Slavery and Marriage in African Societies

ABSTRACT This article introduces a special issue focusing upon linkages between enslavement and marriage within African societies from the 1830s to the present day. The right to make decisions over marriage is one of the core powers which masters have historically exercised over individuals whom they enslaved. The exercise of this right had far-reaching ramifications for lived experiences of enslavement, with slaves – usually women and girls – being forced into conjugal relationships where their labour, sexuality and reproductive capacity were at the disposal of their husband/master. This article introduces two major themes that cut across the issue's contributions: direct connections and comparative analogies. The former refers to scenarios where marriage and enslavement directly intersected and overlapped, while the latter refers to claims that at least some African marriages were analogous to enslavement. Comparisons between marriage and slavery have been a recurring feature of African politics throughout the period considered here. They have also been frequently paired with moral denunciations and calls for change by actors as different as European colonial administrators and African victims-turned-activists. At the same time, the legitimizing cloak of marriage has been repeatedly used to deliberately obscure the continuing legacies of slavery.

The relationship between marriage and slavery in Africa can be approached from two main angles. The first approach revolves around the question of whether the experiences of married women and girls can be analogized as the same asor at least uncomfortably similar toexperiences of enslavement. In the second approach the connection between the two is more immediate, including circumstances where individuals experienced being both married and enslaved at the same time. One of the powers slaveholders have historically exercised over their slaves was the power to decide whom they should marry and on what terms. Within this context, marriage had direct effects upon both lived experiences of enslavement and the larger economic, cultural and legal systems surrounding slavery. Marriage and slavery were sometimes at different ends of a spectrum of social legitimacy. Sometimes they were closely bound together. Some slaves who were forced to marry established kinship ties and were able to somewhat improve their status. Others remained slaves.
The differences between comparative analogies and direct connections were relatively clear-cut when slavery was regarded as a legitimate institution. However, the development of abolitionism has made it harder to distinguish between the two. Opponents of slavery advanced the argument that no one should be enslaved, but once slavery had been legally abolished they had to figure out the status of practices which looked similar to slavery, yet were classified differently. Marriage figured prominently within these political conversations. In many jurisdictions, slave owners and their sympathizers made elaborate efforts to reclassify systems of enslavement under other labels, including efforts to camouflage slaves as spouses. 1 Some enslaved women were reclassified as wives, thwarting the promise of abolition. Others took advantage of limited windows of opportunity to escape their masters.
This process was embedded within the larger history of European colonialism and imperialism. The European men at the coal face of empire were strongly invested in both racial and gender hierarchies; they primarily understood their task in terms of a paternalistic duty to protect their 'inferiors' from slavery. 2 Husbands (or fathers) were assumed to have 'natural' authority over their wives and children, so the main goal was to ensure that 'legitimate' marriage would be disentangled from slavery. 3 As long as the taint of slavery was removed (or said to have been removed), marriages which were otherwise highly unequal in gender terms were frequently regarded as legitimate. Women and girls who were not formally enslaved nonetheless experienced oppression and exploitation owing to patriarchal systems within their societies. Globally protected by law, custom, morality and religion, marriage has beenand remainsinsulated against the kinds of political arguments made against slavery. 4 This special issue explores the relationship between slavery and marriage in different African societies from the 1830s to the present day. This period coincides with both the spread of legal abolition and European colonial conquests and their legacies. This introductory article identifies and analyses the main characteristics and connections that have bound together slavery and marriage over this period. In support of this overall goal, the article is divided into four main sections. The first section further develops the distinction between direct connections and comparative analogies, drawing upon a series of historical examples from Africa. The sections which follow divide the relationship between slavery and marriage into three chronological phases: pre-abolition, abolitionism and colonialism, and post-colonial politics and contemporary legacies. When these sections are put together, a larger historical trajectory emerges. The historical relationship between slavery and marriage has never been straight-forwardeven prior to abolitionsince enslaved individuals had very different experiences of marriage. This slavery-marriage equation was altered by the emergence of anti-slavery in multiple ways. Abolitionism confirmed that slavery was a heinous crime; legitimated intervention against it; enabled Europeans who posed as liberators to elude moral and political scrutiny; and created opportunities for slaves to exercise greater control over their married lives. Despite injunctions against slavery, connections with marriage have persisted in multiple ways to the present day.

Marriage and Enslavement in Africa: Connections and Comparisons
In The Nile Tributaries of Abyssinia (1867), Samuel Baker documents a steady slave trade from Abyssinia and Sudan to Egypt and the harbour towns of the Red Sea coast. In April 1862 he visited a slave market in Gallabat on what today is the border between Sudan and Ethiopia, where he saw a large number of 'lovely captives' sold by local slave dealers: On my return to camp I visited the establishments of the various slave merchants: these were arranged under large tents formed of matting, and contained many young girls of extreme beauty, ranging from nine to seventeen years of age. These lovely captives, of a rich brown tint, with delicately-formed features, and eyes like those of the gazelle, were natives of the Galla, on the borders of Abyssinia, from which country they were brought by the Abyssinian traders to be sold for the Turkish harems. Although beautiful, these girls are useless for hard labour; they quickly fade away and die unless kindly treated. They are the Venuses of that country, and not only are their faces and figures perfection, but they become extremely attached to those who show them kindness, and they make good and faithful wives. There is something peculiarly captivating in the natural grace and softness of these young beauties, whose hearts quickly respond to those warmer feelings of love that are seldom known among the sterner and coarser tribes. Their forms are peculiarly elegant and graceful-the hands and feet are exquisitely delicate; the nose is generally slightly aquiline, the nostrils large and finely shaped; the hair is black and glossy, reaching to about the middle of the back, but rather coarse in texture. These girls, although natives of Galla, invariably call themselves Abyssinians, and are generally known under that denomination. They are exceedingly proud and high-spirited, and are remarkably quick at learning. At Khartoum, several of the Europeans of high standing have married these charming ladies, who have invariably rewarded their husbands by great affection and devotion. The price of one of these beauties of nature at Gallabat was from twenty-five to forty dollars. 5 This reads as a pro-slavery text. It justifies enslavement on multiple grounds, including the then common trope in Africa and the Middle East that the feminine virtues of enslaved girls (adult women tended to be less prominent) protected them from the harshest aspects of captivity, together with the equally dubious assertion that they could aspire to nothing better, especially for new wives of 'Europeans of high standing'. Most women and girls purchased at Gallabat remained enslaved, and thereby became either sexually accessible slaves or enslaved wives, depending on the requirements of their master/husband and the internal dynamics of their household. These relationships were not entirely controlled from above. They were also shaped by the agency of the enslaved under conditions of acute duress. 6 Buyers who perceived slavery as illegitimate, such as the Europeans mentioned by Baker, manumitted the female slaves whom they purchased as wives. This does not mean, however, that manumission translated into freedom for the African women concerned in any meaningful sense of the term. They were instead obliged to once again pursue their own strategies in an effort to improve their fortunes. 7 Despite his idyllic account of the 'lovely captives', Baker was no supporter of slavery. In 1869, the Khedive of Egypt Ismail Pasha (Ismāʿīl Bāshā) despatched an expedition under Baker's command to suppress slavery in the Sudan and Central Africa. Baker's account of this second expedition, Ismailïa (1874), describes a different world, one that Bakeras explorer, abolitionist, political advisor, and commander of antislavery expeditionsplayed a role in reshaping. 8 Slave merchants no longer displayed women for sale in the open, but operated clandestinely. And Baker, having acquired the power to search and arrest slave merchants, set up military stations in key locations along the White Nile to intercept slave traders.
On the 10th of May 1870 a vessel approached one such station on its way to Khartoum. Following Baker's orders, Colonel Abd-el-Kader undertook a search of the vessel. Finding it filled with corn, stowed in bulk, he drew a steel rod from a soldier's rifle and probed through the corn. An enslaved woman who lay underneath the cereals made a faint cry. She felt an arm reaching for her through the loose corn, grabbing her wrist and pulling her out. A human cargo of about 150 enslaved persons, mainly women, girls and boys, some kept in irons, was found hidden onboard. The sail of the main yard appeared abnormally full and heavy: it was unrolled, and a young woman emerged. She had been hidden with special care. In the months preceding her release from the sail, this woman's body had caught the attention of her captors. It promised a higher profit than aged, childish, or male bodies. Her value as a commodity was directly linked to the value assigned to young women as wives, concubines and mothers.
The freed slaves 'began to comprehend that their captors were now captives. They now began to speak, and many declared that the greater portion of the men of their villages had been killed by the slave-hunters'. 9 In keeping with larger trends, more women than men had survived, and their survival via enslavement can be attributed to their greater market value and perceived tractability. The following morning, Baker informed slaves found on the vessel that they were now free people. He gave them freedom papers, which he signed personally, and informed them they could leave and try to return to their former homes. Most stayed. At this critical juncture, Baker's vision of freedom for his captives proved to be highly gendered. Freed boys would be trained as carpenters, blacksmiths, tailors, shoemakers, and domestic servants. Women and girls had only two options: marriage or domestic work for the officers.
