African Conceptions of Enslavement
African involvement in the transoceanic slave trades is today a particular sensitive topic, given the huge scale of the human traffic and the equally huge discrepancies in modern income levels between sub-Saharan Africa on the one hand and on the other the Old-World nations that organized the maritime segment of those trades. John Thornton has grappled with this dilemma in an essay that lays out what might be taken as a major subtheme of the present work as identified in Chapter 1.
African rulers controlled states typical of their era throughout the world. Such states were often based on assumptions of power and authority that are no longer fashionable … the right of governments to tax and take tribute without necessarily delivering any services, judicial systems that unashamedly protected the interests of the rich and powerful … holding people as slaves … and selling these rights. Today … these ideas are repugnant, but rulers nevertheless … had their principles and ethics, even if these were not the same as today.Footnote 1
African states had very clear ideas about who could be enslaved and who among those enslaved could be forced into the hold of transoceanic slaving vessel.
For millennia before the onset of Atlantic slave trading, states with very different cultures across the Eurasian landmass had participated in slave markets with buyers and sellers each having different conceptions of what slavery meant. Such variations in comprehension at the time of the transaction have often meant that modern scholars are unable to agree on how to define slavery. In a sense the Venetian notary in the Sea of Azov’s port of Tana in 1360 who recorded the sale of a Chinese girl – no doubt a “base” person in her country of origin – was no different from the Portuguese buying an African at Arguim a century later. In both instances the enslaved person was moving from one type of enslavement to a very different kind. What made such transactions possible was that the person traded was for both parties an outsider and accordingly eligible for slave status. As discussed in Chapter 1, Africans did not have a sub-continent-wide sense of identity or a shared affiliation with a world religion that prevented them from enslaving and selling other people living in Africa. In contrast, for sixteenth-century Europeans enslaving other Europeans was no longer possible, but for them any African was a potential slave. Yet on the African coast as the Portuguese quickly discovered Europeans could buy only those Africans that the sellers also considered to be outsiders.Footnote 2
In the following centuries European newcomers to the business following in the Portuguese wake typically learned what might be called the eligibility rules for enslavement the hard way. In the absence of a land-based imperial presence in Africa the Hawkins slaving raids in Sierra Leone in the 1560s could not sustain a transatlantic slave trade. Likewise, a century and a half later as the English Royal African Company’s monopoly crumbled, a host of London newcomers, called “ten-percenters” arrived on the coast without fully comprehending the rules. They did not raid African communities, but neither did they fully appreciate the often-narrow African eligibility requirements. Several of them panyarred (captured) free Africans, then sold them in Barbados, thereby temporarily wrecking the trading relationship for all English on the coast.Footnote 3 The RAC ordered its agents on the island to retrieve these individuals and then notified its Cape Coast Castle factors:
the Guinea Hen carries back (from Barbados) Seven free tradeing Negroes Panyared by Capt Jackman wch our factors have wth great difficulty Recovered them & sent home. Pray send us some affidavits of Capt. Jackman and other 10 percent Men’s miscarriages upon ye Coast in this or like manner wch may be of use to us.Footnote 4
A similar situation occurred in the aftermath of South Carolina’s reopening of the African slave trade between 1804 and the end of 1807 when several hundred inexperienced US and Cuban merchants arrived on the African coast to take advantage of this temporary window. What many of them did not know was that Vili societies on the Loango coast, a major source of slaves for transatlantic markets north of the Congo, had a law prohibiting “the sale of anyone born in Loango” other than criminals.Footnote 5 Among the newcomers was a Captain Churchill who had previously “taken off some of the free natives of the Kingdom of Leango [Loango, north of the Congo].” Churchill had the temerity to return to the same port where the Vili Mafouks conspired in retaliation to put on board individuals who would start a rebellion while the vessel was still loading. The rebellion was unfortunately too successful – in the resulting melee the powder on board exploded killing all 240 captives. Churchill escaped but his supercargo later wrote “[t]he people are, in Congo, much exasperated against Captain Churchill, and if they can get possession of him, I have no doubt would murder him.”Footnote 6 Other documented cases of slave-ship captains ignoring African norms of eligibility involved equally inexperienced slave-ship’s officers. Boston captains James Smith and Thomas Silk, on their only trips to Africa in 1645 and 1736 respectively, enslaved African or, more precisely, lançado slave traders with similarly violent outcomes.Footnote 7
But perhaps the clearest and most poignant indicator of African ideas of freedom appears in the record just five years later – when Churchill was still fighting in court in his ultimately futile attempts to claim insurance on the destruction of his human cargo. In 1812 six “natives of Africa” placed on board the condemned slaver S Miguel Triunfante at Ouidah vigorously petitioned against being entered in the Liberated African register. How could they be “liberated,” they argued when they had been wrongfully enslaved in the first instance? They “declared themselves freemen and never that they were slaves.” The court agreed and indeed they do not appear in any of the court’s registers and were not assigned apprenticeship duties. As discussed in Chapter 7, this was indeed a rare instance where rescue from a slave ship really did mean untrammeled freedom.Footnote 8
Yet African conceptions of who was eligible for sale to offshore slave traders could also be porous. The Vai people of the Windward Coast dominated the traffic from Galinhas in the nineteenth century, and occupied territory extending 35 miles inland and comprising 3,300 square miles.Footnote 9 Thanks to the African-Origins database we know the ethnolinguistic associations of 2,756 people who left Galinhas in the hold of a slave ship in this era. Only forty of this sample were Vai – most presumably convicted of offenses within the Vai community. No clearer indication of the dominant Vai role as slave traders exists. British slave-trade suppression policies took effect here earlier than at most points further south. After 1850, only two slaving ventures left from the region – the last one in 1856.Footnote 10 Two years later, by which time slave supply lines to the coast would likely have atrophied, a French ship, the Regina Coeli, showed up off Cape Mount seeking engagés – or captives who would be bought from slave traders to serve ten years of indentured servitude on French Caribbean sugar plantations. In the absence of robust captive supply lines, the Vai elite responded by selling many of their own domestic slaves to the French. Disastrous consequences followed. While held on board off Monrovia the former domestic slaves staged a rebellion fueled, as they later claimed, by resentment at the violation of their status, which they understood to have constituted protection against alienation by sale, certainly into the Atlantic slave trade. Clearly, the Vai did not see themselves as fodder for plantations in the Americas. The captives killed all the crew and escaped, mostly into the Monrovia region.Footnote 11
As the above suggests, the African-Origins database can be used to identify not only protected groups, but also to provide a novel way to identify African slave traders and enslavers, at least for the nineteenth century. Susu peoples had a strong presence in the hinterlands of the Rio Pongos and Rio Nunez. Language identifications exist for 2,959 individuals embarking at these trading nodes.Footnote 12 Yet only thirteen of this group were Susu, thus providing numerical backing for Jorge Felipe’s identification of several major Susu slave traders at both locations.Footnote 13 The same database provides language identifications of 3,645 people embarking at Malembo, Cabinda, and Loango, Atlantic outlets for the three separate states of Loango, Kakongo, and Ngoyo, and where Vili slave traders were dominant. Only one Vili is identified, suggesting that even Vili lawbreakers were not eligible for dispatch overseas. Likewise, among the 2,861 Africans on ships captured leaving Jakin and Whydah after 1807 whose language we can identify, only fifty Fon names appear. Further south, the distinction between slaves that could be alienated for sale overseas and those that could not was already well established in the major polities of Kongo and Ndongo in mid seventeenth-century West Central Africa. A partial ecclesiastical census in 1704 revealed many slaves in Angola who did not expect to be sold away from their families. And at the very end of the traffic Kosoko, king of Lagos, discovered that one of his palace officials had been enslaved and sent to Bahia. As with most attempts at recovery. his efforts to retrieve this formerly free man failed.Footnote 14
The space for cultural misconceptions and malfeasance that might result in the free becoming unfree was undoubtedly greater on the African coast than in most slave markets in Eurasia discussed in Chapter 1. Children of African slave traders regularly accompanied their European counterparts to a vessel’s home port for language and literacy training that could last several years before they returned home. Typically, they lodged in the homes of well-known slave traders in both French and British ports. But the Atlantic was a violent and unpredictable ocean, especially in wartime. Captain Cusack of the Henry (ID 81804), charged with carrying two 12-year-old sons of “kings” of Gabon to Liverpool “to be there educated,” was captured by a Spanish schooner in 1798 and taken into Havana, where the two free African boys were sold along with the other 134 Africans on board, “in spite of all his (Cusack’s) remonstrance.” They probably ended their lives on a sugar plantation.Footnote 15 This and other incidents mentioned here speak to the tacit agreement between buyer and seller that must exist for market transactions to occur and be sustained over time. But they also underscore that the items being traded – human commodities – were not without influence on the transaction’s outcome. Breakdowns in the agreement might result in violence, but the capacity for retaliation on both sides ensured that neither buyer nor seller held the upper hand for long.Footnote 16 Outside Angola White power on the oceans was matched by Black power on land.
The absence of permanent power imbalances between Europeans and Africans at the transaction point on the coast is consistent with the absence of hard evidence of either side imposing unequal trading terms on the other. Descriptions of the exchange of people for merchandise on the African coast have not survived for the first century and a half of transatlantic slave trading. Detailed accounts of the negotiations between African sellers and European buyers begin to show up only in the later seventeenth century. We have seen that in the first two decades after the British joined the slave trade in 1640 almost 70 percent of the Africans they carried off came from New Calabar, Old Calabar, and a few from Bonny, all in the Bight of Biafra.Footnote 17 The earliest accounts of this trade emerge only toward the end of the century, with the logbook of the Vine (1682), accounts from the journal of the Fly (1697), and the much fuller record of the Albion (1699).Footnote 18 We know from the names of the African traders listed in these documents that they were middlemen of mainly Qua and Efik origin, and that the people being traded were almost certainly largely Igbo and Ibibio. How did the Efik and English view each other? These documents clearly communicate a sense of equality between buyer and seller in the transaction. When conflicts developed between African families or houses of traders, Europeans certainly sought to aid one side or the other in their own self-interest.Footnote 19 But whatever the impact on societies from which slaves were drawn, at the point of sale and embarkation Europeans dealt with African traders as one merchant to another. Most compelling are the warm and familiar letters, written from 1760 to 1789, by Bristol and Liverpool merchants on the one hand and important African traders in Old Calabar on the other.Footnote 20 Bargaining over price and commodities, human and material, might include disputes and detention, with even violent deaths always possible, but the risk from such events was there for both, and resolution of disputes was never in the long run one-sided.
