Something New out of Africa: States Made Slaves, Slaves Made States

Abstract

In this article I explain a nexus between slavery and state formation in Africa, proceeding from initial demographic and institutional conditions to an external demand shift, individual state responses, and their collective systemic consequences. Historically, African rulers faced distinctive challenges: low population density prioritized control of people more than territory, and internal disintegration was often a greater threat than external conquest. A massive expansion in the demand for slaves offered African rulers increased opportunities to use external resources for “outside-in” state building. Many did so by creating highly militarized predatory slaving states. The collective consequence was heightened systemic insecurity. Variation in the timing of these developments reflected regional and historical variation in the expansion of the demand for slaves. Slaving states appeared first in West Africa, reflecting the late-seventeenth-century expansion of the trans-Atlantic slave trade, before spreading to East Africa a century later, following the parallel later increase in the Indian Ocean slave trade. This “outside-in” path to state formation both parallels and contrasts with contemporary postcolonial state formation.

Between 1500 and 1850, 12 million enslaved Africans were brought to the Americas (while 2 to 3 million Europeans made the journey voluntarily),¹ with vital political legacies that endure to this day.² The slave trade both from and within Africa was highly consequential for both state formation and its systemic consequences. The argument presented here is based on a four-stage explanation which moves from initial conditions to an external demand shift, to strategic responses to this shift, and finally to the interactive systemic consequences of these strategies. It is built on a foundation of highly insightful scholarship by area specialists and historians of Africa. Rather than being aimed at Africanists, however, the “something new” referenced in the title is for scholars of international politics and state formation in advancing knowledge of international relations.

Drawing closely on a starting point provided by Herbst,³ the first stage of the explanation is that African rulers tended to face different challenges from those in Europe: low population density meant that control of scarce people was generally more difficult and more important than control of abundant land. Fission via the defection of vassals and tributary sub-rulers, and even whole populations simply decamping, were often greater threats to these loose, composite African states than external territorial conquest. Thus, as in many contemporary states in the developing world, anarchy was often more a “domestic” than an “international” problem.

Second, although slaves and long-distance slave trading existed in Africa for centuries before European contact, from the late 1600s there was a massive expansion of trans-oceanic slave trade associated with the rise of the New World plantation economy. The heightened demand for slaves continued even after the abolition of the Atlantic slave trade from the time of the Napoleonic Wars, with the number of enslaved Africans peaking only in the late nineteenth century. The increase in demand for slaves was at first limited to Atlantic Africa, but from the early nineteenth century it spread to Indian Ocean Africa. The export of slaves gave way to the export of slave-produced commodities, but this maintained the elevated demand for slaves.

Third, as a result of this demand shift, the opportunity to amass new external resources via the slave trade provoked the rise of predatory states based on the militarized production of slaves. Expanded demand increased the returns to slave raiding for export, relative to other economic activities. These states followed an “outside-in” logic of state building. Many African rulers mobilized external resources gained from supplying the expanded demand for slaves to counter internal threats to their fissiparous domains. This is in contrast to the conventional “inside-out” state-building path, whereby rulers mobilized internal resources to counter external threats (conquest by foreign powers). Specifically, these African rulers and would-be rulers used the increased returns from predatory violence to build armies and states—directly through conscripting slave soldiers and the trade for guns, and indirectly by using revenues and other foreign trade goods to pay off subordinates. Reflecting the fact that the expansion of the slave trade began earlier in Atlantic Africa than in East Africa, slaving states first arose in the early modern period in the former, but only in the nineteenth century in the latter.

Fourth, increased competition for captives in West and later East Africa tended to bring these predatory states into systemic, fiercely destructive military competition and exacerbated their brittleness. If one state grew more powerful by tapping the new opportunities offered by the expanded slave trade, this posed a security dilemma for neighboring states. Because such states depended on a steady supply of slaves for their military and economic reproduction, internal cohesion, and geopolitical survival, interruptions in the supply of slaves often started a self-reinforcing cycle of disintegration. Thus the slaving states were inherently brittle and unstable. Jointly, therefore, these states constituted a violent and unstable international system characterized by acute security dilemmas and negative-sum competition, and fundamentally dependent on external trade in slaves and slave-produced goods.

What new light does this story shed on general theories of state building and international politics? In Europe the conventional understanding is that external military pressures and Darwinian interstate competition forced rulers to mobilize internal resources via state formation, setting up a self-reinforcing cycle.⁴ In Africa, by contrast, it was external resources, those obtained in exchange for slaves as commodities and their production, that were increasingly essential for countering internal obstacles to state formation. As a result of the combination of long-standing internal weaknesses, heightened security competition, and external resource dependence, African slaving states, despite their highly militarized character, were more brittle and unstable than their European counterparts. Collectively, these features rendered the continent vulnerable to imperial conquest and colonization.

Turning to contemporary and global implications, a thread of continuity between pre- and postcolonial state formation within and beyond Africa is the degree to which internal threats are often more dangerous than external ones, and to which state formation depends on external resources. To this extent, Africa's early historical experience with outside-in or second-image reversed state formation is preconsciously modern. In contrast to the slave-based examples discussed in this paper, however, the relatively benign environment of contemporary international politics allows brittle postcolonial states to endure. Evidence here demonstrates not only that outside-in state formation can occur in insecure international environments, but also that the interaction of such strategies can accentuate this insecurity.

Variation within Africa, however, is as important as these general trends. The examples of Dahomey and Segu in West Africa demonstrate how the massive expansion in the slave trade stimulated the rise of slaving states in the eighteenth century. In East Africa, the later rise of the Omani Swahili and other slaving states is linked with the nineteenth-century expansion of the Indian Ocean slave trade. Dahomey and the Omanis largely managed both the systemic security pressures and the transition in the nature of the slave economy until they were conquered by European imperialists. In contrast, Segu and others succumbed to the combination of external pressure and the inherent vulnerabilities of this form. Throughout this period, the multitude of stateless societies in Africa were also often crucially shaped by the slave trade, though their experience lies outside the scope of this paper.⁵

The argument here follows conventional usage in adopting Charles Tilly's definition of states as “coercion-wielding organizations that are distinct from households and kinship groups and exercise clear priority in some respects over all other organizations within substantial territories.”⁶

Initial Conditions for State Building in Precolonial Africa

A fundamental difference between the theoretical presumption of international relations scholarship and African experience was that for long periods of African precolonial history anarchy was more a “domestic” problem than an “international” one. Rather than there being a domestic sphere of authority constituted by vertical chains of command, and a threatening outside environment of unrestrained competition, area specialists and historians have shown that the main ordering problem was often at home, where loose domains were held together (to the extent that they were held together) by lateral ties. Though external threats certainly existed, until the rise of the slaving states, international anarchy was relatively less threatening because of the reduced salience of territorial conquest. Precolonial Africa was no less violent than Europe. But “internal” warfare, succession and secession struggles, rebellions and civil wars, were more common and more dangerous than international wars of conquest.

