Let's try a little thought experiment, shall we?

One question that atheists invariably must face concerns the sheer prevalence of religious belief: if all religions are fundamentally mistaken about the thing that most concerns them, then why are human beings everywhere and in every time so overwhelmingly religious? Why is this mistake—which many, perhaps most, atheists think catastrophic—so nearly universal?

Since atheism came into its own a hundred and fifty years or so ago, atheists have tried various ways of answering these questions, some anthropological, some sociological, some psychological, some existential. But recently the most common explanation atheists give for religious belief is that such belief is evolutionarily favored: that is—for reasons which some atheists guess at, while others decline to speculate—religious belief in a person increases the chance that that person will pass on his or her genes to another generation. (Earlier this year Robin Marantz Henig wrote a helpful survey of some of these "adaptationist" views of religion for the New York Times.)

Now, an atheist saying this immediately has a new problem, especially if he or she thinks that religious belief produces violence and intolerance—which is what many atheists, most notably Richard Dawkins and Sam Harris, have shouted from the world's rooftops. Anyone who holds both these views is in an interesting position, to say the least. Do we say that if I am violent and intolerant toward others I am more likely to pass along my genes—perhaps because I kill or injure those who do not share my religious beliefs before they can reproduce? If we do say that, then the atheist who protests against violence and intolerance will have to argue that we should behave in ways that do not maximize the likelihood of passing along our genes.

But this is a bad situation for an atheist to be in, since he or she is likely to have trouble grounding that "should" in anything compelling, and in any case is—according to his or her own philosophy—fighting a losing battle. If religiously inspired violence and intolerance are evolutionarily adaptive, and the blind processes of natural selection are the only ones that determine reproductive survival in the long term, then people who argue against religion and its accompanying pathologies are certain to diminish in numbers and eventually become totally marginal—nothing more than the occasional maladaptive mutation. The selfish gene will ultimately, necessarily, win out over the altruistic one.

In his recent book The God Delusion Richard Dawkins has referred to religious belief as a kind of evolutionary "misfiring," but, curiously enough, uses precisely the same term to describe "the urge to kindness—to altruism, to generosity, to empathy, to pity." Now it is true that he thinks religion is a lamentable, disgusting evolutionary error, while altruism and kindness are "blessed, precious mistakes." But in the long run—and evolutionary theory is always and only about the very long run—misfirings are misfirings, and Dawkins clearly doesn't believe that the instinct for altruism is adaptive. So Dawkins had better enjoy those blessed and precious mistakes while they're still here to enjoy: they won't be here forever. (Of course, Dawkins can take comfort in knowing that long before natural selection eliminates altruism he'll have made his own exit from the scene.)

Since this is not a corner that many atheists will want to paint themselves into, it is not surprising that some of them are more inclined to argue that religious belief (with its accompanying intolerance) was once evolutionarily adaptive but is no longer. But it seems to me that this is an even less desirable position for them to take. As I have already noted, the mills of natural selection grind exceeding fine, but they grind very slowly indeed: or, to use another metaphor, in the very long drama of human evolution, atheism has just come on to the stage and has had almost no lines. For it to start squeaking that it has triumphed and that all the other characters on the stage are obsolete seems comically hubristic at best. Even if it were true that circumstances of human culture have changed so that religious belief is no longer evolutionarily favorable, it could take millennia for that change to manifest itself clearly in the human population. (Our reproductive cycle isn't quite as fast as that of the fruitfly.) Moreover, if religious belief doesn't happen to be adaptive at the moment, circumstances can change so as to make it adaptive once more. Some of the favorite stories of biology textbooks involve species whose adaptability has been dramatically affected by relatively sudden changes in environment. For all these reasons, the notion that religious belief has outlived its evolutionary usefulness is pure guesswork.

But such data as we currently have can't be encouraging to the atheist cause. As many studies have shown, atheism is especially associated with higher levels of education and with Western Europe, and highly educated people and Western Europeans tend to have low birth rates—less than replacement level, in many cases. If religious people are having lots of offspring and atheists aren't having many at all, that would suggest that it's almost impossible for atheists to gain ground, evolutionarily speaking.

(Here, I must admit, I am but summarizing the conceit of the movie Idiocracy, which is extremely crass but also quite germane to this topic. Reihan Salam explains why here, not neglecting to note some of the less edifying aspects of the film. You have been warned.)

So here's where I'm headed with this thought experiment: if the evolutionary account of religious belief that many atheists are now promoting is correct, then atheists don't have much of a future. Their own arguments, plus some elementary demographic data, show that their position cannot become dominant. The only real chance that atheism has to flourish is if it's wrong. If the Christian anthropology, for instance, happens to be true, then we will expect people to rebel against God, to act in violation of his will. But we will also expect them not to want to admit that that's what they're doing. So they will try to argue that their actions, however sinful, however violent, intolerant, and cruel, are somehow in keeping with God's will. But eventually the cognitive dissonance of that position is likely to become too much for them, at which point they might find—like that one–time Russian Orthodox seminarian Josef Stalin—that the easier path is simply to deny the existence of the God who otherwise would be their Judge.

So if Christianity is true (and a similar case could be made with regard to some other religions, though not all), then we might well expect atheism to flourish, at least in certain places and at certain times. But if the evolutionary argument against religious belief is true, then atheism is doomed.

Perhaps my thought experiment has gone awry somewhere—I am not convinced of its correctness—but I am sure of one thing. After having spent a great deal of time and energy trying to come up with an evolutionary explanation of religion, atheists now need to turn their attention to a still greater puzzle: What's the evolutionary explanation for atheism?

Alan Jacobs teaches English at Wheaton College in Illinois; his Bad to the Bone: an Exemplary History of Original Sin (HarperOne) will appear in Spring 2008. His Tumblelog is here.

Copyright © 2007 Books & Culture. Click for reprint information.