For the distinguished editors of the definitive edition of Jean Barbot’s writings on Guinea, the journal of the Albion (ID 21073) “was probably the most detailed account of a slave-ship, in any century of the Atlantic trade, arriving at an African port, dealing with African traders, and embarking slaves.” Of Old Calabar, Barbot wrote:
It is well furnish’d with villages and hamlets all about, where Europeans drive their trade with the Blacks, who are good civiliz’d people, and where we get, in their proper seasons, as at New Calabar, all sorts of eatables, yams, bananas, corn, and other provisions for the slaves, which we barter there, as well as elephants teeth.
The very detailed reports about Efik traders in the eighteenth century available in the correspondence of Liverpool traders and in the Diary of Antera Duke: An Eighteenth-Century African Slave Trader demonstrate that this assessment of the slave traffic as a trade between equals in the seventeenth century also holds good for the later period.Footnote 21 In August 1771 the London slave ship Elizabeth (ID 78857) ran aground at the entrance of the Bonny River. As was typically the case for ships in distress in coastal waters anywhere in the world, the local population attempted to exercise salvage rights and threatened to plunder the vessel. Captain Welch sought help from King Varrée of Bonny who immediately came in person with his retainers “s’opposer à ce désordre.” However, fighting broke out and the powder magazine exploded, killing all on board including King Varrée and the captain.Footnote 22 A few decades later, overtly racist effusions in the London press notwithstanding, the shift from the traffic in people to an even higher value traffic in palm oil after 1807 took place without disturbing that balance and indeed the respect of European buyers and African sellers for each other.Footnote 23
Contrast these narratives of African-European interaction with the extreme power imbalance, accompanied of course by arrogance, that emerges from the account of the arrival of another British ship in the region. The flagship of Commodore Wilmot, Commanding Officer of the British anti-slave trade squadron dropped anchor in the Bonny River in 1865 – just a few years after the occupation of Lagos had established the first British land-based presence in what was to become Nigeria. Wilmot wrote a private letter to the head of the Slave Trade Department of the British Foreign Office describing his visit. He began “Bonny is settled satisfactorily” – by which he meant that the transatlantic slave trade had ended. He went on:
I was there the other day and remained in the River 3 days. [King] Pepple was too ill to attend, but his son and principal chiefs attended. The first day they did not come to time, so I fined the King 20 puncheons, [of palm oil] and went out, returning again in 48 hours, when I found half the fine paid, and the other half forthcoming if I desired it. A most humble letter of apology was also sent promising good behavior for the future. I assembled all the traders, chiefs, etc., gave the latter a most severe lecture, which astonished their weak minds, after which I let them off the remainder of the fine …Footnote 24
If there was ever a single document that punctuated the shift from a trading relationship of equals to the master and subaltern connections of imperialism, this is it. This private correspondence between the two most important permanent British officials charged with suppressing the slave trade, suggests that 1865 rather than 1851 – when the British took over Lagos in a close-fought assault – was the year that the British completed their transition from suppressor of the slave trade to the major imperialist power of sub-Saharan Africa.
In the slave trade era, any inequality, at least in terms of both market and military power, was typically experienced by Europeans, not Africans. At Ouidah, Bonny, and the embarkation points north of the Congo River, buyers negotiated a price with the political authority prior to the onset of trade. The Vili Mafouks (officials of the king of Loango) were frequently able to operate as an effective monopolist. For the earliest era of the traffic Ivana Elbl has noted “[t]he iron-handed control that the Oba of Benin exercised over slave supply to the Portuguese.”Footnote 25 When the slave trade was its peak, the king of Dahomey and the king of Bonny played this role. Given the presence of other slave ships off the coast, almost all owned by different buyers of captives, there were frequently several prospective buyers, but only one African authority who could authorize trade. Only on the Upper Guinea coast was there a more openly competitive market structure. It is indeed, possible that the lack of significantly large state structures on the Upper Guinea littoral helps explain why this large and populous region contributed fewer than 12 percent of the nearly 13 million total that left the sub-continent. But even here there is nothing to indicate unequal trade. If on the coast of Western Africa, the advantage could lie with the seller rather than the buyer, in the Americas, European slave traders could sometimes find themselves temporarily as the only seller.Footnote 26
Commodore Wilmot’s report signified that several centuries of mutual respect between European and African on the sub-Saharan littoral had come to an end. Toby Green’s argument that Europeans were the dominant partners in trade from the early eighteenth century is not supported by the thousands of individual transactions discussed below and now available online.Footnote 27 But just who were the African slave traders that participated in these transactions? Like merchants everywhere, slave traders had minimum requirements without which sustained trading activity was impossible. The first was ample supplies of the commodity to be traded, in this case a human commodity. The second was trust between both parties in a transaction that the terms of any agreement would be carried out. This was where the African political authority became centrally important. The crucial role of the African state in the external slave trade lay not with supplying enslaved people, but in underpinning and sustaining mercantile activity. African rulers certainly acquired and supplied captives to Europeans, especially after winning armed conflicts. both civil and interstate. King Agaja of Dahomey did establish a crown monopoly on slave sales in Ouidah that lasted for two decades, but it resulted in declining numbers of enslaved people leaving the port and could not endure. The main income of rulers derived from the fees and taxes levied on transactions in which they were not direct participants.Footnote 28 These prerequisites could be provided by very small polities such as those in Rio Pongo, Galinhas, and Casa Mansa in Upper Guinea, as well as Bonny and Old Calabar in the Bight of Biafra, all located in stateless regions.Footnote 29 But obviously larger states such as the Fante on the Gold Coast, Dahomey, and at different times, the kingdoms of Kongo and of Loango could also provide such an environment. Such polities of course could also refuse to participate in the slave trade as did the Kingdom of Benin in the early years of the transatlantic traffic.Footnote 30
A third prerequisite that receives almost no attention from scholars is the larger population’s acceptance of slavery and the slave trade. Though plantation slavery existed only in São Tomé and later in the Mascarene Islands, early modern Africa was no different from the rest of the world in its acceptance of the complete subordination of one human being to another that constituted slavery. Slavery itself is a matter primarily for the enslaved, the slave owner, and the community or state authority that sanctions it. The slave trade, however, involved a large segment of the African community. Manufacturing and assembling trade goods – or alternatively an army – feeding and transporting captives and manning barracoons until those captives embarked or were sold, demanded many more human resources than maintaining plantations where slaves sustained themselves. While the scholarly focus is very much on the organizational activities of slave-trading elites, many thousands of workers in Europe, the Americas, and Africa had no qualms in consuming slave-grown produce and finding employment in the sector that supplied enslaved labor to American plantations.
While we cannot enumerate such active participants in the business with precision, we can plausibly estimate nearly a million crew and owners of those slave ships leaving ninety-five different European ports over the slave trade era, with at least as many again involved in preparing these vessels for slaving ventures. Iron bars in Sweden, beads in Italy, linens in Silesia and the varied patterns of cotton and woolen textiles manufactured in all European countries created specifically for African markets over a 360-year period employed many times more than those preparing and working on slave ships. A similar exercise for the Americas produces even larger numbers. As argued in Chapter 2, the American population not only dispatched slave ships to Africa but also ensured the reception and distribution of their unwilling occupants at disembarkation. For every captive carried across the Atlantic there could easily have been months of employment for a worker in Europe and the Americas. As the petitions from textile workers in support of the early eighteenth-century slave trade discussed in Chapter 3 indicate, one can see only enthusiasm for the business, accompanied by a desire to avoid enslavement of oneself.
African Slave Traders
The share of economic resources in Africa dedicated to obtaining and delivering captives was at least as great as those in Europe and the Americas. In the last decade researchers have uncovered trade books of sixty-four French, English, and Dutch slaving ventures buying captives at major embarkation outlets ranging from Galinhas on the Sierra Leone–Liberia border to Cabinda in an Angolan enclave north of the Congo River. The trade books record a total of 5,500 transactions, in which 13,900 enslaved people changed hands. For this topic, their importance lies in that each transaction specifies the name of the African seller of the captive. Stacey Sommerdyck has undertaken the laborious task of converting thirty-two Middelburgse Commercie Compangie trade books into a usable database, and this is now augmented by similar records from an equal number of English, Portuguese, and French voyages.Footnote 31
The African sellers of slaves fronted a large labor force ensuring the delivery of captives and supplying provisions. Table 5.1 shows that during the eighteenth century, slave vessels spent on average nearly five months on the African coast seeking to complete their complement of captives. Trading times did drop dramatically after 1810 partly in response to attempts to suppress the traffic but typically the African embarkation phase of the voyage was always of much greater duration than the disembarkation phase in the Americas. Captains hired a range of African help: pilots in riverine environments, canoes and the men to operate them where the surf was severe, and in the era of suppression, guards and canoes to shift embarkation points in response to naval cruiser activity. Kru men, living along what is now the Liberian coast, were essential for navigation off the greater part of West Africa. Where Europeans had a land-based presence at locations ranging from Senegambia to the Bight of Benin, most of the personnel were African, usually Gromettoes and Guardians – to control the enslaved. Guardians usually originated in regions remote from the location of the fort or trading post itself to discourage fraternization with local populations. Some of these were enslaved, but most were not. All European establishments were heavily dependent on local labor for construction and maintenance of physical infrastructure, as well as provisions for both the European residents and most of the vast quantities of provisions that the 12.75 million crammed occupants of more than 40,000 slave ships needed to reach the Americas.Footnote 32
Table 5.1 Time spent in days by vessels trading for slaves on African coast
| Years | Number of days on African coast | Number of voyages | |
|---|---|---|---|
| Mean | N | Std. deviation | |
| 1661−1700 | 93.7 | 98 | 64.6 |
| 1701−1809 | 141.4 | 3,550 | 93.1 |
| 1810−1834 | 99.8 | 143 | 61.4 |
| 1835−1866 | 31.6 | 37 | 73.0 |
| Total | 137.5 | 3,728 | 92.4 |
African sellers of slaves were even more dependent on African labor. The trade books show hundreds of Africans receiving payments for small groups of captives, often just one or two per transaction as the above summary suggests. Indeed, the principals in the slave trade – buyers and sellers of people – were perhaps just as numerous in Africa as their counterparts were in the Americas and Europe. As Obikili has pointed out, “the opportunities and profits from the slave trade were open to everybody.”Footnote 33 The older idea, shared by historians across the ideological spectrum, that mostly elite Africans sold slaves needs to accommodate this new evidence.Footnote 34 The system that held in Portuguese West Central Africa and the southern rivers of Upper Guinea described in Chapter 2, which saw human cargoes assembled on land prior to embarkation may have conformed to this pattern, but we need to know more about how captives entered the Portuguese holding pens. Outside the Portuguese-dominated embarkation sites, the trade books show hundreds of sellers supplying captives to slave vessels, albeit often at a price previously established by negotiations between ship captain or supercargo on the one hand and the African ruler or his representatives on the other.