Marriage was the default option. Having been released from bondage on May 10th, the formerly enslaved women and girls were asked to become wives on the morning of May 11th: 'If any of the women wished to marry, there were many fine young men in the regiments who would make capital husbands'. In the afternoon, Baker asked the officer whether any of the 'negresses' had reached a decision. He was told that all the women wished to marry and had chosen their husbands already. Or perhaps their husbands had chosen them. What actually happenedas experienced by the womenis unknown and unknowable. Baker records events in the following terms: Some of the girls were pretty, and my black troops had shown good taste in their selection. Unfortunately, however, for the Egyptian regiment, the black ladies had a strong antipathy to brown men, and the suitors were all refused. This was a very awkward affair. The ladies having received their freedom, at once asserted "woman's rights". 10 How should Baker's words be read? In his earlier account from 1862, enslavement appears as a pathway to marriage. Eight years later, being rescued from enslavement also appears as a path to marriage. While Baker's treatment of these two events differs in several key respects, he ultimately ends up providing a sanitized picture of marriage on both occasions. It is likely that the 'antipathy to brown men' which Baker describes can be traced to concerns that the North African members of the Egyptian regiment would continue to view black women and girls as slaves, and therefore keep them in bondage or perhaps covertly sell them. The women's choice of husbands can therefore be read as a strategic calculation based on their perception of the 'Arab' soldiers' disposition toward them. But the choice, if there was one, was made within a context where all of the available options were decidedly limited, since re-enslavement at the hands of the 'black troops' these women had allegedly chosen to marry cannot be excluded. Since the available information regarding these specific events comes from Baker, his assumptions, interpretations and omissions as a white European man cast a long shadow over contemporary efforts to grapple with the larger historical dynamics at work here. 11 The 'woman's rights' which Baker refers to were effectively a cruel joke. Despite being nominally 'free', these women were immediately compelled into conjugal relationships with strangers, and their labour and sexuality were placed at the disposal of their new husbands. European abolitionists such as Baker were products of societies which were dominated by models of racial, gender and socio-economic hierarchy. By 'marrying off' the female captives he had freed the previous day, Baker not only ensured that there would be minimal disruption to the status quo ('unattached' women were widely understood to be a potential threat to social order), he also ensured that many of his male subordinates directly benefited, since they were able to acquire new wives (whom they may have also regarded as slaves).
Several articles in this special issue further explore how men were able to control certain categories of women and transfer them to others. Writing on Sierra Leone in the late nineteenth and early twentieth century, Sarah Delius shows that the distribution of women and girls enslaved in raids and wars was integral to both the regional political economy of warfare and related efforts to expand individual households by incorporating enslaved outsiders into existing networks as wives, concubines and dependants. 12 She draws upon Emily Osborne, whose work on Kankan-Bate has analysed how 'political elites use the household as a foundation for statecraft … ', thereby deploying ' … marital bonds and familial ties to build and organize the state'. 13 Baker's 1862 account of the operations of the slave market in Gallabat also points in a broadly similar direction, with enslaved women captured in raids acquiring value as human commodities owing to their sexual and reproductive appeal within elite households.
This relationship between marriage and enslavement was deeply entrenched, and endured in the face of anti-slavery interventions and the establishment of colonial authority. Delius shows that the British-led West African Frontier Force (WAFF) in Sierra Leone were repeatedly accused of taking 'wives and slaves' away from local elites, with escaped slaves sometimes becoming the wives of African members of the WAFF. 14 Crucially, the WAFF was formally responsible for ending the slave trade, although this mission was severely compromised in practice. While the circumstances were different, parallels can again be drawn with Baker's account from 1870, with freed slaves once again becoming the wives of colonial soldiers, and the transfer again taking place under an anti-slavery rubric. A further example of this dynamic comes from Toni Smith, whose contribution to this special issue documents how colonial agents in the Belgian Congo used the violent abductions of women as a strategic tool against African elites whose loyalty they suspected. Smith demonstrates that the violence for which the Congo Free State is renown persisted despite the transition to Belgian rule in 1908, and that sexual violence was intrinsic to its operations. Colonial practices reproduced key features of pre-colonial slave systems in Kasai, with violently abducted women and girls forced to have sex with Europeans and become 'wives' of men and boys conscripted, usually by force, into the Belgian colonial military. 15 These examples help to demonstrate that organized anti-slavery altered the relationship between slavery and marriage without bringing it to an end. Two recurring scenarios can be identified. In the first version, political authorities took active measures to ensure that slave systems continued despite their official embrace of the anti-slavery cause. As Smith observes in the case of Congo, anti-slavery rhetoric was not meaningfully connected to policy, with humanitarian posturing operating alongside decades of state-sponsored enslavement. In some cases, such as colonial Nigeria, this active support for slavery included slave trading continuing under the guise of marriage or concubinage, with men ransoming enslaved women to make them their new wives. 16 In the second version, political authorities attempted to limit the practical effects of anti-slavery reforms. In colonial Africa, this most commonly involved European administrators attempting to dilute their anti-slavery obligations by manipulating the definition of slavery on two key fronts: redefining 'domestic' slavery as 'benign', so that urgent intervention was not required; and restricting the definition of slavery to try and ensure that practices that looked similar to slavery, yet were formally classified differently, were excluded. 17 Marriage was central to these political dynamics. Slave owners threatened by anti-slavery reforms frequently redefined their slaves as wives, concubines or dependants in a strategic attempt to retain as many of their established prerogatives as possible. This strategy often required colonial authorities to look the other way, since the identities of people who remained enslaved were known within their communities. The end result can be best described as a patriarchal bargain, with European colonial authoritieswho were overwhelmingly malereaching a tacit bargain with African political elitesalso overwhelmingly male to minimize disruption to the status quo by establishing pathways that enabled them to uphold their authority within their households. While slavery was recognized as a public matter, intimate relationships between husband and wife were commonly understood to be a private concern.
Efforts to conceal slave status under the guise of marriage were not entirely successful. Enslaved women were sometimes able to take advantage of antislavery reforms to leave their masters and established new lives and new conjugal relationships. In this issue, Morenikeji Asaaju explores vivid examples of these efforts in the court system of Abeokuta in colonial Nigeria during the early twentieth century, where women successfully petitioned for freedom/divorce from their masters/husbands. As Asaaju demonstrates, some women who came before the court based their case on the claim that they were not wives, but slaves, and therefore required their freedom. One major point of contention concerned the status of children, which both husbands and masters sought to maintain their authority over.
Only a minority of the women who came before the Abeokuta court reported that they had been enslaved. The vast majority instead sought divorce on other grounds, thereby compelling colonial authorities to grapple with broader questions regarding how marriage should be both regulated and practiced more generally. In this context, slavery appears as one aspect of larger conversations about marriage which included many other issues, including the status of polygamy, religion, divorce, domestic violence, commercial sex, age of consent, and the morals and economics of bride-price. 18 Colonial outsiders found it difficult to understand the complexities of marriage practices in Africa, especially when it came to households and hierarchies. Writing about Lagos in the later nineteenth century, Kristin Mann observes that 'family relationships were deeply hierarchical. Parents exercised authority over their children, husbands over their wives, and both children and wives were ranked by age and seniority inside the household'. 19 Europeans found these intimate hierarchies and internal household dynamics hard to come to terms with when slavery was legally sanctioned; anti-slavery measures only complicated matters further.
Three distinct lines of inquiry regarding slavery and marriage can be identified in this context. At the level of direct connection, there is the issue of how marriage practices affected the lives of people who were formally enslaved, either while slavery was legal or immediately after slavery had been legally abolished. This is where the connection between these two is strongest. At the level of comparison, there is the further issue of whether specific individual experiences of marriage closely resemble experiences of enslavement. Then, more broadly, we have comparisons between marriage and slavery as institutions. Both individual and institutional comparisons raise questions relating to similarity and equivalence.
One of the core attributes of slavery is unconstrained authority. While different historical slave systems operated in different ways, they nonetheless shared one defining feature in common: masters typically exercised a high degree of personalized and discretionary power over their slaves, including their labour, sexuality, and fertility. Reproductive concerns are regularly cited as a major reason why enslaved women and girls were frequently (but not always) in greater demandand commanded higher pricesin a number of African contexts. 20 Vulnerability and enslavement went hand in hand, since masters and their agents were socially and legally empowered to exercise their authority in violent, abusive and sometimes capricious ways without facing serious consequences. While not everyone exercised their authority in the same way, even rare cases of 'good treatment' carry limited weight, since every interaction between enslaved and enslaver was coloured by the authority of the latter to make decisions over the former that could not be reviewed or revoked.
