L'homme ne peut se réaliser de manière humaine, dans un milieu naturel donné, s'il ne le rationalise d'une façon ou d'une autre. De plus, sa vie en société exige que cette rationalisation, même si elle émane d'un individu céateur, dépasse le cadre d'une conventionnalité purement individuelle—notion à peine concevable—et s'impose, dans la mesure où elle lui convient—c'est là l'origine du mot conventionnel—au milieu social.

The recent appearance of a monograph by a social anthropologist, Peter Lloyd, on The Political Development of Yoruba Kingdoms in the Eighteenth and Nineteenth Centuries is likely to arouse considerable interest among historians of Africa, whose appetites have been whetted by adumbrations of his interpretation in some of his earlier publications. Lloyd traces the political development of the kingdom of Oyo through its period of imperial expansion in the eighteenth century until its collapse in the 1830s, and of five Yoruba states in the nineteenth century—Ibadan, Ado Ekiti, Abeokuta, Iwo, and Ilorin. He seeks to apply to the history of these states a model of the process whereby ‘tribal kingdoms’ develop into ‘highly centralised monarchies’. A ‘tribal kingdom’ is defined as one in which ‘political power…rests with a council of chiefs, each of which is selected by and from among members of a descent group—[and] the king is seen more as an arbiter between the chiefs than as an autocrat’. In a centralized monarchy, on the other hand, power rests with the king, the senior chiefs are appointed by the king, and a concept of ‘citizenship’ develops to replace descent-group loyalties. The Yoruba states discussed in this monograph did not, in fact, develop in this way, and Lloyd's theme is their failure to achieve centralization. The analysis is applied principally to Oyo. Of the nineteenth-century states discussed, relatively little is said of Iwo and Ado Ekiti, while Ibadan, Abeokuta, and Ilorin did not start out as ‘tribal kingdoms’ but as war-camps without kings. Moreover, it is suggested that the failure of Oyo to achieve centralization provided precedents for decentralization which influenced the development of its successors in the nineteenth century.

This paper had its genesis in the recent thoughtful and stimulating article by Robin Horton, on African conversion; the core of his argument is summarized below. Horton's evidence is chiefly drawn from Christian African experience, though he seeks to generalize from it for both Islam and Christianity. My specific purpose here is to look more closely at the phenomenon of Muslim conversion, and to see what light this may throw on Horton's argument. In particular, he has, I think, over-estimated the survival, admittedly in considerably developed forms, of original African elements of religion; and more important, has under-estimated the willingness and ability of Africans to make even rigorous Islam and Christianity their own. My other, general, purpose is to suggest that such Muslim/Christian comparisons may considerably enlarge our perspectives upon black Africa. The comparative approach is extended further, I hope not rashly, with brief considerations of conversion in the ancient world, and of the effects of literacy.

Towards the end of my 1968–9 period of study in Ghana an audience was arranged for me with the late Asantehene, Otumfuo Sir Osei Agyeman Prempeh II. He was ailing at the time and his readiness to see me was largely attributable to his acute interest in the political affairs of Ahafo, the district in which I was working. One of the first questions he asked me was whether I had been told in Ahafo about a certain ‘Asibi Entwi’. I replied that I had heard a great deal about this person. Otumfuo then assured me emphatically: ‘Asibi Entwi is a lie. No such man ever lived.’

This article attempts to carry further the discussion of some aspects of the issues of the analysis of relations between African traditional thought and Western concepts of causation presented by Robin Horton (1967) in earlier articles in this journal. It will also be clear to some readers that its title has been inspired by the theme of Sir Karl Popper's two-volume work, The Open Society and Its Enemies, which first appeared in 1945. In this work Popper makes a systematic attack on modern totalitarians and their predecessors, and in so doing contrasts two types of societies, ‘the open society’, which is the ideal aspired to by liberal democrats, and ‘the closed society’, which he equates with tribalism; and it is to a form of this tribalism, he contends, that the totalitarians would have us return. Popper belongs to a long line of thinkers, including philosophers, lawyers, economists, and sociologists, who have used tribal or so-called primitive man—or rather their conceptions of him—as a means of illuminating their analysis of the problems of society in general and of modern society in particular, and in so doing have often raised the hackles of social anthropologists, who, by virtue of having lived in non-literate societies, tend to consider themselves the profession most likely to know something about the people whom their social-science colleagues cite with deceptive facility. It so happens, I believe, that Popper's conception of tribal or primitive man implied in his picture of ‘the closed society’ is much nearer the mark than, say, Hobbes's picture of man in ‘the state of nature’ or of Rousseau's idea of ‘the noble savage’.