Animism: Comparing Durkheim and Chidester’s analyses of EB Tylor’s theory of religion

The purpose of this research study was to compare the analyses of the anthropologist Edward Tylor’s animist theory of religion in the work of two major scholars of religion. At the beginning of the 20th century, Durkheim refuted Tylor’s classical explanation of the origin of religion, before he would proceed to develop his own sociological explanation. At the turn of the 21st century, from a postcolonial South African location, David Chidester offered a critical analysis of the triple mediation under colonial and imperial conditions that made Tylor’s evolutionary theory possible. By foregrounding definitions, making arguments explicit and comparing these two assessments, the two analyses shed light on each other as well as allowed us to view the issue of animism in a new light. This article concluded by highlighting points that emerged and need continuing attention in the academic study of religion. Contribution: This article, as part of a collection on re-readings of major theorists of religion, offers a comparison of Durkheim and Chidester’s analyses of Tylor’s classical animist theory of religion. By comparison, the analyses shed light on each other and on the theory of animism itself, highlighting critical issues that deserve the continuing focus of students of religion.


Introduction
Reflecting on ways to introduce Durkheim to students, Smith (2005a:1-11) outlined four rules that he followed in his teaching of Durkheim's theory of religion. Smith's essay shows that this influential scholar of religion 1 not only followed these rules in his teaching of Durkheim but that he also applied them in his own research study. His wider oeuvre makes it clear that these rules actually informed his pedagogy and research, in general (cf. Smith 1991Smith , 2005b. What are these four rules and why are they important?
Firstly, definition. The best way to start an argument, Smith held, is to provide a working definition. 2 One must fill one's key terms with the content that limits the scope of what one will analyse (cf. , as that will provide one with a clear indication of the data that one will analyse. In Durkheim's Elementary Forms of the Religious Life, the purpose of the first chapter was precisely to offer a definition of 'religion' as a generic category.
Secondly, argument. The student and scholar must be able to make arguments explicit. Engaging with Durkheim, in Smith's view our most exemplary presenter of sustained argumentation that always keeps one involved, 3 means that one needs to follow Durkheim's argument and be able to clarify its development rather than merely cite his conclusion. In the first chapter of Elementary Forms, for example, Durkheim develops an argument in debate with other major definitions of religion before he offers his own working definition that will offer the point of reference for arguments that Durkheim will develop in the rest of his book.
1.On JZ Smith's profound influence on the academic study of religion, see the collected reflections in Crews and McCutcheon (eds. 2020).
2.Cf. Smith (1986:94): 'As I persist in teaching my students … the proper way to begin an argument is by an act of definition. It is only then that there is something on the table which you and I might agree to take as a datum for purposes of the argument'.
3.Cf. Smith (2004:128): 'I take Durkheim, especially in the Elementary Forms, to be one of the greatest crafters of argument in the history of the study of religion. From the translation of "religion" to "society" in the very first sentence, his work has a rhetorical and intellectual momentum from which it is almost impossible to disengage'. Smith then offers a critical analysis of Durkheim's argument on mana by showing that this concept was not present amongst Durkheim's aboriginal Australians, but was superimposed by him on them from Melanesian data.
The purpose of this research study was to compare the analyses of the anthropologist Edward Tylor's animist theory of religion in the work of two major scholars of religion. At the beginning of the 20th century, Durkheim refuted Tylor's classical explanation of the origin of religion, before he would proceed to develop his own sociological explanation. At the turn of the 21st century, from a postcolonial South African location, David Chidester offered a critical analysis of the triple mediation under colonial and imperial conditions that made Tylor's evolutionary theory possible. By foregrounding definitions, making arguments explicit and comparing these two assessments, the two analyses shed light on each other as well as allowed us to view the issue of animism in a new light. This article concluded by highlighting points that emerged and need continuing attention in the academic study of religion.