Similar kinds of observations have been made regarding the institution of marriage. There is a long history of scholarship and activism, frequently but by no means exclusively from a feminist standpoint, where the example of slavery has been repeatedly invoked as part of efforts to analyse the systemic inequalities and vulnerabilities which have been endured by women in general, and married women in particular. 21 This approach does not necessarily presume a direct connection with enslavement, but instead takes up the question of potential equivalence through 'the woman as slave' analogy. 22 This analogy was mobilized in colonial debates over bride-price/bridewealth. During the late nineteenth and early twentieth centuries, Europeans in Africa routinely expressed concerns that bride-price payments by the family of the groom were a de-facto instrument of purchase and abuse. The following passage from a letter by a Christian missionary in the French Sudan in 1905 is a case in point: The general custom in this country is that any young man of the age to get married must buy the one whom he wishes to marry. He makes arrangements with the parents of the young woman and it is with them alone that the price is discussed. The young bride, irrespective of whether she is of free or slave status, is never consulted. One can say that here women are nothing but slaves [captives]. This is the case before their marriage, and even more the case after their marriage. To provide Christian husbands to these young women is at present the best way to achieve the improvement of women's status and the destruction of the most deeply rooted form of slavery. 23 These ideas were not exclusive to missionaries, but were common within colonial circles. They emerged out of the larger history of binary oppositions between European 'civilization' and African 'backwardness' which were key to justifying colonial conquest. 24 While 'normal' (read European) marriages were tacitly regarded as legitimate, an homogenized model of 'African' marriage was analogized to slavery. Few allowances were made for variations in experience, including matrilineal societies, woman-to-woman marriage, or motherhood outside of marriage. 25 The same kinds of arguments and assumptions used to justify colonialism were redeployed in relation to marriage: African men were reduced to patriarchal abusers of African women, 'backward' cultural practices were framed as root causes, and humanitarian intervention by 'enlightened' Europeans was required to 'rescue' victims. 26 The claim that bride-price was analogous to enslavement proved to be an unsustainable generalization. As historians and anthropologists have demonstrated at length, marriage payments do not mean one thing, but instead carry multiple meanings, and therefore cannot be easily comprehended outside the social orders within which they occur. 27 Nuances which are apparent to insiders may not be evident to outsiders looking in, so the same set of practices can look very different depending on the positionality of the observer. It should not be surprising, therefore, that Westerners 'sometimes interpreted such exchanges as signifying the sale of the woman by her family, [while] Africans valued bridewealth for venerating the relationship and for tying families together in an economic and social community'. 28 Marriage payments also tend to be associated with social status and legitimacy. Marriages involving slaves were frequently defined by the absence of bride-price payments, creating an important point of demarcation between slave-wives and wives more generally. 29 European models which treat marriage payments as a sign of enslavement make little sense in this context. It is also important to keep in mind that children born to free women were typically treated differently from children born to enslaved women (married or otherwise). Being enslaved and being forced to marry as a free person were usually understood and experienced differently, yet whenever marriage is analogized to enslavement these nuances get overshadowed by a singular model of oppression on the basis of gender. One of the main problems with this approach is that freeborn African women who experienced arranged marriages and/or abusive marriages rarely classified their experiences as enslavement, especially when they were familiar with slavery as a living institution in their societies. Comparative analogies run the risk of imposing an external category upon individuals who would not describe their experiences in such terms.

Pre-Colonial Hierarchies and the Politics of the Household
The right to make decisions over marriage is one of the bundles of powers and prerogatives associated with enslavement. Slaves have historically been forced to marry their master or their master's kin, other slaves, or other persons, resulting in a variety of scenarios where the category of 'slave-wife' could be applied. 30 'Slave-husband' was less prevalent but not unheard of, and includes some categories of elite male slaves who married free women. 31 Histories of enslavement in Africa frequently talk in terms of girls being captured, purchased, or given as tribute, and thereby forcibly integrated into their masters' household as wives and conjugal partners, with a further emphasis on children born out of these relationships. Captives could sometimes be married prior to being abducted, so the loss of marital partners was part of the violence of enslavement.
None of these issues are distinctive or exclusive to African history, but instead have applications in many corners of the globe. In the pre-abolitionist world intimate relations between slave and free were common. So common, in fact, that they were subject to legal regulation. Roman law stated that no lawful marriage could be concluded between free and enslaved persons. 32 Early Christianity did not change this. 33 Faced with frequent exceptions to these rules, Byzantine Emperor Justinian decreed that Christian marriage amounted to manumission in the sixth century. This resulted in rising slaveholders' opposition to the marriages of their slaves. In the eleventh century, the Byzantine Emperor Alexius Comnenos I aligned himself with developments in Canon Law that saw all Christians, of whatever status, as eligible for the matrimonial state, and decreed that slaves could marry without altering their status. 34 This also meant that the marriage of slaves would not undermine the slaveholders' proprietary claims upon them. In the Christian world, which until the seventh century included large parts of North Africa, periods when marriage and slavery were seen as incompatible alternated with periods when they weren't.
In the late Medieval Mediterranean, enslaved women provided work and sexual services for their male owners. As summed-up by Hannah Barker, 'in an Islamic context […] sex with slave women produced heirs, while in a Christian context, it produced property'. 35 In the thirteenth and fourteenth centuries, Christian practice started to resemble Islamic practice, although there remained significant variations both within and between different jurisdictions. 36 One fundamental difference between Christian and Islamic contexts was that, in the former, children born from sexual relations with slave women acquired their mother's (slave) status, whereas in Islamic contexts they generally acquired their (free) father's status. As long as the master recognized the child as his own, an enslaved woman's status was altered to that of 'mother of the child' (umm al-walad), giving her more rights than to other slave women: the umm al-walad (usually translated 'concubine') could not be sold and was manumitted at the death of the master. 37 Yet an umm al-walad occupied a lower rank than free wives, creating an intermediate status where she 'owed more service … than a free woman, but less than that required of a female slave'. 38 In Islamic Africa the purchase of enslaved women and girls as concubines was widely accepted as a way of circumventing the 'four wives barrier'. 39 Within elite circles this frequently resulted in huge households, which Europeans never failed to exoticize. In seats of political or mercantile power such as Kano and Zanzibar, hundreds of concubines could be bound to key political figures, and regularly played important diplomatic and administrative roles. 40 There is now a substantial literature on elite women, both enslaved and free, yet the vast majority of households were not organized on such grand terms. 41 It is also important to recognize that relationships were dynamic, not static. In both Islamic and non-Islamic contexts there were many layers when it came to social and legal status which could change over a life-cycle. 42 The ideal types which are found in legal definitions and religious texts rarely capture the full diversity of lived experiences. As part of his analysis of slavery and marriage in nineteenth-century Cairo, Terence Walz observes that Muslim Egypt distinguished between 'marriage, a contractual relationship between free individuals, and concubinage, a legal relationship between a man and a slave woman he owned'. 43 Despite 'slave wife' not being a legally meaningful category, Walz found that it was nevertheless frequently used in practice: in Cairo's census otherwise unmarried men were living with concubines termed, incorrectly, 'slave wives' (zawjat jariya), a terminology that 'cannot be taken literally, for legally a man had to free his slave in order to marry her'. 44 De facto marriages between persons of free and slave status were common, taking a broad range of shapes and forms in both Ottoman North Africa and non-Islamic Africa. It is important to keep in mind, moreover, that the contemporary meaning of the English term 'marriage' regularly ends up obscuring the operations of distinct institutions. A good example of this comes from R.S. Rattray, who in 1929 listed six expressions referring to the different ways in which marriage was conceptualized in Twi-speaking Asante: one of these, afona awadie, referred specifically to marriage between a free man and a slave. 45 The kinds of experiences outlined above look quite different from the experience of Africans enslaved across the Atlantic, where the integration of enslaved persons into freepersons' kin-groups was limited by racist injunctions against 'miscegenation'. 46 Slaves were not forced to marry their master, or master's kin, yet they remained subject to sexual abuse. Voluntarily chosen conjugal partnerships generally lacked formal standing. Being able to legally marry was one of the gains of abolition, but it also proved to be a double-edged sword. As Tara Hunter has demonstrated for the nineteenth-century United States, the right to marry introduced new standards by which the intimate relations of formerly enslaved persons were legally and morally scrutinized. Black couples and families risked persecution when they could not, or would not, abide by rules governing marriage. 47 The history and legacies of Atlantic slavery shaped European views about slavery and marriage in Africa. Many of the key features of slave systems in Africa went against their expectations, since slaves who had been captured or purchased were frequently married by their captors or owners. Newly acquired enslaved wives occupied a low rank in the gradations of married identities as 'slave spouses', but could rise-up in wifely hierarchies by gaining the appreciation of their husband/ master. Kinship and enslavement were not polar opposites, yet as far as European missionaries were concerned terms such as 'slave kin' and 'slave wives' were oxymorons.
