African Conceptions of Gender and the Slave Traffic

THE gender and age structure of the transatlantic slave trade is critical to understanding the development of societies of the Atlantic rim. From the broad perspective of contact between the Old World and the New, two salient characteristics of that structure have emerged from the recent literature. First, as is now well known, males predominated in the Atlantic slave trade, though compared to other branches of pre-nineteenth-century migration, both coerced and free, females and children were well represented. Second, the proportion of African women and children carried across the Atlantic was far from constant or uniform; sex and age ratios varied strongly by region and over time.[1] Attempts to explain these broad patterns have generally focused on the economic functions of slaves on both sides of the Atlantic, but especially the requirements of the plantation complexes of the Americas, without which a transatlantic slave trade would not have existed. Even though New World planters demanded men, they quickly discovered that enslaved African women had a high work rate.[2] Planters forced black men and women alike to labor in the fields, and the price differential between males and females was generally much lower in the Americas than in Africa.[3 African gender studies have rarely focused on the era of the Atlantic slave trade.[4] Economic historians have generally avoided gender questions, and historical demographers who address the issue have usually assumed uniform sex and age ratios for all African regions. To the latter group of scholars, the main question to be resolved was the proportion of men among those forced to leave Africa.[5] The few historians who have explored the gender structure of the slave trade see the predominance of males in overseas export as more a function of supply than of demand. For them, African suppliers of captives channeled women and children away from the Atlantic and men toward it. Women could be sold for more in domestic African slave markets, whereas men commanded higher prices in markets supplying the Atlantic.[6] Women thus constituted the large majority of slaves in Africa. A recent survey of available price data in African regional markets confirms this hypothesis.[7] The discovery of the high value Africans placed on women is especially helpful in suggesting that African conceptions of gender helped shape the structure of the Atlantic slave trade. But this new emphasis on the ability of Africa to shape the pattern of coerced migration still means that it is the economic function of slaves and market forces that receives the most attention. Although social processes are acknowledged, the emphasis remains on market forces, which crystallized in three overlapping markets—Atlantic, Saharan, and domestic—and generated significant price differentials.

The few studies to take up interregional variations in the age and sex of captives follow a similar tack. One view suggests that women were sold in inland markets because they attracted higher prices there and men were moved to the coast for the same reason. Women and children were important in overseas markets only where the major provenance areas were near the coast. Transportation costs are deemed the critical factor in such decisions. If Atlantic markets put a lower value on women and children than on men, then women and children were not worth moving over long distances to reach those markets.[8] This argument makes sense, but it is more useful for explaining the differences between inland and coastal markets than between one part of the coast and another. Coastal regions with nearby provenance zones still exhibited marked differences in the age and sex patterns of those sent into the trade. Most strikingly, West Central Africa, the region with the longest supply lines in Africa, had one of the largest ratios of children entering the trade.

More women entered the transatlantic slave trade at the Bight of Biafra than at any other coastal region of Africa. There, too, the market approach dominates attempts to account for this exception. For Joseph Inikori, outside forces—specifically, planters in the Americas—overwhelmed the impact of trans-Saharan and internal markets. Inikori explains the large proportion of females leaving the Bight of Biafra by suggesting, first, that African suppliers moved male captives to ports in adjacent regions where prices were higher, and, second, that Euro-American buyers were prepared to accept females from this single region.[9] Letters of instruction to Africa-bound supercargoes and captains lend credence to the view that buyers demanded predominantly men; sometimes Euro-Americans made favorable comments on Biafran women.[10] But merchants and planters expressed no consistent preference for Biafran women, nor did the occasional favorable comment translate into effective demand. Males usually outnumbered females from the Bight of Biafra, and after 1730, as the region grew in relative importance to the Americas, the proportion of males increased.

This article does not reject outright the conclusions of any of the recent work on age and gender in the slave trade. It does, however, attempt to recast the question of gender by taking into account African factors to explain not only the overall demographic structure of the trade but also interregional differences.[11] It draws in particular on the insights of the historian Sandra Greene, even though her work has been chiefly concerned with the impact of the slave trade on Africa, whereas the present study examines how African constructions of gender interacted with the Atlantic slave trade. For the Ewe group on the Gold and Slave Coasts, Greene links demographic pressures on land arising from conflict with the evolution of a culture of conspicuous consumption, both of which stemmed ultimately from the slave trade. She traces the impact of this process on constructions of gender and gives a central role to ethnicity in her analysis. In short, she recognizes that gender relations often varied from group to group.[12] Differing conceptions of gender among African peoples are central to explaining the structure of the slave trade as well as its impact.

While markets mediate both economic and noneconomic values, economic behavior has many cultural determinants. There are some obvious patterns of behavior in the Bight of Biafra (and no doubt elsewhere) that are difficult to account for in terms of maximizing profits. It seems unlikely that economics will explain why the Igbo and Ibibio killed twins, when selling two or more children would fetch more money than selling one, or why the twin taboo was deepest in Aro-Chukwu, the headquarters of the commercial Aro.[13] Similarly, why did the region’s specialized warriors decapitate men captured in warfare instead of selling them, especially when Europeans went to the coast looking first and foremost for men?

Profit maximization was only one factor that shaped African conceptions of gender. Three slave markets are known to have coexisted in Africa during the Atlantic slave trade era: the Atlantic, the Saharan, and the domestic. In general, the Saharan market was female oriented, whereas both females and children predominated in the domestic market. The Atlantic market concentrated on dealing in males, preferably adult males. The specific configuration of any of these markets might be expected to have some impact on the gender structure of the others. In the Bight of Biafra, the influence of the Saharan market was remote, and the institution of female slavery was marginal. Interaction between these two markets might result in more females leaving the Bight of Biafra than from other regions that were more susceptible to trans-Saharan influences and put a different value on female labor. This article compares the patterns in the Bight of Biafra with those in other regions of Atlantic Africa.

The Bight of Biafra extends from the Niger Delta (exclusive of the River Nun) in modern Nigeria to Cape Lopez in modern Gabon.[14] The region was highly decentralized politically. The Nigerian section, which accounts for 90 percent of all captive departures from the Bight of Biafra, was largely homogeneous, with Igbo and Ibibio peoples predominating in much of the region south of the Benue River—known today as southeastern Nigeria.[15] The Bight of Biafra became a major captive export region only from the 1730s onward. The departures of captives more than tripled between the end of the seventeenth century and the middle of the eighteenth. The traffic closed down quickly in the 1840s, but for most of the preceding century, the Bight of Biafra had been the second most important region for captives carried to the Americas (though lagging well behind West Central Africa). The major slave embarkation ports were Bonny, Old Calabar, and New Calabar, with the first two probably accounting for nearly four out of every five captives carried from the region. During the rise to importance of the Biafran traffic, Old Calabar, originally the premier departure point, gradually fell behind Bonny. Captive departures rose from all ports as the trade expanded, but expansion was principally associated with Bonny and its hinterland due to Aro incursions into the densely populated central Igboland and establishment there of large dominant diaspora settlements. Centrally located Bonny attracted the trade with the Aro network more than any other port, although the Aro, a rare, if not the only example of a non-Muslim African trade Diaspora, were in fact the principal traders of the hinterland for all ports in the Bight of Biafra.[16]

The Bight of Biafra is of particular interest in understanding African conceptions of gender. If, in other regions, the Euroamerican drive to secure men was not wholly successful, it all but broke down in the Bight of Biafra. In this region, the character of African warfare and the role of women in the indigenous economy and social institutions shaped the age and gender structure of the slave trade differently from elsewhere. Key elements in this process were African conceptions of slavery and the division of labor, reproduction in the context of the lineage, polygyny, and methods of enslavement. 

