Time and difference in the anthropology of religion

This work is licensed under the Creative Commons | © Richard Fardon. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau6.1.034


Preface to the 2000 Frazer Lecture, “Time and difference in the anthropology of religion” by J. D. Y. Peel

Richard FARDON, SOAS, University of London


It was at the suggestion of his friend Tom McCaskie that J. D. Y. (John) Peel submitted his Frazer lecture to HAU for publication. Tom recalls from their weekly conversations that while John was happy with the thrust of the argument, he was not yet satisfied with its nuance or balance. In the event, John worked on it no further so far as we can tell. In large part, this must have been because he had embarked on refining arguments about the Yoruba that would find their fullest expression in the book he brought to completion fifteen years later, published shortly after his death as Christianity, Islam, and Orisa religion: Three traditions in comparison and interaction (2016, University of California Press). Islam was the last dimension to be added to the comparison partly since its importance was greatest in the areas of Yorubaland with which John was least familiar, and partly because John was most at home in Christian and notably Anglican history. In May 2000, John was awaiting the publication that year of the greatest of his trilogy of books on Yoruba religion, one of the great books about African religion in short, Religious encounter and the making of the Yoruba (2000, Indiana University Press), and he was turning his attention to the challenges a three-sided comparison—the classic problematic of the “triple African heritage”—would pose his scholarly grasp of Yoruba history.

As a historian and sociologist sojourning long in a Department of Anthropology, both insider and outsider, John was by turns both stimulated and exasperated. He accepted that an ethnographic method was essential to comparison, but he was frustrated by anthropological theorizing that either reinvented the [542]historical-sociological wheel or apparently ignored or overlooked the fact that wheels had a history. This lecture is valuable for retaining more of his amusedly spikey guying of anthropological interlocutors than would have been likely to survive his editing. Some of the usual suspects make their appearances and are dismissed.

Fundamentally, John felt that anthropologists were poor at temporality in a number of senses: when locating their own researches, the lives of those they met, the sources they used, their own notes, when delineating what they meant by context, what it meant to their subjects, and where it came from, and most germane here, in recognizing the historical trajectories imparted to religions by their histories, discourses and practices. In short, anthropologists for all they wrote about it were practically poor when describing the consequences of humans being beings in time. While John might not have published his thoughts in quite so forthright a manner, I cannot imagine he would have any objection to our doing so posthumously given he was never shy in letting us know his opinions.

I am grateful to Tom McCaskie for advice and to Karin Barber for tidying up the Yoruba to a scholarly halfway house appropriate to a general publication which does not show tones but does use subscripts to denote phonetic values.


Richard FARDON is Professor of West African Anthropology in the Department of Anthropology and Sociology at SOAS, University of London, where he moved in 1988. He was joined by J. D. Y. Peel in 1989, and together they taught courses on West Africa and Social Theory, their mutual interests, and also shared a variety of administrative responsibilities in the course of which they usually discovered more interesting things to talk about.

Richard Fardon
Department of Anthropology and Sociology
SOAS, University of London
Thornhaugh Street, Russell Square
London WC1H 0XG
United Kingdom


This work is licensed under the Creative Commons | © J. D. Y. Peel. ISSN 2049-1115 (Online). DOI: http://dx.doi.org/10.14318/hau6.1.034


Time and difference in the anthropology of religion

The Frazer Lecture, Oxford, 9 May 2000

J. D. Y. PEEL, SOAS, University of London

J. D. Y. Peel’s Frazer Lecture of 2000, published here posthumously, presented his early thoughts about the three-sided comparison that would culminate his trilogy of works on Yoruba religion. Working through these arguments would occupy another decade and a half until the publication of Christianity, Islam, and Orisa religion: Three traditions in comparison and interaction (2016, University of California Press). As a historian and sociologist, John was by turns stimulated and exasperated by anthropologists. An ethnographic method was essential to comparison he accepted, but anthropologists were poor at temporality in a number of senses: when locating their own researches, the lives of those they met, the sources they used, their own notes; and when delineating what they meant by context, what it meant to their subjects, and where it came from; and most germane here, in recognizing the historical trajectories imparted to religions by their histories, discourses and practices. In short, for all they wrote a deal about it, anthropologists were practically poor when describing the consequences of humans being beings in time. The lecture proposes solutions to these lacks.

Keywords: comparison, history, time, religion, Yoruba, Christianity, Islam

I feel I should begin with something of an apology for the rather vague and yet grandiose title I have given this lecture. I was pressed for a title some months ago, when I was still uncertain of the precise focus I wanted to give it. Though the title is not really misleading, my experience has been like that of a traveller who only realizes where he is going when he arrives at his destination. As you will shortly see, my best title might have been “The value of going back for the anthropology of religion.” But still, it is a great honor to have been invited to deliver this Frazer Lecture, and I hope it is auspicious that my subject of this evening has some precedents in the topics chosen by some notable Frazer lecturers of the past. I have in [534]mind Meyer Fortes’ “Oedipus and job in West African religion” (Glasgow, 1956),1 Marshall Sahlins’ “Captain Cook: or, The dying god” (Liverpool, 1982),2 and Robin Horton’s “Back to Frazer?” (Cambridge, 1987).3 For I will follow them in speaking of West Africa, of the moral and cosmological bearings of religion, of Europe’s impact on the non-European world, and of the relations between anthropology and history.

These commemorative lectures are intended, or presumably were when originally endowed, to evoke expressions of pietas. While at Liverpool, I had the opportunity of listening to a Frazer Lecture every fourth year; and over the years I have also heard four or five of the other premier series, the Malinowski lectures, given at the LSE. Reading further back, one has to be struck by the contrast in how the lecturers have tended to treat their respective commemorands. Frazer may be an ancestor, but for many of his lecturers his example is a somewhat embarrassing one, rather as if he was a batty old grandparent who once a year has to be brought down from the attic for his birthday party. Malinowski, by contrast, who actually died three years before Frazer, continues, more than half a century later, to command warm expressions of filial devotion. The reason is not mysterious; Malinowski established the research practices—doing fieldwork, writing ethnography—by which modern social anthropologists know themselves as practising their subject. In recent years, however, judgments have got more nuanced. Marilyn Strathern surprised us in her 1986 Frazer Lecture at Liverpool in discerning postmodernist qualities in Frazer’s texts.4 Her aim was to use the transition from the Victorian mode of anthropology—which aimed to reconstruct “the early history of mankind”—to the “modernist” mode—marked by holism and synchrony—to throw light on the putative transition of the last fifteen years: from “modernism,” the more-or-less established way of doing anthropology, to a new “post-modernist” way whose principal charter was Clifford and Marcus’ collection of papers, Writing culture (1986). As Strathern made very clear, these two transitions are hardly of equivalent weight. The former had solid and lasting achievements to its credit, new ways of doing and writing anthropology, based on a distinctive view of the relations between the ethnographer, his subject and his audience. Postmodernism set out, above all, to challenge and unsettle those relations. But beyond that, its own achievements have been extremely slight: more than a decade after she made it Strathern’s remark that “there is more talk about what postmodernism might be than examples of it” still rings true. In its insubstantiality, postmodernism might be compared to a hauka spirit that likes to possess research students before they do fieldwork; for afterwards its symptoms rarely recur, unless they fall under the spell of one of the well-known American shamans of the cult, like Michael Taussig or Paul Stoller.

