A Brief History of the Anthropology of Sub-Saharan African Homosexualities
by Stephen O. Murray
Published on May 23, 2018
A Tangents Online exclusive
There are many claims, from Edward Gibbon in The Decline and Fall of the Roman Empire (1776–88) and Richard Burton in the terminal essay to A Thousand and One Nights (1994) to contemporary African politicians, denying that homosexuality exists in sub-Saharan Africa, although attestations of all three of the basic organizations (by age differences, by gender differences, and between status equals) were scattered through published literature by travelers and proto-anthropologists. But until nearly a decade into the 21st century C.E. (Gaudio 2009), there was no book-length ethnography focused on those engaged in same-sex sex or on transgendered persons in sub-Saharan Africa.
In leading the exploration and colonization of Africa, the Portuguese became the first Europeans to realize that African sexuality and gender diverged in surprising and—to them, shocking—ways from their own normative ones. In the early seventeenth century, their efforts to conquer the Ndongo kingdom of the Mbundu were stymied by the inspired leadership of a woman warrior named Nzinga. She successfully outmaneuvered the Portuguese for nearly four decades (Sweetman 1984: 39–47). Other early reports from Angola make it clear that Nzinga’s conduct (and role) was not some personal idiosyncrasy but was based on beliefs that recognized gender as situational and symbolic as much as a personal, innate characteristic of the individual.
The reports from Angola set the tone for what followed. When natives like E. E. Evans-Pritchard’s Zande informants told Europeans that men had sex with boys “just because they like them,” Europeans were shocked, incredulous, and confused. They recorded but did not understand sexual and gender practices that epitomized for them how black Africans were different from (and inferior to) them. Although such reporting is short of detail, European accounts document widespread same-sex patterns and roles throughout the continent.
Boy Wives and Female Husbands (Murray and Roscoe 1998, henceforth M&R) examined evidence of same-sex patterns in some fifty African societies. All these societies had words—many words, with many meanings—for these practices. Furthermore, these societies are found within every region of the continent, and they represent every language family, social and kinship organization, and subsistence pattern. There is substantial evidence that same-sex practices and patterns were “traditional” and “indigenous.” Since anthropologists and other observers have rarely inquired systematically into the presence of homosexuality in Africa (or elsewhere), absence of evidence can never be assumed to be evidence of absence.
Before the twentieth century, information about “primitive” (in technology) cultures was extracted from reports by travelers, merchants, and missionaries, only the last of whom spent extended periods of time in close contact with the peoples discussed. In a 1937 Frazer Lecture at Oxford, Henry Balfour analogized spinning and weaving: spinners were the providers of information, weavers were the theorists who took the spun yarn and wove it into theories. A famous “weaver” (ethnologist), Sir James Frazer replied to a query from philosopher William James about personal experience with “natives”: “Heaven forbid!”
In the United States, Frank Hamilton Cushing had pioneered long-term observant participation with the Zuni, though English anthropologists credit Polish émigré Bronislaw Malinowski with “inventing” ethnography—long-term if not very participant observation—in the Trobriand Islands where he was detained during the First World War. Malinowski theorized generalizing from his Trobriand materials. Many of those who would do more sustained ethnographic fieldwork in Africa participated in his seminars at the London School of Economics (LSE) during the late-1930s.
At the LSE along with Malinowski, Charles G. Seligman, who had done survey work in the Sudan (reported in Seligman 1930, Seligman and Seligman 1932), taught, supervising two of the leading Africanists of the next generation, Edward Evan Evans-Pritchard and Meyer Fortes. (The two collaborated on the 1940 book, African Political Systems.)
There were passing references to homosexual roles, as well as gender “deviant” ones, in accounts of African cultures and societies by Francophone and Anglophone anthropologists of the colonial era 1. Although German colonies were taken away and reassigned to other European nations after World War I, there were more substantial accounts of homosexual roles and relations in the German anthropological literature than in those of any other.
In particular, Kurt Falk, who spent a dozen years in Africa, surveyed male-male and female-female sexual conduct within “tribes” of Angola and German Southwest Africa (what is now Namibia). He reported monogamous relationships between young male-male and female-female Mbundu couples (aponji) along with opportunistic homosexual relations between men and boys and men and men when away from their wives. He remarked: “While the act is permitted, speaking about it is considered disgusting.” Usually, both aponji (plural of eponji) remain together a long time and jealously guard that neither commits “infidelities.” If one asks a young boy which intercourse he prefers, that with a boyfriend or that with a girlfriend, the usual answer is: “Either is equally good and beautiful!” (Falk 1920).
Falk also recorded how a Nginé (Ganguela) soldier, when being punished by white sergeant for sexual assault, complained:
Doesn’t the Sergeant know that are there men who from youth on desire women, and others, who are attracted only to men? Why then should he be punished now? After all, he knows not why, God created him like this—that he can only love men!
In Namibia, Falk (1926) documented an Ovahimba, a “true invert, who was active as a wizard and medicine-man,” and Ovambo “ovashengi (singular: eshengi), effeminate men who submit to passive coitus in anum for income. They are usually fat, large persons who go about in female clothing and are employed mainly in women’s work. They bedeck themselves with female jewelry and try to arouse the attention of men through flirting.” Moreover, boys accompanying men to mine work “serve the others as passive pederasts on the journey and during the term of service at the [mine]fields.” Indeed—and this is the remarkable thing about it—these boys are given to the young men usually by their wives or betrothed “to keep the men faithful.” 2 Across Namibia, Falk estimated that about 90% are “able to consort with both sexes, while estimating 3.5% are “born homosexuals,” which leaves 6.5 exclusive heterosexual. 3
Falk reported Herero (Namibian), San (Namibian), and Wawihé (Angolan) women, especially young recently married women, rubbing their genitals together, engaging in manual stimulation of each other, and using dildos.
