Читать книгу The Gun in Central Africa - Giacomo Macola - Страница 13

1

Power and International Trade in the Savanna

THIS CHAPTER offers a preliminary overview of the main drivers of the history of the central African interior. Its principal aim is to contextualize the case studies presented in parts two and three of this book by exploring, first, the workings of power in the central savanna from c. 1700 and, second, the changes in governmentality precipitated by its growing involvement in global trading networks over the course of the nineteenth century. As Steven Feierman and other historians of eastern Africa have argued, such changes were often revolutionary, leading to the emergence of new social groups, new polities, and new ways of enforcing authority.¹ In the troubled nineteenth century, individual charisma and military success became undoubtedly more central to the wielding of political power than they had been in previous centuries. Still, the impact of violent innovation was not the same everywhere. The “hereditary” and “mystical” principles of political organization discussed in the first section of this chapter did not disappear overnight, and an exclusive stress on historical ruptures runs the risk of obfuscating patterns of continuity.² The experiences of dislocation and turmoil were pervasive, but so were attempts to neutralize or adapt to them. Commercially driven violence and the increasing availability of firearms could provide the bases for the growth of new warlord polities and related mercenary groups, and they could bring to a premature end preexisting state-building efforts. But they could also be harnessed by, and thus inject new life into, the latter. Broad generalizations, then, are not the best way to address the changing political culture of the central savanna in the nineteenth century. Neither should the one-sidedly gruesome descriptions that many coeval Western observers indulged in be swallowed hook, line, and sinker. What can be said with certainty is that the intrusion of merchant capital and its African spearheads left central Africa more politically and culturally heterogeneous than it had ever been at any time in its long past.

The chapter also introduces the theme of firearms, describing the timing and modalities of their arrival on the central savanna and offering some initial indications of the disparate reactions that they gave rise to. The interaction between the peoples of the central African interior and firearms must be regarded as an instance of cross-cultural technological consumption. African understandings of guns were as complex as they were contingent, and one of the overarching arguments of this book is that the meanings and functions that the peoples of the central savanna attributed to firearms were shaped by preexisting sociocultural relationships and political interests. Without an appreciation of the multiplicity and diversity of such relationships and interests, it is impossible to grasp the logic behind the heterogeneity in patterns of gun domestication that characterized the region. Guns, as later chapters will show, were appropriated differently by different groups, for different were the sociopolitical contexts into which the new technology came to be fitted.

The final objective of this chapter is to introduce nonspecialist readers to the intricacies of the precolonial history of a macro-region that is frequently overlooked in recent general syntheses. It is therefore unashamedly encyclopedic in tone and structure. Since it paints with a broad brush and covers a wide array of areas, peoples, and themes, the chapter might perhaps be regarded as a kind of legenda, to which readers might want occasionally to refer back as they proceed with the rest of the book.

THE CENTRAL SAVANNA TO THE EARLY NINETEENTH CENTURY

This section concentrates on the workings of power in the interior of central Africa before external trading influences began to make themselves uniformly felt in the nineteenth century. Any such discussion must begin by stressing that the vast stretch of open grasslands and woodlands found between the Congo basin rainforest and the Zambezi River offered an altogether unpropitious environment for political entrepreneurs. In this region of “pedestrians and paddlers,”³ state building involved the consolidation of structures and institutions that brought together for regulatory and extractive purposes several descent groups, the main units in central African political relations over the past thousand years or more.⁴ Commonly, the process was predicated on the recognition of an overarching center of power providing a unifying 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 was instead shaped by the complex politico-economic requirements of the upper Zambezi floodplain. Unlike the Upemba Depression (and much of the central savanna), the upper Zambezi floodplain could support cattle keeping. Yet, in other respects, the two ecosystems were comparable, for the Luyana heartland, too, consisted of a special environment that needed to be closely managed if its economic potential was to be fully realized. Annual floods compelled the people of the plain to build their villages, grow their crops, and herd their cattle on both natural and artificial mounds. At the height of the floods, however, even such mounds had to be temporarily abandoned and temporary residence taken up on the plain’s forested margins. These conditions favored the development of a particularly centralized form of administration. Not only did the Luyana monarchs (Litungas) become “director[s] of the public works of the kingdom,” but control over the allocation of scarce natural and man-made mounds also gave them an important means with which to buttress their power and position.²⁰

