Durkheim, Indoctrinability and the Moral Ambivalence of Human Sociality

Abstract Humans tend to be prosocial towards in-group members, but often hostile towards out-group members. In line with ethology and evolutionary psychology, this study develops the concept of Durkheimian learning, meaning that groups actively integrate individuals by cultural transmission of behavioral norms. Yet humans might have genetically evolved predispositions for indoctrinability and constrained learning of what is adaptive for self-interested individuals living in groups. In-group prosociality might have coevolved with tendencies towards out-group rejection. The mechanism behind this outcome might have been cultural-group selection in combination with gene-culture co-evolution. The feasibility of the argument will be illustrated by two simulation experiments of a selfish-teamwork game. Inspired by the Price equation of multilevel selection, the simulations combine fuzzy group-boundaries and between-group competition with cooperative behavior and intergenerational cultural transmission of prosocial behavior within groups. Results show that group boundaries, e.g. due to cultural markers, can explain the evolution of cooperative behavior. However, efficient Durkheimian learning and intergroup competition might have accelerated its expansion, despite of considerable levels of between-group migration. This insight will be discussed in the light of recent arguments from cultural evolutionary psychology. Introduction Why are humans so ‘groupish’ (Haidt, 2012: 225)? Did human prosociality coevolve with group-boundaries? Have in-group preferences and the (relative) devaluation of out-groups co-evolved with prosociality? These questions are not only of practical relevance given the numerous empirical results (Berreby, 2008; Ellemers, 2017) on intergroup competition, conflict and own-group preference in intergroup relations. They are also crucial to our understanding of human nature. Sociological concepts of social integration and biological studies on the evolution of prosociality share the focus on the essential tension between individual and collective rationality (Bahar, 2018). In humans, group integration is not “naturally” given. According to Maryanski (2018: 230), human nature is not inherently group-oriented, but rather individualistic. As Durkheim’s classic argument suggests (Durkheim, 1915), early humans had to actively generate social integration by collective practices, and individuals had to learn social norms and taken-for-granted knowledge. Given that cultural learning was adaptive not only for the individual but especially for the group—the primary locus of social interaction and cultural transmission—the evolutionary mechanism may have been cultural-group selection, a variant of multilevel selection. Multilevel selection theory (Bourrat, 2021; Boyd, 2018; Price, 1970) suggests that mechanisms of group bonding provided an advantage in intergroup competition and conflict. When within-group prosociality and group bonding exist, between-group selection pressure is typically intensified, thereby driving intra-group cooperation even further. In turn, group bonding is more readily established if individuals already possess cognitive and emotional dispositions that make them receptive to social influence from the group. Cultural-group selection thus required the evolution of ‘cultural brains’ (Muthukrishna, 2025), which facilitate the efficient acquisition of knowledge transmitted via social interaction. By the same token, gene-culture coevolution and ‘cultural brains’ may have fostered individual indoctrinability (Eibl-Eibesfeldt, 1998)—that is, individual susceptibility to social influence and the proclivity for bonding with the group. Who we are inclined to learn from depends partly on neurochemical processes, including oxytocin release. As an individual disposition, indoctrinability might be linked to an evolved psychological predisposition for prepared or constrained learning (Plotkin, 1998) of what is important for a species of group-living, but self-interested, individuals: learning about group-cohesion and the maintenance of group-boundaries (Eibl-Eibesfeldt, 1982, 1998). We must learn social integration. However, through cultural evolution and cultural-group selection, our evolving cultural brains have specifically fine-tuned our genetic disposition toward learning group integration. This will be called Durkheimian learning in the present study. This concept is inspired by Maryanski (2018) who connects Durkheim’s learning-theoretic account of how humans actively shape social integration with recent findings from evolutionary and biosocial research. Building on this, the present study advances the discourse by incorporating indoctrinability as an individual condition of Durkheimian learning, and the evolutionary mechanism of multilevel selection to explain the evolutionary advantage of cooperative and cohesive groups. Such a mechanism, however, requires either more or less clear group boundaries, or the group’s capacity to extent cultural transmission to immigrants so that assimilation is possible. Otherwise, the group cannot function as a unit of cultural-group selection. Following previous studies (Bowles, 2004, 2006, 2009; Choi & Bowles, 2007), simulation models based on the Price equation of multilevel selection will investigate the feasibility of the argument, also given fuzzy group boundaries due to considerable levels of inter-group migration. The following sections are arranged according to Fig. 1: the adaptive value of cultural learning, once in place, is the starting point, and morality-based dispositions towards group boundaries the end point. Section 2 highlights the relevance of morality for group-boundaries, resulting from group-based morality and indoctrinability as evolved predispositions in humans (Sect. 3). Section 4 describes the close relationship between indoctrinability, constrained and Durkheimian learning, efficient norm transmission and gene-culture co-evolution, that emerges from the evolutionary mechanisms of multilevel-selection and cultural-group selection (Sect. 5). Section 6 describes the simulation model, Sect. 7 the results, and Sect. 8 discusses the argument against the background of evolutionary cultural psychology. Group-Boundaries and Morality The distinction between in- and out-group is an integral part of an elementary form of human relations, namely communal sharing (CS) (Fiske, 1993): every individual contributes to the collective good conditional on his or her capabilities; resources are distributed according to individual need rather than achievement or contribution. In order to avoid exploitation by outsiders, who do not contribute to the common good, group-boundaries regulate the access to the group’s resources. Hence, “the negative or inverse side of CS as a distributional principle is that it involves favoritism toward the we-group and ignoring and discriminating against outsiders. (…) It seems to be characteristic of CS groups that they are based on radical contrast of an oppositional sort; all that is good is We, all that is bad is Other” (Fiske, 1993: 61). CS might have been the primary form of social interaction in ancient hunter-gatherer groups (Fiske, 1992: 693). Yet CS does not function flawlessly in hunter-gatherer contexts (Singh & Glowacki, 2022), which ledBoehm (2012: 199–202) to examine how coalitions emerge as mechanisms of social control to suppress free-riders and despotic individuals. In a sense, ostracism and capital punishment(Boehm, 2012: 240–249) defined these cooperators as outsiders. Spanning a social boundary between in– and out–group thus protects the product of within-group cooperation. How do groups span these boundaries? Durkheim (1915) compiled empirical material from Australian Aboriginal tribes on totemism. In Durkheim’s view, the totem is a sacred symbol of the clan. When performing religious rituals organized around the totem, group members attribute the experience of strong and extraordinary collective emotions to the power of the totem. He described these emotions as ‘collective effervescence’, which in fact represent the power of the group (Durkheim, 1915: 221). Hence, Durkheim did not conceptualize human nature as naturally collectivistic or group oriented. His view was later corroborated by cladistic analyses of ape social relationships which revealed rather weak