A NEW THEORY OF RELIGION AND POLITICS

All the evidence points to the fact that most Africans – like most people on the planet, for that matter – understand and interpret the world partly through the prism of religion. In other words, religion, whatever else it may be, is a mode of apprehending reality. Much, of course, depends on what is meant by ‘religion’. There is a wide variety of definitions in existence. For present purposes, we have argued, the best way to proceed is not to assume that religion has the same meaning in all times and places, nor to use whatever definition the writer personally finds most pleasing. A better approach, we suggest, is to study the range of social phenomena observable in Africa, and only then to formulate a definition of religion that incorporates features relevant to its specific context. This then constitutes a working definition – not an attempt to classify religion in general, but a tool adapted for the purpose at hand (Ellis & Ter Haar Reference Ellis and Ter Haar2004: 13–16). A working definition has the advantage of being provisional in nature, meaning that it can be adapted in future as new data become available (Platvoet & Molendijk Reference Platvoet and Molendijk1999). Such a definition will then proceed from local epistemologies. Among the most salient features of African epistemologies, we have argued on the basis of empirical research, is a conviction that the material and immaterial aspects of life cannot be separated, although they can be distinguished from each other, much as the two sides of a coin can be discerned but not parted. To judge from the available evidence, religion in sub-Saharan Africa is best considered as a belief in the existence of an invisible world, distinct but not separate from the visible one, that is home to spiritual beings with effective powers over the material world. This is the definition that we used in Worlds of Power.

Another distinctive feature of religion in sub-Saharan Africa is its use of what may be termed a ‘spirit idiom’. This refers to the widespread belief that the immaterial forces perceived to be operating in the material world consist of, or are controlled by, individual spirits. These spirits are often imagined as having a name and a personality, and to have their abode in an invisible world. This spirit world is perceived to contain power, and for those who believe in it, this power is real. Belief in the existence of immaterial forces is common pretty much everywhere in the world, although these forces are often imagined in secular terms, for example as social or economic ones. No serious social scientist thinks that because something cannot be seen, it therefore does not exist. Furthermore, as we have argued, a social scientist or other analyst does not have to be a religious believer in order to study or understand religion. Our own analysis is not written from the point of view of a religious believer. However, for many observers, taking African religious ideas seriously requires thinking about religion in terms different from those with which they are most familiar.

The historical record suggests that what we term ‘religion’ has always existed in every part of Africa, and that it has been changing continuously. We do not believe that there exists an authentic Africa that stands outside time and that incorporates a primal view of the cosmos. Simply, Africa, like other parts of the world, has a history. It is possible to reconstruct, at least partially, what people in Africa have thought about the world in the past and how their ideas have changed over time (cf. Collingwood Reference Collingwood1993: 228). Any analysis that places distinct phenomena in historical context implies a greater or lesser degree of continuity. In the present case, the historically grounded theory that we propose tends to emphasise continuities rather than ruptures. Hence, we see current charismatic and neo-pentecostal or ‘born-again’ Christianity, for example, as a recent development in a long-existing mode of thinking about the spirit world rather than as a major historical rupture. This is in spite of the fact that ‘born-agains’ often make extravagant rhetorical claims to have made a complete break with the past (Peel Reference Peel2006). In other words, the claim of a born-again Christian to have broken with the past is to be understood as an emic statement, as distinct from the etic position adopted by an academic analyst. Thus, the neo-pentecostal movement in Africa reflects the times we live in, on the one hand, but, on the other hand, it can also be situated in historical context. Religion is a symbolic language, whose evolution may be compared to that of other languages. Like them, it reflects and communicates people's ideas about the world they live in. All languages change over time. Individual items of vocabulary disappear and new ones appear, invented on the spot or borrowed from abroad. Languages, however, also have a grammar, a structure that changes only slowly. Religion, too, has a vocabulary and a grammar, both of which change over time and at different speeds.

