Why Every Human Society Invents Religion

In 1871, in his rooms at Oxford, Edward Burnett Tylor sat with a Maori carving on his desk and the page proofs of his book Primitive Culture spread out around him. He was trying to do something that sounds modest and turns out to be enormous: to write down the smallest possible definition of religion, one that would hold true for every human society anthropologists had then begun to catalogue, from the societies just outside his window to those described in traveler's reports from the far side of the world. The definition he arrived at was four words long. Religion, he wrote, is "the belief in spiritual beings." It was deliberately minimal, and it launched the comparative anthropology of religion as a field anyone could recognize.

What pushed Tylor toward that spare formula was a fact that has only grown more striking in the century and a half since. Anthropologists have now described thousands of distinct human societies, and they have never found one without religion-like practices. The forms vary almost beyond belief, but the presence of religion does not vary at all. That is the puzzle this article is about. Not why one particular people worships one particular god, but why every people, everywhere, ends up doing something we recognize as religion in the first place.

The Universality That Demands an Explanation

The first thing to grasp is how wide the variation actually is. When anthropologists say "religion-like practices," they are pointing to a loose family of things rather than a single shared belief: named spirits, rites for the ancestors, prayer or some functional equivalent of it, ritual specialists who stand apart from ordinary people, the category of taboo, sacred objects set off from everyday ones, and stories about how the world came to be. No two societies assemble these elements the same way. Some are crowded with gods; some have one; some have none in the way a Western reader would expect a god, yet still treat certain places, ancestors, or forces as charged with significance.

So the universality is not a uniformity. It would be far easier to explain if every culture believed roughly the same thing, because then you could simply say they had all inherited it from a common source. Instead, the content differs everywhere while the category persists everywhere. That combination is what makes the fact theoretically interesting rather than merely curious. Something about human beings, or human groups, or human minds, keeps producing this kind of practice independently, again and again, under wildly different circumstances. The rest of this article walks through the major attempts to say what that something is. It is worth flagging at the outset that these are not rival teams where only one can be right. A working anthropologist of religion typically keeps several of them in mind at once, because each one illuminates a different facet of the same stubborn fact.

Tylor and the Soul Inside Everything

Tylor's own answer began with that minimum definition and then asked what the simplest historical form of "belief in spiritual beings" might be. His proposal was animism, the idea that all things, not just people but animals, plants, rivers, and stones, possess souls. He thought he could see how a mind might arrive at such an idea on its own. Dreams present vivid images of absent or dead people; the body of a sleeping or dead person lies inert while something seems to have departed from it. From experiences like these, Tylor reasoned, early people inferred a separable soul, and then extended the inference outward to the whole observed world.

Tylor's account is sometimes called intellectualist, because it treats religion as a kind of early theory, a reasonable attempt by reasonable people to explain genuinely puzzling experiences with the evidence available to them. That framing was, for its time, generous. It refused to dismiss the beliefs of distant societies as nonsense and insisted instead that they were the product of ordinary human reasoning working on ordinary human data. The weakness, which later thinkers seized on, is that it makes religion sound like a private cognitive exercise, a lone mind puzzling over dreams, when so much of religious life is plainly social, performed together, in public, with feeling.

Durkheim and the Society That Worships Itself

It was Émile Durkheim who turned the question inside out. In Les formes élémentaires de la vie religieuse (1912), he reasoned that if religion really is universal, there must be something universal that every religion, underneath its particular content, is actually about. His answer was startling and has never quite stopped being startling: religion, he argued, is society worshipping itself. When people gather for a rite and feel themselves lifted into the presence of something larger, more powerful, and more enduring than any individual, they are not mistaken about the feeling, only about its source. The force pressing on them is real. It is the group itself, made vivid through shared ritual.

The piece of Durkheim's argument that has proved most durable is the distinction he placed at the foundation of all religion, the division of the world into the sacred and the profane. The profane is the realm of ordinary, workaday life; the sacred is whatever a community sets apart and surrounds with prohibition and reverence. For Durkheim this distinction is not one religious idea among others but the basic cognitive infrastructure of every human group, the very act of classification that makes collective life possible. Where Tylor saw a lone mind theorizing about souls, Durkheim saw an assembled crowd generating the categories by which it would understand itself. The two are not really answering the same question, and that is precisely why both survive.

Geertz and the Web of Symbols

By the middle of the twentieth century the discipline had taken what is often called the interpretive turn, and Clifford Geertz brought it to the study of religion. In his essay Religion as a Cultural System (1966, later reprinted as a chapter of his book The Interpretation of Cultures), Geertz set out to capture not what religion is for in some functional sense but what it does for the people inside it, how it shapes their experience from within. His definition is famously dense, and it repays slowing down. Religion, he proposed, is a system of symbols that acts to establish powerful, pervasive, and long-lasting moods and motivations by formulating conceptions of a general order of existence and clothing those conceptions in such an aura of factuality that the moods and motivations seem uniquely realistic.