social ties in the last common ancestor (LCA), except for bonds between mothers and offspring (Maryanski, 2018: 221). Actively generating collective emotions is thus important for social integration and group cohesion. This is an important implication for human nature in Durkheim’s perspective: rather than being guaranteed by an ‘innate trait’ (Maryanski & Turner, 1992: 66), group cohesion must be permanently (re-)produced by performing religious rituals, by generating collective effervescence towards the totem (Maryanski, 2018: 259), and particularly by cultural transmission (Henrich & Muthukrishna, 2021). Herein lies the ambivalence of human sociality: commitment and cultural transmission facilitate cooperation within groups, which implies spanning a boundary against others, who we consider as less cooperative, often as competitive, or even as threatening. When morality and norms are group-specific, we believe it is us, the in-group, who knows the ‘right’ way to do things, but not the other group (Tomasello, 2018: 87). Ellemers (2017: 34) argues that group-specific morality defines who does and does not deserve to belong to the group. Since our well-being depends on group membership (Cacioppo & Patrick, 2009), we are sensitive to our group’s morality, and willing to accept it in order to avoid social exclusion. When groups are confronted with deviating moral principles of out-groups, and out-group moral principles are considered morally wrong, there is a strong inclination to reject ‘dangerous’ out-groups (Ellemers, 2017: 182) “because we have a harder time predicting how they will interpret ambiguous situations, which outcomes they find important, what they will consider ‘right’ or ‘wrong’ choices, and how they are likely to behave” (Ellemers, 2017: 110). If there are cues that a person shares the same moral principles, e.g. by religious symbols (Simmel, 1964: 30), we are more inclined to trust this person and to cooperate (Henrich & Muthukrishna, 2021: 230). Morality as an Evolved Predisposition How do culture and morality enter people’s minds? Instead of assuming a blank slate or genetic determinism, ethologists prefer the term predisposition. Konrad Lorenz’ famous example is the imprinting in geese. After birth, geese follow the first visible, moving object (Eibl-Eibesfeldt, 1989: 79). This algorithm (follow the first moving object you see after hatching) might be genetically pre-organized. Likewise, in most mammalian species, mothers intensively interact with their newborn children after birth, e.g. by licking or vocalization (Eibl-Eibesfeldt, 1998: 36), which correlates with the emission of the neurotransmitter oxytocin (Sapolsky, 2018: 109). Rather than being a universal love-drug, the effect of oxytocin is ambivalent, depending on whether the other person is perceived as an in-group-member or as a potentially threatening out-group-member (Sapolsky, 2018: 116–118). Possibly, humans’ preadaptation for child-parent bonding extended to bonding within “small groups of kin to form broader alliances” (Eibl-Eibesfeldt, 1982: 179), which, according to Eibl-Eibesfeldt, even made group-selection likely (see below). In this view, oxytocin release may serve as a mechanism that bonds us to in-group members, from whom we are highly inclined to learn. We might have become adapted to comply with group norms because sizeable coalitions of group members could easily outnumber individuals and enforce norm compliance (Boehm, 2012: 83). Coordinated punishment is possible due to mental adaptations for the detection and coordination of alliances described by Tooby and Cosmides (2010: 210–211), e.g. the representation of others’ formidability and their social support (Pietraszewski, 2021). According to the gene-culture coevolution theory such cognitions are learned, but humans might be genetically prepared to learn what is adaptive in social environments (see below) (Plotkin, 1998: 166). Indoctrinability and Cultural Groups Group-Boundaries: Who Will Cooperate? By highlighting the importance of within-group morality for the reproduction of the group, Darwin (1871: 166) implicitly believed in group selection (Plotkin, 1998: 77–78; Wilson, 2012: 52): natural selection would operate at the group-level by favoring traits which increase the fitness of the group relative to other groups. But how do cooperators recognize each other – which is a necessary condition to selectively interact with them? Biologists suggested either kin-selection, which means that a group of close relatives is the unit of selection, or reciprocal altruism, which “… occurs when an individual chooses to act altruistically toward another who has already acted altruistically in the past” (Trivers, 1985: 48). Both mechanisms require more or less stable group-boundaries. Group selection of genetically transmitted behavioral dispositions becomes unlikely in case of high rates of migration between groups (Chudek & Henrich, 2011: 222). In primates, females transfer to another group after the onset of puberty (Boesch, 2012: 120), and probably this pattern already existed in australopithecines (Dunbar, 2016: 133), who lived ~3mio. years ago. Likewise, reciprocal altruism requires complex cognitive mechanisms to distinguish between cooperators and ’cheaters’ in the exchange process (Trivers, 1985: 395). In haplodiploid state-building insects sisters share 75% of their genes. Here, the evolution of altruism can be well explained by Hamilton’s rule of kin selection: if the benefit B of an altruistic act weighted by genetic relatedness r out-weights the costs C of an altruistic act (rB > C), genetic dispositions for behaving altruistically will spread in the population. Altruism as a result of genetic group-selection is only possible, however, as long as groups maintain rigid social boundaries.Moffett (2019: 69) reports an example of a researcher who fed ants in a laboratory with cockroaches. When he provided a different kind of cockroach, any “… ant that so much as touched one of these brown-banded cockroaches was summarily slaughtered by her fellows”. Brown-banded cockroaches emit hydrocarbon molecules similar to molecules that foreign ant colonies use as a scent indicating colony membership. As soon as an ant touched the cockroach for feeding, they got contaminated by the odor, were misidentified as a foreigner, and got killed by their fellows. In humans, in contrast, neither genetic group selection nor reciprocal exchange can explain the evolution of altruism, if the latter does not consider a group boundary that increases the likelihood of interacting with cooperators. Gene-Culture Co-Evolution and Indoctrinability in Humans The example of how pheromonic communication automatically initiates the execution of genetically inherited behavioral programs is far from appropriately describing human behavior. Mammals, and particularly primates, are endowed with high levels of agency. According to Tomasello (2022: 83–84), primates are able to simulate their action plans and to evaluate the potential outcomes. Given the high level of agency in homo, the “essential tension” (Bahar, 2018) between unprecedented levels of agency, individual selfishness and the collective good became even more pronounced. How could human groups control individual selfishness? Since norm compliance became crucial for the survival and the social status of individuals (Boehm 2012; Wrangham 2019b: 130), humans became increasingly susceptible to the influence of group norms. Self-control relies on the evolved large neocortex (Boehm, 2012: 173), and is a condition of norm compliance. Norms are elements of culture, and the individual capacity to absorb these crucial cultural gadgets might have been beneficial for survival and reproduction. Moreover, given non-negligible levels of intergroup migration, particular females (Dunbar, 2016: 133), group-level selection (see below) might operate more effectively for culture than for genes. Within their familiar groups, humans learn about norms and taken-for-granted-knowledge by horizontal and vertical transmission (that is, across generations) (Laland, 2017; Richerson & Boyd, 2005). Similar to Chomsky’s concept of a ‘language acquisition device’, Konner (2010: 720) assumes a ‘cultural acquisition device’ (CAD). The emotional dimensions of the CAD consist of attachment, positive identification, fear of strangers, negative identification and emotion management. In line with Konner, and also