Many of these sellers were representatives of the head of state or members of the local nobility and appear in ten or more transactions per voyage often spread over two or three months. But a long tail of sellers always shows up in these documents, with the seller’s name appearing just once or twice per vessel. Sixty-one percent of the transactions involved individuals who sold no more than five captives, usually to more than one ship. Indeed, 300 sellers sold just a single enslaved person. Some may have been employees, given that Stacey Sommerdyk has raised the possibility that the many Africans involved in guarding and delivering the enslaved to the coast were, like the Portuguese sailors on slave ships discussed in the previous chapter, paid in slaves.Footnote 35 But a more likely explanation for this exceptionally long tail is the involvement of numerous small merchants. Such a pattern appears nowhere in the existing literature, which tends to focus on the necessarily larger “warrior-merchants” found especially in the hinterland of the Bight of Benin–Dahomey, and the various factions in the wars following the collapse of Oyo. Perhaps the pattern that Tony Hopkins observed in the post abolition era – the “crisis of adaptation” that African warrior-merchants faced following the decline of the Atlantic slave trade and the shift to “legitimate commerce” was restricted to such hinterlands. Perhaps the small-scale “producer” was always a major component in the market for both enslaved people and African produce on the African coast, both before and after abolition.Footnote 36 Stacey Sommerdyk comments, “a large number of smaller traders also engaged in the trade … there is an underlying story of minor merchants yet to be uncovered in Western African history,” to which Mariana Candido adds in a study of Benguela, “there were also many smaller merchants who handled only a few slaves at a time.”Footnote 37
The largest number of enslaved were sold by Prins Tom in Malembo, north of the Congo River, who between 1749 and 1765 put 467 enslaved individuals on to seven different Dutch slave ships. Only ten African traders managed to sell more than one hundred individuals. But the exceptional size of these sales is highlighted by the overall average per trader, which was just eleven individuals per transaction. No less than eighteen cases involved fractions of an enslaved person. The major sellers established their status via multiple transactions of relatively small numbers of captives spread over the several weeks or months of a vessel’s stay on the coast. But the African ruling authority was not always among this group. The king of Dahomey only appears eight times in our sample for a total of fifteen enslaved, but we know from Robin Law’s work that he was often not a major supplier of captives to vessels trading at Ouidah. King Kosoko (Cocioko), the last independent ruler of Lagos, from which port far more enslaved people entered the transatlantic trade after 1800 than from Dahomey’s outlet of Ouidah, presided over the departure of 32,000 captives during his reign of 1847–1851. Thanks to captured correspondence we know that the king himself sold only 212 of these. Of course, as with the king of Dahomey, direct sales never formed Kosoko’s only income from the slave trade, though it is noticeable that he was never able to pay for his own slaving vessel, which he had ordered through merchants at Bahia.Footnote 38
The new data do show a larger number of individual sellers on the Loango coast than appear further north. Nine hundred Africans sold slaves in the former region, but only 200 are identified in Bonny and Calabar, albeit with a smaller number of voyages. Nevertheless, for Igboland Northrup has argued that “the profits of the slave trade were distributed more widely … than in parts of Africa where force and central control were more important.”Footnote 39 Large caravans of captives did arrive at, say, Galinhas from the Fulani state of Fouta Djallon, and at Bonny as fleets of Ibani canoes arrived from interior markets. Lagos, too, saw caravans arriving, formed from the surges of people displaced by armed conflict as the final collapse of the Oyo Empire got underway in 1817.Footnote 40 Such sudden influxes of large numbers of captives into the coastal outlets represented in our sample must have occurred at times, but they were not the norm. And in centers stretching from the Gambia River to Benguela it seems that there was not just room for the smaller trader, but indeed the smaller trader provided most of the captives.
As in the numerous regions that sent ships to buy slaves in Africa, the principals involved such as ship owners, ship builders and captains, and, in Africa, sellers of slaves – all depended on hosts of support workers. To the smaller African traders must be added the armies of retainers that enabled the handing over of the captives to Europeans. Holding people in subjection and moving them is highly labor-intensive.
Figure 5.1 is a unique contemporary image of the trading post at Loango Bay, through which passed 435 individuals in our transactions dataset. This was painted by an officer of the Marie-Séraphique (ID 30941) in 1771. There are three sections to the painting. The top third comprises the merchandise unloaded from the vessel that prospective sellers are inspecting. The middle shows the trading activity, and the lower portion of the image shows chained captives leaving the compound, presumably on their way to the slave vessel. At the bottom right a tall man captive, and a woman slave, the former more heavily chained than the latter, are being led into the compound. Seventy-seven individuals are depicted, only nine of whom are captives traded, or to be traded. Three figures are European buyers, and a dozen appear to be African sellers inspecting merchandise. Most of the figure are guards, attendants to sellers, or menials helping to display the merchandise.
Figure 5.1 Slave-trading compound at Loango, north of the Congo estuary, 1771.
As this suggests, delivering the captives to this point of sale must have been far more labor-intensive than bringing nonhuman commodities of equivalent value to the coast. The focus of historians on elite slave traders as the drivers of the traffic has resulted in the broad-based labor force, without whom the business could not have existed, not receiving the scholarly attention it deserves. As elsewhere in the Atlantic world, down to the late eighteenth century, no one in Africa wanted to be a slave, but acceptance of the enslavement and trafficking of others must have extended well beyond the elite and continued to do so in most of Africa at least until the traffic ended.Footnote 41 Here, too, we find that until late in the slave trade era a full acceptance of the slave trade in moral terms that was shared across, say, European textile workers and gunsmiths and African canoe men. And as discussed in Chapter 3 personal experience of enslavement made no difference to the attitudes of formerly enslaved Europeans and Africans.
Enslaved Africans of the slave trade era were no different to the European norm on this issue, as noted in Chapter 3. Recent scholarship on freed Africans originally embarked at Lagos and Ouidah and then returning to their homeland from Bahia and Sierra Leone has laid bare the movement of slave traders, many of them formerly enslaved persons. Kristin Mann has identified thirty freedmen in Lagos in the 1840s, most of whom were involved in the slave trade to Bahia.Footnote 42 Joaquim d’Almeida, a Mahi slave freed in Brazil and an associate of major slave trader Domingos José Martins, was key to redirecting the movement of captives through Agoué, thirty miles westward, when British pressure threatened Ouidah’s role as a principal embarkation point.Footnote 43 Mann identifies several other smaller shippers. The typical pattern was displayed in the manifest of the Santana (ID 3850) where two or three owners accounted for more than half of the 311 enslaved persons on board and most of the remaining entries were for owners of just one or two captives. The major dealers were often of mixed Euro-African heritage, but few of these were former slaves. Most shippers are best described by Luis Nicolau Pares as “the better off … among the elite of Africans who managed to prosper … a minority within what was already the minority of freedmen.” Lisa Castillo has identified several freedmen employed on slave ships who bought a captive or two in Africa for resale in Brazil.Footnote 44 In Luanda, Dona Ana Joaquina Santos did not allow her former experience as a slave to prevent her from becoming one of the major slave traders in Luanda, or in attempting to redeploy her human property to a sugar-mill outside the city as the city’s role in slave trade declined in the late 1840s.Footnote 45
The phenomenon of freed persons resorting to trafficking in slaves was not restricted to Afro-Brazilians. Several thousand Africans “liberated” by the Freetown courts returned to their homes in Lagos and Abeokuta after 1840. In May 1857, during judicial proceedings against the recently captured Abbot Devereux (ID 4247) bound from Ouidah to Cuba, many Africans on board stated that they had been captured and sold by members of these relocated Yoruba communities.Footnote 46 Reports of such activity increased from 1855 to 1861 as the slave trade from the Bight of Benin reached its final crescendo, thus providing further evidence of the “precariousness of freedom” in the nineteenth-century Atlantic world, in Sidney Chalhoub’s words.Footnote 47 As argued in Chapter 7, British policy toward former slaves, flawed as it was, aimed at turning Liberated Africans into quiescent workers responding to free market forces. British officials, of course, criticized the Africans they had freed from slave ships for working at slaving establishments from Rio Pongos to the Congo River, but the accused might have responded reasonably enough that following market signals and participating in a market economy was exactly what they were doing.
The largest African slave traders were likely in Angola. Linda Heywood and John Thornton have argued that Queen Njinga of Ndongo, a coastal state in northern Angola south of the Congo River, was associated with the embarkation of 190,000 captives between 1624 and 1663.Footnote 48 Such an exodus constitutes nearly half of those estimated to have left West Central Africa in these years and about one quarter of the population of Ndongo in 1623.Footnote 49 We have no record of Njinga interacting with ship captains, however, and many of the enslaved could have come from Njinga’s periodically losing side in the civil wars of the period. But if only half that number comprised Njinga’s share, her long life probably ensured that she was the African ruler who sold the largest number of people into the transatlantic slave trade.
Finally, scholars need to recognize that even beyond the limits of the Portuguese Atlantic, slave traders of African descent had a significant presence outside Africa. As already noted, major dealers often dispatched their offspring overseas for training in languages and the ways of foreign traders with whom they dealt. Some African slave traders extended their activities to the Americas. John Ormond was the son of a Susu woman and a man from Liverpool, England, who had sailed from Liverpool to West Africa as a cabin boy, but had remained first in Rio Nuñez and later the Rio Pongo, where he prospered from buying and selling slaves. After an education in England, John Ormond, Jr.,Footnote 50 returned to the Pongo from London in 1805 to run his father’s slave-trading business.Footnote 51 His own son, John Ormond III, by another Susu woman, traveled to Havana and Matanzas frequently to further his business. John Ormond II was recognized as a plaintiff in a Cuban court and made an out-of-court settlement with a white Cuban. In 1764, Fenda Lawrence, a female slave trader in the Gambia, negotiated a passage to Charleston together with a letter from the captain affirming her free status. She received a confirmation of that status in a second letter from the acting governor of South Carolina.Footnote 52 The role of people of African descent in the slave trade outside the sub-continent has received scant scholarly attention and is largely confined to exploring the roles of Black sailors on slave ships.Footnote 53 The “Black Atlantic” comprised more than enslaved people and Black crew. It is highly probable that north of Angola at least, there were always more Africans involved in the slave trade in non-African Atlantic ports than there were non-Africans located in African slave-trading centers.