Building on the pioneering work of area specialists, I begin by explaining the scarcity and difficulty of controlling people, and the fissiparous nature of African states.⁷ In general (and relative to their European peers), African states were composite entities, depending on indirect rule, with fuzzy frontiers and overlapping, heteronomous authority claims.⁸ The idea that internal threats of disintegration commonly dominated external threats of conquest is then supported with four brief examples from different periods and regions of sub-Saharan Africa involving a mix of actors: along the Niger River just south of the Sahara in the late sixteenth century; in Angola-Congo slightly later; along the Zambezi River in East Africa c. 1600–1800; and Southern Africa in the nineteenth century. The point of choosing such widely dispersed examples is not to suggest that these trends were unvarying universals in Africa but to argue that they were common, recurring patterns that reflected similar underlying drivers.

The link between war-making and the development and reproduction of political structures was as intimate in Africa as it was in Europe, but took a very different form. Some of the most important differences stemmed from the low population density of Africa compared with Europe.⁹ Because land was relatively plentiful but people relatively scarce,¹⁰ wars were more commonly about control of the latter than the former.¹¹

Because there were few fixed investments in land, territory was often not worth defending, and so frontiers were usually loose zones.¹² As Herbst puts it: “Precolonial African states therefore had precisely the opposite physiology of many in Europe: the power assets were concentrated in the center with gradations of authority extending to the hinterland. The European model of placing significant assets in the hinterland to protect against outsiders and to make the boundaries real was neither viable nor relevant.”¹³ The lower priority placed on control of territory also explains African rulers’ lack of concern about the European outposts scattered along the coast from c. 1500. These rulers could have expelled the Europeans, and sometimes did when it suited their interests.¹⁴

Internally, authority often radiated out in a series of concentric circles from a political core. Authority over sub-rulers became increasingly attenuated as the distance from the core increased, with vassals and suzerain relationships giving way to the occasional payment of tribute.¹⁵ Especially in areas where transport was possible only by foot, the difficulty of communication and control forced a high degree of reliance on various forms of indirect rule.¹⁶ In this regard, African states were akin to Nexon and Wright's conception of empires as rimless hub-and-spoke systems, rather than fitting the conventional understanding of the sharply bordered sovereign state.¹⁷ As one historian puts it, African states “resembled the spokes of a wheel, whereby the center delegated power to provincial governors who were frequently military overlords with more or less direct links to the kingship.”¹⁸

The reliance on autonomous intermediary rulers, and the mobility of populations, gave many African states a fissiparous character. Nominally subordinate rulers often politically or physically distanced themselves from their erstwhile overlords, with larger kingdoms often fracturing or dissolving as a result.¹⁹ Whereas in Europe Elias argues that a competitive “monopoly mechanism” and a “royal mechanism” worked to both centralize the control of violence and bind social groups increasingly closely together within larger units,²⁰ African states were much more likely to retain a composite character. Central rule was a relatively thin and fragile overlay covering subunits that were often the primary focus of loyalty and authority.²¹ Attempts to overcome this fragility could exacerbate it, as centralizing moves provoked resistance and exit from subunits.

Even operating at the highest level of generality, it is important to acknowledge the diversity of political forms in precolonial Africa.²² In particular, perhaps half the African population lived in thousands of stateless societies with populations numbering only in the hundreds.²³ These communities interacted closely with larger counterparts over the centuries, but until the dawn of the twentieth century, neither military nor economic dynamics produced their extinction.

Conquest Versus Entropy: Things Fall Apart

How did these challenges of state building play out in practice? The brief examples that follow demonstrate how would-be conquerors were stymied by the tendency to fission among African states. Though there were foreign invasions, rather than outside security threats entrenching internal state formation, as in the conventional understanding of European dynamics, disintegration proved to be the undoing of the invaders. The examples illustrate this tendency in West, littoral Atlantic, East, and Southern Africa from the late sixteenth to the nineteenth century, and include both local powers and Europeans in Africa.

The most consequential early modern outside military intervention in sub-Saharan Africa was carried out not by Europeans but by the Moroccan sultanate. In 1590 Sultan Ahmad al-Mansur sent an army south across the Sahara to conquer the Songhay kingdom that had arisen along the Niger River in what are now Mali and Niger.²⁴ Yet the Songhay kingdom had in effect unraveled before the Moroccans won the climactic battle in 1591, as Songhay's frontier vassals and tributaries had dropped their allegiance to the center. The Moroccan victors were themselves undone by the same dynamic in their subsequent, futile twenty-year effort to pacify and consolidate their Sahelian conquest. The local Moroccan military leaders in the former Songhay territories became increasingly detached from their distant nominal superiors north of the Sahara. They began to assert their independence, creating—together with the remaining fragments of the former Songhay domain—numerous competing slave-raiding ventures.²⁵ Political entropy prevented both empire building and state consolidation.

Europeans’ ambitions were frustrated in the same manner. The initial establishment of Portuguese Angola from 1575 depended on local tributary rulers’ switching their allegiance from their former African overlords to the Portuguese. But the latter found that these local rulers were prone to shift their allegiance back to the African kings after Portuguese defeats in Kongo or at the hands of the Dutch in the 1640s.²⁶ Other ostensible Portuguese subordinates stayed neutral, waiting to see which way the political winds blew.²⁷ The Portuguese in Angola had acquired allies of convenience, not conquered subordinates.²⁸ For both European and African overlords, the loyalty of nominally subordinate rulers was easily given but even more easily withdrawn, either to be transferred to a new patron or exercised independently.²⁹ For this reason, “until the nineteenth century, the Portuguese colony of Angola consisted of two coastal settlements, three small forts, a tenuous hold on a hundred miles of river, and little else.”³⁰

Evidencing the same dynamic, Portuguese principals on the Indian Ocean coast of East Africa found that their control over agents they had granted estates in the hinterland of the Zambezi River Valley (in what is now Mozambique) became progressively attenuated. From the 1600s these agents hybridized with local African ruling families, forming a new creole group of prazeiros (estate-holders, from prazo, estate). Obedience to the Portuguese authorities became a polite fiction, and later even this pretense was dropped, along with Christianity, European dress and family relations, and Portuguese language and literacy.³¹ Well into the nineteenth century, these prazeiros ruled their own domains, raided for ivory and slaves, and became imbricated in the various overlapping, heteronomous authority relationships of their African peers.³² They also maintained their own slave armies, the chikunda, who, with the expansion of the East African slave trade in the early nineteenth century, became important political actors in their own right.³³