To probe these issues, the new transatlantic slave trade dataset is invaluable. Information on the sex and age composition of captives is available for 630 voyages from the Bight of Biafra, perhaps 15–20 percent of all the slaving voyages that left the region after 1650. More important than the size of the sample, however, is the ability to examine the shares of men, women, and boys and girls. For all quarters between 1651 and 1850 for which adequate data are available, ratios of women are higher and those of men lower than for those leaving other regions of Atlantic Africa. In the second half of the seventeenth century, when the traffic from the Bight first became significant, females were consistently in the majority. Biafra was the only large provenance zone in Atlantic Africa, as well as any part of the Old World from which people sailed for the New, for which this was the case. Moreover, at this stage very few of the coerced migrants were children, so the number of women carried into the trade was usually larger than the number of men. In the early eighteenth century, the female share of coerced migrants slipped slightly but remained very close to parity for the first quarter—a pattern that continued to distinguish the Bight of Biafra from other coastal regions supplying captives. Indeed, between 1701 and 1725, the rest of Africa was already sending two males to the Americas for every female. In this period, the proportion of women was 40 percent higher in the Bight of Biafra traffic than elsewhere, and the proportion of girls twice as high. The sample size permits no further comment until after midcentury, by which time the volume of the Bight of Biafra trade had tripled. As with the rest of Africa, the share of males in the trade as well as the share of children rose steadily in the eighteenth and nineteenth centuries, but the share of females leaving the Bight of Biafra continued well above that in other regions. If more than half the deportees in the seventeenth century were female, by the second half of the eighteenth century, this ratio had fallen to 40 percent, and to about one-third in the nineteenth century—a ratio attained elsewhere in Africa a century earlier (see Table I).

The female ratio does, however, disguise a dramatic change in the distribution of women and girls. Whereas women had constituted nearly half of those forced to leave the Bight of Biafra in the seventeenth century, fewer than one in six captives leaving Biafran ports in the final quarter century of the trade (1826–1850) were adult females. Girls, who had made up less than 10 percent of females in the Biafran traffic before 1700, came to outnumber women by the third quarter of the eighteenth century. The proportion of girls also increased over time in other provenance zones, but women always remained more numerous than girls outside the Bight of Biafra. Among males, the percentage of men remained relatively stable before, during, and after the rapid increase in the trade. It was 46 percent in the seventeenth century, dipped to just under 40 percent in the eighteenth century, then recovered to its earlier level after 1775 and until 1850. Thus, most of the increasing male ratio in the trade came from increased numbers of boys, whose share overall doubled during two centuries. Once more this pattern differentiates the Bight of Biafra from the rest of Africa, where men consistently composed more than half of those entering the trade before 1800 and where, in the nineteenth century, their ratio slipped somewhat. Boys were thus proportionately more important in the Bight of Biafra than elsewhere in sustaining and, indeed, driving upward the high male ratios.

Before examining these intriguing differences between the Bight of Biafra and other regions, it will be useful to evaluate the differences in these ratios among ports within the Bight of Biafra. The evidence here is rather more limited than it is at the regional level. Not only is the sample smaller (544 voyages for all identified ports), but the distribution over time is uneven (see Table II). Of the three major ports of Bonny, Old Calabar, and New Calabar, few observations exist for Bonny before 1776. Further, information on New Calabar is totally lacking for 1700 to 1776 and is meager for the nineteenth century. A continuous series is thus possible only for Old Calabar, which overall embarked far fewer captives into the Atlantic trade than did Bonny. This discrepancy is unfortunate because the dramatic increase in the volume of the trade from the Bight of Biafra is associated with the rise of the port of Bonny to predominance. Bonny appears to have surpassed the other ports in numbers of captives dispatched in the third quarter of the eighteenth century, and between 1776 and 1825, it sent between three and four captives to the Americas for every one who left from Old Calabar and New Calabar together. Old Calabar recovered somewhat in relative terms in the last two decades of the trade but did not displace Bonny as the principal point of embarkation in the Bight of Biafra at any point after 1730 (see Map II).

In the earliest years of significant slave trading in the Niger Delta-Cross River region, no significant differences occurred in the demographic structures of the traffic from New Calabar and Old Calabar despite the distance between the two ports and the reputedly different hinterlands on which they drew. Statistically significant differences among major ports emerge only in the later period, although there are large gaps in the data for the crucial middle half-century of the eighteenth century when the Bight of Biafra trade was increasing so rapidly. From 1776, the flow of captives from Bonny, by now the dominant port of departure, had a different mix of sex and age from those leaving Old Calabar, and New Calabar’s profile was closer to that of the geographically adjacent Bonny than to the more distant Old Calabar. More females, in particular more women, left Bonny and New Calabar than left Old Calabar, and among males, the move to more boys, a striking feature of the nineteenth-century trade, was already underway in the last quarter of the eighteenth century. After 1800, most of the statistically significant differences between the major ports of Bonny and Old Calabar (once more the New Calabar data are inadequate) disappear, except for the last quarter-century of the trade, when the age profiles of departing males are different—more men leaving Bonny and more boys leaving Old Calabar.

In summary, flows of captives leaving the major ports of departure in the Bight of Biafra generally show little difference in age and sex mix, with all three ports together (accounting for 85 percent of all departures from the Bight of Biafra) deviating from other African regions, though the Bight of Biafra, in common with most other African regions with Atlantic trade, sent increasing proportions of males and children into the trade over time. To the extent that there were differences among ports in the Bight of Biafra (and readers should keep in mind the gaps in the data in the important mideighteenth-century period), they show up in the last quarter of the eighteenth century when the slave trade from the Bight of Biafra was at its peak. Bonny in these years (and possibly, too, in the preceding two quarters when data are lacking) sent more females into the trade than did the Cross River port of Old Calabar.

These developments are best understood arising from the interaction of indigenous conceptions of gender and the process of Atlantic slave trade. The gender division of labor is at the core of constructions of gender anywhere. It is in this division that productive roles of gendered persons in freedom and slavery are located. Women likely did have a larger economic role in sub-Saharan Africa than in most other continental groupings of societies. In all the varieties of African societies, women’s and men’s “spheres” were separate, but closely interrelated and complementary. Given the male majority among those sent into the slave trade, women probably formed the majority of many African societies during the era of the overseas slave trade.

In one region in Angola, which seems to be the only place where a census count was taken before the nineteenth century, there were 70.9 male slaves and 81.8 free males to every 100 women.[17] This pattern may be in part due to the importance of women in agrarian regimes, in most of which they performed the bulk of labor. During the late seventeenth and early eighteenth centuries, Benin women were said to “have so much Employment, that they ought not to sit still.” In the same period in Whydah in Dahomey, women “Till[ed] the Ground, for their Husbands only.” In West Central Africa, the Capuchin priest Denis de Carli reported that, while seventeenth-century Kongo men served in large armies and carried ‘great Logs of wood of a Vast weight,’ they nevertheless enjoyed considerable leisure time. Women, on the other hand, worked from morning to evening tilling the ground, sowing all crops, cultivating, and harvesting in addition to the family and household duties. The same was true of contemporaneous coastal societies of the Upper Guinea Coast, a region that also exported small proportions of females, according to reports written in both halves of the seventeenth century. In southern Mozambique, men were said to have done next to nothing.[18]

Yet do all these indications of the major economic role for women mean that the cultural determinants of unbalanced sex ratios were the same in all sub-Saharan societies and remained constant over time?[19] For historian Catherine Coquery-Vidrovitch, the pattern that held in most of West Africa was also valid in Niger Delta and Cross River region hinterlands. Among the Igbo, she claims, “it was women who worked in the fields.”[20] But if the economic role of women was so large, why were more females sent into the trade from this region than elsewhere in Atlantic Africa? Interregional differences in the gender division of labor suggest interregional differences in conceptions of gender that may have affected the sex ratios in both the slave trade and in societies supplying that trade, at least as much as did the requirements of planters in the Americas.