Taking religious ideas seriously – the heart of our endeavour – challenges the academic disciplines in which the study of contemporary Africa is most often conceived. The basic reason for this is that social science has been developed over generations on the assumption of a separation between the secular and the religious realms. Other writers (e.g. Lal Reference Lal1998) have pointed out that this separation reflects the historical experience of Europe, but not necessarily that of the rest of the world. The proper reaction to this awareness is not to reject social science as irredeemably Eurocentric, but to adapt its techniques in such a way as to encompass worldviews that are a product of histories different from those of European countries (Chakrabarty Reference Chakrabarty2000).

With regard to Africa, taking religious ideas seriously seems particularly to challenge the disciplines of anthropology, development economics and political science. If we deal with the latter first, we may observe that political science has no difficulty in accepting our proposal to understand politics as a field of activity that is not associated only with state power, but that can be found in a wide variety of social settings, in the form of politics ‘from below’. If this is a rather uncontroversial observation, it is no doubt because it is a theme that has been quite widely studied in the literature on politics in Africa for over 20 years (Bayart et al. Reference Bayart, Mbembe and Toulabor1992). Far more striking is the problem caused to political scientists by the subject of religion, which, at least until the post-9/11 period, they often omitted from their scope entirely. When political scientists have focussed on religion, with some notable exceptions, it has tended to be on its institutional aspects rather than the ideas and social practices that permeate the institutions (e.g. Haynes Reference Haynes1996). This is rather odd, since there is overwhelming evidence that institutions borrowed from elsewhere or imposed on a particular society by outsiders rapidly gain a distinctive flavour through assimilation into local repertoires of ideas, to the extent that states are individually formed through a multiplicity of procedures (Bayart Reference Bayart and Manor1991). Development economists, too, largely ignore religion, although there are signs that a number of research-funding institutions are beginning to reflect on the importance of religion in African societies, and this may lead to new angles of vision in due course (Ter Haar & Ellis Reference Ter Haar and Ellis2006). Anthropology, on the other hand, has a long record of considering African religions, but has tended to do so as cultural artefacts rather than as ‘real’ religions. Furthermore, anthropology has often considered African societies in an ethnographic present rather than in historical terms. Although many anthropologists nowadays would claim that these problems have been rectified, and that they do now place the phenomena they analyse in historical context, such an assertion remains open to question. It has become quite fashionable for anthropologists to take as objects of study topics related to religion. Prominent examples include the study of pentecostal movements and of witchcraft. Both of these are expressions of religious thought and practice. Pentecostalism is universally acknowledged as a particular form of religion, but is often treated by anthropologists with only scant regard for its global history and theological content. Witchcraft is usually studied without reference to religion at all – although, being concerned with a belief in mystical forces, witchcraft falls within the scope of religion (Ter Haar Reference Ter Haar2007) according to the definition we use. Some influential anthropological studies emphasise the role of these phenomena as modes of negotiating modernity (e.g. Comaroff & Comaroff Reference Comaroff and Comaroff1993; Geschiere Reference Geschiere1997; Meyer Reference Meyer1999), despite the fact that modernity is not a concept of great analytical value (Ashforth Reference Ashforth2005: 116–21). Moreover, analysing religious phenomena in terms of negotiating modernity overlooks the fact that, in Africa, religion has historically been a language for interpreting the world.

Any project of taking African epistemologies seriously, and attempting to incorporate them into a formal theory with the power to explain a range of social and political phenomena, is not an enterprise that concerns Africa alone. It is for this reason that our book Worlds of Power (2004) is interspersed with comparative material from other parts of the world. Given the extent and depth of new alignments or realignments of religion and politics throughout the world, it is possible that a theory developed on the basis of data from Africa could also provide new insights for analysing developments in other parts of the world, including notably Asia and the Middle East. Scholars therefore have a rare opportunity to show that Africa is not disconnected from world affairs and that the study of Africa can help in developing a better understanding of these. Religion and politics are undoubtedly forming new patterns in many parts of the world. This does not mean, however, that a religious revival is taking place worldwide, as is often suggested. The new patterns of religion and politics discernible in Africa and other places are of course affected by phenomena such as state failure, globalisation and economic crises, as many commentators have pointed out, but that is not the heart of the matter. The heart of the matter, rather, is that many people in the world, just as in sub-Saharan Africa, consider power as having its ultimate origin in the invisible world. This, we argue, has a marked influence on the conduct of politics and on political attributes such as authority and legitimacy.

