The religion-as-an-adaptation theory doesn't wash with everybody, however. As anthropologist Scott Atran of the University of Michigan in Ann Arbor points out, the benefits of holding such unfounded beliefs are questionable, in terms of evolutionary fitness. "I don't think the idea makes much sense, given the kinds of things you find in religion," he says. A belief in life after death, for example, is hardly compatible with surviving in the here-and-now and propagating your genes. Moreover, if there are adaptive advantages of religion, they do not explain its origin, but simply how it spread.The article itself covers a lot of territory - Justin Barrett's work on childhood belief, and Paul Bloom's work on innate dualism.
Bloom says the two systems are autonomous, leaving us with two viewpoints on the world: one that deals with minds, and one that handles physical aspects of the world. He calls this innate assumption that mind and matter are distinct "common-sense dualism". The body is for physical processes, like eating and moving, while the mind carries our consciousness in a separate - and separable - package. "We very naturally accept you can leave your body in a dream, or in astral projection or some sort of magic," Bloom says. "These are universal views."
There is plenty of evidence that thinking about disembodied minds comes naturally. People readily form relationships with non-existent others: roughly half of all 4-year-olds have had an imaginary friend, and adults often form and maintain relationships with dead relatives, fictional characters and fantasy partners. As Barrett points out, this is an evolutionarily useful skill. Without it we would be unable to maintain large social hierarchies and alliances or anticipate what an unseen enemy might be planning. "Requiring a body around to think about its mind would be a great liability," he says.
Which feeds into Jesse Bering's findings on assumptions about life after death, and Deborah Kelemen's work on teleology (our natural tendency to assume that objects have a purpose - the sun exists to provide warmth, for example).
True, it concludes that there could be some benefit to religion (from an evolutionary perspective)
Robin Dunbar of the University of Oxford - the researcher most strongly identified with the religion-as-adaptation argument - also has no problem with the idea that religion co-opts brain circuits that evolved for something else. Richard Dawkins, too, sees the two camps as compatible. "Why shouldn't both be correct?" he says. "I actually think they are."
But the general tenor seems to capture the way the wind is blowing in this particular field of research. Evidence continues to accumulate that many of our supernatural beliefs are accidental by-products of brain functions that evolved for other purposes, while empirical evidence that believing in god actually provides measurable benefits (from an evolutionary perspective) remains weak and circumstantial.