Enough is known about women’s work in the Bight of Biafra to question Coquery-Vidrovitch’s observations.[21] In this region, both males and females contributed significantly to agriculture. The hoe per se has never been, as Coquery-Vidrovitch claims, principally a women’s tool among the peoples of the Bight of Biafra. Generally, men used the big hoe to till the ground, and women and children used the small one for weeding.[22] The division of labor was particularly clear-cut among the two groups that supplied the overwhelming majority of the region’s captives—the Igbo and the Ibibio. Females performed a wide range of tasks, such as weeding and planting vegetables and other crops—pumpkins, maize, okra, beans, pepper, and cocoyam (taro). Tilling the ground, planting and stemming yam, building, and climbing trees were exclusively male tasks. Both men and women were involved in clearing, but it was nevertheless a predominantly male activity.[23] The historical role of men in agriculture in several Igbo communities was such that women assumed a major role only in the twentieth century.[24] In the coastal societies, male slaves dominated the manual labor relating to trade and did the same when agriculture became important. Most women performed “non-slave, low-status” domestic, agricultural, and commercial activities. Captain Hugh Crow, who visited the region regularly from the late eighteenth century to 1807, observed that in Bonny it was the women who fetched condiments for the kitchens of European ships.[25] So unimportant was women’s role in agriculture made to seem in the patriarchal order that even in the second quarter of the twentieth century the missionary G. T. Basden omitted it in his otherwise informative chapter on “women’s work.”[26] The major role of males in the agriculture of the Bight of Biafra and in particular Igbo societies is not replicated in other African regions that supplied captives to the Americas.

It is not always easy to determine how much of the gender division of labor antedated the slave trade and shaped its gender structure or how much of this division was a direct result of the trade. Historian John Thornton presents a useful model for measuring change during the Atlantic slave trade. He assumes that the preponderantly male exports affected the women the transatlantic traffic left behind in two major ways. First, they increased the dependency ratio for all working people. Since these were now mostly women, the burden of providing subsistence to the dependent population fell heavily on them. Second, with about 20 percent fewer males to perform male roles, women had to step into the breach or leave them undone. Invariably, this situation resulted in lower production and lower quantities of high-protein food, because women did not take over hunting.[27] Thus the division of labor was altered, and females increasingly assumed tasks and responsibilities hitherto performed by men.

If Thornton’s model holds for regions throughout Atlantic Africa, then Biafran societies, which sent significantly higher proportions of females and children overseas and so must have had more balanced population pyramids, would have escaped some of the ruptures that the model predicts. The African regions where women have been found to constitute the majority of the slave populations were also areas where women cost more, performed the bulk of agricultural labor, and were recognized as doing so. Slavery in those regions centered on women. The Capuchin missionary Michael Angelo reports that, in seventeenth-century Kongo, “Young Women [were] of the same Value as Men.”[28] Unlike these regions, Biafran slavery did not rest on women. Most interesting, research on Bonny, the populous Ngwa of southern Igboland, and the Aro, the last of whom controlled the inland trade, indicates that female slavery for domestic purposes was marginal.[29]

Recent field work among the Aro in particular supports the marginality of female slavery. Asked whether men routinely acquired slave women, respondents indicated that even if a man retained a female trade captive in his household, the relationship invariably changed to that of husband and wife. Alternatively, the man might give her to his wife as a slave or to one of his sons or dependents as a wife. One interviewee reported that, among the Aro, women were a small part of slaveholding: “There were [women slaves], but not so many. A female slave would be given to a slave man because a woman could not establish an ama [lineage].”[30] Moreover, one elder noted, an important reason for the accumulation of male slaves is that they could be used in fighting wars.[31] According to another elder, the Aro believed that
Women were difficult. Their mobility was minimal. It was only practical to buy a woman and she became your wife or that of your son, or she became a fellow woman’s slave. Except if she was sold to a far away place or overseas, it was not very practical. In certain circumstances, people bought female slaves, but they invariably ended as wives. I do not think, however, that a person set out to buy a woman so that she would become his wife. What I am sure of is that women bought women slaves.[32]

Whether women were significantly less mobile, the attitude reflected in this comment seems to have affected the choice of enslavers. Another respondent explained how slave women became wives: “A man could buy a woman from the market and bring her home. You know that man-woman matters are complicated.…If he was the sort that had a soft spot for women, he might decide to keep her longer and might marry her.…A woman who came as a slave was effectively the man’s wife once she got pregnant and bore children for him.”[33] Historian Susan Martin has observed a similar phenomenon with respect to southern Igboland.[34] Female slaves usually belonged to females, whom feminist anthropologist Ifi Amadiume has misidentified as “female husbands.”[35] Although current in present Igboland, this form of slavery-considered-as-marriage likely evolved in the twentieth century and incorporates the basic slave/mistress relation of earlier times.[36] From 1775 onward, when the traffic from Biafra peaked, Bonny sent more females into the trade than did other ports. The rise of Bonny probably meant that the long-term trend of fewer females entering the trade slowed down and that more females entered the trade than would have been the case if Bonny had remained a slave trading backwater. Bonny was the main outlet for central Igboland. The expansion of the Biafran trade, the rise of the Aro network, the shift of trade from Old Calabar to Bonny, and changes in gender ratios were correlated developments. In the Bight of Biafra then, unlike other African regions, pre-existing gender constructions seem to account for the export of a high proportion of females.

The objective here is to recognize, first, that men played an important role in agriculture, second, that scholars of the slave trade have given this male role insufficient attention in the recent literature, and third, that the gender ideology that emerged in some societies in the Bight of Biafra had more to do with preexisting cultural norms (in current jargon, was “constructed”) than with the reality of what women could or could not do. On this last point, initially among the Tiv of the Middle Belt in the hinterland of the Niger Delta, male and female roles in agriculture were carefully defined and to a large extent separated. The Tiv explained this division as a function of physical inequality and the need for females to be modest, even though the differentials in strength requirements in tasks on either side of the divide are not obvious and that modesty is culturally determined.[37]

Cultural, as opposed to biological, factors were important in the allocation of crops and crop tasks in most of Igboland and had important implications for the slave trade. Although yam was the staple food in many societies bordering on both the Bights of Benin and Biafra up to their respective Middle Belts, only among the Igbo and the Ibibio of Biafra was it cultivated exclusively by men and regarded as the king of crops. Among the Igbo and Ibibio, yam, if available, would be eaten before anything else. Alexander Falconbridge, who made trips to the Bight of Biafra during the second half of the eighteenth century as a slaver’s surgeon, observed that “Yams are the favorite food of the Eboe.” This preference remained constant into the twentieth century; Basden observes that yam “stands to [the Igbo] as the potato does to the typical Irishman. A shortage of the yam supply is a cause of genuine distress, for no substitute gives the same sense of satisfaction.” More important, yam provided one of the bases for measuring wealth and had varied ritual functions. Women’s role in yam cultivation was restricted to weeding. The so-called subsidiary crops such as maize, cocoyam, okra, and beans probably had the potential to surpass the yam in nutritional yield, but yam was considered supreme because it was in the male domain. Novelist Chinua Achebe calls yam “a very exacting king” and writes that “for three or four moons it demanded hard work and constant attention from cock-crow till the chickens went back to roost.” Basden further assessed the cost effectiveness of yam cultivation:
From an agricultural point of view, the yam is a very extravagant vegetable to grow. Each tuber requires a full square yard of land which, in itself, is a big demand. For seven or eight months of the year, regular attention must be given to its care, absorbing much time and labor. If wages had to be paid, it is doubtful whether a yam farm would pay its way, let alone yield profit.

Although Basden’s observation reveals the fixation the Igbo and the Ibibio had for the yam, perhaps beyond the point of economic rationality, the crop did yield reasonable economic returns.[38]

Yam production owed something to women’s input, but females occupied a lesser agricultural role in the Bight of Biafra compared to other African coastal regions. In Biafra, anthropologist Victor Uchendu writes, “women’s crops follow the men’s.” Apart from symbolizing the degradation of women’s role in agriculture, this statement indicates that part of women’s labor (principally weeding) went to yam production. Women were made to plant their crops, in Uchendu’s words, “between the spaces provided by the yam hills.”[39] This mentality affected day-to-day decisions and food preference patterns. Because women did not primarily work yam and were not acknowledged as important in its production, this region’s leaders, it would seem, were more willing to countenance the forced migration of females. By contrast, in the Gold Coast, Upper Guinea Coast, and West Central Africa, females were vital to the production of rice or corn, and, as a result, lower proportions of females were sent into the Atlantic traffic. Interestingly, in Yorubaland, where men also had a major agricultural role, the proportion of women sent into the trade is closer to that in the Bight of Biafra than in the Gold Coast, Upper Guinea Coast, and West Central Africa.[40] That the proportion of men sent from Yorubaland remained above Biafran levels is possibly owing to other factors, particularly methods of enslavement.

