What are atheists for? Hypotheses on the functions of non-belief in the evolution of religion

An explosion of recent research suggests that religious beliefs and behaviors are universal, arise from deep-seated cognitive mechanisms, and were favored by natural selection over human evolutionary history. However, if a propensity towards religious beliefs is a fundamental characteristic of human brains (as both by-product theorists and adaptationists agree), and/or an important ingredient of Darwinian fitness (as adaptationists argue), then how do we explain the existence and prevalence of atheists ! even among ancient and traditional societies? The null hypothesis is that ! like other psychological traits ! due to natural variation among individuals in genetics, physiology, and cognition, there will always be a range of strengths of religious beliefs. Atheists may therefore simply represent one end of a natural distribution of belief. However, an evolutionary approach to religion raises some more interesting adaptive hypotheses for atheism, which I explore here. Key among them are: (1) frequency dependence may mean that atheism as a ‘‘strategy’’ is selected for (along with selection for the ‘‘strategy’’ of belief), as long as atheists do not become too numerous; (2) ecological variation may mean that atheism outperforms belief in certain settings or at certain times, maintaining a mix in the overall population; (3) the presence of atheists may reinforce or temper religious beliefs and behaviors in the face of skepticism, boosting religious commitment, credibility, or practicality in the group as awhole; and (4) the presence of atheists may catalyze the functional advantages of religion, analogous to the way that loners or non-participants can enhance the evolution of cooperation. Just as evolutionary theorists ask what religious beliefs are ‘‘for’’ in terms of functional benefits for Darwinian fitness, an evolutionary approach suggests we should also at least consider what atheists might be for.