The EvolutionaryPsychology of Religion

Steven Pinker

Harvard University

Presented at the annual meeting of the Freedom fromReligion Foundation, Madison, Wisconsin, October 29, 2004, on receipt of “TheEmperor’s New Clothes Award.”

Thank you very much; this is a tremendous honor. I lookforward to displaying the Emperor proudly in my office at Harvard. It's aspecial honor to be here on the occasion that is recognizing theaccomplishments of Anne Gaylor and I'd like to express my appreciation for thewonderful work that she has done in this Foundation.

Do we have a “God gene,” or a “God module”? I'm referring toclaims that a number of you may have noticed. Just last week, a cover story of Timemagazine was called "The God Gene:Does our deity compel us to seek a higher power?" Believe it or not, somescientists say yes. And a number of years earlier, there were claims that thehuman brain is equipped with a “God module,” a subsystem of the brain shaped byevolution to cause us to have a religious belief. "Brain's God module mayaffect religious intensity," according to the headline of the LosAngeles Times. In this evening's talk, Iwant to evaluate those claims.

There certainly is a phenomenon that needs to be explained,namely religious beliefs. According to surveys by ethnographers, religion is ahuman universal. In all human cultures, people believe that the soul lives onafter death, that ritual can change the physical world and divine the truth,and that illness and misfortune are caused and alleviated by a variety ofinvisible person-like entities: spirits, ghosts, saints, evils, demons,cherubim or Jesus, devils and gods.

All cultures, you might ask? Yes, all cultures. I give youan example of a culture we're well familiar with, that of the contemporaryUnited States. The last time I checked the figures, 25% of Americans believe inwitches, 50% in ghosts, 50% in the devil, 50% believe that the Book of Genesisis literally true, 69% believe in angels, 87% believe Jesus was raised from thedead, and 96% believe in a god or a universal spirit. You've got your work cutout for you!

So what's going on? In many regards, the human mind appearsto be well-engineered. Not literally well-engineered, but it has the signs or appearance of engineering in the biologist’ssense. That is, we can see, think, move, talk, understand, and attain goalsbetter than any robot or computer. You can't go to Circuit City and buy Rosiethe Maid from "The Jetsons" and expect to it to put away the dishesor run simple errands. These feats are too difficult for human-made creations,though they're things that a five-year-old child could do effortlessly. Theexplanation for signs of engineering in the natural world is Darwin's theory ofnational selection, the only theory we've come up with so far that can explainthe illusion of design in causal terms.

The question is, how can a powerful taste for apparentlyirrational beliefs evolve? H.L. Mencken said that “the most common of allfollies is to believe passionately in the palpably not true. It's the chiefoccupation of humankind.” This poses an enigma to the psychologist.

There is one way in which religious belief could be anadaptation. Many of our faculties are adaptations to enduring properties of thereal world. We have depth perception, because the world really isthree-dimensional. We apparently have an innate fear of snakes, because theworld has snakes and they are venomous. Perhaps there really is a personal,attentive, invisible, miracle-producing, reward-giving, retributive deity, andwe have a God module in order to commune with him. As a scientist, I  like to interpret claims as testablehypotheses, and this certainly is one. It predicts, for example, that miraclesshould be observable, that success in life should be proportional to virtue,and that suffering should be proportional to sin. I don't know anyone who hasdone the necessary studies, but I would say there is good reason to believethat these hypotheses have not been confirmed. There's a Yiddish expression: "IfGod lived on earth, people would break his windows."

There have been other, more plausible attempts to explainreligion as a biological adaptation. Even though I'm far more sympathetic toDarwinian explanations of mental life than most psychologists, I don't find anyof these convincing.