Thirdly, comparison. Nothing in Durkheim must be allowed to stand alone. Durkheim Elsewhere in his work, Smith devoted considerable energy to think through and propose a method, in which comparisons might be performed in a disciplined and productive way. In brief, he argued that after deciding on an interesting issue, one must analyse an example of the issue by contextualising the example historically and embedding it politically, economically, socially and culturally. One repeats the analysis of a second example that illustrates the issue in the same way. Only then would one be able to compare in a productive way the examples of the issue as to similarities and differences so that each example may shed light on the other, and help us to see and redescribe the issue itself in a new light. 5 Fourthly, with this new understanding of the issue, one should be able 'to play with' the idea by applying it to other examples in innovative ways. Students would, for example, be asked to apply Durkheim's theory of religion to an analysis of the Vietnam War Memorial in Washington, or to pertinent court cases in the United States in order to discover discrepancies in judgement.
In this contribution, I follow the first three of Smith's rules. I intend to compare Emile Durkheim and David Chidester's analyses of a classical theory of religion, that is, Edward Tylor's anthropological theory of animism in Primitive culture (1871), 6 to see to what extent we may understand the one analysis in terms of the other and see Tylor's classical theory of animism as the issue in a new light. Highlighting definitions of religion and animism as key terms will be crucial to the argument. Before I offer observations on similarities and differences between Durkheim and Chidester's analyses of Tylor's classical theory of animism, I present a close reading and make the arguments of Durkheim and Chidester on the theory explicit. In conclusion, I outline a number of issues that have emerged as deserving the continuing attention of students of religion.
The primary sources for comparison will be the relevant chapters from Durkheim's Les formes élémentaires de la vie religieuse: Le système totémique en Australie ([1912] 1968), 7 on 4. Smith (2004:1-60), in an autobiographical essay 'When the chips are down' speaks on how he spent his early academic years at the University of Chicago to come to terms with his influential senior colleague Eliade, by first offering an accurate summary of his work before developing a critique of it and alternative to it.
Central to Durkheim's broader definition of religion is the distinction between sacred and profane, where the first concept refers to anything that is 'set apart' from the ordinary. Rather than assuming that something is holy on the basis of the experiential claim of believers that a mystical power revealed itself in it (as would be emphasised in the definitions 8.Empire of Religion (2014) was published as a sequel almost 20 years after Savage Systems (1996). The earlier work focused on comparisons of religion in colonial South Africa, whilst the later work devoted much more attention to British imperial theorists of religion in relation to South African colonial middlemen and indigenous informants.
9.Cf. Capps (1995:159): 'The Elementary Forms of the Religious Life, first published in 1912, still stands as a landmark in sociological theory as well as within the history of the academic study of religion. E.E. Evans-Pritchard, who disagreed with Durkheim on many questions, nevertheless referred to him as "perhaps the greatest figure in the history of modern sociology."' 10.For a fuller contextualisation of Chidester's work, see Strijdom and Scharnick-Udemans (2018).
11.For a contextualisation and summary of Durkheim's argument against definitions that limit religion to transcendental beliefs, see Jones (1986:120-124).

12.Elementary
Forms is divided into three books, which after 'preliminary questions' (Book 1) focuses on elementary beliefs (Book 2) and rituals (Book 3). Each book consists of a number of chapters. The reference here is to Chapter 1 in Book 1.
from Schleiermacher through Rudolf Otto to phenomenologists of religion), Durkheim insisted that anything could be made sacred by adherents. The function of beliefs and ritual practices made sacred by adherents would, in his view, then be to unite those adherents into 'a single moral community'. By focusing one's analysis on sacred beliefs, ritual practices and institutions as basic elements of all religions, Durkheim held that one would be able to analyse all religious phenomena empirically as social facts within their concrete reality.
The subtitle of Elementary Forms indicates that Durkheim would take the totemic system in Australia as primary evidence in his analysis. Considering reports on Australian aboriginal religion as the earliest available form of religion, Durkheim argued that by examining it, one would be able to identify the most basic or simplest form of religion that persisted also in later, more complex forms. Durkheim (2008) explicitly rejected evolutionary theories that denigrated indigenous religions and held religions, such as Christianity, to be superior: All are equally religions, just as all living beings are equally alive, from the humblest unicellular organism to man. We turn to primitive religions, then, not with the ulterior motive of depreciating religion in general, for these religions are no less worthy than others. They answer the same needs, they play the same role, they issue from the same causes. They can effectively serve, as a result, to show the nature of religious life and consequently to resolve the problem we have set for ourselves. (Cosman's translation, p. 5) Durkheim's argument was, in other words, that an analysis of the earliest form of religion would assist in explaining and understanding all religious forms, thus contributing to the development of a general theory of religion.
