New Scientist: Evolutionary anthropologist James Dow has written a program – called Evogod – that simulates the evolution of religion, attempting to determine whether the impulse to pass on unverifiable information might have evolutionary benefits. When run, the software concludes that, yes, the impulse does sustain itself, but only if non-believers help believers out.
Other attempts to explain the origins of religion contend either that A) Religion is an artefact of other brain functions (cf Julian Jaynes’ The Origin of Consciousness in the Breakdown of the Bicameral Mind), or that B) Religion is an adaptation in its own right (my take on #B: when non-believers are persecuted, of course belief becomes a survival benefit).
The article explains the idea that religion only flourishes if non-believers help out believers by suggesting that belief could be an impressive trait to non-believers. I think it could also correlate with the history of non-believers being forced to help build pyramids/cathedrals, or to otherwise participate in believer culture. Religion generally has an imperialistic (evangelistic) trajectory, a tendency to overcome non-believers in the local culture, so that non-believers come under control of believers (even today non-belief carries stigma, which is itself a cultural force that confers evolutionary advantage to believers).
Not addressed in the article is any kind of scrutiny of Evogod’s actual code or algorithms. If the principles in the source code aren’t sound, neither is the theory.