Even when the populations in question remained predominantly of European origin, groups unhappy with their current political arrangements could simply vote with their feet and exit.³⁴ The tendency for things to fall apart politically was thus not due to any African cultural essentialism. The most prominent example is the Boers of Southern Africa. In the eighteenth and nineteenth centuries, the Boers trekked into the African hinterland to escape the control of first the Dutch East India Company and then the British Crown. Aside from economic and linguistic factors, one particular point of friction was the British emancipation of the Boers’ slaves.³⁵ Having exited from their previous political ties, the Boers formed a range of loosely allied republics that engaged in war and diplomacy with local African rulers.³⁶ In this regard, the Boers were adopting the same exit option that had been common in Africa for centuries.³⁷

Thus the Moroccan, Portuguese, and Boer examples demonstrate the problem of fission and the obstacles to state building. They also show that when Africans and Europeans interacted in the organization of rule and authority, Europeans became Africanized much more than the other way around. Rather than marking some primitive baseline soon superseded by a teleology of European-style state formation, the tendency to fragmentation and fission remained pervasive.³⁸

The Expansion of Slavery

Slavery, the slave trade, and slaves as a means or end of warfare were all present in Africa before Europeans arrived.³⁹ With inevitable exceptions, African property regimes over land were commonly weak or absent, but property regimes over people tended to be stronger and more developed.⁴⁰ Slaves were commonly a goal of war.⁴¹ In Muslim Africa in particular, from medieval times, armies were often built from slave soldiers.⁴² In a pattern recognized by scholars as early as Weber, the challenges of segmented, decentralized states with strong and unruly vassals meant that a ruler's slave retainers often constituted the most reliable cadre of political and military followers.⁴³

Thus slaving was a recurrent theme of war and state building in parts of early modern Central and East Africa.⁴⁴ In the sixteenth and seventeenth centuries the king of Kongo depended on slaves for his centralizing efforts vis-à-vis the nobility.⁴⁵ In Ethiopia slaves were an important war aim, and in both Ethiopia and Morocco slave soldiers were important to centralizing rulers exerting power over the armies of their over-mighty nobles.⁴⁶ It is important to acknowledge variation, however. In many other areas of Africa, particularly in stateless societies, slaves were relatively scarce, and they were closely integrated into domestic units, working at home or in the fields alongside their owners.⁴⁷ In some circumstances children of slaves were born free, whereas elsewhere slavery was intergenerational.⁴⁸

The long-distance slave trade had a significant role in some instances of African state-making and war making long before Europeans arrived, especially across the Sahara. Along with gold, slaves were one of the most important sources of income from the long-distance trans-Saharan trade routes that underpinned the Niger River kingdoms of medieval Mali and later Songhay.⁴⁹ The East African slave trade dated back to around the eighth century.⁵⁰ In most other areas of Africa, however, slave trading was much more limited and local before European contact.

For centuries after first contact, Europeans in Africa largely had to conduct their relations according to African rules, institutions, and practices when it came to diplomacy and trade, and the slave trade in particular. As late as 1876 less than 10 percent of the continent was under European rule.⁵¹ Because the relations that are my focus here are not the familiar ones of European domination and non-Western submission and subaltern resistance, they suggest a new and important perspective. This confounds deeply held beliefs about the relationship between Africa and the West, which consistently assume the former is the passive victim and the latter is the dominant party.⁵² For example, dependency theory and world-systems theory see Africa as epitomizing the exploited periphery preyed on by the European core.⁵³ In contrast, the argument here is that European demand for slaves presented many African rulers with new options for achieving their own goals, even if the collective consequences turned out to be disastrous.

From as early as the fifteenth century, foreign slaves captured in war were one of the main commodities exchanged by Africans in return for European goods.⁵⁴ Africans were discerning and assertive traders.⁵⁵ Historian John Thornton's uncompromising verdict is that for the whole of the precolonial period

Africans controlled the nature of their interactions with Europe. Europeans did not possess the military power to force Africans to participate in any type of trade in which their leaders did not wish to engage. Therefore all African trade in the Atlantic, including the slave trade, had to be voluntary.⁵⁶

Thus from the beginning to the end of the slave trade, Africans dominated arrangements up to the water's edge, especially the supply and contacting arrangements.⁵⁷

Even the European supply of guns, one of the main goods traded for slaves, brought Europeans little influence over local African rulers, thanks to fierce competition between the suppliers. As one Dutch trader complained in 1703:

Perhaps you wonder how the Negroes come to be furnished with fire-arms, but you will have no reason when you know we sell them incredible quantities, thereby obliging them with a knife to cut our own throats. But we are forced to do it; for if we would not, they might be sufficiently stored with that commodity by the English, Danes, and Brandenburghers; and could we all agree together not to sell them any, the English and Zeeland interlopers would abundantly furnish them.⁵⁸

This undercuts the earlier historical view of a “gun–slave cycle,” whereby Africans were supposedly forced into the slave trade and internecine warfare at the behest of Europeans because European controlled the supply of firearms.⁵⁹

From 1600 to 1650 the average number of slaves trafficked across the Atlantic fluctuated around 10,000 annually. It then rose to 25,000 per year in the second half of the century, and then more than 50,000 per year from 1700.⁶⁰ From 1500 to 1650, 703,890 slaves were taken across the Atlantic, compared with 2,900,785 in 1651 to 1750, and 6,808,535 in 1751 to 1850. This massive increase reflected the growth of the plantation economy in the New World.⁶¹ Plantation crops, especially sugar, tobacco, and cotton, required a vast slave labor force. Due to death in transit, horrific maltreatment, disease, and a lack of natural reproduction, maintaining the slave population of the Americas required a constant stream of new victims from Africa. It took decades for the Royal Navy to effectively enforce the prohibition on the slave trade announced in 1807, with 2.8 million slaves transported across the Atlantic after this date.⁶² Abolition in the Indian Ocean took even longer.

Significantly, the impact of the expansion of the slave trade varied across the continent. This difference in timing between West Africa and East Africa had crucial implications for the timing of political change within and between states in each region. Not surprisingly, given its closer proximity to the main source of demand for slaves in the Americas, Atlantic Africa was drawn into the expanded trade before East Africa.