Falk (1920) also reported the paying by young (18-ish) Nginé men of brideprice to the fathers of boys of around 12:
If the older one later marries this changes nothing in his relation to his katumua; he merely sleeps alternately with his wife and his boyfriend, until finally the boyfriend, growing up, wants to marry. Then the older one brings the younger one back to his father and pays the agreed price. The younger now, for his part, takes a boyfriend, while the older looks for a new katumua. So it comes to be that almost every man without exception, whether single or married, has his lover.
In a number of Sudanese tribes with “widespread homosexuality,” anthropologist Siegfried Nadel (1947: 300) reported a fear of heterosexual intercourse as sapping virility accompanying a common reluctance to abandon the pleasures of all-male camp life for the fetters of permanent mixed-sex settlement. He noted that he knew “men of forty and fifty who spent most of their nights with the young folk in the cattle camps instead of at home in the village.”
There were already published references to Azande homosexuality at the time Evans-Pritchard undertook his fieldwork in 1926. Larken (1926: 24) had referred to the practice of chiefs having sex with male youths. In traditional Azande culture, before the pacification which made it safe for ethnographers like Nadel and E. E. Evans-Pritchard to work in the Sudan in the 1920s and later,
men used to have sexual relations with boys as they did with wives. A man paid compensation to another if he had relations with his boy. People asked for the hand of a boy with a spear, just as they asked for the hand of a maiden of her parents. All those young warriors who were at court, all had boys. (Kuagbiaru, quoted by Evans-Pritchard 1970:1430)
Homosexuality is indigenous. Azande do not regard it as at all improper, indeed [they regard it] as very sensible for a man to sleep with boys when women are not available or are taboo…. In the past this was a regular practice at court. Some princes may even have preferred boys to women, when both were available…. I was told that some princes sleep with boys before consulting poison oracles, women being then taboo, and also that they sometimes do so on other occasions, just because they like them. (Evans-Pritchard 1971: 183)
Kuagbiaru told Evans-Pritchard that when a prince considered a boy (between the ages of twelve and twenty) appealing, the prince would summon the boy as a page. Later, the prince “provided bridewealth for his pages when they grew up” (p. 185). In addition to princes having pages at their disposal,“many of the young warriors married boys, and a commander might have more than one boy-wife” (p. 199).
Another commander, Ganga, told Evans-Pritchard (1970:1431) that “there were some men who although they had female wives, still married boys. When a war broke out, they took their boys with them,” although they were left in camp, as befitted their wifely status, not their future status as warriors.
The warrior paid bride wealth to the parents of the boys and performed services for them as he would have done had he married their daughter…. Also, if another man had relations with his boy, he could, I was told, sue him at court for adultery. (p. 1429)
Evans-Pritchard believed that “it was on account of the difficulties of getting satisfaction in heterosexual relationships that boy marriage was recognized,” because with easier (and earlier) marriage between men and women. “Boy-marriage has in post-European [colonial] times entirely disappeared” along with the cultural contexts. That is, the military companies and the royal court also disappeared. Even after the custom and much else of the culture had been disrupted, “I have never heard anyone speak of sleeping with a boy with distaste,” he noted (p. 1429).
French colonial administrator Adolphe Cureau (1904:64–5) attributed warrior pederasty among the Sandeh (Azande) to the monopolization of women in the vast harems (bodimoh) of Sandeh royalty. Vassals, soldiers, and servants had to make do with what the rulers left. Boy servants d’armes took the place of women. Wearing their hair artfully parted, with arms and necks loaded with decorations, a woolen skirt around the hips, and their bodies oiled and glistening, the boys were at the disposals of soldiers. These ndongo-techi-la followed the soldiers on their marches, carrying their husbands’ rifle, hammock, and a little bag with pipe, fire stick, and some millet. In the camps they cooked and managed household finance (i.e., away from home, undertook “women’s work”). Among the Mossi, soronés (pages), chosen from among the most beautiful boys aged seven to fifteen, were dressed and had the other attributes (including le rôle) of women in relation to chiefs, for whom sexual intercourse with women was denied on Fridays. There is nothing in Tauxier’s (1912: 569–70) report as indicating sexual unavailability of the boys during the rest of the week.
While serving as a soroné, there were annual tests (of some unspecified sort) to make sure that the boy had not been sexually intimate with any woman. After the boy reached maturity he was given a wife by the chief. The first child born to such couples belonged to the chief. A boy would be taken into service as a soroné, as his father had; a girl would be given in marriage by the chief, as her mother had.
According to the 1907 Taberer Report:
It appears to have become a well-recognized custom among the mine natives recruited from the East Coast to select from the youths and younger men what are termed amankotshane or izinkotshane. An inkotshane may be described as a fag and is utilised for satisfying the passions. Any objections on the part of the youth to becoming an inkotshane are apparently without very much difficulty overcome by lavishing money and presents upon him…. An inkotshane’s duty appears to be to fetch water, cook food and do any odd work or run messages for his master and at night time to be available as bedfellow. In return for these services the inkotshane is well fed and paid; presents and luxuries are lavished upon him. (p. 2, quoted in Moodie et al. 1988: 234)
The institution of boy wives of miners in South Africa has been thoroughly analyzed by Moodie et al. (1988) and Achmat (1993), the latter also considering homosexual liaisons in South African prisons.
