Breaking the Spell

by Daniel C. Dennett

  Words exist. What are they made of? Air under pressure? Ink? Some instances of the word “cat” are made of ink, and some are made of bursts of acoustic energy in the atmosphere, and some are made of patterns of glowing dots on computer screens, and some occur silently in thoughts, and what they have in common is just that they count as “the same” (tokens of the same type, as we philosophers say) in a system of symbols known as a language. Words are such familiar items in our language-drenched world that we tend to think of them as if they were unproblematically tangible things—as real as teacups and raindrops—but they are in fact quite abstract, even more abstract than voices or songs or haircuts or opportunities (and what are they made of?). What are words? Words are basically information packets of some sort, recipes for using one’s vocal apparatus and ears (or hands and eyes)—and brains—in quite specific ways. A word is more than a sound or a spelling. For instance, fast sounds the same and is spelled the same in English and German, but has completely different meanings and roles in the two languages. Two different words, sharing only some of their surface properties. Words exist. Do memes exist? Yes, because words exist, and words are memes that can be pronounced. Other memes are the same sort of thing—information packets or recipes for doing something other than pronouncing—behaviors such as shaking hands or making a particular rude gesture, or taking off your shoes when you enter a house, or driving on the right, or making your boats symmetrical. These behaviors can be described and taught explicitly, but they don’t have to be; people can just imitate the behaviors they see others perform. Variations in pronunciation can spread, and so can variations in cooking methods, doing the laundry, planting crops.

  There are vexatious problems about just what the boundaries of a meme are—is wearing a baseball cap backward one meme or two (wearing a cap, and putting it on backward)?—but similar problems arise for word boundaries—should we count “copping out” as one word or two?—and, indeed, for genes. The boundary conditions are crisp for single molecules of DNA, or their constituent parts such as nucleotides or codons (triplets of nucleotides, such as AGC or AGA), but genes don’t line up cleanly with these boundaries. They sometimes come apart into several separated pieces, and the reasons that biologists call the separated strings of codons parts of a single gene instead of two genes are very much the same reasons that linguists would identify “tickle [my, his, her] fancy” or “read [me, him, her] the riot act” as salient idioms, not just verb phrases composed of several words. Such yoked-together parts raise problems for anybody trying to count genes—not insurmountable, but not obvious, either. And what is copied and transmitted, in the case of both memes and genes, is information.

  I will have more to say about memes in later chapters, and since overeager meme-enthusiasts and equally overeager memedebunkers have made the topic a hot-button issue for many people, I need to protect a (relatively!) sober version of the concept from some of its friends and enemies. Not everybody need participate in this exercise of conceptual hygiene, however, so I have reprinted my basic introduction to memes—“The New Replicators,” from the recent two-volume Encyclopedia of Evolution published by Oxford University Press in 2002—as appendix A at the back of this book.14 For our purposes now, the main reason for taking the memes perspective seriously is that it permits us to look at the cui bono? question for every designed feature of religion without prejudging the issue of whether we’re talking about genetic or cultural evolution, and whether the rationale for a design feature is free-floating or explicitly somebody’s rationale. This expands the space of possible evolutionary theories, opening up room for us to consider multilevel, mixed processes, getting us away from the simplistic ideas of “genes for religion” at one extreme and “a conspiracy of priests” at the other extreme and permitting us to consider much more interesting (and more probable) accounts of how and why religions evolve. Evolutionary theory is not a one-trick pony, and when the Martians set out to theorize about Earthly religion, they have lots of options to explore, which I will swiftly sketch, in extreme versions, just to give a sense of the terrain to be explored more carefully in later chapters.