If social anthropology has been more unsettled by postmodernism than most other subjects, the reason is perhaps that its history shows it to have been so [535]protean, with its view of its subject matter changing more radically over the past century than any other social science. But at least since Malinowski it has been able to hold to one defining feature, basing its distinctiveness less on its subject-matter than on its special method of inquiry: namely, the personal interaction of the anthropologist as human subject with other subjects in the immediate present of the field situation. This has stamped anthropological thought with a pervasive atemporalism, of which one token has been a strong penchant for spatial or co-temporal modes of representation and explanation, for privileging diagrams over narratives. Going with this has been a heavy reliance on one principal strategy of explanation and interpretation: by placing phenomena in their context, relating them to other, coexistent, features of the setting in which they were observed. The contributors to Roy Dilley’s recent collection of essays, The problem of context (1999), subject this strategy to several searching critiques. But though they variously respond to problems arising from the strategy of contextualization by stretching the notion of context from its primary meaning—which is an external, synchronic, objective setting—to something inward, or more to do with the analyst’s perspective than the features of the object under analysis, they pull back from challenging its most problematic feature: its presentist or atemporal character.

For in every context, there is what reaches out of the context, into the past. The only kind of consciousness human subjects have is a historical consciousness, that is one of being-in-time, and it is an essential condition of their capacity to be agents. While this is grounded in personal memory, it connects and merges in various ways with the collective memory sustained by the social groups to which individuals belong, and this largely explains variations in its range, content and significance. The conditions of doing fieldwork, especially for the first time, encourage us to forget that we always enter in medias res, that we only have direct access to the smallest temporal cross-section of our subjects’ lives, and that their actions are informed by a knowledge of the past, their own and their community’s, which is not open to our observation. So stated, this all seems pretty obvious; but it has not proved easy to address the implications. One principal reason for this has been the tendency, ever since anthropology in the 1950s first started to feel it had to do something about its ahistorism, for two lines of approach to get disconnected from one another. The first treated the past as real and different: it first took the form of studies of social change in late-colonial settings, then of anthropological studies of the past, finally of what might be called “world-historical anthropology,” notably practiced by Eric Wolf (1982) and Ernest Gellner (1988). The other was about the past as represented, history as a subjective reality.

Over the past decade, memory and representations of the past have become popular topics—the interest in “heritage” and museums, the debates about Holocaust denial and so-called False Memory Syndrome—not just in anthropology but throughout the human sciences. As a field of inquiry it still faces the same dilemma that the rigorous presentism of Malinowski and Radcliffe-Brown resolved for a long generation of social anthropologists. Are we justified analytically in treating memory and tradition, though they refer to the past, as other than facts of the present? Memory is frail and fallible, tradition is manipulable and often invented, so how can we treat them as mediating the influence of a factual past on the present? No sooner had anthropology put social change firmly on its research agenda than [536]it was swept by a new ahistorical paradigm, in the form of Lévi-Strauss’ structuralism. This urged on us that oral narratives were better used to give access to cultural logics than to any past which they seemed to represent. Jan Vansina’s attempt to recover the early history of the Bantu colonization of central Africa from their legends of heroes and migrations seemed less persuasive than the tour de force of Luc de Heusch’s in using them to decode the categorical foundations of Bantu belief.5 Few works were more generally cited in the 1970s and 1980s than Hobsbawm and Ranger’s Invention of tradition (1983)—a sign that historians were now learning a lesson which anthropologists had never forgotten. And as if to climax this vigorous fight-back of the present over the past, historians started to adopt the spatial metaphors beloved by anthropologists.6 David Lowenthal (1985) entitled his study of the representation of the past in heritage sites, buildings, museums, landscapes, and artifacts The past is a foreign country, shifting the past from its actual otherness in time to an otherness in space, as if it were somewhere that one might go and do fieldwork in. Then there is the concept of lieu de mémoire, “site of memory,” used by Pierre Nora and his collaborators in his great multi-volume study of representations of the French past.7 Nora, significantly, believes that modernity has brought about a profound rupture between past and present: the past no longer flows spontaneously into the present, forming identities and practices along received lines, but has to be deliberately fostered in the form of memories attached to situated objects. Yet though Nora’s argument is premised upon the idea of a rupture from the past—that indeed establishes its affinity with the “modernism” of an anthropology along Malinowskian lines—his data are still eloquent of a human need to establish some sort of relationship with a presumptively real past.

Nora acknowledges the influence of Maurice Halbwachs, and the case treated in Halbwachs’s essay on the sacred topography of the Holy Land illustrates well the point the point I want to make.8 In the early fourth century when the old spontaneous traditions of the minority, persecuted church no longer sufficed for its new mass membership and official status, the empress Helena sought to create mnemonic pegs for the sacred narratives by locating sites where the major incidents of the Gospels took place. It was, if you like, a large-scale exercise in the classical Art of memory (Yates 1966). The fact that most of these sites may have been bogus does not derogate from the fact that it secured a critical transmission of belief, identity, and practice across a major divide in Christian history.[537]

But to return to the two lines of approach. . . . If we cut off the analysis of memory and representations of the past from any concern for what the past was, locating them entirely in the present, we lose the means to address the work of representation itself, because we cannot ask about the relationship between the representation and the represented. The temptation to do this is much greater if we regard social change in strongly discontinuist terms—that is as a sequence of ruptures—or if we identify history simply with social change. It does involve social changes, of course, but it involves continuities no less: it is the mix or balance of these which is the stuff of things. On the other hand, if we try to study social change without attending to the role of history as a subjective reality, then we effectively exclude human agency from it: social change becomes something driven by inherent mechanisms below the level of consciousness, or a naturalization of history. So ethnography, qua the contextual analysis of the present in which fieldwork was done, is not complete in itself; it calls for history to complete it.