Although statistics are much less complete than for the trans-Atlantic trade, historians are nearly unanimous that the late eighteenth and early nineteenth centuries saw massive expansion of enslavement in East Africa, in response to demand from several new sources.⁶³ Around 2 million slaves were taken across the Sahara, the Red Sea, and the Indian Ocean in the nineteenth century alone.⁶⁴ Brazilian and European slavers catering to the American plantations increasingly purchased slaves from Southeast Africa,⁶⁵ as British naval anti-slavery patrols in the Indian Ocean began only in the 1860s.⁶⁶ While 70,900 slaves were sent across the Atlantic from East Africa in the whole of the eighteenth century, in the first half of the following century the total was 407,000.⁶⁷ The establishment of plantation agriculture on the islands of Mauritius and Reunion meant that another 442,000 slaves were shipped to these and other destinations in the Indian Ocean in the nineteenth century.⁶⁸ As the century progressed, a scramble for slaves developed, with slave-raiders venturing ever further inland.⁶⁹ As in West Africa, war and raiding were the most common sources of slaves.⁷⁰

Aside from exporting people to work in the New World, the other major driver of continued high demand even after the end of slavery in the Americas was the increasing use of slaves in African export agriculture. A terrible irony of abolition (a drawn-out process that took place in multiple stages) is that a concurrent reduction in the price of slaves in many parts of Africa, combined with the sharply increased demand from industrializing Europe and North America for crops like palm oil, cloves, peanuts, and cotton, gave rise to massive growth in slave plantation agriculture in Africa.⁷¹ As in the examples of Dahomey and the Omani domains discussed later, the systems designed to produce slaves for export came to be used to produce slaves for local plantation labor. The demand for slaves actually increased through most of the nineteenth century, such that by the mid-1800s there were more slaves in Africa than there ever had been in the Americas.⁷²

If people were scarce in Africa, why did African rulers export them as slaves? One view is that because African labor was scarce and productivity relatively low (due to lack of capital and to environmental factors) it was expensive, making slavery for export and other coercive solutions more economical for rulers than wage labor.⁷³ Reinforcing this tendency to coercive solutions was the problem of mobile populations discussed already: taxing peasants was often a problem of hitting a moving target. Relating to the export of slaves, others argue that a deeper logic was at work: European demand for slaves “encouraged the creation of institutions that favored the extraction rather than the creation of wealth.”⁷⁴

To present these institutional and systemic logics, the remainder of the article explains how the opportunities for outside-in state building provided by the expanded slave trade could lead to the rise of slaving states and to more acute systemic security competition. These two logics are first laid out in general terms, based on historians’ characterizations of common patterns and recurring dynamics. These general patterns and the difference in timing between West and East Africa are then substantiated with more specific evidence. In West Africa, Dahomey on the Atlantic littoral and Segu in the Niger River Valley illustrate these trends from the late seventeenth century onward. In East Africa, the Zanzibar-based Omani domain and the formally Portuguese prazeiro estates in what is now Mozambique illustrate parallel developments in the nineteenth century. As a group, and in their external orientation and militarized character, these states are characterized by crucial “family resemblances,” being defined by “attributes that they share to varying degrees, as contrasted to non-family members who may share few of them.”⁷⁵ The cases that follow exemplify the sequences and mechanisms of the argument, conforming to the logic of within-case analysis of “causal process observations” that defines comparative historical social science.⁷⁶

Slaving States

The slave trades from Africa were the most important trans-oceanic movements of people before the twentieth century.⁷⁷ African slaving states were the indirect products of this flow. (Historians sometimes use cognate terms to describe the same class of states, such as “predatory states,”⁷⁸ “warrior states,”⁷⁹ or “warlordism.”⁸⁰) The potential resources to be gained by feeding this increased demand presented rulers with new options for predatory state-building strategies. The interaction of these individual strategies had an unintended systemic effect: sharper security competition, which undermined states’ survival prospects. Thus, overall, an external demand shift gave rise to new individual state-building strategies in response to long-standing local conditions, the unintended collective consequences of which rendered individual states more vulnerable.

The slave trade was inherently based on violence; supplying more slaves required more violence. Because of the scale of the trade—millions of individuals, over centuries—this violence was institutionalized.⁸¹ Klein speaks of the relationship between the rising appetite for slaves and the rise of a new type of predatory state: “This demand for slave labor called into existence partners who could provide a large number of slaves … The most visible development was the appearance of a series of military states, which used war and raiding to supply larger and larger numbers of slaves.”⁸² Reid supports this same conclusion: “The violence of the slave trade drove the formation … [and] dramatic proliferation of such highly militarized states and communities.”⁸³ Robinson also argues that states adapted themselves toward military predation to secure slaves for external markets.⁸⁴

But there was important regional variation in the timing of this pattern, which first appeared in West Africa from the late 1600s, and then in East Africa in the nineteenth century. Green holds that “there was a direct connection between the globalization of West Africa in the era of the slave trade and changes to the structures of political power.”⁸⁵ As Reid puts it, “The general tendency was towards the militarization of state and society in the Atlantic zone, largely because the supply side of this new economic system [the slave trade] … required militarization and the regular deployment of armies in slave-gathering for export.”⁸⁶ He notes the same trend in East Africa in the nineteenth century, once more because of international trade: “The expansion of the slave and ivory trades from the early nineteenth century onward prompted new ways of organizing violence and heightened levels of militarization.”⁸⁷

How did slaving states work? The empirical examples provide details, but in general they were based on intertwined coercion and commerce: states made wars for slaves, and slaves made the state, by filling the ranks of their armies and by constituting the foundation of their economies, first as commodities to be sold, and later as captive labor forces for export production: “Overall, the slave trade meant the creation of what were essentially war economies, with ruling elites commanding armies which were economic assets as much as they were designed to bring about political aggrandizement.”⁸⁸

The necessity of reproducing the slaving state locked in dependence on the trade, and thus “slavery induced predatory institutions” that fed it.⁸⁹ First, there was the need to strengthen and replenish existing slave armies through forcible recruitment, for internal and external security. Second, slaves were the main commodity by which rulers obtained European goods, most notably guns and what Goody referred to as “the means of destruction,”⁹⁰ but also luxury items.⁹¹ Third, slaves were how subordinates were rewarded to hold the state together, either directly as they were granted slaves (usually women), or via goods and money obtained through the slave trade. Finally, the slave soldiers and the elite needed other slaves (again usually women) to grow their food.⁹² In combination, these imperatives meant that in many cases between a quarter and a half of the people in slave states were themselves slaves.⁹³

Systemic Consequences

The rise of slaving states created new systemic pressures as these states increasingly came into violent competition. The African slave states of the eighteenth and nineteenth centuries could raise larger armies than their predecessors and had better access to modern European weapons, obtained in exchange for slaves or slave-produced commodities.⁹⁴ Slaving states’ military prowess in turn fed their ability to win wars and thereby capture more slaves.⁹⁵ The success of the new slave states posed acute security threats for their neighbors and peers, creating an escalating systemic security dilemma and increasingly negative-sum competition.