Religion remains a prism through which many people view the world. Worlds of Power investigates this proposition in detail in regard to sub-Saharan Africa. We show that, for those who believe in it, spiritual power constitutes real and effective power. Thus, religion and politics become two facets of power that are in constant interaction. This is not always evident to observers of African politics, as African countries since colonial times have been officially governed through institutions based on a Western model of separation of church and state. This institutional architecture of government has tended to obscure the reality of spiritual power in Africa's public life. Like politicians the world over, Africa's political leaders spend most of their time in the pursuit or distribution of material resources, and their cultivation of spiritual power is usually more private than public. But, in any event, cultivating spiritual power is a vital component of a political career, as is widely attested by the popular media and radio trottoir.

This is why no study of African politics can afford to ignore the religious factor. A prime advantage of adopting our suggested definition of religion is that it enables us to consider religion and politics within a single field of power.

SOME ISSUES OF METHOD

Writing about religion and politics in the way we suggest has certain methodological implications. Among other things, it entails writing about religion in Africa in a manner that is objective, in the sense of not containing hidden assumptions about its ontological and moral status.

The latter requires particular attention to terminology. We have already noted (see above) that many reviewers of Worlds of Power assign to such categories as ‘superstition’ and ‘magic’ a range of phenomena that we prefer to classify as ‘religious’. Reference to ‘superstition’ and ‘magic’ amounts to making unscholarly judgements as to what constitutes proper or ‘real’ religion, as opposed to improper types of religion. In similar vein, many anthropologists writing on Africa today use ‘the occult’ as a category to include various phenomena that we describe as ‘mystical’. The word ‘occult’, although having a primary meaning of ‘hidden’, has a pejorative implication when applied to religious matters. All these labels carry a heavy ideological burden that should be set in historical context: for centuries, outsiders have tended to judge religious ideas and practices in Africa as both wrong and morally inferior. The use of such labels as ‘magic’, ‘superstition’ and ‘the occult’ implies that certain phenomena which in other parts of the world would be classified as religion, are better described by different terms in the case of Africa. This amounts to a form of exceptionalism, suggesting that a special vocabulary is needed for analysing Africa by reason of its supposed uniqueness. It is ironic that the charge of exceptionalism is sometimes made against our approach by anthropologists, who, we maintain, continue to struggle with the implications of the ethnographic method that has been so foundational in their discipline (e.g. Green Reference Green2006).

To be sure, it is inevitable that social scientists will attribute data to categories of their own choosing as part of a process of analysis. However, the choice of categories is neither arbitrary nor innocent. In matters of religion, we maintain that scholarly analysis is best done in two stages, known as emic and etic modes of interpretation. This means that the starting point for analysis is to consider religion in a subject's own terms of expression, before analysing it in terms of social science at a later stage. Following a two-stage process of interpretation of religious data implies a translation of the symbolic language of religion into the secular language of social science. In the case of Africa, this process may well involve the literal translation of key concepts from an African to a European tongue. In the end, therefore, analysts can never escape the demands imposed by the language in which they write. Whatever terms they use will always carry some historical and ideological charge, even such apparently universal terms as ‘religion’ and ‘politics’ (Meyer Reference Meyer2004: 466). The knowledge that this is so does not, however, absolve analysts from their obligation to take due precautions in making cultural translations of all types (Crick Reference Crick1976: 109–27), defining, wherever necessary, any category that is being used for analytical purposes (Platvoet & Molendijk Reference Platvoet and Molendijk1999).