Religion is par excellence the institutional site where we may explore the effects of being-in-time and the operation of “the past in the present” in social life.9 To a greater extent than political and economic phenomena, religions don’t merely have a past whose traces they bear, but speak expressly of the past. In most societies their own credit derives from the messages they bring from the beyond; so their anthropological analysis is less plausibly limited to the fuller explication of local context than that of almost any other institution. I say “from the beyond” since the otherness of origin of religions can be represented in both spatial and temporal terms—often in a blend of both, for which the most persuasive image is that of a journey, a movement at once through time and space. World religions above all, for nearly all their adherents, are imports from elsewhere, with their most sacred centres external and distant. And even when they have become well domesticated, it is as vehicles of messages from the past—inscribed in canonical texts or hallowed practices or collective memories—that they continually shape the present. At the same time and as a corollary of this, they continuously generate a specific alienating effect, a sense of disjunction between present realities and ancient foundational ideals. Thus they keep alive the option of reaching back into their past for revival and reformation.

Now strongly context-oriented anthropologists are uncomfortable with the trans-temporal influence that is implied here. It smacks of essentialism, which as an anthropological thought-crime now almost ranks up there with ethnocentrism. But I think we worry ourselves too much here. Consider Talal Asad’s (1993: Chapter 1) critique of Geertz’s well-known definition of religion (Geertz [1966] 1973: Chapter 4), essentially on the grounds that it raises to universal status what is a particularly modern Christian view of religion. He drives his point home by stressing [538]how different it is from medieval Catholicism, with its disciplines to create a regime of religious truth. While I find his historical relativization of Geertz’s definition cogent enough, it does leave us to view the history of Christianity simply in terms of its discontinuities: that is to say, modern Christianity is merely contrasted with that of an earlier period, along the lines of Foucault’s so-called “history of the present.” Asad’s choice of this strategy stands out the more starkly because in a book whose subtitle is Discipline and reasons of power in Christianity and Islam, he makes no attempt whatever to compare the two religions. That would inevitably have suggested a very different line of analysis, toward an analysis of religions as world-historical forces, perhaps along Weberian lines. Unfortunately, there is no development of a suggestion he had made in an earlier essay on the anthropology of Islam, that Islam might be viewed as a “discursive tradition,” referring to the ways in which Muslim subjects draw on their specific past as they respond to new contingencies (Asad 1986). But without something of that kind there is nothing to counteract the appeal of an extreme discontinuism à la Foucault.

Anthropology’s love affair with Foucault has, I suspect, one of its sources in its own attachment to contextual explanation. This proceeds by the establishment of links of synchronic implication between entities coexisting at a particular point of time or period. The tighter and more complete is the whole set of implications, the more satisfactory the explanation is felt to be. Contradictions are not excluded, but they must be of the system, not arising from some tension or lack of fit between it and factors continuing into it from the logical and temporal beyond. In case you may think I am trying to resurrect the nineteenth-century notion of a “survival”—something to be avoided, even in a Frazer Lecture—I emphasize that what I have in mind has quite a different logical status. Survivals are the residues of earlier synchronic systems, which they may be used to reconstruct. So in fact they too imply a fairly discontinuist view of social change: as a succession of discrete contexts, each with its quite distinctly unified character, each one emergent from past influences which no longer produce effects on the continuing present.10

To come closer to the main subject of this lecture, I may illustrate what is missing from two major contributions to the study of African religious change: Robin Horton’s (1971, 1975) essays on “African conversion,” and the Comaroffs’ (1991) study of the role of Christianity in cultural change in South Africa. The point is the more telling because these two bodies of work are so different in other ways. Horton’s theory ignores power relations within the context of conversion, and makes colonialism entirely incidental to his explanation of why it came about; whereas both these are absolutely central to the Comaroffs’ account. But both posit the closest implications between the ideational and the sociological features of the context, and exclude non-contextual influences. For Horton, missionaries are merely “catalysts” of changes that are sufficiently determined by the key features of the context, so the Muslim or Christian baggage they brought with them is treated as having no historical consequences. This criticism was early made by Humphrey Fisher, a historian of Islam (see Fisher 1973); but I think it fair to say that it has been very little taken up, at least by anthropologists. Yet should we not see the context of [539]conversion as preeminently a site where two traditions have converged: a local one, and a highly articulate one from another time and place?11

The Comaroffs do not ignore missionaries, but limit their historical significance to their mediation to their converts of the capitalist civilization from which they arose. “The goods and messages they brought with them to Africa presupposed the messages and meanings they proclaimed in the pulpit and vice versa. Both were vehicles of a moral economy that celebrated the global spirit of commerce, the commodity, and the imperial marketplace.” (Comaroff and Comaroff 1991: 8–9). But we have to ask, is this all the missionaries were about? To make the point with one example, does the redemptive sacrifice of Christ—a long premodern idea which stood at the heart of evangelical preaching—imply capitalist timekeeping and vice versa? This teaching was among many messages transmitted by missionaries from their religious past which stood in no determinate relationship to the colonial context in which they were enunciated, were taken up by local agents following agendas of their own, and produced historical effects which long outlasted the contexts in which they were introduced. World religions are seriously about the cultural longue durée.


In the rest of the lecture, I want to present three episodes in the development of Yoruba religion over the last 150 years. The first, centred around the 1850s and 1860s deals with the triangular engagement of Christianity, Islam, and the indigenous religion. The second deals with some religious aspects of the construction of the Yoruba as an ethnographic category by a cluster of Yoruba pastors around the 1890s. The third is the contemporary wave of religious enthusiasm, known variously as Charismatic, neo-Pentecostal, or Born-again. My aim is not to give a potted religious history, but to show how, at each stage, history was made through a complex interplay between messages received from the past and critical reflection upon them.12

Let me begin by citing two maxims from different points in the sequence, in which Yoruba themselves reflect on this experience of religious diversity and change: a modern one which looks back, and one from the nineteenth century which looks forward. The first, of uncertain origin, has circulated widely in the twentieth century:

“We met Ifa in the world, we met Islam in the world, it was noon before Christianity came to the house.”

The implication is that the novelty of Christianity long presented a problem for its reception. It seems to echo a concern expressed in the second maxim. This comes from the mouth of a small-town chief in 1878, alarmed at a vehement sermon on the Last Judgment given by an African evangelist:

“Let the Ifa man worship his Ifa: let the orisha man worship his orisha; and let the slave follow his Shango priestcraft for his food.”[540]

The linking of slaves with the thunder-god Sango indicates the recent arrival of that cult, its vectors being slaves from its northern Yoruba heartland. It also discloses the terms on which new cults were accepted: by offering religious services which people were willing to pay for. With its active admonition to religious live-and-let-live, it rests on a lucid perception of the problems that the world religions might pose for the harmony of Yoruba communities, and uncannily anticipates problems that have come right to the fore only in recent years.