The direct threat posed by slaving states was predation, perhaps in terms of conquest, but more often in terms of victims’ territories being “swept clean” of people, who were carried off to feed the insatiable demand for slaves.⁹⁶ Given the fixed pool of potential captives, one ruler's slaving gain was another's loss, in addition to those killed as a direct or indirect consequence of slave raids and war. Strategic depopulation to undermine a rival state through enslavement and massacres was referred to as “eating the country.”⁹⁷ Speaking of slaving states at this time, one historian notes that “warfare and capture made for an increasingly insecure world … These were societies undergoing a constant process of violent transformation,”⁹⁸ while another speaks of “a system that perpetuated violence between and within states.”⁹⁹ For rulers, the best way to avoid being the victim of slaving states was often to become one, creating “domino-effect militarization.”¹⁰⁰

The systemic effects of the rise and proliferation of predatory slaving states often put pressure on neighbors to follow the same route, even those that were ideologically opposed to the trans-Atlantic slave trade. For example, from the eighteenth century, a variety of reformist jihadi insurgent movements arose in West Africa with the rationale of overthrowing Muslim rulers who enslaved Muslims (in violation of Koranic prescriptions) and then sold them to Europeans.¹⁰¹ Yet when these movements attained power, in states like Futa Jallon, Sokoto, and Tukulor, they often found themselves drawn into the same trade: “There was no item which could bring as quick and sure a return as slaves. Thus, they had to slave to survive … They found themselves caught up in the same economic and military pressures that shaped their predecessors.”¹⁰² Speaking of these systemic pressures another historian concurs that “enslavement had become so basic to the political economy of West Africa that political survival was impossible without it.”¹⁰³ The timing reflects the regional expansion of the slave trade: “Within a few decades, between the late seventeenth and early eighteenth centuries, polities and elites whose economic survival rested on specializing in slave raiding and slave trading developed across West Africa. As a result, peer-polity rivalry and bitter conflict intensified.”¹⁰⁴ As described later, the same systemic effects were visible in nineteenth-century East Africa.¹⁰⁵ Thus, rather than just individual state formation strategies, it was the interaction of these strategies that created systemic pressures and pushed many (but by no means all) African states toward the slave-state model, remaking their internal military, political, and economic structures as they did so.

The section that follows focuses on the slaving states of Dahomey, Segu, and Zanzibar, but historians have suggested many possible further instances of this “family resemblance” concept,¹⁰⁶ including Kongo, Mbundu,¹⁰⁷ Oyo, Asante,¹⁰⁸ Futa Jallon, Sokoto,¹⁰⁹ Borno, Futa Tooro,¹¹⁰ Ngoni,¹¹¹ Wassloulou, Nyamwezi, Yao, Kano, Darfur, Ndongo, the Yoruba city-states,¹¹² the Wolof kingdoms, Whydah, Lunda, Kaarta,¹¹³ and Imerina Madagascar.¹¹⁴

What evidence is there of more intense and destructive security competition, both over time and compared with other regions? Though records are very fragmentary, Bates demonstrates that there was a sharp rise in the destruction of states in West Africa after 1700.¹¹⁵ In a comparison of nineteenth-century international systems in West Africa, South Asia, and Southeast Asia, Butcher and Griffiths find that the first was the most war-prone.¹¹⁶ Qualitative analysis from historians confirms this judgment of growing insecurity, at first in West Africa, and then later in the East.¹¹⁷

For all the predatory power of the slaving states, both their internal features and the systemic insecurity they engendered meant that they were also brittle. First, their very destructiveness meant that competition between them was often negative-sum:¹¹⁸ “Increasingly there was an emphasis on warfare rather than production as the generator of wealth.”¹¹⁹ Second, the dynamic whereby victories generated more slaves, which provided more wealth and power to win further victories, could and did go into reverse. Military defeat reduced the supply of slaves and the ability to trade them with Europeans, sometimes creating a vicious cycle of reduced military competitiveness and compromised internal political cohesion, leading to the destruction of the slaving state.¹²⁰ The increased military competition meant that there had to be losers as well as winners. Finally, there was something of a dynamic stability to slaving states: if rulers failed to mount raiding campaigns, subordinates often undertook their own “unofficial” raids. “When largesse did not flow the bonds between elites and their followers frayed and caused the states to collapse. Predatory states had the capacity to make war but little ability to transform social realities or to consolidate large areas under lasting political control.”¹²¹ Rulers had to constantly channel the slaving imperative outward, or risk their warriors despoiling their own domain. As a result, even in the nineteenth century “societies continually struggled to maintain stability and cohesion,” with a “continual process of fission and fusion.”¹²² These general institutional and systemic dynamics applied to slaving states in West and East Africa.

Dahomey and the Atlantic Slave Coast

Dahomey was perhaps the largest single exporter of slaves across the Atlantic in the eighteenth century.¹²³ It has been seen as epitomizing the new breed of militarized slaving states “designed more or less to fight wars and capture slaves.”¹²⁴ How does it fit the template I sketched out in general terms before? Dahomey was built on an outside-in basis, whereby external resources from the slave trade were used to build and maintain the military and economic bases of the state. Its fundamental imperative centered on capturing slaves through war, at first for export as commodities, later as labor for export-oriented plantations. Because neighboring rulers faced similar structural incentives and adopted similar strategies, Dahomey found itself in an acutely threatening security competition.

Dahomey began as a small inland tributary of another West African state, Allada, in the seventeenth century.¹²⁵ From the 1670s the rise of the Atlantic slave trade (both in volume and the prices of slaves) began to destabilize existing political arrangements by offering increased returns for slave raiding. In engaging in such “violent entrepreneurialism,”¹²⁶ Law holds that Dahomey's early state building was hard to distinguish from banditry.¹²⁷ The forerunners of the Dahomean kingdom used resources obtained from raiding to conquer its former overlord, Allada, in 1724, aided by the defection of other key Allada vassals.

