembered, whether well or poorly recalled. Their relevance and meaning were actively engaged and reassessed in processes of ongoing change. The source materials point to a long history of knowledgeable commentators viewing the past as something to be openly discussed and argued over. Thus in 1903 Jantshi kaNongila and Thununu kaNonjiya argued in James Stuart’s presence about the precise circumstances of Shaka’s birth.Footnote ⁴⁶ In his earlier press articles, and in his 1922 book, Magema Fuze exhorted his readers to engage in historical deliberation and to do their own historical research.Footnote ⁴⁷ In the 1980s, Dumisa Dlamini, the host on Swazi Broadcasting Services of a radio program featuring expert clan historians, invited his experts and listeners to actively engage in debate about the past.Footnote ⁴⁸

In situations where political entities formed up and re-formed in new configurations relatively rapidly, people identified in new ways with the emerging configurations, while recognizing the possibility of future realignments, and importantly, holding onto knowledge of older configurations and relationships. They maintained consciousness of shared history of past political formations centered in particular places which, historically, gave shape to their identities. They further maintained a strong consciousness of their particular recent ancestors, ancestral lines and ancestral connections in the past as shaping their identities.

Aspects of these past ancestral and political associations were mobilized or downplayed according to the demands and possibilities of their presents and of unfolding futures. In relation to both ancestral connections and past political relationships, the past that was relevant to the present, and associated understandings of identity, was neither absolute nor invented, but, rather, open to ongoing reflective engagement and critical reassessment. Identities were thus dynamic and in process, as well as relational. They were a product of respected ancestral connections and past encounters and associations, were shaped by changing conditions, and took form through interactions. This understanding of identity is a far cry from the primordial tribal, ethnic and national identities in terms of which the history of this region has long been understood. It is also very different from the Zuluist identity politics of colonial times and after, when mobility was no longer an option.Footnote ⁴⁹

Our conceptualization of the dynamic relationship between identity and politics over time brings past historical deliberative activity into focus in a manner that marks our research out from much of the preceding scholarship on the period. Our approach is thus distinctive for the way in which it focuses on the nature of political thought, intellectual activity and shifts in discourse that accompanied ongoing political change. This allows us to foreground the dynamism in the relationship between politics and identities in a way that those scholars who approach this past with a structuralist paradigm do not do, and in a more granular way than archaeologists and historical linguists whose work mostly deals with change on a larger scale. It further suggests that the varying traces of these activities in oral accounts are not so much evidence of their inconsistencies and weakness as historical truth as they are testimony to complex and layered discursive practices in action over time.

The past in political discourse in situations of mobility

In foregrounding dynamic and creative political thinking and activity, and the value of wide-ranging regional knowledge – both historical and up-to-date – that was needed both in rulers’ courts and in commoner homesteads, we seek to move beyond the kinds of questions that historically have occupied scholars steeped in the inheritances of the colonial imagination of tribal life and schooled in thinking about these societies in the terms of African political anthropology. In our analysis, much that was historically conceptualized by European commentators as timeless custom and tradition emerges as responsive intellectual, political, and creative engagement. Consideration of the past as terrain of discussion of present and future possibilities, management of change, and critical consideration of the ethics of rule was, we suggest, generally seen as fundamental to the nature of political praxis.

In summary, this article puts a focus on engaging critically with four entrenched aspects of thinking about the period under study. We focus on movement and political fluidity rather than falling back on habits of settled thinking in terms of firmly demarcated states; we attend to understandings of dynamic and contestable sovereignty rather than lines of hereditary genealogical succession; we explore the political activity of women, moving aside the trope of jealous wives and looking beyond notions of women as “beasts of burden”; and we foreground evidence of processes of historical reassessment as opposed to a habituated accepting of oral tradition as passed down in processes of passive generational relay.

In relation to all these features, our revisionism does not seek wholesale overturning so much as evidence-driven modulation. We foreground the weight of material concerning mobility, political flexibility, and fluidity, as well as regularly and overtly contested successions and appropriations of rulership, and lateral webs of connectivity that were animated by the political activity of women and by curated pasts. In the process we develop new perspectives on sovereignty, political practice, identity-making, and the nature of political thought at the time. Once the first of the assumptions which we unsettle is called into question, the others demand reconsideration. If we pivot away from assumptions about notions of progress and development lying in supposed signs of the evolution of big states and established monarchies, much of the available evidence stops appearing confused, tatty, and fragmented, and reveals the rich logic of its interlocked complexity.