with the concept of phylogenetically programmed “built-in learning mechanisms” (Lorenz, 1966: 47), Plotkin (1998) suggests that cognitive modularization corresponds with a disposition toward constrained learning: from an evolutionary perspective, learning should be adaptive, which is why humans are disposed to narrowing down the search space of possible content to learn (Plotkin, 1998: 166). Being neurologically adapted to imitate and to comply with group norms is one example of gene-culture co-evolution (Haidt, 2012: 241; Henrich, 2016; Henrich & Muthukrishna, 2021: 226; Richerson & Boyd, 2005). Humans lived in extremely diverse habitats, such as the tropical rainforest, the desert or the Arctic (Gintis & Helbing, 2015: 24; Laland, 2017: 190; Richerson & Boyd, 2005: 130–131). This flexibility in niche construction required learning reliably the specific set of taken-for-granted knowledge, the cognitive gadgets (Heyes, 2018), to survive in the respective niche. Cultural evolution might have passed the threshold of becoming autocatalytic ~2mio. years ago. ‘Autocatalytic’ means that culture itself produces “the fuel” which propels its own evolution: the selective advantage of retaining and transmitting (“teaching”) adaptive cultural innovations favored the ‘cultural brain’ (Muthukrishna, 2025: 164), that is, genetic conditions of cognitively processing cultural information (e.g. storage, retrieval and communication), which resulted in a self-reinforcing process of increasing the cognitive capacity to produce, to apply and to reproduce culture (Henrich, 2016: 64; Laland, 2017: 196–197). According to Laland (2017: 130), the ability to attribute thoughts, beliefs and mental states to others (theory of mind) facilitates social learning. In this regard, the ‘social brain’(Dunbar, 2016: 30) is an essential component of the ‘cultural brain’. Due to autocatalytic cultural evolution our norm psychology(Chudek & Henrich, 2011) even became “addicted to culture” (Henrich & Muthukrishna, 2021: 210). Yet our survival depends on the acquisition of the “right” culture (Richerson & Boyd, 2005: 111), adapted to the particular ecological niche of the respective cultural group (Henrich & Muthukrishna, 2021: 224), whereas the “wrong” knowledge could be lethal in a given context (Henrich, 2016: 34–35). Cultural transmission follows basic rules, such as imitating the successful or the most common behavior, so that humans often simply copy cultural practices with high fidelity, without the burden of testing alternatives themselves (Muthukrishna, 2025: 74; Richerson & Boyd, 2005: 114–126). Hence, the increasing importance of culture and the ‘cultural brain’ coevolved with humans’ ‘indoctrinability’ (Eibl-Eibesfeldt, 1982, 1998), with the cognitive and emotional openness to social influence and to adopt the “right” culture. Indoctrinability seems to correspond with an innate disposition towards “social motivation” and “attentional biases” towards faces and vocalizations (Heyes, 2018: 64). Babies might have benefited from an evolved predisposition for rudimentary forms of a theory of mind and for attracting potential caregivers’ attention (Hrdy, 2009: 139–141), not just by crying. When babies are physically separated from their mothers, babbling attracts the attention of caregivers the baby is accustomed to. The slow, high-pitch, and melodic “motherese” speaking of mothers and caregivers (Hrdy, 2009: 112), in turn, signals their presence and commitment to the baby’s well-being (Hrdy, 2009, p. 123). Adults are open to social influence as well. Sherif’s classic experiments on the autokinetic effect showed that when we observe light points in a dark room, after a while our brains generate the illusion that these points move. People are susceptible to the influence of others when evaluating the direction and distance of the move (Sherif, 1935). If the experimenter’s confederates become successively replaced with naive persons, the false belief in the movement seems to remain stable (Jacobs & Campbell, 1961). In other words, the group can establish a false belief which is then perpetuated over subsequent generations – the false belief becomes established as a culture. Sunstein (2019: 87–88) argues that conformity in groups often emerges because we avoid social disapproval. When respected persons confidently express their opinions on complex issues, people tend to keep their differing views to themselves. As a result, deliberation on complex issues can push groups toward more extreme decisions than a simple aggregation of members’ initial views would suggest (Sunstein, 2019: 80). People expect social disapproval if they deviate from shared beliefs of the group. Group-living early humans have depended on others’ social approval for ~2mio. years (Turner, 2021: 200), so we might have evolved capacities for social cognitions and emotions, like shame and guilt, that probably were hominins’ “first language” (Turner & Maryanski, 2024: 185 − 160), as well as for constrained learning and imitation, even though this is still debated (Bonini et al., 2022; Heyes, 2018: 119–121; Meltzoff, 2002; Rizzolatti et al., 2002; Sapolsky, 2018: 538–541). Oxytocin, Group Rituals and Constrained Learning: A Neo-Durkheimian Argument Oxytocin seems to be important in mother-child bonds, and could have been a preadaptation (Turner, 2021: 44) that became extended to regulate the distinction between in- and out-group (Eibl-Eibesfeldt, 1998: 38). Oxytocin can reinforce group cohesion (Cacioppo & Patrick, 2009: 140), in particular in combination with rhythmic group rituals, such as music and singing (Mithen, 2006: 216–217). Being socially embedded into our in-group of people, who we know and trust, has positive effects on our mental and physical health, which is also mediated by oxytocin (Cacioppo & Patrick, 2009). Yet oxytocin seems to make us more sensitive to the boundary between the in-group and the out-group and blinds “ourselves to the humanity of people outside our group when we feel threatened” (Hare & Woods, 2020: 111). The ambivalent effects of oxytocin thus suggest a bio-social foundation of the bonding effect of group rituals and collective effervescence (Maryanski, 2018: 190). Oxytocin could explain why humankind’s “… identification via symbols bears much resemblance to … imprinting phenomena” (Eibl-Eibesfeldt, 1998: 35). Consequently, the question of whom we are more inclined to learn from is likely influenced by oxytocin. In contrast to haplodiploid insects, group-boundaries in humans are obviously not defined by pheromonic communication or other “hardwired bioprogrammers” (Turner, 2021: 48). Oxytocin and other neurochemicals involved in social bonding are predispositions for group integration, social influence and indoctrinability. Their activation requires collective activity to overcome the egoistic individualism (Maryanski, 2018: 230). In Plotkin’s view these mechanisms “… all draw upon a single process that has a single evolutionary origin in maintaining group-level adaptations” (Plotkin, 1998: 250), namely constrained learning in a group living species: learning evolved because of its adaptive benefit in dynamic environments. We tend to learn behavior more easily when it was adaptive in the past (Lorenz, 1966). Driven by our cultural brains, Durkheimian learning is a specific form of constrained learning that enables us to draw group boundaries. Durkheimian learning means selectively learning to bond with the group due to emotional arousal, social influence and cultural transmission of group specific taken-for-granted knowledge, in particular of group-norms and markers of group-membership. Hence, it is likely that humans are biologically prepared to social influence, conformity and indoctrination. Group rituals, collective effervescence and the emission of oxytocin might be important pathways of social influence and indoctrination, which bind the individual to the group. In environments with competing groups, indoctrinability becomes beneficial since it allows groups to function as higher-level units. Already E.O. Wilson (1978: 186–187) argued that group selection favored genetic predispositions for conformity with norms, so that the susceptibility to indoctrination could have become an innate trait. This could be a reason why the origin of humans’ sensitivity to group-boundaries, as elaborated e.g. by social identity theory (Tajfel, 1982), results more from cultural markers rather