  Sweet-tooth theories: First, consider the variety of things we like to ingest or otherwise insert into our bodies: sugar, fat, alcohol, caffeine, chocolate, nicotine, marijuana, and opium for a start. In each case, there is an evolved receptor system in the body designed to detect substances (either ingested or constructed within the body, such as the endorphins or endogenously created morphine analogues) that these favorites have in high concentration. Over the ages, our clever species has gone prospecting, sampling just about everything in the environment, and after millennia of trial and error has managed to discover ways of gathering and concentrating these special substances so that we can use them to (over) stimulate our innate systems. The Martians may wonder if there are also genetically evolved systems in our bodies that are designed to respond to something that religions provide in intensified form. Many have thought so. Karl Marx may have been more right than he knew when he called religion the opiate of the masses. Might we have a god center in our brains along with our sweet tooth? What would it be for? What would pay for it? As Richard Dawkins puts it, “If neuroscientists find a ‘god center’ in the brain, Darwinian scientists like me want to know why the god center evolved. Why did those of our ancestors who had a genetic tendency to grow a god center survive better than rivals who did not?” (2004b, p. 14).

  If any such evolutionary account is correct, then those with a god center not only survived better than those without one; they tended to have more offspring. But we should carefully set aside the anachronism involved in thinking of this hypothesized innate system as a “god center,” since its original target may have been quite unlike the intense stuff that turns it on today—we don’t have an innate chocolate-ice-cream center in the brain, after all, or a nicotine center. God may just be the latest and most intense confection that triggers the whatsis center in so many people. What benefit accrued to those who satisfied their whatsis craving? It could even be that there isn’t and never has been any actual target in the world to obtain, but just an imaginary or virtual target, in effect: it’s been the seeking, not the getting, that has had a fitness advantage. In any case, if the need, or at least the taste, for this still-unidentified treasure has become a genetically transmitted part of human nature, we tamper with it at our peril.

  Theories in this family raise some interesting possibilities. Both sugar and saccharine trigger our sweet-tooth system. Are there religion substitutes to be found or concocted by clever psychoengineers? Or—even more interesting—are religions themselves a kind of saccharine for the brain, less filling or debilitating or intoxicating than the original and potentially harmful target? Is religion itself a subspecies of folk medicine, in which we self-medicate for relief, using therapies honed by thousands of years of trial-and-error development? Is there genetic variation in religious sensitivity, like the huge genetic variation recently discovered among human beings in taste and olfaction? Those of us who can’t stand cilantro have a gene for an olfactory receptor that cilantro lovers don’t share. Cilantro “tastes” rather like soap to us. William James speculated a hundred years ago that he—but not everybody—had a brute need for religion: “Call this, if you like, my mystical germ. It is a very common germ. It creates the rank and file of believers. As it withstands in my case, so it will withstand in most cases, all purely atheistic criticism” (letter to Leuba, quoted in introduction to James, 1902, p. xxiv). James’s mystical germ might actually be a mystical gene. Or it might be, just as he said, a mystical germ, something that spread from person to person not “vertically” (by descent from parents) but “horizontally,” by infection.

  Symbiont theories: Religions might turn out to be species of cultural symbionts that manage to th
rive by leaping from human host to human host. They may be mutualists—enhancing human fitness and even making human life possible just as the bacteria in our gut do. Or commensals—neutral, neither good for us nor bad for us, but along for the ride. Or they might be parasites: deleterious replicators that we would be better off without—at least so far as our genetic interests are concerned—but that are hard to eliminate, since they have evolved so well to counter our defenses and enhance their own propagation. We can expect that cultural parasites, like microbial parasites, exploit whatever preexisting systems come in handy. The sneezing reflex, for instance, is in the first place an adaptation for ridding the nasal passages of foreign irritants, but when a germ provokes sneezing, it is typically not the sneezer but the germ that is the principal beneficiary, getting a high-energy launching into a neighborhood where other potential hosts can take it in. Spreading germs and spreading memes may exploit similar mechanisms, such as irresistible urges to impart stories or other items of information to others, enhanced by traditions that heighten the length, intensity, and frequency of encounters with others who might be likely hosts.