The floodplain’s large labor requirements were met by means of internal slavery—historically much more pervasive in Barotseland than anywhere else in the central savanna²¹—and through the makolo system. The makolo were military and, especially, labor units to which all the inhabitants of the floodplain belonged from birth.²² They, too, worked as a means of royal centralization, for in the eighteenth century members of the royal family were seemingly replaced as makolo leaders by appointive officials. In time, the creation of new makolo and related headships became an exclusive royal prerogative. Thus, makolo leaders ended up forming what Mutumba Mainga calls a “bureaucratic aristocracy,” whose elevated social status depended solely on its alliance with, and loyalty to, the kingship.²³ A system of territorial governorships overlapped the makolo. Like the heads of the latter, the officials in charge of specific districts were also selected on the basis of merit rather than birth. Their functions were probably mainly judicial, since it was the makolo chiefs who controlled people “for purposes of raising an army, collecting tribute and recruiting labour.”²⁴ The marginalization of the members of the royal family to the advantage of the Litungas and their officialdom had one important consequence. Deprived of accessible outlets for their ambition, Luyana princes began fiercely to compete for the biggest prize of all: the position of Litunga. This was especially the case during interregna or following successful conspiracies. In this, the kingdom of the Luyana resembled that of the Luba, among whom royal heirs were similarly barred from assuming positions of territorial responsibility. In the nineteenth century, protracted civil strife would become a feature common to both polities.

All of these centralized organizations—the Luba and Ruund nuclear kingdoms, the “members states” of the Lunda “Commonwealth,” and the Luyana polity—partook of a political economy in which rights over people, including bonds of political loyalty, were predicated on the transfer of material goods.²⁵ Thus, a key function of rulers was to act as redistributors of resources—be they internal resources accumulated through tribute or, increasingly as we shall see, imported commodities resulting from participation in long-distance trading networks. On the upper Zambezi, for instance, the people of Bulozi, the floodplain heartland of the Luyana state, required such forest products as wood for canoes and bark for ropes. They, of course, also needed to secure access to the higher lands to which their villages were moved when the floods made the plain uninhabitable. The residents of the forest, on the other hand, turned to the floodplain for fish, cattle, and milk. This set of converging interests meant that the Luyana kings were strategically placed at the center of networks of accumulation and redistribution that ensured the circulation of the products of complementary ecological and economic zones. By controlling and manipulating such networks, the Luyana Litungas brought into being webs of dependencies and obligations that structured their power and broadcast it beyond the circle of their immediate followers.

Imposing, outward-looking capitals were essential ingredients of the mystique that surrounded most savanna kingdoms. Veritable instruments of rule, they functioned as tangible embodiments of royal power intended to impress their rulers’ subjects.²⁶ But the growth of these urban agglomerates was also clearly related to the tribute-gathering and redistributive functions fulfilled by the royal courts that they housed.²⁷ The Portuguese explorer Antonio Gamitto, for one, was very taken by the administrative sophistication that regulated social life in the capital of Mwata Kazembe IV Keleka in 1831–32. Going by his rough estimates, the population of what he deemed to be the “greatest town in Central Africa” is unlikely to have numbered less than ten thousand.²⁸

As already implied, political power in the central savanna also had important religious dimensions. Its holders were deemed responsible for ensuring the health and fertility of the land by propitiating their (or other relevant) ancestral spirits. A rule of thumb is that the more centralized the state, the more developed kingly cults revolving around royal ancestors were likely to be. Following their installation, peripheral allies of the Luba Mulopwes received a number of special insignia. Paramount among these was the white powder prepared by royal spirit mediums. Through potent symbols such as these, subordinate lineage heads “participated in the aura of Luba sacral kingship,” implicitly acknowledging “the superiority of that kingship over local concepts of chiefship venerated in their home villages.”²⁹ As for the Luyana state, the position of royal ancestors in its cosmology explains why, in 1886, Litunga Lewanika went to great trouble to persuade the missionary François Coillard to pay homage to the grave of his predecessor, Litunga Mulambwa Santulu. As convincingly argued by Gwyn Prins, this step was deemed necessary to establish the monarchy’s ritual superiority over the new arrivals, who, being “perceived by ordinary people . . . as magicians,” posed a threat to a kingly power that was itself infused with mystical attributes.³⁰

Contrary to what a host of modern scholars tend to imply, processes of ethnic consolidation were not solely a colonial phenomenon.³¹ While precolonial central Africa was most definitely not inhabited by discrete, impermeable, and mutually hostile collectivities (the “tribes” of colonial parlance), the fact is undeniable that the roots and some at least of the building blocks of present-day ethnicities are to be found in the regionally uneven processes of political integration with which this chapter has so far concerned itself. Thus, as Andrew Roberts pointed out several years ago, to be Bemba in precolonial times meant not only to speak chiBemba, but also to consider oneself a subject of the holders of the Chitimukulu, the probably Luba-derived title that, after becoming the preserve of one single lineage of the Bena Ngandu royal clan late in the eighteenth century, embarked on a process of sustained territorial expansion, gradually imposing its sway over much of present-day northeastern Zambia in the following century.³²