For the comparative student of religion this is a situation of unusual interest: we can compare Islam and Christianity in a context where they are already being compared by the Yoruba themselves. What this shows up at once is the contrasting historicity of each religion: by this I mean the principles of each religion’s self-realization over time, laid down most definitely in their earliest decades and centuries but always subject to later additions and modifications to produce a cumulative sense of its collective being-in-time. Nineteenth-century Yorubaland was frontier territory for both of them, and there comes to mind a remark of Braudel’s which Wendy James and Douglas Johnson quote in their introduction to African Christianities (1988: 5): that it is “most often on the border that the most characteristic aspects [and] phenomena [of a civilization] can be found.” If we can unpack the implications of this remark, it suggests that religions tend to assume their most distinctive form when they are at the edge of their range, set in antithesis to another way of being. On the other hand, competition between two religions, especially when the criteria of choice are independent of either, may push them toward a common ground proper to neither. We see these two possible outcomes constantly working against one another in nineteenth-century Yoruba.

In their earliest relations with one another, Christianity had stood to Islam as precursor, a religion encompassed in thought and substantially in practice too, as Islam overran Christian lands and drew most of its earliest converts from it. In their chronological asymmetry, Islam was manifest to Christianity as heresy and threat, requiring to be refuted and repulsed. By the time that the two religions came face to face in nineteenth-century Yorubaland, these old mutual perceptions (not forgotten by their clerical professionals) had been overlaid by two paradoxical reversals. Now Christianity had become intimately associated with Western Europe’s rise to world hegemony through applied science and capitalism; and Christianity found itself to be following in the footsteps of Islam as its monotheist precursor.

The Yoruba themselves were more directly affected by another contrast between Christianity and Islam, which pivoted on language and strategies for conversion. Again they derived from links between culture and power in the religions’ early years. Graduating to an imperial ideology within a few years of its founder’s death, Islam fused religion with civilization of the basis of its language of revelation, Arabic. By contrast, Christianity took three centuries on the path to political power, so was unable to fuse religion and civilization, which remained essentially pagan. Without the means of secular power, the diffusion of the Word depended on its translation into the vernacular languages of the late classical world: from the spoken Aramaic into first the common Greek of the day, and then into Latin, Syriac, Armenian, Gothic . . . and where these ancient tongues led, Yoruba, Setswana, and Luganda would in the fullness of time follow, with far-reaching cultural consequences.[541]

The project of Yoruba translation had the rare advantage that it was initiated by a Yoruba, the Rev. (later Bishop) Samuel Crowther; but it still involved some rather difficult choices. It has been suggested that he looked to find “deep Yoruba” equivalents of Christian terms and drew especially on the advice of oriṣa priests. But if we focus on the terms that he actually used in his translations of the Bible, we find he was much more drawn to take maximum advantage of the monotheistic bridgehead that Islam had already created. Some of this vocabulary—mostly from Arabic mediated through Songhay—had already by the mid-nineteenth-century passed into the religiously unmarked vocabulary of Yoruba: words such as aanu (“mercy, regret”) or alaafia (“peace, well-being”). The words for prayer, preaching, saint, pagan, evil, and priest were all borrowed from this Islamic lexicon. The last of these, alufa, originally meant “Koranic teacher,” and Crowther used it in preference to the indigenous term aworo, which meant a priest or medium in the pagan cults. In this he seems to have followed the prompting of ordinary Yoruba, who often greeted Christian missionaries as “alufa.” On one occasion a Yoruba evangelist called Macaulay was even greeted—by an eegun, or masked ancestral spirit!—with the words As-salaam alaykum. Sometimes the prior assimilation of Muslim and pagan conceptions created insuperable problems of translation. If one asks why Crowther introduced a neologism from English, angẹli, to render “angel,” rather than make use of the Islamic maleka, a word which started its career in Yoruba as an Arabic-derived neologism, the answer almost certainly is that maleka had by the nineteenth century come to be widely used as a synonym of oriṣa.

So what Christianity met in Yorubaland was not a static “traditional religion” but a local religion already starting to move in the direction where missionaries wanted to take it further. We find more support for this from variations in the religious beliefs encountered by missionaries in different parts of the country, especially between two broad zones: the savannah zone to the north and west, where Islam had been long present; and the forest zone of the south and east, from which Islam was absent until after the first missionary visits in the 1860s or later. Hypothetically we may treat the south-east as an Ur-Yoruba baseline, and use regional differences to suggest a model of historical development. The differences fall in two distinct areas.

The first has to do with concepts of human destiny and time. In the north and west (where Islam was present), the missionaries found two barely compatible sets of belief. One held that after death human beings (at least those who had attained social maturity) would be reborn in one of their descendants. The other—often described by missionaries in terms like “vague beliefs” or “some notion”—held that there was some kind of divine judgement after death, which would send people to one or the other of two heavens, the “good heaven” or the “heaven of potsherds.” In the south and east, conceptions of this latter kind were not found, and the emphasis on the rebirth of the dead was exclusive. So, in the area where Islam was present, the conception of human life as a unique, one-way process which would determine one’s eternal destiny had begun to circulate—to be seized on for further development by Christian missionaries.

The second has to do with the concept of God. It is nowadays the established consensus amongst Yoruba, authoritatively stated in the Yoruba dictionaries, that God, the supreme being associated with the sky, does not belong to the class of [542]oriṣa. But when we turn back to the nineteenth century, a much more complicated picture emerges. The journals of Yoruba evangelists to the east and south (who themselves all came from the north-west area) often noted with surprise and disapproval that, as one of them put it, “there were many people who could not distinguish between the word oriṣa (idol) and God the Supreme Being.” Moreover, several deities regarded in the north and west as oriṣa were in the east and south treated as aspects or refractions of the Supreme Being. These included Ọramfẹ, a deity of thunder; Ọlọjọ, “lord of the day”; Ọbatala or Oriṣaala, who created humanity; and Oduduwa, regarded as the ancestor of the human race (latterly of the Yoruba nation), who was also widely represented as female. If this complex set of attributions was anything like the original Yoruba theology—without sharp cut-offs between God and the cosmos or humanity, this God being ancestor as well as creator, female as well as male—then the influence of Islam makes it intelligible why the God of the north and west should have been simplified as He was: to a Creator radically set off from the oriṣa, and particularly divorced from any idea that He might be consubstantial with humanity through being its ancestor. This was a conception that Christians were very happy to pick up and run with; and it was one that Yoruba Christians in particular were anxious to regard as immemorially traditional. They might take a historical view of their religion, but not of their God.

