Women in African Muslim Societies, 1400-1700
This study might well be subtitled "Looking for the Invisible in the Dark", for the period under review is not one recognized by African historians (who prefer 1500-1800) nor has even the 'overlap' attracted the attention of those working either in Islamic or women's history. Between the 'truly Medieval' and the 'emergent Modern' lies one of those historical moments which has too long remained in the shadows of presumed 'more important' historiographical eras.
By the early fifteenth century, West African Saharans and the peoples of the sahel-savanna to the south were engaged with Islam.Most of the East African coast, its neighbouring islands, and scattered hinterland communities would have referred to themselves as 'Islamic' by culture, if not Muslim by faith. Sources, dominated by Arabic accounts originating in North Africa and the Middle East, reflect a few first-hand observations, but are mostly compostites of other authors relying in turn on unattributed information gathered haphazardly from Arab merchants. Not surprisingly, given the commercial focus of their interests, they rarely commented on culture, except as it compared favourably (or more often unfavourably) with their own experiences of Islam. Women appear as evaluations of mens' honour and virtue, as elements of domestic organization, and as measures of 'proper' social hierarchy.
The 16th century partial conquest of the coast by the Portuguese, and the 19th imposition of Arab Omani rule from their new capital in Zanzibar complicate the task of using any subsequent sources. The Portuguese, frustrated by their own military weakness and inability to access the wealth they observed, presented East African "Moors" as if they were still engaged in crusades; women were victims of these infidels ('proof' of Islam's oppression) and potential Christian converts. The powerful Omanis, in contrast, strongly influenced how people wished to see and have themselves seen with respect to their social origins, ethnic identity and religious affilications. Wealthy, Arab-centred and orthodox, the Omanis became a kind of 'internal other' against whom many East Africans wished to define themselves; some of the literature and histories we might otherwise regard as being fully indigenous and relatively transparent are thusly somewhat problematic.
Historians accustomed to such written texts are uncomfortable moving into the genre of art and literature, but it is here that questions of gender and culture are going to be most fruitfully pursued. In the East African context, Swahili poetry (utendi, (f.) utenda) was written in Arabic script beginning in the1600s. It reflected centuries of interaction with Islam and the Middle East (including Persia and India), while simultaneously building on African traditional oral structures. Working through Swahili, a language which was itself the creation of historical processes evolving from an oral to a written expression during the period under consideration, allows us to better penetrate the culture in which we seek women's reality.
For West Africa, historians have exploited oral traditions (epics and stories of origin) to complement the paucity of written records for this era, but not to explore the specific themes of Islamization and women. Saharan stories intended to legitimate a clan's economic or political position, take Muslim heritage for granted; they are buttressed by geneaologies that omit female family members. Women appear as slaves, concubines or the 'cause' of conflict generating tribal divisions. Here too, a major art-form is poetry composed in the Arabic-Berber dialect hassanya; relatively little has been recorded, even less plumbed by historians, although it has long been Saharans’ preferred manner of 'remembering' and teaching about the past . A whole genre is composed in the context of 'love', offering insights into gender relationships which could yet be useful. South of the Sahara, traditions are more often of the poetic, 'epic' genre, rich in metaphor, nuanced in meaning but potentially more revealing of gender relations and women's roles than the external sources interested in primarily in male-dominated trade and politics. However, the literacy which accompanied Islam meant that as some of these epics were written for the first time, they acquired an 'Islamic gloss': patrilineal practices, women's veiling and seclusion, lawful treatment of slaves and concubines were emphasized. This tendency was accentuated from the late 18th century as reformist movements, jihads, spread across West Africa. Colonial European sources, in turn, frequently attributed the weight of authenticity to these written variations in their own interpretations, further complicating our task of interrogating early 'oral' traditions.
Contemporary secondary histories of these regions argue that the most significant change occuring prior to the 18th century was the strengthening of Islam: as Islam increasingly ‘gained hold’ of society, women increasingly ‘lost access to’ personal and political power. This personifying of ‘Islam’ simultaneously privileges male voice and obscures female actions of 'choice'. Between 1400 and 1700, West and East African societies were challenged by the destablizing effects of internal warfare, prolonged droughts, and population fluctuations; they also addressed the transforming impact of international trade (including the slave trades). Islam was absorbed and nurtured in very particular, indeed personal, ways by societies and peoples whose experiences in this context varied widely. Class and gender feature prominently in explaining these differences.