Centralizing, hierarchical state systems had thus undoubtedly come into being in the central savanna by the eighteenth century. Still, away from hubs of dynastic power, the lives of most central African peoples revolved around more fragmented, smaller-scale sociopolitical structures. On the frontiers of emerging state formations, political life was confined within the boundaries of the village, or, as in the case of the “stateless” Tonga of present-day southern Zambia, the ritual territory of rain priests. Alternatively, it gravitated around competing chiefly titles, the relationships between which were fluid and subject to frequent renegotiation. In the closing decades of the eighteenth century, this latter kind of political landscape was characteristic of the Chokwe and Luvale (or Luena) of the upper Zambezi and Kasai Rivers, and of the kiKaonde-speaking sub-clans to their east.

Lacking overarching centers of power, some of these communities found it problematic—or unnecessary—to mobilize resources on a large scale. This accounts for the comparative slowness with which they reacted to long-distance trading opportunities, or the fact that, when coastal traders did appear on the scene in the nineteenth century, small-scale societies—such as, for instance, the Ila and Lenje along the Kafue River—could end up being regarded as reservoirs of slaves to be raided at will. In contrary cases, however, well-developed hunting and forest-harvesting economies could actually facilitate and expedite involvement in the market economy. Thus it was that, beginning in the late eighteenth century, the highly mobile Luvale and Chokwe spawned aggressive market-oriented dynasties (as would, some decades later, the Kaonde) that were ready to take advantage of the new circumstances ushered in by the expansion of the Angolan slave and ivory trading frontier. Their polities, however, never entirely lost their original traits, consisting of unstable coalitions rather than centralizing kingdoms. Near the sources of the Zambezi, for instance, competition between the Chinyama, the Kakenge, and other Luvale titles remained an important factor throughout the nineteenth century.³³ Among the Luvale and similarly mobile groups, power—as a perceptive visiting missionary observed in 1895—was more “diffused” than in settled monarchical communities, such as the neighboring Lozi (the Luyana’s successor state).³⁴ Partly because of this, on the frontiers of large-scale state systems, ethnic cultures did not coincide with coherent political unities, and the influence of the colonial “creation of tribalism” would be profound and transformative.³⁵

THE IMPACT OF MERCHANT CAPITAL

The incorporation of the central savanna into globalizing trading networks in the nineteenth century was bound to have serious repercussions on the political landscape and patterns of governance described above. The increasingly pervasive impact of the outside world and the complex changes it gave rise to form the subject of this section. Merchant capital and its African representatives followed two broad axes of penetration into the central interior of the continent. Their convergence in the middle decades of the century turned the central savanna into an arena of accelerated economic and political exchanges. To understand how this came about, the development of both eastern and western itineraries and the activities of their chief protagonists need to be discussed in some detail.

The Indian Ocean Trading Frontier

Until the closing years of the eighteenth century, much of eastern and central Africa had been relatively isolated from the Indian Ocean coast, where Zanzibar was then emerging as a major commercial entrepôt under the leadership of the Bu Sa’idi dynasty of Oman. Indirect trading links, based on local and regional networks of exchanges, had certainly been in place, but these had not been of such magnitude as to transform the sociopolitical landscape of the interior. From the end of the eighteenth century, however, a growing international demand for east-central Africa’s ivory and slaves resulted in the forces of commerce gathering unprecedented momentum. During the first few decades of the nineteenth century, the bulk of the long-distance trade through present-day central Tanzania remained in the hands of African caravaneers—most notably the Nyamwezi and related groups, such as the Sumbwa (see map 1.2). It is probable that by the 1800s some Nyamwezi caravans had already begun trading with the kingdom of Kazembe, to which they were attracted not only by ivory and slaves, but also by the copper that the Mwata Kazembes received as tribute from their subordinates in southern Katanga.³⁶ Direct Nyamwezi trade with Katanga itself dates to the 1830s or so.

