This analysis is inspired by two recently published studies, one by Cornelia Giesing and Valentin Vydrine on the origin of the Kaabu Kingdom, the other by Emily Lynn Osborn on state formation in Kankan.Footnote 1 Both of these studies feature oral traditions on the origin of an elite's political rule that refer to a migrating, in-marrying Muslim. This is a variation on the common myth ‘that outsiders brought statecraft to sub-Saharan Africa’.Footnote 2 The relationship between the marriage pattern and the development of the storyline in each account suggests, however, an alternative to this familiar reading. Such accounts may have little to do with a historical migration by an in-marrying Muslim, but instead reveal a social code, by which I mean a set of ideas, conditions, and premises necessary for successful political rule. This alternative, political reading of accounts of in-marrying Muslims also appears to be applicable to the most celebrated oral tradition of the Western Sudan, the Sunjata epic, and offers strong support for the argument that most of that epic as we know it is of post-medieval origin.
By demonstrating that the narratives from Kaabu and Kankan can be understood as variants of the same social code, it will be argued that this social code was of major importance in both matrilineal and patrilineal societies. This principle will be applied to the Sunjata epic, which is why the epic serves well for both patrilineal and matrilineal groups. It is also applicable to narratives that appear to be lists of successive warrior kings, but that reveal implicitly a change in marriage practices and in descent systems. Reading accounts of in-marrying Muslims as conveying a social code to be observed by the political elite sheds light on the principles guiding the relationships between states in the Western Sudan in the period c. 1600–c. 1850.
This analysis of the relationship between marriage and Muslim identity was inspired by Jarich G. Oosten's anthropological study of social codes and marriage patterns in Indo-European mythology, and by a crucial observation by Robert Launay in his ethnography of the town of Korhogo, in Côte d'Ivoire.Footnote 3 Launay observed that the town's population consisted of two groups, both of which were devout Muslims, and, on the basis of what he knew about their family histories, these groups had been experiencing similar processes of Islamization. Yet they considered themselves to be different: one group was cast as ‘unbelievers’, the other as ‘Muslims’. Launay then notes:
In effect, the distinction between Muslims and non-Muslims, at least within the confines of the region, coincided exactly with the distinction between Dyula and non-Dyula. … These distinctions were reflected in marriage patterns. Tun tigi [‘warrior’] and mory [‘Muslim cleric’] groups [who together form the Dyula] could give and receive wives to and from one another; they could also receive wives from ‘unbelievers’ (non-Dyula), but not bestow their own women in marriage to them, since Muslim women ought only be given in marriage to Muslim men.Footnote 4
Thus, Launay demonstrated, the label ‘Muslim’ or ‘unbeliever’ did not refer in reality to an individual or group's religion, but to whether one was a ‘bride-taker’ or a ‘bride-giver’. ‘Unbelievers’ could also be adopted by ‘Muslims’ and change their patronymic in order to acquire permission to marry a ‘Muslim’ woman, thus becoming a ‘Muslim marrying a Muslim’. But, generally speaking, within the logics of Islam, non-Muslim men had to marry women from a non-Muslim background. In sum, there was an identity component to being ‘Muslim’ whereby the group defined status by means of marriage constraints; ‘Muslim’ identity can thus be analyzed as a position in a matrimonial system.Footnote 5
My analysis has been inspired by Launay's ethnographic observation, but it is not an analysis of marriage rules and marriage practices: it will use Launay's logics of thinking in terms of the binary ‘Muslim’ and ‘unbeliever’ as the key to a social code related to political power and political concerns. When the labels ‘Muslim’ and ‘non-Muslim’ are read as political reflections on alliance-making, one can begin to sense the political complexity of a historical narrative that traces the origins of a society back to a matrimonial conflict or a collaboration between two men, one of them a Muslim (marabout), the other a non-Muslim (warrior), thus integrating kinship, religion, and warfare. Such integrative narratives are common in the Western Sudan; when read in terms of marriages, historical foundation narratives such as those from Kaabu and Kankan will reveal the concerns of the political elites of the Western Sudan in the period c. 1600–c. 1850. Whether they also document real marriage practices is an issue that goes beyond the scope of my analysis.
I read oral traditions as referring not to characteristics attributed to an individual (often an héros civilisateur), but as referring to idealized social relationships between persons occupying particular social roles or statuses.Footnote 6 This structural reading treats a source as a narrative that represents the crucial relationships contemporary to the period and area in which the narrative is told. This approach has been well established in historical research on Africa for decades. Joseph Miller, for instance, in his pivotal study of the early states of Angola, Kings and Kinsmen, interpreted his oral sources similarly: ‘The traditions … never deal with individuals. They further refer selectively to only certain parts of the past, those which have evident analogues in the present.’Footnote 7 And also: ‘The historical genealogies … consist of sets of personal names linked to the other by … conventional relationships of descent and affinity.’Footnote 8 My study of sources from the Western Sudan also follows Miller's approach by declining to reconstruct processes of state formation on the basis of oral evidence alone. Such reconstructions are the simplified products of Western assumptions and preconceptions – which Miller analyzes convincingly with the concept of ‘the Hamitic myth’.Footnote 9
ON THE ORIGIN OF POLITICAL POWER IN KAABU
Giesing and Vydrine's critical text edition of the Bijini Ta:rikh Mandinka presents a detailed and rich text that contributes to the study of the history of the famous kingdom of Kaabu.Footnote 10 In their lengthy introduction, the two authors offer little interpretation of their own, and, in their defence, one might argue that analysis and interpretation are not the main objectives of a critical text edition. Giesing and Vydrine present the manuscript texts as accounts of what happened to individuals and groups, thus following the local (and national) appreciation of the texts. However the authors miss crucial dimensions of the narratives by understating the social relationships (matrimonial and otherwise) expressed in the protagonists’ actions.