Reflecting the crucial importance of external resources, Dahomey's early wars focused on getting direct access to the coast, and therefore to European slave traders.¹²⁸ Such access not only expanded revenues from the slave trade but also gave a better supply of guns (of which Dahomey was an early and effective user), while also cutting off rivals further inland.¹²⁹ As in other slaving states, slave export revenues and European trade goods were crucial to holding the state together through equipping the army and paying off subordinates. Dependence on European goods did not mean dependence on Europeans, however. Dahomey repeatedly destroyed European trading posts and forts when its ruler's wishes were disregarded, and exercised tight control of European traders.¹³⁰

Originally Dahomey closely fitted the model of a composite state and army, in that provincial chiefs governed autonomously and were responsible for raising and supplying their own forces. As Dahomey established its hegemony over surrounding vassal rulers, they paid regular tribute in slaves. Later, however, there was a trend toward centralization via the use of royal slave soldiers, including a substantial female contingent.¹³¹ The internal security function of these armed slaves was at least as important as their external role. Archaeological evidence demonstrates state consolidation as more strongholds, garrisons, and roads enabled a transition from relying on tribute from subordinates, to directly governing and extracting resources from the countryside.¹³² After the decline of the trans-Atlantic trade, Dahomey put its slaves to work on plantations (especially palm oil) to generate export revenue.¹³³ Its slave soldiers continued their slave raiding in the interior to replenish both the army and the labor force. With a majority of the population being slaves well into the nineteenth century,¹³⁴ the kingdom never lost its militarized character until its conquest by the French in the 1890s.

Rather than seeing the Dahomean slaving state in isolation, it is important to see the systemic context. Dahomey's neighbors faced the same structural incentives, and often adopted the same strategies:

The increase in the volume of slave exports, therefore, led necessarily to an increase in warfare and disorder. The disintegration of the existing political order on the Slave Coast [now the Bight of Benin] which was evident by the end of the seventeenth century, involving both wars among states and private banditry, was in the final analysis due to the commercialization of violence by the rise in the export trade in slaves. The effects were seen not only in the increasing levels of disorder, but also in the increasing prominence of groups for whom violence was a profession.¹³⁵

Thornton critiques some historians who advance a “predatory state hypothesis” because they ignore systemic pressures on Dahomey as neighboring rulers adopted similar strategies.¹³⁶ Not all of Dahomey's wars were prosecuted just to take slaves; many were waged for control of all-important external trade routes.¹³⁷ And as Dahomey campaigned to destroy coastal rivals it also faced predation from inland, as the cavalry forces of another slaving state, Oyo, repeatedly swept through Dahomean territory, taking slaves and destroying the capital.¹³⁸ Oyo had arisen around the same time as Dahomey and was built on the same foundation: “Slave-taking driven by warfare was the mainstay of the entire … system.”¹³⁹ Oyo's raids proved to be so destructive that Dahomey was forced to pay tribute from the 1740s until Oyo disintegrated in the early nineteenth century. Oyo's collapse and fragmentation then opened up new slave-hunting grounds for the Dahomeans to replace those they had emptied.¹⁴⁰

Segu and the Niger River Valley

Segu (or Segou) arose in what is now Mali, in the Niger River Valley. The Segu state created in 1712 has been described as “an enormous machine to produce slaves.”¹⁴¹ Meillassoux and Barry similarly see Segu as the epitome of the slaving state, the former arguing that its “function was war and the capture of men.”¹⁴² A more recent study comes to the same conclusion, based on archaeological evidence, that Segu was a “military kingdom whose expansionist project was geared primarily towards enslavement and looting.”¹⁴³ How, then, did this state function?

Segu was founded by younger men excluded from the gerontocratic structures of power in existing agrarian societies. They took advantage of the new opportunities provided by the expanded slave trade. Like their Dahomean counterparts a few decades earlier, they grouped together to prey on surrounding societies, such that “warfare and enslavement became the foundations for state power.”¹⁴⁴ Many of those captured were sold via networks of Muslim traders to Europeans on the coast. Segu became a supplier of slaves to the Atlantic, benefiting as both the volume and prices of slave exports rose.¹⁴⁵ In return, Segu's rulers obtained guns, horses, and other military supplies essential for continued raiding, as well as trade goods to reward subordinates. Just as important, Segu also sold slaves across the Sahara, especially after the decline of the Atlantic trade.¹⁴⁶ But many slaves were retained. Of these, younger men and boys were conscripted as slave warriors, while women were forcibly taken as wives who became responsible for agricultural work.¹⁴⁷ Older men were killed. In its form, the Segu state closely followed the template described earlier: “Segou exhibits the ‘bull's-eye’ structure typical of many historic West African states. There was a consciously defined and well-protected core … with rings of diminishing political domination and tribute beyond it, giving way to peripheral areas exploited by raiding.”¹⁴⁸

How did this state reproduce itself? The king regularly sent his forces on formal large-scale slaving campaigns, with half of all booty and slaves captured taken as the royal share, and the remainder kept by the slave soldiers and their leaders. Smaller raids of a few dozen horsemen were also regularly authorized to attack the periphery or villages defying the king's writ, with half the plunder and captives again handed over to the king.¹⁴⁹ Much of the royal share had to be distributed to maintain the loyalty of the slave lieutenants in charge of the various war-bands, who themselves had to buy their subordinates’ continuing obedience. The king also maintained a special cadre of palace slaves who “played the role of a political police” in repressing internal threats.¹⁵⁰

In the absence of successful campaigns and authorized raids, warriors engaged in small-scale but highly destructive “unofficial” raids within the kingdom, targeting the very slave plantations and long-distance trade routes on which the kingdom's survival ultimately depended. Thus, “Throughout the eighteenth and nineteenth centuries, the Segu state constantly struggled to maintain state power through external conquest lest it perish through internal brigandage.”¹⁵¹ Aside from tributaries on the periphery that required regular reconquest,¹⁵² the state was prone to lapsing into feuding fragments and then having to be militarily reconstituted, as occurred in the 1750s and 1790s.¹⁵³

As with Dahomey, it is important to see Segu as part of a system. Like Dahomey's, the rulers of Segu went to war to control and protect the long-distance trade routes essential to the reproduction of the state.¹⁵⁴ Again like Dahomey's rulers, they faced a hostile setting of states founded on similar, predatory principles. Thus in 1861 Islamic jihadi forces defeated and destroyed Segu, as well as the neighboring Caliphate of Hamdallahi. But the new rulers reproduced the same features of the slaving state, especially the tendency toward constant slaving wars in order to reproduce the state.¹⁵⁵ In another illustration of entropy over conquest, in the few decades before the French conquest the new rulers largely failed to manage the tendency toward disintegration via “unofficial” slaving, while also suffering invasions from other Islamic slaving states, like Timbuktu.¹⁵⁶ A French observer in the 1880s observed of the Niger Valley: “The inhabitants … conduct incessant warfare. The only object of these incessant combats is to capture women, children and young men in order to sell them … The chiefs sell their proper subjects in order to replenish their supplies of firearms and gunpowder and to buy beautiful ornaments.”¹⁵⁷ The pervasive insecurity in the region can also be seen in archaeological remains, which show that settlements were built for defense due to “the insecurity that reigned in eighteenth-century West Africa in the wake of the accelerating Atlantic slave trade.”¹⁵⁸