The first is that religion gives comfort. The concepts of abenevolent shepherd, a universal plan, an afterlife, or just deserts, ease thepain of being a human; these comforting thoughts make us feel better. There's anelement of truth to this, but it is not a legitimate adaptationistexplanation, because it begs thequestion of why the mind should findcomfort in beliefs that are false. Saying that something is so doesn't make itso, and there's no reason why it should be comforting to think it so, when we have reason to believe it is notso. Compare: if you're freezing, being toldthat you're warm is not terribly soothing. If you're being threatened by amenacing predator, being told that it's just a rabbit is not particularlycomforting. In general, we are not that easily deluded. Why should we be in thecase of religion? It simply begs the question.

The second hypothesis is that religion brings a communitytogether. Those of you who read the cover story of Time might be familiar with this hypothesis because thegeneticist Dean Hamer, whose new book The God Gene inspired the cover story, offered this as hisDarwinian explanation of religion. Again I think again there's an element oftruth in this. Religion certainly does bring a community together. But again itsimply begs the question as to why.Why, if there is a subgoal in evolution to have people stand together to faceoff common enemies, would a belief in spirits, or a belief that ritual couldchange the future, be necessary to cement a community together? Why not justemotions like trust and loyalty and friendship and solidarity? There's no apriori reason you would expect a belief in a soul or a ritual would be asolution to the problem of how you get a bunch of organisms to cooperate.

The third spurious explanation is that religion is thesource of our higher ethical yearnings. Those of you who read the book Rockof Ages by Steven Jay Gould, who arguedthat religion and science could co-exist comfortably, are familiar with hisargument: since science can't tell us what our moral values should be, that'swhat religion is for, and each “magisterium” should respect the other. A bigproblem for this hypothesis is apparent to anyone who has read the Bible, whichis a manual for rape and genocide and destruction. God tells the Israelitesinvading all Midianite villages, “Kill all the men, kill all the kids, kill allthe old women. The young women that you find attractive, bring them back toyour compound, lock them up, shave their heads, lock them in a room for 30 daystill they stop crying their eyes out because you've killed their mom and dad,and then take her as a second or third or fourth or fifth wife." So theBible, contrary to what a majority of Americans apparently believe, is far froma source of higher moral values. Religions have given us stonings,witch-burnings, crusades, inquisitions, jihads, fatwas, suicide bombers,gay-bashers, abortion-clinic gunmen, and mothers who drown their sons so theycan happily be united in heaven.

To understand the source of moral values, we don’t have tolook to religion. Psychologists have identified universal moral sentiments suchas love, compassion, generosity, guilt, shame, and righteous indignation. Abelief in spirits and angels need have anything to do with it. And moralphilosophers such as Peter Singer (one of tomorrow’s honorees) who scrutinizethe concept of morality have shown that it is logically rooted in theinterchangeability of one's own interests and others. The world's enduringmoral systems capture in some way the notion of the interchangeability ofperspectives and interests, the idea that "I am one guy among many":the golden rule; the categorical imperative; Singer's own notion of “the expandingcircle,” John Rawls' “veil of ignorance,” and so on. A retributive, human-likedeity meting out justice doesn't have a role in our best explanations of thelogic of morality.

To answer the “why is Homo sapiens so prone to religious belief?” you first have to distinguishbetween traits that are adaptations,that is, products of Darwinian natural selection, and traits that are byproductsof adaptations, also called spandrels orexaptations. An example: Why is our blood red? Is there some adaptive advantageto having red blood, maybe as camouflage against autumn leaves? Well, that’sunlikely, and we don't need any other adaptive explanation, either. Theexplanation for why our blood is red is that it is adaptive to have a moleculethat can carry oxygen, mainly hemoglobin. Hemoglobin happens to be red whenit's oxygenated, so the redness of our blood is a byproduct of the chemistry ofcarrying oxygen. The color per se was not selected for. Another non-adaptiveexplanation for a biological trait is genetic drift. Random stuff happens inevolution. Certain traits can become fixed through sheer luck of the draw.

To distinguish an adaptation from a byproduct, first of allyou have to establish that the trait is in some sense innate, for example, thatit develops reliably across a range of environments and is universal across thespecies. That helps rule out reading, for example, as a biological adaptation.Kids don't spontaneously read unless they are taught, as opposed to spoken language, which is a plausible adaptation, because it does emergespontaneously in all normal children in all societies.