Having clarified his stance towards and definition of religion, Durkheim proceeded to first refute the anthropologist Edward Tylor's animist theory of religion (Book 1, Chapter 2) and the philologist Max Müller's naturist theory of religion (Book 1, Chapter 3), before he turned to the development of his own argument on the totemic origin of religion in the rest of Elementary Forms. Our focus here is on Durkheim's presentation of Tylor's animist theory of religion. 14 According to Tylor (1832Tylor ( -1917, religion originated when humans started to worship spirits, that is, when early humans set out to construct a cult in which beliefs and ritual practices centred on spirits as agents ('animism' derives from Latin animus 'spirit(s)'). Durkheim held that Tylor's animist explanation of the origin of religion postulated three steps, moving from a belief in a soul to a belief in ancestral spirits to a belief in nature spirits. 15 After surveying these three steps, 14.Durkheim, in this chapter, considered Tylor as the most important exponent of animist theory, with Spencer's theory as essentially the same except in one respect. Animist theory, in the survey here, therefore, refers to Tylor's theory as presented by Durkheim. 15. Durkheim ([1912Durkheim ([ ] 1968Durkheim ([ :52-53, 1915 held that proponents of the naturist theory, with Max Müller (1823-1900) as the most important exponent, on the other hand postulated that the first cult was addressed to natural phenomena (whether cosmic forces, such as sun, moon and stars, or objects on earth, such as rocks, mountains, rivers, animals and plants) and only later to human spirits.
Durkheim would raise his objections corresponding to these steps.
The first step occurred, when early humans came up with the idea that a soul inhabited each body as a double -a belief that was born in order to explain dreams, whose images they mistakenly thought were identical to those seen whilst awake. In dreams, a far-away person could be met, or visiting a far-off land was possible, whilst the body remained in the same place. By positing a soul that could leave the body, the animist argument goes, the earliest humans thought they could explain their dreams. Although a decisive step in the development of religion, the invention of the idea of a soul was not yet religion in the sense of an animist cult addressed to spiritual agents.
The second step, the actual step towards an animist cult, occurred when the idea of spirits emerged. According to the animist explanation of the origin of religion, this belie, emerged because of the fact of death, which transformed souls into spirits. If the soul animated the living body, leaving it temporarily during dreams (or similar states, such as fainting or ecstasy), it separated definitely from the body at death. The earliest humans then thought that these spirits of their ancestors took on an independent existence, with the ability to cause good or harm to their descendants. By addressing the ancestral spirits in prayers and ritual practices, such as sacrifices, descendants believed that they could appease the ancestral spirits so that the latter might avert evil and act in the interest of their descendants. The first religion was according to the animist theory thus born, with the first rites being food offerings at ancestral tombs. By ascribing such power to spirits of their own making, early human beings in Tylor's judgement became the unfortunate captives of their own imaginations.
As a third and final step in the evolution of the earliest form of religion, according to animist theorists of religion, human beings started to believe that inanimate matter was inhabited by spirits. By ascribing life and agency to cosmic phenomena, such as the sun, moon and stars, or to other natural phenomena on earth, such as plants, animals, rocks, mountains and rivers, early human beings projected their own image onto and anthropomorphised their environment. Children and animals served in the animist argument as analogies for the mentality of the earliest humans, as if animals and children were not able to distinguish animate from inanimate objects, and ascribed life to the latter in their own image. This phase derived from and was according to the animist theory secondary to the previous step. Early human beings now, at this final stage, came to believe that they could influence and placate nature spirits, which was extremely important to survive in a hostile natural environment.