Rosine Opo's story offers an illustrative example of these complex relationships between slavery and kinship. Rosine's father, Kwaw Kutanku, was probably a man of slave descent who had acquired wealth and fathered many children with his female slaves in the matrilineal Akan society of the Gold Coast in the first half of the nineteenth century. 48 Rosine, born around 1832, was sold to a Euro-African on the coast by Adum Tokori, Rosine's father's 'brother' in 1840. Rosine escaped in 1844 and returned to her homeland. She was not forced to go back. She married a respected elder, royal treasurer Kwaku Sae, who paid the traditional 'head money', thereby acquiring rights to have a sexual relationship with her. Yet, since Kwaku Sae neither bought nor redeemed her, Rosine remained a slave in possession of her father's family. In 1856 Kwaku Sae was baptized together with several members of his family, including Rosine, by the Basel missionaries. Shortly afterwards fights broke out between Rosine, her husband, and her free co-wives. During one of these fights Rosine attacked Kwaku Sae with a wooden pestle; she later had a relationship with Charles Irinkye, a man pawned to Kwaku.
Rosine was expelled from both her Akan family and the Christian community, which would not accept her lack of repentance for committing adultery. Her children from Kwaku Sae were taken from her by the free members of his family and she was put in chains as punishment. The missionaries intervened in her defence. She was released, but her husband Kwaku Sae swore an oath against her and Charles Irinkye. Forgiven by the missionaries, Rosine hid herself for several months on the Mission's grounds, fearing that as a rebellious slave she might be sacrificed at a chiefly funeral. She was later re-sold several times by her father's family until the Basel missionaries bought her and her child's freedom in 1868, eight years before the British passed an abolition decree in the Gold Coast in 1874. Rosine was slave, wife, and kin. These multiple statuses mutually influenced each other and shaped her interactions with relatives and owners.
Slaves also married other slaves. Many of these unions were informal from the perspective of slaveholders. Unless masters had been involved in the arrangements as sponsors, formal recognition or protection would be withheld. Lower-ranking slaves in some slave systems could ask for permission to marry, which suggests that they were able to exercise at least some degree of choice, albeit within narrowly circumscribed parameters. In most cases, however, slaves had little or no input when it came to their conjugal partner. They instead entered into marriage owing to decisions made by their enslavers. Take, for example, the following narrative recountedby a descendant of slaveholdersto Alice Bellagamba, which describes an agreement between two slaveholders regarding slave marriage in the Gambia: Only the two of us would talk. The agreement would be some money and a female goat. That is what I would pay and then we would tie the marriage. When you go, you would say to your slave woman: 'your husband will be Bakoyo's so and so slave'. When I return home, I would say to my male slave: 'your wife is Mamadi's so and so slave woman'. This is how they were married. My male slave, if he married your slave woman, their children belonged to you. They would be with you and work for you. If I had a female slave, and one of your male slaves took her, all her children would come and work for me. This is how slaves married in those days. 49 Testimonies by descendants of nomadic Tuareg slaves in Niger similarly distinguish between marriages between slaves of the same slaveholder and slaves of different slaveholders. Further distinctions were also drawn between the marriages of slaves who lived with the masters (and followed them in their travels), and marriages of slaves who led a sedentary existence in separate hamlets. Slaves who lived attached to their masters married mostly with other slaves of their masters. When they married outside the group of their masters' slaves, the bride's master was contacted by her father if he was there. Then the bride's master would inform the master of the husband that one of his male slaves wanted to marry one of his female slaves. Then they would agree on the arrangements. The master of the groom had to pay bridewealth for his slave's future wife. Bridewealth went to the master, not the father. The husband, who was a slave, would spend the night at the camp of his wife's master, with his wife, and the day in the camp of his own master. But when the marriage was between two slaves of one master, usually there was no bridewealth involved. If bridewealth was paid, the amount was undetermined. When 'the distant slaves' [the slaves who lived in slave hamlets] wanted to get married, the [leader of the free Tuareg group] of the groom only had to give him his authorisation, but would not contribute to bridewealth. 50 These kinds of testimonies come from parts of Africa where the legacies of slavery have lasted into the present day, with descendants of slaveholders and slaves reminiscing about the living conditions of, respectively, their slaveowning and enslaved ancestors. There are methodological challenges associated with these kinds of sources, yet they remain valuable since there are few precolonial or early-colonial African sources on the intimate life of enslaved persons.
As these examples demonstrate, slaves were usually married via processes which were quite different from the marriages of free people. It should also be clear, moreover, that marriage did not necessarily bring about an end to or even a dilution of enslavement. When many slaves got married they continued to be enslaved and children from their union were considered the property of the female slave's owner. However, there were also times where marriage could be associated with positive changes in how slavery was experienced, including cases where men could marry female captives and slaves who left slave status behind in the process of becoming wives. Toyin Falola suggests that this was the norm in parts of Yoruba, where slave women 'could become "dowry free" wives to the freeborn and thus became free'. 51 Sexual abuse of slaves is a tragic yet ubiquitous attribute of enslavement, so there was no need to get married in order to ensure sexual access. Marrying a slave meant more than sex. Slave wives were valued not only for their capacity to have children, but also as the mothers of heirs, and giving birth to an heir invariably improved their status. 52 In Africa before the twentieth century, where new land was relatively accessible and labour was the scarce resource, assimilative social logics prevailed over strategies to minimize the number of heirs who could make claims on familial resources. Assimilation prevailed where a man's power depended largely on his control over people. Counting the offspring of slave wives as heirs enlarged the size of households, lineages, and clans. 53 This had the effect of intensifying the gendered nature of slavery by making enslaved women particularly valuable compared to enslaved men. It also made slave marriages quintessentially political, since they influenced the dynastic structure of society. 54 For elite men polygamy was a feature of kinship that served political ends. 55 The cultural value placed on a large number of children turned fertility into a political strategy for both men and women.
Men rose in status by heading large families; women could use their fertility for the purpose of social advancement. Wives in polygamous households could be hierarchically ranked from first, free, senior wife to junior, enslaved wife or official concubine. This resulted in multiple lineages descending from the same polygamous man and his higher-ranking wives. Lower-ranking wives were often unable to create separate branches (stirpes) of the family and their offspring would be attached to the stirpes of official wives, where low-ranking wives played a menial role. These complex hierarchies are taken up by Ettore Morelli, whose contribution to this special issue focuses upon the household dynamics and conjugal strategies of Lesotho's king, Moshoeshoe (died 1870). After 1833 Moshoeshoe is reported to have been married to between 30 and 100 wives (estimates varied with time and source). Eugène Casalis, a missionary who visited Moshoeshoe in June 1833, was introduced to 31 of his wives, only one of which he characterized as 'queen' and head of the first and most distinguished of three 'houses'. These 'houses' corresponded to three ranked stirpes descending from three wives, mirrored in the topography of Moshoeshoe's capital. Each wife and her offspring occupied separate neighbourhoods, and 'houses' functioned as administrative units of the kingdom, endowed with separate fields, servants, military regiments, and public functions in government. 56 Powerful men sought to acquire a large number of wives for themselves. They also provided wives for male clients unable to do so on their own, or through their lineage. Claude Tardits wrote that at the court of pre-colonial Bamun kings, in what is today northwest Cameroon, hundreds of courtiers and clients benefited from the gift of thousands of enslaved women. 57 Maurice Bazémo argued that in pre-colonial Burkina Faso people from all ethnic groups could acquire a [slave] wife. The father could buy one for himself, for one of his sons, or for any other family member. The female slave could be donated as a gift and thereby used instrumentally to strengthen social relations. 58 Subordinate men at the receiving end of these gifts could be free junior kin, domestic slaves, or persons of slave descent who lacked support from elders of their own lineage willing to commit the lineage's corporate capital (usually livestock) to the payment of bride-price for junior members.
These logics help to explain why slaveholders sometimes sponsored their male slaves/subordinates to acquire a wife: this created a debt between them. This debt could mean different things, and could sometimes go as far as the sponsor claiming as his own the offspring of a man whose union he had enabled. Morelli shows that Moshoeshoe claimed the offspring of subordinate men whose wives he had procured (mostly by paying bride-price for them) as his own descendants. Since descent followed bride-price (bohali), elite men multiplied their offspring by acting as patrons for low-ranking men for whom they paid bohali. Male clients who had obtained wives as gifts lacked guardianship over their own offspring. This is similar to the fate of slave fathers who were denied social paternity because their biological offspring belonged to the owner of the enslaved mother. However, in this scenario the determinant tie of dependence was between the male patron and his male client rather than between male owner and female slave. Together with adoption and pawnship, these practices manipulated kinship to reproduce and strengthen political hierarchies.
These relationships and dynamics have a strong intergenerational component, and could result in very different outcomes for the offspring of mixed-status unions, depending on the specific hierarchical position of their slave and free genitors. Some persons of mixed descent were able to secure the honours and chiefly prerogatives of their fathers. Others remained marginal because the status of their enslaved mothers was particularly 'sticky', and therefore carried social stigma for themselves and their children and even children's children. As a general rule, slave status was inheritable through descent lines, but what this implied in practice, and how many generations (if any) would be affected, changed across societies and periods. Organized anti-slavery added a new political dimension to this already complex situation.