Questioning Some Big-picture Interpretations
So far in this chapter we have used data from www.slavevoyages.org to argue that neither side in the negotiations that put millions of Africans into the hold of a slave ship held a permanent advantage over the other. The concept of unequal trade is, however, only one of the strongly held but questionable beliefs about this horrific business that have become “established” in recent research. The balance of this chapter will examine some of the most prominent of these against the backdrop of new data that has become available in the last two decades. First, the conviction that African depopulation, either overall or regional, was offset by the introduction of New World crops, particularly maize and manioc. Second, that while slavery in Africa predated the Atlantic slave trade, exposure to the external demand for both people and commodities fundamentally altered the nature of Indigenous slavery. Third, that African economies were gutted by an influx of cheap European goods. Fourth, that the slave trade destroyed interpersonal trust to the point at which economic growth lagged the non-African world. Fifth, that as the slave trade expanded, slave routes penetrated ever further into the interior. Sixth and last, that as the slave trade came to a gradual close during the nineteenth century this non-Indigenous form of slavery in Africa expanded dramatically. According to this viewpoint, the abolition of the slave trade substantially increased the number of slaves within Africa, and those slaves worked in conditions closer to Americas-style plantation slavery. No single scholar espouses all these views together and it is certainly not possible to mention all those who embrace at least one of them. It is nevertheless worth querying the extent to which these six propositions hold in the face of new evidence embodied in databases on population trends, the size and direction of the slave trade and the ethnolinguistic identities of those who were enslaved in Africa prior to their forced exodus across the Atlantic.
Most Africanist scholars have only partially absorbed the implications of the mass of this new data. In my 1987 book, I pointed to the disconnect between what we knew then about the quantitative patterns of the slave trade on the one hand, and on the other, the positions assumed by most major scholars in the field as they developed their master narratives on why the slave trade happened, and, more particularly, the economic, political, and social impact of the traffic on African societies. In the thirty-seven years since that book appeared, the disconnect between new data and interpretations of relations between Africa and the Atlantic world has made that gap even wider. The more recent data in effect poses a numeracy test that is seldom applied to the work of Africanists. Africanists do draw on the voyage details available at www.slavevoyages.org, as well as the estimates of the volume and direction of the slave trade that the site provides, but they tend not to use this information to address the larger questions about the impact of the slave trade on Africa.
Population Decline?
Did the African population decline because of the slave trade? The discussion of global demographic patterns in Chapter 1, concluded that the overall African population was unlikely to have declined in response to the exodus of people via the Atlantic, the Indian Ocean, and the Sahara Desert, although these slave trades probably prevented the population growth that the rest of the globe experienced between 1500 and 1900. A fresh approach to these demographic issues is possible by distributing the numbers of captives carried off from Africa across modern countries and then computing the ratio of total departures to the population of each country in 1850 (as estimated by Frankema and Jerven).
Table 5.2 shows the results. Three obvious caveats are in order. Boundaries are a later and largely colonial addition, but it is easy to group the geocodes of embarkation points within them for the benefit of the general reader. Nation-states have the obvious advantage of familiarity and there is no intention here of projecting modern consequences of historical events. A second potential problem is that the ratios of people to population do not reflect the full demographic impact of the trade on African populations, given the deaths associated with warfare and kidnapping. We do not know how many casualties slave traders left in their wake. Age/sex data culled from slave ships suggest the very young and the old must often have been killed or left to die, with prime-age males comprising battlefield deaths. The abolitionist Thomas Fowell Buxton suggested that such casualties amounted to one-half of those captured, so 25.5 million, of whom only 12.75 million survived to embarkation, but we will never know.Footnote 54 Final caveats are that the table does not include the trans-Saharan and Indian Ocean traffic; Southeast African states are excluded from the table. In addition, the impact of the trans-Saharan trade on Upper Guinea is not assessed. The two leading slave-trading nations, Britain and Portugal, also happened to be the nations in Europe whose cumulative total of transatlantic emigrants leaving their own shores comprised the largest share of their respective 1850 populations. These two European countries are included in Table 5.2 for comparative purposes, but it should be noted that the large British population of 1851 was of very recent origin. Between 1500 and 1800, it was in the 5–8 million range and the ratio of migrants to populations was accordingly greater.
Table 5.2 Transatlantic movement of people across the Atlantic from regions that became Western African nation-states, and from Portugal and Britain, 1519–1850
| Source country (modern) | Population in 1850 | Cumulative number of people embarked | Column 2/column 1 |
|---|---|---|---|
| Mauritania | 288,461 | 653 | 0.00 |
| Senegal | 931,646 | 202,966 | 0.22 |
| Gambia | 112,067 | 236,971 | 2.11 |
| Guinee-Bissau | 210,508 | 277,433 | 1.32 |
| Guinee | 1,028,675 | 132,406 | 0.13 |
| Sierra Leone | 778,717 | 329,117 | 0.42 |
| Liberia | 330,073 | 159,946 | 0.48 |
| Ivory Coast | 1,033,829 | 62,734 | 0.06 |
| Ghana | 1,981,047 | 1,551,034 | 0.78 |
| Togo | 506,578 | 68,262 | 0.13 |
| Benin | 791,150 | 1,308,401 | 1.65 |
| Nigeria | 13,539,371 | 1,837,883 | 0.14 |
| Cameroon | 2,767,547 | 91,853 | 0.03 |
| Equatorial Guinea | 108,342 | 2,907 | 0.03 |
| Gabon | 297,400 | 98,919 | 0.33 |
| Congo-Brazzaville | 507,653 | 226,454 | 0.45 |
| Congo-Kinshasa | 8,222,558 | 262,107 | 0.03 |
| Angola | 2,428,947 | 4,660,574 | 1.92 |
| Total | 35,864,569 | 11,510,620 | 0.24 |
| Portugal | 3,450,000 | 1,600,000 | 0.46 |
| England | 21,121,967 | 2,000,000 | 0.07 |
Even after allowing for the caveats, column 3 of Table 5.2 suggests that the overall demographic impact of the slave trade varied greatly by region. The greatest effects appear to have been on three areas occupied by the modern countries of the Gambia, Benin, and Angola. As column 3 shows, the area occupied by these modern nations lost more people to the slave trade over 350 years than lived there in 1850. The results are somewhat misleading, however. For Angola, combined populations, and population densities in the neighboring countries of Congo Kinshasa and Congo Brazzaville, were much greater than in Angola itself. Half the captives leaving the two Congo countries boarded ship in Cabinda and Malembo, both of which are now located in the Angolan enclave north of the Congo River.Footnote 55 It makes sense, therefore, to combine the data for Congo Kinshasa, Congo Brazzaville, and Angola. Doing so brings the ratio of total departures to the 1850 populations in this vast West Central African region to 0.46, or about the same as that imposed on Portugal by its migrants over the same period. A similar situation exists for Benin, many of whose captive deportees originated in its heavily populated neighboring region, now Nigeria. Applying a similar adjustment to captives and populations of these two countries shows that 350 years of cumulative departures amounted to approximately one-quarter of the combined 1850 population. This falls between the emigrants/population ratio of Portugal (0.46) and of Britain (0.07). Likewise, the Gambia, whose river was navigable for up to 200 miles for oceangoing vessels, channeled many captives from what are today neighbors, especially Senegal. After these adjustments, Table 5.2 shows the region occupied by Ghana to have been most severely affected by the slave trade, with a departures-to-population ratio of 0.78, far greater than that of any country in Europe in the same period.
The slave trade from what is now Ghana spanned a shorter range of years than held in every other major coastal region in Africa. It began very gradually at just a few hundred people a year in the mid seventeenth century, intensifying only after 1700 before ending abruptly in 1808 in the aftermath of British and Danish abolition of the slave trade. Annual departures over 120 years averaged 13,000. West Central Africa and Benin/Nigeria saw larger numbers leaving annually, but their populations were several times the size of the Ghana region. Even in Ghana, however, the demographic impact of the traffic in those 120 years cannot have been great. Accepting the scholarly consensus that the sub-Saharan African population remained constant between 1650 and 1850, departures of 13,000 a year meant that natural population growth (i.e. births over deaths) was unlikely to have wholly offset transoceanic departures from eighteenth-century Ghana.Footnote 56 But the difference between departures and recovery rates was small. In all other regions affected by the transatlantic slave trade shown in Table 5.2, including Nigeria/Benin and the grouping of three West Central African nations, the average annual departure rate fell well below the Malthusian natural population growth recovery rate. This assessment does not reckon with the high male ratios in the traffic. Only one in three hauled off from Africa were female, a ratio that could only have enhanced recovery rates in the source population. The slave trade could therefore not have eroded populations in the region, unless the act of creating a captive resulted in extremely high death rates. For most of Africa, as with Europe, departures across the Atlantic did not mean population decline. The findings here offer new confirmation of a long-held scholarly consensus.
Comparisons between sub-Saharan Africa and Western Europe might not seem appropriate at first sight given the devastating impact of wars and slave raids on the former. But while the Atlantic slave trade carried off only Africans, violence and social disruption were not confined to Africa. Transatlantic migrants from Europe were not enslaved, and before 1820 were far fewer in number than slaves carried off from Africa, but within Europe widespread warfare, dispossession, and violent persecution, usually associated with religion, caused disruption, and triggered migration. Peoples of European descent not only organized a slave trade in Africans but between 1500 and 1850 were at war with each other almost continuously. Such wars killed 13 million combatants within Europe (or 7 million in Western Europe alone), and many more millions of civilians. The Thirty Years’ War, part of which was fought in West Central Africa was particularly brutal. African casualties, in all conflicts, including noncombatants, were likely lower than in then contemporary Europe.Footnote 57 But Europe’s population did grow between 1650 and 1850, as indeed did that of Asia. The overseas slave trades may have contributed to Africa’s inability to share fully in the global increase in population of these two centuries, though disease factors within Africa would also have played a role.Footnote 58
Impact of the Slave Trade on African Slavery
A second potential impact of the slave trade is that while slavery in Africa predated the Atlantic slave trade, exposure to the external demand for both people and commodities fundamentally altered the nature of Indigenous slavery. The core idea here can be found in Walter Rodney’s influential work.Footnote 59 For him, slavery scarcely existed in Africa before European slave ships arrived. Paul Lovejoy, on the other hand, accepts its prior existence but the transformation he has in mind mirrors Rodney’s argument on the slave trade’s effect on African social structures, including slavery. The third edition of Lovejoy’s Transformations in Slavery published in 2012, fully embraces the data in www.slavevoyages.org first made available at the end of 2008, the appearance of which may indeed have been one reason for this new edition. But nowhere does the author recognize that the now widely accepted profile of the slave trade projected by this new data undermines his core argument. The transformation that Lovejoy sees happening in Africa was one in which various forms of Indigenous slavery turned into a “slave mode of production,” a phrase without a precise meaning, but which here is taken to refer to slaves employed in the production of commodities for a market.