African epistemologies often pose a particular challenge to academic categorisation due to their tendency to work on the basis of addition or incorporation rather than replacement. In other words, people show a marked preference for adopting innovations by assimilating new elements into existing repertoires, without necessarily making a choice between ideas or practices considered in Western epistemologies to be mutually contradictory. The tendency in African religious traditions to innovate through incorporation rather than exclusive selection does not in itself pose a conceptual problem, but it does require attention from social scientists used to categorising by reference to distinct typologies. It is relevant to note that the social science literature on Africa has often been marked by an insistence on filtering data by reference to dualistic categories of analysis or binary oppositions, including notably tradition/modernity and African/non-African, whereas, in African epistemologies, ideas or institutions can belong to several of these categories simultaneously.

A final comment on method is to ascertain whether sub-Saharan Africa constitutes a viable category for analysis of the matter at hand. Opinion seems divided on this issue, with some reviewers of Worlds of Power objecting to the inclusion of material from outside Africa on grounds of relevance, or arguing that Africa is in any case too diverse to fit a single model of analysis. Others, meanwhile, warm to our suggestion that no aspect of African religion is so unusual that it cannot be understood by reference to universal categories suitably defined, and that a model of religious-political interaction developed on the basis of African data can perhaps be usefully applied to other parts of the world. It is not unusual to take one sub-continent, one continent, or even the entire world as a unit of analysis, depending on the matter under consideration. Few people seem to object to the idea of analysing economics or development in pan-African terms (as implied, for example, by the existence of NEPAD, the New Programme for Africa's Development). The question is whether the great variety of religious practices and ideas to be found in sub-Saharan Africa has enough in common as to support this level of generalisation.

In our view, sub-Saharan Africa has at least four common elements that make it a viable analytical unit for our purpose. These are, first, that religious cultures throughout the region show a marked tendency to posit the existence of a spirit world. Second, all of sub-Saharan Africa has a robust oral culture, notwithstanding the use of writing for religious purposes, in some places for many centuries. Third, African religious cultures have a strong idea of evil as a transcendental force. Fourth, all sub-Saharan countries have undergone a similar experience of colonialism in some shape or form, even Ethiopia and Liberia. All four factors have been significant in shaping people's religious ideas. It should go without saying that within an area of some 600 million people, there will always be a great variety of opinions – including, in this case, the existence of professed atheists – and that any given religious idea is always liable to be associated with different material outcomes, depending on various factors. Taking an entire sub-continent as a unit of analysis is therefore not unusual, nor does it amount to ‘cultural essentialism’ (Green Reference Green2006), in other words identifying a cultural element that has remained unchanged over time. On the contrary, the identification of common elements over a variety of societies permits comparison, while insistence on the historical processes that have affected Africa indicates that religious change has always taken place there.

In summary, then, we maintain that religion in Africa is grounded in modes of acquiring knowledge that both reflect and shape the ways in which people have viewed the world, past and present. If only for this reason, religion has an important bearing on politics, and indeed politics in Africa cannot be fully understood without taking its religious dimension into account. Although African epistemologies involve concepts that may be unfamiliar to many Europeans and North Americans, there is nothing in them that cannot be analysed by the conventional methods of social science, provided both the scope of investigation and the terms of analysis are considered with sufficient rigour.

RELIGION AND DEVELOPMENT

Up to this point, we have considered the relation between religion and politics in Africa as a purely theoretical matter, striving to avoid terms that imply either approval or disapproval of the data at hand. The question arises, however, how a theory of religion and politics might be applied to specific situations.

Some observers seek to go beyond a consideration of the theoretical relationship between religion and politics and its close relative, economics. Interested in practical solutions to real problems, they want to know whether particular religious trends help or hinder development. For example, it is sometimes argued that the emphasis on deliverance found in neo-pentecostalism diminishes individuals' sense of personal responsibility for their own financial circumstances, as does the rhetoric of miracles. The religious views of charismatics – the argument goes – do nothing to encourage productivity or a work ethos. The spiritualisation of politics can thus play into the hands of the worst dictators, who are able to use religion as a channel of political support (e.g. Gifford Reference Gifford1993, Reference Gifford2004). ‘In what way have religious revivals affected resources or made life more successful, and how could one show that?’, Gifford (Reference Gifford2005: 247) asks in a review of Worlds of Power. ‘Are modifications of religious ideas in fact leading to Africa's economic and political progress?’

