Published October 7, 2013

Religion is puzzling.

Religion—at least the organized religions with which people in the West are familiar—cause people to do odd things, such as adopt (or, at least, appear to adopt) a large number of false beliefs, engage in rituals that are at best useless and at worst costly, and give up their resources to strangers.

Probably because these things all seem strange from an evolutionary point of view, scholars have spilled considerable ink trying to explain religion, both its component parts (supernatural belief, costly ritual) and the whole enchilada. Book length treatments include, in no particular order, Religion Explained (Boyer), The Belief Instinct (Bering), The God Delusion (Dawkins), Breaking the Spell (Dennett), and Darwin’s Cathedral (Wilson).

Arguably the predominant argument surrounding religion goes something like this. People with religious beliefs are less likely than people without religious beliefs to lie, cheat, steal, or mix dairy with meat because they are afraid of supernatural punishment. Doing less of (at least the first three of) these things is good for the individual—because supernatural-agent-believing-in people get punished less than they otherwise would for breaking rules—and for the group—because there is less welfare-destroying behavior than there would be if people didn’t believe in punitive gods.

Of course a link exists between religion and cooperation, if only insofar as members of organized Western religions really do tend to cooperate with their co-religionists. Members of religious organization cooperate in any number of ways, of course, from bake sales to fund renovations of the nave to cooperative child care to going on Crusades.

And there are data supporting a link between religion and cooperation using various techniques. For example, some have investigated the issue by looking at individual differences. Atkinson and Bourrat (2011), for instance, looked at people’s self-reported religious beliefs and their self-reported endorsement of a number of items relating to morality. (For instance, people’s views on whether lying is never, sometimes, or always justified.) They conclude:

The results we present here are consistent with and provide support for specific predictions of the supernatural monitoring and fear of supernatural punishment hypotheses. As predicted by both theories and consistent with our Hypothesis 1, individuals who professed belief in God rated moral transgressions as less justifiable than those who did not. Consistent with Hypothesis 2 and the supernatural punishment hypothesis, stronger beliefs about the unjustifiability of moral transgressions were present in individuals who professed belief in heaven or hell. And consistent with Hypothesis 3 and the supernatural monitoring hypothesis, among those who believe in God, those who professed belief in a personal God rated moral transgressions as less justifiable than those who professed belief in a Spirit or Life Force.

But maybe the cooperative face of religion isn’t the whole story, or even the main story. After all, people organize themselves into all kinds of cooperative groups that have few features of organized religions (supernatural beliefs, rituals, etc.). Perhaps people organize themselves into religious groups for some other reason and then, by virtue of finding shared interests, cooperate toward common goals.

new paper now in press in Evolution and Human Behavior by #cough# Jason Weeden and Robert Kurzban with the laconic title, “What predicts religiosity? A multinational analysis of reproductive and cooperative morals” diverges from the dominant religion-is-for-cooperation view.

We used the same database that Atkinson and Bourrat used, but our interest was in the possibility that people used religious organizations less as a locus of cooperation, generally, but more as a means of advancing their reproductive goals. Previous work showed strong relationships between religious variables such as church attendance and planned church attendance with self-reported sexual behavior and attitudes, such as sociosexuality—i.e., roughly, promiscuity—and moral views on stuff pertaining to sex and reproduction, such as views on abortion, birth control, and pre-marital sex. So, we looked at the relationships among religiosity, reproductive morals, and cooperative morals drawing on the vast amount of data available from many nations in the World Values Survey.

As one might have guessed, both the cooperative items and the reproductive items correlate reasonably well with the religious items. Still, just looking at these relationships, we find that, internationally, the relationship between cooperative morals and religious items was about one quarter of the size of the relationship between reproductive morals and religious items.

Though this suggests to us that the reproductive issues might be driving participation in religion, in themselves these relationship don’t settle much because both accounts of religion allow, in principle, for other relationships to exist. One could imagine that religion was for cooperating, but then once cooperative religious groups form, they take on systematic properties, such as moralizing sexual behavior. So, the next step is to run analyses in which we regress religiosity measures on both the reproductive items and the cooperative morals items. When we ran these analyses, we found that the coefficients on the relationship between reproductive morals and religion stayed about the same but the coefficients on the relationship between cooperative morals and religion were mostly eliminated, in some cases becoming slightly negative. This pattern existed, with variations on the theme, worldwide.

In short, while there is, for sure, a relationship between cooperative morals and religiosity, this relationship is attenuated or eliminated when controlling for reproductive morals, and this attenuation occurs pretty much everywhere we have looked.

This is not to say, of course, that various aspects of religion might not have evolved because of their role in facilitating cooperation. I myself am skeptical of such accounts for a number of reasons, not the least of which is that, for reasons I’ve discussed before, natural selection should punish systems that give rise to false beliefs about supernatural punishment in favor of systems that give rise to true beliefs. Still, these patterns of data suggest to us that people might be joining religion because of religion’s role in facilitating their favored reproductive strategy.

So of course religious groups cooperate, and often do so effectively and successfully. But maybe people join religious groups for reasons that go beyond cooperation per se, having more to do with the role religion plays in people’s sexual and reproductive behavior.