Although relationships between women of equivalent (age, gender, and class) statuses certainly existed in “traditional” black Africa, the limited data suggest that the most common culturally marked form of female homosexuality there was a role for a woman who becomes a sort of husband to another woman and claims some of the social prerogatives of a man. Gender is the major idiom of homosexuality in Africa for both men and women, though the husband is often older than the wife in same-sex as in mixed-sex relationships. In Lesotho mummy-baby relationships, the partners are differentiated by age, though Gay (1985: 102) reported that “the most frequently given reason for initiating a particular relationship was that one girl felt attracted to the other by her looks, her clothes, or her actions.” In the eleven instances in which Gay specified the age of both, the mean difference was 4.8 years and the median difference was 5. (The range was from 1 to 12.) Gay attempted to deny the erotic component in her data; later work by Kendall (idn MR) and Siza Khumalo (in Morgan and Wieringa 2005) made a sexual component clear.
Although the coding of the female husband as a man (jural) is widespread, not all woman-woman relationships in Africa are so conceptualized. Huber (1968: 746) reported that among the Simiti (a Bantu-speaking group on the eastern shore of Lake Victoria), the natally female pair are termed “mother-in-law” and “daughter-in-law.” There are also self-identified “lesbians” and butch-femme couples (the butch one often called some variant of “tomboy” and considering themselves to be men whose vaginas should not be penetrated).
Although female husbands often take on the parental, kinship, and work roles of men, financial responsibility, and decision-making, they do not generally cross-dress (in contrast to tomboys who do).
Woman-woman marriages—in which one woman pays brideprice to acquire a husband’s rights to another woman—have been documented in more than thirty African populations (O’Brien 1977: 109). There has been lively academic controversy about local conceptions of the gender of female husbands, while none of the observers (other than the research supervised by Morgan and Wieringa 2005) seems to have asked about sex between such husbands and their wives. (Most ethnographic accounts of female husbands do not even report where or with whom the partners usually slept.)
O’Brien argued that “a female husband does not engage in sexual interaction with her wife.” Her rationale, however, is that “if homosexual behavior were a regular component of female-husband marriages, the association would have been noted in the ethnographic record” (1977: 109, 123, n. 1). Such an assumption attributes more acuity (and data collection) to anthropologists on sexual matters than is warranted, overlooks the fact that sexual behavior is generally invisible to outsiders (especially hostile ones), and ignores the history of avoidance of consideration of “the sin not named” among Christians in anthropological discourse (see Murray 1997). Moreover, British social anthropologists (notably the contributors to Fortes and Evans-Pritchard 1940) were preoccupied with formal jural relations; little interested in messy, actual behavior; and totally uninterested in psychological motivations.
Some have questioned ethnographic reports that indicate (or simply allow the possibility of) a sexual dimension to woman-woman marriages, e.g., Christine Obbo (1976: 372) exhibiting horror that
Melville Herskovits seems to imply that mutual masturbation is necessarily homosexuality (1938: 289) and that there was no doubt that prosperous female “husbands” utilize the relationship in which they stand to the women they marry to satisfy themselves (1937b: 338). His passing remarks cannot be taken as conclusive evidence, and therefore we cannot know whether the Dahomean women who practiced woman-to-woman marriage were lesbians. She did not explain what might be “heterosexual” in female-female mutual masturbation or consider local language terms.
Herskovits (1937b: 340) carefully outlined aspects of the institution of woman-woman marriage, explicitly noting it does not necessarily imply a homosexual relationship between the “husband” and the “wife.” Herskovits (1938: 277–90) presented a detailed discussion of Dahomean (Fon) sex education and sex experience, reporting that between the time of puberty and marriage, when the young girls have been withdrawn from the boys of their age, “Homosexuality is found among women as well as men; by some it is claimed that it exists among women to a greater extent” (289).
He also noted that prior to the onset of puberty, because of the value attached to thickened vaginal lips in that society, “girls between the ages of 9 and 11 are assembled by compounds in groups of eight or more…and these engage in the practice known as axoti, which consists of massaging and enlarging the lips of the vagina” (282). This practice is widespread, and no anthropologist seems to have inquired about any (sexual) pleasure being a part of the procedures.
Given the broader context of Dahomean sexual behavior, no great leap of the imagination is required to suggest, as did Herskovits, that some of the females involved in woman-woman marriage in Dahomey might also use the relationship as a means of obtaining sexual satisfaction.
Rudi Gaudio (1996:136n17) noted that there are Hausa lesbians (in Kano of the 1990s), some of whom know of and use the terms for male homosexual roles, and that gay Hausa males assert that “lesbians engage in the same types of relationships as they themselves do, i.e., involving the exchange of material gifts and the attribution of [dominant/submissive] roles.”
