How Religion Began: From the First Burials to the Axial Age

Follows: The World's Religions, One Page Each, Ritual: Why Humans Can't Quit It and Indigenous and Folk Religions: The Oldest Faiths on Earth. Companion pieces: How the Sacred Texts Were Written and Atheism, Agnosticism & Humanism, Fairly.

About 75,000 years ago, in a cave in the Zagros Mountains of what is now Iraqi Kurdistan, someone laid a dead Neanderthal behind a tall slab of rock that had fallen from the cave roof. Over generations, others were placed in the same small space, about a metre and a half long. When archaeologists published the newest of these bodies in October 2026, they counted at least five people in the cluster, and wrote that the repeated use of one marked spot for the dead supports the idea of "complex, potentially meaningful" mortuary behaviour among Neanderthals.

Potentially meaningful. Those two words are the honest heart of this subject. We can dig up what early people did with their dead, their pigments and their stones. We cannot dig up what they believed. Every account of how religion began is a bridge built from a little hard evidence and a great deal of inference, and the people who build those bridges disagree, often sharply.

There is also a second kind of answer. For billions of believers, religion did not "begin" as a human invention at all. It began when God spoke, or when the gods were always there and humans learned to see them, or when the first ancestors laid down the law of the land. That answer deserves a hearing too, and it gets one here, in its own terms, beside the scholarship and not beneath it.

This long read covers what counts as religion, what the archaeology of the deep past does and does not show, the great theories of why humans are religious, the newer science of the religious mind, the shift from forager spirits to the gods of kings and cities, the famous idea of an "Axial Age", and how believers themselves tell the story.

A note on approach: this piece describes scholarly theories and religious accounts without deciding between them. Where evidence is contested, it says so and gives the main positions. It does not rank traditions, defend them or debunk them. The scientific theories here try to explain why humans believe; none of them can settle whether what they believe is true, and most of the scientists involved say so.

In this long read

1. What counts as religion?
2. The first graves (about 100,000–34,000 years ago)
3. Painted caves and the shaman hypothesis (about 51,000–12,000 years ago)
4. Göbekli Tepe: what it may and may not show (about 9500–8000 BCE)
5. The founders' theories: Tylor, Frazer, Marx, Durkheim, Freud and James
6. The cognitive science of religion: minds built for gods?
7. Costly signals and big gods: the argument over cooperation
8. From the spirits of the forest to the gods of the city
9. The Axial Age and its critics (800–200 BCE)
10. In their own words: revelation, and how believers read the science
11. What we know, what we don't

Part 1 — What counts as religion?

Before you can ask how religion began, you have to say what it is. That turns out to be one of the hardest questions in the field.

A word with a history

The English word comes from the Latin religio. As the Stanford Encyclopedia of Philosophy's entry on the concept puts it, religio originally meant something like "scrupulousness" or "felt obligation", the conscientious care you owed because of a taboo, a promise or a curse, whether or not any god was involved. Only much later did "religion" come to mean a kind of thing in the world, a category with members called Christianity, Islam, Buddhism and so on.

Many languages had no equivalent word. Classical Chinese, Sanskrit and the languages of many Indigenous peoples had rich vocabularies for duty, ritual, law, teaching and the way of the ancestors, but no single term that drew a line around "religion" and set it apart from the rest of life. As Indigenous and Folk Religions in this library notes, for many communities what outsiders call religion is simply how you live with your land, your ancestors and the beings around you.

Three famous definitions

Scholars have tried for 150 years to pin the thing down. Three attempts frame the debate.

Belief in spiritual beings. In 1871 the English anthropologist Edward Burnett Tylor proposed, in Primitive Culture, what he called "a minimum definition of Religion, the belief in Spiritual Beings." It is simple and works for most traditions. But it struggles with Theravada Buddhism, which does not centre on a creator god, and with forms of Confucianism focused on ritual and ethics rather than on beings.

A community around the sacred. In 1912 the French sociologist Émile Durkheim defined religion by what it does: "a unified system of beliefs and practices relative to sacred things, that is to say, things set apart and forbidden — beliefs and practices which unite into one single moral community called a Church, all those who adhere to them" (The Elementary Forms of the Religious Life, trans. J. W. Swain, 1915). No spirits required. The weakness is the opposite of Tylor's: by this test, a nation with its flags, anthems and war memorials might count as a religion too.

A system of meaning. In the 1960s the American anthropologist Clifford Geertz described religion as a system of symbols tied to a conception of "a general order of existence", a picture of what reality fundamentally is, which makes people's moods and motivations feel uniquely realistic. This captures how religion shapes a whole way of seeing, but critics say it makes religion sound like a set of ideas in the head rather than something people do with their bodies, their food and their calendars.

Families, not essences

Many scholars now avoid a one-line definition. Borrowing from the philosopher Ludwig Wittgenstein's idea of "family resemblance", they treat religion as a cluster of overlapping features, none of which every religion must have. The British scholar Ninian Smart listed seven dimensions: ritual; narrative and myth; experience and emotion; doctrine and philosophy; ethics and law; social and institutional life; and material culture (buildings, art, sacred places). A tradition can be strong in some and thin in others and still clearly belong to the family.

Is "religion" a Western invention?

A more radical group of scholars argues that the very category is a product of modern European history. Jonathan Z. Smith wrote in 1982 that "there is no data for religion": the category is created by scholars' acts of comparison. Talal Asad (Genealogies of Religion, 1993) argued that the modern concept carries Christian, and specifically Protestant, assumptions, above all the idea that religion is mainly a matter of private belief. Brent Nongbri (Before Religion, 2013) traced how the word changed meaning. Their point is not that nothing is going on in temples, shrines and festivals, but that slicing human life into "religion" and "not religion" can distort cultures that never sliced it that way.

The deeper we go into the past, the less sure we can be that the people we are studying would have recognised our word "religion" at all.

That matters for everything that follows. When an archaeologist asks whether a Neanderthal burial is "religious", the question smuggles in a category built in early modern Europe. This piece uses the word anyway, in a broad family-resemblance sense: practices and beliefs that relate people to sacred, more-than-human or unseen realities, such as gods, spirits, ancestors or a cosmic order. But keep the caution in mind.

Part 2 — The first graves (about 100,000–34,000 years ago)

Burial is the oldest place where archaeologists look for something like religion, because it is one of the few behaviours that leave traces. A body deliberately placed in a dug pit, with care, perhaps with objects, suggests that the dead mattered to the living. Whether it suggests belief in an afterlife is another matter.

The Neanderthal debate

The first claimed Neanderthal burial was found in 1908 at La Chapelle-aux-Saints in southwest France. For a century sceptics argued that the body might simply have ended up in a natural hollow. In 2014 a team that had spent twelve years re-excavating the site reported in PNAS that the pit was dug by people and the body was covered quickly, evidence they said "support[s] the hypothesis of an intentional burial."

