Rumsey, Orpheus.jpg

Beyond Borderlands is a forum for the artistic and critical exploration of topics relating to esotericism, paranormality, and the culturally weird.


Trends in the Anthropology of the Supernatural

Attention: open in a new window. PDFPrintE-mail

Definitions, Origins, Functions and Experiences:
Trends in the Anthropology of the Supernatural from Tylor to Turner

The branch of anthropology that most frequently encounters the supernatural is the anthropology of religion. Religion has been a key concern for anthropologists since the very dawn of the discipline in the mid-Nineteenth Century. In light of the apparent diversity of forms that religion takes throughout the world—from the monotheism characteristic of Christianity, Judaism, and Islam to the polytheism of Hinduism and Buddhism and all the many variations and varieties in between—one of the first tasks facing the early anthropologists was to try to develop a standard definition of religion. At first glance this might seem like an easy task, but a generally accepted definition has, even today, yet to be devised. The problem lies in the complexity of the various phenomena usually classified as religious in nature, and in understanding how all of these disparate parts relate to one another. For example, religion may be defined in terms of the beliefs of a certain people or in terms of their practices—that is, their rituals, rites and performances. Religion might also be interpreted politically and economically, or described using the language of psychology and philosophy. This complexity, combined with huge cultural variation, makes the development of an all–encompassing definition and theory of religion a particularly difficult task (Boyer 2001, 2–3).

Defining Religion:

Edward Burnett Tylor (1832–1917), widely regarded as one of the founding fathers of modern anthropology, realized that any definition of religion would have to be inclusive of the broad spectrum of religious ideas present throughout the human world. Tylor was shocked to read in the reports of pioneering explorers and missionaries that many newly discovered societies were described as possessing no religion at all, despite their apparent preoccupation with spirits, demons, and ancestors. This, he thought, arose from too narrow a definition of what religion entails. He  argued that if our definition of religion is built around belief in a supreme deity, judgment after death, or the adoration of idols (which are hallmarks of both classical and contemporary European religions), the beliefs of a great many non-European people would immediately be excluded from the category of religion. To Tylor this simply did not make sense. The problem with this sort of definition was that it was based upon a particular development of religion, namely a Judeo-Christian development, and not upon religion itself. In order to counter this apparent bias, Tylor defined religion, in its simplest terms, as the belief in spiritual beings, a common trait of which he found ample evidence in the vast majority of ethnographic documents he read. Tylor’s definition of religion, therefore, highlighted the significance of belief, as well as the supernatural objects of these beliefs (Tylor 1930). Another definition of religion was offered by Émile Durkheim (1858–1917), a contemporary of Tylor’s, who argued that belief in spiritual beings, although common to many religions, could not be considered a minimum definition of religion because there are religious traditions that do not hold such beings as centrally important. To illustrate this point Durkheim gives the example of Buddhism, which does not consider gods and spirits to be central to its beliefs (though it does not exclude them), but rather emphasizes the Buddha’s Four Noble Truths (dukkha, “suffering”; samudaya, “craving”; nirodha, “the end of suffering”; and the Eightfold Path to end suffering) as its main creed. Durkheim suggested, therefore, that religion could best be defined as a unified system of beliefs and practices relative to sacred things (rather than supernatural beings specifically) and as beliefs and practices that unite its adherents in a single moral community. To Durkheim, religion was to do with the sacred, which he defined as things set apart and surrounded by prohibitions. For Durkheim, then, the sacred did not have to include supernatural concepts; his definition of religion could, for example, equally be applied to other social phenomena, as long as they were treated with a special kind of respect and separated from normal everyday life by certain prohibitions.  Above all of this, Durkheim thought of religion as a social and communal phenomenon (Durkheim 2008); we will discuss the influence of Durkheim’s sociological view of religion shortly.