Gill Shepherd (1978:133) reported, “In Mombasa, both male and female homosexuality is relatively common among Muslims; involving perhaps one in twenty-five adults.” Shepherd (1987:240—with no data nor discussion of the basis for either the earlier estimate or its revision based on the same fieldwork) raised the estimated rate to one in ten. She wrote that “lesbians [in Mombasa] are known as wasaga (grinders) and that “the dominant partner. . . is not seen as a man.” She had earlier claimed, on the basis of an unspecified sample of wasaga,” that “there is almost always a dominant and subordinate economic relationship between them.” Shepherd (1987:254) elaborated: “The verb kusaga (to grind) is commonly used for the grinding of grain between two millstones…. The upper and lower millstones are known as mwana and mama respectively: child and mother.” Perhaps Swahili is unlike English in spatializing hierarchy in terms of top and bottom, since the bottom millstone is predominant, but there is something of the mama and her dependent mwana to lesbian relationships in Mombasa, paralleling basha/shoga for male homosexual pairs of differing ranks. Clear status distinctions characterize msagaji relationships, with the dominant woman usually being older as well as wealthier. Whether these distinctions affect sexual behavior inside the relationship, Shepherd did not report. Msagaji seem to be bidding for some male privilege beyond that of having sex with submissive women. That is, they do not (entirely) conform to Mombasa conceptions of how women should look and act. The prototype misago differs from “normal” women in both gender patterning and sexual partnering.
Gustave Hulstaert (1938:95–6) wrote that (in what was then the Belgian Congo):
Nkundó girls play at ‘husband and wife’ and even adult married women engage in this vice. According to my informants, the causes are as follows: first, an intense and very intimate love between two women, second and above all, the fact that wives of polygamists find it difficult to satisfy their passions in a natural way. Often they engage in this practice with co-wives of the same man.
He further noted that “in establishments where girls are too securely kept away from the opposite sex, there has been an increase in sexual relationships between girls. The latter often engage in sex with co-wives.”
The Gender of the Female Husband
In her survey of the status of women, Niara Sudarkasa argues that there is a general de-emphasis on gender in “traditional” African societies and a corresponding emphasis on status (“personal standing”) which usually, but not always, is determined by wealth (Sudarkasa 1986: 97). Robertson also argues that age and lineage override gender in traditional African societies (1987: 111), while Matory distinguishes “gender” from “sex” and stresses (in reference to the Yoruba) that “far stronger than the ideology of male superiority to the female is the ideology of senior’s superiority to junior” (1994: 108). In the title of her 1980 article on woman-woman marriage among the Nandi, Oboler poses the question, “Is the female husband a man?” Evans-Pritchard (some years earlier) had already answered the question in the affirmative:
A woman who marries in this way [i.e., takes a wife] is generally barren, and for this reason counts in some respects as a man. She acquires cattle through the marriage of kinswomen, including some of those due to uncles on the marriage of a niece, or by inheritance, since she counts as a man in these matters. A barren woman also often practices as a magician or diviner and thereby acquires further cattle; and if she is rich she may marry several wives. She is their legal husband and can demand damages if they have relations with men without her consent. She is also the pater of their children and on marriages of their daughters she receives “the cattle of the father,” and her brothers and sisters receive the other cattle which go to the father’s side in the distribution of bridewealth. Her children are called after her, as though she were a man, and I was told they address her as “father.” She administers her home and herd as a man would do, being treated by her wives and children with the deference they would show a male husband and father. (1951: 108–9)
Rivière added: “It is clear that one finds in numerous societies women acting out male roles, including that of taking a wife” (1971: 68). Thus, an Efik-Ibibio (Nigerian) woman who grew up during the nineteenth century (and whose husband had eleven other wives) recalled:
I had a woman friend to whom I revealed my secrets. She was very fond of keeping secrets to herself. We acted as husband and wife. We always moved hand in glove and my husband and hers knew about our relationship. The village nicknamed us twin sister. (in Andreski 1970: 131)
In other words, African marriages are between individuals in male and female roles, not necessarily between biological males and females.
The very possibility of a formal status for female husbands reflects the divergence between gender and sex in African societies. The exact nature of this status, however, is still the subject of debate. Krige, for example, argues that Lovedu female husbands are neuter, neither masculine nor feminine within the terms of a binary gender system (1974: 32–33). At the same time, she offers no evidence that this status amounted to an intermediate or distinct third gender role, involving special dress, social roles, or religious functions. The view of the societies she considers (southern Bantu groups) appears to be that the female husband becomes a “social male” within a system of binary genders—which are, however, not homologous with biological sex.
In Nigeria, John McCall interviewed an elderly Ohagia Igbo dike-nwami (brave-woman) named Nne Uko. Early on, she told McCall, she “was interested in manly activities” and felt that she “was meant to be a man” and so “went as my nature was given to me—to behave as a man” (McCall 1996: 129). She was initiated as a woman, but after being married for a time and producing no children, she was divorced. She subsequently farmed and hunted while dressed as a man, was initiated into various men’s societies (including the most exclusive one), and took two wives of her own. She did not, however, take her wives to bed (130). Her wives’ children (who were sired by her brother) referred to her as “grandmother.” She was no longer active in men’s societies at the time she was interviewed, having “risen to a position as a priestess of her matrilineage’s ududu shrine, offering sacrifices to the ancestresses in her maternal line” (131). Upon her death, she will join those ancestresses to whom sacrifices are made in the ududu shrine by those seeking good crops and children. In other words, despite her departures as an adike-nwami from the conventional role for women, she remained a woman in the view of the Ohagia Igbo. As McCall concludes, “Throughout her life Nne Uko was recognized in her community as a woman, socially and otherwise” (131).
Oboler argued that among the Nandi of Kenya, “The key to the question of the female husband’s gender lies in her relationship to the property that is transmitted through her to the sons of her wife.” The Nandi believe that the most significant property and primary means of production should be held and managed exclusively by men. Thus, a female husband is categorized as a man — “promoted” to male status (kagotogosta komostab murenik, literally “she has gone up to the side of the men”) (Oboler 1985: 131). According to Oboler’s informants, “A woman who takes a wife becomes a man and (except for the absence of sexual intercourse with her wife) behaves in all social contexts exactly as would any ordinary man” (1980: 69, 70).