My analysis is centred around the foundation narrative on Kaabu, which is only one part of the Bijini Ta:rikh Mandinka. Bijini is a small town in present-day Guinea-Bissau; the Bijini Ta:rikh Mandinka relates its history and the history of its wider context, which is the Kaabu kingdom. The Bijini Ta:rikh Mandinka is a compilation of three manuscripts dating back to the nineteenth century that were previously guarded and transmitted separately.Footnote 11
The second manuscript in the compilation relates the origin of three matrilineal warrior groups (ñànco). These three groups are said to descend from one mother (Tenenbaa), who lived in a cave in the wilderness, and prayed daily. After being caught, she was imprisoned in a locked windowless house, where she mysteriously gave birth to three daughters. The narrative suggests that the daughters were the result of intercourse with a jinn. Footnote 12 After being liberated from the house, Tenenbaa married the king, and subsequently seized the land from the king by treason. Each of the three daughters married, thus creating Kaabu's three warrior groups. Each group had access to the throne in turn. Tenenbaa clearly represents Islam and its introduction, through her practice of praying outside the cave in which she had formerly lived.Footnote 13
Political accounts in polygamous societies often start with one mother, as a ‘conventional relationship of descent’, when these accounts express sound alliances. The image of uterine kinship through a single mother represents the harmony at the base of a stable social organization. By contrast descent from ‘brothers’ or, even worse, from half-brothers, represents rivalry and conflict, or at the least constant renegotiation. A narrative such as the Kaabu story casts the origin of warrior groups in terms of descent from three mothers and one grandmother: it emphasizes unity and stresses the necessity for harmony among warriors, and thus in the organization of the army.Footnote 14 Their father, while technically necessary to produce the daughters, is structurally irrelevant and in this instance portrayed as a jinn.
More important, though, the Bijini account of the origin of the ñànco is a variation on a well-known regional narrative theme of the bush woman: marriage with a stranger-princess (fled from her home, living in the wilderness), an autochthonous slave, a woman with the characteristics of a bush animal, or a strong (‘puissante’) woman. Giesing and Vydrine note the unique identity of these women in marriage – ‘la mère fondatrice … est un don’ (the founding mother … is a gift) – and elaborate on the characteristics of these women. Here I will instead consider the political position that these women effectuate for male rulers and their lineages.Footnote 15
What matters in these narratives is not the woman's characteristics, but that in marrying a ‘bush woman’ a man bypasses the obligations relating to a regular marriage. Marriage with such a woman entails no reciprocal obligations towards a different lineage of bride-givers; the ruler who marries such a woman is indebted to no other lineage. In short, these stories reflect the matrimonial concerns of rulers straining to retain control of their family lines. The code ‘behind’ these concerns argues that power is most autonomous and least threatened when it can transcend conventional matrimonial obligations; an unexceptional bride would have engendered social and political obligations in relation to the bride-givers.
I agree with Giesing and Vydrine that it is important to analyze marriage in terms of the classical dichotomy of symmetry and asymmetry, thus stressing the debt relationship between a bride-giver and a bride-taker. Though the idea may be tempting, it is, however, not possible to analyze these marriages between Muslims and warriors in the Western Sudan in terms of hypergamy: there is no evidence that bride-takers were either poorer or richer than the bride-givers or that Muslims by definition had a higher status than warriors. Such genealogical accounts have to be read as conventional relationships of descent, in this case to in-laws; they are similar because they carry the same message: avoid personal and lineage debt that might lead to alternative claims to power, offspring, and property.
WHY SUNJATA NEVER MARRIED
The analysis of the tale of Tenenbaa reveals a principle that holds throughout the Western Sudan, namely the advantages to be had from avoiding the loss of autonomy as a result of marriage relations. This principle explains why Sunjata never married in the numerous versions of the epic collected so far. The evasion of marriage obligations (encoded in the wild woman theme) thus explains the political logic of the Sunjata epic, which is the macro-framework for history across a large part of the Western Sudan. In all its variants the epic closely considers how Sunjata's father, the king of Manding, acquired as a bride Sogolon the Ugly, Sunjata's mother, as a reward (!) for having the buffalo of Do killed. Although a princess of Do by birth, Sogolon is not considered a marriage partner by her father, because she is an ugly hunchback. In numerous versions of the epic, Sogolon transforms herself into a giant porcupine on her way from Do to Manding.Footnote 16 Because of this transformation into a wild animal, the hunters who won Sogolon for killing the buffalo of Do cannot have sexual intercourse with her, and they decide to give her to their king, who has a son by her.Footnote 17 After the king's death, he is first succeeded by his firstborn son, born to his first wife. She carries the familiar Muslim patronymic Berete and apparently could only have been acquired through conventional exchange with the bride-givers – no version of the epic elaborates on her origin. This succession by the son born from the senior wife is, politically speaking, a failure, since the weak king is defeated by the blacksmith king Sumaoro Kante. Sogolon's son Sunjata, though, is successful in establishing Manding society. Again, as in the Kaabu case, a marriage without obligations to the group of bride-givers is the basis for political success as well as for the establishment of autonomous power.
Reproduction of society, though, would require different concepts and images, namely those that recognize one or more groups of bride-givers: in-laws, who by definition threaten a ruler's autonomy. That is why Sunjata never married in the epic: marriage would have undermined the core idea in the epic of Sunjata representing the autonomous power of all his descendants, the numerous rulers in the Western Sudan with the patronymic Keita who all trace descent to Sunjata.
Hence only full male siblings descending from Sunjata are mentioned in the ‘conventional relationships of descent’ that griots recite when listing the generations after Sunjata.Footnote 18 These genealogies do not mention women, because they deal with issues of both legitimacy and autonomy. When reciting genealogies with only full male siblings griots emphasize implicitly a kind of unity in the society founded by Sunjata, and they avoid the issue of obligations towards bride-givers/in-laws.
REPRODUCTION OF RULE: MATRILINEAL WARRIORS VERSUS PATRILINEAL MUSLIMS?