But although Christianity thus borrowed from and built upon the changes introduced into the Yoruba religious repertory by its monotheist precursor, it still needed to differentiate itself from Islam. Moreover, since the two monotheisms were rivals before a pagan court of opinion, Christianity had now to find a way of appealing to those whom it had colluded with Islam in disparaging as idolators and polytheists. A complex three-way negotiation of differences was called for. Negotiation takes place in contexts, but it is conducted in the light of strategies, which imply temporal schemas by which the deposits of the past might be converted into desired future outcomes. Missionaries sought to criticize Muslims on their own ground for failing as bearers of a monotheistic religion of the Book: they could recite Arabic but not understand it, they used Koranic texts mainly to make charms and to divine, they even attended pagan rituals. But while this might discomfort Muslim alufa, it did little to strengthen the Christian case in the eyes of pagans. A subtler and more effective strategy took its rise from one of the age-old staples of inter-faith polemic, exploiting one consequence of the historical asymmetry of Christianity and Islam, which rendered them respectively the precursor, and a heresy, to the other. For whereas Mohammed was wolii eke (“a false prophet”) to Christians, to Muslims Christ was Anabi Yisa (“the Prophet Jesus”): someone whose divinity and crucifixion were denied, but whose virgin birth and divine inspiration were conceded. This enabled the figure of Christ to be assimilated to the role that Yoruba pagans ascribed to their oriṣa, mediators who transmitted blessings whose ultimate source was God in heaven.


I now come to the second episode, in the 1890s, when missionaries and African Christians were able to look back at the first fifty years of their project. In its initial assault on “heathenism,” Christianity had been characterized by the Yoruba as [543]doubly strange: as the religion of the Oyinbo (“Europeans”), and as an iconoclastic monotheism like Islam. Toward Islam a two-track strategy evolved: exploit the monotheistic bridgehead Islam had created within Yoruba culture; then differentiate itself, as offering a superior level of individual and collective empowerment through its links with European modernity. But this was still too external; the missionaries (and especially African ones) wanted to see Christianity fully integrated into Yoruba culture. The trouble was that the much-denounced heathenism was firmly ensconced at its heart. But here also there were two ways to go at it. The first was what we might call instrumental secularization: that is, to make the culture safe for Christianity by trying to extract the heathenism from it. Since Yoruba did not recognize religion as a discrete sphere of activity, this meant their “religion” had to be discursively constructed. (Similar things were going on all over the mission field at this period; and Frazer, of course, was a great beneficiary of it, through his network of missionary correspondents).13 But a more challenging step was envisaged by a number of the Yoruba clergy: to present Christianity, not just as a rebuttal of paganism but as its fulfilment, a simultaneous negation and completion.

Now the fruits of this ambition are to be found in a literary and cultural movement which flourished among educated Yoruba Christians between the late 1880s and the First World War. Whilst it took many forms, including an enthusiasm to re-Africanize personal names and dress-styles, its main monuments were the first major published studies of Yoruba history and religion. It is no exaggeration to say that Yoruba studies, as a scholarly pursuit, began here, and that, to an extent equalled by no other African people, the basis of their later history and ethnography was laid by insiders. Indeed, so crucial was this body of work in creating the Yoruba as an ethnographic object and so pervasive are some of its assumptions to the contemporary self-image of the Yoruba, that its proper understanding is a vital part of the critical equipment of any would-be ethnographer of the Yoruba today. One continuing obstacle in the way of our doing this is encapsulated in the very name by which the movement has come to be known in recent years, namely “cultural nationalism,” which interprets the movement as a cultural anticipation of the political nationalism of fifty years later. It also employs a strongly contextual analysis: the movement is seen as a direct, reactive response to the imperialist attitudes and racial discrimination which accompanied the imposition of colonial rule on the Yoruba in the late 1880s and 1890s. There is surely some value in this interpretation, but it ignores the way that the movement was a natural extension of the strategies of evangelization that had been hammered out on the anvil of a myriad of religious encounters over the preceding fifty years. This was less a reassertion of tradition in the face of Christian evangelism, than a reconstitution of it in order to ease the task of making Christianity, as the Revd. E. M. Lijadu (one of the most prominent cultural nationalists) put it, “the home religion of our nation.”[544]

To this end, five distinct strategies were used, all of which can be seen in journal accounts of evangelism before they appear in such key texts of cultural nationalism as Samuel Johnson’s History of the Yorubas, or James Johnson’s Yoruba heathenism:

  1. Prophecy: It was common for evangelists to hear prophecies—typically from the babalawo or priests of Ifa—predicting such things as the coming of the white man or even that people should become Muslims or Christians. Ifa worked by offering mythical precedents for courses of action that its clients felt disposed to take: it legitimated moving with the times. But by thus accepting the predictive power of Ifa as genuine (as most Christians certainly did), Ifa itself was magnified. So this leads to:
  2. Anticipation: As I have described elsewhere, missionaries were impressed and fascinated by babalawo, the intellectuals of traditional society. As a result, they effectively colluded with their version of what tradition was, and particularly with their aggrandisement of their particular oriṣa, Ọrunmila, whom they treated as an especial confidant of God and the mediator supreme between heaven and earth. But this suited evangelists too: for it created a niche in the Yoruba pantheon which approximated to Christ; which they might then exploit to argue that Christ was the fulfilment of a potential in Yoruba religion. E. M. Lijadu even argued that in Ọrunmila “our ancestors had a foreknowledge of Jesus Christ.”
  3. “Back to the Future”: This was the argument that the truly ancient Yoruba religion was a monotheism which became corrupted by the introduction of idols requiring sacrifice. This was argued on the basis of a comparative religion, based first on casual parallels between various details of Yoruba ritual practice and those of the ancient Hebrews, then stated systematically in James Johnson’s Yoruba heathenism. This was a neat way of turning round a standard response to Christian preaching, that “we cannot desert the religion of our forefathers.”
  4. Religious evolution: Again this drew its evidence from comparative religion, but of a more modern kind. Johnson argues that in “heathenism” the Yoruba share a generic type of religion, common to many peoples at many other times and places, from which advance can be expected. The advancement of Yoruba heathenism over several neighbouring varieties shows that in their case the journey to higher enlightenment has already begun, and with Christianity will be taken further.
  5. Euhemerism: The view that the gods were once human, raised after their death to divine status for their great deeds, arose spontaneously among African missionaries. Why them, and not Europeans? It enabled them to challenge the idolatrous status of oriṣa, without denying that they represented something real and valued—for they often figured in “secular” forms like genealogies and foundation narratives. So it was for their own social identity, and not just to reduce the area of unhelpful disagreement with pagans, that they embraced this way of retaining the ancestral, heroic or royal, if no longer the divine, character of oriṣa.