We now know that West Central Africa had the longest association of any African region with an external slave trade – almost three centuries – and was the source of 45.7 percent of all captives dispatched to Atlantic destinations. The sheer volume of captives carried away from this one region over so long a period suggests that any change in African slavery should have happened first and be most apparent here. Yet this area that supplied by far the most enslaved people shows the least evidence of changes in the nature of slavery established in the sub-continent. The Congo Basin and Angola provide little evidence of captives employed in commodity production either before or after the external slave trade in the region came to an end. Coffee production in Angola expanded very slowly, and in the words of one recent reassessment based on Angolan slave census data “the widely accepted theory that the size and composition of slave populations in Western Africa changed under the influence of the maritime slave trade” is questionable.Footnote 60 To this we could add that except for limited parts of West Africa, the introduction of the “slave mode of production” was unlikely to have been an inevitable or generalized effect of transatlantic slave trading. Lovejoy’s latest work tracks the connections between jihad in Africa and the Atlantic World, but overall, Paul Lovejoy might be seen as the least African-centered of the major Africanists in his ideas, with John Thornton perhaps occupying the opposite pole.Footnote 61
Were Africans Subjected to Unequal Trade?
Were African economies gutted by an influx of cheap European goods, a third potential impact of the slave trade? Just as the scholarship on the impact of slavery and the slave trade on Western societies has come to foreground the ideas of Eric Williams and his now eighty-year-old book, Capitalism and Slavery, so assessments of the impact of the external slave trade and slavery on Africa are underpinned by the only slightly more recent works of Rodney. His second book focused on the slave trade’s destruction of African social structures and economies. Some of the most prolific and influential scholars on precolonial Africa’s relations with the rest of the world since Rodney’s early death continue to show more than just traces of his core ideas. And outside the scholarly community Rodney’s influence had an immediate impact. Leaders of the Suriname maroon communities on a trip to Ghana in 1971 asked Akan leaders why their forefathers had been sold into slavery across the ocean – indeed “they felt some retribution was owed them.”Footnote 62 Within the scholarly community we consider the work of Paul Lovejoy, Toby Green, Joseph C. Miller, and John Thornton in the light of the expanded version of slavevoyages.org.Footnote 63 Most Africanists see these consequences as strongly deleterious, socially as well as economically. If, as argued above, Africans and Europeans traded for the most part as equals, it is not likely that the merchandise exchanged for captives was inferior. Walter Rodney’s account of the exchange now looks unsophisticated, especially its view of African traders implicit in his description of the merchandise they accepted as “cheap gin, cheap gunpowder, pots and kettles full of holes and assorted rubbish.”Footnote 64 A half-century later, this now appears both inaccurate and patronizing to African traders and consumers. Like buyers and sellers in most locations, Africans were careful in their evaluation of quality and precise in defining their needs.Footnote 65 Yet Toby Green has recently reworked Rodney’s position. Europeans, he argues got more from transactions on the African coast than did Africans. They dumped their goods in Africa (“the words dump, “dumping,” and “dumped” appear frequently in his text) and were able to extract “capital and surplus value through circuits of long-distance trade [that] came to hold greater value than these material objects themselves.”Footnote 66
This statement is not supported with documentary evidence even though it is amenable to testing. Apart from the abundant papers of merchants, there is an extensive secondary literature on what trade goods cost, first at source, and second on the African coast as well as what captives cost on both sides of the Atlantic along with the prices of plantation produce in both the Americas and in Europe.Footnote 67 Without defining “dump,” “surplus value,” and “capital,” much less providing even an approximate idea of their size and source, Green repeatedly restates the broader thrust of Rodney’s argument: trading on the coast devastated African societies and in so doing generated Western economic growth. But given all the abundant documentation on transactions on the African coast now available, the reader is entitled to know how much “surplus value” was extracted and from where.
Green argues that vast quantities of capital were accumulated by Europeans in their trading activities with Africans. We could assume that the author is looking at differences in the price of slaves, gold, and African produce between their purchase in Africa and their sale in the Americas and Europe. But the literature makes clear that such differences are accounted for by transportation costs, not “surplus value,” capital accumulation, or profits.Footnote 68 We should ask exactly what the sources were of “growing imbalances in capital accumulation.” Green disparages “formalist economic rationality,” but what he offers in its place to explain what was, for all participants, a transatlantic market for slaves is unclear. Most surprisingly, Green makes little use of Curtin’s extensive work on Senegambia, which shows the terms of tradeFootnote 69 turning strongly in Africa’s favor as the Industrial Revolution drove down the prices of all manufactured goods between the 1780s and 1850. It now seems that this pattern holds for the whole of West Africa and for the whole of the eighteenth century – the very period when Europeans were supposedly extracting most from Africa.Footnote 70
The counterarguments to this position are straightforward. First, Green ties “dumping” by which he seems to mean selling cheaply, together with widespread depreciation of African commodity currencies such as the shells in the book’s title.
Repeatedly throughout the book Green sees cheap goods flooding the African domestic markets but points out that they were in fact used as currencies in many African states – cowries, and pieces of cloth as well as gold, iron bars and copper. At the same time, Green argues these large quantities of commodity currencies stimulated rampant inflation far beyond that experienced in sixteenth and seventeenth century Europe as American silver and gold arrived. The economic tension in linking these two arguments is striking – to the point of impossibility. If currencies are depreciating because of inflation, then foreign-sourced goods cannot possibly be cheap for the domestic consumer. One argument makes sense by itself and so does the other, but not both at the same time. The reader is never told of the relative strength of each effect. If Africans found themselves spending more domestic currency to buy a given foreign good (depreciation) and at the same time the foreign seller was lowering prices (dumping), then the transaction price to the consumer might well have remained constant.Footnote 71 In addition, Green goes to great lengths to establish the economic, political, military, and cultural equivalence of sub-Saharan Africa and Europe at the outset of Atlantic contact between the two in the fifteenth and sixteenth centuries. This is plausible, but the amount of bullion flowing into Europe from the Americas between 1520 and 1800 was vastly greater than the total value of merchandise imported to Africa in the same period. Even if all African imports comprised shells, textiles, and gold (in other words currencies in Green’s sense) how is it that devaluation on the scale posited by Green for Africa did not occur in Europe where on a per capita basis the boost to the money supply from American gold and silver was an order of magnitude greater?Footnote 72
A second argument against Green’s extraction of surplus position is that the slave trade does not appear any more profitable than other branches of long-distance trade. Averaging out wealthy merchants and planters in the Caribbean with bankruptcies in their respective businesses, investments in the slave trade and the plantation colonies were no more profitable than other sectors of comparable risk in Europe’s complex economies.Footnote 73 Africa accounted for 3.6 percent of British exports in 1784–1786 and less than 1 percent in 1824–1826. Adding these African figures to exports to the West Indies to measure the impact of whole slave sector increases the share to 14 percent and 11 percent respectively.Footnote 74 Trade statistics for other European nations show much smaller shares of their exports going into the Atlantic slave systems. As shown in Chapter 3, some, like Germany and Italy, had shares close to zero; others, such as Spain and Portugal, had been drawing on the Atlantic slave system far longer than Britain without exhibiting any trace of a step change in economic development. And the surge in productivity in the domestic French economy occurred well after significant plantation activity had collapsed in the wake of the Haitian Revolution.Footnote 75 The question posed in Chapter 3 is again relevant here. Is it possible that Western Europe’s economic development might have occurred without an Atlantic slave system, drawing on sub-Saharan Africa? Whether we take Africa alone or the whole Atlantic trading system, the answer must remain very likely yes.Footnote 76
Did the Slave Trade Destroy Trust between Africans?
A fourth and quite different malevolent impact of the slave trade on the African economy focuses on loss of trust in commercial transactions. In two papers, the second coauthored with Leonard Wantchekon, Nathan Nunn has argued that countries that supplied the most captives during the slave trade era are also the most underdeveloped today. The second paper provided a possible explanation: Slave trading undermined interpersonal trust and thereby inhibited economic development so that “Africans whose ancestors were heavily raided during the slave trade era are less trusting today.”Footnote 77 But raiding was not the major source of captives. Thornton’s survey of the evidence on enslavement indicates that slightly more than one-third were prisoners of wars between African states or civil wars, and a further one-fifth entered the traffic via debt or judicial procedures. Extensive conflicts in Europe, already discussed, as well as in the Mediterranean and the Americas (within both Indigenous and settler societies) do not appear to have generated a similar lack of trust among modern descent groups. In Thornton’s survey kidnapped individuals and those sold by relatives together – activities more likely to undermine trust – accounted for less than half of those dispatched. Also, scholars have yet to entertain the possibility that some unknown fraction of the exodus were already enslaved, although as above, we do have cases where African-owned “domestic” slaves in Liberia became outraged at their sale to overseas buyers.
A second problem with the Nunn/Wantchekon position is that it ignores the critical role of trust in both Indigenous African trade networks and in the organization of the slave trade itself both during and after the slave trade era. Curtin, Lovejoy, Richardson, and others have established the centrality of trust in the business of slave trading. In many parts of sub-Saharan Africa human pawns secured goods advanced on credit. European and African slave traders “appropriated and adapted” what was a local practice for the slave trade. A chain of credit could extend far into the interior and was based on the expectation that a debtor would pay off a creditor within an allotted time. Human pawns or not, this continent-wide system could not have worked without trust. This was one of the reasons that the shift from the slave trade to what quickly became a much more valuable trade in commodities, especially palm oil and groundnuts, occurred so seamlessly in the aftermath of the demise of slave trading. The trading networks, including the practice of human pawnship, were already in place.Footnote 78 Seymour Drescher states that Africanists have shown that “there were no more complex chains of trust, interlocking trade diasporas, long-term lines of credit … and other recipes for economic transactions, than in the world of the slave trade,” but this assessment also applies to the wide range of African products that gradually took the place of the trade in people after 1807.Footnote 79
A third and more fundamental problem with the diminution of trust issue has to do with the econometrics that the two scholars employ in reaching this conclusion. In the last two decades the most prestigious journals in economics have published a range of studies covering a wide range of topics in world history all sharing the common theme of comparing an often-distant past with the same area in the present day. This research is now grouped under the label “persistence studies,” all of which tend to come to the same conclusion: specifically, modern outcomes strongly reflect characteristics of the same places in the distant past. Morgan Kelly has subjected twenty such papers from the most prestigious journals, including Nunn and Wantchekon’s work, to robustness checks.Footnote 80 He finds that in more than half the studies, including the two Nunn papers, the high scores on significance tests are likely the outcome of underestimating standard errors or of fitting spatial trends. Kelly is at pains to distance himself from any intention to “disprove” the results of the twenty-five studies, but his corrective procedures do mean that the t-statistics in both slave trade papers are no longer significant. Indeed, Kelly’s standard error adjustments and robustness checks generate results for the two slave trade papers that are among the most dramatic corrections in the twenty studies he examines. Finally, in this quantitative critique it is worth pointing to the miniscule ratio of the value of external trade per person in the Western African sub-continent shown in Table 5.3. Is it credible that the nature of a trade, even the slave trade, could have had such a profound long-range impact on attitudes to risk a century and a half after the cessation of that trade, especially given the very modest per capita values of the traffic compared to the rest of the Atlantic world?Footnote 81
Table 5.3 Value of imports per person in select regions in the Atlantic world c. 1800 in pounds sterling
| The Netherlands* | France | United Kingdom | United States | Western sub-Saharan Africa | |
|---|---|---|---|---|---|
| Imports (millions £) | 11.4 | 20.6 | 56.0 | 20.6 | 25.2 |
| Population (millions) | 2.1 | 27.5 | 10.5 | 5.46 | 35.9 |
| Imports per Person | 5.4 | 0.75 | 5.3 | 3.77 | 0.08 |
Was There a Slaving Frontier that Moved Inland over Time?