As the century progressed, better capitalized and better armed coastal merchants—normally referred to as Arab-Swahili, that is, residents of Zanzibar or the Islamized Swahili towns under its loose sway—began to encroach on Nyamwezi routes.³⁷ One of the effects of this competition was to increase the destructiveness of the trade and to push its frontiers further and further inland—a development facilitated by the establishment of Zanzibari bases in Unyamwezi country and at Ujiji, on the eastern shore of Lake Tanganyika, in the 1840s. By then, as attested by Gamitto in 1831–32, coastal merchants had already begun to visit the Lunda of the Mwata Kazembe, meeting part of the royal demand for cloth and other “ostentatious” possessions, including “forty very clean shot guns, and six hunting carbines wrapped in lace-trimmed cloth.”³⁸ Initially clearly deployed solely as charismatic symbols of royal wealth, the guns imported by Mwata Kazembe IV Keleka (“Chareka”) eventually left his Wunderkammern (cabinet of wonders): c. 1850, Keleka’s subjects—the Zanzibari trader Said ibn Habib reported—were “armed with muskets,” some of which must have been used for military purposes.³⁹ These muskets were almost certainly flintlocks—either cast-off “Brown Besses” or trade guns modeled on the latter that Zanzibar was then beginning to ship to the mainland in sizeable quantities. It has been estimated that the number of guns passing through Zanzibar rose from about five thousand per year in the 1840s to nearly a hundred thousand in the 1880s, by which time the Bu Sa’idi capital had become one of the “main conduits for small arms transfers” in the entire western Indian Ocean.⁴⁰ After visiting the lower Luapula Valley, Said ibn Habib proceeded to the “great copper mines” of Katanga—where “a great many people [were] employed” and whose product was “taken for sale all over the country”—and the “petty states” of the Ila (“Boira”) and Lenje (“Warengeh”).⁴¹ It was on the middle Kafue that Said ibn Habib first came across a party of long-distance traders from the far west: the Indian Ocean and Atlantic trading frontiers had intersected.⁴² Taken together, they were set to overlay the bulk of the central savanna.

MAP 1.2. Main nineteenth-century trade routes.

Further to the south, the ivory trade between Lake Malawi and the Portuguese colony on Mozambique Island, in northern Mozambique, had been in the hands of the Yao, then living predominantly to the south of the Ruvuma River and along the eastern shore of Lake Malawi, since c. 1700. In the latter part of the eighteenth century, as conditions of trade deteriorated on Mozambique Island, the Yao reoriented their activities towards the coast of southern Tanzania and, specifically, the ancient port of Kilwa, which had accepted Bu Sa’idi overrule in the 1740s. It was from about this time that the Yao, besides ivory, also began to deal extensively in slaves, a good number of whom were now acquired by French traders in Kilwa and imported into their Indian Ocean island colonies of Mauritius and Réunion. The Yao obtained some of their ivory from the Bisa, another group of semiprofessional traders who acted as intermediaries between the Lake Malawi region and political brokers further inland, such as, once again, the Mwata Kazembes of the lower Luapula Valley, with which the Bisa had been in contact since shortly after the foundation of the kingdom in about 1740. Eventually, in the middle part of the nineteenth century, the Bisa began to travel themselves all the way to Kilwa, where they offered their ivory at a cheaper price than the Yao did. Bisa competition in the ivory market—as Edward Alpers has suggested—further intensified Yao involvement in the slave business. The search for fresh supplies of slaves—many of whom were now deployed internally, on Arab- and Indian-owned plantations on the islands of Zanzibar and Pemba, and the Tanzanian coast opposite them—eventually led rival Yao gunmen to encroach manu militari upon Mang’anja territory along the southern shores of Lake Malawi and the heavily populated Shire Highlands of southern Malawi in the 1850s and 1860s.⁴³

As in the Nyamwezi case, the Yao sector, too, witnessed increasing coastal activity as the century progressed. By the middle of the nineteenth century, an Arab-Swahili base had been established at Nkhota Kota, among the Chewa near the southwestern shore of Lake Malawi.⁴⁴ In the late 1870s, a smaller coastal settlement also sprang up among the Senga of the upper Luangwa River. Both sets of traders fed the traffic in slaves along the Malawi-Kilwa caravan routes, where violence—as Livingstone famously reported—had become a pervasive fact of life from at least the 1860s.⁴⁵

Political relationships between Arab-Swahili traders and preexisting African authorities can be roughly subdivided into two phases. Since the vertical distribution of resources was an essential ingredient of their power, savanna kings and chiefs normally did their best to ensure that foreign trade took place under monopolistic conditions, or something that closely approximated them. Initially, given their precarious position, trailblazing coastal traders had little choice but to cooperate with domineering African authorities. It was only during a second phase, roughly beginning in the 1850s–1860s, that the Arab-Swahili resolutely moved against central African trading monopolies. This change of strategy was dictated both by their newfound military strength in the interior and by the increasingly competitive nature of the long-distance trading environment. The assumption of an explicitly political and military role on the part of coastal traders followed thereupon.