Omani Zanzibar and Southeast Africa

As the expansion of the slave trade in East Africa lagged that in the West, so too did the rise of slaving states. From the early nineteenth century, the Omanis set up a loose mercantile empire along the East African Swahili coast (now mainly Kenya and Tanzania). Earlier the Omanis had expelled the Portuguese from the Indian Ocean coast north of Mozambique.¹⁵⁹ The Omanis then increasingly reoriented themselves away from their original Arabian home to Zanzibar, which culminated when the sultan moved the court to the East African island in 1840 (from 1856 Omani Zanzibar became a separate sultanate in its own right). Once again, the expansion of foreign trade was the lifeblood of the Omani realm, and slavery was the foundation of its political economy. Most of these slaves were forcibly taken by Arab and allied raiders, whose predations extended further and further into the African interior as the nineteenth century went on. But due to their later rise, the Omanis made the switch from slaves as an export commodity to slaves as producers for Western markets more quickly than our West African examples did. This timing reflected the narrower window for trans-oceanic slave trading given the onset of the process of abolition, but also the huge expansion in the demand for African primary products from industrializing Europe and North America in the nineteenth century.

Though the Omanis had slaves before the 1800s, they were relatively few.¹⁶⁰ In the early nineteenth century the Omanis learned the model of the slave plantation economy from the French Indian Ocean islands, a model that became spectacularly lucrative with the end of the Dutch East Indies’ monopoly on clove production.¹⁶¹ From the 1820s to the 1870s the slave population of Zanzibar rose from 15,000 to 200,000,¹⁶² necessitating a constant flow of at least 10,000 replacements annually to make up for the horrifyingly high death rate on the clove plantations.¹⁶³ Later the emphasis switched to slave-farmed grain and ivory, cementing a “slave mode of production” throughout the Omani Swahili coast.¹⁶⁴ By the 1870s the rise of slave plantations meant that around 40 percent of the people in the area were slaves.¹⁶⁵ In exchange for the slave-produced commodities, the Omanis imported Western manufactures, particularly guns. The Omanis benefited from the steeply rising prices for ivory, cloves, and grain in the West.¹⁶⁶

The Omani state was a conglomerate. The core of the sultan's domain in Africa was the island of Zanzibar. Along the port cities of the coast, ruling families pledged suzerainty to the sultan, who supervised and levied the crucial customs duties on foreign trade through his former or current slave officials.¹⁶⁷ Around the ports were plantations, worked by slaves and producing for export. Further into the hinterland, armed traders hunted for ivory and people. Tippu Tip, the most successful of the late-nineteenth-century Omani slavers who became dominant in the interior, owned 10,000 slaves, some armed, many working on his twenty plantations. In the 1880s he claimed much of the eastern portion of the Congo for the sultan.¹⁶⁸ The peak of the Omani state may have been in the 1870s, but even at this time it was falling under the sway of British hegemony, which was spreading west from India and the Persian Gulf. The Omanis were eventually undone by the British, who forced the abolition of slavery, and thus kicked out the props of the sultanate.¹⁶⁹

To what extent were the systemic features discussed earlier replicated in East Africa? The shifting incentives created by the expansion of the slave trade were as important for other actors as they were for the Omanis. As in West Africa, these incentives rewarded military predation for slaves. This shift had undermined old states and created new ones in the Atlantic coast and the Niger River Valley in the early 1700s, and it did the same in East Africa a century later.

An important example is the prazeiros of the Zambezi River Valley, who had gradually achieved independence from the Portuguese crown and hybridized with local African societies. With the sharp rise in the demand for slaves, the prazeiros began raiding further into the interior. They also made the suicidally short-sighted decision to sell their own armed slaves, the chikunda, who might number several thousand on a single estate.¹⁷⁰ This hereditary group of slave soldiers, organized in regiments and kept separate from the rest of the slave population, were the forces the prazeiros depended on to keep the rest of their slaves in line and to defend against external threats. Faced with sale to the Americas, the chikunda often revolted, which then prompted outside predators to attack the undefended estates. By the 1830s, twenty-eight of the forty-six prazos had been overrun.¹⁷¹ The chikunda then set up their own small slaving states, each of which depended on slaving and looting to maintain its numbers and to obtain weapons, trade goods, and food.¹⁷² By the 1880s some chikunda states had their own slave armies of up to 10,000.¹⁷³ Thus over the longer term this region of what is now Mozambique saw a double disintegration of political authority: first the Portuguese crown lost control over the prazeiros, and then the prazeiros were overthrown by their chikunda slave retainers. The latter were eventually defeated by Portuguese and Belgian imperial forces, themselves often comprised of former slaves.¹⁷⁴

The Omanis and the independent chikunda operated on broadly the same model in mainland East Africa.¹⁷⁵ Furthermore, they fought and competed against each other in the same slave hunting grounds (around what is now Malawi) for the same diminishing pool of victims.¹⁷⁶ For both parties, “military slavery was an engine of economic production.”¹⁷⁷ More specifically:

Economic motivations for war became paramount. Predatory traders and warlords fought wars to gain exclusive control over trade routes, to raid for slaves and ivory, and to secure food and supplies as bands that specialized entirely in militarized strategies abandoned cultivation. A system of extraction that relied on warfare drove the demand for arms and ammunition: the most important factor of production became guns and gunpowder.¹⁷⁸

By the late 1800s “Warlordism had become the dominant feature of the East African interior from Tanzania to South Africa.”¹⁷⁹

As lands were stripped of their population, competition for slaves became more and more acute.¹⁸⁰ Contemporary historians echo the verdict of nineteenth-century European observers that the region became “a picture of destruction and despair.”¹⁸¹ Rather than this violence being random or senseless, however, it was closely linked to the international economy.¹⁸² Alpers explains the general pattern of destruction in the region as the systemic consequences of competition between slaving states: “Inexorably fettered to the international slave trading economy, they were unable to do otherwise, since their mid-[nineteenth-]century strength had been built on the profits of the slave trade.”¹⁸³ But the structural predicament of each actor was military as well as economic: those which could not engage in predation lost access to military supplies and thus became vulnerable to destruction by their peers.¹⁸⁴ Thus, in general, in the nineteenth century, the political effects of external trade “locked much of Africa into a spiral of violent competition.”¹⁸⁵