Contestation and Complicity under Colonial Rule
There were many ways of being a slave and many ways of being a wife within African societies. This created constant confusion for European observers and administrators who conceptualized both in binary terms: one was, or wasn't, married; one was either slave or free. This logic overrode considerations about how a person was enslaved, or how she was married. Europeans struggled to conceive of a 'slave spouse' and tended to see all conjugal slaves as simply slaves. If marriage was like slavery, then it was slavery. 59 European ideas of marriage were also bound up in idealized notions of sexual 'purity', Christian virtue, and masculine protection. Unmarried women were seen as posing a danger to society's mores and were pressured to marry, including through fiscal policy. 60 Polygamous marriages in Africa were judged against this idealized picture of marriage in Europe and found wanting. 61 European actors also tended to interpret the supposedly fuzzy boundary between slaves and wives as evidence of the abusive nature of African marriage. Once this idea became established, it found corroboration in biased observations of multiple African practices and customs. Bride-price was a case in point. As Kopytoff and Miers argued back in the late 1970s, the established European definition of slavery as totalizing control over persons translated into a popular perception of slaves as persons treated like commodities. The transactions of wealth that accompanied marriage, which Anglophone observers termed 'bridewealth' or 'bride-price', were assimilated to the purchase of a wife and therefore to slavery. 62 Seeing 'African marriage' as slavery prompted interventionism by missionaries and administrators and, later, by international organizations. The analogy with slavery was not morally neutral: it invited correction and reform based on the view that marriage should not resemble slavery. When it did, colonial administrations were supposed to intervene. 63 Colonial administrators were often uncertain about how to approach bride-price. On many occasions, bride-price was portrayed as a sign of slavery, yet there were also other times when it was cited as proof that specific cases of marriage were not slavery, since established protocolswhich included the payment of bride-pricehad been followed. Marriage was also widely regarded as a safeguard against the allegedly loose mores of 'the African woman', so it could sometimes be portrayed as a positive alternative even if it was said to resemble slavery. 64 African women tried to leverage these colonial approaches to marriage to their advantage. There were contexts where free wives felt maltreated and compared their circumstances to those of slaves. 65 This comparison was all the more consequential where free wives could point to 'slave wives', a less fortunate category they did not belong to and, in their view, should not be confused with. Used by freeborn wives, the analogy accused their husbands of failing to respect their prerogatives as free wives. But women who had actually been enslaved described their marriage as slavery in a literal sense. Asaaju's article shows that in colonial Abeokuta slave wives sought to convince native courts that they were 'slaves, not wives' in an attempt to end relations that they perceived as abusive. Women who had faced the humiliations of slavery convinced judges that they had been acquired through purchase or capture; had been illtreated and forced into unwanted intimate relations; and that no bridewealth or other customary gifts had been provided by their masters/husbands. These contests often focused on who would retain custody of a woman's children. This involved a concern for which social units (lineages, clans) children would belong in and to; what their social status would be; who would have rights on their labour; and (if the children enjoyed free status) who would one day contribute to their marriage payments and inheritance. Women could elope but, if they wished to bring children from previous unions with them, they often had to turn to litigation in courts, which did not always favour formerly enslaved mothers. New partners and new relatives who had the means to pay redemption fees had to intervene in support of slave mothers whose rights to the custody of their children were precarious. Delius discusses the case of Mantene, formerly enslaved wife of paramount chief Balla in Sierra Leone, whose three-year-old daughter was redeemed by the husband of Mantene's recently liberated mother. The payment could be interpreted either as a refund for Mantene's bride-price (which would have been the course of action required if Mantene was considered free), or as the redemption fee for a slave child (required to redeem Mantene's child if Mantene was deemed a slave).
Not having a lineage of their own disadvantaged slave women. There were also times, however, where enslaved women could rely on the partial protection of unconventional lineages following abolition. As abduction and slave trading became riskier for potential slaveholders, subordinate women gained confidence and ran away more easily. 66 They did not hesitate to denounce those who sold and bought them and those close to them even when doing so exposed them to backlash by powerful men. In colonial Niger, for example, the enslaved woman Tadoutchi accused her owner Barau when he sold her younger brothers. Tadoutchi denounced his slave trading activities to the colonial administration. In revenge, Barau sold her and her baby into the trans-Saharan slave trade. But thanks to the efforts of a committed District Officer and collaborating local chiefs, she was eventually retrieved from the depths of the Sahara and returned to testify against Barau together with the brother she had saved. 67 Female slaves knew that they were particularly valued for their sensuality and fertility, and they used these characteristics to attract suitors and thereby reduce their social isolation and vulnerability. 68 Bwanikwa's story, as she narrated it in Luba to missionary Dugall Campbell, who translated and published it in 1934, is a vivid illustration of this. Bwanikwa was a slave living in the area to the south-west of Lake Tanganyika when Europeans arrived and challenged the power of the local warlord Msiri; At this, some of us slaves saw our chance, and fled. We scattered. Men, tired of [Msiri's] despotic rule, would take some or other woman slave, and both would head north, south, east, or west, in search of freedom and a new start in life. When possible, each headed for the old homestead. A well-known elephant hunter and fellow slave in the same village, whose name was Kabongo, took me, and we ran off east. 69 Kabongo and Bwanikwa, having seized this opportunity to run away and regain free status, sought refuge in the territory of another local ruler, chief Kazembe. Kazembe accepted them but, attracted to Bwanikwa, he wanted her for himself. Kabongo, unable to antagonize the chief, preferred parting from Bwanikwa than letting Kazembe have her. He sold her to Bihean traders who would in turn sell her off. Bwanikwa escaped and reached a missionary station. I met a man named Wafwilwa, who, seeing me alone, asked me to be his wife. I refused at first, but he persisted, and would not leave me. I had need of a protector, so finally gave in, and became his wife. We lived near to the mission at Lufoi. […] On arrival at Launguisi, in British territory, he sold me secretly to some Arabs for calico. 70 Bwanikwa was unable to resist successive captures. At each re-enslavement, she used 'marriage' to tie to herself male partners who thereupon were under a social obligation to protect her. Admittedly, the terminology of 'marriage' in Bwanikwa's account is ambiguous. It is unclear whether the marriages she describes in passages such as 'I had need of a protector, so finally gave in, and became his wife' were formal marriages contracted following prescribed rituals, or were informal but relatively stable unions. The latter interpretation seems more likely. Yet it is meaningful that Bwanikwa and other women in similar circumstances would use the terminology of 'marriage' to refer to relations that accrued protection to them. This choice of wording suggests that they recognized the sexual and reproductive potential of these relationships.
Colonial measures to prevent slaving and the slave trade made it harder for African men to acquire female slaves. Wealthy free men who had hitherto been able to access a large number of wives and concubines by integrating enslaved women in their households now had to adapt to the decreased availability and higher cost of slave wives. The decline of slave trading was usually gradual rather than immediate, but it nonetheless gave people who were already enslaved additional options and greater leverage. This influenced internal household dynamics. Senior wives who relied on the services of low-ranking wives and/or slaves also saw their access to subordinate female labour threatened by anti-slavery measures. As Barbara Cooper showed for the town of Maradi in today's Niger, this resulted in an increased demand for concubines who were more intimately integrated in the household than slave women and therefore harder for European administrators to identify as slaves. 71 The options available to former slaves should not be overstated. Formerly enslaved men who walked away from their owners lacked access to capital and lineages. They struggled to amass the wealth required to marry free women. The stigma associated with slavery followed them even when they were able to resettle in locations far from the sites of their erstwhile enslavement and access independent incomes and land. Their new landlords still treated them as low-ranking strangers and gave them slave women as wives. Based on research on migrants classified as slave descendants in the Upper River Gambia, Paolo Gaibazzi has showed that, in a matrilineal context, this meant that the offspring born of these unions acquired their mother's slave status in spite of their fathers' struggles to emancipate themselves. 72 In the first decades after abolition formerly enslaved men frequently married formerly enslaved women who, like Bwanikwa who ran away with Kabongo, valued marriage as a potential source of support and protection. As a general rule, available sources do not provide insights into people's feelings. It is possible to surmise that empathy, shared experiences, and common objectives may have resulted in mutual affection. But this was not always the case: as noted, Bwanikwa was resold by Wafwilwa. Although colonial abolition broadened the range of options accessible to (former) slaves, no formula can predict how this altered the marriage strategies of free and unfree persons. People's emotions, so important in relation to marriage, influenced this history in unpredictable ways.