Political rule requires political autonomy, but rule has to be reproduced. Political elites of the Western Sudan practised a strategy to keep obligations towards others to a minimum as a firm base for successful reproduction of rule. Nevertheless in the context of the political, military, and economic reality in the Western Sudan in the period c. 1600–c. 1850, no ruler could survive without making alliances with other rulers. Marriage was the diplomatic tool for this alliance politics – one cannot avoid indebtedness, generation after generation, by marrying a woman without a family, since that would politically isolate a ruler and his descendants.
Issues of marriage and inheritance are organized in a descent system. The Western Sudan has known at least two descent systems: matrilineality and patrilineality. Matrilineality guarantees that a ‘bastard’ will never be on the throne, since the candidate is guaranteed to emerge from a royal womb, namely the king's sister, and thus royal power is safely transmitted, symbolically speaking.Footnote 19 For precolonial West Africa it has been argued that the slave trade had an impact in favour of matrilineality, while Islam strongly favoured patrilineality.Footnote 20 Patrilineality has political advantages for a ruler, since it helps him retain direct control of power. In the pre-1850 Western Sudan, matrilineality and patrilineality might have oscillated or transitioned into each other, with perhaps some space, too, for double descent systems, because the slave trade and Islam were each vital economic and political forces in the Western Sudan prior to 1850.Footnote 21 I therefore integrate in my analysis examples from a patrilineal and a matrilineal descent system, thus illustrating that a similar social code featured in both of these systems.
For the Western Sudan, the accounts of matrilineal succession should be studied in relation to historical developments in which Muslims increasingly competed for political power, leading in the nineteenth century to the dominant political position of Islamic reformist movements. Levtzion's explanation for why, before 1700, rulers married Muslims but, at the same time, had to keep them at a distance has a compelling logic for the warrior rulers’ relationships with Muslims.Footnote 22 Levtzion's description of rulers who migrated from the medieval empire of Mali also holds for other states in the Western Sudan:
The founders of the Mandingue states on the Gambia migrated from the territories of the empire of Mali. They were accompanied by clerics, who played roles similar to those in other states. The Muslim traders on the Gambia carried the slave trade in response to a growing demand by the Europeans on the coast. They controlled also the supply of firearms, bought from the Europeans, which made them an asset but also a potential threat to the rulers. Footnote 23
Levtzion sees the subsequent rise in Islamic militancy ‘as a radical departure from earlier patterns of relationships between Muslim clerics and scholars, who had been outside the field of political competition, and chiefs, who though not practicing Muslims were not considered unbelievers’.Footnote 24
Muslims were thus kept out of political office and warfare – either in practice or in the imagination of society, or in both – because they were either useless clerics, or dangerous suppliers of firearms. That the political impact of Muslims was growing and increasingly seen as a threat is, however, illustrated well by the first Wolof rulers, who refused to marry into clerical families, and who marginalized their own sons with too great an interest in Quranic studies.Footnote 25 It was only in the eighteenth century that, again according to Levtzion, ‘clerics were given, for the first time, territorial chieftaincies. The royal family also sought to cement its relations with the clergy through political marriages.’Footnote 26 At the same time, however, the Wolof rulers positioned themselves as ‘unbelievers’:
Confrontation with the militant Islamic movements changed political perceptions toward Islam. Whereas earlier European accounts referred to the Wolof as Muslims, later European travelers, from the end of the eighteenth century and throughout the nineteenth century, say that the Wolof were Muslims but their rulers were ‘pagans’. It has been only since the end of the nineteenth century that the whole Wolof society converted to Islam.Footnote 27
When Islam became more influential, political elites first seem to have stressed their claim to warrior status, but at a certain moment they faced the challenge of how to integrate Islam. Islam became such a prominent political factor that it became difficult to reproduce rule solely through the ‘warrior’ elite. That is why one elite positioned itself with a Muslim identity, and the other maintained a warrior identity. Both choices were politically inspired.
The political elite of Kaabu developed a matrilineal interpretation of the social code that came under pressure in the mid-nineteenth century. It was not until then that Islam became such a strong religious discourse in the Western Sudan that it could be used to justify political rule, and no longer needed a matrimonial component. This strength of Islam is nicely illustrated in, for instance, the tone and content of the historical narratives on Islamic jihads in the Pakao region in southern Senegal. These narratives, which deal with a nineteenth-century situation, are mainly about the punishment of peoples and villages for religious reasons; there are no references to marriage as a source of either power or weakness.Footnote 28
Indeed, in the eighteenth century, ‘political perceptions [changed] toward Islam’, as Levtzion writes, but there was not yet a radical departure from patterns of relationships that were valued, since the importance of marriages as a basis for political success continued to be emphasized. Given the political situation of that era, the stories of ‘bride-givers’ versus ‘bride-takers’ have a religious-political-economic-military relevance.Footnote 29 And rulers always sought to represent their power and authority as something for which they were not indebted to either another ruler or a trader/leader of armed mercenaries/supplier of firearms. The rulers of Kankan developed a patrilineal variant that successfully met these conditions, but first I will illustrate how the Kaabu political elite dealt with them.