than from racism (boundaries according to phenotypic markers) (Kurzban et al., 2001). If two-days old babies learn to choose among audio tracks, they show a clear preference for their native language, or intonation patterns characteristic of this language (Moon et al., 1993). Although pre-natal and early learning surely influenced these results (Heyes, 2018: 64), it is likely that they indicate a predisposition for a coalitional psychology (Bloom, 2013: 109–111; Tooby & Cosmides, 2010), which interprets language as a cultural indicator of group-boundaries. In this view, and in line with the concept of constrained Durkheimian learning, humans might be innately prepared to learn about group cohesion, membership and group-boundaries. Group boundaries based on persistent cultural differences between groups are a condition of cultural-group selection (Boyd, 2018: 98–99; Henrich, 2004; Muthukrishna, 2025: 195–197). For instance, morality and particular ways of regulating social interaction, e.g. by punishing norm violators(Eibl-Eibesfeldt, 1982: 187; Henrich, 2016: 196) or executing anti-social bullies (Boehm, 2000: 72–73; Wrangham 2019a: 146–153), determine the probability and the quality of collective goods. This, in turn, affects the outcomes of intergroup competition and violent intergroup conflict (Richerson & Boyd, 2005: 208). Groups become culturally distinct, despite of inter-group migration, if cultural transmission efficiently follows the rules of ‘imitate the successful or the most common type’ (Richerson & Boyd, 2005: 162). Cultural markers indicate who belongs to the same or similar cultural group and who is likely to share one’s taken-for-granted knowledge, norms and morality (Henrich, 2016: 201; Tomasello, 2018: 91). Cultural Groups, Multilevel-Selection and Intergroup Violence The issue of prehistoric violence is hotly debated in anthropology, in particular regarding hunter-gatherers before the dawn of agriculture ~ 12,000 years ago (Fry, 2015, 2025; Singh & Glowacki, 2022; Xiao et al., 2026). Pinker (2011: 47–49) presents data suggesting a high prevalence of intra- and intergroup violence from the pre-historic archaeological record. “Pinker’s list” has been harshly criticized for not being representative (Ferguson, 2015), but the “Quasi-Rousseauans” seem to be selective in their reception and interpretation of the state of research as well (Glowacki, 2025; Overy, 2024: chp. 3). According to the Quasi-Rousseauan argument, ancient hunter-gatherers lived almost without possessions, so that intergroup conflict was not beneficial given the risks and costs. Yet being a mobile hunter-gatherer does not necessarily mean that territoriality is unimportant. Richerson and Boyd (2005: 221) cite an illustrative example of an! Kung San man who was confronted by an arrow tip of unfamiliar style. In his view, the arrow tip had been lost by unfamiliar, and therefore dangerous, people who entered the territory. This example illustrates the relevance of symbols for the territorial demarcation of group-boundaries also in hunter-gatherers (Singh & Glowacki, 2022). Moreover, a crucial resource hunter-gatherers might have been fighting for, also between groups, could have been women (Chagnon, 1988: 986; Ehrlich, 2000: 212; Richerson & Boyd, 1998: 81). Institutions and rituals for peace building and conflict resolution (Boehm, 2015) as well as precautious behavior indicates at least that there was a problem to be solved (Glowacki, 2025: 9). Rather than being either a biological drive or entirely learned, humans’ capacity for violence “… is deeply ingrained in us as a means or tool, ever ready to be employed” (Gat, 2022: 234), depending on the respective social situation (Ehrlich, 2000: 211–212; Glowacki et al., 2020). Choi and Bowles (2007) show in their simulation study that the combination of parochialism and altruism might have increased the success in violent intergroup conflict, and thereby increased the reproductive benefit of the parochial altruists. Taking intergroup migration into account, simulations reveal that groups consisting of many parochial altruists is at least one equilibrium state (Choi & Bowles, 2007: 638). Generally, competition and conflict might increase between-group migration (Richerson & Boyd, 1998: 81). In their concluding remarks, Choi and Bowles call for integrating cultural learning processes and peer influence into the simulation models in order to take cultural group selection into account. Bowles (2009: 1297) shows that “lethal group conflict may have been frequent enough to support the proliferation of quite costly forms of altruism”. Again, the transmission process in Bowles’ study is basically genetic. In an earlier study, the effect of between-group competition was combined with the presence or absence of institutions protecting altruists from being exploited by egoists (Bowles, 2004: 448). The relative share of altruists in the population increased with the probability of intergroup conflict, and this increase was even stronger in a scenario where such institutions existed (Bowles, 2004: 464). Moreover, cultural practices of reducing differences in reproductive success, e.g. by food sharing, could have led to genetic group differences strong enough to allow for group-selection, given intergroup competition and despite of between-group migration (Bowles, 2006). However, the efficiency of cultural transmission has not yet been the focus of these studies. In a study on the evolution of self- and other-regarding preferences, Grund et al. (2013) considered cultural socialization, but did not take intergroup conflict into account. The same holds for the models elaborated by Gintis and Helbing (2015: 21–22). These studies did not simultaneously combine the effect of violent intergroup conflict or competition with more or less efficient cultural transmission. Xiao et al. (2026) scrutinize the evolution of parochial altruism by intergroup conflict. Instead, they assume that some individuals engage in violent intergroup aggression in order to feather their own nests, which Xiao et al. (2026) call ‘individual exploitation’. This inclination, however, could have led to a severe disadvantage within the own group, since other group members were afraid of violent retaliation and therefore avoided contact with individual exploiters. Given the prehistorical data, however incomplete it is, there is evidence of inter-group exchange and cooperation (Glowacki, 2024; Grueter, 2026), but one should not assume that encounters between different groups generally proceed smoothly and without tension (Glowacki, 2025). Besides, multilevel selection still works if we replace the assumption of violent conflict with the assumption of intense competition between groups. Based on the Price equation (Bourrat, 2021; Price, 1970), multilevel selection theory combines individual and group selection by adding the intergenerational transmission bias E(wi●Δzi) to the covariance cov(wi,zi) of a trait z and its fitness wi (Bourrat, 2021: 10–15): In this equation, Δzi is the average change in character trait z between parent i and its (several) offspring. Given the data in Table A1, Appendix A1, the covariance of wi and zi is 0.96, the transmission bias E(wi●Δzi) is 0.2, so that w̅Δz̅ = 0.96 + 0.2 = 1.16 (see appendix for details of the computation and derivation). We get a more intuitive interpretation by dividing both sides of the equation by the average fitness w̅. The average change in character trait Δz̅ indicates an average increase in the character value z for each offspring (0.44615, Appendix A1). Yet the parental population can be part of an overarching set of populations. Now, zi will be replaced in the Price equation by z̅, and Δzi will be replaced by Δz̅ on the right-hand side of the equation. The explanandum, the average change in character trait in the offspring population is now ΔZ̅. Note that units of the parental population are in this case not individuals, but populations, or groups, at the next higher level: Since we can now insert the lower-level Price Eq. (2) for Δz̅ in the transmission bias term in (3), this leads to the multilevel version, which can be extended into a theoretically infinite number of levels: In the following simulation, the fitness wi of cooperative behavior might differ between the individual and the group level: defection in the selfish teamwork-game (see below) is beneficial to the individual within groups, but