We share Norenzayan et al.’s ambition to understand religious representations in terms of evolved human dispositions. But a central part of their argument is based on misunderstandings of the historical record.
In describing societies with Big Gods, the target article perpetuates a common but misleading confusion between the religions of large-scale archaic societies – for example, Egypt, Mesopotamia, Mesoamerica – and Axial Age religions with moralizing and spiritual doctrines that appeared only in a small subset of these societies. In the former societies, gods were described as seeking obedience and sacrifices, as enforcing political norms and authority, but not as interested in people's prosocial behaviors (Bellah Reference Bellah2011). The latter feature is characteristic of various movements that appeared roughly at the same time in a small subset of large-scale archaic societies, in the Ganges Valley, China, and in the Eastern Mediterranean (Baumard & Boyer Reference Baumard and Boyer2013). The appearance of moralistic, ascetic movements with highly similar features may be related to a much higher level of wealth in these regions at the time (Baumard et al. Reference Baumard, Hyafil, Morris and Boyer2015). Axial Age doctrines were then adopted by political elites and spread through conquest and coercion to the rest of the world (Bellah Reference Bellah2011).
This difference between fierce archaic gods and Axial Age moralizing doctrines means that we just cannot use features of the latter and project them, anachronistically, onto the former. That unfortunately may be the case in the scenario proposed in the target article. So, for example, the experimental evidence we have for religious primes triggering prosocial restraint (Gervais & Norenzayan Reference Gervais and Norenzayan2012a) comes from people familiar with Axial Age, moralizing gods. There is no historical evidence that people in religions with Big Gods, outside Axial Age movements and their offshoots, were more prosocial, cooperative, and so forth, than before or elsewhere.
The proposed historical scenario also projects features of very recent religious developments onto human history and prehistory, often against the documented record. For example, the authors claim that an important part of religious behaviors in large societies stems from the need for people to demonstrate belief in gods and commitment to the group and the doctrine. But the evidence mentioned is vague and confusing. The authors write of belief and commitment signals such as “sacrifices, painful initiations, celibacy, fasting [… that] more effectively transmit commitment to others” (sect. 2.5, para. 1) and mention large temples or monasteries, as well as martyrdom. But the list mixes elements from very different types of societies and unduly attributes the features of some to the others. For example, animal sacrifice is not costly (the animals are consumed and the owners gain reputation), so it does not belong in the list; painful initiations are often nonreligious, generally do not express any beliefs at all (Barth Reference Barth1975; Reference Barth1987; Bloch Reference Bloch1974), and are motivated by the dynamics of coalitional recruitment (Cimino Reference Cimino2011); celibacy, fasting, and martyrdom do not appear in archaic large-scale societies, but typically in Axial Age religious movements (Bellah Reference Bellah2012), so they cannot be relevant to pre-Axial developments; temples or massive offerings to the gods do not show that the populace was committed to these religious symbols, but more prosaically that coercive authorities could rely on high taxes and large amounts of forced labor.
One explanation for this confusion may be that the very notion of commitment and credible belief displays is mostly found in the historically atypical, but (to us) very familiar circumstances of large-scale, industrial, relatively liberal social orders, what are sometimes called “WEIRD” people (Henrich et al. Reference Henrich, Heine and Norenzayan2010b). That credible displays of belief are irrelevant to most forms of human religious activity is quite clear in the anthropological and historical record. The religious activities found in bands and tribes (the social environment of human evolution) consist in propitiatory ancestor-cults, shamanistic-like healing, and various forms of magic. People perform these activities for pragmatic purposes (curing illness or ensuring good crops). There is no evidence in the record that people engaged in these activities ever engage in signaling to demonstrate to others that they do believe in ancestors or spirits. Indeed, being a believer in such contexts would be a useless signal, as that belief does not index any special prosocial inclinations. In large-scale archaic societies, a recent development in human evolution, a literate elite is closely allied to political authorities and enforces standardized ceremonies, an official doctrine, and so forth. The populace is mobilized, and often coerced, into participating in costly performances or giving away resources and labor. Participation, being mandatory, is precisely not a signal of anything beyond obedience. Indeed, the only people who sent credible belief signals in such archaic societies were people who did not accept the religious order, such as Jewish Messianic heretics, early Christians in Rome, iconoclasts in Byzantium, and so forth – that is, people who made it clear that they could not be trusted to abide by the common norms. Finally, in some recent social orders, because of the diversity of available competing doctrines, and relatively powerless religious institutions, it can make sense for some people to signal to others their specific adherence to particular beliefs and their commitment to a religious group. This is extremely familiar to most Western people but should not be seen as the common lot of humankind.
Once we discard ethnocentric or anachronistic assumptions, the model proposed may point to useful hypotheses in the study of religious movements. Specifically, as the authors suggest, people are often extremely interested in what others believe as far as gods and spirits are concerned. But, we would add, this occurs when expressed beliefs may serve as recruitment tools and commitment signals for specific moral and political projects. Humans need coalitions, and they need commitment from other coalition members. They use information about potential threats, superhuman powers, and moral violations as a way to elicit commitment in others (Tooby & Cosmides Reference Tooby, Cosmides and Høgh-Olesen2010). The study of such dynamics is indeed crucial to social science, but it requires that we stop believing, against the evidence, that religion is in any way special in human cognition or is central to human evolution.