These questions cannot be answered satisfactorily without prior consideration of what the people concerned – in this case, Africans – understand to constitute ‘progress’ and ‘development’ (Ter Haar Reference Ter Haar2006). It is a mistake to assume, as some authors do, that an improvement of material and institutional conditions would necessarily reduce ‘the hegemonies of the spirit’ (Green Reference Green2006) that are apparent in Africa, and would instead stimulate the language of liberal secularism. Nor can the symbolic language of religion be reduced to a cultural tradition reflecting the continuation of certain ancestral practices. Rather, a religious mode of apprehending reality (even one couched in a spirit idiom) constitutes an epistemology that is simultaneously traditional and modern, capable of updating and renewing itself as times change.

African epistemologies, then, include religious perspectives affecting popular understandings of concepts such as progress and development, but also justice, prosperity and others (Ter Haar & Ellis Reference Ter Haar and Ellis2006). Consequently, Africa's economic and political progress should not be considered exclusively in terms of technical criteria such as macro-economic indicators, nor should all the standard formulae of development be taken at face value. Nor should the ways in which Africans debate these matters be assumed to be apolitical because they are expressed in spiritual idioms. One example is ‘good governance’, an expression widely used in development circles, but which serves poorly as an analytical concept due to its implicit value-judgement. The word ‘good’ in this expression invites judgement in regard to technical excellence, but it also conveys a moral concept. For many Muslims, for example, good governance implies a society that is ultimately ruled in conformity with divine law. Among both Muslims and non-Muslims in Africa, the moral nature of power depends on the manner of its exercise, as we have suggested in Worlds of Power (Ellis & Ter Haar Reference Ellis and Ter Haar2004: ch. 7). Charismatic preachers preoccupied with Satan are not necessarily externalising responsibility for the misfortunes of the society they live in, but may rather be considered as condemning the actual presence of evil within their society. Such criticisms are typically expressed in a spirit idiom and are often extended to national politics, as we have demonstrated at length in Worlds of Power. The Nigerian scholar Ogbu Kalu (Reference Kalu2006) has pointed out how Western opinions on these matters contrast with the analyses of African scholars. These include Matthews Ojo (Reference Ojo2006), Kwabena Asamoah-Gyadu (Reference Asamoah-Gyadu2005) and Asonzeh Ukah (forthcoming). This underlines the importance of taking into account studies by scholars living and working in Africa.

Development, from a religious perspective, is more than a set of technical benchmarks. For many religious believers, the road to development is at least as important as the final destination. For them, development is an integral process that implicates the full range of human existence, including its perceived spiritual dimension. In such a view, lasting and effective social change is dependent on the inner change or transformation of individuals. This is clearly an opinion shared by charismatic Christians and by Islamists, for example, both of them representing important and popular movements worldwide, and not just in Africa. As one observer has noted, many of the major flaws in the development process arise from a failure to consider the metaphysical questions concerning human life (Tyndale Reference Tyndale2001: 3). In investigating the connection between religion and development, therefore, the first requirement should be to understand what religious believers are saying and thinking about the nature of society and its defects, and what ideas they have about its possible improvement. Academic researchers should scrutinise these issues within the context of religious actors' worldviews. Only then is it helpful to investigate the matter from a social science perspective, and to investigate, for example, whether a given religious group has issued a statement on development or human rights or any associated matter, or has taken money from a corrupt head of state.

