The analysis of ‘social stratification’ (a polite way of speaking of inequality) is crucial to that of political power, though not easy, regardless of the social group investigated. At the cost of oversimplification, die-hard theoreticians minimise the problem by using ready-made typologies and causal links purporting to explain all living social structures.

I describe a people who, in aspiring to Swahili Islamic values and life style, marginalise themselves economically and politically.

The title of the article both indicates the dilemma of their identity and, for scholars who study Swahili-speaking peoples, is deliberately provocative. How can people whom we call or who call themselves Swahili possibly also be referred to as Mijikenda? After all, these two peoples have in other descriptions been shown to be sharply distinct, either as enemies, as traders in separate goods, or as employers and labourers respectively. Yet, though this is a reasonable characterisation of a difference of great historical significance for the creation of Kenyan coast culture and hegemony, it forgets the areas of overlap between the two peoples, where, in effect, it becomes sometimes impossible to posit consistent differences. The sharp delineation conventionally recorded by writers between the Swahili and the Mijikenda (and the various names by which they have been known before the use of this term in 1945) is as much a result of the conditions under which such writings were produced: they unconsciously reproduce the very distinctions by which pre-colonial, colonial and post-colonial authorities administered and ruled the area and its peoples.

Among the Arab tribes of Oman the only acknowledged nobility is that reckoned by tribal ancestry (asl), affirmed by genealogies and oral history, and known by all. This goes not only for the Sultanate of Oman, but also for Omani tribes which have settled in East Africa. One is either a noble or a member of a group of inferior status, that is, the clients (mawâlî) who comprise the labour force in village communities. Slaves ('abid) by definition have no status. This society is thus socially divided between tribal organisations and client groups. On the other hand, within the tribe itself clans or sections appear to be equal or egalitarian, by law and in reality.

As one goes through the numerous publications concerning East Africa, it becomes apparent that certain subjects have rarely been approached, having been neglected or even totally ignored. Such is the case concerning the problems linked with the Hadrami diaspora, the extended study of which—apart from a few notable exceptions—has not yet aroused the interest of historians or the curiosity of anthropologists.

In this article I consider gender as a form of stratification on the coast of East Africa, with specific reference to northern Mafia Island where I have carried out fieldwork between 1965 and 1967, and again in 1976 and 1985. I am interested in the way in which this has changed over the twenty-year period in which I have been observing this area.

At the turn of the century Veblen (1899) wrote on the socio-economic base of the leisure class. He suggested that, in addition to the ‘bourgeois’ and working classes identified in the nineteenth century, there existed an upper strata which was more than a mere anachronistic remnant of the preindustrial aristocracy, but was in fact reproduced by the developed industrial society.