The most visible women are slaves: domestic slaves prepared and served food; female slaves spun cotton in Senegambia (West Africa), and slaves (including women) outnumbered "Moors" on farms and in palm groves around Kilwa and Mombassa (East Africa). But most commented upon are concubines. In West Africa’s Ancient Mali, royal concubines in the 13th century wore fine clothes and jewels; 1000s of them accompanied Mansa Musa on his famous pilgrimage through Egypt. A late 15th century ruler of Kano (Hausaland) announced his 'commitment' to Islam by ordering the catching of girls and women to populate his new harem. And a famous 17th century Swahili utendi, 'Lament to Greatness', spoke of a declining urban civilization which had once known "harem chambers" ringing with laughter and the talk of slaves. There is nothing intrinsically Islamic about this upper-class concubinage, but there may have been a growing sense that the taking of female slaves should follow Muslim law, sharia. Mansa Musa was informed by Egyptian scholars that if he 'possessed' the beautiful daughters offered to him by his subjects, he must marry them; however, he was permitted only four such wives. Only slaves could be concubines and free women could not be treated as slaves. A 15th century ruler of Songhay (Mali's successor) consulted a North African scholar: slave girls, sold and then married to the purchaser, were frequently already pregnant; quarrels then errupted between the merchants. For the theologian, most important was the fact that sharia stipulated sold slave girls should be placed in the care of a trustworthy man until their next menstration in order to assure that they were not already pregnant (muwada’a); and it was the responsibility of good Muslim rulers to enforce this law. The implication here is a recognized responsibility on the part of the 'father' not to sell the mother of his child, umm al- walid. We can also infer from this exchange that female slaves were sometimes married, becoming actual 'wives'; both marriage and concubinage provided immediate social mobility and the potential of freedom if sharia was respected. That this was not always the case is revealed in the epic of a Malian ruler who had not treated his slave concubine well. She was forced to raise her son a slave. As a man, he brought civil war to the land in the quest for his 'rightful' power.
The 'moral' of the otherwise traditional story of succession reflects a strong cultural recognition of Islamic values betrayed, a 'gap' between internalizing and implementing sharia which may have become especially significant where it involved the treatment of women. One of the first signifiers of a 'truly' converted ruler was the restriction of his wives to four -- and invariably the expansion of his slave harem. Early accounts from both East and West Africa reveal a more subtle but equally important aspect of 'Islamic' gender relations: foreign merchants established themselves locally by marrying free women to gain access to land and power; they took concubines in order to have children. Whether women were destined to wear silks and jewels was not a function of being free or slave; it mattered more whether they were rural or urban, rich or poor -- slaves had more opportunity for social mobility than poor, free women. The role of the concubine was also critical: her children were the progeny of their father; she had no family to claim either child or inheritances on his/her behalf. Local noble women may have become increasingly less valued as mothers; frequent descriptions of dark and 'black Moors' in East African accounts, as well as observations in later centuries of 'black' bidan (Saharan nobles) are suggestive of such a trend.
The significance of these observations is underscored by the fact that these host African societies were matrilineal. Ibn Battuta noted that in the Sahara a man's heirs were the sons of his sister, and women had higher status than men; 'legitimacy' in the oral epics to the south was almost always expressed through female kin relations; and in East Africa, coastal traditions reveal early matrilineal inheritance patterns, while research in the hinterland shows land was inherited through female lineages. It is argued however, that as these societies became more Islamic, they became patrilineal; women lost the powers associated with matrilineality. But even somewhat coloured 'evidence' suggests otherwise. Oral traditions recounting the decline of Ancient Ghana reflect continuing matrilineal succession in spite of adoption of Islam by the state. Seventeenth-century Saharan genealogies were rewritten to 'prove' patrilineal descent, but their oral histories and poetry continued to betray matrilineal values. One well known East African epic about a struggle over power between a brother and half-brother is centred on the tension between the older African tradition of succession and the newer Islamic one, the latter dominates only belatedly. In one coastal region, the Muslim 'sultan' descended through the African female line for seven generations before a daughter took an Arab husband and it appears that patrilineal succession begin.
But it is not clear that women were excluded entirely from politics, even over time. In West Africa, marriages continued to cement alliances between Saharan clans, as well as powerful 'emirs' through the 19th century. Clan histories of origin and migration feature women as 'causes' of conflict but also as the means by which reconciliation later occurs. A late 15th century Queen of Hausaland, immortalized in a poem referencing her mortar of 'scented Guinea wood' and her pestle of 'solid silver', is called both (Muslim) "Amina" and (African)"Gumsa"; in asking Allah to give her the long life of a frog and the dignity of an eagle, the poet collapses into one cultural identity and definition of power in belief systems of two different but not necessarily competing worlds. Equally telling are the oral epics in which 'political power' remains inextricably tied to occult powers: battles are won with magic and the source of magic is inevtiably a powerful woman. Women (or their symbols -- food, mortars and pestles) appear as mothers, daughters, sisters and sirens, dynamic catalysts to men's actions who, like Saharan women, re-appear to 'accept the responsibility' imposed by such power. Some later manuscript versions, like those of the famous Sonjata, founder of Ancient Mali, attempt to displace the importance accorded to the hero’s mother and female kinfolk. In this case, crippled Sonjata's miraculous cure derives from a token taken from his father, rather than his mother; his sister ceases to have a role in his final taking of power. This kind of 'Islamisation of the past' is revealing, suggesting (as do traditions that retain a matrilineal social discourse) that societies may not yet have been as reformed in the Islamic patrinlineal image as historians have assumed. Sometimes it was necessary for traditional historians to re-write matrilineal conceptions of origin in order to create contemporary patrilineal identities and facilitate the legitimizing of 'proper' Muslim power.
