on the excavations at the two sites, no material was recovered that would permit positive identification of them as the sites of Awdaghost or the capital of Ghana described by early chroniclers. The strong case for these identifications is archaeological rather than epigraphic in nature.

Among the major questions that remain unanswered is the nature of the settlement system at Koumbi Saleh and Tegdaoust. The “city-centric” approach of the Devisse team, which mainly focused on the central area of stone-built ruins, involved little systematic investigation of sites in the hinterland of these two towns. We thus do not know if Koumbi Saleh and Tegdaoust existed essentially as ports of call, supported by trade and local agriculture but largely isolated from other sites of appreciable size, or whether they functioned as part of a well-integrated network of villages and hamlets. Bonnel de Mezières (1923) commented on the dense zone of archaeological material that extended around Koumbi Saleh in a zone thirty-five kilometers in diameter. Because there is no pottery sequence for the region, it is impossible to know which of the sites in this zone were contemporaneous with Koumbi Saleh at various points in its evolution. Thus, we know little of the developing settlement system, which is fundamental to any reconstruction of the nature of the Ghana polity.

The emphasis on the medieval paradigm on North African artifacts and influences has also left us remarkably uninformed as to the Sudanic aspects of the Ghana polity. Al-Bakri’s information establishes firmly, however, the very sub-Saharan nature of the kingdom. In his description of the pagan cults of Ghana we readily recognize the forest shrines, fetishes, and sorcery that are widespread in the Mande religion and elsewhere in West Africa.

French archaeological activity in the middle Senegal Valley (MSV), in the area of the early kingdoms of Takrur and Sila mentioned by al-Bakri, was relatively rare during the colonial period and most often took the form of casual surface reconnaissance. The unsystematic collection of a small number of appealing artifacts from surface walking or small-scale sondages was a common method of fieldwork into the 1980s. The animating spirit of these collections, registered with minimal notes or accompanying observations at IFAN in Dakar, was strongly antiquarian.

Internationalization (1970s Onward)

The general site inventory accomplished between 1971 and 1973 under the leadership of Charles Becker (a historian by training) was an important step toward a more systematic approach to archaeology in the MSV (Martin and Becker 1974, 1984). Contemporaneous with the inventory by Martin and Becker were large-scale excavations undertaken by the IFAN physical anthropologist Guy Thilmans and IFAN research associate Annie Ravisé at Sincu Bara. Important not only for highly significant finds of brass and silver artifacts, the excavations also ushered in an era of modern archaeology in the Senegal Valley. The excavation report includes specialized analyses of pollen, sediments, metals and slags, and animal bone, and the spectacular brass and silver artifacts, which suggest that Sincu Bara functioned as an elite site within the Takrur or Sila polity, are thoroughly described. Pottery did not receive equal attention, however, so the establishment of a basic pottery sequence for the region was relegated to a future time. Other important excavations in the MSV were conducted by Bruno Chavane (1985) in the mid-1970s.

With an increasing internationalization of archaeology in the western Sudan after the late 1970s, new paradigms, perspectives, and methodologies were applied to research on the Sudanic kingdoms. Partly as the result of the participation of Americans familiar with research paradigms in mesoamerica and mesopotamia, a diachronic and regional perspective on the origins and development of complex societies was introduced along with methodologies appropriate to those concerns (McIntosh and McIntosh 1984). The fundamental building blocks of this approach are the establishment of detailed chronological sequences through controlled stratigraphic excavation and a detailed study of domestic pottery assemblages, systematic site survey and surface investigation using