A watershed moment in the history of the region to the west of Lakes Tanganyika and Malawi was the defeat of the important Tabwa leader Chipioka Nsama I in 1867.⁴⁶ Chipioka, who some twenty years earlier had scattered an Arab-Swahili party, was attacked by Tippu Tip, the notorious Zanzibari trader and empire-builder, and his comparatively small but heavily armed following. In the encounter, Tabwa bowmen “died like birds! When the guns went off, two hundred were killed instantly and others were trampled to death! They fled. In one hour, more than a thousand died. Our casualties were only two slaves killed and two wounded.”⁴⁷ Following Nsama I’s defeat, his country entered a state of continuous civil war in which the Arab-Swahili—now stably settled between Lakes Tanganyika and Mweru—repeatedly played the role of kingmakers.⁴⁸ After the Tabwa, it was the eastern Lunda’s turn to bear the brunt of what Marcia Wright and Peter Lary aptly termed the “Swahili version of British gun-boat diplomacy.”⁴⁹ Controlled trade on the lower Luapula came to an end between the 1860s and 1870s, as coastal entrepreneurs worked systematically towards undermining royal monopolies and inaugurated an era characterized by successive foreign military interventions in the internal affairs of the much weakened kingdom.⁵⁰

The Atlantic Ocean Trading Frontier

Having examined the chronology and prime movers of the Indian Ocean trading frontier, let us now turn our attention to the network of trade routes that developed in the opposite direction: from west to east. Since the seventeenth century, the Portuguese coastal towns of Luanda and Benguela had been the main Angolan termini of a thriving slave trade. The itineraries flowing out of Benguela are especially important for the regions with which the next three chapters of this book are concerned.⁵¹ Throughout the eighteenth century, most of the slaves exported via Benguela came from its immediate hinterland and the Umbundu-speaking, politically divided, central Angolan highlands. In the second half of the century, Luso-African traders settled on the plateau, especially in the kingdoms of Bihe (Viye) and Mbailundu. It was these merchants and independent local operators who extended the Benguela trading complex in an easterly direction, eventually reaching the upper Zambezi River in the closing decades of the eighteenth century.⁵² At this time, the Luyana—as Livingstone was later to learn⁵³—refused to export the slaves on whose labor the political economy of their core area depended. More willing commercial partners were found among Chokwe and Luvale hunters, who quickly learned to minimize the technical deficiencies of the mainly Belgian-made trade muskets imported by Bihean and other traders, to appreciate their versatility as tools for the production of animal capital and human booty, and to draw on their gendered symbolism. The direct incorporation of the upper Zambezi into the Atlantic economy went hand in hand with the spread of both social insecurity and new opportunities for marketing the region’s primary products. These factors worked towards consolidating the sway of competing merchant chiefs. Internally divided Luvale trading principalities came into their own at a somewhat earlier date than the aforementioned Yao merchant dynasties in southern Malawi, but they obeyed similar market-driven logics and appropriated guns in similar fashion: as means of production, territorial expansion, and masculine affirmation. Slaves remained the upper Zambezi’s dominant export until at least the 1850s, though the rise in ivory prices at the coast from the mid-1830s meant that ivory (and wax) exports gradually grew in importance. Later, during the 1870s and 1880s, they would be joined by natural rubber.⁵⁴

In the mid-nineteenth century, Umbundu-speaking traders were probably the main carriers of the Angolan slave and ivory trade.⁵⁵ By then, their reach extended well beyond the upper Zambezi area. Having established solid relationships with the Kololo—Sotho migrants who temporarily overran the Luyana state in the middle decades of the century—the Ovimbundu trading (and, when circumstances permitted, raiding) sphere embraced the Ila and Lenje of the middle Kafue River and, eventually, the neighboring Kaonde of present-day Solwezi and Kasempa districts. Ovimbundu caravans—sometimes comprising several thousand free and enslaved porters (see figure 4.1)—also became very active in southern Congo. Since c. 1700, slave-dealing Ruund Mwant Yavs had been in contact with the Portuguese capital of Luanda through the mediation of the Imbangala of Kasanje.⁵⁶ For about a century, the rulers of Kasanje had been able to protect their middleman position. The bottleneck, however, had exploded early in the nineteenth century, as a result of the crisis of Kasanje.⁵⁷ From that point onwards, the Ruund capitals on the upper Mbuji-Mayi became a key destination for both Luso-Africans from the hinterland of Luanda and Ovimbundu from the Angolan plateau.⁵⁸ With the appearance of new actors on the scene, trade became less regimented, and the Mwant Yavs found it increasingly problematic to enforce such monopoly over foreign exchanges as they had enjoyed in earlier decades.⁵⁹

