Breaking the Spell

by Daniel C. Dennett

  Why do people want to be stewards of their religions? It is obvious, isn’t it? They believe that this is the way to lead a moral life, a good life, and they sincerely want to be good. Are they right? Notice that this is not the question of whether religions have enhanced human biological fitness. Biological fitness and moral value are entirely different issues. I have postponed the fitness question until we could see that, although it is a good, empirical question, a question that we ought to try to answer, answering it will still leave wide open the question about whether we ought to be stewards of religion. With that point firmly established, let us at last consider—not answer—the question of whether, in the end, folk religions, and the organized religions they have morphed into, have conferred fitness benefits on those who practice them. This question has preoccupied anthropologists and other researchers for centuries, often because they confused it with the question of the ultimate (moral) value of religion, and there is no dearth of familiar hypotheses to explore once we’ve cleared the decks. Two of the most plausible will receive further attention in later chapters, so for now I will just acknowledge them. Dunbar (2004) summarizes one of them well:

  It is surely no accident that almost every religion promises its adherents that they—and they alone—are the “chosen of god”, guaranteed salvation no matter what, assured that the almighty (or whatever form the gods take) will assist them through their current difficulties if the right rituals and prayers are performed. This undoubtedly introduces a profound sense of comfort in times of adversity. [p. 191]

  Notice that comfort, in and of itself, would not be a fitness booster unless it also provided (as it almost certainly does) the practical advantages of resolution and confidence, in both decision-making and action. May the Force be with you! When you are faced with the often terrifying uncertainty of a dangerous world, the belief that somebody is watching over you may well be a decisively effective morale booster, capable of turning people who would otherwise be disabled by fear and indecision into stalwart agents. This is a hypothesis about individual effectiveness in times of strife, and it may—or may not—be true.

  An entirely distinct hypothesis is that participation in religion (in harrowing initiation rites, for instance) creates or strengthens bonds of trust that permit groups of individuals to act together much more effectively. Versions of this group-fitness hypothesis have been advanced by Boyer, Burkert, Wilson, and many others. It may or may not be true—indeed, both hypotheses could be true, and we should try to confirm or disconfirm them both if only for the light they will shed—no more—on the question of the moral value of religion.

  2 The ant colony and the corporation

  Religions exist primarily for people to achieve together what they cannot achieve alone.

  —David Sloan Wilson, Darwin’s Cathedral

  But what are the benefits; why do people want religion at all? They want it because religion is the only plausible source of certain rewards for which there is a general and inexhaustible demand.

  —Rodney Stark and Roger Finke, Acts of Faith

  Why do people join groups? Because they want to—but why do they want to? For many reasons, including the obvious: for mutual protection and economic security, to promote efficiency of harvesting and other necessary activities, to accomplish large-scale projects that would otherwise be impossible. But the manifest utility of these group arrangements does not in itself explain how they ever came to pass, for there are barriers to overcome, in the form of mutual fear and hostility, and the always looming prospect of opportunistic defection or betrayal. Our inability to achieve truly global cooperation in spite of persuasive arguments demonstrating the benefits to be had, and in spite of many failed campaigns intended to create enabling institutions, shows that the limited cooperation and loyalty we do enjoy is a rare achievement. We have somehow managed to civilize ourselves to some degree, in ways no other species has even attempted, so far as we can tell. Other species often form populations that cluster together in herds or flocks or schools, and it is clear why these groupings, when they occur, are adaptive. But we are not grazing animals, for instance, and among the foraging (and predating) apes that are our nearest animal relatives, the largest stable groups are generally restricted to close kin, extended families into which newcomers are admitted only after a struggle and a test. (Among chimpanzees, the newcomers are always females emigrating from their home groups to find mates; any male that tried to join another group would be summarily killed.) There is no mystery about why we, like other apes, would have evolved a craving for the company of conspecifics, but that instinct for gregariousness has its limits.

  It is remarkable that we have learned to be comfortable in the company of strangers, as Seabright (2004) puts it, and a perennially persuasive idea about religion is that it works to promote just such group cohesiveness, turning otherwise hapless populations of unrelated and mutually suspicious people into tightly knit families or even highly effective super-organisms, rather like ant colonies or beehives. The impressive solidarity achieved by many religious organizations is not in doubt, but could this explain the rise and continued existence of religions? Many have thought so, but just how could this work? Theorists of all persuasions agree that the R & D required to set up and maintain such a system has to be accomplished somehow, and there seem at first to be just two paths to choose between: the ant-colony route and the corporation route. Natural selection has shaped the design of ants over the eons, tooling the individual ant types into specialists that automatically coordinate their efforts so that a normally harmonious and vigorous colony results. There were no heroic individual ants who figured it out and implemented it. They didn’t have to, since natural selection did all the trial and error for them, and there is not now and never was any individual ant—or council of ants—to play the role of governor. In contrast, it is precisely the rational choices of individual human beings that bring a corporation into existence: they design the structure, agree to incorporate, and then govern its activities. Individual rational agents, looking out for their own interests and doing their own individual cost-benefit analyses, make the decisions that shape, directly or indirectly, the features of the corporation.