  When we look at religion from this perspective, the cui bono? question changes dramatically. Now it is not our fitness (as reproducing members of the species Homo sapiens) that is presumed to be enhanced by religion, but its fitness (as a reproducing—self-replicating—member of the symbiont genus Cultus religiosus). It may thrive as a mutualist because it benefits its hosts quite directly, or it may thrive as a parasite even though it oppresses its hosts with a virulent affliction that leaves them worse off but too weak to combat its spread. And the main point to get clear about at the outset is that we can’t tell which of these is more likely to be true without doing careful, objective research. Your religion probably seems obviously benign to you, and other religions may well seem to you to be just as obviously toxic to those infected by them, but appearances can deceive. Perhaps their religion is providing them with benefits that you just don’t understand yet, and perhaps your religion is poisoning you in ways that you have never suspected. You really can’t tell from the inside. That’s how parasites work: quietly, unobtrusively, without disturbing their hosts any more than is absolutely necessary. If (some) religions are culturally evolved parasites, we can expect them to be insidiously well designed to conceal their true nature from their hosts, since this is an adaptation that would further their own spread.

  These two families of theories, sweet tooth and symbiont, are not exclusive. As we have already seen with the example of the alcohol-excreting yeast, there are symbiotic possibilities that may combine several of these phenomena together. It may be that an initial craving is exploited by cultural symbionts that include both mutualist and parasitical forms. A relatively benign or harmless symbiont may mutate under some conditions into something virulent and even deadly. For millennia, people have imagined that other religions might be a form of disease or sickness, and apostates often look back on their earlier days as a period of affliction which they have somehow survived, but the evolutionary perspective allows us to see that there are just as many positive as negative scenarios once we start looking at religion as possibly a cultural symbiont. Friendly symbionts are everywhere. Your body is composed of perhaps a hundred trillion cells, and nine out of ten of them are not human cells (Hooper et al., 1998)! Most of these trillions of microscopic guests are either harmless or helpful; only a minority are worth worrying about. Many of them, indeed, are valuable helpers that we inherit from our mothers and would be quite defenseless without. These inheritances are not genetic. Some of them may be passed on via the shared bloodstream of mother and fetus, but others are picked up by bodily contact or proximity. (A surrogate mother who makes no genetic contribution to the fetus implanted in her womb nevertheless makes a major contribution to the microflora that the infant will carry with it for the rest of its life.)

  Cultural symbionts—memes—are similarly passed on to one’s offspring by nongenetic pathways. Speaking one’s “mother tongue,” singing, being polite, and many other “socializing” skills are transmitted culturally from parents to offspring, and infant human beings deprived of these sources of inheritance are often profoundly disabled. It is well known that the parent-offspring link is the major pathway of transmission of religion. Children grow up speaking their parents’ language and, in almost all cases, identifying with their parents’ religion. Religion, not being genetic, can be spread “horizontally” to nondescendants, but such conversions play a negligible role under most circumstances. A dim appreciation of this has led in the past to some crude and cruel programs of “hygiene.” If you think that religion is, all things considered, a malignant feature of human culture, a childhood disease of sorts with lingering aftereffects, the public-health policy to deal with it would be politically drastic but quite simple: inoculation and isolation. Don’t let parents give their own children a religious upbringing! This policy has been tried, on a major scale, in the former Soviet Union, with dire consequences. The rebound of religion in post-USSR Russia suggests that religion has roles to play and resources undreamt of by this simple vision.

  A completely different sort of evolutionary possibility is represented by sexual-selection theories. Perhaps religion is like a bowerbird’s bower. Male bowerbirds devote extraordinary time and effort to building and decorating elaborate structures designed to impress females of the species, who choose a mate only after assessing rival bowers carefully. This is an example of runaway sexual selection, the subvariety of natural selection in which the pivotal selective role is played by the choosy female, whose preferences may, over many generations, snowball into highly specific and onerous demands, such as the whims of peahens that oblige peacocks to grow spectacular—and spectacularly expensive and awkward—tails. (See Cronin, 1991, for a fine overview.) The bright coloration of male birds is the best-studied example of sexual selection. In these cases, an initial bias in the innate whims of females, such as a preference for blue over yellow, gets amplified by positive feedback into intensely blue males, the bluer the better. Had a majority of the females in an isolated population of the species just happened to prefer yellow over blue, the runaway selection would have ended up with bright-yellow males. There is nothing in the environment that makes yellow better than blue or vice versa except for the reigning taste of the species’ females, which exerts a powerful, if arbitrary, selection pressure.