What then was the impact of methods of enslavement on the gender and age composition of the Atlantic slave trade from the Niger Delta and Cross River regions? To some extent, enslavement strategies might be expected to have adjusted to whatever means supplied the market best, but some patterns in the Bight of Biafra are hard to explain in such terms. People were made captive through various means: judicial processes, warfare, kidnaping, political disputes, and economic necessity.[41] The first three of these accounted for probably three quarters of the captives.

Warfare was an important source of captives everywhere, but its role varied from region to region.[42] A calculation from a nineteenth-century sample of enslaved Africans liberated from ships bound to the New World suggests that war captives accounted for 34 percent, the largest single category.[43] Warriors taken in war were always likely to be exported. Even militarized states tended to send prisoners of war into the external slave trade. In the Gold and Slave Coasts, where the evolution of professional armies has been charted, states also employed raiders by the seventeenth century. William Bosman reported that “Prisoners and Ornaments of Gold” formed the booty there. By the eighteenth century, captives became the region’s principal export. Although wars were not necessarily made for capturing people, the bulk of the captives sold into the Atlantic market were war victims—mostly combatants in regions that supplied mostly females and children to the Saharan and domestic markets. The same was true for the Upper Guinea Coast and eighteenth-century West Central Africa.[44]

After the decline of the overseas slave trade, this outlet was no longer available. In the Western Sudan, where there was a high incidence of militarist states, conquerors henceforth massacred the men they defeated because they could no longer sell their captives.[45] In contrast to the Western Sudan, where the killing of prisoners became widespread only following the decline of the Atlantic slave trade and as a security measure, the character of Biafran warfare was such that the specialized warriors of the region focused on head-hunting. They cut off heads as a matter of honor, an action that was both the raison d’être for fighting and an avenue to full citizenship status and prestige in their communities—provided the victims were men. Surviving prisoners tended to be women and children.[46] That the Upper Guinea Coast, through which the Western Sudan supplied captives to the Atlantic, recorded the highest proportion of men and the Bight of Biafra the lowest, indicates fundamental differences in social processes, including enslavement mechanisms. The Gold Coast falls between these two regions.

Whereas warfare in the Bight of Biafra was biased against the taking of men captives, kidnaping tended to focus on children. Kidnappers knew that they were breaking the law, and the punishment for kidnaping was often harsh. Victims had to be carried far away to prevent them from escaping and to avoid discovery. Children were easier to capture and confine than adults. Kidnap victims accounted for the second largest group—30 percent—in the sample of liberated Africans and were probably even more important in the Bight of Biafra.[47] In many cases, the means by which a person became captive was more important than his or her market value in determining where that person was sold.[48]

Despite these pressures leading to the large number of women and children among those forced into the slave trade in the Bight of Biafra, men were always among the war victims and the kidnaped, and other methods of enslavement ensured the predominance of males overall. Judicial processes and, more specifically, political struggles guaranteed that men were well represented.[49] Those sent away on account of rivalry, dissent, or other acts considered deviant were usually men. If the Atlantic slave trade served African rulers as an avenue for punishing offenders, it also served as a way of getting rid of political enemies with little or no cost.[50] This practice was distinct in the Bight of Biafra, according to two Liverpool ship’s masters, Captains John Adams and Hugh Crow, both of whom specialized in the Biafran trade in the late eighteenth and early nineteenth centuries. They identify a group of high-born Igbo regularly sent into the overseas traffic, despite a tendency to revolt and incite revolt and ships’ masters’ aversion to taking them aboard.[51] These would likely have been mostly political deportees.

If the share of females entering the trade in the Bight of Biafra was apparently always larger than that of other major African regions, why did it decrease over time? In Niger Delta and Cross River ports, the percentage of females remained near parity until the first quarter of the eighteenth century, fell to about 41 percent in the third quarter, when it stabilized, and fell again in the fourth quarter to just above one-third, a figure that held for as long as the trade lasted in the nineteenth century. The second-quarter plateau between these downward shifts coincides with the rise of Bonny to its position as a major departure point. The decline in the share of females affected all African regions, but in the Bight of Biafra it appears to be associated with another aspect of the expansion of the Aro network. As the Aro expanded into central Igboland in the late eighteenth and early nineteenth centuries, they began to draw on women from the Nri-Awka region of central Igboland for wives. This practice left fewer females for the Atlantic traffic.

The central importance of reproducing the lineage in most African societies accounts for this behavior. Association with large groups of people in the form of kinship networks conferred a prestige separate from the economic rewards that the labor of those people generated. Lineages could expand through the addition of outsiders or through natural increase. In most parts of Atlantic Africa, procreation was the primary means of expansion. In late nineteenth-century Benin, the Dutch witness David Van Nyendael observed that “the fruitful Woman is highly valued, whilst the Barren is despised.”[52] Similarly, according to Coquery-Vidrovitch, the African “woman’s value and status depended first on her fertility and second on her cooperativeness, initiative, and ability to work.”[53] Among Igbo women, bearing children was a source of joy: daughters would strengthen existing ties and create new ones, and sons would consolidate a woman’s place in her husband’s lineage. Not having children often led to a miserable life, and Igbo folk tales suggest that infertile women were mocked. Other patrilineal societies would have replicated this pattern, albeit in varying degrees. A key characteristic of lineage expansion is the importance attached to sons, especially in patrilineal societies.

Although such values are scarcely unique to the Igbo, the extension of Aro influence had some exceptional features that have implications for the falling female ratios in the Bight of Biafra. The Aro network was the only non-Muslim trade Diaspora and the only large African organization espousing what Uchendu describes as “big compound” ideals to acquire slaves primarily by trade.[54] Military activity was probably less important a means of getting captives than in other regions. The Aro did not rule directly over the source populations, and their influence spread primarily through trade and cultural prestige, not conquest. They augmented their own population through the massive incorporation of people as immigrants, refugees, clients, and slaves. The Aro acquired women principally as wives and overall absorbed more females from central Igboland than males. The men they did take in—by whatever means—invariably found wives from elsewhere. In other words, for each new male entrant, there was at least one female entrant as a wife. A principal mechanism ensuring this practice was that immigrants, including ex-slaves, could not normally marry Aro women. Among the rest, they often chose to marry from their natal homes. In this, they had the encouragement of Aro men who themselves saw the region as a prime reservoir of wives. The preference for Nri-Awka women derived from the Aro drive to expand business ties and a belief, firmly entrenched by the nineteenth century, that Nri-Awka women made good wives. In a situation where most marriages originated from third-party introductions, commercial and nuptial ties reinforced each other. Many Aro from the central Igbo settlements were the sons or grandsons of Nri-Awka women. It is difficult to find a man who did not marry from the Nri-Awka region.

