What should we believe about atheists?

Catherine Caldwell-Harris provides a stimulating review of the psychological correlates of atheism. The broad conclusion is compelling evidence in support of the hypothesis that atheism results from natural variation along a continuum of belief (hypothesis 2 in my target article of this issue). I called this my ‘‘null hypothesis’’ because it just seems the most likely and least complicated or speculative: we observe people who believe in God (for example) and people who do not. We also know that there is variation in the strengths of such beliefs. Therefore, like any other biological trait ! whether physiological or psychological ! atheism is probably just one end of the spectrum of naturally occurring variation in the propensity for religious belief. However, significant doubt remains, for several reasons: we do not know much about causality (more experimental work is needed); we have very little data on atheism in non-western and indigenous societies (are western populations representative?); we must not rely too heavily on self-report data (people may declare themselves atheists but nevertheless display supernatural thinking); and there have been almost no studies on the other nine evolutionary hypotheses for atheism (speculative as they are, we have no data to judge them). This is not to emphasize disagreement. At the end of my target article I said I found the null hypothesis (natural variation) the most compelling, and CaldwellHarris’s excellent study only reinforces this view. However, in the same spirit of questioning everything that inspired my original article, let me now attack the null hypothesis itself.