principle of hierarchy: a chief or a king holding a dynastic name or title. The title was vested in a specific kin group, but its sway was also acknowledged by other lineages, who were themselves the keepers of subordinate titled positions and who might sometimes compete among themselves for the topmost dignity. But this was easier said than done, for the region’s scattered population, vast and easily traversable spaces, and relative scarcity of natural resources magnified the challenges of state building. As John Darwin aptly put it, where “rebelling meant no more than walking away to found a splinter community,” the job of leaders was very tough indeed.5 In the central savanna, even more than elsewhere in sub-Saharan Africa, the key objective of aspiring big men and state builders was always to establish durable claims over the labor and loyalty of unrelated people. Given the frequent absence of standing armies until the latter part of the nineteenth century, violent conquest was only effective in the short term and when it proceeded alongside less disruptive, “softer” forms of rule. Different societies came up with different solutions to overcome the parochialism of localized descent groups. Invariably, such solutions were related to the ecological specificities of their respective areas.
An early center of political experimentation in the interior of central Africa was certainly located in the Upemba Depression. Archaeological evidence in the form of copper ornaments and small iron bells suggests that processes of social and political differentiation were at work in this comparatively densely populated floodplain on the upper Lualaba River, in present-day southern Congo, since at least the first centuries of the second millennium.6 The rise of wealthy ruling groups in the area may have had something to do with the need to contain inter-lineage competition focusing on access to the floodplain’s rich, but finite, alluvial soils and to its game and fish resources. The same authorities might have also been responsible for coordinating such hydraulic works as were required to keep local economic life viable.7 Experiments in conflict management and political integration on the upper Lualaba are likely to have influenced developments in its immediate surroundings, beginning with what would become the heartland of the Luba “Empire,” the district located between the Lualaba and Lomani Rivers. Alternatively (or additionally), it is also possible to speculate that control over access to scarce trading resources—the salt and iron with which the future Luba core area was endowed—enabled one specific descent group to emerge as locally dominant and to regulate and tax the visits of outsiders seeking the same resources.8
MAP 1.1. The central savanna, c. 1800.
By c. 1700, the time that a Luba dynastic kingdom becomes recognizable in the oral historical record,9 elaborate political hierarchies, revolving around the Mwant Yav (Mwata Yamvo) royal title, had also come into being among the Ruund, to the southwest of the Luba.10 The first substantial written mention of the Ruund state, by the Angolan slave trader Manoel Correia Leitão, dates to 1756. By then, the “Matayamvoa” was being described as a “powerful” conqueror and his followers as “terrestrial Eagles,” raiding “countries so remote from their Fatherland only to lord it over other peoples.”11 Well-known traditions expounding on the marriage between the Ruund princess Ruwej and the wandering Luba hunter Chibind Yirung (Chibinda Ilunga) have frequently been interpreted as implying some form of Luba military conquest or, at a minimum, strong Luba influences on the genesis of the Ruund kingdom. In fact, the linguistic data examined by Jeff Hoover in the 1970s and the objective differences between the Luba and Ruund political systems suggest the playing out of more complex and longer processes of mutual borrowing than the conquest state model allows for.12
The twin institutions of positional succession and perpetual kinship—the Ruund trademark contribution to the precolonial political history of the central savanna—were certainly endogenous innovations.13 Jan Vansina has recently called them “a stroke of genius.”14 Positional succession and perpetual kinship established permanent links between offices rather than individuals, and this meant that the Ruund kingdom “ideally consisted of a web of titled positions, linked in a hierarchy of perpetual kinship” and occupied by people of different background.15 Because these institutions could be adopted without disrupting preexisting social structures, they became wonderfully effective means of imperial expansion. Subordinate hereditary positions could be created for real or honorary sons of a given Mwant Yav; their descendants—no matter who they were, or how far they lived from the Ruund heartland on the upper Mbuji-Mayi River—would continue to acknowledge the original connection, quite independent of the actual biological relationship that would eventually obtain between them and the successors of the Ruund king by whom the appointment had first been made. The integrating effects of positional succession and perpetual kinship were reinforced by another Ruund technique of rule: the recognition of the role of the “owners of the land.” The distinction between “owners of the land” and “owners of the people” was rooted in the ancient political culture of the savanna, but the Ruund systematized it and broadened its application in the context of an imperial strategy. Both in Ruund and Ruund-influenced Lunda states, the leaders of autochthonous groupings were not eliminated or marginalized. Rather, they were granted important ritual prerogatives. Ruund and Lunda political rulers were the “owners of the people,” dealing with the nitty-gritty of daily governance. But, though they were largely excluded from the sphere of temporal government, the “owners of the land” were still accorded a glorified position in the new dispensation. This was partly because they were believed to be in contact with the spirits of their ancestors, who exercised forms of supernatural authority over the districts they had first colonized.16
Thus, between the eighteenth and the nineteenth century, while the Luba sacred kings (Mulopwes) expanded their sway by favoring the accession of select peripheral lineage heads and incorporating them into the bambudye, a cross-cutting secret society which they controlled, the workings of positional succession and perpetual kinship helped bring into being a Lunda “Commonwealth.” The “commonwealth”—a definition which Vansina advocates in preference to the more traditional one of “empire”17—consisted of a network of independent, though interconnected, polities. While their leaders claimed real or putative origins among the Ruund, and though they recognized the Mwant Yavs as the fountains of their prestige, these Lunda kingdoms were not ruled from a single center and did not form a single cohesive territory. The “commonwealth,” in fact, consisted of a series of dominions ruled less by Ruund proper than by elites who had adopted Ruund symbols of rule and principles of political organization.
This Lunda sphere of influence—the extent and workings of which were first reported upon by the famous Angolan pombeiro, Pedro João Baptista, at the beginning of the nineteenth century—was crisscrossed by tributary and exchange networks and covered a large swathe of the central savanna. Its easternmost marches were occupied by the kingdom of Kazembe, founded as a result of the collapse of a Ruund colony on the Mukulweji River towards the end of the seventeenth century and the subsequent eastward migration of a heterogeneous group of “Ruundized” title holders.18 In the west, the holders of the Kinguri, the royal title of the Imbangala kingdom of Kasanje, dominating the middle Kwango River since the seventeenth century, also claimed Ruund origins and—as will be seen below—became the Mwant Yavs’ principal trading partners in the eighteenth century. In the south, smaller Ruund-inspired Lunda polities took roots along the Congo-Zambezi watershed. Early written evidence shows that “travelers” from the Lunda-Ndembu polity of the Kanongesha (“Canoguesa”), near the present-day border between Zambia, Angola, and Congo, were wont to take “tribute” to the Mwant Yavs in the 1800s.19 There is no reason to believe that the southern Lunda of the Shinde and related titles, further to the south, would have behaved any differently.
Political change was not necessarily the result of diffusion or borrowing. Processes of state formation could be more insular and self-contained than in the Luba and Ruund/Lunda cases. The Luyana (later Lozi) state is a good case in point. Its rise owed very little to external influences, but