As the slave trade expanded, did slave routes penetrate ever further into the African interior? This constitutes the fifth questionable view of the slave trade’s impact on the sub-continent. Joseph Miller made the strongest case for this pattern in Angola, though other historians had established this as a central theme of their work previously. Among the older generation of Africanists, only Patrick Manning questioned the emergence of more remote provenances of captives over time.Footnote 82 New data generated by Philip Misevich and Daniel Domingues da Silva provides a basis for revisiting Manning’s position.Footnote 83 Further evidence of the proposed pattern is presented below in Chapter 7’s maps. Generally, historians have taken insufficient notice of the time profile of regional departures of enslaved people. West Central Africa not only saw far more captives leave its shores than any other of the eight regions depicted on slavevoyages.org, but the volume of its departures built up steadily over time. From 150,000 in the half-century 1551 to 1600, the region reached its all-time peak of nearly 2 million between 1801 and 1850.Footnote 84 Just as with the transformation thesis, the impact of the slave trade on the slaving frontier and slave supply systems generally surely would have been at its most severe in West Central Africa during the first half of the nineteenth century. Miller tracks the slaving frontier as moving intermittently eastwards between 1730 and 1830.Footnote 85 Given the nearly 2 million captives carried off from the region in 1801–1850, his argument might lead us to expect that the frontier would have been at its furthest point from the “floating tombs” on the coast in say the second quarter of the nineteenth century. He also links this movement with state formation and disintegration. But, as with the transformation thesis, the historical evidence of this in West Central Africa is questionable.
Since 2010 when www.african,origins.org went live,Footnote 86 students of the slave trade have had new material at their disposal to evaluate and refine interpretations of where people were enslaved. As noted in Chapter 1 and analyzed more fully in Chapter 7, one of the by-products of the British, then international, campaign against the slave trade after 1807 was the creation of registers of Africans released from detained slave vessels or held in coastal barracoons. Details of 95,000 individuals with 65,000 unique names are now available for analysis, plus 11,264 included in the Portuguese slave census counts of the mid 1850s, most of whom were originally intended for the Atlantic slave trade. For 35,300 individuals in this combined group, it is possible to link the African name with a language grouping. Such links throw new light on the sources of captives in three broad regions of the sub-continent for the period, 1808 to 1856: Sierra Leone and the Windward Coast (now Liberia) treated as a single region, and the hinterlands of the Bight of Benin and Biafra, including in the latter case, the Cameroon Highlands.Footnote 87 With the help of the www.ethnologue.com database of languages, we can locate the major groups in relation to their ports of embarkation.
For Sierra Leone and the Windward Coast Philip Misevich has drawn on this material for the period 1808–1844. The time profile of departures from these areas was quite different from that of the overall traffic. Although closer to both Europe and the Caribbean than most of sub-Saharan Africa, for two centuries they supplied relatively few transatlantic captives, but as the overall traffic peaked after 1750 slave traders moved into these regions to the extent of dispatching 422,000 people into New World slavery in the second half of the eighteenth century. After 1800, the volume of the traffic fell by 50 percent, with the creek systems of Rio Pongo and Galinhas replacing the offshore islands, Cape Mount, and the Sierra Leone, all of which were now more vulnerable to British attacks. Nevertheless, the half-century total was still the second highest of the slave-trading era. In this region at least, the Miller model appears to explain shifts in origins. Drawing on Barry Higman’s analysis of the British Caribbean’s slave registration data, plus a range of more conventional contemporary observations, Misevich finds a much wider range of origins before 1800 than after. He also notes much higher prices for captives in the earlier period. After 1800, lower prices and lower volumes meant that the local Mende/Sherbro language speakers predominated, with two-thirds of all captives embarked linked by language to locations within 100 miles of the coast. More recently, Jorge Felipe Gonzalez has tightly tied the rise and fall of the Kingdom of Galinhas under King Siaka to embarkations of captives in southern Sierra Leone.Footnote 88 Galinhas was thus one of the clearer, if smaller scale, examples of the transoceanic slave trade driving state formation. This pattern of proximity to the coast also emerges for a different segment of Upper Guinea which is now Guinea-Bissau and Guinea for the sixteenth and early seventeenth centuries.Footnote 89
The slave-trading profile of the Cameroons shared some of the characteristics of Upper Guinea. Although some of its captive people had always left via Old Calabar, direct departures from the “Cameroons Estuary” or “Cameroon River” jumped after William Davenport, one of the largest Liverpool traders, moved a significant part of his business there in the late 1760s.Footnote 90 Identification of the ethnolinguistic links of 929 captives embarked between 1822 and 1837 indicate a more scattered range of linguistic origins but with four groups, Tikari, Douala-Bimbia (admittedly a coastal location rather than a language), Banyangi and Bakossi together, accounting for 60 percent of the total. But only one-third of the total likely traveled more than 125 miles. The coastal regions and grasslands of the northwest highlands supplied most captives, with much of the Cameroons, particularly what is now the French-speaking segment, not represented at all in this sample. As in the Igbo and Ibibio-speaking areas to the west, before, during and after the slave trade, political structures here were “stateless,” and warfare was not a major source of slaves.Footnote 91
The evidence for West Central Africa is much less clear-cut. As we have seen, this was by far the largest sub-Saharan African region that felt the impact of the transatlantic slave trade. Domingues da Silva has identified the languages of 7,612 individuals who were enslaved here in the nineteenth century.Footnote 92 He has been able to group his sample into twenty-one languages and 116 ethnicities distributed across the region’s interior. His sample is small, but drawing mainly on vessels captured at sea, probably random. However, some of his identifications are questionable. The largest group in his sample, comprising one-sixth overall, is Ndongo. They are labelled as Kimbundu speakers, but a careful survey of eighteenth- and nineteenth-century contemporary sources by François Bontinck has them located in the vicinity of Lake Mai-Ndombe – far away from Kimbundu speakers. This region is 500 miles from the sea, rather than the 250 or less assumed as the center of linguistic gravity for Kimbundu speakers. We do not have the numbers to correct the distribution, but it does seem likely that the pattern established for West Africa of shorter distances from enslavement to the sea was, as a minimum, less pronounced in Angola.Footnote 93
For the Bights of Benin and Biafra west of the Cameroons after 1807 we now know that Yoruba names comprised 88 percent of those leaving from points lying between Keta in the west and the Kingdom of Benin in the east. Igbo names were not quite so dominant among those departing from ports located from the Niger Delta to Old Calabar coastline, but still comprised 69 percent of the total embarking. The maps in Chapter 7 use the www.ethnologue.com database to locate the geographic central point of these two languages and thus gives us a crude proxy for distance traveled on average by the typical Yoruba or Igbo captive. For the Yoruba this turns out to be a point northeast of Ibadan. No less than 87 percent of those forced to board a slave ship in the Bight of Benin passed through Lagos, Ouidah, and Grand Popo.Footnote 94 For the Igbo the modern central language point is 100 miles due north of Bonny – the major embarkation point in the Bight of Biafra. Of course, we cannot be sure that these modern language coordinates hold for the first half of the nineteenth century, but the major population shifts in the last two hundred years have been towards cities, not toward the coast. Lagos is already a mega-city, but cities in the interior of Nigeria have also grown rapidly. A striking feature of the identifications is how few could have come from north of the Yoruba–Igbo belt. Islamic, Hausa, Fulani, and Nupe names comprise less than 9 percent of the Bight of Benin embarkations and just 2 percent of the Biafran group. Non-Yoruba and non-Igbo were mostly people from either east of Yorubaland, such as Akan, Ewe, Fon, or Gbe – or west of Igboland, the Efik, Ibibio, and some from the Upper Cross River. The Ethnologue central coordinates for all these languages lying to the east and west of the Igbo and Yoruba are much closer to the coast than are the equivalent Igbo and Yoruba geocodes.
But while the thrust of the new evidence is that most captives came from regions relatively close to the coast, there is at least one possible exception. After 1740 as we have seen in Chapter 3, Liverpool and Bonny, an island in the Niger Delta, developed close-knit trading ties that catapulted both to an extraordinary sixty-year dominance in the West African segment of the Atlantic slave trade. Liverpool dispatched more slaving voyages than any other port in the Atlantic world, while Bonny dispatched the second largest number of captives of any African port after Luanda.Footnote 95 Three-quarters of Bonny’s captives left Africa on Liverpool ships between 1761 and 1807. The port’s access to the interior via the Imo and Niger rivers was unparalleled, giving it effectively the largest slaving hinterland in sub-Saharan West Africa. But even here most of the Niger flowed through lightly populated regions and Igbo peoples from the forest zones further south and east always dominated Bonny’s deportees. The Congo River was potentially of similar importance, but there was no port on its estuary that played a role like Bonny’s. Instead, captives left Africa from several embarkation points north and south of the Congo.Footnote 96
Did the End of the Slave Trade Drive up the Number of Enslaved Africans within Africa?