Female rulers and regents, some combining African titles with Arab Muslim names, others being either purely African or Arab, continue to appear through the early 18th century in East African chronicles. One tradition celebrates the winning of power from one such Queen by a sultan who was himself the son of a humble woman. She had been the daughter of a fisherman, serendipitously 'discovered' by the sultan in answer to her father's prayers to Allah -- a 'legitimation' of the throne through a mother which makes sense only when read through a matrilineal lens. Less obtuse are five seventeenth-century coastal epitaphs commemorating woman rulers, convincing evidence, suggests one scholar, that we may have be mistaken in regarding women as a "suppressed class".
We need to dispense with both notions: women as 'class', women as 'suppressed'. It was precisely the diversity of classes to which women belonged which shaped their varied adoptions of, and adaptations to, Islam. "Suppression" reflects a contemporary (mis)valuing of Islam which in turn distorts the way we ask our questions of women's experience. In East African sources we glimpse wealthy wives, royal women and the female slaves of their households. The poor (free and non-free), the rural, the kin of artisans, petty traders, fisherman and sailors are invisible except when fate brings them into the realm of the 'important'. Thus, we meet a fisherman's daughter because she became a queen; the wive of an artisan because she was taken concubine to the Sultan, the Pate chronicle remembers her 'freeman' husband seeking revenge; the anonymous women of Malindi and Mombassa because the Portuguese recounted how they resisted or embraced (respectively) conversion to Christianity. West African Arabic sources rarely describe anyone who is neither royal nor slave and in oral epics, excepting slaves and servants mentioned incidently, women are uniformly identified by their kin roles. They remain undifferentiated except in terms of magical power, significant but still leaving the majority of women invisible.
But we also aggrevate these intrinsic problems by rooting questions in contemporary perceptions -- like the notion of Muslim women as an 'oppressed class'. We measure Islam by the practises of seclusion (purdah) and veiling, although these were clearly not signifiers even for the prudish Ibn Battuta. The introduction of purdah is attributed to the same 15th century Hausa-Kano king who ordered the 'catching of women and girls' as concubines and the taking of 1000 'first-born virgins' as wives -- a notable contrevention of sharia, in spite of his 'purifying' reputation. Female seclusion among wealthy Hausa survived into the early twentieth century; yet by the 19th century, another Muslim purist bewailed the fact women in Hausaland no longer knew the basic teachings of their religion, including prayers. Men had 'abused' the notion of submission to Allah in demanding women’s unquestioning obedience in the guise of housework! In East Africa, there is no evidence that Islam brought veiling or seclusion to women. One Portuguese source mentioned that the 'Moors' "shut up" their wives, but the many accounts of finely dressed women, observant of skin colour and jewellery, belie the existence of general veiling and purdah even among the wealthy Swahili. The utendi "Lament for Greatness" implies the existence of gendered living space ("men's halls", "harem chambers") as does historical architectural research in the Kano palace, but this is neither synonymous with seclusion nor particular to Islam; its significance in terms of our subject matter has yet to be explored. And as for veiling, one scholar suggests that it was not Islamic but cultural in origin: early "Shirazi" settlers did not veil but the Arab immigrants seeking social and political ascendancy during the 16th and 17th centuries did. Veiling was a measure of their social and cultural status vis-à-vis local Swahili and Africans.
Recent research raises other issues. One famous 17th century utenda appears to be instructions from a mother to her daughter on how to be a good, submissive wive. Once received as 'evidence' of female oppression, it has recently been interpreted by a scholar who understands the literary form (as well as the words) as being a work of irony and sarcasm -- an expression of mockery for the 'power' supposed to be in the hands of male spouses and rulers -- 'veiled' by literary tradition to all but those who were intended to understand it. Not only does this approach open new windows on the accomplishments of women poets, for it appears that upper class women were generally educated and often adept at utenda, it suggests the existence of a female discourse negotiating the arena of social relationships in ways and with goals we have yet to understand. Research in parts the more rural hinterland shows a continuity of matrilineality: women did not marry 'down', marriages strengthened female lines, town and village 'wards' were defined and dominated by female lineages, and land remained in their control. Men used their religious roles to establish a kind of parallel system of power rooted in the mosque, which ultimately came to complement the strength of matrilineality.
The received wisdom that Islam's 'penetration' was about male traders and clerics, and Islam's 'strength' was in the hands of sultans and ulama overlooks the fact that in Africa's matrilineal societies, women were in the position to shape how Islam would be integrated into local culture and politics. As mothers, wives, sisters, daughters and slaves they had many ways to influence society. That matrinlineality both competed with and accomodated Islam's patrilineal infrastructure, and that the 'extra-Islamic' power of magic and local cults associated with women remained important to the practice of Islam, should not be surprising. Integrating different belief systems and creating new, sometimes parallel power structures within society and culture were not evidence of 'inferior' or 'regressive' Islam. More often, they reflected responses of women to real social change generated by combinations of internal dynamcis and external factors. We should not be looking at Islamic culture in Africa as something imposed, measured against signifiers like seclusion and veiling; rather, we should be seeking women's historical experiences, as we have tried to do here, in order to understand just what constituted Islamic culture in any given time and place.