Military institutions also played a central role. Men of slave descent who enrolled in colonial military units exited a hierarchy of status and entered a new hierarchy of military valour. This gave them opportunities to loot villages and take women and children away from the households of freemen as booty. The 'women of the tirailleurs' who accompanied the tirailleurs sénégalais involved in the French occupation of West Africa were an internally diverse group. They included former slave women who had run away from their owners' camps and followed the tirailleurs when the latter reconnoitred the region; women abducted by the tirailleurs in the course of military campaigns against hostile groups; and women who escaped from the harems of defeated local rulers. 73 Some women were complicit in their own abduction. They agreed to be their abductors' 'wives' and to run away with them in exchange for protection and support.
In this issue, Seymour, Apio, and Rossi's article describes the zeribas of Nubian soldiers who served under Egypt's Governor Emin Pasha and in the British IBEAC's Uganda Rifles. 74 Frederick Lugard mentions that each soldier was accompanied by an average of eleven dependents, mainly wives, concubines, and female and male slaves acquired in raids or inherited from others in their military corps. 75 In Sierra Leone, Delius shows that the men who joined the WAFF after having been enslaved attacked the households of elites and took away the chiefs' (enslaved) women during their military expeditions. In Congo, Smith shows that European officers and locally recruited soldiers and police regularly engaged in sexual violence. This was an attack on both the bodies of African women and the authority of male slaveholders, chiefs, and household heads. Sometimes formerly enslaved men who became colonial soldiers supported the emancipation strategies of female slaves, but they also used their newly acquired autonomy and power to abduct and enslave women. European officers who posed as abolitionists sometimes felt entitled to subject African women to sexual abuse and control tantamount to possession.
As these examples show, European attitudes towards violence and exploitation were self-serving. All kinds of abuses were associated with colonial rule, yet Europeans were adept at excusing and justifying their own abusive conduct while denouncing African 'backwardness'. A great deal of hypocrisy is evident here. Colonial authorities legally prohibited slavery, but then made extensive use of forced and indentured labour schemes that closely resembled the slave systems they had formally abolished. The promise of anti-slavery was repeatedly undercut and subverted by colonial interests. Much the same can be said in relation to marriage. If marriage and slavery were found together, then authorities were faced with a challenging social and legal problem. If marriage and slavery could be cleanly separated, then the problem went away. Many colonial administrators were keen to declare that the problem had gone away, which enabled connections between slavery and marriage to continue in covert forms.
The relationship between slavery and marriage in Africa was pulled in two competing directions during the 1950s. At an international level the two were pushed closer together, with a new United Nations Convention in 1956 which formally connected 'servile' marriage with slavery. At the same time, there were also countervailing developments which pulled them apart, as colonial classifications came under increasing scrutiny as part of moves towards national self-determination. The most immediate issue was the status of brideprice payments, which some colonial missionaries and administrators continued to regard as a de facto instrument of enslavement.
The most important development during this period was the drafting of the 1956 Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery. This Convention was primarily designed to both 'augment' and 'intensify' the earlier 1926 Slavery Convention, which had taken the key step of establishing an authoritative definition of slavery under international law. 76 As the title of the 1956 Convention suggests, one of the main points at issue was the status of practices which were regarded as similar to slavery. This was by no means a new issue, but most attention had been previously directed towards colonial forced labour systems (the 1926 Convention calls for 'measures to prevent compulsory or forced labour from developing into conditions analogous to slavery'). 77 The 1956 Convention took the further step of formally defining four new practices as 'similar to slavery': debtbondage, serfdom, servile marriage, and the delivery of 'a child or young person under the age of 18 years … to another person … with a view to the exploitation of the child or young person or of his labour'.
Servile marriage was formally defined in the following terms: (c) Any institution or practice whereby: (i) A woman, without the right to refuse, is promised or given in marriage on payment of a consideration in money or in kind to her parents, guardian, family or any other person or group; or (ii) The husband of a woman, his family, or his clan, has the right to transfer her to another person for value received or otherwise; or (iii) A woman on the death of her husband is liable to be inherited by another person; Further provisions in Article Two also call for the 'consent of both parties … to be freely expressed in the presence of a competent civil or religious authority', together with suitable measures for marriage registration and minimum ages for marriage. 78 These provisions are heavily indebted to colonial conversations regarding bride-price payments in Africa. Especially notable here are the references within (i) and (ii) to a woman being 'promised or given in marriage on payment' and being transferred 'to another person for value received'. This definition also specifically focuses upon women, despite the fact that men and boys can also be compelled to marry. The language found in (i) is particularly revealing, since it makes servile marriage contingent upon payment. As discussed above, slaves in Africa could be forced into marriage without bride price payments being made, yet this scenario falls outside the scope of the 1956 definition. This is also true of violent abductions in conflict settings which result in forced marriages, which usually do not feature marriage payments either. Making payment a precondition has the effect of narrowing the scope of the definition (both of these examples still fall within the remit of the 'without right to refuse' provision).
The importance attached to 'marriage on payment' offers an insight into the kinds of thinking which informed the Convention. The drafting process which led to the 1956 Supplementary Convention can be traced back to the deliberations and outputs of an earlier Ad Hoc Committee on Slavery (1950)(1951) established by the United Nations. One of its most important contributions was the identification of specific practices as similar to slavery. Its rationale for formally connecting marriage with slavery is worth quoting at length: The Committee then considered what has commonly come to be known as 'brideprice'. It noted that, according to the information it has received, a special custom exists in many parts of the world whereby a man cannot obtain a wife for himself without offering money or services to the family of his prospective wife … in certain parts of the world the bride to-be is regarded as the property of her father, or that at least his right is recognised to dispose of her in marriage; and that as a widow she is regarded as part of the heritable property of her husband and as such passes to the person who inherits that property on his death … The fact that the bride is sometimes married without her consentand often when too young to consent even if she were askedadds to the deleterious practical effects … The Committee decided, therefore, that it would consider as a form of servitude the practice whereby a woman is given in marriage, without the right to refuse, at a price or under conditions which give to the husband, to his clan or family, a right of disposal over her or over her children and permit her exploitation for the advantage of others. 79 The language used here was later refined via the drafting process that led to the 1956 Convention, but there can be little doubt about the connection between them. 80 As Suzanne Miers has documented, the work of the Ad Hoc Committee was strongly informed by one of its five members, Charles Greenidge, who served as Secretary of the London based Anti-Slavery Society from 1941 to 1955 (when he then became Honorary Director). 81 Born into a privileged family from the planter class of Barbados, Greenidge was a Cambridge trained lawyer whose career included a posting as Solicitor General of colonial Nigeria (1936)(1937)(1938)(1939)(1940)(1941) and other prominent positions within the British Empire. These experiences in Africa strongly informed his insistence that bride-price in Africa was a vehicle for enslavement: In customary marriage in Africa, the union becomes a marriage on payment by the husband of the bride price to the father or of the guardian of his wife. On doing so he is entitled, according to their customary law, to sell her to another man or if he predeceases her, she is inherited by his heir. If either of these acts is not an exercise of ownership over a person I do not know what it is. In my own experience in Nigeria I know an African who was only stopped from selling his wife to buy a railway ticket by his English employer giving him a railway ticket. 82 This passage also captures some of the exasperation which Greenidge and others of a similar mind experienced throughout this period. Whenever they made efforts to draw attention to bride-price in Africa as a form of slavery, they invariably encountered public opposition and political pushback.
In 1957 the Anti-Slavery Society invited Dr Lucy Mair, an anthropologist at the University of London, to give a high profile public speech on 'African marriage as servile institution'. 83 They had hoped that her speech would build upon the link with slavery established via the 1956 Convention, but things did not go as planned. Mair instead maintained that 'the extent to which African girls are coerced into distasteful marriages is very much exaggerated … [yet] it does happen sometimes'. She further declared that references to '"buying wives", "servile marriage", "bond-wives" and so forth', were unhelpful and misguided. They were instead characterized as part of a 'mythology of misinterpretation. If a woman is purchased, she must be her husband's property; therefore, she is a slave'. When asked how these misconceptions regarding bride price had arisen, Mair further responded that 'the first people [Europeans] who went to Africa were rather prejudiced. They really did think that anything which was different from the way they did things must be wrong'. 84 By the late 1950s the Anti-Slavery Society had belatedly recognized that it has acquired 'a bad name for uninformed condemnation of all forms of African marriage', and that much greater cultural sensitivity would be required in a post-colonial world. 85 These differences of opinion over bride-price in Africa were emblematic of broader changes in this period. By the 1950s the political legitimacy of colonial rule in Africa was rapidly eroding. National self-determination was the order of the day. As part of this transition the core features of the colonial project were recast in a new and different light. Many colonial era classifications were reevaluated or rejected as part of this transformation. Past European evaluations of African societies and peoples came to be stronglyand understandablyassociated with racism and ethnocentrism, and therefore lost much of their earlier legitimacy. The connection between bride-price and slavery was acutely vulnerable in this context.