How and why did Durkheim object to the animist explanation of the origin of religion? Taking each of the three steps in turn, Durkheim began by expressing his reservations about the animist explanation of the emergence of a concept of soul.
http://www.hts.org.za Open Access Although he appreciated the effort by animist theorists to view the concept as a historical and mythical construct rather than a given, he was for several reasons not persuaded by their argument that the concept of a soul was created by the earliest human beings to explain their dreams. Not only was the soul commonly thought of by indigenous people as much more intimately entangled with the body (e.g. by being located in specific organs and carrying the same injuries to a body) than animist theorists would assume, but dreams themselves could have been seen by the early humans as simply another way of seeing afar. Travelling to the past in dreams would, furthermore, be difficult to explain by animist theorists. A much simpler hypothesis would be that the first humans would have considered such dreams as similar to memories that they had whilst awake, although more intense. Also, if meeting a friend in dreams presupposed the encounter of two souls, Durkheim held that checking with the friend would have revealed quickly that the friend did not have the same encounter in his dreams, and that the encounter in the dream was, therefore, clearly imagined. Finally, it was not clear to Durkheim why the earliest humans would have devoted so much time to their dreams, if their struggle during daytime would have required the primary focus for reflection. It was, in his view, more likely that the belief in spirits preceded the explanation of dreams in terms of a soul leaving the body during sleep rather than dreams providing the basis for the creation of the notion of a soul. This was in his view confirmed by the fact that indigenous Melanesians clearly distinguished between ordinary, profane dreams as a mere play of images and dreams in which souls featured.
Turning to the heart of the animist explanation of the origin of religion, the transformation from the belief in a soul to a cult addressed to ancestral spirits, Durkheim centred his critique on the absolute difference between the sacred and profane as the distinctive feature of religion in his definition of religion. Instead of accepting the animist theory that death transformed the belief in a soul into a cult that is addressed to spirits, Durkheim argued that ethnological data showed that early human beings thought of aging as weakening the soul and of death as erasing the soul as life principle of the body. Death would, therefore, not have rendered the soul sacred amongst early human beings. What would make a number of special ancestral spirits sacred and objects of ritual attention, however, was the belief that they possessed mana (a Melanesian concept that Durkheim thought applied to all earliest religions) 16 already during their lives, whereas this extraordinary power of mana was considered to be absent in ordinary spirits and profane things. Taking aboriginal Australians as earliest evidence of religion, he underlined that an ancestral cult simply did not exist amongst them. Their funerary and mourning rites did not involve periodic, recurrent ceremonies with prayers and sacrifices at graves that would constitute a cult but were only limited to once-off rites. In so far as ancestors did play a part in their religion at periodic ceremonies, they were not honoured as spirits who had been transformed by death into sacred beings but honoured as heroes with mana who created the land and human beings at the beginning of time. A cult addressed to ancestral spirits, Durkheim argued, was not early, but rather the characteristic of more advanced civilisations, such as those of China, Egypt, Greece and Rome.
Although Durkheim's refutation of the second step in the animist explanation of the origin of religion would render a critique of the third step superfluous, Durkheim nevertheless continued to critically engage with the thesis that the cult addressed to natural phenomena was an extension of the cult addressed to ancestral spirits. He considered it necessary as its presupposition was uncritically accepted by many scholars of religion that early human beings, like animals and children, were not able to distinguish between animate and inanimate objects, with the latter being imagined as alive like themselves. Durkheim argued that these analogies should be discarded, as evidence showed that both animals and children were able to distinguish between animate and inanimate objects. The cat knows whether a mouse that he has caught is alive or dead. The child plays with his or her toys, imagining that they are alive; however, the child will be the first to be astounded, if one of the toys were to bite him or her. Instead, one should focus on the content of the earliest belief in nature spirits, and ask whether it retained the traits of the human soul and ancestral spirits from which it was supposed to be derived. Durkheim failed to detect such a continuity, as nature spirits unlike human souls did not reside in the objects, which served as their base, but were thought to frequent and influence their base (e.g. causing storms at sea). Moreover, Durkheim found the animist argument invalid that early human beings ascribed life to inanimate things in their own image, as the evidence from aboriginal Australians as well as North American indigenous peoples showed that they imagined their ancestors and themselves in the form of animals rather than projecting their own image onto animals. The evidence rather showed that anthropomorphism was a later development, with Greece still retaining some animal characteristics in Gods, such as Dionysus and Demeter, and Christianity exemplifying the culmination of anthropomorphism with God constructed wholly out of human elements.