WHEN MARRYING A MUSLIM – KAABU'S MATRILINEAL STRATEGY
The narratives published by Giesing and Vydrine reveal from a matrilineal perspective the social code that underlies the logics of the reproduction of political rule.Footnote 30 Giesing and Vydrine give an excellent mid-nineteenth century example of matrilineality under pressure in practice. This lengthy case relates in detail how marriage relations negotiated the structural tensions in society. To me, this case illustrates the dynamics of the rulers’ social code. Giesing and Vydrine write (my translation):
Hecquard [1855] reports that in Kaabu at the turn of the eighteenth century, marriages between marabouts and women of royal families were unexceptional. Around 1850/1851, he learned that the king of Pakis, Mansa Baakar, was the son of a marabout and of a ñànco, and that his legitimate successor could only be the son of a ñànco – either of his own wife or of his sister.Footnote 31
… the asymmetry of relations with local guests [Muslim strangers] is disliked in strictly patrilineal societies like those of the Balanta who, as a principle, didn't accept that the invited Muslims married their daughters without giving their own in return. The oral traditions that we collected among the Balanta in the Woyi-region, people who, as was already written by Labat [in 1728], didn't marry their daughters to strangers or neighbours, often imply that they were not able to develop good relations with the Mandinka [Muslims], because these didn't meet the reciprocity expected by the Balanta regarding the exchange of women.Footnote 32
If, on the contrary, the strangers were matrilineal, an absence of asymmetry or reciprocity was of lesser weight: the sons born from relationships between mothers from a matrilineal society and a patrilineal father had a choice. Depending upon the status of their mother, they could become sòoninkee kings, or follow their fathers’ heritage, as was opted for in the case of the ‘first imam’ of Bijini.Footnote 33
Moreover, Mansa Baakar, the son of a marabout and a ñànco woman of the Pakis royal lineage, is an example among many others of legitimization through the mother's side. He has a right to kingship because his mother is a ñànco. The practicing of this function obliged him, according to Hecquard, to adhere to sòoninkee rites. On the other side, the same author reports that Mansa Baakar, not having children from a ñànco mother, didn't have the right to abdicate, in favour of his non-ñànco son, because of the matrilineal succession rules according to which the sons of his sister were the only legitimate inheritors of the throne. Hecquard's testimony enables us to note the fact that the king of Kaabu was the son of a Muslim and, though a member of the Islamic community, he didn't impede at all the preservation of the pagan religion and the succession rights that were in practice. The reproduction of the sòoninkee identity is not challenged, keeping the responsibility for this continuity in the hands of the king's sister, which Hecquard noted as well.
To summarize, daughters of mixed marriages became, according to the rules, mothers of Muslims, while the sons had the possibility to choose the side to which they wished to belong or to fuse the advantages of the two. But the impact of reformist Islamic movements that accompanied the Foutanké who conquered Kaabu resulted in a tendency to favour patrilineal succession.Footnote 34
While the foundation of a society has to be based on a marriage without obligations towards the bride-givers, for the reproduction of society marriage obligations cannot be avoided. However, efforts were made to minimize obligations and alternate claims in order to avoid undermining the ruler's position. However, the reproduction of legitimate authority had to include both its military (warrior) and the religious (Muslim) aspects; both were indispensable for successful reproduction of political rule. The question then is: how can such success be framed conceptually? The social code reveals here a discussion about when to marry a Muslim, and who to marry to a Muslim. The Kaabu case illustrates the strategies for a matrilineal warrior ruler. Through his sister the ruler has to produce a successor with a warrior identity, but with a Muslim father. Regarding the future king, the Muslim dimension through the father is kept minimal thanks to the matrilineal system of succession of rule. But what about a patrilineal Muslim ruler? How does he integrate the warrior dimension necessary to reproduce his rule?
WHEN MARRYING A MUSLIM – KANKAN'S PATRILINEAL SUCCESS STORY
The increasing influence of Islamic reformist movements resulted in political power among groups that represented themselves as Muslims and also transmitted rule through patrilineal succession – the Islamic preference. This model of transmission was preferred in Kankan, and such a model would remove the rulers’ option of producing an heir with a ñànco status, as was the case in Kaabu. The social code in Kankan's narratives of the origins of power reveals that Kankan's political elite's strategy was initially based on the same concerns as those of their matrilineal, warrior counterparts. I suggest this by a rereading of the narratives published in Osborn's monograph on Kankan's history; to me, these reveal the patrilineal political elite's ideological problems and matrimonial strategies.
As a place of political importance Kankan entered the scene in the second half of the eighteenth century. The first reference to Kankan in a European written document is from 1820, when it was a flourishing commercial centre. The town's prosperity and growth were then of recent date.Footnote 35 Osborn argues compellingly that Kankan's rise is related to the economic rise of Freetown. Certainly, Kankan and its hinterland were confronted with increasing warfare in the period after 1750. In practice, therefore, the political elite of Kankan had to survive in a context of both increasing warfare and the increasing influence of Islam.
The oral traditions that Osborn presents for Kankan reveal matrimonial issues and political concerns similar to those of Kaabu's political elite discussed above. I will now analyze these matrimonial issues by offering a structural reading of the oral traditions, thus elaborating an analytical alternative to Osborn's more literal use of these traditions. The Kaba are Kankan's ruling family. Their ancestor is Kaba Laye, who is portrayed as an Islamic cleric in Jafunu, a centre of Suwarian Islam, far away north of Kankan.Footnote 36 As a reward for an impressive deed, a warrior gives his daughter to Kaba Laye.Footnote 37 This marriage results in two sons. While still living in Jafunu, Kaba Laye marries the daughter of a chief. This marriage yields a daughter and a son. The son's name is Abdurahamane Kaba, and he is the ancestor of the Kaba family that will become rulers of the polity of Baté, and that will establish, later, the famous town of Kankan, at the banks of the Milo River. Osborn writes:
[Abdurahamane Kaba] was diligent in his studies and humble to his elders, and he provoked the jealousy of his half-brothers. The sister was a gifted person who saw that the life of her full brother was under threat. She advised him to escape to a village located between the rivers. … She told him, ‘Go and wait for me in that place. My husband has seven years left, I will follow you when he goes to the next world.’ The arrival of Abdurahamane on the banks of the Milo River marks the birth of the state of Baté.Footnote 38
In the next story presented by Osborn, Abdurahamane Kaba's son – Fodé Moudou Kaba – is the protagonist. Unfortunately for the analysis I undertake here, there are no data on Fodé Moudou Kaba's mother. Consistent with the structural analysis I present here, however, this absence of information is not a coincidence: for successful military rule one needs a warrior mother, but this would present an unacceptable image for the ruling Kaba family of Kankan. Thus, a silence is preferred.