The sexuality of the female husband does not appear to have been a factor in defining her role. Female husbands were expected to discontinue sex with men and, indeed, abstain from sexual intercourse with either gender because if “she should conceive, both the issue of inheritance and the dogma that she is a man would be too thoroughly confounded to be withstood” (85). Female husbands’ most strenuous attempts to conform to male behavior occur “in contexts that are closely connected with the management of the heirship to the family estate” (86).
Oboler characterizes the Nandi female husband as “a woman of advanced age who has failed to bear a son” (69). Since the purpose of the union is to provide a male heir, the wife of the female husband has a male consort whose sole function in the relationship is to serve as the progenitor. Oboler suggests several motivations for females marrying female husbands: (1) the somewhat higher bridewealth paid by female husbands, (2) greater and more casual companionship, (3) somewhat greater participation in household decisions, (4) sexual autonomy, (5) less quarreling with and physical violence than with male husbands, (6) and a dislike of men (76, 78).
In her 1980 article, Oboler concluded that the Nandi wives are under the control of their female husbands and are not promiscuous. In her 1985 book (based on the same 1976–77 fieldwork), however, she wrote that the wife of a female husband “is free to engage in sexual liaisons with men of her own choosing” and suggests that this represented a change from earlier practice: “Formerly, it is said, it was the right of the female husband to arrange a regular consort for her wife…. Today, sexual freedom is cited as one of the advantages of marriage to a female husband” (132; see also 1980: 76). On the other hand, some women were reluctant to become female husbands, since they were expected to forgo sexual intercourse. However, Oboler does not report any data about the actual sexual conduct of Nandi females, and she acknowledges that at least some Nandi claims regarding their sexual restraint reflect ideal rather than actual behavior. In any case, the institution is still flourishing: 10 of the 286 households in the community she studied in the 1970s had female husbands. However, Oboler anticipated that Christian disapproval and the spread of bilateral inheritance will eventually undermine the institution.
Although the coding of the female husband as a man is widespread, not all woman-woman relationships in Africa are conceptualized as those of a husband and wife. Huber has reported that among the Simiti (a Bantu-speaking group on the eastern shore of Lake Victoria), “Neither in the view of the people, nor in the terminology, nor in the wedding ritual itself is there any suggestion that a woman assumes the role of a husband in relation to another woman. Rather, the two main persons concerned are related as mother-in-law and daughter-in-law.” Huber translates the usual expression for the relationship, okoteta mokamona wa nyumba ntobu, as “to give cattle for a daughter-in-law on behalf of a poor house”—a “poor house” being one that has produced no male offspring (or none that reached child-producing age) (Huber 1968: 746).
Marriage without Sex?
Anthropologists and Africans alike have been almost unanimous in denying the possibility that woman-woman marriages includes sex or even emotional attachment. (See, for example, the comments of Ifi Amadiume [1987: 7] quoted by Kendall earlier in this volume.) Krige (1974), Obbo (1976), Sudarkasa (1986: 97), and others have all warned against assuming that women who married each other, even when they slept together, had sex. Few of these denials, however, are based on actual inquiries with or observations of the individuals involved—and certainly not observations of sexual behavior. (Most ethnographic accounts do not even report where or with whom the partners usually slept.)
O’Brien, for example, argued that “a female husband does not engage in sexual interaction with her wife; indeed, nowhere do the African data suggest any homosexual connotation in such marriages.” Her rationale, however, is that “if homosexual behavior were a regular component of female-husband marriages, the association would have been noted in the ethnographic record” (1977: 109, 123, n. 1). Such an assumption attributes more acuity to anthropologists on sexual matters than is warranted, overlooks the fact that sexual behavior is generally invisible to outsiders (especially hostile ones), and ignores the history of enforced silence on homosexuality in Western discourse.
Even ethnographers not hostile to the phenomenon may overlook it. Robert Brain, for example, spent two years doing fieldwork among the Bangwa of Cameroon (1976: 31). His “best woman friend in Bangwa was Mafwa, the chief’s titled sister,” with whom he often traveled to distant rites and with whom he spent many happy hours drinking and talking (55). Nonetheless, “it was not until some months after I first knew Mafwa that I found out that my rather androgynous princess had her own wife” (56). In fact, she had inherited two wives when her father died. Aside from accumulating bridewealth when her (wives’) daughters married, Brain concluded that Mafwa had become a female husband “to have her own compound, with a wife to cook for her and her own children around, without the overbearing or annoying presence of a husband” (57). Once he became aware of the relationship between Mafwa and her co-resident wife (whose daughter called Mafwa “father”), Brain observed “their obvious satisfaction in each other’s company.” He took this as “evidence of a perfect alliance between two mature women who felt no need of the presence of a male husband” (58).
Brain correctly notes that cross-culturally the “rights of sexual access and domestic companionship are only some of the aspects involved in marriage. . . . ‘Husband and wife’ are joined in marriage for many purposes—not merely to enjoy sexual intercourse, companionship, or even have children” (58). The latter two are clearly part of the Bangwe “bundle of marital rights” in marriages between men and women as well as woman-woman marriages.