Another definition of religion was proposed more recently by Clifford Geertz (1926–2006), who suggested that religion is primarily a symbolic cultural system. For Geertz, religion is a set of symbols that perform particular functions within a society. Geertz’s definition reads:

Religion is (1) a system of symbols, which (2) acts to establish powerful, pervasive and long-lasting moods and motivations in men by (3) formulating conceptions of a general order of existence and (4) clothing these conceptions with such an aura of factuality that (5) the moods and motivations seem uniquely realistic (Geertz 1973, 87–125).

In considering religion as a set of symbols, Geertz shifts the focus of defining religion away from purely sociological and psychological factors (though these are also included in his definition) towards a more holistic view of the phenomenon. Geertz emphasizes religion as a lived–in system, comprising the worldview of its adherents, and shaping their experience of the world.

It is clear, then, that a definition of religion must include a broad range of supernatural conceptualizations, address both social and psychological factors, and emphasize the holistic, and embedded, nature of any given religious worldview in the lives of its adherents.

Theories on the Origins of Religion:

Perhaps more pertinent to the anthropology of the supernatural than definitions of religion are the various theories that have been proposed on the origin of religion, because these theories seek to explain how and why religion arises. In his book Theories of Primitive Religion, anthropologist E.E. Evans-Pritchard (1902–1973), split such theories into two distinct sets: psychological theories and sociological theories. These categories are still broadly applicable today (Evans-Pritchard, 1972). For the benefit of simplicity, then, we will look at a selection of anthropological theories on the origin of religion within Evans-Pritchard’s framework, and will begin with the psychological theories.

1) Psychological & Cognitive Theories:

As we have already seen, E.B. Tylor’s minimum definition of religion was the belief in spiritual beings. He called this belief animism, from the Latin root word anima (meaning “soul”), and suggested that it was from this belief in the existence of spirits that all religious ideas ultimately stemmed. In order to account for the origins of the spiritual beings in which early humans believed, Tylor proposed a model based on the misinterpretation of experiences such as dreaming and trance states. For example, Tylor argued that early humans might have mistaken their meetings with deceased acquaintances in dreams as real encounters with real people. From such experiences, he suggested, early humans posited the existence of a non-physical component of the person (spirit/soul) that could continue to exist after the death of the physical body. Tylor further reasoned that primitive humans expanded this idea to apply to other aspects of the world: attributing spirits or souls to animals, plants, and other natural phenomena such as the wind, lightning, mountains, rivers, and sun, amongst many others, which often seemed to possess a consciousness of their own. Tylor’s definition of animism, therefore, suggests that supernatural beliefs arise from an attempt to make sense of unusual experiences and to explain the seemingly conscious activities of animals, plants, and other natural phenomena.

Another quite different psychological theory of the origin of religion is the psychological functionalism of Bronislaw Malinowski (1884–1942). For Malinowski, supernatural beliefs developed out of a need for psychological stability in an unpredictable world. To illustrate this point, Malinowski considered the performance of fertility rituals, which he suggests serve the function of reassuring the individual that they have done all that is within their power to ensure a good harvest. Malinowski is eager to point out, however, that dependence upon magic is not evidence of a lack of “practical,” “scientific” knowledge, but rather that magical rituals are a supplement to this knowledge. Supernatural beliefs, according to this functionalist perspective, serve an evolutionary purpose in enabling humans to better cope with the psychological stresses of everyday life. They are, therefore, purely psychological phenomena, and might best be understood as psychological defense mechanisms (Malinowski 1948).

Structuralist anthropology, pioneered by the French anthropologist Claude Lévi-Strauss (1908-2009), seeks to understand human culture in terms of deep structures within the human mind. These structures inform the way in which we experience the world around us, and become apparent in our cultural constructions (Leach 1970, 21). The classic example of such a ‘deep structure’ is the propensity towards interpreting the world through binary oppositions: positive and negative; good and evil; light and dark; hot and cold; raw and cooked; clean and unclean; and so on. For Lévi-Strauss, then, religions, and mythology more specifically, are cultural manifestations of the underlying structures of the human mind. Religious and mythological systems represent a means of understanding the world through symbols and conceptual constructs that have their roots deep within our minds, and are expressions, at a very fundamental level, of the way in which we, as human beings, categorize and understand the world around us (Lévi-Strauss 1986).