The remainder of this section provides a summary of Osborn's account on the son Fodé Moudou Kaba and my reinterpretation of the material she presents.Footnote 39 Fodé Moudou, who lived in the Muslim town of Jankana, received a sign that the first person who found a certain baobab tree would bless his family with the leadership position. The chief of Jankana also learned this, and he instructed his wife to wake him up early the next day. She failed to do so and to make it worse she served his breakfast extremely late. In the meantime, Fodé Moudou had found the baobab tree, and there he built the shelter that became the site of the Kaba settlement. Osborn adds to this: ‘Indeed, some accounts make this point [that the wife's behaviour was the cause of the husband's political failure] by describing the chief's wife as being of a different ethnicity or of slave status.’ Later, Abdurahamane Kaba's sister joins her brother, and she brings Islam to the place: when she drops her bag in the Milo river, it becomes completely immersed, but the Quran in the bag remains dry. Because of her Islamic knowledge, her brother offers her and all her descendants leadership, but she refuses.
A few generations later, the oral traditions mention the legendary Alfa Kabiné, the founder of Kankan, after Baté's destruction in the invasions by a Wasulu ruler, which must have taken place sometime in the eighteenth century. Osborn dates Alfa Kabiné to c. 1770 – which is on the eve of Kankan's ascendance. Alfa Kabiné is remembered in Kankan as an unmarried man who lived in celibacy and dedicated his entire life to Islamic study.Footnote 40 He is also remembered for having introduced a system of furunyògònya (‘marrying each other’) as well as founding the nabaya system, which is, according to Osborn, a specific practice of ‘statecraft’:Footnote 41
… an elaborate system of welcome, or nabaya, meaning literally ‘come all here’. … Baté's elites gave traction to the principle of nabaya by granting migrants plots of land on which to settle and by arranging marriages for them with local women. … [T]he newly arrived ‘stranger’ married into Baté's leading families. … [T]he migrants who arrived in Kankan married into families of their male hosts, and it was through their new wives that male settlers gained access to the extended households that constituted the state.Footnote 42
What is described in these traditions in terms of matrimonial relationships and kinship relationships? I discern the following scenario concerning the political history of Kankan's early days, before the destruction of Baté. Abdurahamane Kaba, the legendary ancestor of the Kaba, has a father who marries two wives with a non-Muslim identity: one wife is a warrior's daughter, the other one a chief's daughter. The legendary founder himself is born from the womb of the chief's daughter. For his father, marriage with the first wife, the warrior woman, does not lead to political power for his family in Jafunu, the polity of his warrior in-laws, which indicates that Abdurahamane Kaba's father professes a Muslim identity. Abdurahamane Kaba then decides to go into exile, and thus gives up whatever claims he has in relation to his in-laws. He marries – which must have been in-marrying, and thus he is a bride-taker – since he fathers a son called Fodé Moudou. His alliance through in-marrying leads to political success within one generation by the son. This son is not married (or perhaps not yet), and therefore totally committed to his family's cause (and not indebted to in-laws). Fodé Moudou's rival, on the contrary, fails because of an error in his marriage choice: he takes a slave for a wife, or he marries a stranger woman – both are ‘category mistakes’. As a result, the rival loses access to rule.
Abdurahamane Kaba's sister accompanies her brother only after she is post-menopausal, so he cannot make alliances as a bride-giver by giving her in marriage to someone. Thus, he cannot alleviate the debt relationship with his in-laws through a direct reciprocal exchange. His sister brings Islamic knowledge to the family, which suggests that after the migration the family's Muslim identity had not been stressed. Abdurahamane Kaba then proposes to his sister that he establish his society on the basis of a matrilineal kinship system, but she rejects this (and thus a patrilineal option is chosen). That is how the protagonist of Kankan's successful patrilineal society became cast as a male stranger (Fodé Moudou) who had no competing obligations to his family, far away in Jafunu. He is, in Miller's words, ‘an outsider with no connection to local interest groups’ – who successfully institutionalized Islam after having seized political power (as a warrior). One sees here the mirror of Kaabu's successful matrilineal society, where the protagonist (Tenenbaa) is a female without obligations to her family and where the theme of the introduction of Islamic knowledge through a woman demands a less complicated storyline.
But how is one to marry, in order to reproduce rule, after the successful start? In the narrative of Kankan's glory days, the period after Alfa Kabiné had ‘rebuilt’ society, I read a rigid solution to the problem of indebtedness or dependence on in-laws. Alfa Kabiné follows Fodé Moudou's strategy to avoid incurring obligations towards others. He himself is unmarried and therefore has no marriage obligations. His autonomy is therefore not threatened and he is indebted to no one; he can be totally dedicated to his own lineage's cause. But in spite of his personal strategy, Alfa Kabiné can still be a bride-giver, since, as the head of his patriline, he can marry off his sisters (or his younger brothers’ daughters) to allies. And as such Alfa Kabiné is remembered by the successful nabaya system he introduced. This is the key to the reproduction of rule – Kankan's success – and it turns Kankan's political elite into bride-givers, and not bride-takers. One has to note that, in practice, a man can avoid the bride-taker's position of indebtedness not only by not marrying – Alfa Kabiné did not marry – but also by marrying a parallel cousin (father's brother's daughter), thus keeping the heritage within the clan/lineage. However, politically speaking this practice is dangerous, since it does not extend a family's political network of alliances.