groups compete against each other, so that groups with higher in-group cooperation might have a selective advantage in intergroup competition. According to the Price equation, the transmission bias E(wi●Δzi) is crucial for the change in character across generations, just as is the selection pressure at the higher level, which is in this case the costs of losing in a situation of intergroup competition. Both the level of transmission bias and the costs of intergroup competition are experimental treatments in the simulation of cultural group selection (see below). The Evolution of Humans’ Ambivalent Morality: A Simulation Model The following simulation is based on two experiments based on a set of Nk agents moving randomly in a two-dimensional Netlogo world (Wilensky and Rand 2015) of 51 × 51 patches. For each treatment, the experiment runs 100 repetitions, which results in 300 runs in Experiment #1 (3 treatments) and 800 runs (2 × 4 treatments) in Experiment #2. In both experiments, agents start with a baseline level of energy, drawn from a uniformly distributed random number between 10 and 20. While randomly moving, agents regularly meet other agents on a particular patch. When meeting, they interact, but the character of the interaction depends on whether they meet in-group or out-group agents in Experiment #2 (see below). In Experiment #1, there is no clear distinction between in- and out-group, but a fuzzy boundary exists between egoists and team-workers. The Nk = 100 agents consist of 60% team-workers and 40% egoists. In both experiments, egoists and team-workers generate the payoff matrix shown in the last row in Fig. 2. Team-workers have a probabilistic preference for a spatial area in experiment #1, as described below. Experiment #2 explicitly distinguishes three different groups of Nk = 100 each: high team-workers (T) with 60% cooperators vs. 40% egoists, egoists (E) with 40% cooperators vs. 60% egoists, and a group with norm transmission (N) initially consisting of only 40% cooperators vs. 60% egoists, but whose egoists can turn into cooperators by cultural transmission. Group-Boundaries Group-boundaries show three different degrees of fuzziness in experiment #1, defined by the cooperators’ probabilities of moving to a particular area, and thereby increasing the probability of interacting with other team-workers. The destination area shows considerable fuzziness: the location of the preferred area in the world of 51 × 51 patches is the world’s midpoint (coordinate 0;0) plus a random number uniformly distributed from 0 to 15, added to the midpoint as x and y coordinates. The location of the preferred area is thus somewhere in the upper-right of the squared world of 51 × 51 patches. The treatments in Experiment #1 are the 10% and 20% probabilities of moving to this area, compared to 0% in the control situation of a totally unstructured population. Given the arguments in the theoretical section, team-workers’ tendency towards a preferred location increases the probability of interacting with other team-workers, and thereby reduce the risk of meeting and becoming exploited by egoists. Likewise, the probabilistic withdrawing of team-workers from a major part of the world, where egoists randomly move, reduces egoists’ odds of exploiting cooperators. In Experiment #2 group-boundaries are clearly defined between the groups T and E, but randomly chosen agents of group E can immigrate into N, so that N consists of ~ 15–20% immigrants from group E. Is important to note that emigration out of E does not reduce E’s population size, because the respective agent hatches one offspring before emigrating. Likewise, group N does not grow due to immigration because a randomly chosen actor of N vanishes when an immigrant enters. Due to cultural differences the probability of successful transmission of cooperative norms is only 80% of the transmission probability in the “native” population. Interaction In both experiments two randomly meeting same-group agents play a cooperation game, which is similar to a prisoner’s dilemma. Yet the suckers-payoff (ego’s loss of cooperation if alter defects) is not higher than the individual loss when both agents defect. It is thus a “selfish teamwork” game (Dugatkin 1999: 21, 108). Since the focus is on cooperation, the term ‘team-workers’ is more appropriate than ‘altruists’. Agents do not have a memory so that egoists are able to exploit team-workers in repeated encounters. The payoff increases or decreases the agents’ energy, which is important for reproduction as well as for success in intergroup competition. Intergroup Competition in Experiment #2 From a multilevel selection perspective, groups compete against each other for scarce resources, which is likely to turn into conflict. Groups carried out conflict often by ambushing and small-group hit-and-run attacks, whereby the perpetrators try to minimize their own risks (Wrangham 2019b: 237). The simulation avoids the debate on parochial altruism (Choi & Bowles, 2007) and individual exploitation (Xiao et al., 2026) by modeling social interaction in way that is open for both competition and conflict: If two agents from different groups randomly meet at any patch in Experiment #2, they compete and either have a 25% probability of losing 1 point each, or a 75% probability of carrying out a costly, fierce competition, where the superior loses 1 and the inferior loses either 2 or 5 points (stages of experimental treatment). Winning or losing the competition is determined by comparing the energy of both agents, which is in turn a result of intra-group cooperation. The agent with the lower energy always loses, which implies an advantage for groups whose team-workers generate much group-benefit. Intergenerational Norm-Transmission in Experiment #2 Cultural transmission happens in one out of three groups in Experiment #2. At the beginning, group N has the same share of team-workers as the egoistic group (E), namely just 40%. When meeting an own-group cooperator who is at least 40 time-units older, the egoists in group N have a probability to adopt the norm of cooperation and thereby becoming a team-worker. This transmission probability tp is 20% in the first treatment and 80% in the second (stages of experimental treatment). Although there already is a kind of ‘education system’ in group N in the 20% condition, the system might be inefficient, whereas indoctrination it is highly effective in the 80% condition. Reproduction and Death If agents’ energy is greater than 60 plus a uniformly distributed random number between 0 and 10, they reproduce. The offspring starts with 10 energy units plus a uniformly distributed random number between 0 and 20, and an age of 10 time-units. Since hatching offspring is costly, parents’ energy is set back to 10 units. Whether Group N’s offspring is an egoist or a team-worker does not depend on the parents’ current character, but the trait of team-working again is given with a 40% probability at birth. Over time, agents die for two reasons: They die when they become older than 1000 time-units plus a uniformly distributed random number between 0 and 200. Death also occurs when agents run in a considerable deficit in energy. The threshold of this deficit is energy lower than − 50 plus a uniformly distributed random number between 0 and 50. Unlimited growth of the population would considerably decelerate the simulation. Results Figure 3 shows the share of the team-workers in the population by different levels of boundary-fuzziness, which results from different probabilities of moving to the upper right area of the squared world of 51 × 51 patches. If team-workers can somehow recognize each other or share a preference for a spatial location, the possibility of mutually generating benefits of selfish teamwork, and at the same time becoming less often exploited by egoists, might increase. This is exactly what we find in Fig. 3, although in this illustrative scenario, team-workers do not outcompete egoists if the group boundary is extremely fuzzy (10% probability of moving into the preferred area). Given a spatial boundary of 20%, team-workers quickly start dominating the population. This scenario shows that populations structured by e.g. spatial boundaries or cultural markers might have been sufficient to stabilize cooperative traits in the human population. Experiment # 2 enhances experiment # 1 with the possibility of intergroup competition