Why has religion appeared in so many human cultures?

How widespread is religion, and how old?

Very widespread. The Stanford Encyclopedia of Philosophy asks why religion arose and spread to “virtually every known human society.” In Pew’s figures for 2020, 28.8% of humanity was Christian, 25.6% Muslim and 14.9% Hindu, while 24.2% identified with no religion, though Pew notes some of them hold religious or spiritual beliefs. See our guide to atheists, agnostics and “nones.”

At Qafzeh, the Smithsonian reports the remains of as many as 15 people with 71 pieces of red ochre, a material researchers often treat as a marker of symbolic behavior. At Göbekli Tepe, UNESCO links the monuments to social events and rituals, most likely funerary ones, though recent digs also found what appear to be the remains of homes. We label these ritual readings Supported, still debated.

How have ideas about gods changed? In a 2016 study of 33 hunter-gatherer societies, Hervey Peoples and colleagues found animism in every one. Belief in an afterlife appeared in 79%, all-powerful creator gods in 39%, and such gods active in human affairs in only 15%. From a family tree built with genetic and language data, the authors inferred that animism was the oldest trait they studied, and that gods or ancestor spirits active in human affairs were absent in early humans: a reconstruction, not a record. The Stanford Encyclopedia asks why concepts of a moralizing God, one who cares how people behave, have become “so culturally successful in recent millennia.” In 1757, David Hume argued that belief in many gods came before belief in one.

What do scientists think explains it?

The Stanford Encyclopedia says there is likely no single explanation. This page sorts the main proposals into three kinds.

Ordinary minds. The cognitive science of religion often treats religion as a by-product: a side effect of mental tools that evolved for other jobs. In a 2000 review, Justin Barrett wrote that “seemingly extraordinary thoughts and behaviours can be supported by quite ordinary cognition.” One proposed tool is agency detection: we quickly sense a being with purposes behind an unclear sound or movement, because missing a real threat may cost more than a false alarm. Another is theory of mind, our habit of guessing what others think, which may make unseen minds easy to imagine.

Group life. Ara Norenzayan, author of the 2013 book Big Gods, and his colleagues argue that belief in powerful, moralizing gods, along with credible displays of faith, helped strangers cooperate, so groups holding these beliefs grew, spread or were copied over the last 10,000 to 12,000 years. (On morality, see where right and wrong come from.) Costly signaling theory adds that demanding rituals are hard to fake, so they reveal who is committed. A study of 83 U.S. communes from the 1800s found some support: in the Stanford Encyclopedia’s summary, secular communes were about three times as likely as religious ones to dissolve in a given year.

Learning from others. Researchers also propose that people take up faith from those whose actions show they mean it (credibility-enhancing displays). In five studies with more than 67,000 people, David Maij and colleagues found that seeing credible acts of faith while growing up strongly predicted belief in supernatural beings. These surveys asked people about their parents’ or caregivers’ practice, so they show a link, not cause and effect, and the authors note that shared genes could explain part of it.

Hybrid models that combine these are on the rise, the encyclopedia reports. We label the first and third ideas Supported, still debated. For the second, costly rituals have some support, but whether big gods drove the growth of large societies is an Open question.

How well have these ideas held up?

Psychology is a science, and some well-known findings here did not hold up when tested again.

• Agency detection. In six experiments with 233 people, frightening pictures, horror music and virtual reality did not make people sense more agents that were not there, and their scores were unrelated to supernatural belief.
• Mind reading. A 2012 study linked autism and weaker mentalizing, the skill of reading other minds, to less belief in God. Later studies of more than 67,000 people in three countries did not find that link, apart from a curved pattern in the U.S.
• “God is watching” and generosity. In a 2007 study, people gave more money to an anonymous stranger after unscrambling sentences with religious words. A 2016 meta-analysis (a study that pools many studies) by the original authors and colleagues found reliable effects across 93 studies, but not among nonreligious people. A preregistered replication with 650 people found no significant effect, and a 2018 follow-up found none from the hidden prime used in 2007, only a small effect when religion was named openly, and only among religious people.
• Big gods or big societies first? A 2019 Nature study that put big societies first was retracted in 2021 over its handling of missing data. Its authors stand by a corrected version (see the deep dive).