More recent works have also contested colonial portrayals of men in Africa 'as the custodians of tradition and women its hapless victims'. 86 One major focal point has been the degree to which 'age-old' traditions concerned with gender and sexuality in Africa are colonial inventions arising from the imposition of European gender binaries. 87 In addition, further questions have been raised regarding the portrayal of women in/of the Global South as an 'homogeneous, undifferentiated group leading truncated lives, victimized by the combined weight of their traditions, cultures, and beliefs'. 88 These critiques are applicable to many issues, but they have particular resonance when it comes to the link between bride-price and slavery, which was strongly associated with the portrayal of married women and girls in Africa as disempowered victims and the reduction of culture to 'pre-modern' pathology.

Continuing Legacies and Contemporary Reconfigurations
Critiques of colonial discourses have had a major effect upon the subsequent reception of the formal link between slavery and marriage which was established via the 1956 Convention. Many states have ratified the Convention (there were 123 signatories by 2022), but there has also proved to be relatively little appetite for further conversations about servile marriage. Countless discussions have taken place regarding abusive marriage practices in Africa during the post-colonial era. The vast majority of these conversations have taken place without engaging with slavery. This includes research and advocacy focusing upon sensitive issues such as forced and early marriage, bride abduction, age of consent in marriage, and intimate partner violence within marriage. 89 Efforts to reform abusive marriage practices also found their way into numerous international treaties, including the 1962 Convention on Consent to Marriage, the 1966 International Covenant on Civil and Political Rights, and the 1979 Convention on the Elimination of All Forms of Discrimination against Women. None of these instruments take up the link between marriage and slavery. The 1956 Convention successfully pushed marriage and slavery together, but relatively little came of this connection during the second half of the twentieth century owing to the post-colonial political dynamics pulling them apart. However, there has been a notable resurgence of this connection over the last two decades. More attention has recently been paid to abusive marriage practices as forms of slavery than at any point from the 1960s onwards. 90 Three main themes can be highlighted: forced marriage as 'modern slavery', forms of abduction and conjugal slavery in recent African wars, and the intergenerational inheritance of slave status in postslavery societies in Africa. 91 Forced marriage has recently been incorporated into the larger pantheon of issues associated with modern slavery and human trafficking. Marriage does not attract the same kind of interest as commercial sex, but it is now usually included in lists of practices said to be forms of modern slavery. 92 In 2012, the United Nations Special Rapporteur on contemporary forms of slavery, Gulnara Shahinian, published a thematic report specifically focusing upon marriage which observed that 'the idea that forced and early marriages are forms of slavery and, therefore, servile marriages has been lost'. 93 Recent modern slavery campaigns have resulted in this connection being 'found' again. This includes the 1956 Convention, which is now regularly invoked to authorize the classification of forced marriage as slavery (its specific clauses are rarely discussed in any depth, and the provision regarding widow inheritance is regularly omitted entirely). A further connection has also been made in terms of marriage as a vehicle for human trafficking. 94 One important recent example of forced marriage as modern slavery comes from the International Labour Organisation (ILO). In 2017, the ILO published research with the Walk Free Foundation which estimated that 40.3 million people were subject to modern slavery. This global estimate was divided into two main subcategories, with 24.9 million people said to be subject to forced labour and 15.4 millionor 38% of the overall totalsaid to be subject to forced marriage as a specific form of slavery. 95 Their forced marriage research was structured around two deceptively straightforward questions. National household surveys were administered asking respondents whether 'they had ever been forced to marry' and 'if they had consented to the marriage'. 96 An affirmative answer to both was coded as forced marriage and therefore modern slavery. By applying complex algorithms to these answers the ILO concluded that the prevalence of forced marriage and modern slavery was 'highest in Africa'. 97 This exercise raises analytical and political questions. To get to their 15.4 million estimate, the ILO also had to assume that it was possible to make a cross-cultural determination when it comes to consent, but this is much easier said than done. It is relatively easy to point to the absence of consent in scenarios featuring direct violence, high levels of overt coercion and/or young children, but there can also be many shades of grey when it comes to how and why people end up being married. 98 The approach favoured by the ILO is organized around two alternatives: forced and free. Everything on the 'free' side of this binary is classified as legitimate based upon consent at the point of marriage, which can have the effect of tacitly excluding issues which take place within marriage from scrutiny. Any marriage that rises above the threshold of enslavement is regarded as a legitimate marriage. At the same time, everything on the 'forced' side of the binary gets classified as a form of slavery, thereby collapsing different kinds of marital relationships and social positions into a singular modelenslavementwhich tends to be associated with disempowered victimhood. Here the primary rationale for linking slavery and forced marriage is political rather than analytical. By invoking slavery, campaigners and organizations hope to raise awareness regarding issues which might otherwise be marginalized.
A second theme revolves around experiences of wartime sexual violence. One of the main entry points is the now familiar idea of 'rape as a weapon on war', which gained traction from the mid-1990s onwards as part of larger political efforts to strengthen responses to experiences of sexual violence within both peacebuilding efforts and transitional justice mechanisms. 99 Thanks to these efforts, much greater attention and investment has focused upon sexual violence and its effects, with the Women, Peace and Security Agenda (WPS) serving an a organizing framework. One key focal point is criminal justice reform, especially internationally, where there has been a sustained effort to hold high profile perpetrators criminally accountable for acts of sexual violence via the International Criminal Tribunal for Rwanda, the Special Court for Sierra Leone (SCSL), and the International Criminal Court (ICC).
Legal discussions have increasingly circled back to both slavery and marriage. What legal status should be assigned to individuals who were violently abducted during conflicts and held against their will? Does this constitute enslavement? What legal status should be assigned to women and girls who endured long-term patterns of sexual violence after being violently abducted? Does this constitute sexual slavery? What legal status should be assigned to women who are violently abducted and forced to 'marry' one of their captors, and to thereby live with them in a domestic setting? Is this enslavement? Forced marriage? Does labelling it 'marriage' debase marriage? If 'marriage' is the wrong term (it is nearly always italicized), is it preferable to talk in terms of conjugal slavery, a form of slavery resembling marriage? 100 These questions have generated intense discussions over specific legal questions, and there have been several important recent rulings from specific cases.
Two key examples are the Armed Forces Revolutionary Council (AFRC) trial before the SCSL (key rulings in 2007 and 2008) and the Dominic Ongwen trial before the ICC (key ruling 2021). In the Appeals Judgement the SCSL described the relations between AFRC soldiers and Sierra Leonean abducted women in the following terms: throughout the conflict in Sierra Leone, women and girls were systematically abducted from their homes and communities by troops belonging to the AFRC and compelled to serve as conjugal partners to AFRC soldiers. They were often abducted in circumstances of extreme violence, compelled to move along with the fighting forces from place to place, and coerced to perform a variety of conjugal duties including regular sexual intercourse, forced domestic labour such as cleaning and cooking for the 'husband,' endure forced pregnancy, and to care for and bring up children of the 'marriage.' In return, the rebel 'husband' was expected to provide food, clothing and protection to his 'wife,' including protection from rape by other men, acts he did not perform when he used a female for sexual purposes only. 101 The SCSL initially ruled that forced marriage was 'subsumed by the crime of sexual slavery', but this decision was appealed, resulting in a final determination that forced marriage was a distinct offence under international law where a perpetrator 'compels a person by force, threat of force, or coercion to serve as a conjugal partner resulting in severe suffering, or physical, mental or psychological injury to the victim'. 102 The more recent case features Dominic Ongwen, a commander in the Lord's Resistance Army (LRA) in northern Uganda, who was convicted of charges of enslavement, sexual slavery, and forced marriage. 103 As a commander within the LRA, Ongwen was responsible for leading raiding parties which resulted, amongst other abuses, in the violent abduction of women and girls who were forced to become the conjugal partners, or 'bush wives', of male members of the LRA. These patterns of violence in Uganda share a number of features in common with wartime abuses in Sierra Leone and other conflicts. Violent abductions for the purpose of forced marriages have also recently been documented in Liberia, Rwanda, Nigeria, and the Democratic Republic of Congo.
There are now many first person testimonies of experiences of abduction. 104 The best known example is Evelyn Amony, who was abducted at 11 and forced to become the 'wife' of Joseph Kony, the LRA commander, at 14. Amony had three children with Kony before being recaptured by the Ugandan military after spending 11 years with the LRA, and later served as a liaison with Kony as part of peace talks at Juba in 2006. Amony's place within society is still chiefly defined by her position as 'Kony's wife', despite having been violently 'married' under threat of death and further abuse. 105 Abducted women and girls remain tied to their 'husbands' for decades, and their children are also stigmatized. 106 'Marriages' are usually regarded as marriages even when there was clearly no way to meaningfully consent.