Durkheim's final objection to the animist theory of religion was directed at its reductive explanation of the emergence of beliefs in souls, spirits and sacred beings as based in the erroneous judgement of early humans regarding images in their dreams. He considered this explanation far from satisfactory, given the fact that religion had such an enormous influence on morals, law and society. It is, therefore, to social realities that he would turn to develop an alternative explanation of the origin of religion. According to Chidester (1996Chidester ( , 2014, it is crucial to locate Tylor's theory of animism not only within contemporary European debates but also in relation to colonial mediations. 17 If the former Eurocentric narrative is typical of histories of the academic study of religion, Chidester views his examination of the triple mediation of theory formation from indigenous through colonial to imperial comparative religion as an innovative retelling of this history. What has enabled this change of perspective, Chidester holds, has been his relocation to South Africa since the mid-1980s, from where he could take a new look at the history of comparative religion from a postcolonial periphery (cf. Strijdom & Scharnick-Udemans 2018).

Chidester on Tylor's animist theory of religion
Before he commences with his innovative analysis of the complex triple mediation, Chidester provides a few important pointers that contextualise Tylor's animist explanation of the origin of religion, defined by Tylor as 'the belief in spiritual beings', within imperial British centres. It particularly debates on the measurement of skulls, as well as the relationship between human and animal psychology amongst evolutionists, which are highlighted as the European context within which Tylor's theory needs to be understood.
If British phrenologists shared the same conference podiums as Tylor, arguing that the form and measurement of the skulls of Catholics and indigenous peoples ('savages' in their terms) indicated their intellectual and moral inferiority to Protestants and civilised Westerners, Tylor took a different approach to argue his evolutionary thesis of animism. For Tylor, the belief in spirits lied at the basis of all religions -it was particularly evident not only in the religion of indigenous people (termed 'savages' by him too) but also amongst spiritualists with their scéances in his own Victorian Britain. Religion, as Tylor argued, retained a 'primitive' mentality and psychology, which was to be erased in the interest of the evolutionary progress of civilisation.
The continuity between the psychology of animals and early humans (commonly referred to as 'primitives' in discourses of the time), posited by Darwin and his followers, appealed to Tylor. According to Darwinists, both animals and indigenous peoples (the thinking of contemporary 'savages' was taken to be the best indicators of the thinking of 'primitives' or the earliest humans) shared the same mentality of ascribing life to inanimate objects -the very mistaken judgement that Tylor considered at the origin of religion.
17.The term 'triple mediation' is a key concept in Chidester's analysis, focusing on the fact that imperial theorists of religion did not simply develop their theories on the basis of debates amongst themselves but also by drawing on colonial mediators' reports, who, in turn, depended on the reports of indigenous mediators who as converts were situated between indigenous and Christian traditions. Chidester shows, in other words, that the production of imperial theory was not 'immediate' but depended on a complex process of mediation from indigenous, through colonial to imperial mediators.
After sketching this European context, Chidester continues with his innovative unravelling of the way in which Tylor constructed his thesis of animism. By focusing on Tylor's use of reports on Zulu dreaming from colonial middlemen in South Africa, Chidester shows how Tylor's explanation of the origin of religion was based on a distortion and decontextualisation of the data that he quoted. How does Chidester's argument proceed?
The principal evidence that Tylor quoted for his theory of animism was the dreams 18 of Zulu speakers, particularly of a Zulu diviner, as reported in the Anglican missionary to Natal, Henry Callaway's The Religious System of the Amazulu (1868-1870). In Tylor's citation, the diviner was reported by Callaway to have continually seen spirits in his dreams to the point of becoming a 'house of dreams', which for Tylor indicated that the diviner -like Zulu people, in general, but more intensely -could not distinguish between images in his dreams and images in a waking state. Chidester underlines that Tylor chose to omit the crucial phrase in the text, which stated that the ancestral spirits in the diviner's dreams came to kill him. It is this phrase that will form the crux of Chidester's argument to locate Zulu dreaming within its colonial context and to show how Tylor constructed his theory by abstracting the data from its colonial context and thus distorting it.