Does explaining religion explain it away?

Some people assume this science is meant to show that religion is false. The Stanford Encyclopedia calls the reality “more nuanced and more metaphysically neutral,” that is, closer to neutral about what ultimately exists, and notes that scientists who believe in God, including Barrett, work in the field.

The key idea is the genetic fallacy: judging a claim by where it came from instead of by the reasons for it. The encyclopedia reports wide agreement that a belief’s origins will not, by themselves, show that it is false. Its contrast works as an analogy: your eyes evolved, yet that does not make what you see false, while a belief picked up from an unreliable tabloid deserves doubt. The analogy’s limit is the real debate: which case is religious belief more like? Some philosophers argue that the causes studied here could produce belief whether or not any god exists. Others reply that some believers have independent reasons, such as arguments or religious experience.

How do believers and skeptics read the same facts?

• Christian. John Calvin (1509–1564) wrote that the human mind has, “by natural instinct, some sense of Deity,” and that no nation lacks the conviction that there is a God. This is often called the sensus divinitatis, or sense of divinity. The Catechism of the Catholic Church teaches that “the desire for God is written in the human heart” and calls religious expression “so universal that one may well call man a religious being.”
• Muslim. The Quran (30:30) speaks of the fitrah “upon which He has created [all] people,” which a footnote in the Sahih International translation explains as “the natural inborn inclination of man to worship his Creator.” A saying of the Prophet in Sahih al-Bukhari teaches that every child is born with this natural faith.
• Hindu. A verse of the Rig Veda (1.164.46) reads, in Ralph Griffith’s 1896 translation: “To what is One, sages give many a title.” In Chicago in 1893, Swami Vivekananda quoted a hymn saying that people’s different paths, “crooked or straight, all lead to Thee.”
• Skeptical readings. In 1757, David Hume wrote that belief in “invisible, intelligent power” is very widespread but perhaps not universal, and not uniform, so it does not spring from “an original instinct.” He traced the first religious ideas to “hopes and fears” about the events of life. Some philosophers today ask: if God cares which faith people hold, why would human minds produce so many rival religions?

Other philosophers argue that this science may support a “common consent” argument, which treats widespread belief in God as evidence for God. Which reading is right is Beyond science. For the classic arguments on each side, see the main arguments for God’s existence.

What do people often get wrong, and where can believers and skeptics agree?

“Everyone is religious.” Not quite. About a quarter of the world identifies with no religion. And if religion is defined by what it does, such as holding a society together, then finding it everywhere is “a product of one’s definition,” another Stanford Encyclopedia entry notes.

“Science has explained religion away.” Origins alone cannot show that a belief is false. “Religion is everywhere, so it must be true.” Wide belief is no guarantee; whether it counts as evidence is still debated.

• Agree: origin and truth are separate questions. Hume separated religion’s “foundation in reason” from its “origin in human nature.”
• Agree: example matters. The Catechism speaks of “the witness of others who teach him to seek God,” and the research on credible displays points the same way.

See also Can someone love science and still believe in God? and our story Science and Faith: Enemies, Strangers, or Partners?

How can you check this yourself?

• See how religion is counted. Pew’s 2012 report put 84% of the world as religiously affiliated in 2010. Its 2025 report revised that 2010 figure to 76.7%. The most important change, Pew says, was in how it measures religion in China.
• Separate the find from the reading. Ochre near bones is an observation. “Ritual” is an inference.
• Look for replications. Weigh large preregistered repeats above one striking study. In the priming research, larger studies tended to find smaller effects.
• Read each tradition in its own words. The texts quoted here are free online.