But, having become bride-givers, how could the Kaba ruling family prevent their new allies from seizing power, as their ancestor Abdurahamane Kaba and his son Fodé Moudou had done? The figure of Alfa Kabiné, I argue, represents the solution to how to neutralize the bride-takers without losing the superior role in the debt relationship established by giving away the bride. Alfa Kabiné's life story is the starting point of a social process, and it solves this issue of marriage that was not discussed in the account of Fodé Moudou, an account that is instead concerned with foundation and autonomy. It might therefore be more than a coincidence that their stories are told separately; the two accounts deal with two different political processes, one about how political autonomy is acquired and one about how politics are practiced.Footnote 43
Kankan's elite apparently opted for an asymmetrical relationship through the nabaya system, thus incorporating strangers who had to be, by definition, ‘Muslim’. This is implicit in the narrative and meant that a stranger who looked for a relationship with Kankan's political elite had to convert to Islam.Footnote 44 One notes that the children of these marriages would never have access to a ruling position in the Kaba family – they belong to another patrilineage. This is a structural difference with the marriages documented for Kaabu, where the daughters of the political elite produced the future political elite. How, then, was one to ensure that the bride-taking stranger's son was as loyal to the bride-givers as his father was? That was solved by reversing the asymmetrical relationship, and marrying the stranger's daughter – who is also a cross-cousin (his father's sister's daughter) – and by continuing to marry cross-cousins. Thus a furunyògòn (‘marry each other’) relationship between families is established, and the introduction of that marriage system is what – according to Osborn – the oral traditions attribute to Alfa Kabiné as his major contribution to Kankan society. Furunyògònya is a system of reciprocal asymmetrical relationships, but with the family that started the cycle as slightly superior and less indebted. The nabaya system could be successful only in combination with furinyògònya. Thus, the implicit structure of the narrative confirms its explicit message. This internally logical account of Alfa Kabiné is a powerful explanation of how Kankan works as a patrilineal society, and therefore Kankan's present-day leading families trace their descent to Alfa Kabiné (and not further back in time).Footnote 45 The story of Abdurahamane Kaba and Fodé Moudou cannot be used for explaining how society actually works. Rather, the story legitimizes the political base of Kankan's autonomy while illustrating what happens when you make a category mistake in marriage: you are cut off.
The social code expressed in the stories about Kaabu and Kankan echoes similar concerns regarding the reproduction of rule among political elites of Western Sudan. The main concern of both groups was to minimize obligations to others and keep others indebted, thus maintaining a sound network of alliances. The cultural solution, however, was different. It had to be, since – in contrast to Kaabu's political elite – the Kankan political elite was patrilineal and Muslim. Therefore, Kankan rulers had to be bride-givers to other men with a Muslim identity (since a Muslim is forbidden to marry his daughter to an unbeliever), and, as Muslim rulers themselves, they could never produce sons who could choose a warrior identity, as was possible in Kaabu.
Immediate reciprocity by exchanging daughters with strangers was, apparently, not attractive. Immediate reciprocity is also, sociologically speaking, less complex and less dynamic than the system of furunyògònya practised by Kankan's rulers from Alfa Kabiné onwards. Since the Kaba rulers tended towards Suwarian doctrine, they had to innovate in order to survive politically and militarily in the eighteenth century. Furunyògònya was Kankan's patrilineal (Suwarian) Muslim implementation of the social code for successful rule in the Western Sudan in the period c. 1600–c. 1850. It was a system in which they secured their allies’ support, generation after generation.
Kankan's foundation narrative and Kaabu's foundation narrative circle around the same problems and solutions. A society can be successful only when it manages both religious and political-military affairs. When a society is founded, in Kaabu as well as in Kankan, isolated and exceptional women figuratively bring in Islam, conceptually reserving warrior activity for men. There would be no indebtedness to Muslim bride-givers. When a society is reproduced, it matters whether one is Muslim or not. In Kaabu's case, the rulers are warriors, and they need in addition Muslim status/Islam, which they accomplish through marriage. In Kankan, the rulers are already Muslim, and thus they need the status of the warriors’ strength, which they obtain through the nabaya system.Footnote 46 These strangers, however, must have a Muslim identity, since a Muslim is forbidden to marry his daughter to an unbeliever. In eighteenth-century Western Sudan, the number of ‘warrior Muslims/Muslim warriors’ was definitely increasing – ‘Islamic militancy’ becoming a growing factor.Footnote 47
Kankan's model coped with the new reality both practically and conceptually. The increasing presence of Muslim army leaders in combination with the conceptual inclusion of them might be the reason why Kankan flourished and Kaabu's position was under pressure.
The oral traditions from Kankan confirm the necessity and inevitability of warfare and violence in the organization of society. Though Alfa Kabiné is pictured in oral tradition as remaining faithful to the Suwarian tradition of not preaching jihad and of living in harmony with non-Muslims, we do know that Kankan rose to power in the second half of the eighteenth century, a period of increasing warfare. Though I agree with Osborn that the stories about Kankan express a Suwarian identity – I fully accept the idea that the Suwarian ideology of cohabitation/accommodation is represented in the stories – I disagree with Osborn that Kankan was pacifist. The stories relate to the politics of warfare: they legitimize autonomy and politico-military relationships with allies.Footnote 48
FROM MATRILINEALITY TO PATRILINEALITY – THE WARRIORS’ CHALLENGE FOR RULE
The stories discussed in this article are those of the political elite, and they should therefore be analyzed in the first instance as narratives of elite men and women's concerns at the time of their telling. These concerns are revealed in the structure of the narratives. The polities their grandfathers lived in had faced similar problems: warfare increasing justified through claims to defend Islamic orthodoxy, the slave trade, and marauding bands. Their accounts reveal the social code for survival in these circumstances: marry your son to the right person, marry his children to the right persons, give priority to the army, and minimize indebtedness to or dependence on others. While in the narratives individual men act univocally, of course in real life a man's broader entourage – including both senior men and senior women – would have had a role to play in the political stakes of marriage. Yet in the narratives themselves women feature largely as pawns in marriage politics. The oral traditions on state formation are best seen as ‘coded texts’ in which gender relations serve to convey lessons about the perils to the reproduction of lineage power.