If Tylor's theoretical construct constitutes the final mediation of theory, Callaway's presentation of a paper in London demonstrates the extent to which this colonial middleman did not simply transmit knowledge about indigenous Zulu religion to the imperial centre but also used imperial theory to interpret indigenous Zulu practices. In his paper, Callaway too abstracted Zulu dreams from its colonial context by arguing that Zulus were prone to confuse brain sensations in dreams with real images when they were awake.
Although The Religious System of the Amazulu was published under Callaway's name, Chidester emphasises that its actual author was Mpengula Mbande, Callaway's Christian convert and indigenous informer. This voice of Mbande, however, was not unambiguous, as Mbande was located between indigenous and Christian identifications. Coming from an indigenous religious family, Mbande had left his home and converted to Christianity. From this new missionary location, he portrayed traditional Zulu religion negatively, which was also evident in his account of James the diviner.
Like Mbande, James himself had converted to Christianity. But after receiving a call to become a diviner, James left the Christian mission. As a Christian convert loyal to the Christian mission, Mbande remarked that James like other traditional Africans blindly believed what the ancestors had transmitted. Instead of understanding this as a lament from an indigenous Christian convert within a changed situation, Tylor quoted the passage to prove the essential ignorance and intellectual inferiority of all 'primitive' people.
18. Chidester (2014:118-119) considers, in addition to dreaming, Tylor's citation of Callaway on Zulu sneezing as indicative of a belief in spirits amongst the Zulu to construct his theory on the animist origin of religion.
How does Chidester interpret the phrase, as reported by Mbande in Callaway, that according to the diviner James, the ancestral spirits came to kill him? The crucial phrase should, Chidester argues, be understood within the context of 19th-century colonial dislocation and dispossession of indigenous people. Not only was the land taken from indigenous people by British and Dutch settlers but the dispossession of land also meant that they lost their cattle.
In traditional Zulu religion, descendants were expected not only to bring back the spirits of deceased ancestors to their homesteads (the ukubuyisa ritual) but also to sacrifice cattle for them. Ancestral spirits appeared in dreams demanding from descendants that they act on their obligations, but colonial dislocation and dispossession rendered those duties increasingly difficult to fulfil. As a consequence, ritual techniques were devised to stop the dreaming. In the case of James, the diviner, however, he kept dreaming, with the ancestral spirits threatening to kill him. According to Chidester, the colonial context of dislocation and dispossession of indigenous people must then, serve as a proper context to understand those Zulu dreams in the 19th-century South Africa, rather than as data for the construct of an animist explanation of religion.
Although the complex process of triple mediation raises important ethical questions, Chidester does not pursue this further in Empire of Religion. In his keynote lecture at the International Association for the History of Religion Conference in Durban in 2000, however, he was much more explicit in taking a stance on the 'horrible' history of the imperial study of religion that served the political and capitalist interests of the British empire and the colonies, which should be 'engage [d] critically and creatively with the possibility of new horrible histories that might be on the horizon'  2009:117).

Comparing Durkheim and Chidester's analyses of Tylor's animist theory of religion
Reflecting on his place within the history of the academic study of religion, Chidester (2012:xii-xiii) traces his 'devotion' 19 to the field 'ultimately' back to Durkheim. From his oeuvre, it is clearly Durkheim's broad definition of and sociological approach to religion that have profoundly influenced Chidester's analysis. Not limiting the term religion to the belief in supernatural beings, Chidester (1991) would agree with Durkheim that the modern nation state with its sacred stories, rituals and institutions serves the same function as conventional religions in uniting a group, applying the insight to the apartheid system as well as to the construction of a post-apartheid South Africa. Chidester (2005Chidester ( , 2012 would, furthermore, also 'play with' Durkheim's understanding of religion by applying it to forms of popular culture that behave like conventional religions, such as Coca-Cola, Tupperware, Disney, Rock 'n' Roll and the World Cup in South Africa. 19. Chidester (2012:xii) uses the verb 'devoted to' probably in the sense of 'committed to', but I thought it is appropriate to change it here to the noun 'devotion', which would encourage an analysis of Chidester's academic study of religion itself as a kind of religion in the broader sense.