Shanidar Cave in Iraq produced the most famous case. In 1960 Ralph Solecki's team excavated a cluster of Neanderthal skeletons, one of which, Shanidar 4, became known as the "flower burial" because soil samples contained clumps of pollen. The image of Neanderthals laying flowers on their dead travelled around the world. It did not survive scrutiny. Later work suggested the pollen could have been carried in by burrowing animals or bees, and the 2026 paper from the new Cambridge-led excavations states plainly that the heavily criticised flower-burial theory "has not stood up to further scrutiny."

What has survived is something subtler. New excavations since 2015 found more individuals right beside the 1960 ones, including Shanidar Z, a woman in her mid-forties or older who lived more than 70,000 years ago and was laid in a shallow, hand-dug depression. The cluster, behind a prominent fallen boulder, was used again and again within what may have been a few generations. The researchers think that pattern is the strongest sign yet of deliberate, perhaps meaningful, treatment of the dead by Neanderthals. Other archaeologists remain unpersuaded that Neanderthal "burials" show anything symbolic.

Early Homo sapiens

Our own species has its own early graves. At Qafzeh, a cave in what is now Israel, early Homo sapiens buried their dead around 100,000 years ago. Seventy-one pieces of ochre, a red mineral pigment, were found associated with the burials, and a boy of about 12 or 13 was laid in a grave cut into the bedrock with fallow-deer antlers beside his chest. In 2025 a team reported further burials with ochre at Tinshemet Cave, also in Israel, dated to about 100,000 years ago, and argued that both Homo sapiens and Neanderthal-like groups in the region shared a set of behaviours that included "intentional human burial and the use of ochre in burial contexts."

In Africa, the oldest known human burial is a child of about two and a half to three years old from Panga ya Saidi cave in coastal Kenya, dated to 78,300 years ago, plus or minus 4,100. The study, published in Nature in 2021, showed the body was placed in a deliberately dug pit, tightly flexed, and decomposed in place under grave earth.

The Homo naledi controversy

The boldest claim concerns a different species. Homo naledi, a small-brained hominin found in the Rising Star cave system in South Africa and dated to between 335,000 and 236,000 years ago, left over 1,500 bone specimens deep inside nearly inaccessible chambers. Lee Berger and colleagues argued in eLife (2023, revised 2025) that the evidence is most compatible with "cultural burial", which would mean that meaning-making around death predates large brains. Many specialists reject this. A 2024 paper in the Journal of Human Evolution by María Martinón-Torres and colleagues was titled, bluntly, "No scientific evidence that Homo naledi buried their dead and produced rock art." This debate is live and unresolved.

What a grave can tell you

Even where burial is secure, it does not by itself reveal belief. People bury their dead to protect bodies from scavengers, to deal with smell, to keep a loved one close, or to mark a place. Grief needs no theology. The archaeologist's honest claim is narrower: by roughly 100,000 years ago some humans were treating the dead with deliberate care, sometimes with pigment and objects, and by about 34,000 years ago, at Sungir in Russia, three people were buried with more than 13,000 ivory beads, which by one estimate would have taken some 10,000 hours to make. That is consistent with ideas about what happens after death, and with social status, and with love. The trowel cannot choose between them.

A grave proves that the dead mattered. It does not tell you whether the mourners thought they were going somewhere.

Part 3 — Painted caves and the shaman hypothesis (about 51,000–12,000 years ago)

The second great body of evidence is image-making: carved figures and painted caves.

The oldest pictures

For most of the 20th century, the great painted caves of France and Spain seemed to be the beginning of human art. That picture has changed. In 2021 researchers dated a painting of a Sulawesi warty pig in Leang Tedongnge cave, Indonesia, to at least 45,500 years ago. In 2024 a new laser-based dating method pushed things back further: a scene at Leang Karampuang on the same island, showing human-like figures interacting with a pig, was painted at least 51,200 years ago, which the team called the earliest known example of representational art and visual storytelling.

In Europe, the Lion-man of Hohlenstein-Stadel in southern Germany, a figure 31 centimetres tall carved from mammoth ivory with a human-like body and a cave lion's head, dates to between about 35,000 and 41,000 years ago. Chauvet Cave in France holds animal paintings tens of thousands of years old; Lascaux's famous bulls and horses are usually dated to around 17,000–22,000 years ago.

A creature that is part human and part animal is not an image of anything anyone has seen. That is one reason the Lion-man is so often cited as evidence of imagination reaching beyond the visible world, and perhaps of beliefs about beings that cross the line between species. Perhaps. It could also be a costume, a story character or a clan emblem. Art as the Mirror of Consciousness, in this library's art section, takes up this art as a window on the human mind.

Three theories of the caves

Hunting magic. The French priest and prehistorian Henri Breuil, who sketched Lascaux soon after its discovery in 1940, championed the idea that cave paintings were sympathetic magic: painting the animal helped capture it. A problem soon emerged: the animals on the walls were not the animals on the menu. At Lascaux, reindeer were the artists' main source of food, yet only a single reindeer appears among hundreds of images.

Structure. From the 1960s André Leroi-Gourhan mapped where different animals appeared in caves and argued for a deliberate organisation of the painted space, built on a masculine/feminine duality seen in pairings such as bison and horse. It drew attention to the caves as organised wholes, but its specific scheme has proved hard to confirm.

Shamanism. The most influential modern theory came from the South African archaeologist David Lewis-Williams, whose doctoral research was on the symbolic meanings of southern San rock paintings in southern Africa. Read alongside San testimony and ethnography, much of that art, he argued, depicts what healers saw and did in trance. With Thomas Dowson, in a 1988 paper in Current Anthropology titled "The Signs of All Times", he argued that the human nervous system produces similar visual patterns in altered states of consciousness everywhere: first geometric "entoptic" shapes (dots, grids, zigzags), then those shapes turned into meaningful objects, and finally full immersion in a vivid other world. Ice Age art, they proposed, mixes all three stages, and the cave wall itself may have been seen as a membrane between this world and the spirit world. Working with the French prehistorian Jean Clottes, he applied it to the painted caves of France and Spain, and set out the full argument in The Mind in the Cave (2002).

The critics

The shamanism theory has fierce opponents. Patricia Helvenston and Paul Bahn, in a series of papers including one in the Cambridge Archaeological Journal (2003), tested the "three stages of trance" model against the neuropsychology it relies on and argued that it does not hold up. Others object to stretching an analogy from the 19th- and 20th-century San across tens of thousands of years to Ice Age Europe, and to the theory's ability to explain almost any image, which makes it hard to test.

