Early in the twentieth century, French and British colonial scholars developed rigid, descent-based models of African pastoral societies. These models emphasized stasis partly because the scholars relied on unrepresentative samples of the pastoralists' views of their own societies, and partly because the scholars simply misinterpreted data. By the 1970s anthropologists had radically revised these models, arguing that although pastoralists generally defined themselves in terms of descent, their societies were nevertheless quite dynamic. In their view, descent was an idiom of social discourse; while pastoral societies may have operated according to the idiom in the past, the economic changes brought about by colonialism had ruptured the connection between the ideology and social practice. More recently, historians have begun to argue that pastoral societies were also dynamic before colonialism, and that there was great flexibility in the ways pastoralists reckoned social identities.

This essay draws on evidence from the nineteenth-century western Sahara to argue that pastoral societies were dynamic long before the colonial period, and that many Saharans perceived their society in this way. This evidence was neglected by the early colonial scholars and many post-colonial anthropologists in favor of those descriptions that emphasized stasis. Saharan accounts that described social dynamism were often based on the explicitly Islamic model of the Prophet Muhammad and his diverse community of supporters. This model, then and throughout Islamic history, has offered the possibility of social improvement, and therein lies the explanation for why some Saharans interpreted society as static while others saw it as dynamic. Social models that fix groups into specific ranks according to descent serve the interests of those at the higher ranks, while dynamic models serve to legitimize social mobility.