and cultural transmission. Again, boundaries are fuzzy since there is some migration from E to N (see above). Each panel in Fig. 4 shows the share of the respective group. This is row-wise from top to down: Group N (with Norm transmission, starting with 40% team-workers), Group T (with a high share of Team-workers of 60%), and Group E (with a low share of Team-workers of 40%, and 60% Egoists). The two columns in Fig. 4 represent the treatments of low (20%) and high (80%) norm-transmission in group N. Although this treatment is administered only to Group N, the development of this group affects the remaining two groups, since groups are related by intergroup competition. Each of the six graphs represents an additional treatment: the fierceness of intergroup competition. The pink scattergram and the blue fitted-line represent a scenario where the inferior agent loses 5 points, but also the superior agent loses 1 point. In the scenario with the green scattergram and the red fitted-line, the superior agent of intergroup competition loses 1 point, whereas the inferior agent loses 2 points. This scenario is rather conservative since the superior agent does not benefit from winning the competition, that is, he or she does not acquire goods or reproductive resources. Graph A in the upper left panel shows that under the condition of low norm-transmission probability of 20%, the share of Group N remains stable under fierce intergroup competition, that is, when it is costly to lose intergroup competition (-5 points). Yet if the consequences of losing intergroup competition are less severe (-2 points), the norm-transmitting Group N starts to increase its share after ~ 300 simulation steps. Hence, there is already a scenario of Group N‘s persistence even if the norm-transmission probability is low, or in other words, the indoctrination is inefficient. But better indoctrination helps in intergroup competition. In the scenario with a high norm-transmission probability (80%), both treatments of competition fierceness allow Group N to thrive (Graph B). A high fidelity of cultural norm-transmission allows Group N to compensate its disadvantaged ‘genetic’ starting conditions (compared with T) and thereby increase the share of team-workers, who generate the collective goods required to succeed in intergroup competition. Nevertheless, there is an enormous degree of fanning out in Graphs A and B when the fierceness of group competition is high. Cultural transmission as a strategy to generate within-group cooperation needs thus many attempts. Even high-fidelity cultural transmission does not always generate a sufficient share of team-workers to bring the overall group above the threshold to succeed in intergroup competition. If the norm-transmission probability in Group N is high, Group T starts to decline after ~ 300 simulation steps. Once Group N has passed the threshold of generating a sufficient number of team-workers, group T declines when the norm-transmission probability in Group N is high. The question of whether cultural group selection enables Group N to spread in the overall population thus depends on the competing high-team-workers T, rather than on the low-team-workers/high-egoists of Group E. Graphs E and F show that Group Ealways declines relative to the other groups due to the strong selection effect at the group-level, and this decline is slightly accelerated when the norm-transmission probability in Group N is high. Consequently, Δzi indicates a change in favor of culturally transmitted cooperative behavior zi. The result of cov(wi,zi) is clearly positive, so cooperative behavior zi increased fitness wi. Cultural group selection is a promising path in the evolution of cooperation and prosociality. The more egoistic Group E is inferior anyway, and the fierceness of intergroup competition accelerates its decline. Discussion Since humans do not draw rigid boundaries by pheromonic biomarkers, they must actively create group integration by collective rituals and cultural transmission. Evolution might have prepared humans to quickly learn about cultural markers, norms and symbols of their in-group, which the present study describes by the concept of Durkheiminan learning. Groups and group boundaries are by no means naturally given, but require active maintenance by their members. Conceptualizing groups as fuzzy sets—whose inherent tendency toward disintegration must be actively countered—overcomes common metaphors that view groups as clearly bounded containers (Pietraszewski, 2021: 9). Such a view is further challenged by the concept of multilevel selection, according to which selective forces act across different levels with varying weights. In humans, selection at the group-level primarily works as cultural-group selection, which might be, however, based on evolved psychological dispositions at the individual level: the ‘cultural brain’ allows high-fidelity cultural transmission, our emotionality enables us to experience ‘collective effervescence’ (Durkheim, 1915), to respond with shame and guilt when violating norms (Turner, 2000: 55–56, 2021: 169–170), and also to dehumanize others in intergroup-competition and conflict (Hare & Woods, 2020: 112). This is the moral ambivalence of human sociality. It is likely that social influence and group conformity rely on innate predispositions for constrained learning, of which Durkheiminan learning is a special case. Human sociality is rather a product of nurture rather than nature (Maryanski, 2018: 214), but what might have genetically evolved are dispositions to indoctrinability. Indoctrination is the way how humans substitute pheromonic biomarkers of group membership. This argument is in line with e.g. assuming mental dispositions to adopt “cultural attractors” (Sperber, 1996: 110–112), meaning that “even from the earliest stages of cognitive development human minds seem designed to acquire useful knowledge about their environment”, in particular from other people (Boyer, 2018: 70). To Durkheim, the evolutionary mechanism of multilevel selection was not available. According to the argument in the present paper, once the adaptive value of cultural learning within the own group is in place, the group level gains in importance and individuals become increasingly prepared to Durkheiminan learning. Whether intergroup relations were primarily violent or not is among the most controversial issues in anthropology (Fry, 2015; Gat, 2022: 161–163; Glowacki 2025; Pinker 2011; Wrangham 2019a; Xiao et al. 2026). Group-boundaries and costly competition between groups could be an equivalent mechanism of how cognitive and behavioral dispositions of in-group favoritism and out-group devaluation have evolved. Nevertheless, future biosocial-evolutionary research on group boundaries should more systematically consider the role of positive intergroup relationships (Glowacki, 2024; Grueter, 2026), which could be integrated into an overarching framework of multilevel selection (Bourrat, 2023). The same argument might hold with respect to varying intensities of parochial altruism across different groups (Pisor & Ross, 2024), which is, in the end, an additional expression of the moral ambivalence of human sociality. The argument in the present paper has been developed from the “Californian” gene-culture co-evolution school’s perspective (Heyes, 2018: 31). It conceptualizes Durkheimian learning as an evolved disposition shaped by gene-culture co-evolution, treating it here as a distinct model of constrained cultural learning: Durkheimian learning assumes a genetic preparedness to learn bonding with the group. Accordingly, Durkheimian learning is a component of the social and cultural brain and, only in this regard, the argument assumes modularized cognitive specialization. This is not a ‘radical’ view: it neither assumes a highly differentiated modularization nor the evolution of, for example, a domain-specific ‘species identification’ module generalized to ‘ethnie detection’ via exaptation (Gil-White, 2001: 519). Heyes (2018) challenges the theories of evolved cognitive modularization by her own ‘cognitive gadget’ model: in line with Turner and Maryanski (2024: 79), she assumes the evolution of general-purpose cognitive capacity in increasingly complex cultural environments. This enables humans to quickly and easily adopt ‘cognitive gadgets’ during childhood. According to Heyes’ strong argument, these gadgets are results of