Most specialists today would say the caves probably did not have one meaning across 30,000 years and a whole continent. Some images may have been sacred, some instructional, some decorative, some all three. What the art does establish is that by 50,000 years ago humans were making images of things that matter to them, and sometimes of things no eye has seen.

Part 4 — Göbekli Tepe: what it may and may not show (about 9500–8000 BCE)

In 1995 a German archaeologist named Klaus Schmidt began excavating a hill in southeastern Turkey called Göbekli Tepe, "Potbelly Hill". What he found rearranged textbooks.

The site

Beginning around 9500 BCE, before pottery, before metal and before fully domesticated crops, people here carved and raised great T-shaped limestone pillars inside round and oval enclosures about 20 metres across. The tallest pillars stand up to 5.5 metres high. Many are carved with arms, hands and belts, so that the pillars seem to be stylised human or superhuman figures, and with reliefs of wild animals: foxes, boars, snakes, vultures, scorpions. The site was used until about 8000 BCE and was added to UNESCO's World Heritage List in 2018. Related sites with T-pillars, including Karahan Tepe, have since been found across the region.

Schmidt's interpretation was dramatic. He saw the enclosures as the world's first temples, a sanctuary where scattered hunter-gatherer bands gathered for ritual and feasting. His preliminary report of 2000 carried the title "Zuerst kam der Tempel, dann die Stadt": first came the temple, then the city. If religion drew people together to build monuments, he suggested, the need to feed the builders may have helped drive the invention of farming. Religion would be not a result of civilisation but one of its causes.

What changed

Since Schmidt's death in 2014, work led by the German Archaeological Institute, with Lee Clare and Turkish colleagues, has complicated the picture.

• People lived there. Excavations have found smaller rectangular buildings around the enclosures that look like houses, rainwater channels feeding cisterns cut into the bedrock, and huge numbers of everyday tools.
• They processed cereals on a large scale. A 2019 study of more than 7,000 grinding stones and related tools found that most were used for processing cereals, confirmed by plant microfossils. The builders were not "pure" hunter-gatherers living only off wild game.
• The enclosures were not deliberately buried all at once, as Schmidt thought. Later work suggests landslides repeatedly damaged them, followed by repairs and rebuilding.
• Human remains were handled in special ways. Fragments of three human skulls carved with deep grooves, reported in 2017, suggest a form of "skull cult", though no graves have been found in the enclosures.

Göbekli Tepe shows that people without farming or writing could organise the labour to raise monuments of stone. It does not show what they believed about them.

How many hands?

The site's meaning depends partly on a practical question: how many people did it take to build? Schmidt argued that quarrying, moving and raising the pillars, cut from bedrock pits about 100 metres away with flint tools, was beyond the capability of a few people, and, using experiments with the stone statues of Rapa Nui as a guide, estimated that moving the pillars alone must have involved hundreds. That supported his picture of a regional gathering place where many bands came together. Later researchers working at the site have suggested instead that the structures exposed so far could have been built by 12 to 24 people in under four months, within the capacity of a single extended family or village. The difference matters. Hundreds of builders suggests a pilgrimage centre for a whole region; a dozen or two suggests a community building something special in its own backyard.

The argument also feeds a bigger one. Prehistorians agree that settled life and farming arose together in southwest Asia, but disagree about which came first. A monumental complex on a rocky hilltop, built before clear evidence of crop cultivation, has made Göbekli Tepe a centrepiece of that debate.

What it may and may not show

The deepest lesson of Göbekli Tepe may be the one drawn in The Pecking Order, in this library's sociology section: large-scale cooperation did not need farms or kings. Whether that cooperation was organised around gods, ancestors, feasts, clan identity or all of these, the stones cannot say. The archaeologist Edward Banning asked in 2011 whether the enclosures might have been elaborate houses rather than temples at all, and noted that in many traditional societies the line between a sacred building and a dwelling barely exists. That warning echoes Part 1: the people who carved these pillars may never have drawn our line between religion and everything else.

Part 5 — The founders' theories: Tylor, Frazer, Marx, Durkheim, Freud and James

The archaeology is new. The big theories are older. In the 19th and early 20th centuries, a handful of thinkers proposed answers to the question "why religion?" that still frame the debate. Each saw something real. Each has been heavily criticised. Here they are, as their defenders would put them, followed by their critics.

An 18th-century forerunner: David Hume

Before any of them, the Scottish philosopher David Hume, in The Natural History of Religion (1757), pointed to fear, hope and a tendency of the mind:

"There is an universal tendency among mankind to conceive all beings like themselves ... We find human faces in the moon, armies in the clouds." — David Hume, The Natural History of Religion, Section 3

Hume argued that, anxious about health, harvest and death and ignorant of their real causes, early people imagined invisible powers with minds like their own. Remember that sentence about faces in the moon. Two centuries later it became a research programme (Part 6).

Tylor: the soul and animism (1871)

Tylor thought religion began with reasoning about experience. Early humans saw the difference between a living body and a corpse, and saw the dead and the absent in dreams. The natural inference was that each person has a soul that can leave the body. Extended to animals, plants and places, this produced animism, which Tylor considered the root of all religion, from which polytheism and monotheism later developed. His great strength was taking early peoples seriously as thinkers, not as fools. His weakness, by modern lights, was the ladder: the assumption that "primitive" societies represented an early stage on a road that led to Victorian England. Indigenous and Folk Religions describes how scholars such as Graham Harvey have since reclaimed the word "animism" without the ladder.

Frazer: magic, religion, science (1890)

James George Frazer's The Golden Bough gathered a vast catalogue of myths and rites and proposed three stages of thought. In magic, people try to control nature through supposed laws (like produces like); when magic fails, they turn to religion, appealing to powerful beings; finally science replaces both. "The movement of the higher thought," he wrote, "has on the whole been from magic through religion to science." Frazer was enormously influential on literature and popular culture. Anthropologists later criticised his "armchair" method, which pulled customs out of context from all over the world, and his three-stage scheme, since magic, religion and science coexist in most societies, including modern ones.

Marx: the sigh of the oppressed (1844)

Karl Marx is often quoted as dismissing religion in five words. The full passage, from the introduction to his Contribution to the Critique of Hegel's Philosophy of Right (written 1843–44), is more complicated:

"Religious suffering is, at one and the same time, the expression of real suffering and a protest against real suffering. Religion is the sigh of the oppressed creature, the heart of a heartless world, and the soul of soulless conditions. It is the opium of the people." — Karl Marx (1844)

For Marx, "man makes religion, religion does not make man." Religion is a human product that both expresses real misery and soothes it, and in doing so can reconcile people to unjust conditions instead of changing them. Opium, in the 1840s, was a common painkiller as well as a drug. Marx's real target was the social world that "requires illusions". Critics answer that religion has as often fuelled rebellion as quieted it (from medieval peasant revolts to the American civil rights movement), and that a theory explaining religion purely by economic conditions cannot explain why it flourishes among the comfortable too.

