Over the last century, a considerable body of literature has asserted that the site of Niani, in the Republic of Guinea, conceals the ruins of the capital of the medieval kingdom of Mali at the height of its splendour, in the 14 th century AD. This hypothesis was constructed under the influence of a number of factors, but always independently of the archaeological data. This article provides an examination of the archaeological data available for Niani, and concludes that there are no arguments in favour of this hypothesis.
My thanks to Serge Robert for several pieces of information and for figure My thanks for Bertrand Poissonnier for his suggestions, observations on the beads and the calibration of the dates.
My thanks to Nicolas Valdeyron and Bertrand Hirsch for their reading and comments. It is a two-day car journey from the capital, Conakry. The village is located in the far east of Guinea. The watercourse marks the frontier with the neighbouring Republic of Mali. Was it here that a palace was once erected, with a domed audience chamber looking over the parade ground and the Friday Mosque described by Ibn Battuta and another contemporary Arabic author, al-Umari Cuoq, ?
If we walk on the plain, though we can observe a few hillocks with an anthropogenic appearance figure 3 , nothing leads us to think that this was once the site of the political and commercial capital of a great medieval African kingdom. Figure 1 — Niani, Guinea: view of the village from the hill of Nianikoro. The watercourse is the Sankarani, a tributary of the Niger, which here constitutes the frontier with Mali.
Figure 2 — Niani, Guinea: circular houses made of banco and straw. Figure 3 — Niani, Guinea: immediate approaches to the village left of image. It was also discovered a little later that the name of this village is not completely incompatible with that of the capital indicated by the Arabian authors of the 14 th century, to the point that the name of Niani was sometimes substituted, through a kind of philological coup , for the original term in the various editions of these texts for example Cuoq, , While toponymic games are never risk-free, their relevance here is still more diminished by the fact that, in the absence of any accurate geographical location for the kingdom, the area in which we must search the capital is enormous: nothing less than the whole area of the Mandinka language.