Similarly, cognitive theories on the origin of religion seek to understand religion as the result of innate functions within the human mind. In certain respects, cognitive approaches can be seen as descendants of Tylor’s approach, which considers supernatural concepts to be the products of misunderstood but natural experiences such as sleep and trance; however, they also bear important similarities to structuralist theories. One of the most influential cognitive theorists of religion is the anthropologist Stewart Guthrie, who has proposed the hypothesis that the innate human propensity for recognizing patterns in random formations (for example, seeing faces in the clouds) ultimately led to the belief that the world was populated by other minds similar to our own (Guthrie 1993). This is often referred to as the hyperactive, or hypersensitive, agency detection hypothesis (Barrett, 2000, 31–32; Boyer 2002, 76–77). Like Tylor’s theory, then, Guthrie’s cognitive theory emphasizes the misattribution of consciousness to non-conscious natural phenomena, and like Lévi-Strauss’ structuralism, cognitive theories understand religion as an expression of innate processes within the human mind.

2)  Sociological Theories:

Sociological theories of religion are not primarily concerned with the experiences of individuals as such. Rather, their emphasis is on social groups and the role that religion might perform within them. Sociological theories can, therefore, be broadly categorized as functionalist, in that they focus on understanding what it is that religion does in a given social group. Émile Durkheim’s writings are usually regarded as the starting point in the development of sociological theories of religion. Durkheim conceived of religion as a purely social phenomenon and was not primarily concerned with trying to understand the nature of the experiences that gave rise to it. For Durkheim, religion, in its very essence, was a sort of social mechanism aimed at maintaining group cohesion. In other words, by providing a group of people with a distinct set of beliefs, religion enables distinctive cultures to develop with their own shared sense of social identity. This shared identity would include, for example, belief in the same gods and performance of the same rituals, and would ultimately lead to a stronger, more unified, society.

The British anthropologist A.R. Radcliffe-Brown (1881–1955) drew on Durkheim’s sociological interpretation to develop his structural-functionalist approach to understanding social systems. According to Radcliffe-Brown’s model, human societies consist of many interrelating component parts (or social systems) that combine to produce a fully functioning whole, much like a living organism (Radcliffe-Brown 1968, 178–187). Social institutions such as religion are, therefore, interpreted as components in the social organism that serve to ensure that it does not fall apart. For Radcliffe-Brown it is not the beliefs of a society that are important; rather, it is their rites and rituals, because it is the performance of rituals that brings the social group together. Indeed, he suggests that, contrary to the approaches of psychological theorists (who tend toward the position that rituals developed from specific supernatural beliefs), it is the performance of essential rituals (such as the burial of the dead) that are ultimately explained and given meaning to by the development of supernatural beliefs. According to Radcliffe-Brown, the origins of religion can be seen in the necessary rituals performed by human groups: ritual comes first, followed by explanatory beliefs. From this perspective, then, religion is nothing more than a specialized social process with no underlying reality other than its sociological and structural efficacy.

Phenomenological and Other Theories:

The psychological and sociological theories outlined above essentially suggest that there is no underlying reality to religious and paranormal experiences (other than their psychological and sociological functions); phenomenological approaches do not. Indeed, phenomenological approaches ignore, or bracket out, the question of the reality of the experience under investigation. In other words, phenomenological approaches treat experience as experience, without attempting to explain it away. Such approaches place significant emphasis on understanding what is often called the religious experience, considering it an irreducible category in itself and an essential component in understanding religious belief. Prior to the nineteenth century, the term “religious experience” was used primarily within a theological context. Religious experience was something that happened within the framework of established religion, and as such was not investigated by social scientists as a distinct phenomenon. The American psychologist and philosopher William James (1842–1910) was one of the first to discuss religious experience from a perspective external to theology. James considered the religious experience to be a distinct class of experience defined by its fruits, or the moral transformation it produces in the experiencer. He reasoned that the characteristics of  “philosophical reasonableness and moral helpfulness,” when resulting directly from a religious experience, should be taken as evidence in favour of defining that experience as genuinely religious in nature (James[1902] 2004, 27–28).