Durkheim: society worshipping itself (1912)

Durkheim studied reports of Aboriginal Australian totemism, which he took to be the simplest religion known. His key observation was that what makes something sacred is not its nature but how a community treats it: set apart, protected, forbidden. Why? Because the sacred stands for the group. In the great seasonal gatherings, when dispersed clans came together to sing and dance, people felt a surge of energy he called collective effervescence, a power greater than themselves. They attributed it to the totem. But the power was real, he argued: it was society itself. "The god of the clan, the totemic principle, can therefore be nothing else than the clan itself, personified and represented to the imagination under the visible form of the animal or vegetable which serves as totem."

Durkheim's insight that ritual creates and renews social bonds is one of the most fertile ideas in social science, and modern research on synchrony and group ritual (see Ritual: Why Humans Can't Quit It) builds on it. The criticisms: his Australian data were secondhand and patchy; Aboriginal religion was not "elementary" but as complex as any; and reducing God to society seems to assume the conclusion. Believers can also turn the argument around: if communal worship produces a real experience of something greater, why must that something be only the group?

Freud: the longing for a father (1913, 1927)

Sigmund Freud offered two theories. In Totem and Taboo (1913) he proposed a speculative prehistory: a primal horde ruled by a jealous father, killed by his sons, whose guilt became the first taboos and the totem feast. He concluded "that the beginnings of religion, ethics, society, and art meet in the Oedipus complex." Almost no anthropologist accepts this story, and Freud's own sources were thin.

His second theory, in The Future of an Illusion (1927), is more lasting. Humans face a terrifying nature, death and the frustrations of civilised life. Religion answers with a protective father in heaven, a life after death and a just order. Its beliefs are "fulfillments of the oldest, strongest, and most urgent wishes of mankind." Freud called them illusions, and was careful to define the word: a belief is an illusion when wish-fulfilment is a prominent factor in its motivation, regardless of its relation to reality. That leaves open, by his own definition, whether the belief is true. Critics point out that the argument cuts both ways: people may also wish there were no God, and wishes do not make a belief false. Defenders reply that Freud explained why religion feels so compelling, which is all he claimed to do.

James: experience first (1902)

The American psychologist William James took a different route. In The Varieties of Religious Experience (1902) he defined religion, for his purposes, as "the feelings, acts, and experiences of individual men in their solitude, so far as they apprehend themselves to stand in relation to whatever they may consider the divine." He studied conversions, mystical states and saintly lives, judged them by their fruits in people's lives, and refused to reduce them to illness or to social function. Mysticism 101 in this library follows his line. His critics say he neglected the communal, ritual side that Durkheim stressed; his admirers say he took seriously what believers actually report.

The anthropologist E. E. Evans-Pritchard, whose own fieldwork began in 1926 among the Azande of the upper Nile, delivered a famous critique of this generation in Theories of Primitive Religion (1965): most of them speculated about the inner lives of peoples they had never met, often by imagining how they themselves would have reasoned in an early human's place. The field that came next tried to replace guessing with experiments.

Part 6 — The cognitive science of religion: minds built for gods?

From the 1990s, a group of anthropologists and psychologists, working independently at first, began asking a new question. Not "what is religion for?" but "what is it about human minds that makes religious ideas so easy to think, remember and pass on?" The field became known as the cognitive science of religion.

Seeing agents everywhere

The anthropologist Stewart Guthrie, in Faces in the Clouds (1993), revived Hume's insight with an evolutionary twist. Imagine hearing a rustle in the grass. If you assume it is a predator and it is only the wind, you lose a moment. If you assume it is the wind and it is a predator, you may lose your life. Natural selection should therefore favour minds that over-detect agents, that guess "someone is there" at the slightest hint. Guthrie argued that religion is, at root, systematic anthropomorphism: seeing human-like intention in the world.

The psychologist Justin Barrett named the proposed mechanism a hyperactive agency detection device, or HADD, in a 2000 paper in Trends in Cognitive Sciences, alongside a broader "theory of mind" that leads us to attribute beliefs, desires and purposes. Developmental psychologists added evidence that young children readily explain natural things by purpose and intention. Reviewing that research in 2004, Deborah Kelemen asked whether children are "intuitive theists", disposed to see nature as designed, and suggested that by around five years of age the description may have explanatory value.

Ideas that are easy to remember

The French-American anthropologist Pascal Boyer, in Religion Explained (2001), focused on the shape of religious concepts. Successful supernatural ideas, he argued, are minimally counterintuitive: they fit almost all our ordinary expectations about a category but break one or two in a striking way. A spirit is like a person in every respect (it sees, wants, remembers, gets angry), except that it has no body. A sacred tree hears prayers. Such ideas are memorable because of the violation and usable because of everything else. Boyer's claim was that religion is not one thing with one explanation but a collection of by-products of mental systems that evolved for other purposes: detecting agents, tracking social obligations, avoiding contamination, dealing with dead bodies.

The anthropologist Scott Atran, in In Gods We Trust (2002), developed a similar picture and stressed what a list of mental quirks leaves out: religious belief is bound up with emotion, ritual and deep commitment to a group, in a way that other striking fictions are not.

The cognitive science of religion asks why gods are easy to believe in. It cannot tell you whether any of them exist.

By-product or adaptation?

This produced the field's central divide.

• By-product view: religion itself was not selected for. It is a side effect, like a spandrel in architecture, of mental tools that evolved for other jobs.
• Adaptation view: religion, or parts of it, spread because it gave groups or individuals real advantages, above all by promoting cooperation and trust.
• Cultural evolution view: a synthesis increasingly popular since the 2010s. The raw materials may be by-products of cognition, but particular religious beliefs and practices were then selected by cultural, not genetic, evolution, because the groups holding them out-competed or out-reproduced others.

The cracks

Like much of psychology, the field has been tested by replication and large samples.

• In five studies with over 67,000 participants across the Netherlands, Switzerland and the United States, a team led by David Maij and Michiel van Elk found in 2017 that measures of mentalizing skill did not predict belief in supernatural agents, except for a curvilinear effect in the US. What did predict belief was having seen credible religious behaviour while growing up.
• Will Gervais and Joseph Henrich posed the "Zeus problem" (2010): Zeus has every feature that makes a god cognitively attractive, and was once worshipped by millions, yet almost nobody believes in him now. Content alone cannot explain which gods people believe in. Context, above all who you learn from, matters more.