The second criterion is the causal effects of the traitwould, on average, have improved the survival or reproduction of the bearer ofthat trait in an ancestral environment -- the one in which our species spentmost of its evolutionary history, mainly the foraging or hunter-gathererlifestyle that predated the relatively recent invention of agriculture andcivilization.

Crucially, the advantage must be demonstrable by some independentlymotivated causal consequences of the putative adaptation. That is, the laws ofphysics or chemistry or engineering have to be sufficient to establish that thetrait would be useful. The usefulness of the trait can't be invented ad hoc; ifit is, you have not a legitimate evolutionary explanation but a “just-so story”or fairy tale. The way to tell them apart is to independently motivate theusefulness of the trait. An example: Via projective geometry, one can show thatby combining images from two cameras or optical devices, it is possible tocalculate the depth of an object from the disparity of the projections. If youwrite out the specs for what you need in order to compute stereoscopic depth,you find that humans and other primates seem to have exactly those specs in oursense of stereoscopic depth perception. It's exactly what engineers woulddesign if they were building a robot that had to see in depth. That similarityis a good reason to believe that human stereoscopic depth perception is anadaptation.

Likewise for fear of snakes. In all societies people have awariness of snakes; one sees it even in laboratory-raised monkeys who had neverseen a snake. We know from herpetology that snakes were prevalent in Africaduring the time of our evolution, and that getting bitten by a snake is notgood for you because of the chemistry of snake venom. Crucially, that itself isnot a fact of psychology, but it helps to establish that what is a fact ofpsychology, namely the fear of snakes, is a plausible adaptation.

Our sweet tooth is yet another example. It’s not terriblyadaptive now, but biochemistry has established that sugar is packed withcalories, and therefore could have prevented starvation in an era which foodsources were unpredictable. That makes a sweet tooth a plausible adaptation.

In contrast, it's not clear what the adaptive function ofhumor is, or of music. I think theexplanations of religion that I've reviewed have the same problem, namely nothaving an independent rationale, given an engineering analysis of why thattrait should, in principle, be useful.

The alternative, then, is that just as the redness of bloodis a by-product of other adaptations, so may our predisposition to religiousbelief. A crucial corollary of the theory of evolution is that conflicts ofinterests among organisms, of different species or of the same species, lead tothe biological equivalent of an arms race. An organism evolves more clever orlethal weapons, another organism evolves even more ingenious defenses, and soon, spiraling the process spiral. At any given stage in an arms race, a featurecan be adaptive for one organism but not for its adversaries, as long as thefirst is overcoming the defenses of the second. That's another reason why noteverything in biology is adaptive, at least not for every organism. What'sadaptive for the lion is not so adaptive for the lamb.

So a way of rephrasing the question “Why is religious beliefso pervasive?” is to ask, Who benefits? Another way of putting it is that onemust distinguish the possible benefits of religion to the producers of religious belief – the religious establishment ofshamans and priests and so on—from the benefits to the consumers of religion -- the parishioners, the flock, thebelievers. The answer might be different for the two cases. One mustdistinguish the question "What good is an inculcation of religious beliefby priests, shaman, and so on?" from the question "What good is anacceptance of religious belief by believers?"

A number of anthropologists have pointed out the benefits ofreligion to those causing other peopleto have religious beliefs. One ubiquitous component of religion is ancestorworship. And ancestor worship must sound pretty good if you're getting on inyears and can foresee the day when you're going to become an ancestor. Amongthe indignities of growing old is that you know that you're not going to bearound forever. If you plausibly convince other people that you'll continue tooversee their affairs even when you're dead and gone, that gives them anincentive to treat you nicely up to the last day.