If, however, Chidester shares Durkheim's broad definition of and sociological approach to religion, in which ways does Chidester's analysis of religion differ and take Durkheim's analysis further? Chidester (2012:xii-xiv) himself emphasised that it was his relocation to South Africa from the United States since 1984, which has forced him to study religion and retell the history of the academic study of religion in a new way from a South African perspective. We may then, with reference to Durkheim and Chidester's respective analyses of Tylor's animist theory of religion, highlight a number of differences that will throw into relief crucial issues that demand the continuing attention in the academic study of religion.
Firstly, definition. Both Durkheim and Chidester considered Tylor's definition of religion as 'belief in ancestral beings' too narrow. Instead, Chidester would accept Durkheim's broader definition of the sacred as including anything that is 'set apart' from the ordinary by a group of people in such a way that it would unite them as a group.
Secondly, origins. Although Durkheim disagreed with Tylor's evolutionary view that depreciated indigenous religions, he shared with him the conviction that by finding the earliest form of religion, one would be able to study religions scientifically. Chidester, however, would disagree with both Tylor and Durkheim's key assumption on the importance of finding the origin of religion in order to understand all religions. Already in one of his earliest publications, Chidester (1990:2) stated that 'we can conclude nothing about the birth of religion'. Instead, Chidester has turned to an analysis of recurrent patterns in religious phenomena (cf. Chidester 1987Chidester , 1988Chidester , 1990, characteristic of phenomenological studies of religion, but with due emphasis on the changes of religious phenomena in recorded history. Instead of taking snapshots frozen in time as phenomenologists of religion would do, Chidester (2018:152-165) had insisted on the historical analysis of religious phenomena as 'moving pictures'.
Thirdly, critique. In assessing Tylor's theory of animism, Durkheim engaged directly with Tylor's arguments by offering objections to them before he would develop his own counter proposal. Chidester's primary interest, however, is not to develop an alternative thesis on the origin of religion but to demonstrate how Tylor constructed his theory by decontextualising and distorting data obtained from colonial middlemen who were dependent on ambiguous indigenous informants. What Chidester does, and Durkheim failed to do, is to focus our attention on the power relations that were at work in the production of an imperial theory of religion, such as that of Tylor. As pointed out in this study, this innovative retelling of an important phase in the academic study of religion has been made possible by Chidester's migration to South Africa. Chidester has been able to shift the standard intra-European accounts of the history of the academic study of religion to an innovative and critical retelling from the vantage point of South Africa with its colonial and postcolonial, apartheid and post-apartheid history.
Fourthly, ancestral spirits. The main objective of Durkheim in his discussion of Tylor's theory was to discard each of Tylor's arguments on animism as earliest religion that developed from belief in a soul (caused by a failure to understand dreams) to the worship of ancestral spirits (effected by death) and on to the worship of nature spirits. Chidester does not discuss Tylor's theory in terms of these three stages. However, if Durkheim did not offer in his assessment of Tylor any serious attempt to analyse the belief in ancestral spirits, Chidester does so explicitly by interpreting 19th-century Zulu dreams under colonial conditions. We return in conclusion to the way in which Chidester has continued his analysis of indigenous dreams and ancestral spirits within the changing political and economic context of South Africa.