Later, the predicament of Ruund royals was compounded by the arrival of Chokwe migrants. Beginning as elephant hunters and, later, rubber gatherers, the Chokwe moved down the Kasai Valley and settled near the Ruund heartland.⁶⁰ Historically, the Ruund had been poorly disposed towards firearms. In the eighteenth century, according to Leitão, they had regarded them as a “handicap to valor.”⁶¹ One century later, guns were still scarce among the nuclear Ruund.⁶² This explains the ease with which musket-wielding Chokwe invaders carved out a dominant role for themselves during the internecine wars that marred the political life of the Ruund state after the death of Mwant Yav Muteb in 1873. By 1887, the year in which Carvalho visited the capital of a much enfeebled Mwant Yav Mukaz, the Ruund state was a shadow of its former self. Mudib, Mukaz’s predecessor, had been killed by his erstwhile Chokwe backers and his capital destroyed. On that occasion, Chokwe warlords had captured “more than 6,000 people”—adults, children, and, especially, women, whom they incorporated into their expanding matrilineages.⁶³ By the late 1880s, Chokwe raiders were the real masters of the Ruund heartland and had initiated a phase of indiscriminate slaving, “leaving a virtual desert where the Lunda empire had once stood.”⁶⁴ The southern members of the Lunda “Commonwealth,” too, struggled to adjust to the conditions of the frontier market economy. As will be further seen in the next chapter, the southern Lunda of the Shinde and others were preyed upon by heavily armed Luvale slavers throughout the better part of the nineteenth century.

The effects of international trade on the central Luba state were, if possible, even more pernicious. This was largely a question of timing. Since they had remained more or less unaffected by the long-distance trade until the 1870s, the Luba only experienced it in its most disruptive, late-nineteenth-century guise. From about 1870, the Ovimbundu had become the main trading partners of Garenganze, Msiri’s newly formed warlord state, the brutally extractive methods of which were then producing unprecedentedly large quantities of ivory and slaves for export. The Luba fish was hooked as a by-product of the commerce between Angola and Garenganze. Exploiting Luba internal divisions and lack of familiarity with foreign trade, Luso-Africans and Ovimbundu quickly established themselves as the dominant powerbrokers between the Lualaba and the Lomani Rivers. The civil wars that accompanied each royal succession became especially destructive, with different Luba factions drawing on the support of competing Angolan entrepreneurs. The early stages of this spiral of violence—one which would eventually result in the dissolution of the old Luba state as a cohesive territorial and political entity—were witnessed by Cameron in the mid-1870s. At the time of Cameron’s passage through the Luba heartland, the followers of Alvez—one of the traders with whom Mulopwe Kasongo Kalombo had allied himself with a view to defeating his many internal opponents—were given license to plunder the most vulnerable of the king’s subjects. “Any cultivated spot they at once fell on like a swarm of locusts, and, throwing down their loads, rooted up ground-nuts and sweet-potatoes, and laid waste fields of unripe corn, out of sheer wantonness.” Cameron was certain that, “had they not been armed with guns,” Alvez’s followers “would never have dared to act thus, for on entering countries where the people carried firearms these truculent ruffians became mild as sucking doves.”⁶⁵

Between Political Innovation and Continuity

The interaction between the forces of global commerce and preexisting authorities forms one of the master themes of the history of the central savanna in the nineteenth century. The political effects of this increasingly close connection with the outside world, as I am presently going to argue, were profoundly ambivalent and contradictory. To be sure—as has already been noted—central African ruling elites often struggled to control the dynamics unleashed by the onset of the long-distance trade in ivory and slaves and the militarization of social relationships that it precipitated. Among the central Luba and Ruund and in most Lunda states, the erosion of royal monopolies over foreign commerce and the distribution of imported commodities led to fragmentation, increased violence, and enhanced social differentiation and levels of slave exploitation.⁶⁶ Under these circumstances, the growth of international trade resulted in the weakening of long-established elites, whose dominance was now challenged by the rise of “new men,”⁶⁷ often initially installed on the already contracting peripheries of the old political formations.

In a number of cases, these efforts at state-building in a period of widespread turmoil were entirely shorn of traditional legitimacy. The aforementioned Yeke state of Msiri, the latter a Sumbwa caravan leader turned empire-builder in the region lying between the collapsing Luba and Ruund states and the shrinking eastern Lunda sphere of control, is an especially clear example of the spread of warlordism in the central savanna. Better equipped to face the trials of the era of large-scale trading than were the old Luba and Ruund/Lunda aristocracies, Msiri and numerous other political opportunists—not least Lusinga, in northern Katanga, whose tragic story has recently been masterly told by Allen Roberts⁶⁸—gave birth to violently entrepreneurial conquest states, sometimes ephemeral, sometimes more enduring. In these new political organizations, power rested less on religious sanction, heredity, and redistribution than on their leaders’ personal achievements, successful involvement in commerce, and preparedness to resort to sheer violence to achieve their aims. The diffusion of firearms, and their recasting into primary means of military and economic domination, were often central to these processes of political realignment, which also drew part of their impetus from the related emergence of semiprofessional and cosmopolitan standing armies of brutalized young men.⁶⁹