Food taboos are also common in religious belief, and mightbe explained by the psychology of food preference and dispreference, inparticular, disgust. If you withhold a food, especially a food of animalorigin, from children during a critical period, they'll grow up grossed out atthe thought of eating that food. That's why most of us would not eat dog meat,monkey brains, or maggots, things that are palatable in other societies. Thereare often ecological reasons why food taboos develop, but there are probablyalso reasons of control. Since neighboring groups have different favored foods,if you keep your own kids from having a taste for the foods favored by yourneighbors, it can keep them inside the coalition, preventing them fromdefecting to other coalitions, because to break bread with their neighborsthey'd have to eat revolting stuff.

Rites of passage are another intelligible feature ofreligion. Many social decisionshave to be made in categorical, yes-or-no, all-or-none fashion. But a lot ofour biology is fuzzy and continuous. A child doesn't go to bed one night andwake up an adult the next morning. But we do have to make decisions such aswhether they can vote or drive or buy a gun. There's nothing magical about theage of 13 or the age of 18 or any other age. But it's more convenient toarbitrarily anoint a person as an adult on a particular, arbitrarily chosenday, than to haggle over how mature every individual is every time he wants abeer. Religious rites of passage demarcate stages of life, serving the functionthat we have given over to driver's licenses and other forms of ID. Anotherfuzzy continuum is whether someone is available as a potential romantic partneror are committed to someone else. Marriage is a useful way of demarcating thatcontinuum with a sharp line.

Costly initiations or sacrifices are also present in almostall the world's religions. A general problem in the maintenance of cooperationis how to distinguish people who are altruistically committed to a coalitionfrom hangers-on and parasites and free-riders. One way to test who's genuinelycommitted is to see who is willing to undertake a costly sacrifice. To take anexample close to home: To see whether someone is committed to an ethnic group Iam familiar with, you can say, "You've just had a baby. Please hand overyour son so I can cut some skin off his penis." That's not the kind ofthing that anyone would do unless they took their affiliation with the groupseriously. And there are far more gruesome examples from the rest of the world.

Yet another explicable feature of religion is signs ofexpertise in occult knowledge. If you're the one who knows mysterious butimportant arcane knowledge, then other people will defer to you. Even innon-religious contexts, most societies have some division of labor inexpertise, where we accord prestige and perquisites to people who know usefulstuff. So a good strategy for providers of religion is to mix some genuineexpertise -- and indeed, anthropologists have shown that the tribal shaman orwitch doctor really is an expert in herbal medicine and folk remedies – with acertain amount of hocus-pocus, trance-inducing drugs, stage magic, sumptuous robesand cathedrals, and so on, reinforcing the claim that there are worlds ofincomprehensible wonder, power, and mystery that are reachable only throughone's services.

These practical benefits take some of the mystery over whypeople like to encourage religious belief in others, without committing oneselfto a specific biological adaptation for religion. The inculcation of religiousbelief would be a byproduct of these other, baser, motives.

What about the other side of these transactions, namely the consumers?Why do they buy it? One reason is that in most cases we should defer to experts. That's in the very nature ofexpertise. If I have a toothache, I open my mouth and let a guy drill my teeth.If I have a bellyache, I let him cut me open. That involves a certain amount offaith. Of course, in these cases the faith is rational, but that deferencecould, if manipulated, lead to irrational deference, even if the larger complex of deference can be adaptive onthe whole.

There are also emotional predispositions which evolved forvarious reasons and make us prone to religious belief as a by-product. Theanthropologist Ruth Benedict summed up much of prayer when she said,"Religion is universally a technique for success.” Ethnographic surveyssuggest that when people try to communicate with God, it's not to share gossipor know-how; it’s to ask him for stuff: recovery from illness, recovery of achild from illness, success in enterprises, success in the battlefield. (And ofcourse, the Red Sox winning the World Series, which almost made me into abeliever.) This idea was summed up by Ambrose Bierce in The Devil'sDictionary, which defines "topray" as "to ask that the laws of the universe be annulled in behalfof a single petitioner, confessedly unworthy." This aspect of religiousbelief is thus a desperate measure that people resort to when the stakes arehigh and they've exhausted the usual techniques for the causation of success.