The economic realities of Aro expansion and the social and cultural constructs that went with it thus account for at least part of the decrease in the female export ratio during the apogee of the slave trade. The incidence of people from this region in export samples varies inversely with Aro incursion into central Igboland. The Aro preferred Nri-Awka people to other groups when taking immigrants into their own society.[55] The Aro tendency to accumulate women would have contributed to the decline in the ratio of women leaving the Bight of Biafra. Whereas women made up 46 percent of departures between 1650 and 1700, they accounted for a mere 15 percent in the last quarter of the trade. Yet the proportion of girls leaving increased over the same period, and, while their share of the trade moved erratically, about a third of that increase occurred in the nineteenth century. Despite the dramatic fall in the portion of women, polygyny—an institution that both fascinated and remained largely beyond the comprehension of European visitors to Atlantic Africa—does not appear to have been much of a factor either in shaping or being shaped by the demographic composition of the slave trade, notwithstanding the strong opinions of some modern scholars.[56] There is no evidence in the Bight of Biafra of a rising incidence of polygyny during or after the expansion of captive departures from the region. The one qualification to this assessment is the rise of what historian Nakanyike Musisi in another context calls “elite polygyny”—defined as marrying more than four wives—in the mid-nineteenth century, by which time the palm oil trade had become more valuable than the slave trade.[57] In the slave trade era, dramatic examples of polygyny tend to come from the Slave Coast and West Central Africa rather than from the Bight of Biafra and the Gold Coast. Bosman reported that in the late seventeenth and early eighteenth centuries Gold Coast men “content[Ed] themselves with one, two, three, and the most considerable Men, with eight, ten or twenty Wives,” while men in the Bight of Benin port town of Whydah had “forty or fifty, and their chief Captains three or four Hundred, some one Thousand, and the King betwixt four and five Thousand.” Joseph Miller documents that “the numbers of young wives surrounding older males…astonished visitors to the interior of Angola” in West Central Africa.[58] This difference between the Bight of Biafra and the Gold Coast on the one hand and the Slave Coast and West Central Africa on the other could be either cause or effect of the higher proportions of females among the captives sent from these regions.

The sex ratio does not, however, appear to be the principal factor driving polygyny. In one region in Kongo (West Central Africa), for instance, a baptismal register for the years 1774 and 1775 reveals that only fifty-four (10.6 percent) of the 507 families recorded were polygynous. Among this polygynous group, perhaps forty-six (9.2 percent of the whole register) had only two wives. Only one man, described as “the lord of the area,” had as many as four. As John Thornton notes, “the majority of marriages were monogamous.”[59] His conclusion accords with conditions in the Bight of Biafra beyond the trading states. Yet, while West Central Africa sent into the Atlantic slave trade the lowest proportion of females after the Upper Guinea Coast, the Bight of Biafra sent the highest. This comparison suggests that polygyny was not universal in Africa and that men married women, not simply because of the abundance of women, but because men needed women’s labor and reproductive resources.

In the final decades of the slave trade from the Bight of Biafra, the declining female ratio was reinforced by a change in the economic role of women in the Niger Delta and Cross River areas. Palm oil production became a major activity in the region. Because palm oil was a labor-intensive industry in which the Aro had a major role (though less so than the slave trade) and because by the late 1830s the value of palm oil exports surpassed those of captives sent to the Americas, the pressures on the Aro to hold women at home are evident. Less easy is a quantitative assessment of this effect.

African conceptions of gender shaped the sex and age structure of the overseas slave trade. These constructs emerged from factors altogether more profound than the merely economic. The unusual Bight of Biafran sex ratio is perhaps best explained by the cultural determinants of the male and female roles in agriculture and by enslavement methods in an environment of decentralized political power. The convergence of these factors, together with marginal contact with Saharan markets, resulted in large numbers of women being sent into the overseas trade. The only other area with a gender division of labor similar to that of Biafra was the Yoruba region in the Bight of Benin. It recorded a lower percentage of female departures, which may be attributed to two causes. First, unlike in Biafra, Yoruba men did not play a major role in agriculture, at least to the extent of reserving for themselves the cultivation of the most significant crops. Second, warfare was a more important means of procuring captives in Yorubaland, especially during the nineteenth century. For the gradual decrease in female ratios over time, the peculiar nature of Aro expansion and the associated drive to incorporate new members into the lineage appear to be important in the Bight of Biafra. The Atlantic slave trade intensified gender inequality everywhere, a development that also created a wider role for women (and therefore an increased demand for women in the domestic economies) so that this process may have been self-correcting in the long run.

It is not the intention here to ignore the economic impulse, or indeed the economic basis of African slaving, but, as anthropologist Peter Gose notes, culture is played out in the understanding and practice of the economic process. [60] In this respect, there is no contradiction between the demand conditions and prices in Africa encouraging the expansion of trade, on the one hand, and cultural elements shaping the categories of export captives, on the other. These are complementary, not contradictory, variables. Two implications may be drawn from this interplay of supply and demand (broadly defined). First, whereas slavers decided on whom they sold, they directed their captives to the best available markets. Second, demand from the Americas affected the composition of export captives but so also did supply conditions in Africa. Without Euro-American demand for captives, many fewer persons would have been kidnaped, captured in war, or fallen victim to judicial processes. Such captives were procured in response to demand. But the processes by which people became captive took form in the framework of African culture and political economy, touching on the general division of labor, slavery, the lineage principle, and changing ideas about gender.

Ugo Nwokeji is a member of the Department of History and the Institute for African American Studies, University of Connecticut. He is indebted to David Eltis for many incisive suggestions and to Martin Klein, Robin Law, and Victor Madeira for helpful comments on the first draft. He also benefited from discussions with Barrington Walker and the feedback from conference participants, especially Kristin Mann.


1 David Eltis and Stanley L. Engerman, “Fluctuations in Sex and Age Ratios in the Transatlantic Slave Trade, 1663–1864, Economic History Review, 46 (1993), 308; Eltis and David Richardson, “The Structure of the Transatlantic Slave Trade, 1595–1867,” paper presented at the Social Science History Meeting, Chicago, 1995.

2 Robert William Fogel and Engerman, Time on the Cross: The Economics of American Negro Slavery (Boston, 1974), 75–77; Deborah Gray White, Arnt I a Woman?: Female Slaves in the Plantation South (New York, 1985), 67 68, 98–105; Hilary McD. Beckles, Natural Rebels: A Social History of Enslaved Black Women in Barbados (London, 1989), 7–23; B. W. Higman, Slave Population and Economy in Jamaica, 1807–1834 (Cambridge, 1976), 194; Michael Craton, Searching for the Invisible Man: Slaves and Plantation Life in Jamaica (Cambridge, Mass., 1978), 142–47; James Walvin, Black Ivory: A History of British Slavery (London, 1992), 119–21; Bernard Moitt, “Behind the Sugar Fortunes: Women, Labour and the Development of Caribbean Plantations during Slavery,” in Simeon Waliaula Chiliungu and Sada Niang, eds., African Continuities (Toronto, 1989); Ira Berlin, Many Thousands Gone: The First Two Centuries of Slavery in North America (Cambridge, Mass., 1998), 111.

3 Eltis and Engerman, “Was the Slave Trade Dominated by Men?” Journal of Interdisciplinary History, 23 (1992), 253.

4 African gender studies were the subject of special issues of 3 major Africanist journals since the early 1970s: Audrey Wipper, ed., The Roles of African Women (Canadian Journal of African Studies, 6, No. 2 [1972]); Edna G. Bay and Nancy J. Hafkin, eds., Women in Africa (African Studies Review, 18 (1975), published as Women in Africa: Studies in Social and Economic Change (Stanford, Calif., 1976); and Shula Marks and Richard Rathbone, eds., The History of the Family in Africa (Journal of African History, 24 [1983 ]). Only the last has contributions (Donald Crummey, “Family and Property amongst the Amhara Nobility,” 207–20, and Anne Hilton, Family and Kinship among the Kongo South of the Zaïre River from the Sixteenth to the Nineteenth Centuries, 189–206) dealing with the era covered in this paper, and only Hilton’s deals with Atlantic Africa. Most other historical works on this sphere have concentrated on the 19th-century transition from the Atlantic slave trade to other economic activities, made most explicit in Richard Roberts, “Women’s Work and Women’s Property: Household Social Relations in the Maraka Textile Industry of the Nineteenth Century,” Comparative Studies in Society and History, 26 (1984), 229–50. See also Catherine Coquery-Vidrovitch, African Women: A Modern History, trans. Beth Gillian Raps (Boulder, Colo., 1997); Claire C. Robertson and Martin A. Klein, eds., Women and Slavery in Africa (Madison, 1983); Robertson, Sharing the Same Bowl: A Socioeconomic History of Women and Class in Accra, Ghana (Bloomington, In., 1984); and Susan M. Martin, Palm Oil and Protest: An Economic History of the Ngwa Region, South-Eastern Nigeria, 1800 1980 (Cambridge, 1988). Other exceptions include Sandra E. Greene, Gender, Ethnicity and, Social Change on the Upper Slave Coast (Portsmouth, N. H., 1996); Edna G. Bay, “The Royal Women of Abomey” (Ph. D. diss., Boston University, 1977); Edward A. Alpers, “State, Merchant Capital, and Gender Relations in Southern Mozambique to the End of the Nineteenth Century: Some Tentative Hypotheses,” African Economic History, 13 (1984); and Onaiwu W. Ogbomo, When Men and Women Mattered: A History of Gender Relations among the Owan of Nigeria (Rochester, N. Y., 1997).