None of this means the cognitive approach has failed. Its proponents point out that it explains why certain kinds of supernatural ideas recur across unrelated cultures, while cultural learning explains which particular gods a given person comes to believe in. The two levels fit together.

That shift led to the idea of credibility-enhancing displays. Henrich argued (2009) that humans are wary of copying beliefs from people who do not act on them. A preacher who lives modestly, fasts and gives generously is more persuasive than one who does not. Children raised among people whose actions are costly displays of faith are more likely to believe themselves. This takes us from the individual mind to the group, and to the most argued-over theory of all.

Part 7 — Costly signals and big gods: the argument over cooperation

Humans cooperate with strangers on a scale no other primate does. How did that become possible, and did religion help? This question has produced the most ambitious, and most combative, research in the field.

Costly signalling

Why would religions demand things that are expensive: fasting, painful initiations, dietary restrictions, distinctive dress, long hours of worship? Evolutionary anthropologists, notably Richard Sosis, developed the costly signalling theory of religion. Hard-to-fake displays of commitment signal to others that you are a reliable member of the group, which makes trust and cooperation easier.

Sosis tested it on an unusual population: 19th-century American communes, all of which had to solve the problem of getting people to share labour. In a study published in 2003 with Eric Bressler, using data on 83 communes, they asked whether communes that imposed costlier requirements survived longer. The results, they wrote, "support aspects of the costly signaling theory." In Sosis's broader analysis of about 200 communes, as summarised in later reviews, about 39 percent of religious communes were still functioning twenty years after founding, against about 6 percent of secular ones, and the number of costly demands predicted longevity only in the religious communes. The data are historical, and communes are a special case, but the pattern is striking.

Big gods

The social psychologist Ara Norenzayan, in Big Gods (2013), and a team including Joseph Henrich and Edward Slingerland (in a major 2016 paper in Behavioral and Brain Sciences) proposed a bigger story. Forager spirits, they noted, are often not very interested in human morality. But many later religions centre on powerful, all-knowing, moralising gods who watch and punish. Such "big gods", the argument runs, helped groups expand beyond the size at which everyone can watch everyone: "watched people are nice people." Groups with these beliefs cooperated better, grew and spread, and their gods spread with them.

There is experimental support for part of this. In 2016 Benjamin Purzycki and colleagues ran economic games with 591 people in eight communities, from Tanna in Vanuatu to the Hadza of Tanzania and the Tyva Republic in Siberia. The more participants rated their moralistic god as punitive and knowledgeable, the more coins they allocated to distant co-religionist strangers.

The Seshat fight

The historical test proved explosive. In 2019 a large team using Seshat, a databank coding hundreds of past societies, published a paper in Nature that coded 414 societies from 30 regions over 10,000 years. Its headline: moralising gods followed, rather than preceded, large increases in social complexity, appearing only after "megasocieties" of around a million people had formed. Big gods did not build big societies. If anything, rituals that standardised religion across large populations came first.

Then came the counter-attack. Bret Beheim and colleagues, including Norenzayan, Henrich and other big-gods researchers, argued in Nature that the result depended on how missing data were treated: in large parts of the dataset, "unknown" had been converted into "absent", which biased the timing. In July 2021 the authors retracted the 2019 paper, accepting that they should have coded those portions differently. Three of the paper's authors did not agree with the retraction. The original team then published a corrected and extended reanalysis (Religion, Brain & Behavior, 2023), which they said confirmed their headline finding: the largest increases in social complexity precede big gods. A separate Seshat analysis in Science Advances (2022) found that agricultural productivity and military technology best explained the rise of complex societies.

Historians joined in too, questioning how any database can code something as slippery as "moralising gods" across Egypt, China and the Andes. Where does this leave the question? Many researchers now hold a middle position that the Austronesian study pointed to: belief in some kind of supernatural punishment, by ancestors, spirits or impersonal forces like karma, may be very old and may have helped cooperation, while the specific belief in a single all-seeing moral god may have spread later, with empires. This is one of the field's most open questions.

In 2019 a big-data study said big gods came after big societies. In 2021 it was retracted. In 2023 its authors said the conclusion still stands. Science, in other words, working as it should and not yet finished.

Part 8 — From the spirits of the forest to the gods of the city

Whatever the mechanisms, the historical shape of religion changed dramatically as humans moved from foraging to farming to states.

Religion among foragers

Hunter-gatherer societies observed in the last two centuries are not fossils of the past; they have their own long histories. But they are the best window we have on how religion works in small, mobile groups. A 2016 study by Hervey Peoples, Pavel Duda and Frank Marlowe in Human Nature coded seven traits across a global sample of hunter-gatherer societies and reconstructed their likely history on a family tree of cultures. Their results suggested:

1. Animism, the sense that animals, plants and places have spirit or personhood, was present in the most recent common ancestor of present-day hunter-gatherers.
2. Belief in an afterlife came next.
3. Then shamanism, ritual specialists who enter altered states to heal or travel among spirits, and ancestor worship.
4. Ancestors or high gods who are active in human affairs, rewarding and punishing, were absent in early humans, consistent with the egalitarian character of forager life.
5. Belief in a high god, a single creator, stood apart: it appeared in some forager societies regardless of their other religious traits.

That last finding is intriguing, because it echoes an old dissent from Tylor's ladder. In 1898 the Scottish writer Andrew Lang argued that some of the simplest societies known believed in a Supreme Being, and the Catholic priest and ethnologist Wilhelm Schmidt developed this into a theory of Urmonotheismus, an original monotheism, in his multi-volume Der Ursprung der Gottesidee (1912–1955). Most scholars rejected Schmidt's broader claim that monotheism was humanity's first religion. But the observation that high gods appear among some foragers has held up.

The first farmers and their dead

With settled life came new religious forms, many centred on the dead. At Çatalhöyük in central Anatolia, a large settlement occupied from around 7100 BCE, people buried their dead beneath the floors of their houses, under platforms and hearths. In the Levant, between about 8000 and 6000 BCE, people at Jericho and other sites removed skulls from the dead, covered them with plaster modelled into faces and sometimes set shells in the eye sockets. Ancestors, it seems, stayed at home.

The idea of a single prehistoric "Mother Goddess" also comes from this era, mainly from the excavations of James Mellaart at Çatalhöyük in the 1960s. Later excavations under Ian Hodder found little support for a Mother Goddess cult, and Hodder points out that the site gives little indication of either matriarchy or patriarchy. Female figurines are real and common; what they meant is not known.