Within the development business, there are signs of a growing acceptance of the possibility of different paths to political and economic development. It has been noted that the success of many East Asian countries was achieved largely by ignoring the advice of the international financial institutions on specific economic policies (Stiglitz Reference Stiglitz2002: 91). More importantly, it was also achieved by some creative marriages between the technical requirements for development and local political resources that invariably express themselves in cultural form (Bayart Reference Bayart1994). The realisation that development has an inescapable cultural dimension has, over time, helped to mellow some of the rigid and even dogmatic theories of modernisation that were current in the middle of the last century. This is not to suggest the existence of fixed, development-friendly or development-unfriendly cultures, but only to underline that political choices affecting development contain a cultural dimension.

In recent years the relationship between the macro-economic aspects of development and the social processes involved has been articulated through the concept of human development, now widely used by both policy-makers and theorists. According to the United Nations Development Programme (2006), human development, ‘is about creating an environment in which people can develop their full potential and lead productive, creative lives in accord with their needs and interests’. It thus refers to people's resources beyond any purely material aspect. Many policy-makers today accept that sustainable development can be achieved only if people build on their own resources, including the quality of relationships in society, often construed as ‘social capital’. In Africa, we have noted, communication with a perceived spirit world is common religious practice. In other words, social relationships extend into the invisible world and the latter hence becomes part of people's ‘social capital’. For those who believe in its existence, the invisible world is another human resource. For analytical purposes, religious resources may be divided into four major categories. Religious ideas (what people believe) are one such category. Others are religious practices (what people actually do on the basis of such belief), religious organisation (how religious communities are formed and function), and religious – or spiritual – experiences (such as the subjective experience of inner change or transformation) (Ter Haar Reference Ter Haar, ter Haar and Busuttil2005: 22–7). All of these elements produce knowledge that, in principle, could be made beneficial to the well-being or the development of a given community.

No person or institution has thus far been able to identify mechanisms for eliminating corruption or injustice, or for improving the quality of governance in Africa, so effective that they cannot be subverted. Effective action to improve the quality of governance, and thus of economic performance, can come only from a variety of factors that include a widespread moral commitment to this goal. The latter can take a religious or a non-religious form: the overwhelming preference of Africans seems to be to express these matters in a religious idiom. In any event, improvements of this nature do not come from legislation alone, but from a changed moral climate (Wraith & Simpkins Reference Wraith and Simpkins1963). This is a point where religion and development may meet.

SPIRITUAL KNOWLEDGE AND HISTORY

Conclusions as to how religion may either help or hinder development can be situated on a continuum. At one extreme is the proposition that Africa has to adopt specified institutional forms if it is to achieve the ‘good governance’ considered necessary for development. This is a point of view favoured by aid donors and by the international financial institutions, which generally base their ideas about the types of institutions suitable for their purpose on models drawn from an idealised reading of European and North American history. At the other end of the spectrum is an opinion that indigenous institutions and epistemologies can potentially adapt themselves to a wide variety of purposes, and that development is indeed most likely where the fundamental technical skills or qualities required have been incorporated into local social patterns. However, whether indigenous or of foreign origin, all institutions depend crucially on the mental dispositions of those who staff them (Douglas Reference Douglas1986).

It is useful in this regard to consider the widespread sense of powerlessness that is often said to pervade African populations. Several sources – from Achille Mbembe (Reference Mbembe2000: 25–6) to the US National Intelligence Council (2005) – have noted the importance in Africa of the belief that the continent is threatened by vast forces that amount to something resembling a cosmic conspiracy. It is undeniable that Africa is indeed threatened by some massively destructive forces that can be quite precisely identified, including disease, debt and underinvestment. At the same time, however, many Africans also consider these conditions in a spirit idiom, seeing themselves as beset by evil forces that have a known material cause but that also have a spiritual dimension (Ellis & Ter Haar Reference Ellis and Ter Haar2004: ch. 3). The roots of this conviction that economic and political power has a dangerous spiritual aspect, which Africans can no longer control, can be located in the continent's history. Ultimately, this perceived lack of control dates from the colonisation and evangelisation of the nineteenth and twentieth centuries, that regarded indigenous spiritual forces as harmful, backward or demonic. A combination of ignorance and prejudice caused European colonisers and missionaries to despise many indigenous concepts of the spirit world (Ilesanmi Reference Ilesanmi1995: esp. 54). It is rather disturbing to note that many secular analysts today appear to react in a comparable way, displaying bias or intolerance regarding manifestations of an interaction between sacred and secular aspects of reality as it is perceived by many Africans.

