Reverse side of the New Hampshire quarter, featuring the Old Man of the Mountain. “We find human faces in the moon, armies in the clouds,” wrote David Hume in The Natural History of Religion.
Memes aside, what are some other evolutionary hypotheses for religion? It is not sufficient simply to say that people worldwide turn to religion to meet certain psychological needs, otherwise unmet: explaining great mysteries such as death, or the meaning of life, or because it provides solace, a sense of belonging, meeting our “spiritual” needs, and so forth.
The problem is that these don’t suffice as biology, which requires us to ask: Why do people need explanations for death, or for the meaning of life? Why do people need the solace that religion evidently provides, etc.? Why do people have spiritual needs? And by why, we mean: What is the evolutionary payoff? We might ask, for example, why do people eat? Answer: because they get hungry. But why do they get hungry? Because hunger is a sensation generated by natural selection, a mechanism to get people to nourish themselves when such nourishment is necessary; which is to say, when it is adaptive to eat. If people have a universal hunger for God, why is that? Maybe God is a worldwide tapeworm, generating hunger for His own sake. Or maybe—warning: snark attack!—God yearns for the kind of worldwide worship that religion generates, so He has instilled a need for religion in human beings because He is fundamentally lacking in self-esteem. If God felt better about Himself, we’d all be atheists.
Getting serious, now, lets first examine possible payoffs to individuals. Once again, it isn’t sufficient to conclude that religion provides answers to “deep questions” not otherwise answerable – unless we accept that these answers are more accurate and thus more fitness-enhancing, than those otherwise available. If that is your perspective, then you have your “scientific” answer and there’s nothing more to say.
For the rest of you, lets start with something basic: A definition of religion. Daniel Dennett came up with a good one, not only useful but pleasantly simple: Religions are “social systems whose participants avow belief in a supernatural agent or agents whose approval is to be sought.” Each part of this deserves to be “unpacked”—religions involve social systems rather than solitary activity, commitment to the supernatural as opposed to nature generally, and the seeking of approval—as opposed to, say, black magic or voodoo, for which the goal is to manipulate the supernatural for personal benefit.
A series of closely related hypotheses present themselves, all starting with traits that are otherwise adaptive, which then overshoot their mark. Start with the importance of being attuned to the world outside ourselves. As useful as it is to know when it is getting dark, or where to find the nearest water hole, these are things that nonhuman animals can manage. A higher level of cognition—and, presumably, of benefit—involves knowing (or making a good guess) about what others are doing, and why, interpreting underlying motives and attributing significance to things. A cracking sound might mean a predator stepping on a branch while sneaking up on you, or it might simply be a twig breaking in the wind. Better to assume the more consequential, even if there is a cost if it is in fact, a false alarm.
Statisticians refer to two different kinds of error. Type I errors are false positives, thinking that something is true or significant when in fact it isn’t. Type II errors are false negatives, thinking that something isn’t genuine or meaningful when in fact it is. When it comes to interpreting underlying meaning or pattern in something, a type I error is inconvenient and even potentially costly, but not that big a deal–at least, not compared to a type II error. For example, if you hear a noise and figure it’s a murderous villain when its just a tree branch scraping against your window, the resulting Type I error may cause you to lose some sleep, or to get out of bed unnecessarily. The alternative, a type II error, occurs if you hear a noise and decide it’s merely a tree branch when in fact it’s a lethal threat. For our ancestors on the African savannah who interpreted a rustling in the grass to be a snake when it was just a small rodent, such a type I error would have been troublesome but hardly lethal, whereas those who committed the opposite error—thinking it’s a mouse when its really a venomous snake—would have left fewer descendants. Hence, it’s a good bet that we’re predisposed to err on the side of false positives rather than false negatives.
Early human beings may thus have been especially prone to expand such an adaptive tendency to protect one’s self, if need be by anticipating the worst and, in the process, being prone to over-interpret the world. Add to this, as well, the payoff of delving deeply into a version of Lenin’s famous question: “Who, whom?” Who is doing what to whom? Who is planning what with—or against—whom? The result is a powerful inclination to see “agency” in the world, not only when it is really there but even when it isn’t, especially when potentially directed at ourselves and thus important to us. “We find human faces in the moon, armies in the clouds,” wrote David Hume in The Natural History of Religion, “and by a natural propensity, if not corrected by experience and reflection, ascribe malice and good will to every thing that hurts or pleases us.” Sometimes, not just to those things that hurt or please us, but to every thing, period.
Renowned anthropologist Bronislaw Malinowsky argued similarly, suggesting that religion evolved as a consequence of humanity’s restless intellect: “Like magic it [religion] comes from the curse of forethought and imagination, which fall on man once he rises above brute animal nature.”
The idea, in brief, is that human beings are especially prone to detect, or imagine that these worldly agents are directed toward ourselves, because sometimes they are, and when this is the case, better safe than sorry. The result is a human penchant for wielding an array of Hyperactive Agent Detection Devices, which aren’t devices for the detection of hyperactive agents, but rather, detection devices that are themselves hyperactive, readily perceiving “agency” in the universe, whether or not actually present. Like the viral meme hypothesis, this one is uncongenial to believers since it suggests that although Agency Detection Devices were adaptive (and probably still are), when it comes to their hyperactivity as manifested in religion, we’ve been HADD.