cultural evolution. Cognitive gadgets usually fitted to the various different human niches, and could be comparatively quickly replaced with other, more appropriate cognitive gadgets in dynamic environments. For instance, the fact that immigrant assimilation is possible in group N in Experiment #2 is in line with the cognitive gadget model: assimilation implies that cultural transmission between the groups is possible. Probably, Heyes’ view can be reconciled with a minimalist approach which limits domain-specificity to Durkheimian learning. In Heyes view, it is likely that genetic evolution generated dispositions to “enjoy response-contingent stimulation, and therefore to approach and learn from interaction with other agents” (Heyes, 2018: 59). In this sense, the social brain (Dunbar, 2016: 64–66) implies a minimum degree of (networked) modularization. It is difficult, however, to prove the genetic predestination of neural subnetworks (part of the ‘starter kit’ in Heyes (2018: 53) view) for social cognition. On the other hand, if late Australopithecines (~ 2mio years) already lived in comparatively stable groups (Turner, 2021: 42–51), it is unlikely that this social environment did not have any impact on genetic selection of social cognitive and emotional capacities. Turner and Maryanski (2024: 89) argue that it was the selection pressure on enhanced emotions which increased the social bonding among hominins. Refinements of social emotions resulted in dispositions to feel shame and guilt, that are involved in “the moral social control of social relations” (Turner & Maryanski, 2024: 100). The cognitive gadget theory assumes that language comes first and mentalizing (theory of mind) is a result of learning and acquiring cognitive gadgets. Turner and Maryanski (2024: 96–98) highlight, in contrast, that visual and vocal expression of emotions were the ‘first languages’ in early hominins. Given that, if we accept 1. the existence of mirror neurons active in emotional displays and reaction (Bonini et al., 2022: 776), that 2. the mirror neuron machinery develops “genetically canalized” during ontogeny (Bonini et al., 2022: 769), and 3. the fact that hominins evolved in social contexts for more than 2 million years (Dunbar, 2016), we are very close to considering Durkheimian learning via the social and cultural brain an evolved specialization. The interesting debate on evolved domain-specific vs general-purpose cognition, and the degree and scope of specialization, is not yet settled, and a “weak” modularization approach limited to Durkheimian learning does not contradict Heyes’ cognitive gadget model. Future research should continue this debate and also focus more systematically on migration between groups. Is there a limit of immigration that a group can cope with by efficient cultural assimilation? How much inefficiency can be tolerated in the assimilation process? Conclusion According to the first experiment, cohesive and cooperative groups can prevail through natural selection if group-boundaries are sufficiently clear, so that team-workers have an increased probability to meet and to act, and thereby to increase their own reproduction (Wilson, 2014: 38–39). Cultural group selection works despite of fuzzy group boundaries (intergroup migration). If groups actively transmit cooperative norms, they become stronger in intergroup competition. Yet norm transmission requires high fidelity, also with respect to the assimilation of immigrants. The morally ambivalent Durkheimian learning is thus a comparatively fast way of adapting to particular social environments. Further research should more systematically investigate the threshold of how efficient assimilation must be, so that immigration does not undermine team-working and the success of inter-group competition. Nevertheless, as argued with reference to ethology and gene-culture co-evolution theory, this process requires a biological foundation. The cultural brain enables us to become easily indoctrinated by the group. Durkheimian learning is a good candidate to explain how humans compensate their lack of pheromonic markers of group membership (Maryanski, 2018): humans easily transmit culture and bond individuals to the group by collective effervescence and strong emotions. In this view, gene culture co-evolution made us not just capable of developing culture, we became even dependent on and “addicted to culture” (Henrich & Muthukrishna, 2021: 210), and on acquiring cognitive gadgets. Data Availability Code for the agent based modeling, NetLogo and R files available from: https://github.com/mwindzio/durkheimian_learning. References Bahar, S. (2018). Essential tension: Competition, cooperation and multilevel selection in evolution. Springer. Berreby, D. (2008). Us and them: The science of identity. University of Chicago Press. Bloom, P. (2013). Just babies: The origins of good and evil. Broadway Books. Boehm, C. (2000). Hierarchy in the forest: The evolution of egalitarian behavior. Harvard University Press. Boehm, C. (2012). Moral Origins: Social Selection and the Evolution of Virtue and Shame. Perseus Books Group. Boehm, C. (2015). The biocultural evolution of conflict resolution between groups. In D. P. Fry (Ed.), War, peace, and human nature: The convergence of evolutionary and cultural views (pp. 315–340). Oxford University Press. Boesch, C. (2012). Wild cultures: A comparison between chimpanzee and human cultures. Cambridge University Press. Bonini, L., Rotunno, C., Arcuri, E., et al. (2022). Mirror neurons 30 years later: implications and applications. Trends in cognitive sciences, 26(9), 767–781. Bourrat, P. (2021). Facts, conventions, and the levels of selection. Cambridge University Press. Bourrat, P. (2023). A Pricean Formalization of Gaia. Philosophy of Science, 90(3), 704–720. Bowles, S. (2004). Microeconomics: Behavior, institutions, and evolution. Russell Sage Foundation. Bowles, S. (2006). Group competition, reproductive leveling, and the evolution of human altruism. science, 314(5805), 1569–1572. Bowles, S. (2009). Did warfare among ancestral hunter-gatherers affect the evolution of human social behaviors? Science (New York N Y), 324(5932), 1293–1298. Boyd, R. (2018). A different kind of animal: How culture transformed our species. Princeton University Press. Boyer, P. (2018). Minds Make Societies: How Cognition Explains the World Humans Create. Yale University Press. Cacioppo, J. T., & Patrick, W. (2009). Loneliness: Human nature and the need for social connection. W. W. Norton. Chagnon, N. A. (1988). Life Histories, Blood Revenge, and Warfare in a Tribal Population. science, 239(4843), 985–992. Choi, J-K., & Bowles, S. (2007). The coevolution of parochial altruism and war. science, 318(5850), 636–640. Chudek, M., & Henrich, J. (2011). Culture-gene coevolution, norm-psychology and the emergence of human prosociality. Trends in cognitive sciences, 15(5), 218–226. Darwin, C. (1871). The Descent of Man, and Selection in Relation to Sex. Murray. Dunbar, R. I. M. (2016). Human evolution: Our brains and behavior. Oxford University Press. Durkheim, E. (1915). The elementary forms of religious life. George Allen. Ehrlich, P. R. (2000). Human natures: Genes, cultures, and the human prospect. Island. Eibl-Eibesfeldt, I. (1982). Warfare, Man’s Indoctrinability and Group Selection. Zeitschrift für Tierpsychologie, 60(3), 177–198. Eibl-Eibesfeldt, I. (1989). Human ethology. Routledge. Eibl-Eibesfeldt, I. (1998). Us and others: the familial roots of ethnonationalism. In I. Eibl-Eibesfeldt, & F. K. Salter (Eds.), Indoctrinability, ideology, and warfare: Evolutionary perspectives (pp. 21–54). Berghahn Books. Ellemers, N. (2017). Morality and the regulation of social behavior: Groups as moral anchors. Routledge. Ferguson, R. B. (2015). Pinker’s list. Exaggerating pre-historic war mortality. In D. P. Fry (Ed.), War, peace, and human nature: The convergence of evolutionary and cultural views (pp. 113–131). Oxford University Press. Fiske, A. P. (1992). The four elementary forms of sociality: Framework for a unified theory of social relations. Psychological Review, 99(4), 689–723. Fiske, A. P. (1993). Structures of social life: The four elementary forms of human relations. Free. Fry, D. P. (Ed.). (2015). War, peace, and human nature: The convergence of evolutionary and cultural views. Oxford University Press. Fry, D. P. (2025). The quest for scientific objectivity: comment on Luke Glowacki’s (2024) The