Stripped of its scaffolding, the claim is this: religion works by means of symbols that simultaneously describe how the world is and tell you how to feel and act within it. It offers a picture of the general order of things, and because that picture comes to seem simply true, the dispositions it recommends, the moods of awe or comfort, the motivations to act in certain ways, feel like the only sensible response to reality. Geertz's contribution was to insist that you cannot understand a religion from the outside as a checklist of beliefs; you have to read it the way you read a text, attending to the meanings it carries for those who live inside it.

Magic in the Garden, and Minds Built to Believe

Two further threads complicate and enrich this picture. The first comes from Bronisław Malinowski, who lived in the Trobriand Islands between 1915 and 1918 and watched closely how islanders combined practical knowledge with ritual. He noticed that they used sound, sober technique when they were confident of the outcome, and turned to magic precisely where technique ran out and danger took over. They navigated the calm inner lagoon by skill alone; they performed elaborate magic before sailing the open, unpredictable sea. From this he argued, in the essays collected as Magic, Science and Religion, that magic, science, and religion coexist within any society and address different domains of life. Science handles what can be controlled, magic addresses the anxious gap where control fails, and religion speaks to the larger questions of meaning, death, and the order of the cosmos. People are not confused about which is which; they deploy each where it belongs.

The second thread arrived at the very end of the twentieth century from the cognitive science of religion. In Religion Explained (2001), Pascal Boyer offered a different kind of answer to Tylor's old question, locating the source of religion not in society or in symbols but in the ordinary architecture of the human mind. His central idea is that religious concepts succeed because they hit a precise cognitive sweet spot. They are what he calls minimally counterintuitive: mostly ordinary, with just one or two violations of our everyday expectations. A ghost is a person in nearly every respect except that it passes through walls; a statue is an ordinary object except that it hears prayers. Concepts like these are memorable because the single strange feature makes them salient, yet they are easy to reason about because everything else behaves normally. Fully ordinary ideas are forgettable, and wildly bizarre ideas are too unwieldy to transmit, so it is the minimally counterintuitive ones that spread from mind to mind and lodge in the culture. On this view, religion is universal because the mind that produces it is universal.

What the Classical Theories Cannot See

It would be tidy to stop there, with four answers neatly arranged, but contemporary anthropology has turned a more skeptical eye on the very categories these theories use. Saba Mahmood's Politics of Piety (2005), based on fieldwork with women in the mosque movement in Cairo between 1995 and 1997, makes the sharpest version of the criticism. The women she worked with cultivated piety as a deliberate project of self-formation, training their desires and dispositions toward devotion as an end in itself. Mahmood argues that a secular liberal frame, quietly built into Tylor, Durkheim, and Geertz alike, cannot quite see this for what it is. That frame tends to read piety as something else underneath: as instrumental social cohesion, or as false consciousness, or as cognitive error. It assumes that what people really want is freedom understood as the absence of constraint, and so it misreads the willing cultivation of religious constraint as either oppression or confusion.

Her point is not that the classical theories are worthless but that they carry assumptions they do not announce, and those assumptions shape what they are able to notice. This is also the moment to state plainly what the anthropology of religion does not do. It does not take a position on whether any particular gods exist. Its task is descriptive and analytic, not theological. The discipline asks what religion does, how it functions, and what it means to the people who practice it; it does not adjudicate the truth-claims of any tradition. A scholar can spend a career explaining why humans build religions without ever pronouncing on whether any of them is right, and that restraint is a feature of the method rather than a dodge.

Key Takeaways

Every documented human society possesses religion-like practices, and the puzzle is not the uniformity of religion but its universality amid staggering variation, which is why no single theory has displaced the others. Tylor (1871) defined religion minimally as belief in spiritual beings and located its origin in animism, the reasoned inference of souls from experiences like dreams and death; Durkheim (1912) inverted the question and argued that religion is society worshipping itself, with the sacred-profane distinction serving as the basic cognitive infrastructure of every group; Geertz (1966) read religion as a system of symbols that establishes durable moods and motivations by making a conception of the general order of existence feel uniquely real. Malinowski showed that magic, science, and religion coexist and divide the labor of life between control, anxiety, and meaning, while Boyer (2001) grounded religion's spread in the mind's preference for minimally counterintuitive concepts. Mahmood (2005) then turned the lens back on the theories themselves, exposing the secular liberal assumptions they smuggle in and that can misread sincere piety as cohesion, false consciousness, or error. Through all of it, the discipline holds a firm line: it explains what religion does and means for its practitioners, and it leaves the question of whether any gods exist entirely alone.