  Is the robustness of a religion, its ability to persevere and thrive in defiance of the second law of thermodynamics, like the robustness of an ant colony or a corporation? Is religion the product of blind evolutionary instinct or rational choice? Or is there some other possibility? (Might it be a gift from God, for instance?) The failure to ask—let alone answer—this question is the charge that has long been used to discredit the functionalist school of sociology initiated by Emil Durkheim. According to its critics, functionalists treated societies as if they were living things, maintaining their health and vigor by a host of adjustments in their organs, without showing how the R & D required to design and adjust these super-organisms was accomplished. This criticism is essentially the same criticism aimed by evolutionary biologists at the Gaia hypothesis of Lovelock (1979) and others. According to the Gaia hypothesis, Earth’s biosphere is itself a sort of super-organism, maintaining its various balances in order to preserve life on Earth. A pretty idea, but, as Richard Dawkins succinctly puts it:

  For the analogy to apply strictly, there would have to have been a set of rival Gaias, presumably on different planets. Biospheres which did not develop efficient homeostatic regulations of their planetary atmospheres tended to go extinct…. In addition we would have to postulate some kind of reproduction, whereby successful planets spawned copies of their life forms on new planets. [1982, 1999, p. 236]

  Gaia enthusiasts, if they want to be taken seriously, have to ask, and answer, the question of how the presumed homeostatic systems got designed and installed. Functionalists in the social sciences must assume the same burden.

  Enter David Sloan Wilson (2002) and his “multi-level selection theory” to try to save the day for
a brand of functionalism by grounding the design process in the same R & D algorithms that account for the rest of the biosphere. According to Wilson, the design innovations that work systematically to bind human groups together are the result of Darwinian descent with modification guided by differential replication of the most fit, at many levels, including the group level. In short, he accepts the challenge of showing that competition between rival groups led to the extinction of the ill-designed groups in failing competition with the better-designed groups, which were beneficiaries of free-floating rationales (to put it my way) that none of their members needed to understand. Cui bono? The fitness of the group must trump the individual fitness of its members, and if groups are going to be the ultimate beneficiaries, groups must be the competitors. Selection can go on at several levels at once, however, thanks to competitions at several levels.

  Critics have long scoffed at the functionalists’ invocation of something like mystical societal wisdom (like the imagined wisdom of Gaia), but Wilson is right to insist that there need be nothing mystical or even mysterious about Durkheimian group-friendly functions’ getting installed by evolutionary processes—if he can demonstrate group-selection processes. The distributed wisdom of an ant colony, which really is a sort of super-organism, has been analyzed in depth and detail by evolutionary biologists, and there is no doubt that evolutionary processes can shape group adaptations under special conditions like those that prevail among the social insects. But people aren’t ants, or very much like ants, and only the most regimented religious orders approach the fascistic lock step of the social insects. Human minds are hugely complex exploration devices, corrosive questioners of every detail of the world they encounter, so evolution had better add some remarkable bells and whistles to its adaptations for human groupishness if there is to be any chance of success by the group-selection route.

  Wilson thinks that competition between religious groups, with differential survival and replication of some of those groups, can generate (and “pay for”) the excellent design features we observe in religions. The opposite theoretical pole—the only alternative, or so it first appears—is occupied by the rational choice theorists, who have recently arisen to challenge the widespread presumption by social scientists that religion is some kind of lunacy. As Rodney Stark and Roger Finke (2000) note with scorn, “For more than three centuries, the standard social scientific wisdom was that religious behavior must be irrational precisely because people do make sacrifices on behalf of their faith—since, obviously, no rational person would do such a thing” (p. 42), but as they insist:

  One need not be a religious person in order to grasp the underlying rationality of religious behavior, any more than one need be a criminal in order to impute rationality to many deviant acts (as the leading theories of crime and deviance do)…. What we are saying is that religious behavior—to the degree that it occurs—is generally based on cost-benefit calculations and is therefore rational behavior in precisely the same sense that other human behavior is rational. [p. 36]

  Religions are indeed like corporations, they claim: “Religious organizations are social enterprises whose purpose is to create, maintain, and supply religion to some set of individuals and to support and supervise their exchanges with a god or gods” (p. 103). Demand for the goods that religion has to offer is inelastic; in a free market of religious choice (as in the United States, with no state religion and many competing denominations) there is vigorous competition among denominations for market dominance—a straightforward application of “supply-side” economics. But as Wilson notes in a useful comparison between his theory and theirs, even if we were to grant that now it is rational for church members to make what are basically market decisions about which religion to invest in (an assumption we will soon examine), this doesn’t answer the question about R & D:

  But how did the religion acquire its structure that adaptively constrains the choices of utility-maximizers in just the right way? We must explain the structure of the religion in addition to the behavior of individuals once the structure is in place. Were the bizarre customs consciously invented by rational actors attempting to maximize their utilities? If so, why did they have the utility of maximizing the common good of their church? Must we really attribute all adaptive features of a religion to a psychological process of cost-benefit reasoning? Isn’t a process of blind variation and selective retention possible? After all, thousands of religions are born and die without notice because they never attract more than a few members (Stark and Bainbridge, 1985). Perhaps the adaptive features of the few that survive are like random mutations rather than the product of rational choice. [p. 82]

  Wilson is right to stress the alternative of a blind variation and selective retention process, but by clinging to his radical group-selection version he misses a better opportunity: the evolutionary design process that has given us religions involves the differential replication of memes, not groups.3 Wilson briefly mentions this as an alternative, but dismisses it with hardly a glance, largely because he views its defining doctrine to be that religious features must be dysfunctional. He thinks the meme theory requires that all religious memes be (fitness-reducing) parasites, and seldom if ever fitness-neutral commensals or fitness-enhancing mutualists.4 Here Wilson is led astray by a common misunderstanding: Richard Dawkins, who coined the term meme, is no friend of religion and has often likened memes—religious memes in particular—to viruses, stressing the capacity of memes to proliferate in spite of their deleterious effects on their human hosts. Although this jarring claim needs to be considered as a major possibility, we should not forget that the vast majority of memes, like the vast majority of bacterial and viral symbionts that inhabit our bodies, are neutral or even helpful (from the perspective of host fitness). Here, then, is my mild memetic alternative to Wilson’s group-level hypothesis:

  Memes that foster human group solidarity are particularly fit (as memes) in circumstances in which host survival (and hence host fitness) most directly depends on hosts’ joining forces in groups. The success of such meme-infested groups is itself a potent broadcasting device, enhancing outgroup curiosity (and envy) and thus permitting linguistic, ethnic, and geographic boundaries to be more readily penetrated.

  Like Wilson’s more radical group-selection theory, this hypothesis can in principle account for the excellence of design encountered in religion without postulating rational designers (the religion-as-corporation route). And it can account for the fact that individual fitness is apparently subordinated to group fitness in religions. According to this theory, we don’t need to postulate group-replication tournaments but only a cultural environment in which ideas compete. Ideas that encourage people to act together in groups (the way Toxoplasma gondii encourages rats to approach cats fearlessly) will spread more effectively as a result of this groupishness than ideas that do a less effective job of uniting their hosts into armies.5

  Using the meme’s-eye view, we can unite the two “opposite” poles of theory—ant colony versus corporation—and explain the R & D of human groupishness as a mixture of blind and foresighted processes, including intermediate selection processes of every flavor of knowingness. Since people are not like ants but really quite rational, they are unlikely to be encouraged to invest heavily in group activities unless they perceive (or think they perceive) benefits worth the investment. Hence the ideas that maximize groupishness will be those that appeal, just as Stark and Finke say, to “rewards for which there is a general and inexhaustible demand.”

  An unexpected bonus of this unified perspective is that it makes elbow room for an intermediate position on the status of religion that modifies one of the most troubling features of the rational choice model. Stark and Finke and the other rational choice theorists of religion like to portray themselves as defenders of those with religious faith, saying in effect: “They’re not crazy, they’
re smart!” However, this deliberately cold-blooded rational analysis of the market for religious goods deeply offends many religious people.6 They don’t want to see themselves as cannily making a sound investment in the most effective purveyor of supernatural benefits. They want to see themselves as having set aside all such selfish considerations, as having relinquished their rational control to a higher power.

  The meme theory accounts for this. According to this theory, the ultimate beneficiaries of religious adaptations are the memes themselves, but their proliferation (in competition with rival memes) depends on their ability to attract hosts one way or another. Once allegiance is captured, a host is turned into a rational servant, but the initial capture need not be—indeed, should not be—a rational choice by the host. Memes sometimes need to be gently inserted into their new homes, overcoming “rational” resistance by encouraging a certain passivity or receptivity in the host. William James, a memeticist ahead of his time, notes the importance of this feature for some religions, and usefully draws our attention to a secular counterpart: the music teacher who admonishes the student, “Stop trying and it will do itself!” (1902, p. 206). Just let go and clear your mind, and let that little information packet, that little habit-recipe, take over!

  One may say that the whole development of Christianity in inwardness has consisted in little more than the greater and greater emphasis attached to this crisis of self-surrender pp. 210–11]…. Were we writing the story of the mind from the purely natural-history point of view, with no religious interest whatever, we should still have to write down man’s liability to sudden and complete conversion as one of his most curious peculiarities. [p. 230]