The final ingredient of this overview of the slave trade’s impact on Africa is that the suppression of the slave trade increased the number of slaves held within Africa. The value of sub-Saharan African produce exports certainly came to exceed the value of captive embarkations in the immediate aftermath of slave trade suppression.Footnote 97 The shift to a produce trade happened first in the Bight of Biafra and last in the West Central Africa region. But the African labor that made this possible worked under a variety of labor regimes ranging from small farmers to slave owners. Perhaps slavery within Africa expanded temporarily as the shock of the 1826 Anglo-Brazilian Treaty took effect in 1830 and again as Brazil imposed serious measures to end the traffic twenty years later. Yet there is no evidence that the supply lines of the hinterlands that sent approximately 800,000 captives from Africa to Brazil during the two last crescendos of the Brazilian traffic – 1826–1830, and 1846–1850 respectively – suddenly found alternative markets in Angola and Benin in the succeeding quinquennia of 1831–1835 and 1851–1855.
We do not have sufficient census and other data to conclusively evaluate the impact of slave-trade suppression on slaveholdings within Africa, but we do have some evidence of movements of the enslaved from one part of Africa to another. We know that groundnuts, palm oil, and palm kernels comprised the major Western African commodity exports in middle decades of the nineteenth century.Footnote 98 Minor crops such as coconuts, wax, coffee, cloves, and an assortment of dyes and timber were much less important than these three. Moreover, slavery was only one form of labor used to produce these crops. Some slave labor was employed in the southern part of the Upper Guinea’s groundnut region but was much less prevalent in the palm oil belt in the hinterlands of the Bights of Benin and Biafra.Footnote 99 Enslaved people could also be found cultivating cloves in Zanzibar, coffee in São Tomé, and sugar in the Mascarene Islands.Footnote 100 And in regions producing for domestic markets, the Sokoto Caliphate used slave labor to produce cotton, indigo, kola and shea nuts, grain, rice, and tobacco. An intra-African traffic in people did develop to the groundnut regions, and to São Tomé, as the transoceanic slave trade came under attack, and this, too, was susceptible to British disruption. Between 1820 and 1844, the British detained eighty-two small shipments of enslaved people destined for the groundnut areas of Upper Guinea, many in flotillas of canoes, carrying a total of 965 enslaved people, and a further sixteen voyages sailing for São Tomé with 697 on board. In the transatlantic trade the loss ratio due to captures was 20 percent.Footnote 101 If we assume that loss ratios were the same in these branches of the intra-African traffic, then we can estimate that Upper Guinea’s groundnut region was the intended destination of 4,825 departures, and 3,485 left West Central Africa for São Tomé in these years.Footnote 102
Such figures are, however, almost trivial when compared to the adjustment to suppression of the slave trade on the western side of the Atlantic. When plantation owners in the US and in Brazil had difficulty obtaining enough forced labor from Africa, after 1790 in one case and 1850 in the other, they quickly resorted to the intra-American traffic to supply their needs. Planters in Brazil’s expanding coffee sector immediately turned to domestic sources of enslaved people via an intra-American traffic. They purchased 200,000 captives from the rest of the country in the three decades down to 1881. More than triple this number were moved from the Upper to the Lower Southern States of the US in the seventy years after 1790.Footnote 103 And within the British Caribbean, a further 25,000 were moved from long-established islands such as Jamaica to Trinidad, Grenada, and British Guiana between 1807 and 1833.Footnote 104 Mid nineteenth-century sub-Saharan Africa could well have contained at least as many bond-persons as Brazil and the United States put together, but the only substantial movement of slaves had been to the offshore islands such as the Mascarenes and São Tomé, where plantation complexes were established, and to the Cape of Good Hope.Footnote 105
Table 5.3 displays the available data on populations and movement of people. Even if we include the enslaved labor introduced to grow cotton, indigo, kola and shea nuts, grain, rice, tobacco, and onion in the Sokoto Caliphate, the trade in enslaved persons within Africa could not have come close to the 850,000 enslaved people forced to move within the Americas over the same period. Nor could it have matched the size of the earlier traffic to Réunion and Mauritius, or the now expanded deportation of people across the Indian Ocean and the Sahara Desert that lasted until the early twentieth century. There is no evidence of an extensive intra-African traffic in people – either maritime or land based – during the nineteenth century to match these numbers either before or after European colonial powers carved up the continent. We can conclude that no surge in slave-produced African commodity exports in these years from any of these African regions was comparable to what happened in the Americas in the sugar, cotton, and coffee sectors in response to the transatlantic slave trade closing. This comparison suggests it is highly unlikely that the ending of the Atlantic slave trade triggered a massive increase in African slaveholdings.
Nevertheless, a substantial shift within Africa was triggered by suppression of the transatlantic slave trade, but it was not toward regions producing commercial produce for Europe, as happened in the Americas. Rather it was directed to locations that the British intended as sites safe from slavery – just how safe is taken up in Chapter 7. As part of the efforts to suppress the slave trade after the passage of the 1807 abolition acts of Britain and the US, 96,000 Africans were disembarked in Sierra Leone, 24,100 in St. Helena, 6,700 in Cape of Good Hope 5,100 in Liberia, and 2,200 in Luanda. While, as we shall see, many did not remain in these locations, this was not an economically rational deployment of labor by the standards of any of the imperial powers on the global stage referenced in Chapter 1, or indeed by the standards of early modern Europeans. On the contrary, did the labor power of 134,000 Africans diverted from the plantation Americas by the British navy constitute yet another significant foregone opportunity for British plantation owners? Perhaps not. Later in the century when Europeans divided up the continent, shedding most of their humanitarian sensibilities in the process, the position of African commodity exports relative to the rest of the world (and certainly the Americas) did not improve. From a long-run perspective, the outcome of the colonial partition of Africa was no different from the attempts of Europeans to bring sugar-growing technology to the African mainland in that compared to the Americas at least neither brought substantial gains to the imperialists.Footnote 106
Suppression of the Atlantic slave trade might have had some small effect on slaveholdings south of the Sahara. Of course, overseas exports increased but not because more slave labor had become available. And major slave societies did appear in the form of the Sokoto Caliphate and in parts of French West Africa, but from an international perspective these made only modest contributions to African overseas exports. If slavery within Africa increased as the slave trade ended, that, too, must have been of modest proportions if it happened at all.
In summary, this rather lengthy review of how the positions of some leading Africanist scholars are at odds with recent scholarship on the slave trade points to a depressing conclusion. Major Africanists have exaggerated the impact of the Atlantic world on sub-Saharan Africa. They see cheap European goods devastating African economies, fundamental social practices such as slavery irrevocably altered, trust between strangers permanently undermined, states rising and falling in response to the slave trade, with the impact of all steadily increasing over time. What all this means is that the effect of much of the post-Rodney historiography as represented by Miller, Lovejoy, and Green is to severely reduce the agency of Africans in shaping their own history. In a sense, it suggests that for those not dispatched to the Americas, the impact of the Atlantic world was different from, but nevertheless comparably malevolent, for those that were forced to leave. In this sample of the post-Rodney literature Africans remain predominantly victims. Certainly, the argument that “a western ‘capitalist mode of production’ was fed by the African ‘slave mode of production’ over the course of the nineteenth century,” becomes possible only by ignoring seventy years of painstaking data collection on the eighteenth- and nineteenth-century movement of commodities and people around the Atlantic world.Footnote 107
Of the four influential scholars who have contributed most to current views of the effect of the slave trade on Africa, John Thornton is the one who has offered arguments most consistent with the wealth of new evidence that has recently become available.Footnote 108 He has long decentered the external slave trade in his interpretation of African history. In the words of David Gordon, Thornton sees “conquest and expansion from the sixteenth to the nineteenth centuries … driven by efforts to control textile, copper, and salt production” within Africa rather than the external slave trade.Footnote 109 His long list of publications on specialist topics such as religion, warfare, and the transatlantic influence of the African diaspora, interspersed by his two major works of synthesis, have often been at odds with other major Africanists discussed here.Footnote 110 His most recent book generally avoids the big questions discussed in this chapter. Instead, it comprises a meticulous piecing together from primary sources of the political and military histories of the states of West Central Africa from Loango in the north to Banda and Ambuela in the south and stretching to Lake Tanganyika in the east.Footnote 111 Nevertheless, without drawing much on www.slavevoyages.org and related research it is more consistent with recent findings than are the works of other major figures in the field.
For Thornton, sub-Saharan Africa was abundant in land and wealth was expressed in ownership of people. Africans by and large took care of themselves before and during the Atlantic slave trade era without developing any dependency on overseas commerce whether in merchandise or people. As early as 1600 he can identify “some 200 known independent states along the African coast or in the inland areas where people who came into the slave trade lived.” The involvement of these political entities in trafficking people was a choice, sometimes only occasionally exercised, as with the Kingdom of Benin. And as Thornton points out “most systems of African law that we have knowledge of today supported slavery and the sale of people into slavery … [which] … was both widespread and legally accepted, albeit with important reservations and safeguards.”Footnote 112 Thornton makes frequent reference to transoceanic trade and politics as well as to population loss, but except for Portuguese diplomatic, military, and religious interventions, the outside world is rarely the driver of, or even a participant in, African economic, social, and cultural change. In addition, Thornton is a strong proponent of the economic and demographic equivalence of Africans and Europeans at the onset of Atlantic slave trading, by which he means the economic well-being, life expectancy and military conflicts experienced by ordinary people in their respective sub-continents.Footnote 113
John Thornton is not an economic historian and, except in his discussion of demographic trends, largely eschews numbers, but there is substantial support for his Africa-centered view of sub-Saharan history in the newly available quantitative evidence. Table 5.3 compares Western African imports with that of most of its main transoceanic trading partners – only the Brazilian data is missing. Just prior to British and US abolition of the slave trade the quantities of both slaves and African produce – mainly palm oil and ivory – as well as the prices of both commodities were close to their all-time peaks. African access to offshore merchandise was thus also at its peak and especially so given that, as already noted, the terms of trade were also swinging in Africa’s favor. Yet on a per capita basis and, in comparison to other large regions bordering the Atlantic, the imports that these exports enabled were trivial even at their peak. African ratios fell well below those of their chief overseas trading partners and amounted to just over one-tenth of those of the West European nations that were least dependent on international trade, namely the French. In terms of per-capita income, Gemery and Hogendorn have calculated subsistence costs of £1 a year for eighteenth-century West Africa, so perhaps £1.20 after allowing for a margin beyond subsistence.Footnote 114 Thus, a little over 6 percent of average income was spent on ocean-borne merchandise and for most of the slave trade era before and after this peak late eighteenth-century period, import ratios would have been much smaller than this. As Thornton points out, Africans could feed, clothe, and house themselves, as well as sustain the savings and investment necessary to support these activities to a greater extent than European nations and their offshoots.
