As Norenzayan et al. cogently argue, religions that proliferated most successfully did so because they facilitated prosociality and cooperation in large-scale, anonymous groups. One important way that religion promotes cooperation may be through improving self-control. In this comment, we cover some potential obstacles to implementing self-control and how religion can overcome them.

Six ideas explaining the existence of moralizing gods are mentioned, and I discuss the words prosocial and antisocial.

We argue that general social psychological mechanisms (e.g., common group identity) can account for prosocial behavior and cooperative norms without the need for punishing Big Gods. Moreover, prosocial religions often do not prevent conflict within their religious groups. Hence, we doubt whether Big Gods and prosocial religions are more effective than alternative identities in enhancing high-level cooperation.

Norenzayan et al. propose that Big God (BG) religions are large-group cooperative enterprises that promote internal harmony and higher fertility, resulting in “mutually beneficial exchanges” for those involved. We examine the possible distributions of costs and benefits within BG religions and propose that they are, instead, successful coordinating mechanisms that rely on intragroup competition and exploitation between the classes and sexes.

After raising some doubts for cultural group selection as an explanation of prosocial religiosity, we propose an alternative that views it as a “greenbeard effect.” We combine the dynamic constraints on the evolution of greenbeard effects with Iannaccone's (1994) account of strict sects. Our model shows that certain social conditions may foster credulity and prosociality.

Ritual cognition builds upon social learning biases that may have become specialized for affiliation within social groups. The adaptive problems of group living required a means of identifying group members, ensuring commitment to the group, facilitating cooperation, and maintaining group cohesion. We discuss how ritual serves these social functions.

We welcome Norenzayan et al.’s claim that the prosocial effects of beliefs in supernatural agents extend beyond Big Gods. To date, however, supporting evidence has focused on the Abrahamic Big God, making generalisations difficult. We discuss a recent study that highlights the need for clarity about the causal path by which supernatural beliefs affect the evolution of big societies.

One of the central claims of Norenzayan et al.’s article is that supernatural monitoring and intergroup competition have facilitated the rise of large-scale prosocial religions. Although the authors outline in detail how social instincts that govern supernatural monitoring are honed by cultural evolution and have given rise to Big Gods, they do not provide a clear explanation for the success of karmic religions. Therefore, to test the real scope of their model, Norenzayan et al. need to seriously engage with questions concerning the evolution of karmic prosocial religions.

In our response to the 27 commentaries, we refine the theoretical claims, clarify several misconceptions of our framework, and explore substantial disagreements. In doing so, we (1) show that our framework accommodates multiple historical scenarios; (2) debate the historical evidence, particularly about “pre-Axial” religions; (3) offer important details about cultural evolutionary theory; (4) clarify the term prosociality; and (4) discuss proximal mechanisms. We review many interesting extensions, amplifications, and qualifications of our approach made by the commentators.

Human cooperation is highly unusual. We live in large groups composed mostly of non-relatives. Evolutionists have proposed a number of explanations for this pattern, including cultural group selection and extensions of more general processes such as reciprocity, kin selection, and multi-level selection acting on genes. Evolutionary processes are consilient; they affect several different empirical domains, such as patterns of behavior and the proximal drivers of that behavior. In this target article, we sketch the evidence from five domains that bear on the explanatory adequacy of cultural group selection and competing hypotheses to explain human cooperation. Does cultural transmission constitute an inheritance system that can evolve in a Darwinian fashion? Are the norms that underpin institutions among the cultural traits so transmitted? Do we observe sufficient variation at the level of groups of considerable size for group selection to be a plausible process? Do human groups compete, and do success and failure in competition depend upon cultural variation? Do we observe adaptations for cooperation in humans that most plausibly arose by cultural group selection? If the answer to one of these questions is “no,” then we must look to other hypotheses. We present evidence, including quantitative evidence, that the answer to all of the questions is “yes” and argue that we must take the cultural group selection hypothesis seriously. If culturally transmitted systems of rules (institutions) that limit individual deviance organize cooperation in human societies, then it is not clear that any extant alternative to cultural group selection can be a complete explanation.

We examine the cultural group selection (CGS) hypothesis in light of our fieldwork in Northern Ghana and Oaxaca, highly multi-ethnic regions. Our evidence fails to corroborate two central predictions of the hypothesis: that the cultural group is the unit of evolution, and that cultural homogenization is to be expected as the outcome of a selective process.

Richerson et al. make a compelling case for cultural evolution. In focusing on cultural group selection, however, they neglect important individual-level accounts of cultural evolution. While scientific discourse typically links cultural evolution to group selection and genetic evolution to individual selection, this association is due to historical accident only. We thus call for more consideration of individual-level cultural evolution.

Richerson et al. establish cultural group selection as a plausible force in human social evolution. However, they do not demonstrate its causal precedence for any trait, let alone its “essentialness.” To do so, they must show that a particular group trait was caused by cultural transmission, and directly caused differences in group fitness.

The cultural group selection (CGS) approach provides a compelling explanation for recent changes in human societies, but has trouble explaining why our ancestors, rather than any other great ape, evolved into a hyper-cooperative niche. The cooperative breeding hypothesis can plug this gap and thus complement CGS, because recent comparative evidence suggests that it promoted proactive prosociality, social transmission, and communication in Pleistocene hominins.

Richerson et al. argue that “cultural group selection plays an essential role in explaining human cooperation.” We believe that cooperation came first, making culture and thus cultural group selection possible. Cooperation and culture began – and begins – in mother–infant interaction.

Because intergroup interactions often are mixed-motive rather than strictly zero-sum, groups often negotiate settlements that enable both cultures to thrive. Moreover, group prosperity rests on in-group love (rather than out-group hate) that emerges also absent intergroup competition or comparison. It follows that cultural group selection (CGS) reflects group effectiveness in organizing in-group trust and cooperation, rather than winning (in)direct intergroup competitions.

Richerson et al. provide a much needed roadmap for assessing cultural group selection (CGS) theory and for applying it to understanding variation between contemporary human groups. However, the current proposal lacks connection to relevant evidence from the human evolutionary record and requires a better integration with contemporary evolutionary theory. The article also misapplies the Fst statistic.

As scholars have rushed to either prove or refute cultural group selection (CGS), the debate lacks sufficient consideration of CGS's potential moderators. We argue that pressures for CGS are particularly strong when groups face ecological and human-made threat. Field, experimental, computational, and genetic evidence are presented to substantiate this claim.

Culture-led gene-culture coevolution is a framework within which substantive explanations of human evolution must be located. It is not itself an explanation. Explanations depend on such concrete historical evolutionary factors such as the control of fire, collective child-rearing, lethal weapon technology, altruistic cooperation and punishment, and the mastery of complex collaboration protocols leading to an effective division of social labor.

Driven by intergroup competition, social norms, beliefs, and practices can evolve in ways that more effectively tap into a wide variety of evolved psychological mechanisms to foster group-beneficial behavior. The more powerful such evolved mechanisms are, the more effectively culture can potentially harness and manipulate them to generate greater phenotypic variation across groups, thereby fueling cultural group selection.