So Yoruba religion came to be represented for the first time as a totality: that is, as an entity homologous to the world religions, constructed so as to ease its supersession [545]by Christianity. Within this novel totality, its elements were given new, differentiated values:

  1. Ọrunmila, the oriṣa of Ifa, became an anticipation of Christ as saviour and mediator, while the oral corpus of divination verses, which pastors like Lijadu had started to collect and publish, became “the Yoruba Bible,” or at least imagined as the Yoruba cultural archive. When I was editor of Africa in the 1980s, a characteristic type of article was submitted by Yoruba scholars, which combined a presentation of fieldwork data on some subject, linked to the citation of selected Ifa verses to give authority to their interpretation.
  2. The contrastive fate of Oduduwa, under the imperative of euhemerism, has been even more surprising. From being God in the aspect of ancestor and usually gendered female, She has been become “the chief of the founders of the Yoruba nation” as Samuel Johnson put it in his History: rather a come-down, to be metamorphosed from a Great Goddess to something like George Washington in barely more than a century! The cultural society founded in London in 1947, which led to the main nationalist party among the Yoruba, was called “of the Sons of Oduduwa,” and the political grouping which over the past year has raised the question of whether the Yoruba should remain part of Nigeria is called the Odua Peoples Congress.


For my last episode I want to fast-forward over a century, to the 1980s and 1990s. Colonialism has come and gone, as least in its political form; and nearly all Yoruba have become Christians or Muslims. One might say we have entered a post-nationalist age, insofar as one of its key features is a radical discontent with nationalist Nigeria and its cultural representations. After some years’ absence, during which I was (so to speak) doing fieldwork in late nineteenth-century Yorubaland, I returned to Nigeria in 1994, thirty years to the month after I had first been there. The religious atmosphere was strikingly different—indeed the social saliency of religion (that is, of Islam and Christianity) was much greater than previously. Most prominent of all was the boom of neo-pentecostal or “born-again” Christianity, which was vastly outstripping the older-style independent churches that I had first gone to study in the early years of independence.

It was as if a pendulum of cultural change that for nearly a century had swung Christianity in the direction of Africanization, of localizing a world religion, now started to swing the other way. Born-again leaders are very aware of belonging to the most dynamic international movement of world Christianity; they travel a lot (particularly to the USA), and their adept use of electronic media means that ideas circulate widely and rapidly. The whole style of born-again religion is oriented to international modernity: in dress and musical style, in a preference for the use of English over Yoruba, and American English at that. One thing I found especially grating was a penchant for saying Ay-men, not Aa-men or A-meen at the end of their prayers! Despite strong continuities in the demands placed on religion at the personal level, the older Aladura churches, with their Yoruba-style music, their white prayer gowns, and their use of local expressive symbolism (like water for healing), are mocked and their leaders accused of compromising with idolatry![546]

Throughout the twentieth-century, the cultural trajectories between Christianity and Islam in West Africa seemed to run in opposite directions: the one seeking to Africanize itself, the other seeking to validate itself by the standards of a more international Islam. Now the two trajectories run essentially in parallel, each oriented to what it takes to be the most prestigious global centres of their religion. A century and a half ago, what is now Nigeria stood on the far frontiers of Islam and Christianity; now it sits astride a major fault-line between them, and outbreaks of religious conflict have become endemic. What for long held conflict at bay among Yoruba was precisely the old cultural matrix in which both had to win converts: as the old Ijẹbu chief said to the vehement James Johnson in 1878, we each get on with our own devotions, and allow others to get on with theirs. But though the oriṣa cults are deep in decline, practically marginal to the lives of most people (especially the young), they do still live vividly in their representations. I don’t here mean their representation by academics for whom they are of the essence of Yoruba culture, but by those who are most opposed to them. Born-again Christians, and likewise Muslim rigorists, need icons of moral alterity, they need their demons. In a collaborative study I am involved with on religion and the contemporary media, set in Agbowo, a sprawling new suburb of 40,000 people north of Ibadan, our survey turned up some 178 churches and over 28 mosques, but no public shrines or signs of oriṣa worship at all—save for a handful of “herbalists” (who might also have done some divination). Yet in video material and TV programmes of a Christian or Muslim character, the pagan or occult world was well in evidence; and “deliverance” from the ill effects of demons or ancestral spirits is a big demand on the pastors of born-again churches.14

This new wave of religious feeling and practice is sustained, indeed it vitally depends on, narratives which rework materials received from the past—memories, antecedents, traditions, exemplars—in the light of current needs and aspirations. Pentecostal history takes two principal forms: one places them in a Christian or sacred, the other in a Nigerian, history. The cardinal experience of Pentecostalism—being “born again”—has as its diagnostic sign to speak with tongues. The prototype of this experience is found in the account in Acts 2 of how the Apostles were enabled to speak to the heteroglot mass of visitors to Jerusalem so that “every man heard them in his own language.” It is a story of the weak made strong and of the mandate of mission. Acts 2 celebrates a temporary remission of the post-Babel state of humanity. Earlier we saw it was a key missionary text, justifying the imperative of translation; now it is the charter for what is taken as an empowering kind of super-language. But the Born-again movement recognizes its relationship both with its missionary ancestors, mainly British, and with the tradition of Protestant revivalism, mainly American—which themselves were, of course, closely cognate. The Born-agains have, in effect, renewed the modernist challenge to African institutions presented a century ago by the great Evangelical missionary societies. Their [547]attitude to the mainline churches—Anglican, Methodist etc.—founded by those missions, is notably less hostile than to the Aladura churches. They are charged with having become “dry” or “dead” in their faith—the language here is reminiscent of what was used by missionaries of their complacent congregations in the nineteenth century—but they are not charged with sliding back to idolatry. Increasingly, there are born-again tendencies within the mainline churches; and there is always the possibility of revival.

The other Pentecostal history treats of Nigeria. Pentecostalism’s promise of being born again properly applies to individuals, though it can be hoped that many regenerate individuals will together renew their society. In some other African countries where it has been won many adherents, such as Kenya, Ghana, and Zambia, Pentecostalism has not yielded much of a political reading—or if it has, it has been supportive of the status quo (especially where the head of state claims to be born again himself). But in Nigeria born-again language was strongly linked with Yoruba opposition to the military dictatorship which only ended last year (1999) with the election of President Ọbasanjọ. Their story goes that Nigeria is a land spoiled, its God-given bounty squandered by those who, through force and fraud, have ruled since independence. And the struggle to redeem Nigeria is a spiritual struggle, waged against principalities and powers. In the born-again view of Nigerian history, certain dates stand out as drawing the country away from its proper destiny. One of these is 1804, the date of the foundation of the Sokoto Caliphate, the Islamic state later incorporated into Northern Nigeria, with whose ruling class most of independent Nigeria’s military rulers were connected. The other is more directed at the whole political culture of nationalism, which is seen by Born-agains to have so let the country down. It is 1977, when the symbolic revival of tradition that legitimated nationalism reached its climax in the Second World Festival of Black and African Arts and Culture, funded at great expense from the boom in oil revenues. For the born-again view is that Nigeria needs to be redeemed from its past. It is not much of an irony that, working from different premises, the Odua Peoples Congress comes to a similarly negative assessment of what the Yoruba should do about Nigeria.