Temples, priests and kings

The next leap came with cities and writing in Mesopotamia and Egypt in the 4th and 3rd millennia BCE, covered in depth in the next long read, The Vanished Gods. In Sumer, each city belonged to a god: Enlil at Nippur, Nanna the moon god at Ur, Enki at Eridu. The temple was literally the god's house, with a household of priests, kitchens and lands. Kings ruled as the god's stewards. The Sumerian King List opens: "After the kingship descended from heaven, the kingship was in Eridug." In Egypt, the king was himself a divine figure, the living Horus. For the first time there were full-time religious specialists, written prayers, fixed calendars of festivals, and temple economies that controlled land and labour.

The first named author in history may belong to this world: Enheduanna, daughter of Sargon of Akkad, high priestess of the moon god at Ur around 2300 BCE, to whom a cycle of temple hymns is attributed (scholars debate how literally). Already, religion had professionals, texts and a canon of places to praise. And the very idea of kingship was cast in religious terms: it was something that had "descended from heaven".

The table is a tendency, not a law. Some forager peoples had moral high gods; some states had capricious gods. But the broad pattern helps explain the next and most famous claim about religious history.

Part 9 — The Axial Age and its critics (800–200 BCE)

In 1949 the German philosopher Karl Jaspers published Vom Ursprung und Ziel der Geschichte (The Origin and Goal of History). In it he proposed that the period from roughly 800 to 200 BCE was an "axis" of world history, an Achsenzeit, when, in several places at once, human thought turned in a new direction.

Jaspers's list

Jaspers pointed to an astonishing cluster:

• China: Confucius and Laozi, and the "hundred schools" of philosophy, including Mozi and Zhuangzi.
• India: the Upanishads, the Buddha, and a full range of philosophies from devotion to scepticism and materialism (and, others would add, Mahavira and the Jains).
• Iran: Zarathustra, teaching the world as a struggle between good and evil.
• Israel: the prophets, from Elijah through Isaiah and Jeremiah to the anonymous prophet scholars call Second Isaiah.
• Greece: Homer, the philosophers Parmenides, Heraclitus and Plato, the tragedians, Thucydides and Archimedes.

What they shared, Jaspers argued, was a new reflectiveness: people began to question inherited myths, to imagine a transcendent reality beyond the visible world, to ask about the individual's moral life and to think in universal terms about all humanity. The great traditions most people still live by, he suggested, have their roots here. The Good Life Across Traditions in this library meets four of these teachers.

Why it might have happened

Explanations range widely. The sociologist Shmuel Eisenstadt, building on Max Weber, studied the social and economic conditions of the "axial civilisations". Robert Bellah, in Religion in Human Evolution (2011), saw the Axial Age as the moment when "theoretic" culture, abstract and critical thinking, joined older ritual and mythic culture. In 2015 Nicolas Baumard, Pascal Boyer and colleagues argued in Current Biology that rising prosperity explained the timing: statistical modelling of "energy capture", a proxy for living standards, suggested that economic development, not political complexity or population size, predicted when ascetic, moralising traditions appeared in the Yellow and Yangtze valleys, the eastern Mediterranean and the Ganges valley.

The Axial Age may be one of history's great discoveries, or one of its great optical illusions. Serious scholars hold both views.

The critics

The concept is influential, and heavily contested.

The dates don't line up. Most scholars now place Zarathustra around 1000 BCE, give or take a century or more, with some still arguing for anywhere between 1500 and 500 BCE. On the majority view, he lived well before 800 BCE. Jesus and Muhammad, arguably as transformative as anyone on Jaspers's list, fall well after 200 BCE. Akhenaten's experiment with a single sun god in Egypt came around 1350 BCE. The edges of the "age" keep moving.

The places don't fit together. The church historian Diarmaid MacCulloch called the thesis "a baggy monster" that bundles together four very different civilisations, only two of which had much contact during the period. The Old Testament scholar Iain Provan devoted a book, Convenient Myths (2013), to arguing that the idea flattens what actually happened.

The supposed novelties appeared earlier. In 2018 a team using the Seshat databank (Daniel Mullins, Daniel Hoyer and colleagues, in the American Sociological Review) systematically tested the claim. They found that critical traits offered as evidence of an "axial" transformation appeared across Afro-Eurasia "hundreds and in some cases thousands of years prior to the proposed Axial Age." Their investigation, they wrote, "undercuts the notion of a specific 'age' of axiality."

It can carry a bias. Some critics argue that the concept quietly ranks "axial" traditions above others, casting Indigenous, Egyptian, Mesopotamian or Mesoamerican religions as "pre-axial" or merely ritualistic. Defenders reply that Jaspers himself presented the axis as shared by several civilisations precisely to avoid a Christian-centred history.

A fair summary is the one Bellah and Joas offered: the idea that the middle centuries of the first millennium BCE marked a significant transition "has become widely, but not universally, accepted." What almost everyone agrees on is narrower: in several literate civilisations of that period, teachers emerged whose words were written down and became the foundation texts of long-lasting traditions. Whether that was one global event or several local ones that look alike from a distance is still argued.

One more caution, which brings us back to Part 1. The Axial Age story, like Frazer's ladder, can make it seem that religion "progressed" from ritual to ethics, from many gods to one, from myth to philosophy. Every tradition in this series would tell its own history differently, and many would reject the very idea of progress in religion.

Part 10 — In their own words: revelation, and how believers read the science

Everything so far has been the outsider's view: religion as a human phenomenon with causes to be found. For most of the world's religious people, the most important thing about their faith's origin is not what it reveals about human minds but what it reveals about God, the gods or the nature of reality. Those accounts are not evidence in the archaeologist's sense. They are the meaning the traditions themselves give to their beginnings, and they deserve to be stated in their own terms.

Revelation

The Abrahamic traditions trace religion to God's initiative. In Jewish and Christian scripture, God calls Abraham, speaks to Moses from a burning bush and gives the covenant at Sinai (Exodus 19–20). Revelation is often quiet: when the prophet Elijah waits on the mountain, the Lord is not in the wind, the earthquake or the fire, "and after the fire a still small voice" (1 Kings 19:12, King James Version).

In Islam, the Quran is God's speech, revealed to the Prophet Muhammad through the angel Gabriel from around 610 CE. Muslim tradition holds that the first words revealed, in the cave of Hira near Mecca, were the opening of Sura 96:

"Recite thou, in the name of thy Lord who created; created man from clots of blood. Recite thou! For thy Lord is the most Beneficent, who hath taught the use of the pen; hath taught Man that which he knoweth not." — Quran 96:1–5 (trans. J. M. Rodwell, 1861)

Hindu tradition regards the Vedas as śruti, "that which is heard": eternal truth perceived by ancient seers rather than composed by them. Buddhists trace the Dharma to the Buddha's awakening, a discovery of how things are rather than a message from a god. Many Indigenous traditions locate the origin of law and ceremony in the actions of ancestral beings who shaped the land, as in the Aboriginal Australian Dreaming, a reality described by many as continuing, not past.