As we know from matrilineal Kaabu, which was ruled by ‘warriors’, and patrilineal Kankan, which was ruled by ‘Muslims’, this social code expressed the necessity for minimal indebtedness towards others as a firm basis for rule. These two cases are each at the extreme of a gradually changing spectrum. The Sunjata epic, the grand narrative for the history of the Western Sudan, confirms the issues at stake in the social code. Sunjata's father is ‘implicitly’ a Muslim, since his first wife is a Berete, which is a patronymic representing a strong Muslim identity. The children of such a woman would have a completely Muslim identity; a marriage to the Muslim Berete woman will never produce a successful warrior as an heir.
The lack of political-military success attributed to a Muslim status in the social code in the Western Sudan in the period c. 1600–c. 1850 is in sharp contrast with contemporary descriptions of the medieval Mali kings, who are described as powerful rulers as well as practising Muslims. Given the prominence of a social code for political success in narratives of origin in the Western Sudan, I deduce that the Sunjata epic as we know it, with its focus on Sogolon the Ugly, bears traces of this transition in ideas of political rule. The Mali empire flourished economically thanks to the gold trade, and its decline and political transformation are historically related to the discovery of gold in the Americas as well as to the introduction of blade arms, and later, firearms in the Western Sudan from the late Middle Ages onwards.Footnote 49 In addition, the slave trade increased from the sixteenth century onwards, and warrior kings profited from both the internal slave trade and the Atlantic slave trade. Warfare became endemic and prompted the political elite of the Western Sudan to restructure conceptually its power base by including sustained access to the means of conducting warfare. This might be the historical condition in which the political elite developed a social code about marrying a Muslim and avoiding indebtedness (irrespective of the transfer of royal rule through matrilineality or patrilineality). This might explain why the stories about Sunjata's father and Sogolon the Ugly, so central in the epic as we know it today, are not referred to in the medieval sources that mention Sunjata only; the stories about the father are a later addition, because they discuss a post-medieval situation.Footnote 50
The elite's conceptual reorientation, with an emphasis on an identity as a warrior (needing firearms to rule successfully, but keeping indebtedness to rivals as well as to Islamic networks at a distance), should be read as a political social code and not as a religious statement against Islam. Perhaps this is why European visitors describe the Wolof chiefs as practising Muslims, while (probably not much later) these same chiefs rejected demands from ‘Muslim militants’ to convert to Islam.Footnote 51
The historical fact that reformist movements favoured patrilineal succession and that these movements gained increasing influence made me wonder whether there are oral traditions that refer to a transition from matrilineality to patrilineality. The figure of Fakoli, one of the major protagonists in the Sunjata epic and a figure considered to be the legendary ancestor of many families in West Africa, might provide information on such a transition. Most versions of the Sunjata epic explain Sunjata's victory over Sumaoro Kante by referring to the fact that Sumaoro's important general Fakoli deserted from Sumaoro's army to Sunjata, thus changing the military balance. Sumaoro was Fakoli's uncle (his mother's brother) and he took Fakoli's only wife to add to his existing 99 wives. This insult pushed Fakoli into Sunjata's camp. This is a widely known narrative in the Western Sudan.
The variant that caught my attention is based on a recording from Guinea.Footnote 52 In this text, Fakoli grows up in the land of Soso because his mother becomes a widow at an early age and therefore returns to her native land of Soso, ruled by her brother Sumaoro Kante.Footnote 53 This creates a standard situation for growing up for a royal successor in a matrilineal society. Another hint at the matrilineal transmission of rule comes, in this text, from the acts ascribed to Sunjata, after his defeat of Sumaoro: he decides to give the rule over Soso to Fakoli, thus factually transmitting royal rule through the matrilineage.Footnote 54
When the story of Fakoli's breach with his uncle's army is read as an account from a patrilineal society, it is ‘just another’ remarkable and shocking ‘family affair’. One does not expect his mother's brother to behave like that, since this relationship is very positively valued in patrilineal societies.Footnote 55 However, when read as a story from a matrilineal society, it touches at the heart of society: when a king takes his nephew's wife in a matrilineal society, this destroys the system. It is like a father taking his son's wife in a patrilineal society. The prominence of the narrative theme of Fakoli's wife in the Sunjata epic suggests that it tells a very important message; it may tell us something about once crucial issues of legitimate leadership in a matrilineal system of rule rather than refer to internal family questions in a patrilineal society.Footnote 56
When narrating the histories of their polities, the politically leading patrilineages of the Western Sudan needed stories about marriages to identify their relationships vis-à-vis rival lineages. However, issues of warfare, military organization, and inheritance within the lineage could be expressed in terms of ‘brothers’, ‘fathers’, and ‘sons’ – without a reference to a woman. A story about Farin Kamara, the ancestor of the ruling Kamara family in Moussadougou, taken from a 100-page typescript from the archives of Yves Person, illustrates this principle.Footnote 57 The text tells of dozens of wars and political conflicts between men; women are largely absent. The story of the rise of Farin Kamara, and his succession to his uncle Toumani Komè, is the only exception to this, since it specifically mentions marriages. This story can be read as an account of a political rupture in which the old system of matrilineality is used to ensure its own breakdown. I will simply summarize the narrative here. Farin Kamara's maternal uncle intended to respect matrilineal succession in naming his nephew successor, but the local elites were opposed to this on the grounds that Farin Kamara was a ‘stranger’ who would destroy their ways. When in power, Farin Kamara himself marries a marabout's daughter, he gives his daughter away in marriage to another marabout, and he invites a third marabout from Timbuktu. Henceforth it is the patrilineal offspring of these Muslim men who form the core of the state. The reluctance of the local elites appeared to be well founded: through these marriages Farin Kamara had introduced a completely different basis for political rule, namely patrilineal Islam.
