L'Empire de Ghana: le Wagadou et les traditions de Yéréré

Although we must revisit this work and engage in a deeper reading of the different recorded versions of Soninke oral traditions on Wagadu, Germaine Dieterlen's L'Empire de Ghana: le Wagadou et les traditions de Yéréré raises a number of intriguing questions about the extent we can reconstruct this distant past. Beginning with a translation of Diarra Sylla's rendition of this past as a gessere, Dieterlen then provides additional context and attempts a wider analysis of the ancient empire of Ghana (Wagadu). To some extent, she drew on the ethnographic present as well as cultural practices and rites among the Bambara, Dogon, and other West African peoples. This makes her interpretation of the institutions, social structure, and culture of Wagadu necessarily speculative.

In spite of some of these pitfalls in researching a period in pre-colonial West African history that is only partly covered in external Arabic sources, Dieterlen confirmed some of our own past theories. For instance, a cult of sorts to the vulture and hyena was part of the Soninke ritual calendar. According to Dieterlen, on the second moon of the Soninke's calendar, the wage (noble clans) gathered with their gessere to offer sacrifices to the hyena, an animal tied to their oath to Kaya Maga. Remember, the hyena played a role in assisting Diabe in Soninke tradition. There was also a special regard for the fox, it being a taboo for wage to kill this particular animal. Even the dog was the subject of taboos for the wage. For instance, they were not allowed to marry people who ate or sacrificed dogs, presumably due to the dog that accompanied the legendary ancestor, Dinga. One cannot help but wonder if the dogs with collars of gold and fine jewelry in the palace of the Ghana kings was in memory of this? Unsurprisingly, al-Bakri was likely ignorant of this, but informants told him of the palace dogs with richly decorated collars.

Besides these aforementioned taboos, Dieterlen offered some additional information on Kumbi and the Wagadu kingdom. Presumably based on her informant, she reported that Kumbi, the capital, had fifty-one quarters. Koumbi Saleh was allegedly the quarter for the market associated with caravan trade. The imperial quarter was supposedly called Kalaata, presumably the al-Ghaba of al-Bakri. There was even a quarter or section associated with the gessere. A glimpse of the different social classes or ranks in Wagadu can be seen in the description of four types of royal drums. Essentially, the major groups were the nobles, clan chiefs, slaves, and the people (commoners?). A distinct drum was used to call each of these types of residents in the capital city. Although it is difficult to see where the number of fifty-one quarters or neighborhoods came from, the number is probably meant to be symbolic or merely to suggest a huge concentration of people living within a larger capital district. That is very likely to have been true, as al-Bakri's description of the capital as two cities, one for Muslims and another for the king, with continuous habitations between, fits this.

In terms of the administration of the state, Dieterlen provided additional clues. The Koussa, for instance, participated in collecting gold for the king. Other Koussa were Koussa Tounkara, who assisted in organizing official ceremonies for the rulers of Wagadu. Another role attached to spiritual powers like access to ancestors through trance and the fabrication of ritual objects began with Djourou Tandjigo, an artisan in the times of Diabe. In addition, there were fade and hida, the governors or chiefs of provinces within Wagadu. Dumbe and djuwara were said to have been generals or military leaders. The Cisse clan which allegedly produced the most kings. A number of wage or noble clans were tied to specific provinces or regions, such as a figure remembered as Diamera Sokona, son of a fade named Birama, at Nakhata (interestingly, this figure was also said to have been the first to offer a woman to the Bida snake). Supposedly, control of territories included administration of land, markets, livestock, hunting grounds, and water sources, meaning these administrators had a lot of power over the areas under their remit. As for the military chiefs, Dieterlen suggested that these were the nana, a word possibly distorted in Arabic sources as Ghana.

But what of the known history of Wagadu as a state? Or its dynastic genealogy? Here contradictions abound as different versions of oral traditions disagree on details. Dieterlen presented Wagadu as dominated by the Cisse, beginning with Diabe as the first king based at Koumbi. Inexplicably, there were thirty-three of these Cisse kings before a member of the Toure clan became the 34th king. The Toure, who were associated with the power to control water, seized the kingdom with Moudou Toure at the head. A Muslim, Moudou Toure enjoyed support from some of the local Kakolo clans and Diane and Koma clans. Elsewhere, it is claimed that Madyou Toure was the last to reign, though it is unclear what his relationship was to Moudou Toure. Prior to this, one of the 5 Cisse clans apparently dominated rulership, the Kharesi. But the switch to the Toure may have led to a change in capitals and greater Islamization as the first Muslim kings (in the late 11th century, according to Arabic sources). One may interpret this as the rise to prominence of an Islamic noble clan in Wagadu, evidence of which can be found in external Arabic sources on Wagadu in the 1100s and 1200s. Even Zafun allegedly entered the picture, where the older brother of Diabe left when Diabe succeeded Dinga. This Cisse brother married a woman of the Dyafone clan of Dyafounou, suggesting links between Zafun and Wagadu began early in Ghana's history.

However, a looming question remains. Can one reconcile Dieterlen's reconstruction of the kingdom's past with other sources? Her text includes a few other variants of the Soninke tale of Kumbi and the dispersal from Wagadu after the killing of Bida. Some of these other sources provide a different king list for Wagadu. One named the Wagadu kings as Khaya Makha Sise, Wakane Sakho, Damera Sokhna, Gumane Fade, Sero Khumma (all brothers of Diabe, the first king). Then leadership passed to the family of D'ariso, beginning a new line of kings: Kambine, Suleyman, Bannabubu, Wagadu Makha II, Gane, Musa, Biramu, and then a disastrous civil war. Needless to say, some of the people considered kings here were supposedly noble governors or warrior leaders in Sylla and Dieterlen's version of events. We also wonder about Biramu, a name which sounds close to Birama. But once the D'ariso lineage succeeded, one notes father to son inheritance (6 generations) and the appearance of Islamic names. If the civil war which followed the reign of Biramu, among his own sons no less, involved a warrior from D'ara in Tiraka, then this could have been the moment of an Islamic faction seizing the throne. Tiraka, after all, had been an important center for trade to Ghana, Tadmakka, and other groups, with a Wangara population also there attested by some external Arabic sources. This raises the legitimate question of how to interpret the dynastic and/or religious change in Wagadu's leadership in the 11th century. It is tempting to see traditions of a Toure clan Muslim becoming king with the support of some elements within the kingdom (and possible support from the Almoravids).

Ultimately, Dieterlen's book raises more questions than it can possibly answer. Yet the connections she makes to other West African cultures and the Soninke ethnographic present can give the reader new ways to think about the history of pre-Islamic Wagadu. The clear role of the "traditional" religion as the legitimizing ideology behind kingship and the worldview of the people, as well as the strong role played by lineage in creating and distributing hierarchies, rank, and identities is worth exploring.