  How might something like the runaway sexual selection process shape the extravagances of religion? In several ways. First, there might have been straightforward sexual selection by human females for religion-enhancing psychological traits. Perhaps they preferred males who demonstrated a sensitivity to music and ceremony, which could then have snowballed into a proclivity for elaborate rapture. The females who had this preference wouldn’t have had to understand why they had it; it could just have been a whim, a blind personal taste that prompted them to choose, but if the mates they chose just happened to be better providers, more faithful family men, these mothers and fathers would tend to raise more children and grandchildren than others, and both the sensitivity to ceremony and the taste for those who loved ceremony would spread. Or the same whim could have had a selective advantage only because more females shared that whim, so that sons who lacked the fashionable sensitivity to ceremony were passed over by the choosy females. (And if an influential sample of our female ancestors had happened, for no good reason, to have a taste for males who jumped up and down in the rain, we guys would now find ourselves unable to sit still whenever it rained. Girls might or might not share our tendency to jump under these conditions, but they would definitely go for guys who did—that is the implication of the classic sexual-selection hypothesis.) The idea that musical talent is the royal road to the embrace of a woman is certainly familiar; it probably sells a million guitars a year. And there may well be something to it. This could be a
genetically transmitted proclivity, with significant variation in the population, but we should also consider cultural analogues of sexual selection. The potlatch ceremonies found among the Native Americans of the Northwest are striking: ceremonial demonstrations of conspicuous generosity, in which individuals compete with one another to see who can give away the most, sometimes to the point of ruin. These customs bear the marks of having been created by a positive-feedback escalator like those that establish peacock tails and giant Irish-elk antlers. Other social phenomena also exhibit inflationary spirals of expensive and essentially arbitrary competition: tail fins on cars of the 1950s, teen-agers’ fashions, and outdoor lighting displays at Christmas are among those most often discussed, but there are others as well.

  For more than a million years, our ancestors made beautiful “Acheulean handaxes,” pear-shaped stone implements of varying size, lovingly finished and seldom showing any sign of wear and tear. Clearly our ancestors spent a lot of time and energy making these, and the design hardly changed over the eons. Large caches of hundreds and even thousands of these have been found (Mithen, 1996). The archeologist Thomas Wynne (1995) has opined that “it would be difficult to over-emphasize just how strange the handaxe is when compared to the products of modern culture.” “They’re biofacts,” said one archeologist, coining a new term, and inspiring the science writer Marek Kohn (1999) to come up with a striking hypothesis. Geofacts are what archeologists call stones that look like artifacts but aren’t—they are just the unintended product of some geological process. Kohn proposes that these handaxes may not be artifacts so much as biofacts, more like a bowerbird’s bower than a hunter’s bow and arrow, conspicuously expensive advertisements of male superiority, a ploy that was transmitted culturally, not genetically, in a tradition that dominated the battle of the sexes for a million years. The hominoids who worked so hard to participate in this competition no more needed to understand the rationale of the enterprise than do the male spiders who catch an insect and wrap it neatly in silk to present as a “nuptial gift” to females during courtship. This is a highly speculative and controversial claim, but it is not yet disproven, and it usefully alerts us to the possibilities that might otherwise elude us. Whatever the reasons for it, our ancestors lavished time and effort on apparently unused artifacts whenever they could, a precedent worth remembering when we marvel at the expense of tombs, temples, and sacrifices.