The text of the 100-page typescript narrates in detail from the origin of the Kamara – who in the Western Sudan are cosmologically linked to the powers and spirits of the earth – to the rise and fall of Samori, who professed a distinct Muslim identity, and with whom the author's father collaborated. Through the figure of Farin Kamara, who uses the rule he acquired through matrilineal inheritance in effect to destroy this matrilineal system, the narrative links together earth power and Islam into one coherent narrative, explaining how a Kamara became involved in Samori's enterprise.
CONCLUSION: STATE FORMATION INTEGRATES ISLAM, WARFARE, AND THE SUNJATA EPIC
This analysis may deepen our understanding of the processes of state formation in the Western Sudan in the period c. 1600–c. 1850, because it uncovers a structural relationship between dimensions that were hitherto studied separately: religious notions for correct marriage, logics of military leadership and political rule, and the Sunjata epic. Narratives about the origins of a state, the failure of its predecessors, and the period of its greatest success reveal (among other things) a marriage code. The marriage code sets out strategies demanded by the political situation facing polities in the Western Sudan in a period characterized by growing reformist Islam on the one hand and the intensification of warfare fed by increasing access to European arms and the growing demand for slaves on the other. For the development of the story line, the point of departure matters: for rulers with a matrilineal warrior identity, the options are different than for rulers with a patrilineal Muslim identity. But for each group the available choices and solutions are finite because they contend with the same political concerns: each group claims political autonomy and seeks to minimize its indebtedness towards others. In the crucial period of the foundation of the state the founder marries an ‘unrelated’ female, thus avoiding an indebtedness to a bride-giver's group. The success of the state, in a later period, depends on successfully balancing military and religious dominance. A narrative that explains the workings of the state appears rather different for matrilineal warrior rulers than for patrilineal Muslim rulers. Nevertheless, on a closer look, the political elites’ ‘success stories’ are based on comparable social codes.
The importance of marriage to the success of a state attributes a crucial position to women, but not as central agents in this political game. Women are relevant to diplomatic arrangements between male rulers. Women feature in the political narratives of the Western Sudan in the period c. 1600–c. 1850 only when complex situations or processes of change have to be explained; accounts on the internal organization of a state are usually in terms of, in Miller's terminology, a conventional relationship of kinship between brothers or half-brothers. A set of fraternal relations is a relatively simple hierarchical model compared to a model for marriage in which feature issues of indebtedness and reciprocity.
A convincing account of the state's origin is at the same time, of course, a strong ideological tool in diplomacy. As the social code reveals that alliances are crucial to survive, diplomacy requires both historical knowledge and social expertise. These dimensions of diplomacy are ‘materialized’ by the strong presence of griots at the elites’ courts in the West African Sudan. For griots, history and diplomacy are the same activity.
This reference to griots bridges my argument to the Sunjata epic, West Africa's most acknowledged oral tradition. I argue that the Sunjata epic serves, for griots, as their template for diplomacy. The major story line of the Sunjata epic as we know it nowadays (the quest for Sunjata's mother Sogolon the Ugly, and an absence of marriage for Sunjata himself) perfectly illustrates the social code relevant when marrying a Muslim. The fact that these major themes represent the political issues of the period c. 1600–c. 1850 suggests that they have been integrated into the history of Mali ‘recently’; they may not pertain to medieval versions of the Sunjata epic.Footnote 58 The fact that both state foundation stories and the Sunjata epic convey the same social code suggests an integration of regional narratives and the meta-narrative, and it illustrates that oral historical accounts are produced in a process of diplomatic fine-tuning.
This study may therefore shed new light of the history of the Mali Empire. This empire is usually described as being in a process of disintegration after 1600, thus leaving the Western Sudan ruled by a large number of highly dynamic warrior states. In his classical study of the warrior state, Richard Roberts noted that the Middle Niger – Roberts's area of research – was in an almost permanent state of war between 1700 and 1850, yet the area nevertheless had a flourishing economy based on the slave trade and war. Roberts explains:
[My] political economic analysis examines not only the relationship between warfare and the economy, but the way warfare became an expression of the internal cohesion of the state. Warfare was one of the elements of societal construction as well as destruction. … But Umarian conquest [in the 1850s and 1860s] and subsequent forms of warfare and state intervention in the economy led to economic collapse in the riverine districts of the Middle Niger valley. Instead of acting as an engine of economic growth as it had under the Bambara, warfare now led to economic decline.Footnote 59
Roberts's analysis for the Middle Niger may hold for the entire Western Sudan. His study of the Segu warrior state, true to the available sources and the state of the historiography of the moment, focuses on economic issues and military conflicts.Footnote 60 The economy and the military may indeed explain the warrior state's dynamic, temporary character, but there must have been further stabilizing factors that saved the polities from permanent disintegration. The oral narratives offer us access to another dimension: the social code conveyed in the matrimonial strategies that structure the traditions of first-comers and strangers, of warriors, and Muslims. The narratives, to slightly modify Roberts's words, express the internal cohesion of the state through warfare. These narratives provide the stabilizing mechanisms and diplomatic templates through which peace and rule could be legitimized and reproduced. Thus, it may be incorrect to describe the Mali Empire after 1600 as having disintegrated; it seems more appropriate to see in the Western Sudan a complex network of polities that subtly integrated Mali's political heritage into contemporary military and religious concerns.Footnote 61
On the subject of making alliances with rival polities through marriage, Muslim and non-Muslim status appear to have been crucial themes in narratives that addressed concerns about political autonomy and dependence. Both terms of ‘Muslim’ and ‘warrior/unbeliever’ are more than religious identifiers: they have a political-military-economic status component. These terms, and the narratives in which they featured, negotiated important political issues and have probably never functioned as rigid, formally binding marriage codes.Footnote 62 Reading these narratives as accounts of internal political cohesion and/or attempts to political integration of military, religious, and economic agendas enriches our understanding of the history of the Western Sudan. The omnipresence and prestige of the Sunjata epic in the Western Sudan, right up until today, may confirm that the social code to be observed when marrying a Muslim was, in this area, crucial for political survival for several centuries.