Fifthly, intervention. Tylor considered it his task as a scientist to eradicate beliefs in spirits as remnants of an erroneous 'primitive' mentality in the interest of the progress of civilisation. Durkheim differed forcefully from Tylor by insisting that one should rather acknowledge and analyse religion as a social fact that fulfils the crucial function of binding a group together. Chidester, similarly to Durkheim, would agree on the necessity of analysing the social functions of religion, but would also take the analysis further by exposing power relations at work in religious beliefs, practices and institutions. In most of his work, Chidester uses critical theories to describe the role of religion in creating, sustaining and contesting asymmetrical power relations of class, race and gender within changing political and economic contexts, although he does occasionally take an ethical stance against the 'horrible' history of the academic study of religion.

Conclusion
Which are the issues that emerged from the comparative analysis in this study deserving the continuing attention of students of religion? Following on the above points, I highlight four issues that appear to me of importance for the study of religion, in general, and of religion from a South African perspective, in particular.
Firstly, again, origins. Is the search for the origin of religion still relevant? Chidester doubts its relevance, and has not engaged with Tylor's and Durkheim's proposals, although he has accepted, adapted and applied the latter's definition of religion that was based on Durkheim's argument on the earliest form of religion. Except for his book on Christianity that traces its history from its origins , Chidester has clearly chosen to focus on the changes that religions, including South African indigenous religions, have undergone in the modern colonial and postcolonial period.
The search for the origin of religion is, however, continuing as is evident from Robert Bellah (2011)'s Religion in Human Evolution. In his response to Bellah, JZ Smith (2011) emphasised that despite the need for criticising particular accounts, the question of religion in the evolution of humanity remains important in so far as accounts are informed by pertinent theories and based on empirical data, such as those developed by cognitive theorists of religion.
Secondly, animism. Is the concept of animism still useful for analytical purposes? In Religion: Material Dynamics, Chidester (2018:23-29) challenges Tylor's dualism of soul and body, arguing that Tylor's emphasis on beliefs and soul at the expense of body (the general tendency of the founders of the academic study of religion because of a Protestant bias) should now be corrected by attending to material mediations of religion, that is, to bodily rituals, objects and the senses, including dreams and visions. Instead of abstracting religion from its context, these material mediations always need to be embedded and critically analysed within political, economic and cultural relations of power.
Thirdly, ancestral spirits. How we might then analyse beliefs, rituals and institutions that centre on ancestral spirits? We have already seen how Chidester interpreted 19th-century Zulu dreams, in which ancestral spirits instructed descendants to sacrifice to them and bring them home (the ukubuyisa ritual), as part of his analysis of the triple mediation of Tylor's animist theory. Elsewhere in his work, specifically in Religions of South Africa (1992) and Wild Religion (2012), he analysed, in an exemplary way, not only the domestic and political function of ancestral religion in traditional rural settings but also its transformation as black migrants moved to urban peripheries to work in mines, up to its resurgence in post-apartheid projects of nation building (e.g. at Freedom Park and the 2010 World Cup). Emphasising the mobility of sangomas in traditional and urban settings, he has paid particular attention to the changing role of ancestral religion in Credo Mutwa's long life from collaborating with apartheid ideologues up to his engagement with conspiracy theorists in cyberspace and the contestations around white sangomas.
Fourthly, ethics. Do researchers need to take an ethical stance, in cases where religions -including African indigenous religions -create and sustain asymmetrical power relations of class, gender and race? Using critical theories from a postcolonial South African location, Chidester has certainly exposed the inhumane presentation of indigenous religions in Tylor's evolutionary theory and has moved us beyond Durkheim's naive functionalist view of religion as a binding force. Chidester, by using South African examples, has demonstrated the potential of religion not only to include but also to exclude. He has, furthermore, shown from this location and history how sacred beliefs, practices and institutions have justified political oppression and economic exploitation, but have also been instrumentalised to challenge such systemic injustices. In a response to Gerardus van der Leeuw's phenomenological approach to religion, Chidester (1994) argued for a critical phenomenology that would focus on precisely these contested power relations. However as mentioned in this study, Chidester only occasionally made his ethical stance explicit. It is at this point, as might be argued with reference to African indigenous religion and monuments as sacred places (Strijdom 2011(Strijdom , 2018, that scholars of religion need to argue and make their normative frameworks transparent rather than pretend to offer a neutral description of a range of critical judgements of religious phenomena within changing political and economic contexts.