Some have questioned ethnographic reports that indicate (or simply allow the possibility of) a sexual dimension to woman-woman marriages. Feminist anthropologists Eileen Krige and Christine Obbo, for example, criticized Melville Herskovits, whose 1937 and 1938 reports on “woman marriage” in Dahomey suggested that homosexuality might be involved. Krige complained:
Herskovits (1938) imputed to it sexual overtones that are foreign to the institution when, after stating quite definitely that such marriage did not imply a homosexual relationship, he went on to add ‘although it is not to be doubted that occasionally homosexual women who have inherited wealth…utilize this relationship to the women they marry to satisfy themselves’ (1937b: 338). He made no attempt to substantiate his statement. (Krige 1974: 11)
Obbo echoes Krige’s criticisms:
She [Krige] questions Herskovits’ unexplained, unelaborated assumption that such a marriage was a homosexual relationship. Krige asserts that homosexuality is foreign to the marriage as practiced among the Lovedu today…. Herskovits seems to imply that mutual masturbation is necessarily homosexuality (1938: 289) and that there was no doubt that prosperous female “husbands” utilize the relationship in which they stand to the women they marry to satisfy themselves (1937b: 338). His passing remarks cannot be taken as conclusive evidence, and therefore we cannot know whether the Dahomean women who practiced woman-to-woman marriage were lesbians or whether it was a slip of Herskovits’ pen. (Obbo 1976: 372)
In a note to this passage, Obbo added, “While no one can categorically dismiss the possibility that a woman-to-woman marriage may involve a homosexual relationship, there is no excuse for assuming it” (385, n. 4). She does not explain, however, what might be “heterosexual” in female-female mutual masturbation, a practice that does not seem to involve gendered roles.
A careful reading of Herskovits, however, shows that Krige, O’Brien, and Obbo exaggerate his remarks. (Moreover, Obbo misrepresents Krige, as well as Herskovits, in her criticisms.) Herskovits carefully outlined aspects of the institution of woman-woman marriage, explicitly noting it does not necessarily imply a homosexual relationship between the “husband” and the “wife,” and concluded:
The motivating drive behind it—the desire for prestige and economic power—reflects the dominant Dahomean patterns of thought and the fundamental forces that underly Dahomean patterns of thought and behavior in all aspects of life. (1937b: 340)
In his 1938 work on the Dahomeans, Herskovits presented a detailed discussion of the sex education and sex experience of females (and males) (1938: 277–90). He reported that between the time of puberty and marriage, when the young girls have been withdrawn from the boys of their age, “Homosexuality is found among women as well as men; by some it is claimed that it exists among women to a greater extent” (289). He also noted that prior to the onset of puberty, because of the value attached to thickened vaginal lips in that society, “girls between the ages of 9 and 11 are assembled by compounds in groups of eight or more…and these engage in the practice known as axoti, which consists of massaging and enlarging the lips of the vagina” (282). An older woman, “whose most desired qualification is an age not too far removed from that of the girls so that she will not have forgotten her own experiences when undergoing this regimen,” assists in the task, which Herskovits described as follows:
With a shaped piece of wood, this woman manipulates the lips of the vagina of each girl, pulling at them, stretching them, and lightly puncturing the vaginal tissues in several places. This she does eight or nine times for each of her charges during the first year of instruction, and during the next year the girls do this for each other…. For two years at the very least this is continued, and in addition there is the outer massaging of these “lips” to cause thickening and muscular development, for “thin-lipped” women are considered lacking in comeliness. (282)
The end product of this practice is the “enhancement of pleasure in sex play…the roughened surface of the inner vagina lips heightens pleasure during coitus, since the scarifications are not unlike the body cicatrizations” (282–83).
One final bit of information provided by Herskovits on the sex experience of Dahomean girls is of interest. When the sex education of the girls begins, “They no longer go about the compound naked but are given small cloths to wear” (283). They are also warned to avoid hidden play with boys and told not to allow the boys to approach them sexually. Herskovits noted that at this stage, “When the girls find themselves with others who have been given small cloths to wear, and with those older than themselves, they compare sex organs, each boasting of the size of the lips of her own. When none of the older people are about, the girls may indulge in mutual masturbation” (285).
Males may also be involved in long-term homosexual relationships:
One situation which arises in the sex life of the boys during middle adolescence deserves some consideration, for it is at this time of life, when the young girls have been withdrawn from the boys of their age, that any tendencies to homosexuality develop; when, indeed, according to one account, homosexuality, which is ordinarily looked upon by the Dahomeans with distaste, is recognized as normal. Once the games between boys and girls are stopped, the boys no longer have the opportunity for companionship with the girls, and the sex drive finds satisfaction in close friendships between boys of the same group. (289)
Given the broader context of Dahomean sexual behavior, no great leap of the imagination is required to suggest, as did Herskovits, that some of the females involved in woman-woman marriage in Dahomey might also use the relationship as a means of obtaining sexual satisfaction.
Which touches are “sexual” and which are not varies from person to person, and “sex” (as an activity) is not a clearly bounded domain with universally agreed-upon criteria even in one society, let alone cross-culturally. People who sleep together tend to touch each other, and that touches that are experienced by some as erotic are not experienced as erotic by others. Of course, there are economic bases and economic motivations for woman-woman marriages as there are for man-woman marriages in Africa—and elsewhere. Yet no one questions whether men and women in mandatory, arranged marriages have or desire sex with each other or, indeed, even “prefer” the opposite sex in general. The practice of mandatory marriage does not require it. At the same time, alliances that are arranged, as well as alliances contracted for reasons other than love and sexual attraction, may eventually result in erotic attraction and sex. This would seem to be especially likely for those not having sex with other partners. And this is a category that is shaped at least to some degree by individual desires.