As a psychologist James was most interested in the private thoughts and feelings associated with religious experience. He did not consider religious experiences to be “supernatural,” but rather a natural fact of human life. In James’s view, religious belief systems developed around individuals who had had religious experiences themselves. James took a phenomenological approach to his investigations and compiled numerous narrative accounts of religious experiences in his book The Varieties of Religious Experience ([1902] 2004). As part of his phenomenological approach, James undertook a comparative analysis of the experiential narratives he collected and highlighted the similarities between different accounts of religious, spiritual, and mystical experiences. In noting these similarities James was proposing what has now been called the “common core hypothesis”: the idea that there is a fairly standard “religious experience” that is interpreted differently according to the experiencer’s cultural background. Rudolf Otto (1869–1937) was a German theologian and philosopher with a particular interest in the issue of religious experience. Having traveled extensively, Otto was aware of the role of religious experience in non-Christian traditions and so had a wider perspective than many of his contemporaries in theology. Arguably, Otto's greatest contribution to the study of religious experience was his book The Idea of the Holy (1958). In it, Otto attempts to explore what he calls the non-rational element of the concept of the “holy,” that is the element of the notion that is free from morality, goodness, and other modern additions to the idea. Beneath this cultural baggage, Otto suggests, lies a  “unique original feeling response, which can be in itself ethically neutral” (Otto 1958, 6). Otto uses the term numen, or numinous, to describe this sensation, which he defines as a sui generis and irreducible mental state. For Otto, the numinous possessed a dual nature, being at once beautiful and terrifying. He referred to these two aspects as the mysterium fascinans (“beautiful mystery”) and the mysterium tremendum (“terrifying mystery”), and saw these as fundamental components of the religious experience. In distinguishing between the numinous experience and subsequent rational conceptualizations of it, Otto suggests that at its most fundamental level, religion is concerned with a particular kind of experience, and as such cannot truly be understood without an appreciation of it (Otto 1958, 6–7).

The same might also be said of other forms of supernatural belief. David J. Hufford’s examination of the Newfoundland “Old Hag” tradition is a classic example of a phenomenological approach to the study of supernatural beliefs and experiences. Hufford argued in favour of the experiential source hypothesis as a useful tool in the study of supernatural belief traditions, suggesting that rather than being purely the product of cultural influence, supernatural beliefs often have some basis in lived human experience, regardless of whether that experience was genuinely supernatural or not (Hufford 1982). Such an approach does not deny the significance of experience, nor does it necessarily pass judgment on the nature of the experiences it discusses. In other words, it remains ontologically neutral. An alternative to the standard psychological theories we have discussed in this chapter can be found in the writings of the Scottish folklorist and anthropologist Andrew Lang (1844–1912). Lang suggested, in contrast to Tylor’s misinterpretation theory, that supernatural beliefs might have their foundations in genuine anomalous experiences. Indeed, in his book The Making of Religion ([1913] 1995), Lang went so far as to hypothesize that paranormal experiences might have been major contributing factors in the early development of religious ideas. In other words, Lang suggested that supernatural beliefs need not be considered irrational if they are founded upon genuine paranormal experiences, rather than on misinterpreted experiences (de Martino 1972, 183). Lang opts, therefore, for an approach that takes seriously the possibility of ontologically real supernatural phenomena and beings. More recently, anthropologist Edith Turner has called for an approach to the study of ritual that takes seriously the beliefs and experiences of informants when conducting ethnographic fieldwork. Following her own experiences during a healing ceremony in Zambia, during which she witnessed the extraction of a malignant spirit from the back of an afflicted patient, Turner concluded that in order to truly understand and appreciate a particular belief system the anthropologist must learn to “see what the native sees” through a process of active and emotional participation in their belief systems and rituals (Turner 1993).  The approaches employed by the likes of Otto, James, Lang, Hufford, and Turner lead to an appreciation of the significant role of direct personal experience in the development of supernatural belief systems, and do not attempt to “explain them away” in overly reductionist terms. This shift towards a serious appreciation of subjective experiences in anthropology has been a gradual one inspired, in many cases, by the personal experiences of anthropologists in the field (cf. Young and Goulet 1994; Hunter 2011).