5 John Thornton, “The Slave Trade in Eighteenth-Century Angola: Effects on Demographic Structures,” Canadian J. African Studies, 14 (1980), 417–28, and “The Demographic Effect of the Slave Trade on Western Africa 1500–1850,” African Historical Demography, II, Proceedings of a Seminar held in the Centre of African Studies, University of Edinburgh, Apr. 24, 25, 1981; Joseph Inikori, “Under Population in Nineteenth-Century West Africa,” ibid.; Patrick Manning, “The Enslavement of Africans: A Demographic Model,” Canadian J. African Studies, 15 (1981), 499–526, and Slavery and African Life: Occidental, Oriental, and African Slave Trades (Cambridge, 1990), 61.

6 Robertson and Klein, eds., Women and Slavery in Africa; Manning, Slavery and African Life, 42; Paul E. Lovejoy and Richardson, “Competing Markets for Male and Female Slaves: Prices in the Interior of West Africa, 1750–1850,” International Journal of African Historical Studies, 28 (1995), 261–93, and “The Initial ‘Crisis of Adaptation’: The Impact of British Abolition on the Atlantic Slave Trade in West Africa, 1808–1820,” in Robin Law, ed., From Slave Trade to “Legitimate” Commerce: The Commercial Transition in Nineteenth-Century West Africa (Cambridge, 1995), 32–56.

7 Lovejoy and Richardson, “Competing Markets for Male and Female Slaves.”

8 Jan Hogendorn, “Economic Modelling of Price Differences in the Slave Trade between the Central Sudan and the Coast,” Slavery and Abolition, 17, No. 3 (1996), 209–22; Manning, “Louisiana Slavery in Atlantic Context: Demography, Economy, and Culture, 1720–1850,” paper presented at the American Historical Association Annual Meeting, Seattle, Jan. 10, 1998. I am grateful to Professor Manning for giving me a copy of his paper.

9 Inikori, “Export versus Domestic Demand: The Determinants of Sex Ratios in the Transatlantic Slave Trade,” Research in Economic History, 14 (1992), 134–38, 141–43.

10 John Adams, Sketches Taken during Ten Voyages to Africa, Between the Years 1786 and 1800 (London, 1832), 41; Hugh Crow, Memoirs of the Late Captain Hugh Crow (London, 1830), 198; Richard Lander and John Lander, Journal of an Expedition to Explore the Course and Termination of the Niger with a Narrative of a Voyage Down that River to its Termination (New York, 1832), 2:240–41; A. G. Leonard, “Notes on a Journey to Bende.” Journal of the Manchester Geographical Society, 14 (1898), 207.

11 Significant steps have already been taken in this direction. See Eltis, “Fluctuations in the Age and Sex Ratios of Slaves in the Nineteenth-Century Transatlantic Slave Traffic,” Slavery and Abolition, 7 (1986), 257–72; Eltis and Engerman, “Was the Slave Trade Dominated by Men?” 237–57, and ‘Fluctuations in Sex and Age Ratios,” 308–23; Eltis and Richardson, “West Africa and the Transatlantic Slave Trade: New Evidence of Long-Run Trends,” Slavery and Abolition, 18, No. 1 (1997), 29–33; David Geggus, “Sex Ratio, Age, and Ethnicity in the Atlantic Slave Trade: Data from French Shipping and Plantation Records,” J. African Hist., 30 (1989) 23–44; and G. Ugo Nwokeji, “Household and Market Persons: Gender, Deportees, and Biafran Society, c. 1750–1860,” in Deborah Gray White and Mia Bay, eds., Black Atlantic: Race, Nation, and Gender (forthcoming).

12 Greene, Gender, Ethnicity, and Social Change .

13 The taboo against twins in Aro-Chukwu is so strong that it survives to date. For instance, proper care was taken in February 1996 when this writer was on fieldwork to ensure that he was not a twin before he was admitted to a historical site the Nigerian government had declared a national monument in 1972. The largest Aro Diaspora settlement had abandoned this practice by the 1830s; Nwokeji, ” ‘Did We Bring Land with Us from Aro?’: The Contradictions of Expansion among the Aro of Nigeria during the Atlantic Slave Trade” in Lovejoy, ed., Identifying Enslaved Africans: The “Nigerian” Hinterland and the African Diaspora (forthcoming).

14 Major works on the Bight of Biafra include Daryll Forde and G. I. Jones, The Ibo and Ibibio-Speaking Peoples of Southeastern Nigeria (London, 1950); K. Onwuka Dike, Trade and Politics in the Niger Delta, 1830–1885 (Oxford, 1956); G. I. Jones, The Trading States of the Oil Rivers (London, 1963); Victor C. Uchendu, The Igbo of Southeast Nigeria (New York, 1965); Kannan K. Nair, Politics and Society in South Eastern Nigeria, 1841–1906: A Study of Power, Diplomacy, and Commerce in Old Calabar (London, 1972); A.J.H. Latham, Old Calabar, 1600–1891: The Impact of the International Economy upon a Traditional Society (Oxford, 1973); Elizabeth Isichei, A History of the Igbo People (London, 1976); David Northrup, Trade Without Rulers: Pre-Colonial Economic Development in Southeastern Nigeria (Oxford, 1978); Adiele Afigbo, Ropes of Sand: Studies in Igbo History and Culture (Ibadan, 1981); Martin, Palm Oil and Protest; and Dike and Felicia I. Ekejiuba, The Aro of Southeastern Nigeria, 1650–1980: A Study of Socio-Economic Formation and Transformation in Nigeria (Ibadan, 1990).

15 A 1953 census shows that the Igbo and the Ibibio made up respectively 68.56% and 10.36% of much of the region south of the Benue River known today as southeastern Nigeria; “Population Census of Eastern Region of Nigeria 1953,” International Population Census Publications, Africa, Nigeria: 1950, 1952, 1953, 1957 (Woodbridge, Conn., 1981), 15–16, 36–37.

16 For details see Nwokeji, “Biafran Markets and Slave: The Aro and the Slave Trade, c. 1750 1890,” paper presented at the conference “West Africa and the Americas: Repercussions of the Slave Trade,” Feb. 20–22, 1997, and “The Atlantic Slave Trade and Population Density: A Historical Demography of the Biafran Hinterland,” Canadian J. African Studies, 34 (forthcoming). Let us take as a graphic example, one 1821–1822 voyage in which all the captives were declared Igbo and, because the vessel embarked at Bonny, presumably mostly central Igbo. The ship, l’Amelie, was captured by the French navy in early 1822 off the Martinique coast in the Caribbean with 211 (126 adults and 85 children) surviving captives. All were declared Igbo. There were 126 adults (60% of the captives) and 85 children (40%). While women made up only 36% of the adult proportion, girls made up 59% of the children. Calculated from Françoise Thésée, Les Ibo l’Amélie: Destinée d une Cargaison de Traite Clandestine à la Martinique, 1822–1838 (Paris, 1986), app. For trade diasporas in their various historical settings, see Philip D. Curtin, Cross-Cultural Trade in World History (Cambridge, 1984).