Although these developments were especially sudden in the interior of central Africa in this era of long-distance trade, they were by no means unique to it. The coastal societies of Angola, large-scale exporters of slaves from a very early period, had already witnessed the overthrow of established authorities and the rise of heavily militarized polities from the seventeenth century onwards.⁷⁰ Northeastern Tanzania offers a more proximate example. There, control of rain medicine and kinship relations, and the redistribution of internal tribute in livestock and labor, were the most important weapons in the political arsenal of the Kilindi rulers of Shambaa. In the closing decades of the nineteenth century, however, these ways of relating rulers to people were largely superseded by a new political culture. Revolving around Semboja, chief of Mazinde, closer to long-distance routes than Shambaai, the mountain heartland of the kingdom, such culture drew on trading connections and the slave-gun cycle to pose an ultimately unanswerable challenge to old Kilindi politics. A drawn-out civil war and a “complete victory for the forces of decentralization” were the end result. In Shambaa political thought, kingly power had always been “ambivalent. It could be used to bring life or to bring death.” In the Pangani Valley in the third quarter of the nineteenth century, death became dominant, as the monarchical order ceased to be the guarantee of fertility, unity, and stability. Political leaders began to systematically prey upon their subjects, who were often left with no choice but to seek some degree of protection by joining the ranks of the slave raiders and warlords themselves.⁷¹

Still, apocalyptic descriptions of the central savanna on the eve of the European conquest should be rejected.⁷² Revolutionary transformation was not on the cards everywhere, and the continuities in political tradition should not be underestimated. After all, as Ian Phimister has pointed out with reference to late nineteenth-century Zimbabwe, merchant capital sometimes “[modifies] existing social relations without decisively altering them.”⁷³ Centrifugal forces and/or the onslaught of the new entrepreneurs of violence were sometimes effectively resisted. When this was the case, preexisting forms of authority and governance could actually be strengthened by participation in long-distance trading networks. The Bemba and, especially, the Lozi polities exemplify such processes. Unlike the neighboring Tabwa and eastern Lunda, the Bemba kept both the Arab-Swahili of Itabwa and the so-called “Senga Arabs” at bay. As a result, while the overall cohesiveness of the Bemba “federation” might not have increased greatly, each Bena Ngandu chief, dealing on his own terms with coastal traders, used imported commodities to augment his “ability to attract and reward supporters.”⁷⁴ As will be seen in the next chapter, after overthrowing the Kololo in the 1864, the Lozi maintained their erstwhile conquerors’ open-door policy vis-à-vis foreign traders, but reverted to being a slave-importing—as opposed to slave-exporting—society. This, and Litungas Sipopa and Lewanika’s skills in retaining ultimate control over foreign trade and the circulation of charismatic goods—firearms included—that came with it, help explain why the kingdom did not experience the same decline as the Ruund and Luba states underwent, and why it would eventually negotiate its incorporation into emerging colonial structures at the end of the century from a position of comparative strength.

FURTHER CHALLENGES FROM THE SOUTH

A significant role in extending the sway of merchant capital over parts of the central savanna was also played by Portuguese settlements in Zambezia, in the lower Zambezi Valley in present-day central Mozambique. In the eighteenth century, perhaps the majority of the slaves obtained by the lower Zambezi’s prazeiros—the Africanized descendants of the land-grant holders originally recognized by the Portuguese Crown in the seventeenth century—were retained on their estates (prazos) and put to productive uses.⁷⁵ Some of their number, however, were organized into standing armies and entrusted with the task of policing the sprawling concessions and enforcing the subjugation of both their free and unfree cultivators. Occupying a most ambiguous social location, armed slaves, known locally as chikunda, were both “the objects of domination and the means by which the prazeiros controlled the peasantry and accumulated wealth.”⁷⁶ In time, the chikunda—who also worked for their owners as elephant hunters, slave raiders, and long-distance traders—gave birth to a distinctive ethnic identity, one structured around such cultural markers as a “disdain for agricultural labor”—which the Chikunda construed as the preserve of women, common slaves, and subjugated peasants—and the glorification of masculine, martial pursuits such as warfare and hunting, with the notable aid of imported and partly homemade firearms.⁷⁷

