Many religious rituals are clearly non-rational. Sacrificing goods or animals to supernatural agents, praying long hours or enduring pain during initiation rituals are hardly reasonable undertakings from the perspective of a non-believer.
While many scholars were content with attributing these behaviors to “false beliefs”, evolutionary religious studies scholars look for possible functions for survival and reproduction. And thankfully, some scientists did what scientists are for (besides teaching & writing): they designed and tested empirical hypotheses on the matter!
In 2003, Richard Sosis and Eric Bressler achieved a breakthrough by applying the “costly signaling theory”. According to this theory, religious rituals are able to promote intragroup cooperation exactly by bringing up purely or partially non-rational behaviors – signaling to onlookers that true believers are at hand. In “Cooperation and Commune Longevity: A Test of the Costly Signaling Theory of Religion”, Sosis and Bressler found strong empirical support for this assumption by comparing endurance of religious and secular as well as more and less “costly” communities within the USA of the 19th century.
Nevertheless, the debate didn’t end here (as evolutionary studies never do). After all, not all religious rituals (such as speaking the Muslim basmallah at the start of an action or bowing to a cross) are that “costly”. And obviously, the very fact that we are able to discern “religion” in quite different cultural environments indicates that there are certain inter-religious commonalities among ritual forms.
Therefore, delight was great as John Henrich managed to sharpen the theory from the rather vague “costly signals” to the more specific “credibility enhancing displays (CREDs)” in 2009. His paper “The evolution of costly displays, cooperation and religion: credibility enhancing displays and their implications for cultural evolution” was able to establish new standards for anyone interested in the bio-cultural evolution of religiosity and religions.
In the new “Evaluating ritual efficacy: evidence from the supernatural”, Cristine Legare and André L. Souza conclusively demonstrated in four empirical studies (three in Brazil, one in the US), that individuals rated the efficacy of rituals corresponding to two main factors: (1) Procedural repetition and specifity and (2) Association with a supernatural observer. Respective rituals were cognitively accepted (“felt”) to be (more) effective by religious and non-religious people performing them.
Religious rituals as a means of forming cooperative networks and groups
We should be ready to acknowledge that many religious rituals are clearly non-rational – and that this is the very foundation of their psychological, social and finally evolutionary effects. Sports fans ready to dress in weird and colorful dresses, to chant more or less meaningful rhymes and to assemble at specific times for the sake of identifying with their teams are forming cultural groups “just for fun”. Religious congregations are doing the same with the added beliefs of specific supernatural agents watching from above. Religious networks, communities and families are thus strengthened in promoting in-group cooperation, survival and reproduction.
Religious rituals are effective because they are seemingly defying rationality!