What they also have in common is that their capacity to reject the recent past depends on their ability to reach out to a plausible future . . .


In conclusion, let me briefly return to the issue of the relations between ethnography and history, this time by a somewhat postmodernist route, in that it has to do with the character of ethnographic writing. Ethnography, like history, is an ambiguous term, but not in quite the same way. As has often been noted, in English and French (but not in German or Latin), the term “history” refers both to the past and to written accounts of it. Ethnography refers to written accounts too; but also to the way that the subject-matter of those accounts was collected, rather than the subject-matter itself. When we call someone a “great ethnographer,” it is their fieldwork rather than their writing that we have in mind; for it is from this that their authority as an anthropologist is typically held to derive. If it is the assurance of immediate copresence, “Being There,” that most of all underwrites the truth claims of anthropology, then the fact that ethnography qua writing up usually takes place [548]miles away and years later presents something of a epistemological problem for it. Fieldnotes, the immediate raw-material of written ethnography, become a record of encounters in a now-vanished present; they turn into an archive; and ethnography seems to be revealed as nothing more than a species of historiography. If so, it is surely a different kind of historiography, with the author more implicated in the subject-matter. If the past is indeed a foreign country, then we must re-read our fieldnotes—whose contents we have started to forget—as quasi-strangers to our former selves. Yet we have the faculty of memory, and re-reading (we trust) will stimulate memories which will restore the immediacy of the original field encounters. What I like about this view—which (to give credit where it is due) was proposed to me by a SOAS colleague who herself has taken more than a year or two between her fieldwork and her book—is its Proustian quality: one’s old fieldnotes as a plate of madeleines.

But what if we work this argument in reverse: if fieldnotes can become an archive, perhaps an archive can be treated as fieldnotes? The material on nineteenth-century Yoruba religion that I have drawn on in this lecture mostly came from the letters and journals of missionary agents, the great majority of them Africans, preserved in an archive of exceptional richness. Because it is an archive, any use of it will need to be judged by the usual canons of historical research. But I must say that as I immersed myself in it over several years, I came to think of it as being like a set of fieldnotes. Partly this was because I wanted to write in the genre of “historical ethnography,” a study of the encounter of religions contextualized in terms of the lifeworlds and social relations of those turbulent nineteenth-century Yoruba towns. Then too, the journals themselves seemed very like fieldnotes, and often very good ones: at their best vivid, detailed and informed accounts, written up shortly afterwards, of what their evangelist authors had seen, done, heard, argued about, experienced, as they went about their business amid the daily flow of life. But if they were thus like fieldnotes, they were not my fieldnotes, and it was only a fancy that I could really use them as if they were.

But of course no anthropologist of the Yoruba could read those documents without feeling that his capacity to interpret them was enhanced, and in some places made possible, by his own field experience. And they do sometimes work, like my colleague’s fieldnotes, to activate personal memories. Let me give just one example. Toward the end of my first year’s fieldwork in 1964–65, someone quoted to me a well-known proverb (it was the first time I was aware of it): Bi ewe ba pẹ lara ọṣẹ, yio di ọṣẹ, “If the leaf stays long upon the soap, it will become soap itself.” (The proverb alludes to the solvent properties of the soft blackish “native” soap, made from ash and palm-oil). It was a kindly way of complimenting me on my adaptation to the ways of the Aladura churches I was then studying. I had all but forgotten this until I came across the proverb—twice in fact—in the CMS archive, referring to the mutual adaptation of Christianity and Yoruba culture. While such a continuity is gratifying, it is treacherous, and needs to be balanced by the evidence of change. More valuable than the temps retrouvé of 1960s fieldwork was the effect of a three-months visit in 1994, when I had finished work on the archives and was gearing up to write. To then encounter with amazement a new metamorphosis of Yoruba Christianity—one which rejected a great deal of the past from which it arose yet still had to construct a narrative connecting itself [549]with that past—forced me to look anew at the latent potential of the earlier period I had come to know from the archive. Here too the modest time-depth of an anthropologist’s personal acquaintance with another society seems to sharpen his capacity to appreciate the balance of continuity and change in its history over the longer span. What we need to learn is the ethnographic value, not just of “Being There,” but of “Going Back.”


Ardener, Edwin. 1989. “The construction of history: ‘vestiges of creation.’” In History and ethnicity, edited by E. Tonkin, M. McDonald and M. Chapman. London: Routledge.

Asad, Talal. 1993. Genealogies of religion. Baltimore: Johns Hopkins University Press.

———. 1986. The Idea of an anthropology of Islam. Washington, DC: Center for Contemporary Arab Studies, Georgetown University.

Clifford, James, and George E. Marcus, eds. 1986. Writing culture: The poetics and politics of ethnography. Berkeley: University of California Press.

Comaroff, Jean and John L. Comaroff. 1991. Of revelation and revolution: Christianity, colonialism and consciousness in South Africa. Chicago: University of Chicago Press.

Dilley, Roy Martin, ed. 1999. The problem of context: Perspectives from social anthropology and elsewhere. New York: Berghahn Books.

Farrow, Rev. Stephen S. 1926. Faith, fancies and fetich; or Yoruba paganism. London: Society for Promoting Christian Knowledge.

Fisher, H. J. 1973. “Conversion reconsidered: Some historical aspects of religious conversion in West Africa.” Africa 43: 27–40.

Fortes, Meyer. 1959. Oedipus and Job in West African religion. Cambridge: Cambridge University Press.

Geertz, Clifford. (1966) 1973. “Religion as a cultural system.” In Interpretation of cultures: Selected essays New York: Basic Books.

Gellner, Ernest. 1988. Plough, book and sword: The structure of human history. London: Collins Harvill.

Halbwachs, Maurice. 1941. La Topographie légendaire des évangiles en terre sainte. Paris: Presses Universitaires de France.

———. 1992. On collective memory. Chicago: University of Chicago Press.

Heusch, Luc de. 1982. Les rois nés d’un coeur de vache. Paris: Gallimard.

Hobsbawm, Eric J., and Terence O. Ranger, eds. 1983. The invention of tradition. Cambridge: Cambridge University Press.