One consequence of this history is that many traditional spiritual experts have lost prestige and are often regarded as unable to deal with the evil forces threatening the African continent today. While vast numbers of people continue to consult local healers, the feeling is widespread that they cannot provide effective remedies for the key problems of modern times (Ter Haar Reference Ter Haar1992: ch. 4; Ashforth Reference Ashforth2005: esp. 295–300). The quest for effective healing is evidenced by the enormous variety of priests, clerics, prophets, diviners or self-proclaimed experts that exists, contributing to what we term a ‘spiritual confusion’. By this, we mean a situation in which spiritual forces are considered powerful, but where there is little consensus on what precisely these forces are, and how they may be channelled or controlled. This is one main reason why religious or spiritual healing continues to have such importance in Africa.

As an illustration of what spiritual confusion can lead to, we may take the Rwandan genocide of 1994. Some reviewers of Worlds of Power wondered how a theoretical model of a political world that is deemed inseparable from the religious realm might help explain arguably the greatest tragedy of the late twentieth century. The extensive research carried out on the Rwandan genocide (Des Forges Reference Des Forges1999) has revealed the extent to which it was planned and implemented with a high degree of bureaucratic efficiency by a political and administrative elite associated with President Juvénal Habyarimana and his wife Agathe. However, research making use of Kinyarwanda-language sources and popular iconography indicates the overwhelming importance in the genocide of ideas about the nature and proper use of power that are derived from the sacred kingship of precolonial times (Taylor Reference Taylor and Whitehead2004). In the circumstances of the early 1990s, many Rwandans were inclined to believe that their country was not able to enjoy the prosperity and fertility that, according to local religious and political ideas, emanated from the mystical force of imaana, traditionally channelled through the government. The work of the génocidaires associated with the army and the ruling party consisted in persuading many Rwandans to accept that the blockage of the necessary flows of virtue and fertility, simultaneously political and spiritual, was the fault of political opponents of the government. The interest of this point is not that it suggests an alternative explanation for the 1994 genocide. Rather, its importance is that it adds an extra dimension to explanations that are normally couched in uniquely secular-political terms. Ordinary Rwandans were moved to acts of genocide, spurred on by propaganda and fear, because they believed their society to be threatened not only by political and military upheaval, but by malevolent spiritual forces also. Only by taking this spiritual dimension into account does it become easier to understand the awfulness of genocide: why otherwise ‘good’ people are moved to do unspeakably evil things (Juergensmeyer Reference Juergensmeyer2003).

We hope to have demonstrated in this article that the theoretical model concerning the interaction of religion and politics that we have set out in our in book Worlds of Power has the power to explain a wide variety of political phenomena, not only in the conventional sphere of state politics, but in African societies at large. The main obstacle to the use of this model as a tool for analysing African politics in the broadest sense appears to be the novelty of some of its analytical categories, which are drawn from African epistemologies rather than from mainstream social science. In our view, this presents no insoluble problems to social scientists – it only requires them to adopt some new angles of analysis and reflection. Moreover, we maintain that such a theory of religion and politics may well be useful for analysing societies in other parts of the world in which these two forms of power are recomposing in ways that do not accord directly with the classical models of social science. There is an urgent need for analysts to consider how spiritual power works as a political force in contemporary societies.

Religion is the emerging political language of our time. Already, in our original article (Ellis & Ter Haar Reference Ellis and Ter Haar1998: 201), we noted that African politicians were challenged to use this language in a manner comprehensible to outsiders. We also noted that non-Africans needed to learn this language. This remains truer than ever: as with any language, learning to understand a spirit idiom takes time and application, but it can be done.