Of course, large quantities of foreign textiles, metals, alcohol, guns, and decorative goods did enter Africa in the slave-trade era. Their distribution across African regions was clearly skewed toward regions where the slave trade was most active, and Dahomey, accounting for one-fifth of the captives leaving the Bight of Benin in the mid eighteenth century, probably had a ratio of imports to total product as high as 15 percent. Yet Manning’s careful assessment of the Dahomean economy concludes that while “the export of slaves [cannot] be dismissed as lacking real impact [neither] can it be taken as Dahomey’s primary economic activity.”Footnote 115 But such a skewed distribution also implies that most ordinary Africans living away from the major coastal trading sites would have had little experience of overseas merchandise.
Thornton addresses the question of why so many African states voluntarily participated in the overseas slave trade when that traffic had a clearly deleterious impact on their populations and general security. His answer lies in the political fragmentation of the sub-continent in coastal regions south of the Senegal and Niger valleys. Hundreds of mini-states or sometimes ethnolinguistic groups that could coalesce temporarily into larger polities or federations were more nearly the norm than were major state formations thriving on slave trading.Footnote 116 The lack of enduring overarching state structures meant that the potential pool of captives was probably greater in sub-Saharan Africa than at any point in the Old World since the Mongol expansions in Central Asia. In early nineteenth-century Upper Guinea a missionary observed that “[E]ach town has its own independent chief, who in time of war sometimes unite for the sake of mutual assistance and defence.”Footnote 117 But to return to the demographic issue, might not an additional response to Thornton’s question be the very fact that while local impacts of the slave trade could be and were devasting, the demographic consequences of the business on larger regions and in the longer term were less so? There is no clear evidence of a language group disappearing from West Africa during the slave trade era.Footnote 118
The northern interior of sub-Saharan Africa did contain larger states centered on the Niger, Gambia, and Senegal rivers. Thousands of non-Muslims were enslaved by the Islamic states Kaabu, Ségou, and eventually Futa Jallon, among others. Yet we know that after 1807 16 percent of those leaving Upper Guinea had Islamic or Arabic names, a ratio that shrinks to 5 or 6 percent if we broaden our base to West Africa. A simple backward projection into say, the sixteenth and seventeenth centuries, when very few Muslims could have entered the transoceanic trade, would suggest that probably fewer than two in every hundred of those boarding a slave vessel over the whole period could have been Islamic. Moreover, closer inspection of the nineteenth-century sample reveals a significantly higher ratio of men compared to the Yoruba, Igbo, and Mende diasporas. The jihads raging across central and western Sudan in the first half on the nineteenth century would certainly have raised the count of prisoners of war among those leaving Upper Guinea and Bight of Benin ports. In addition to the small numbers of Muslims, we can assume that the high ratio of men among them worked further to reduce the possibility of community-building and reproduction in those regions of the Americas receiving the Muslim influx. These sex ratios also comprise strong evidence of internecine Islamic conflict, some certainly jihadic in nature.Footnote 119 Such patterns were not the West African norm. Fewer than 9 percent of total transatlantic departures from Africa left from Senegambia and Sierra Leone together, and the majority of these came from coastal societies, not the interior.
Considerable scholarly effort has been put into tracking the African origins of captives brought to or intended for the Americas and how those captives came to be enslaved in the first instance. Thornton fully exploits the works of contemporary observers such as Christian Georg Andreas Oldendorp, Alonso de Sandoval, and Sigismund Köelle, as well as the many Portuguese officials in their ventures inland from Luanda. Frederick Bowser’s book on early colonial Peru is also foundational.Footnote 120 But the range of “nations” that they found is such that perhaps the real challenge is to identify African ethnolinguistic groups that were not well represented in the Americas, other than known enslavers like the Vai, Fon, and Vili. If war and banditry were prime sources of captives, then as Thornton concludes, African states were “selling their opponents’ subjects, and whatever damage [was] done [was] not done in their own country.”Footnote 121 It is not surprising that, as in Europe and the Americas, a wide range of places on the sub-Saharan African coast were involved in putting captives on board a slave ship. The transatlantic slave trade database allows us to identify no fewer than 254 different sites on the sub-continent where vessels embarked captives.Footnote 122 In addition, while the sub-Saharan African population, comprised about 3 percent of the global total in 1850, it was home to at least one-quarter of the world’s languages – and language is surely a key proxy for cultural diversity. No other region in the world of comparable size could come close to the degree of human diversity and political fragmentation that existed in Western Africa; sadly, these characteristics correlate well with potential for enslavement. The central point is that, in effect, everyone living in the sub-continent was an “outsider” and thus eligible for enslavement by one or more African polities or ethnolinguistic groups that were different from their own.Footnote 123
The latest evidence for West Central Africa suggests that war was not the primary source of captives, at least after 1780. Domingues da Silva finds that minor conflicts, abduction, and judicial proceedings together accounted for most enslavements.Footnote 124 In other words Africans were at risk in their own communities, eligibility rules notwithstanding. These two views of the enslavement process – wars between states and slaves generated within the state – are not necessarily inconsistent with each other in the West Central African case. A large share of the conflicts that Thornton describes in his 2020 book were succession disputes, or civil wars after which many of the losing side found themselves in the hold of a slave ship.Footnote 125
The recent data throw up three important specific questions for Thornton. First, his estimate that 7 million people were carried off from West Central Africa to the Americas is certainly too high. The figure that Paul Lachance and I generated is 5.7 million and we provided an essay, a separate spreadsheet and a database all downloadable from www.slavevoyages.org to explain its derivation. Thornton does not engage with this research. He relies instead on a “belief that the absence of crucial records makes the period before 1700 subject to serious undercounting which has not been fully considered in the database.” In fact, the 5.7 million figure does attempt to take account of “serious undercounting” in the original sources.Footnote 126 The second issue is Thornton’s assessment of the main demographic impact of the slave trade on West Central Africa, which for him was on the age and sex structure of its population rather than its size. He concludes “the slave trade had only a limited effect on the number of people living in the area.”Footnote 127 While this is consistent with the current research, he nevertheless reports major imbalances in the sex ratio and a skewed age distribution in the remaining West Central African populations because of the removal of a preponderance of people of working age.Footnote 128
But the numerical basis of this position is questionable. The estimated population of West Central Africa in Table 5.2 is 11.2 million (the sum of rows 16–18), and annual departures between 1776 and 1850, when these peaked, averaged 36,560 people. The overall region thus lost just under one-third of 1 percent of its population annually to the Atlantic slave trade (36,560/11.2 million) and most of these were males. Natural population growth must have at least replaced this loss because historians (including Thornton) see the African populations holding steady during the slave-trade era. On this Thornton is in line with the current consensus in arguing that slave trading did not reduce the total population. But it follows that births over deaths must have at least matched slave-trade losses and those births must necessarily have had an approximately equal sex ratio, thus offsetting the skewness of the group forced to leave.Footnote 129 It is possible to have a skewed sex and age ratio as a result of the slave trade as well as a constant population, but it would take a very peculiar population pyramid for this to happen. Under these circumstances, the ratio of seventy-one males per hundred females that Thornton calculates for the Portuguese Angolan slave population in 1777/1778 is not likely to have held for West Central Africa as a whole. Finally, new evidence from census data for 1797–1870 (with gaps) for Ambaca, São José de Encoge, Golungo, and Cazengo located in the coffee-growing belt shows no consistent indication that the slave populations left behind were predominantly female. In Encoge, the sex ratio was in most years somewhat biased to women, but in the other districts female captives were not predominant.Footnote 130 Either the assumptions underlying Thornton’s demographic estimates are problematic or the communities for which he has found data are not representative of the wider region.
The latest version of slavevoyages.org includes databases that go beyond recording voyages and the people on board them. This new information permits an emphasis on African agency in the slave-trading business both in terms of participation in, and defense against, the ravages of the slave trade. Several Africanists have written interpretations consistent with the voyages data even before the full database became available online. In the stateless or politically “decentralized” areas of Upper Guinea as well as the hinterland of the Bight of Biafra, Walter Hawthorne, David Northrup, and Ugo Nwokeji give a strong sense of equality of status between African and European slave traders that is consistent with the new data.Footnote 131 Hawthorne finds that not only were there many sellers of slaves, but that the Balanta people developed strategies to protect their communities against the impact of the Atlantic traffic. In a similar vein, at the end of his study of the Aro network in the Bight of Biafra hinterland, Nwokeji stresses “Africa’s role in shaping the Atlantic slave trade.”Footnote 132 For the areas dominated by what are often called “predatory states” we must dip into the older literature to find discussions that recognize the autonomy of African polities and the decision-making of their slave traders. Though written in the 1970s, the work of Boubacar Barry and Pat Manning recognizes the agency of the states of Kaabu and Dahomey respectively. While Barry writes within a Walter Rodney-type framework, both studies are broadly consistent with the data generated over the years by www.slavevoyages.org.Footnote 133
As must now be obvious, the intent of this chapter is not to provide an alternative to the grand interpretations of Africa’s involvement in the transoceanic slave trades that have dominated the historiography in recent years, but rather to ask more scholars to take on board new information, some of which contains a quantitative element. As Thornton has repeatedly pointed out, Africa had more in common with the Americas and Europe than it had differences during most of the slave-trade era. If this had not been the case, then the slave trade would not have happened. A widespread acceptance of slavery and the slave trade with, for most of the period, no evidence of a desire to abolish either, meant that hundreds of ports and millions of people on all continents bordering the Atlantic were available to organize and sustain the business. Further, each continent or sub-continent could generate the food, clothing, and shelter it needed to sustain its respective populations without reliance on long-distance trade, and for most of the period and with the dramatic exception of the Indigenous populations of the Americas and the 12.75 million people carried off from Africa, life expectancy and vital rates were not much affected by the transoceanic movements of people, except, obviously, those and their families caught up in the traffic. When a divergence in vital rates and income began to emerge between continents in the early nineteenth century, this, too, cannot easily be attributed to slavery and the transatlantic slave trade or, indeed, their abolition.
The major figures in the historiography of course do not ignore the numbers, but the balance of the evidence they employ favors the opinions of contemporary observers – usually visitors to the sub-continent – and, to a lesser extent, African oral traditions. A consequence of relying on such evidence is that books for the general reader, blogs, and press reviews can promote the view that European exploitation of Africa was central to Western economic development. A proper response to the large questions raised in this chapter requires a synthesis that integrates these different methodologies or rather that recognizes that quantitative and qualitative approaches cannot be separated. My assessment of the current historiography is that we are still a long way short of attaining such a synthesis.