Horton, Robin. 1971. “African conversion.” Africa 41: 85–108.

———. 1975. “On the rationality of conversion.” Africa 45: 219–35, 373–99.[550]

———. 1993. Patterns of thought in Africa and the West: Essays on magic, religion, and science. Cambridge: Cambridge University Press.

Hutton, Patrick, H. 1993. History as an art of memory. Hanover, NH: University Press of New England.

James, Wendy and Douglas H. Johnson, eds. 1988. Vernacular Christianity: Essays in the social anthropology of religion presented to Godfrey Lienhardt. Oxford: JASO.

Last, Murray. 1993. “History as religion: de-constructing the Magians ‘Maguzawa’ of Nigerian Hausaland.” In L’Invention religieuse en Afrique: Histoire et religion en Afrique noire, edited by J. P. Chrétien, 267–97. Paris: Karthala.

Lowenthal, David. 1985. The past is a foreign country. Cambridge: Cambridge University Press.

Nora, Pierre, ed. 1996–1998. Realms of memory: The construction of the French past, three volumes, English language edition. Edited by Lawrence D. Kritzman, translated by Arthur Goldhammer. New York: Columbia University Press.

Ortner, Sherry B. 1989. High religion: A cultural and political history of Sherpa Buddhism. Princeton: Princeton University Press.

Parkin, David. 1991. Sacred void: Spatial images of work and ritual among the Giriama of Kenya. Cambridge: Cambridge University Press.

Peel, J. D. Y. 2000. Religious encounter and the making of the Yoruba. Bloomington, IN: Indiana University Press.

———. 2016. Christianity, Islam, and Orisa religion: Three traditions in comparison and interaction. Berkeley: University of California Press.

Sahlins, Marshall. 1985. Islands of history. Chicago: University of Chicago Press.

Strathern, Marilyn. 1987. “Out of context: the persuasive fictions of anthropology.” Current Anthropology 28: 251–81.

Roscoe, Rev. John. 1911. The Baganda: Their customs and beliefs. London: MacMillan.

Vansina, Jan. 1965. Oral tradition: A study of historical methodology. Translated by H. M. Wright. London: Routledge and Kegan Paul.

———. 1985. Oral tradition as history. Madison, WI: University of Wisconsin Press.

———. 1994. Living with Africa. Madison, WI: University of Wisconsin Press.

Wolf, Eric R. 1982. Europe and the people without history. Berkeley: University of California Press.

Yates, Frances. 1966. The art of memory. London: Routledge & Kegan Paul.

Temps et différence dans l’anthropologie de la religion: Frazer Lecture, Oxford, 9 mai 2000

Résumé : La Frazer Lecture de 2000 donnée par J. D. Y. Peel, publiée ici de manière posthume, présente ses premières reflexions au sujet de la comparaison triptyque qui culmina avec sa trilogie de travaux sur la religion Yoruba. L’exploration des idées qui y sont évoquées occupa la décennie suivante, jusqu’à la publication de Christianity, Islam, and Orisa religion: Three traditions in comparison and interaction (2016, University of California Press). Historien et sociologue, John était tour à tour stimulé et exaspéré par les anthropologues. Une méthode ethnographique était essentielle aux comparaisons qui l’intéressaient, mais les anthropologues étaient incompétents à ses yeux en matière de temporalité, à plusieurs niveaux: quand ils situent leurs propres recherches, les vies de ceux qu’ils rencontrent, les sources qu’ils utilisent, leur propres notes; et lorsqu’ils délimitent ce qu’ils entendent par contexte, ce qu’ils présentent à leurs sujets, et l’origine de ces conceptions. De plus, ils ne semblent pas se préoccuper des trajectoires historiques conférées aux religions par leurs histoires, les discours religieux, et les pratiques religieuses: ceci revêt une importance cruciale dans cet essai. En bref, même s’ils ont écrit abondamment sur ces sujets, les anthropologues sont concrètement dépourvus d’une forme d’attention qui rendraient compte du fait que les être humains sont des êtres dans le temps. Cette conférence propose de s’adresser à ces carences.

J. D. Y. PEEL, FBA, (1941–2015) was Professor of Anthropology and Sociology with reference to Africa at SOAS, University of London. His publications were concerned predominantly with social theory and its history, including a biography of Herbert Spencer (1971), and with the social history of West Africa, particularly the Yoruba about whom he wrote four monographs, including a trilogy on religious change: Aladura: A religious movement among the Yoruba (Oxford University Press, 1968), Religious encounter and the making of the Yoruba (Indiana University Press, 2000), and Christianity, Islam, and Orisa religion: Three traditions in comparison and interaction (University of California Press, 2016), as well as the political study Ijeshas and Nigerians: The incorporation of a Yoruba Kingdom (Cambridge University Press, 1983).


1. See Fortes 1959.

2. Published in Sahlins 1985, chapter 4.

3. Published in Horton 1993, chapter 4.

4. See Strathern 1987.

5. Vansina restated his position in Oral tradition as history (1985), and gave a useful brief review of the debate in his autobiography, Living with Africa (1994).

6. As David Parkin (1991: 1) remarked: “Quietly and undeclared as such, a loose assemblage of thinking has entered social science and the humanities. . . . It is the language and study of positions, stances, moves, panoptic views and close or distant gazes, in short of spatial orientation and separation . . .”

7. See Nora 1996–1998, which is an authorized English compression of the original seven volumes of Les lieux de mémoire.

8. See Halbwachs 1941, of which a translation is appended to Halbwachs 1992. For a useful discussion see Hutton (1993: 73–90).

9. A striking case of the centrality of religion to a people’s sense of the history that matters to them is provided in Ortner (1989: 8). In comparison to her earlier, more “anthropological” project on the symbolism and practice of Sherpa rituals, this historical approach, “made sense to people. To a great extent, Sherpa history is a history of their religion.”

10. As astutely noted by Ardener (1989: 22).

11. See, for example, Last 1993.

12. The bulk of material presented below is taken from Peel 2000.

13. See, for example, the Rev. John Roscoe’s The Baganda: Their customs and beliefs (1911), dedicated to him. The nearest the Yoruba came, was the Rev. Stephen S. Farrow’s Faith, fancies and fetich; or Yoruba Paganism (1926), not a great work, and with an introduction by R .R. Marett.

14. “The role of the media in the constitution of new religious publics in Yorubaland” was funded between 1996–99 by the AHRB. It was a joint project between SOAS, University of London, and the University of Birmingham, of which the co-directors in addition to J. D. Y. Peel were Louis Brenner, Paulo Farias, and Karin Barber. Its archive is held in Birmingham [RF based on information from Karin Barber].