controversial origins of war and peace: Apes, foragers, and human evolution. Evolution and Human Behavior, 46(1), 106650. Gat, A. (2022). Ideological fixation: From the Stone Age to today’s culture wars. Oxford University Press. Gil-White, F. (2001). Are ethnic groups biological species to the human brain? Essentialism in our cognition of some social categories. Current Anthropology, 42(4), 515–554. Gintis, H., & Helbing, D. (2015). Homo Socialis: An Analytical Core for Sociological Theory. Review of Behavioral Economics, 2(1–2), 1–59. Glowacki, L. (2024). The controversial origins of war and peace: apes, foragers, and human evolution. Evolution and Human Behavior, 45(6), 106618. Glowacki, L. (2025). Robust evidence that mobile hunter-gatherers participated in war: Comment on Fry (2025). Evolution and Human Behavior, 46(2), 106658. Glowacki, L., Wilson, M. L., & Wrangham, R. W. (2020). The evolutionary anthropology of war. Journal of Economic Behavior & Organization, 178, 963–982. Grueter, C. C. (2026). The Evolutionary-Ecological Basis of Xenophilia. Human nature. https://doi.org/10.1007/s12110-026-09521-4. Epub ahead of print 4 May 2026. Grund, T., Waloszek, C., & Helbing, D. (2013). How Natural Selection Can Create Both Self- and Other-Regarding Preferences, and Networked Minds. Scientific reports 3(1480 (2013)). Haidt, J. (2012). The righteous mind: Why good people are divided by politics and religion. Allen Lane. Hare, B., & Woods, V. (2020). Survival of the friendliest: Understanding our origins and rediscovering our common humanity. Random House. Henrich, J. (2004). Cultural group selection, coevolutionary processes and large-scale cooperation. Journal of Economic Behavior & Organization, 53(1), 3–35. Henrich, J. (2016). The secret of our success: How culture is driving human evolution, domesticating our species, and making us smarter. University. Henrich, J., & Muthukrishna, M. (2021). The Origins and Psychology of Human Cooperation. Annual Review of Psychology, 72, 207–240. Heyes, C. M. (2018). Cognitive gadgets: The cultural evolution of thinking. Harvard University Press. Hrdy, S. B. (2009). Mothers and others: The evolutionary origins of mutual understanding. Harvard University Press. Jacobs, R. C., & Campbell, D. T. (1961). The perpetuation of an arbitrary tradition through several generations of a laboratory microculture. Journal of Abnormal and Social Psychology, 62, 649–658. Konner, M. (2010). The evolution of childhood: Relationships, emotion, mind. Harvard University Press. Kurzban, R., Tooby, J., & Cosmides, L. (2001). Can race be erased? Coalitional computation and social categorization. Proceedings of the National Academy of Sciences of the United States of America, 98(26), 15387–15392. Laland, K. N. (2017). Darwin’s unfinished symphony: How culture made the human mind. University. Lorenz, K. (1966). Evolution and modification of behavior. Methuen & Co Ltd. Maryanski, A. (2018). Émile Durkheim and the birth of the gods. Routledge Taylor & Francis Group. Maryanski, A., & Turner, J. H. (1992). The social cage: Human nature and the evolution of society. Stanford University Press. Meltzoff, A. N. (2002). Elements of a developmental theory of imitation. In A. N. Meltzoff, & W. Prinz (Eds.), The imitative mind: Development, evolution, and brain bases (pp. 19–41). Cambridge University Press. Mithen, S. (2006). The singing Neanderthals: The origin of music, language, mind and body. Weidenfeld & Nicolson. Moffett, M. W. (2019). The human swarm: How our societies arise, thrive, and fall. Basic Books. Moon, C., Cooper, R. P., & Fifer, W. P. (1993). Two-day-olds prefer their native language. Infant Behavior and Development, 16(4), 495–500. Muthukrishna, M. (2025). A Theory of Everyone: The New Science of Who We Are, How We Got Here, and Where We’re Going. MIT Press. Overy, R. J. (2024). Why war? Pelican. Pietraszewski, D. (2021). Toward a computational theory of social groups: A finite set of cognitive primitives for representing any and all social groups in the context of conflict. Behavioral and Brain Sciences, 45, e97. Pinker, S. (2011). The better angels of our nature: Why violence has declined. Viking. Pisor, A. C., & Ross, C. T. (2024). Parochial altruism: What it is and why it varies. Evolution and Human Behavior, 45(1), 2–12. Plotkin, H. (1998). Evolution in mind: An introduction to evolutionary psychology. Cambridge. Price, G. R. (1970). Selection and covariance. Nature, 227(5257), 520–521. Richerson, P. J., & Boyd, R. (1998). The evolution of human ultrasociality. In I. Eibl-Eibesfeldt, & F. K. Salter (Eds.), Indoctrinability, ideology, and warfare: Evolutionary perspectives (pp. 71–96). Berghahn Books. Richerson, P. J., & Boyd, R. (2005). Not by genes alone: How culture transformed human evolution. University. Rizzolatti, G., Fadiga, L., Fogassi, L. (2002). From mirror neurons to imitation: facts and speculations. In A. N. Meltzoff, & W. Prinz (Eds.), The imitative mind: Development, evolution, and brain bases (pp. 247–266). Cambridge University Press. Sapolsky, R. M. (2018). Behave: The biology of humans at our best and worst. Vintage Books. Sherif, M. (1935). A study of some social factors in perception. Archives of Psychology, 187, 5–60. Simmel, G. (1964). Conflict and the Web of group affiliation. Collier-Macmillan. Singh, M., & Glowacki, L. (2022). Human social organization during the Late Pleistocene: Beyond the nomadic-egalitarian model. Evolution and Human Behavior, 43(5), 418–431. Sperber, D. (1996). Explaining culture: A naturalistic approach. Blackwell. Sunstein, C. R. (2019). Conformity: The power of social influences. New York University. Tajfel, H. (1982). Social psychology of intergroup relations. Annual Review of Psychology, 33, 1–39. Tomasello, M. (2018). A natural history of human morality. Harvard University Press. Tomasello, M. (2022). The evolution of agency: Behavioral organization from lizards to humans. The MIT Press. Tooby, J., & Cosmides, L. (2010). Groups in mind: The coalitional roots of war and morality. In H. Høgh-Olesen (Ed.), Human morality and sociality: Evolutionary and comparative perspectives (pp. 191–233). Palgrave Macmillan. Trivers, R. (1985). Social evolution. Benjamin/Cummings. Turner, J. H. (2000). On the origins of human emotions: A sociological inquiry into the evolution of human affect. Stanford University Press. Turner, J. H. (2021). On human nature: The biology and sociology of what made us human. Routledge. Turner, J. H., & Maryanski, A. (2024). Biosocial evolutionary analysis: Reconciling biology, psychology and sociology. Edward Elgar Publishing. Wilensky, U., & Rand, W. (2015). An introduction to agent-based modeling. Modeling natural, social and engineered complex systems with NetLogo. MIT Press. Wilson, E. O. (1978). On human nature. Harvard University Press. Wilson, E. O. (2012). The social conquest of earth. Liveright. Wilson, D. S. (2014). Does altruism exist? Culture, genes, and the welfare of others. Yale University Press. Wrangham, R. W. (2019a). Hypotheses for the Evolution of Reduced Reactive Aggression in the Context of Human Self-Domestication. Frontiers in psychology 10. https://doi.org/10.3389/fpsyg.2019.01914 Wrangham, R. W. (2019b). The goodness paradox: How male resentment created tolerance, morality and homo sapiens. PROFILE Books LTD. Xiao, Q., Columbus, S., & Böhm, R. (2026). Self-serving intergroup aggression escalates and prevails over parochial cooperation. Evolution and Human Behavior, 47(1), 106801. Acknowledgements I thank two anonymous reviewers for their very helpful comments, and special thanks to Johannes Huinink and Michael Koch. Funding Open Access funding enabled and organized by Projekt DEAL. This research did not receive funding. Author information Authors and Affiliations Contributions M.W. is the single author and did all the work. Corresponding author Ethics declarations Competing interests The authors declare no competing interests. Additional information Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Supplementary Information Below is the link to the electronic supplementary material. Rights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. About this article Cite this article Windzio, M. Durkheim, Indoctrinability and the Moral Ambivalence of Human Sociality. Hum Nat (2026). https://doi.org/10.1007/s12110-026-09531-2 Received: Accepted: Published: Version of record: DOI: https://doi.org/10.1007/s12110-026-09531-2