Woman-woman relationships in Africa are conceptualized as those of a husband and wife. Huber has reported that among the Simiti (a Bantu-speaking group on the eastern shore of Lake Victoria), “Neither in the view of the people, nor in the terminology, nor in the wedding ritual itself is there any suggestion that a woman assumes the role of a husband in relation to another woman. Rather, the two main persons concerned are related as mother-in-law and daughter-in-law.” Huber translates the usual expression for the relationship, okoteta mokamona wa nyumba ntobu, as “to give cattle for a daughter-in-law on behalf of a poor house”—a “poor house” being one that has produced no male offspring (or none that reached child-producing age) (Huber 1968: 746).
Historian Marc Epprecht (2004) reported testimony in Southern Rhodesia (now Zimbabwe) courts in the 1920s that indicates that English colonial authorities were beginning to employ psychiatric language—and that natives were refuting it. Epprecht quotes one native as saying, in reference to his cross-dressing son: “I have never thought him mentally affected.” Once Anglo-American anthropologists adopted sexological terminology, they were slow to abandon it. When Evans-Pritchard finally published his observations of Azande homosexuality in 1970, he did so under the rubric of sexual inversion—a term from pre-Freudian sexology. The psychiatric vocabulary of “inversion” was not used when anthropologists began observing and questioning those engaged in same-sex sex.
There continues to be little research on same-sex sex or transgendered persons in 21st-century Africa. Sodomy laws remain in many former British colonies, even if the post-apartheid constitution of South Africa included a ban on discrimination based on sexual orientation. In addition to laws left form colonial powers (that removed sodomy laws at home), new waves of Christianist missionaries to Africa and increasingly sex-negative Islamist imans have targeted homosexuality/homosexuals and increased penalties for those convicted of homosexuality. Official claims that homosexuality is un-African or nonexistent are perilous to local and to foreign researchers wanting to study homosexuality (and to those studied). Evidence of homosexual networks emerge fairly often, but systematic research on present-day homosexualities remains exceedingly rare.
The research by Rudolf Gaudio in Kano, the largest city in northern, Muslim Nigeria, is an exception. ’Yan daudu there are biological males who have sex with males (mostly conventionally masculine ones), but mostly also marry and sire children (which culturally makes them “men” as well as male). In private they don an item marked as women’s (either a headscarf or a wrapper around their waist). In private they play with female pronouns, but in public almost always use male pronouns in address and reference for self and for other ‘yan daudu. Both ’yan daudu and their ethnographer were circumspect in talking about sexual conduct—at least the text of Gaudio’s (2009) book is.
Gaudio also reported that some conventionally masculine males doing appropriately masculine work who “do the deed” (have sex with other males) have enough sense of identity as “homos” (a term in Hausa discourse along with masu harka—“men who do the deed”) and socialize together. As in other Muslim societies, introspection and any identity politics other than as Muslims are discouraged, and as long as duty (siring children) is done, a large zone of what is “nobody’s business” exists for Hausa men. Moreover, the attribution of agency for most everything to Allah provides scope for self-defense by ’yan daudu. If everything that is, everything that happens, is dependent on Allah’s will, so must being a ’yan daudu— and even the hypocrisy of those who proclaim their piety, slander ’yan daudu, but also seek sex with them in the large zone of non-public homosociality.
In an East African “Old Town,” Amory (in M&R) found that
an identifiable community of individuals who engage in homosexual behavior and to a certain extent identify as “homosexuals” exists in Old Town. The individuals I interviewed acknowledged that they were mashoga but also referred to themselves (and their friends) as magai, that is, gay and lesbian, using the English word with the Swahili prefix, ma-, to indicate the plural” and there was a conception of being “out: the Swahili term yuko wazi, “(s)he is open/clear,” stands in opposition to anajificha, “(s)he is hiding,” and “the active partner to the shoga is known as the basha (pl. mabasha) or haji (pl. mahaji)…. Because the basha is seen as a “real man,” his identity remains unmarked unless he is paired with a shoga.
Networks of males seeking sex with other males exist in African cities, even among the supposedly exclusively heterosexual Kikuyu. Public self-identification as gay is much rarer, even in South Africa where some legal safeguards exist.
If “sex” is understood as behavior that might result in pregnancy, homosex is not “sex,” and there is evidence from a number of African sites that it is regarded as “play” and not “sex.” What is considered as “sex” by those who deny sex takes place is noticeably unspecified in the sex-denying literature on African same-sex relationships (e.g. O’Brien 1977), nor is there recognition of anything like a lesbian continuum (Rich 1980) in Africa encompassing affect as well as genital stimulation.
As in North America of four and more decades ago, many of those who love members of their own sex marry members of the other and have children. A distinct, oppositional gay/lesbian community is little developed in African countries, though the possibility of exclusively homosexual conduct is not altogether missing in southern African cities. Still, many of those in ongoing same-sex sexual relationships or engaged in recurrent same-sex sex feel that their safety requires that these liaisons be kept secret, which perpetuates the widespread belief that there is no homosexuality in their society. The demand to procreate is paramount, and necessarily covert sexual behavior largely forestalls homosexual identity construction and keeps subcultures invisible.