Religion as natural to humans: the believer's version

Some of the most striking religious accounts make the same observation as the cognitive scientists, and draw the opposite conclusion. Many traditions teach that humans are made to seek God, so of course religion is universal.

• The Quran speaks of an original human nature (fitra) shaped for faith: "Set thou thy face then, as a true convert, towards the Faith which God hath made, and for which He hath made man. No change is there in the creation of God" (Quran 30:30, Rodwell). Another verse (7:172) describes all of Adam's descendants, before birth, being asked "Am I not your Lord?" and answering "Yes, we witness it."
• In the New Testament, Paul tells the Athenians that he found an altar inscribed "TO THE UNKNOWN GOD", and that God made all nations so "that they should seek the Lord, if haply they might feel after him, and find him, though he be not far from every one of us" (Acts 17:23, 27, KJV). He also writes that God's "invisible things ... are clearly seen, being understood by the things that are made" (Romans 1:20).
• The Hebrew Bible's Ecclesiastes says God "hath set the world in their heart" (3:11, KJV), a word many modern translations render as "eternity".
• The Rigveda's famous Creation Hymn (10.129), by contrast, ends not with certainty but with a question: whether even the one who surveys creation from the highest heaven knows how it began, or perhaps does not. Doubt and wonder are themselves part of religious traditions.

How believers respond to the science of religion

Believers respond to the theories in Parts 5 to 7 in several ways.

"The explanation does not explain away." This is the most common philosophical response. Showing that a capacity has natural causes does not show that its objects are unreal. Humans have evolved capacities for mathematics and for perceiving colour; nobody concludes that numbers or colours are illusions. If God exists and wished to be known, a mind inclined to detect agency and seek meaning is just what one might expect. Justin Barrett, who named the "hyperactive agency detection device", is himself a practising Christian and has argued exactly this. Christian theology, he has written, teaches that people were made to be in relationship with God, so why wouldn't God design us to find belief in the divine natural?

"A sense of the divine." The Reformer John Calvin taught that humans have a sensus divinitatis, an implanted awareness of God. Some modern philosophers of religion, notably Alvin Plantinga, have argued that belief formed by such a faculty could be rational without arguments, if the faculty was designed by God. Sceptics reply that the same reasoning would license every tradition's beliefs, including those that contradict each other.

"Religion was there first." Some believers read the evidence of very early burial and art as confirming that humans have always been religious, and see this as consistent with their scriptures, rather than as a story of invention.

"The science is welcome, but limited." Many believers take the scholarship in their stride as an account of how religion works in human societies, while holding that why there is something to seek is a different question. How the Sacred Texts Were Written in this library makes the same distinction for scripture.

Sceptics, for their part, argue that once you can explain why people would believe in gods whether or not gods exist, belief loses its evidential support. That argument, and the replies to it, are explored in Atheism, Agnosticism & Humanism, Fairly.

The scientist asks why humans everywhere seek gods. The believer asks what it is that humans everywhere are seeking. The same evidence sits beneath both questions.

Part 11 — What we know, what we don't

Key dates

Dates before writing are approximate, many are minimum ages, and several are disputed; the parts above give the arguments.

Who's who

How to read a claim about religion's origins

New discoveries about early religion make headlines every year. A few questions help sort them.

• What exactly was found? A pit, a body, a pigment, a carving? The physical evidence is usually far less dramatic than the headline.
• Who dated it, and how? Dates in this field are often minimums, and methods change: the Sulawesi paintings got older when the dating technique improved.
• Is "religious" doing hidden work? "Burial", "shrine", "temple" and "shaman" are interpretations, not observations.
• Who disagrees? For almost every bold claim in this piece, from Homo naledi burials to Göbekli Tepe's temples to the Seshat big gods, there is a published reply. Read both.

Five open questions

1. Did Neanderthals and Homo naledi have religious ideas? The burial evidence for Neanderthals is stronger than it was twenty years ago; what it meant to them is not known. The naledi claims remain disputed.
2. Was religion an adaptation, a by-product or a cultural invention that spread because it worked? The cultural-evolution synthesis is popular, but it is hard to test against the deep past.
3. Did belief in punishing gods help build large societies, or follow them? After the Seshat retraction and reanalysis, both sides are still publishing.
4. Was there an Axial Age? Or several unrelated transformations that only look connected from a distance?
5. Can a natural history of religion say anything about religious truth? Most scientists in the field say no; philosophers on both sides still argue that it says something.

The more evidence we gather, the further back the story goes, and the more the line between "religious" and "human" blurs.

The takeaway

Religion has no single starting date. By about 100,000 years ago some humans were burying their dead with care and pigment; by 50,000 years ago they were painting scenes and carving beings that no eye had seen; by 11,000 years ago foragers were raising carved stone pillars at Göbekli Tepe. None of this tells us exactly what they believed. The great 19th- and 20th-century theories each caught something true, about reasoning (Tylor), society (Durkheim), suffering (Marx), wishes (Freud) and experience (James), and each overreached. The cognitive science of religion suggests that minds which readily detect agents and remember striking ideas find gods easy to think, while cultural evolution suggests that costly rituals and, later, moralising gods spread with the groups that held them, though when and why is still fiercely argued. The "Axial Age" is an illuminating pattern to some scholars and an illusion to others. Believers, meanwhile, tell the story as revelation and as a human longing placed there on purpose, and the science, by itself, cannot settle which story is true.

Try this: next time you catch yourself seeing a face in a cloud, a car's "expression", or feeling that a run of bad luck is "out to get you", notice it as the mental habit Hume described in 1757 and the cognitive scientists study today. Then ask the question this piece cannot answer for you: is that habit only a quirk of a primate brain, or could it also be, as many traditions say, a capacity for something real?