International criminal justice can be best understood as an exception within a political environment which is otherwise defined by widespread impunity. The SCSL successfully convicted nine defendants. The ICC has only prosecuted Ongwen in relation to the Uganda conflict. This is a drop in the ocean given the scale of these conflicts, and the vast majority of combatants in Sierra Leone, Uganda and other recent African conflicts have received amnesties, resulting in a situation where survivors of extreme violence live in societies where their abusers are free. At least some forms of wartime behaviour end up building upon, rather than breaking with, established customs and gender roles. As Seymour, Apio, and Rossi explore in their contribution to this special issue, the history of marriage and enslavement in Uganda features a number of historical examples of violent abductions and enslavements for the purposes of marriage. 107 This is not a case where the past entirely defines the present, but it is a case where the past informs the present in significant ways.
Contemporary legacies are especially prominent when it comes to inheritance of slave status within post-slavery societies, which brings into focus issues which were at least partially concealed under decades of official denials and deflections. As discussed earlier, one of the main defensive responses to the legal abolition of slavery in Africa was to try and reclassify slaves as wives. This strategy was never entirely successful, but it was successful at least some of the time, paving the way for a series of official declarations that slavery was no longer an ongoing issue.
Take, for example, the following case from northern Cameroon. As Ahmadou Sehou has demonstrated in the case Adamaoua, local rulers and powerful men were repeatedly challenged by the members of the Norwegian mission of Ngaundere and by local women resisting enslavement during the 1940s and 1950s. These contests were subsequently concealed under a web of official denials. 108 After decolonization a 1966 United Nations report on slavery included a lengthy discussion of northern Cameroon which described the 1961 flight of 'seventy wives' from the harem of the Lamido (Islamic ruler) of Ngaundere. These wives were eventually forced to return to the palace by the police, and to swear on the Qur'an that they would not run away again. They were said to number amongst 'thousands left toiling as slaves or cooped up in harem walls'. 109 In response to this submission the newly independent government of Cameroon responded as follows: The author of the note undoubtedly confuses slavery with polygamy; polygamy is not peculiar to northern Cameroon, but is practiced in many other African countries. It is in fact the custom of the Lamibe of North Cameroon to have several wives and to also have a number of men in their service (their servants). The informant considers this to be slavery. 110 Some of the women who tried to escape were both wives/concubines and slaves, yet this did not prevent marriage from being invoked to deflect discussion of slavery. By manipulating the definition of slavery in this way the government of Cameroon was following a script developed during the colonial era. This script continued into the post-colonial period, resulting in a situation where the persistence of slavery was denied internationally, yet was well known locally. As part of fieldwork carried out from 2007 to 2009, Adam Mahamat interviewed a number of people of slave descent, including several who had spent their lives in palaces as enslaved functionaries ('royal slaves') and concubines. These interviews revealed that slave status remains as a significant marker of social identity, to the point that some slaves continue to make ransom payments to be manumitted by their master. 111 Northern Cameroon is one of many examples of a post-slavery society. 112 Within Africa, this category is commonly applied to societies where slavery has been legally abolished, yet the social identities, governing ideologies, religious practices, and/or interpersonal connections which it established continue to have pronounced effects upon how both societies and families operate. Slave descent can take a number of forms. People known/believed to descend from unions between two slaves have created new lineages, which are sometimes derogatively known as 'slave lineages'. This means that there are specific surnames which denote slave ancestry, and people who carry these surnames can sometimes find it difficult to secure marriage partners from amongst 'freeborn' lineages, owing to the social stigma, or 'macule servile', which slave descendants endure. 113 Men seek to avoid marrying women of slave descent for fear that descendants of former masters might one day reappear and claim privileges over their wives' sexuality or offspring. 114 There are also examples of mixed descent, a concept which Lotte Pelckmans uses to examine forms of social mobility available to the descendants of royal concubines in the Hayre region of Mali. This does not denote a single status, but can refer to a number of different strategies and outcomes. 115 Yet, in general, slave ancestry functions as a family secret: rarely talked about in public yet still consequential in private. Samuel Lempereur's contribution to this special issue takes up this theme in relation to southern Benin. He notes that in Fongbe language, people who descended from the offspring of enslaved persons can be referred to as akoflangbe. This means the lineage that stammers, or the lineage that does not speak the language well. 116 This formula suggests that they are outsiders who have failed to fully assimilate, yet does not explicitly mention slavery.
Lempereur focuses upon land conflicts between lineages of former slaveholders and lineages originating from an enslaved apical ancestor, who formed separate families. He shows that in post-slavery contexts intermarriage between persons of slave and free descent turned the former into in-laws. This makes it difficult for the latter to assert their rights over disputed lands when doing so would harm their own children from mixed unions. This is captured by an elite elder who chose to settle a land dispute amicably without recourse to courts, ultimately conceding some of the requests of lineages of slave descent on the following grounds: The people we're talking about, there's already been marriages between the children, we can no longer bring them to the king like that. There are still slaves who sell but we do not dare open our mouths to say that. Your child has married his child and you are going to stand up and say that he is your slave, but what is your child going to eat then? […] You cannot say it's your slave when your daughter [who married him] asks for land for her child. 117 Three related points can be highlighted here: marriage remains at the core of the politics and practices of slave-descent; slave ancestry undoubtedly remains socially important, but its effects are not always straightforward; and there is a widespread desire to minimize conversations about slavery due to the stigma surrounding slavery. The language of a 'family secret' is instructive. Insiders within families and communities where slave descent remains an ongoing issue are well acquainted with the frequently subtle webs of status and obligation which define interpersonal relationships, but they can often be reluctant to talk about these issues to outsiders.
That being said, taboos surrounding slavery have been increasingly challenged, as slave descendants in a number of countries have successfully organized politically. This process began in the late 1970s, gathered steam in the 1990s, and has more recently been strengthened by the consolidation of networks linking together an expanding number of local organizations and international partners. 118 Until recently, the narratives of concubines who derived a sense of pride from their role in the families of their masters/husbands emphasized the potential for social mobility that existed in forms of female slavery where enslaved women could become valued as wives and mothers. 119 But women of slave descent and former victims of enslavement are increasingly becoming activists and challenging the structures of power that have been giving enduring legitimacy to the violence and coercion that afflicted them, their mothers before them, and that today threaten their daughters. Reports by Timidria, a Nigerien anti-slavery NGO, reveal the extent of a form of slavery known in Niger as the phenomenon of the wahaya, or fifth-wife. 120 The most high profile example of this dynamic is the case of Hadijatou Mani, which saw the Community Court of Justice of the Economic Community of West African States (ECOWAS) rule in 2008 that the Nigerien government had failed to protect one of its citizens from slavery. This case is usually understood as a slavery case, but it is actually a case study of intersections between slavery and marriage. Born into slavery in 1984, Hadijatou Mani was sold at the age of 12 for 240,000 West African CFA Francs. When Souleymane Naroua, her new master, was given the suggestion that he pay bridewealth for Hadijatou to try and pass the union as a legitimate marriage and that he beg her to accept ('on va la supplier pour qu'elle accepte la dote … '), Souleymane's reply exposed the vitality of pro-slavery ideologies: 'How so? I bought a (slave) woman and I am expected to pay bride-price?!' ('Comment, j'ai acheté une femme et on va me réclamer la dote?'). 121 Souleymane saw the form of conjugal slavery that tied him to Hadijatou as sacrosanct. 122 Yet Hadijatou and other women who faced similar circumstances contradict him. They compare their experiences to the purchase of a goat at the market. Their testimonies tell of beatings, incessant labour, unwanted intimacy, suffering, and dehumanization endured during their time in slavery. 123

Concluding Remarks
There are few institutions more important to human history than slavery and marriage. It should not be surprising, therefore, that these two can be found together so frequently. As this introductory article has shown, their relationship has taken many different forms. Within Africa, the common denominator which connects slavery to marriage is kinship. Throughout African history, both slavery and marriage have played foundational roles in the forcible integration of social outsiders into established hierarchies and household structures as dependants and, eventually, as kin. Enslaved women and girls were particularly valued for their labour, sexuality and reproductive capacity, and it was within this context that some of them became conjugal partners. For some women and girls, marriage added a further dimension to their experiences of enslavement. In other cases, marriage provided a pathway to manumission and qualified forms of social improvement. Both of these variants greatly expanded the range of conjugal relations and marriage strategies available within African societies.
It should also be clear, moreover, that the union between slavery and marriage within Africa has never been fixed or straightforward. Legal and religious texts rarely capture the complexities of lived experiences, and conjugal slaveslike all slavesinvariably had their own strategies and goals to improve their individual fortunes and resist their dehumanization. While the emergence of organized anti-slavery introduced a new element into this equation, the promise of abolition was repeatedly undercut by official complicity, patriarchal bargains, and cultural imperialism. Windows of opportunity were created, but they tended to be narrow. Female slaves who escaped their marriages were usually obliged to seek out new husbands, and they frequently had limited choices regarding potential partners owing to larger social hierarchies. Both colonial and post-colonial authorities strategically manipulated the definition of slavery in order to minimize the scope of their obligations, and thereby helped to partially obscure a range of underling issues. The relationship between slavery and marriage has been transformed by these and other related developments, yet the connection between them remains a live political issue within a number of African societies.