Those are some of the emotional predispositions that makepeople fertile ground for religious belief. But there also are cognitivepredispositions, ways in which we intellectually analyze the world, which havebeen very skillfully explored by the anthropologists Dan Sperber, Pascal Boyer,and Scott Atran. Anyone who is interested in the evolutionary psychology ofreligion would enjoy Pascal Boyer’s Religion Explained and Scott Atran called In Gods We Trust. Hamer's The God Gene is also good, but I am more sympatheticto Boyer and Atran.

The starting point is a faculty of human reason thatpsychologists call intuitive psychology or the “theory of mind module” –“theory” here not referring to a theory of the scientist but rather to the intuitivetheory that people unconsciously deploy inmaking sense of other people's behavior. When I try to figure out what someoneis going to do, I don't treat them as just a robot or a wind-up doll respondingto physical stimuli in the world. Rather, I impute minds to those people. I can't literally know what someoneelse is thinking or feeling, but I assume that they're thinking or feelingsomething, that they have a mind, and I explain their behavior in terms oftheir beliefs and their desires. That's intuitive psychology. There is evidencethat intuitive psychology is a distinct part of our psychological make-up. Itseems to be knocked out in a condition called autism: autistic people can beprodigious in mathematics, art, language, and music, but they have a terribletime attributing minds to other people. They really do treat other people as ifthey were robots and wind-up dolls. There's also a concerted effort underway tosee where intuitive psychology is computed in the brain. Parts of it seems tobe concentrated in the ventromedial and orbital frontal cortex, the parts ofthe brain that kind of sit above the eyeballs, as well as the superior temporalsulcus farther back.

Perhaps the ubiquitous belief in spirits, souls, gods,angels, and so on, consists of our intuitive psychology running amok. If youare prone to attributing an invisible entity called “the mind” to otherpeople’s bodies, it's a short step to imagining minds that exist independentlyof bodies. After all, it's not as if youcould reach out and touch someone else's mind; you are always making aninferential leap. It's just one extra inferential step to say that a mind isnot invariable housed in a body.

In fact the 19th-century anthropologist Edward Tyler pointedout that in some ways, there is good empirical support for the existence of thesoul, or at least there used to be, until the fairly recent advent ofneuroscience, which provides an alternative explanation for how minds work.Think about dreams. When you dream, your body is in bed the whole time, butsome part of you seems to be up and about in the world. The same thing happenswhen you're in a trance from a fever, a hallucinogenic drug, sleep deprivation,or food poisoning.

Shadows and reflections are rather mysterious, or were untilthe development of the physics of light with its explanation of thosephenomena. But they appear to have the form and essence of the person butwithout any of their actual matter.

Death, of course, is the ultimate apparent evidence for theexistence of the soul. A person may be walking around and seeing and hearingone minute, and the next minute be an inert and lifeless body, perhaps withoutany visible change. It would seem that some animating entity that was housed inthe body has suddenly escaped from it.

So before the advent of modern physics, biology andespecially neuroscience, a plausible explanation of these phenomena is that thesoul wanders off when we sleep, lurks in the shadows, looks back at us from asurface of a pond, and leaves the body when we die.

To sum up. The universal propensity toward religious beliefis a genuine scientific puzzle. But many adaptationist explanations forreligion, such as the one featured in Time lastweek, don't, I think, meet the criteria for adaptations. There is analternative explanation, namely that religious psychology is a by-product of manyparts of the mind that evolved for other purposes. Among those purposes one hasto distinguish the benefits to the producer and the benefits to the consumer.Religion has obvious practical effects for producers. When it comes to theconsumers, there are possible emotional adaptations in our desire for health, love and success, possiblecognitive adaptations in our intuitive psychology, and many aspects of ourexperience that seem to provide evidence for souls. Put these together and youget an appeal to a mysterious world of souls to bring about our fondest wishes.