There has been little research on homosexuality or transgender by black Africans from north of the Republic of South Africa, with the exception of U.S.-trained Cameroonian sociologist Charles Gueboguo, though, with mentorship of Marc Epprecht, Gays and Lesbians of Zimbabwe have published. Ruth Morgan and Saskia Wieringa trained women from Uganda, Kenya, Tanzania and South African to collect life-histories from 3–5 women-loving women, including some female husbands reporting sex with their wives. Plus one Ghanian wrote about other African males who engaged in same-sex sex in Murray and Roscoe’s 1998 collection. The body of anthropological research on homosexuality and transgender in Africa is too sketchy for a critical (re-)evaluation.
Achmat, Zackie. 1993. “‘Apostles of civilised vice’: ‘immoral practices’ and ‘unnatural vice’ in South African prisons and compounds, 1890–1920.” Social Dynamics 19:92–110.
Balfour, Henry. 1938. Spinners and Weavers in Anthropological Research. Oxford: Clarendon Press.
Cureau, Adolphe Louis. 1904. Essai sur la psychologie des races nègres de l’Afrique Tropicale.” Revue générale des sciences pures et appliquées 15:638–695.
Epprecht, Marc. 2004, 2013. Hungochani: The History of Dissident Sexuality in Southern Africa. Montreal: McGill-Queen’s University Press.
Evans-Pritchard, Edward E. 1951. Kinship and Marriage among the Nuer. Oxford University Press.
—. 1970. “Sexual inversion among the Azande.” American Anthropologist 72:1428–1434.
Falk, Kurt. 1920. “Gleichgeschlechtliches Leben bei einigen Negerstämme Angolas.” Archiv für Anthropologie 20:42–45. (tr. In M&R)
—. 1925. “Homosexualität bei den Eingeborenen in Südwest-Afrika.” Geschlecht und Gesselschaft 13: 209–211.
Fortes, Meyer, and E. E. Evans-Pritchard. 1940. African Political Systems. New York: Oxford University Press. (tr. In M&R)
Gaudio, Rudolf P. 2009. Allah Made Us: Sexual Outlaws in an Islamic African City. New York: Wiley-Blackwell.
GALZ. 2008. Unspoken Facts: A History of Homosexualities in Africa. Harare, Zimbabwe : Gays and Lesbians of Zimbabwe.
Gay, Judith. 1985. “‘Mummies and babies’ and friends and lovers in Lesotho.” Journal of Homosexuality 11:97–116.
Gueboguo, Charles S. 2006. La question homosexuelle en Afrique: Le cas du Cameroun. Paris: Editions L’Harmattan.
Herskovits, Melville J. 1937. “A note on ‘woman marriage’ in Dahomey.” Africa 10:335–341.
—. 1938. Dahomey. An Ancient West African Kingdom. New York: Augustine.
Huber, Hugo. 1968. “Woman-woman marriage in some East African societies.” Anthropos 63:745–52.
Hulstaert, Gustave. 1938. Le mariage des Nkundó. Brussels: Librairie Falk.
Larken, P. M. 1926–7. “Impressions of the Azande.” Sudan Notes and Records 9:1–55; 10:85–134.
Moodie, T. Dunbar, Vivienne Ndatshe, and British [Mpande Wa] Sibuyi. 1988. “Migrancy and male sexuality in the South African Gold Mines.” Journal of Southern African Studies 14:228–256.
Morgan, Ruth, and Saskia Wieringa. 2006. Tommy Boys, Lesbian Men and Ancestral Wives. Johannesburg: Jacana.
Murray, Stephen O. 1997. “Explaining away same-sex sexuality when it obtrudes on anthropologists’ attention.” Anthropology Today 13(3):2–5.
Murray, Stephen O., and Will Roscoe [MR]. 1998. Boy Wives and Female Husbands: Studies of African Homosexualities. New York: Palgrave.
Nadel, Siegfried F. 1942. A Black Byzantium: The Kingdom of Nupe in Nigeria. New York: Oxford University Press.
—. 1947. The Nuba: An Anthropological Study of the Hill Tribes in Kordofan. Oxford: Oxford University Press.
Obbo, Christine. 1976. “Dominant male ideology and female options.” Africa 46:371–89.
O’Brien, Denise. 1977. “Female husbands in Southern Bantu societies.” In Sexual Stratification: A Cross-Cultural View, ed. by Alice Schlegel, pp. 109–126. New York: Columbia University Press.
Rich, Adrienne. 1980. “Compulsory heterosexuality and lesbian existence.” Signs 3:11–48.
Seligman, Charles Gabriel. 1930. Races of Africa. London: Thornton Butterworth.
Seligman, Charles Gabriel, and Brenda Z. Seligman. 1932. Pagan Tribes of the Nilotic Sudan. London: Routledge Paul.
Shepherd, Gill. 1978. “Transsexualism in Oman?” Man 13: 133–134.
—. 1987. “Rank, gender and homosexuality: Mombasa as a key to understanding sexual options.” In The Cultural Construction of Sexuality, ed. by Pat Caplan, 240–270. London:Tavistock.
Sweetman, David. 1984. Woman Leaders in African History. London: Hienemann.
Tauxier, Louis. 1912. Les Noirs du Soudan: Pays Mossi et Gourounni. Paris: Émile LaRose.
©2018, Stephen O. Murray
- notes in passing rather than being the focus of any anthropologists’ research in English or French; see Murray 1997 on this pattern around the world ↩
- This is not to say, of course, that there are no same-sex sexual relations, but these do not count as being instances of “sex.” ↩
- Falk challenged the estimate by Dutch sexologist Lucien von Römer that 23 percent of southwest African men were “inverts.” ↩