Though space does not allow their full consideration, other East African state-building projects of the nineteenth century were also fundamentally reliant on slaves. In Merina Madagascar around half of the people were slaves,¹⁸⁶ as their rulers formed “the army into a regular, well-trained, and well-armed slave raiding organization from the early 1820s.”¹⁸⁷ Ottoman Egypt-Sudan also relied on slaves to staff its army and to labor in the cotton plantations that expanded massively after the end of the US Civil War.¹⁸⁸

Conclusion

This article has presented an important new path of slave-based outside-in state formation. Slaving states were formed with external resources provided by an expanded demand for slaves, both as commodities and as export producers. The proliferation of militarized but brittle slaving states sharply increased systemic competition, and hence insecurity. These states were crucially dependent on slavery for the continuing flow of external resources that held them together, and hence were vulnerable to any interruption of these flows.¹⁸⁹ A shift in the international economic environment thus wrought change in both the states and the system created by their interactions. The ultimate expression of the brittleness and vulnerability of slaving states was their destruction, and the destruction of the entire African international system, with the European imperial conquests of the late nineteenth and early twentieth centuries.¹⁹⁰ An important factor in the Europeans’ victories was their ability to recruit Africans displaced by the endemic violence and destruction of the slaving-states system.¹⁹¹

The idea that intercontinental, transnational flows of trade and people could exercise powerful “behind the border” effects, and even provoke the transformation of states, might seem to be limited to the contemporary era of globalization. In fact, the second-image reversed account presented here suggests that this broad dynamic first obtained in Africa, and in some parts of the continent as early as the close of the seventeenth century. To this extent, the slaving states were precociously modern. The story told here is thus simultaneously one of distant and alien historical experience—centered on mass enslavement, the slave trade, and a violent and unstable regional international system—but also one of surprising parallels, in the way that the external economic and security environment shaped domestic politics and institutions.

Recently some scholars have argued that the scars of slavery and precolonial violence can be traced through to current political outcomes in Africa.¹⁹² Given the huge differences between Africa's slaving states and those in existence today, what might link the two? One of the most persuasive answers has been presented by Jean-Francois Bayart. He argues that Africa has a recurring “history of extraversion” extending to the present, whereby African rulers “mobil[ize] resources derived from their (possibly unequal) relationship with the external environment” to facilitate “political centralization and economic accumulation.”¹⁹³

The ideas of extraversion and outside-in state building chime with the writing of many other scholars on the postcolonial state, particularly in Africa, but also more broadly.¹⁹⁴ For these scholars, the decline of international war and conquest from 1945 means that, unlike in the classic European “bellicist” account, external military competition is no longer the primary driver of state formation.¹⁹⁵ Instead, the main threats to rulers and to state formation are internal. Inverting the conventional account, the external environment is a crucial and perhaps even the primary source of resources with which to counter these internal threats in building and maintaining the state. These external resources might range from international legitimacy to foreign aid and cheap credit, and from great power military patronage to commodity exports to international markets.

Thus Jackson argues that postcolonial African states “turn Hobbes inside out: the state of nature is domestic, and civil society is international.”¹⁹⁶ These “quasi-states” are supported from above and outside by international law and foreign aid, rather than domestic resources. In the same vein, Henderson speaks of a “political inversion” in Africa whereby domestic politics is violent and anarchic compared with a relatively pacific international realm.¹⁹⁷ Tilly agrees that postcolonial state formation has followed a very different path from that of early modern Europe. He too stresses the decline of international military competition, the greater prominence of internal threats to the state, and the shift from internal to external sources of support for state formation.¹⁹⁸ Rather than these external and internal aspects being independent, it is argued that the very lack of the discipline and toughening exerted by Darwinian international military competition is responsible for internal vulnerabilities.¹⁹⁹ There are important differences between scholars of postcolonial state formation, but nevertheless the outside-in character of their arguments is striking. Given these common themes, how is the thesis presented here distinctive? The most important points of contrast relate to historical sequencing, the centrality of slavery, and the place of international war.

The first and most straightforward point of distinction is that of historical timing and sequencing. The evidence presented here suggests that slaving states were the trailblazers of outside-in state formation, anticipating some (but only some) of the dynamics followed centuries later by their postcolonial successors. If some African states led the world in this outside-in strategy, it reflects the fact that Africa's external economic and security environment exerted pressures and provided opportunities earlier and more strongly than in other regions thanks to Africa's pioneering place in globalization. It bears remembering, for example, that until the mid-nineteenth century, five of every six people arriving in the Americas were African slaves, not Europeans, and thus colonization in the Western Hemisphere was largely the Africanization of the Americas.²⁰⁰

Second, Africa's prominence in globalization and engagement with the external environment from the seventeenth to the nineteenth century was sui generis because of the centrality of mass slavery. As we have seen, this encompassed not just the Atlantic and Indian Ocean slave trades but also the nineteenth-century boom in slave-produced exports from Africa. Thus rather than the benign external influences of post-1945 norms of sovereign equality and international development, it was the deadly temptations of internationalized slavery that shaped state formation in this earlier era.

The last fundamental point of distinction between the slave-based and contemporary outside-in route to state formation concerns war. The postcolonial path is said to crucially depend on a low and declining rate of interstate war, such that states do not exit the system. As noted, domestic insecurity is often said to be a direct product of this international systemic security. In stark contrast, the rise of slaving states created a violent and unstable international system in which state death was a real and present danger.²⁰¹ Postcolonial states appeared in an international system where the norms and rules had to a large extent already been set by their former imperial masters. The slaving states, however, substantially altered the system dynamics in accentuating insecurity through the interactive effects of their behavior. Thus the outside-in state formation discussed here not only occurred centuries earlier than previously considered, but also occurred under radically different systemic conditions. Contrary to the central tenet of the postcolonial state formation scholarship, not only can outside-in state formation proceed in circumstances of acute international insecurity and instability, but this strategy can actually sharpen this predicament.

Historians rightly caution about reading current-day concepts and politics far back into the past. Even the term “Africa” is a European-imposed anachronism; the “Africans” written about here did not think of themselves as such.²⁰² In part for this reason, there was no normative prohibition on Africans enslaving “other” Africans. Religion, not race, was the more salient normative restraint on slavery in large areas of the continent, because of the Koranic prohibition on Muslims enslaving fellow Muslims.²⁰³ It is impossible to understand slavery without reference to norms. The end of slavery reflected above all a delegitimation of the idea that one person can own another as property.²⁰⁴ Thus by looking into the past and trying to learn something new from it, we are looking into a different normative world.