17 Manning, Slavery and African Life, 42, and “Enslavement of Africans,” 501. For Angola, see Thornton, “Demography and History in the Kingdom of Kongo, 1550–1750,” J. African Hist., 18 (1977), 507–30, and “Slave Trade in Eighteenth-Century Angola,” esp. 422n; and Joseph C. Miller, Way of Death: Merchant Capitalism and the Angolan Slave Trade, 1730–1830 (Madison, 1988), 159–67.

18 For Dahomey, see William Bosman, A New and Accurate Description of the Coast of Guinea.. (London, 1705), 344; for Benin, David Van Nyendael, “A Description of Rio Formosa, or the River of Benin,” ibid., 463; for Kongo, de Carli, “A Voyage to Congo,” in Awnsham Churchill and John Churchill, comps., A Collection of Voyages and Travels…, 4 vols. (London, 1704) 1:622, 629, 630–31; for southern Mozambique, Alpers, “State, Merchant Capital, and Gender Relations,” 39–40; and for Upper Guinea Coast, Walter Rodney, A History of the Upper Guinea Coast 1545–1800 (Oxford, 1970), 103, and Thornton, “Sexual Demography: The Impact of the Slave Trade on Family Structure,” in Robertson and Klein, eds., Women and Slavery, 44, and Africa and Africans in the Making of the Atlantic World, 107.

19 Greene, Gender, Ethnicity, and Social Change; Hilton, “Family and Kinship among the Kongo South of the Zaire River,” 193; Martin, Palm Oil and Protest, 23.

20 Coquery-Vidrovitch, African Women, trans. Raps, 11.

21 Felix K. Ekechi, “Aspects of the Palm Oil Trade at Oguta (Eastern Nigeria), 1900–1950,” African Economic History, 10 (1981), 41; Susan Hargreaves, “The Political Economy of Nineteenth-Century Bonny: A Study of Power, Authority, Legitimacy and Ideology in a Delta Trading Community from 1700–1914” (Ph.D. diss., Centre of West African Studies, University of Birmingham, 1987), 94.

22 Sylvia Leith-Ross, African Women (New York, [1939] 1959), 89–92; Forde and Jones, Ibo and Ibibio-Speaking Peoples of Southeastern Nigeria, 13, 70; Uchendu, Igbo of Southeast Nigeria, 24; Paul Bohannan and Laura Bohannan, Tiv Economy (Evanston, Ill, 1968), 66; Onwuka N. Njoku, “Igbo Economy and Society,” in Afigbo,, ed., Groundwork of Igbo History (Lagos, 1991), 120.

23 In the 20th century, these crops included the new varieties of yams, which were considered as inferior: Forde and Jones, Ibo and Ibibio-Speaking Peoples, 13, 70; M. M. Green, Ibo Village Affairs (New York, 1964; orig. pub. 1947), 170–71; Uchendu, Igbo of Southeastern Nigeria, 24–25; Chieka Ifemesia, Traditional Humane Living among the Igbo: An Historical Perspective (Enugu, Nigeria, 1979), 62; Nina Emma Mba, Nigerian Women Mobilized (Berkeley, 1982), 29–30; T. Uzodinma Nwala, Igbo Philosophy (Lagos, 1985), 178.

24 Gloria Chuku, “Women in the Economy of Igboland, 1900 to 1970: A Survey,” African Econ. Hist., 23 (1995), 39–41.

25 Hargreaves, “Political Economy of Nineteenth-Century Bonny,” 95–97 (quotation); Crow, Memoirs of the Late Captain Hugh Crow, 44; Nair, Politics and Society, 37, 42–43; Latham, Old Calabar, 91–96.

26 See Basden, Niger Ibos (London, [ 1938] 1966)), 325–33.

27 Thornton, “Sexual Demography,” 41.

28 Michael Angelo, “Voyage to the Congo,” in Collection of Voyages and Travels,, 1:620.

29 For the entire Igbo, see Nwokeji, “The Slave Emancipation Problematic: Igbo Society and the Colonial Equation,” Comp. Studies Soc. & Hist., 40 (1998), 325–26; for the entire Biafra, Nwokeji, “Household and Market Persons;” for the Aro, Nwokeji, ” ‘Did We Bring Land with Us from Aro?’ and The Biafran Frontier: Trade, Slaves, and Aro Society,” c. 1750–1905 (Ph.D. diss., University of Toronto, 1999), 121–25; for southern Igboland, Martin, Palm Oil and Protest, 25; and for Bonny, Hargreaves, “Political Economy of Nineteenth-Century Bonny,” 94–97.

30 Ukobasi Kanu-Igbo to author, interview (Enugu, Mar. 7, 1996). See also interviews with Eneanya Akpu (Aro-nde-Izuogu, Mar. 10, 1996); Nwankwo Anicho Anyakoha (Aro-Chukwu, Feb. 24, 1996); Michael Sunday Igwe (Aro-nde-Izuogu, Feb. 18, 1996); Kevin Maduadichie (Aro-nde-Izuogu, Feb. 11, 1996); Emmanuel Nwankwo (Aro-nde-Izuogu, Mar. 10, 1996); Clifford Okoli (Aro-nde-Izuogu, Feb. 11, 1996); Elizabeth Okoro (Aro-Chukwu, Feb. 25, 1996); Jonathan G. Okoro (Aro-nde-Izuogu, Feb. 4, 1996); Jacob O. Okoro (Aro-Chukwu, Feb. 24, 1996); Eze Jonas E. Uche (Nde-Uche, Mar. 17, 1996). Transcripts are in possession of the author.

31 Igwe to author, interview.

32 Anyakoha to author, interview.

33 Igwe to author, interview.

34 Martin, Palm Oil and Protest, 25.

35 Amadiume, Male Daughters, Female Husbands: Gender and Sex in African Society (London, 1987). See also Chuku, “Women in the Economy.” Contrary to Amadiume’s claim, many respondents indicated that slave women were eventually labeled “wives”; interviews with Akpu; Anyakoha; Igwe; Maduadichie; Clifford Okoli; Elizabeth Okoro; Emmanuel Nwankwo; and with Kanu Nwankwo Igwilo, Aro-nde-Izuogu (Mar. 2, 1996); Martina Ike, Enugu (Enugu, Sept. 20, 1995); Aloy Nwankwo (Boston, Dec. 18, 1996); Margareth Nwambego Okoli (Aro-nde-Izuogu, Mar. 8, 1996); Georgina Umunnakwe (London, July 8–9, 1995).

36 The change must be associated with slavers’ response to the antislavery effort of the British colonial administration, which peaked in the 1930s. This was when child-dealing, especially in girls, became endemic. Most of the transactions were cloaked as “marriage, confusing the colonial administrators who came to regard all bridewealth payment as slave-dealing; Nwokeji, “Slave Emancipation,” 337. Buchi Emecheta’s novel Slave Girl (London, 1977), 50, 58, 60–61, set in the 1900s through the 1920s, confirms that these marriages were indeed slavery: “Many of the market women had slaves in great number.” The female slaves referred to in the novel are all owned by mistresses. Even Chiago, who was sold by her father and purchased by another man, is presented to her buyer’s wife. Yet selling girls to mistresses is actually expressed in terms of “marry[ing] her away.” Marriage differed from slavery because, unlike a slave, a married woman and all of her labor power was not alienated from her lineage; Roberta W. Kilkenny, “The Slave Mode of Production: Precolonial Dahomey,” in Crummey and C. C. Stuart, eds., Modes of Production in Africa (Beverly Hills, Calif., 1981), 158–59. In her important biography of the wealthy, late 19th-early 20th-century Onitsha market woman Omu Okwei, Felicia Ekejiuba writes that Okwei “acquired beautiful girls mostly ‘adopted’ children or children pawned to her by her debtors”; Ekejiuba, Omu Okwei, the Merchant Queen of Ossomani: A Biographical Sketch, Journal of the Historical Society of Nigeria, 3 (1967), 633–46. In 1921, Okwei admitted that one Uyanwa was her slave as Okwei struggled to retain the right of inheriting the property that Uyanwa’s husband, a European United African Company manager to whom Okwei had given her in “marriage,” would leave behind; ibid., 637, 643–44. As in Emecheta’s novel, the story revolves around girls. The idea that the “female husband” relationship is an ancient institution seems to have sprung from the unfortunate tendency, noted by Gloria Chuku, “Women in the Economy,” 37, of much existing literature to view Igbo women as unchanging.