We share Norenzayan et al.’s ambition to understand religious representations in terms of evolved human dispositions. But a central part of their argument is based on misunderstandings of the historical record.
In describing societies with Big Gods, the target article perpetuates a common but misleading confusion between the religions of large-scale archaic societies – for example, Egypt, Mesopotamia, Mesoamerica – and Axial Age religions with moralizing and spiritual doctrines that appeared only in a small subset of these societies. In the former societies, gods were described as seeking obedience and sacrifices, as enforcing political norms and authority, but not as interested in people's prosocial behaviors (Bellah Reference Bellah2011). The latter feature is characteristic of various movements that appeared roughly at the same time in a small subset of large-scale archaic societies, in the Ganges Valley, China, and in the Eastern Mediterranean (Baumard & Boyer Reference Baumard and Boyer2013). The appearance of moralistic, ascetic movements with highly similar features may be related to a much higher level of wealth in these regions at the time (Baumard et al. Reference Baumard, Hyafil, Morris and Boyer2015). Axial Age doctrines were then adopted by political elites and spread through conquest and coercion to the rest of the world (Bellah Reference Bellah2011).
This difference between fierce archaic gods and Axial Age moralizing doctrines means that we just cannot use features of the latter and project them, anachronistically, onto the former. That unfortunately may be the case in the scenario proposed in the target article. So, for example, the experimental evidence we have for religious primes triggering prosocial restraint (Gervais & Norenzayan Reference Gervais and Norenzayan2012a) comes from people familiar with Axial Age, moralizing gods. There is no historical evidence that people in religions with Big Gods, outside Axial Age movements and their offshoots, were more prosocial, cooperative, and so forth, than before or elsewhere.
The proposed historical scenario also projects features of very recent religious developments onto human history and prehistory, often against the documented record. For example, the authors claim that an important part of religious behaviors in large societies stems from the need for people to demonstrate belief in gods and commitment to the group and the doctrine. But the evidence mentioned is vague and confusing. The authors write of belief and commitment signals such as “sacrifices, painful initiations, celibacy, fasting [… that] more effectively transmit commitment to others” (sect. 2.5, para. 1) and mention large temples or monasteries, as well as martyrdom. But the list mixes elements from very different types of societies and unduly attributes the features of some to the others. For example, animal sacrifice is not costly (the animals are consumed and the owners gain reputation), so it does not belong in the list; painful initiations are often nonreligious, generally do not express any beliefs at all (Barth Reference Barth1975; Reference Barth1987; Bloch Reference Bloch1974), and are motivated by the dynamics of coalitional recruitment (Cimino Reference Cimino2011); celibacy, fasting, and martyrdom do not appear in archaic large-scale societies, but typically in Axial Age religious movements (Bellah Reference Bellah2012), so they cannot be relevant to pre-Axial developments; temples or massive offerings to the gods do not show that the populace was committed to these religious symbols, but more prosaically that coercive authorities could rely on high taxes and large amounts of forced labor.
One explanation for this confusion may be that the very notion of commitment and credible belief displays is mostly found in the historically atypical, but (to us) very familiar circumstances of large-scale, industrial, relatively liberal social orders, what are sometimes called “WEIRD” people (Henrich et al. Reference Henrich, Heine and Norenzayan2010b). That credible displays of belief are irrelevant to most forms of human religious activity is quite clear in the anthropological and historical record. The religious activities found in bands and tribes (the social environment of human evolution) consist in propitiatory ancestor-cults, shamanistic-like healing, and various forms of magic. People perform these activities for pragmatic purposes (curing illness or ensuring good crops). There is no evidence in the record that people engaged in these activities ever engage in signaling to demonstrate to others that they do believe in ancestors or spirits. Indeed, being a believer in such contexts would be a useless signal, as that belief does not index any special prosocial inclinations. In large-scale archaic societies, a recent development in human evolution, a literate elite is closely allied to political authorities and enforces standardized ceremonies, an official doctrine, and so forth. The populace is mobilized, and often coerced, into participating in costly performances or giving away resources and labor. Participation, being mandatory, is precisely not a signal of anything beyond obedience. Indeed, the only people who sent credible belief signals in such archaic societies were people who did not accept the religious order, such as Jewish Messianic heretics, early Christians in Rome, iconoclasts in Byzantium, and so forth – that is, people who made it clear that they could not be trusted to abide by the common norms. Finally, in some recent social orders, because of the diversity of available competing doctrines, and relatively powerless religious institutions, it can make sense for some people to signal to others their specific adherence to particular beliefs and their commitment to a religious group. This is extremely familiar to most Western people but should not be seen as the common lot of humankind.
Once we discard ethnocentric or anachronistic assumptions, the model proposed may point to useful hypotheses in the study of religious movements. Specifically, as the authors suggest, people are often extremely interested in what others believe as far as gods and spirits are concerned. But, we would add, this occurs when expressed beliefs may serve as recruitment tools and commitment signals for specific moral and political projects. Humans need coalitions, and they need commitment from other coalition members. They use information about potential threats, superhuman powers, and moral violations as a way to elicit commitment in others (Tooby & Cosmides Reference Tooby, Cosmides and Høgh-Olesen2010). The study of such dynamics is indeed crucial to social science, but it requires that we stop believing, against the evidence, that religion is in any way special in human cognition or is central to human evolution.