The task of interpreting and understanding humanity’s supernatural beliefs is a particularly difficult one. There is a constant tension between those theorists who take a psychological approach, focusing on the experiences and thoughts of the individual, and those who take a sociological approach, with an emphasis on the function that supernatural concepts and their associated practices perform for the wider community. Each approach, however, has something significant to offer to our understanding, but none of these approaches, taken alone, is able to provide a complete explanation. The complexity of the issue at hand demands that we take a more pluralistic approach that emphasizes the interaction between social and psychological functions, but that is also open to an examination of the phenomenology of the supernatural (that is, how the supernatural is experienced), as well as the implications of such experiences.


Barrett, J.L. 2000. “Exploring the Natural Foundations of Religion.” Trends in Cognitive Sciences, 4 (1): 29–34.

Boyer, Pascal. 2001. Religion Explained: The Human Instincts that Fashion Gods, Spirits and Ancestors. London: William Heinemann.

Boyer, Pascal. 2002. “Why Do Gods and Spirits Matter At All?” In Current Approaches in the Cognitive Science of Religion, 68–92. London: Continuum.

Durkheim, Émile. 2008. The Elementary Forms of Religious Life. Oxford: Oxford University Press.

Evans-Pritchard, Edward E. 1972. Theories of Primitive Religion. Oxford: Clarendon Press.

Geertz, Clifford. (1966) 1973. “Religion as a Cultural System.” In The Interpretation of Cultures, 87–125. New York: Basic Books.

Guthrie, Stuart. 1993. Faces in the Clouds: A New Theory of Religion. Oxford: Oxford University Press.

Hufford, David J. 1982. The Terror that Comes in the Night. Philadelphia: University of Pennsylvania Press.

Hunter, Jack. 2011. “The Anthropology of the Weird: Ethnographic Fieldwork and Anomalous Experience.” In Darklore, vol. 6, edited by Greg Taylor. Brisbane: Daily Grail Publishing.

James, William. (1902) 2004. The Varieties of Religious Experience. New York: Barnes & Noble.

Lang, Andrew. (1913) 1995. Myth, Ritual and Religion. Vol. 1. London: Senate.

Leach, Edmund. 1970. Levi-Strauss. London: Fontana.

Lévi-Strauss, Claude. 1986. Structural Anthropology. Vol. 1. Harmondsworth: Penguin Books.

Malinowski, Bronislaw. 1948. Magic, Science and Religion. Boston: Beacon Press.

de Martino, Ernesto. 1972. Magic: Primitive and Modern. London: Tom Stacey.

Otto, Rudolf. 1958. The Idea of the Holy: An Inquiry into the Non-Rational Factor in the Idea of the Divine and its Relation to the Rational. Oxford: Oxford University Press.

Radcliffe-Brown, Alfred R. 1968. Structure and Function in Primitive Societies. London: Cohen & West Ltd.

Turner, Edith. 1993. “The Reality of Spirits: A Tabooed or Permitted Field of Study.” Anthropology of Consciousness 4 (1): 9–12.

Tylor, Edward Burnett. 1930. Anthropology: An Introduction to the Study of Man and Civilization. London: C. A. Watts and Co. Ltd.

Young, David E. And Jean-Guy Goulet. 1994. Being Changed by Cross-Cultural Encounters: The Anthropology of Extraordinary Experience. Ontario: Broadview Press.