Sources & Further Reading

Sources
- Kevin Schilbrack, "The Concept of Religion", Stanford Encyclopedia of Philosophy (first published 28 March 2022): religio, Tylor, Durkheim, Geertz, family resemblance, Smart's dimensions, J. Z. Smith, Asad, Nongbri and the critical turn.
- E. B. Tylor, Primitive Culture, vol. 1 (1871), Project Gutenberg #70458; É. Durkheim, The Elementary Forms of the Religious Life, trans. J. W. Swain (1915), Gutenberg #41360; W. James, The Varieties of Religious Experience (1902), Gutenberg #621; J. G. Frazer, The Golden Bough (abridged 1922), Gutenberg #3623; S. Freud, Totem and Taboo, trans. A. A. Brill (1918), Gutenberg #41214; D. Hume, The Natural History of Religion (1757), Section 3, Hume Texts Online; K. Marx, A Contribution to the Critique of Hegel's Philosophy of Right, Introduction (1844), Marxists Internet Archive. Freud's The Future of an Illusion (1927) is summarised from the Wikipedia article and the translation it quotes.
- E. Pomeroy et al., "Evidence for funerary behaviour and lethal trauma in a new Neanderthal individual from Shanidar Cave", Scientific Reports 16 (October 2026), and the Shanidar Cave article (lead).
- W. Rendu et al., "Evidence supporting an intentional Neandertal burial at La Chapelle-aux-Saints", PNAS 111 (2014): 81–86.
- M. Martinón-Torres et al., "Earliest known human burial in Africa", Nature 593 (2021): 95–100.
- Y. Zaidner et al., Tinshemet Cave, Nature Human Behaviour 9 (2025): 886–901.
- L. R. Berger et al., "Evidence for deliberate burial of the dead by Homo naledi", eLife 12 (2023/2025); M. Martinón-Torres et al., "No scientific evidence that Homo naledi buried their dead and produced rock art", Journal of Human Evolution 195 (2024).
- A. Brumm et al., "Oldest cave art found in Sulawesi", Science Advances 7 (2021); A. A. Oktaviana et al., "Narrative cave art in Indonesia by 51,200 years ago", Nature 631 (2024): 814–818. Lion-man, Lascaux (including Breuil and Leroi-Gourhan), Sungir and Qafzeh details from the corresponding Wikipedia articles (leads).
- J. D. Lewis-Williams and T. A. Dowson, "The Signs of All Times", Current Anthropology 29 (1988): 201–245; P. A. Helvenston and P. G. Bahn, "Testing the 'Three Stages of Trance' Model", Cambridge Archaeological Journal 13 (2003): 213–224.
- L. Dietrich et al., "Cereal processing at Early Neolithic Göbekli Tepe", PLOS ONE 14 (2019); J. Gresky, J. Haelm and L. Clare, "Modified human crania from Göbekli Tepe", Science Advances 3 (2017); K. Schmidt, "Zuerst kam der Tempel, dann die Stadt", Istanbuler Mitteilungen 50 (2000); E. B. Banning, "So Fair a House", Current Anthropology 52 (2011). Site chronology and later research via the Göbekli Tepe article (lead, citing Clare 2020).
- J. L. Barrett, "Exploring the natural foundations of religion", Trends in Cognitive Sciences 4 (2000): 29–34; D. Kelemen, "Are children 'intuitive theists'?", Psychological Science 15 (2004): 295–301; D. L. R. Maij et al., "Mentalizing skills do not differentiate believers from non-believers, but credibility enhancing displays do", PLOS ONE 12 (2017); W. M. Gervais and J. Henrich, "The Zeus Problem", Journal of Cognition and Culture 10 (2010); J. Henrich, "The evolution of costly displays, cooperation and religion", Evolution and Human Behavior 30 (2009).
- R. Sosis and E. Bressler, "Cooperation and Commune Longevity", Cross-Cultural Research 37 (2003): 211–239; the 39%/6% figures from R. Sosis, "Religion and Intragroup Cooperation", Cross-Cultural Research 34 (2000): 70–87, as summarised by Jonathan Haidt, The Righteous Mind (2012), ch. 11, and in the Wikipedia article "Evolutionary psychology of religion".
- A. Norenzayan et al., "The cultural evolution of prosocial religions", Behavioral and Brain Sciences 39 (2016); B. G. Purzycki et al., "Moralistic gods, supernatural punishment and the expansion of human sociality", Nature 530 (2016); J. Watts et al., Proceedings of the Royal Society B 282 (2015); H. Whitehouse et al., Nature 568 (2019), and Retraction Note, Nature 595 (7 July 2021); B. Beheim et al., Nature 595 (2021); H. Whitehouse et al., "Testing the Big Gods hypothesis with global historical data: a review and 'retake'", Religion, Brain & Behavior 13:2 (2023): 124–166; P. Turchin et al., Science Advances 8 (2022).
- H. C. Peoples, P. Duda and F. W. Marlowe, "Hunter-Gatherers and the Origins of Religion", Human Nature 27 (2016): 261–282. Çatalhöyük, plastered skulls and Urmonotheismus: corresponding Wikipedia articles (leads).
- Sumerian King List, Electronic Text Corpus of Sumerian Literature (University of Oxford), 2.1.1.
- D. Mullins et al., "A Systematic Assessment of 'Axial Age' Proposals Using Global Comparative Historical Evidence", American Sociological Review 83 (2018): 596–626; N. Baumard et al., "Increased affluence explains the emergence of ascetic wisdoms and moralizing religions", Current Biology 25 (2015): 10–15; Axial Age article (lead) for Jaspers's list, MacCulloch's and Bellah and Joas's words; Zoroaster dating per the Zoroaster article citing Malandra (Encyclopaedia Iranica, 2005) and Kellens (2011).
- Scripture: King James Version (Project Gutenberg); the Quran, trans. J. M. Rodwell (1861), Project Gutenberg #2800. Justin Barrett's views as quoted (from the New York Times) in his Wikipedia biography; sensus divinitatis, Muhammad's first revelation and the life dates in the Who's who table from the corresponding Wikipedia articles (leads). E. E. Evans-Pritchard, Theories of Primitive Religion (Oxford, 1965). Golden Bough first edition 1890 per its Wikipedia article.

Go deeper
- 📘 Religion Explained — Pascal Boyer (the founding popular book of the cognitive science of religion: witty, provocative and clear on why some supernatural ideas stick)
- 📘 Big Gods: How Religion Transformed Cooperation and Conflict — Ara Norenzayan (the big-gods thesis at its strongest, by its main author; read it alongside the Seshat debate above)
- 📘 Religion in Human Evolution: From the Paleolithic to the Axial Age — Robert N. Bellah (a vast, humane synthesis by a sociologist who took both science and religion seriously)
- 📘 The Mind in the Cave — David Lewis-Williams (the shamanism theory of cave art, argued in full)
- 📘 Born Believers: The Science of Children's Religious Belief — Justin L. Barrett (the developmental evidence, by a cognitive scientist who is also a Christian)
- 📘 The Varieties of Religious Experience — William James (still the best book on religious experience; free on Project Gutenberg)
- 🌐 Stanford Encyclopedia of Philosophy — The Concept of Religion (the definitional debate, thoroughly and fairly)
- 🌐 Companion pieces in this library: The World's Religions, One Page Each, Ritual: Why Humans Can't Quit It, Indigenous and Folk Religions, Mysticism 101, Atheism, Agnosticism & Humanism, Fairly, The Pecking Order, and the next long read, The Vanished Gods