Because supernatural beliefs giving rise to religion are nearly universal among humans, many researchers suspect — with considerable justification — that the propensity to harbor such beliefs and adhere to religions is the product of evolution and natural selection. Researchers disagree, however, on whether the cognitive architecture supporting supernaturalism and religion was itself selected for or whether it is the byproduct of selection for something else. These disagreements sometimes lead to arguments over whether religion is adaptive or maladaptive. Because these evolutionary theories of religion are so different (and identified with particular researchers), I have created a Category for each: Religion as Adaptation and Religion as Byproduct.
Those (such as David Sloan Wilson, Nicholas Wade, and Matt Rossano) who assert that religion is an adaptation frequently rely on group-level selection and the role that religion plays in promoting morality-sociality. Richard Sosis makes similar arguments within the context of ritual behavior and costly signaling which solidifies one’s commitment to a group.
Those (such as Scott Atran, Pascal Boyer, and Justin Barrett) who assert that religion is a byproduct of other adaptations point to various aspects of ordinary cognition which give rise to supernatural thinking. This view accords with Stephen Jay Gould’s contention that religion is an exaptation or “spandrel.”
Although these evolutionary theorists of religion differ significantly in their approaches and conclusions, these differences do not seem to be mutually exclusive. There appears to be no reason, in other words, why religion cannot be partly explained as an adaptation and partly explained as a byproduct.