Slave exports from Quelimane boomed early in the nineteenth century, when the port to the north of the mouth of the Zambezi began to attract Brazilian and other ships bent on escaping abolitionist efforts along the Atlantic coast of Africa.⁷⁸ The intensification of the southern slave trade meant that the Chikunda were no longer regarded as members of a comparatively privileged slave community, but, increasingly, as mere chattel for export. The Chikunda reacted violently to the new state of affairs, and slave insurrections and large-scale flights became the order of the day on the lower Zambezi. By the middle of the century, the prazo system had completely collapsed. Its fall resulted in the de facto emancipation of thousands of former armed slaves.⁷⁹ As Allen and Barbara Isaacman have expertly shown,⁸⁰ newly freed Chikunda had a number of options open to them in the new circumstances. Some continued to operate within the Portuguese sphere, working as professional hunters and porters for Luso-African traders (collectively known as muzungus). Other Chikunda—whom the Isaacmans refer to as “transfrontiersmen”—moved permanently away from the lower Zambezi. While some of these migrants ended up selling their military and hunting skills to vulnerable Chewa, Nsenga, and Gwembe Tonga communities, other Chikunda turned into state builders. Led by muzungu adventurers, Chikunda polities sprang up near the Luangwa-Zambezi confluence from the 1860s. Some at least of these new formations—especially those of the warlords Kanyemba and Matakenya—were veritable conquest states, resembling in many respects the creations of Msiri and other Congolese warlords. The products of the slave and ivory frontiers, Chikunda warlord states resorted to large-scale raiding and taxed local inhabitants mercilessly. With guns being deployed as their principal tools of commodity production, Chikunda conquerors made a significant contribution to regional instability, laying waste to large areas of present-day southern, central, and eastern Zambia.

Besides contending with the forces of merchant capital, the societies of the central savanna also had to deal with the extensive ripples of the South African Mfecane. Several reasons have been adduced to explain the turmoil that affected present-day KwaZulu-Natal and neighboring areas early in the nineteenth century. It has been argued, for instance, that conflict over scarce resources increased from about 1750 as a result of either demographic growth, climatic change, or both. Other scholars have preferred to relate competition among northern Nguni-speakers to the expansion of the trade flowing out of Portuguese-dominated Delagoa Bay in southern Mozambique.⁸¹ Whatever its ultimate causes, increasing tension in KwaZulu-Natal precipitated processes of centralization and enlargement of political scale. Among the northern Nguni, state formation took one very specific course. Age sets, or amabutho, had been a distinctive feature of social organization in the area for decades, if not centuries. In their original form, the amabutho consisted of groupings of young men “brought together by chiefs for short periods to be taken through the rites of circumcision and perhaps to engage in certain services, such as hunting.”⁸² In the deteriorating political landscape of the late eighteenth century, local leaders transformed these age sets into labor and war regiments. So reconfigured, the amabutho took on the character of standing armies: they were both a consequence and a cause of the increasing level of militarism obtaining among the northern Nguni.

By the late eighteenth century, a number of opposing power blocs—the principal of which were the Ngwane, the Ndwandwe, and the Mthetwa—had emerged. Conflict between them eventually span out of control, precipitating the so-called Mfecane, a series of wars and migrations that transformed the sociopolitical landscape not only of KwaZulu-Natal, but of the broader southern African region as well. Among the protagonists of the turmoil were Sebitwane—the Sotho-speaking chief of the aforementioned Kololo, the migrant group who conquered and ruled the upper Zambezi floodplain and the Caprivi Strip between the early 1840s and the Luyana/Lozi reconquista of 1864—and Zwangendaba Jere, the war leader to whom the principal Ngoni groups of present-day eastern Zambia and Malawi trace their origin. Coming into their own in the second half of the nineteenth century, the conquering Ngoni kingdoms spawned by the Mfecane affected roughly the same areas as those into which Chikunda “transfrontiersmen” were then expanding. As will be seen in chapter 5, the Ngoni approach both to international commerce and to firearms, its most fundamental of by-products, differed from that of the Chikunda. But, even though the Ngoni’s military preparedness had nothing to do with access to firearms, and was rather the result of the “meritocratic” aspects inherent in Ngoni age-set regiment systems, its effects undoubtedly magnified the violence and insecurity that accompanied the inland advance of the frontier of merchant capital.

. . .

The primary aim of this chapter was to supply the reader with enough background data to engage with the chapters that follow. Given their imbrication in large-scale sociopolitical developments, firearms have already repeatedly cropped up in the discussion, providing some indication of the varied reactions they elicited, and the different meanings and functions attributed to them, across the central savanna. It is now time to explore these reactions and productions more systematically.