37 Bohannan and Bohannan, Tiv Economy, 66.

38 Falconbridge, An Account of the Slave Trade on the Coast of Africa, London, 1788), 21; Achebe, Things Fall Apart (New York, 1994; orig. pub. 1958), 33–34; Basden, Niger Ibos, 389–90, 394.

39 Uchendu, Igbo of Southeastern Nigeria, 24–25; Phoebe Ottenberg, “The Changing Economic Position of Women among the Afikpo Ibo,” in William R. Bascom and Melville Herskovits, eds., Continuity and Change in African Cultures (Chicago, 1959), 207; Ifemesia, Traditional Humane Living among the Igbo, 62; Amadiume, Male Daughter, Female Husbands, 28–30, 37–38. Among the Owan, northwest of the Igbo, where yam was also gendered, men and women appeared more powerful in localities where they respectively controlled this crop; Ogbomo, When Men and Women Mattered, 97.

40 Toyin Falola, The Political Economy of a Pre-Colonial African State: Ibadan, 1830–1900 (Ile-Ife, Nigeria, 1984), 54; Coquery-Vidrovitch, African Women, trans. Raps, 11. It is important to note that recent studies of the Yoruba have suggested the fluidity of gender; J. Lorand Matory, Sex and the Empire That Is No More: Gender and the Politics of Metaphor in Oyo Yoruba Religion (Minneapolis, 1994). Oyeronke Oyewumi has gone as far as denying any notions of the gender division of labor and suggests that the concept of gender is alien to Yoruba culture. See her The Invention of Women: Making an African Sense of Western Gender Discourses (Minneapolis, 1997), esp. 64–77. Even Jane I. Guyer, “Food, Cocoa, and the Division of Labor by Sex in Two West African Societies,” Comp. Studies Soc, & Hist., 22 (1980), 362, who deals specifically with gender division of labor in Yorubaland, notes in a comparative context that the notion was “phrased in terms of pragmatism rather than metaphysics.” These are useful pointers. Nevertheless, we must still come to terms with why one sex was over-represented in some activities and not in others, and why the sexes were sent into the Atlantic traffic in a significantly unequal distribution.

41 For economic necessity, see John Barbot, A Description of the Coasts of North and South Guinea.. (London, 1732), 352, and William Snelgrave, A New Account of Some Parts of the Coast of Guinea, and the Slave Trade (London, [ 1734] 1971)), 159.

42 Lovejoy, Transformations in Slavery: A History of Slavery in Africa (Cambridge, 1983), 56, 61, 68–78; Curtin, The Rise and Fall of the Plantation Complex: Essays in Atlantic History (Cambridge, 1990), 37–38, 119; Thornton, Africa and Africans in the Making of the Atlantic World, 99, 110.

43 P.E.H. Hair, “The Enslavement of Koelle’s Informants,” J. African Hist., 6 (1965), 196–97. The sample is culled from S. W. Koelle, Polyglotta Africana (Graz, Aus., [ 1854] 1963)).

44 Bosman, New and Accurate Description, 182. For the Gold Coast, see Ray A. Kea, Settlements, Trade, and Polities in the Seventeenth-Century Gold Coast (Baltimore, 1882), chap. 4, and Edward Reynolds, Trade and Economic Change on the Gold Coast, 1807–1874 (New York, 1974), 9–15; for the Slave Coast, Law, “Warfare on the West African Slave Coast, 1650–1850,” in R. Brian Ferguson and Neil L. Whitehead, eds., War in the Tribal Zone: Expanding States and Indigenous Warfare (Santa Fe, N. M., 1992), 103–26; and for Upper Guinea Coast, Thornton, “Sexual Demography,” 42, 45.

45 Klein, “Women in Slavery,” 72; Claude Meillassoux, “The Role of Slavery in the Economic and Social History of Sahelo-Sudanic Africa,” in Inikori, ed., Forced Migration: The Impact of the Export Slave Trade on African Societies (New York, 1982), 89, 90.

46 Nwokeji, “Household and Market Persons.”

47 Edna Bay, Wives of the Leopard: Gender, Politics, and Culture in the Kingdom of Dahomey (Charlottesville, 1998), 106–07; Snelgrave, New Account of Some Parts of the Coast of Guinea, 37–38.

48 It nevertheless is unlikely that kidnaping was the predominant means of enslavement in the region as some scholars have maintained. See Isichei, History of the Igbo People, 45–47; Northrup, Trade Without Rulers, 77–80; John N. Oriji, “The Slave Trade, Warfare and Aro Expansion in the Igbo Hinterland,” Transafrican Journal of History, 16 (1987), 161–63; and Geggus, “Sex Ratio, Age and Ethnicity,” 40. For examples of kidnaping see Olaudah Equiano, Interesting Narrative of the Life of Olaudah Equiano Written by Himself, ed. Robert J. Allison (Boston, 1995), and Vernon H. Nelson, ed., “Archibald John Monteith: Native Helper and Assistant in the Jamaica Mission at New Carmel [1853],”Moravian Historical Society Transactions, 21 (1966), 29–52.

49 Nwokeji, “Biafran Markets and Slaves,” and “Household and Market Persons.” The following summary is from Nwokeji, “Household and Market Persons.”

50 Curtin, Rise and Fall of the Plantation Complex, 119, 120, and “The Tropical Atlantic in the Age of the Slave Trade,” in Michael Adas, ed., Islamic and European Expansion: The Forging of a Global Order (Philadelphia, 1993), 177.

51 Adams, Sketches Taken during Ten Voyages to Africa, 41; Crow, Memoirs of the Late Captain Hugh Crow, 199. For a similar phenomenon on the slave coast, see Edna Bay, “Dahomian Political Exile and the Atlantic Slave Trade,” in Lovejoy, Identifying Enslaved Africans, who finds that Dahomian princes of the Slave Coast utilized deportation to humiliate defeated rivals a punishment deemed worse than execution.

52 Van Nyendael, “Description of Rio Formosa,” 447, 462, 463.

53 Coquery-Vidrovitch, African Women, trans. Raps, 16.

54 Uchendu, Igbo of Southeast Nigeria, 54–56.

55 Nwokeji, “Household and Market Persons.”

56 Curtin has observed that “polygyny stood out as a special evil” to Europeans of the Enlightenment in Image of Africa: British Ideas and Actions, 1780–1850 (Madison, 1964), 252. The British slave trader John Hippisley pointed to the gender imbalance of export captives and men lost to wars. But the tendency for the richest men to have many wives “does not prevent the poorest from having one or two.…The number of women must, therefore, exceed that of men” in On the Populousness of Africa (London, 1764), 14–16, 16 n. For a modern opinion that the trade encouraged polygyny see Manning, Slavery and African Life, 42, and “Enslavement of Africans,” 501, 503.

57 Musisi, “Women, ‘Elite Polygyny,’ and Buganda State Formation,” Signs, 16, (1991), 757–86; Nwokeji, “Biafran Frontier,” chap. 6.

58 Bosman, New and Accurate Description of the Coast of Guinea, 344. The importance of sex ratios in polygyny is well recognized by modern scholars, for example, Manning, Slavery and African Life, 42, and “Enslavement of Africans,” 501, 503; Miller, Way of Death, 163; and Thornton, “Slave Trade in Eighteenth-Century Angola,” 425.

59 Thornton, “An Eighteenth-Century Baptismal Register and the Demographic History of Manguenzo,” in Christopher Fyfe and David MacMaster, eds., African Historical Demography (proceedings of a seminar held in the Center of African Studies, University of Edinburgh, Apr. 29, 30, 1977), 411–12.

60 Gose, Deathly Waters and Hungry Mountains: Agrarian Ritual and Class Formation in an Andean Town (Toronto, 1994), xii.