2.1. Accounts that do not specifically target historical myths

Our theory does not seek to account for the totality of culturally successful historical narratives. The concept of historical myths specifically refers to a subset of social representations of the past. The peculiarity of historical myths lies in the co-occurrence of highly specific features – they celebrate the group's ancestral origins, a long history of interactions and shared experiences, in a coherent narrative – and their perceived role as a major ideological justification for group solidarity. This peculiarity is sometimes overlooked in existing research on social representations of history and their role in human politics. As a result, some theories may explain why some particular historical elements (characters, events, narratives, etc.) can gain social and political salience, but do not address the specific puzzle of historical myths. To illustrate this point, we review three such theories.

1. (i) First, for instance, authors frequently note that historical narratives are typically used to justify territorial claims. Territorial expansions and the resulting counterinsurgencies are almost systematically supported by historical narratives that present the land of interest as the historical property of a particular group. Overwhelmingly, land property is justified by claiming precedence (“we were here before you”) (see, for instance, Gori, Reference Gori2013). In such cases, historical narratives seem strategically designed to activate specific features of human moral psychology – in particular, the cross-culturally recurrent intuition that ownership – including collective ownership – ought to be assigned based on first possession (Kanngiesser, Rossano, & Tomasello, Reference Kanngiesser, Rossano and Tomasello2015; Verkuyten & Martinovic, Reference Verkuyten and Martinovic2017; Verkuyten, Sierksma, & Martinovic, Reference Verkuyten, Sierksma and Martinovic2015).

2. (ii) A second recurrent explanation for the political use of history stems from the observation that it is increasingly used as a tool to redress historical grievances. In particular, historical narratives are frequently used by minorities to reclaim compensation from states or companies – usually in the form of financial resources or affirmative action. For instance, in the second half of the twentieth century, social movements representing European Jews, African Americans, or immigrants of African descent in Europe have mobilized to obtain compensation for their history of oppression by various states (Henry, Reference Henry2009; Laforcade, Reference Laforcade2006; Traverso, Reference Traverso2016).

3. (iii) Third, as observed by historians, elites have long promoted historical figures as role models to be imitated by the masses (Weber, Reference Weber1976). For instance, the French Third Republic is famous for celebrating role models such as Voltaire, Rousseau, Victor Hugo, Louis Pasteur, and Emile Zola, organizing national funerals and regular commemoration on their birth and death dates (Ben-Amos, Reference Ben-Amos2000). The commemoration of such prestigious individuals probably aims to incentivize similar behaviors by sending a public signal that they are highly valued in society.

All these psychological mechanisms certainly explain why some historical elements can elicit public interest and play an important role in human politics. However, we argue, these theories do not answer our main puzzle, as they do not account for the core features of historical myths and why they seem to play such a foundational role in establishing group cohesion.

It is certain that historical material can be strategically mobilized to claim possession over a territory and seek compensation for past grievances. However, from a legal and ideological perspective, claiming territory only requires a narrative showing historical precedence on a land; and seeking compensations would only require historical evidence of past torts. Yet, historical myths typically do much more in portraying the ancestral bonds that have tied a people together, and the long history of salient cooperative events that have “made the nation” (Liu & Hilton, Reference Liu and Hilton2005; Smith, Reference Smith1999; Thiesse, Reference Thiesse2021; Weber, Reference Weber1976, see Fig. 1 for an overview). It is this sense of deep connectedness and solidarity through time, not mere occupation of a land or specific historical grievances, that most characterizes historical myths (Smith, Reference Smith1999). Likewise, it is very likely that history can be used to incentivize specific behaviors or norms by celebrating historical figures that embody them. However, this theory does not explain the most important features of historical myths, such as why they emphasize the ancient roots of the nation or why a sense of shared history is perceived as a driver of social cohesion.

The discussion of these mechanisms highlights the specificity of historical myths compared to other politically salient information about history. Although we acknowledge the importance of the latter, they were not included in the scope of this article. The main reason is that they have quite straightforward explanations. In all of the listed alternative accounts, the historical material has a relatively clear connection to pressing issues in the present. For example, the interest of minority movements for historical grievances is readily explained by the fact that they have an immediate interest in advertising them to obtain fair compensation. Historical myths, by contrast, are especially puzzling because they commemorate a very distant past or aspects of the group's history with much less obvious impacts on current affairs – such as West African ethnic groups advertising their ancestral existence (Horowitz, Reference Horowitz1977, Reference Horowitz1985) or Indian nationalists reclaiming the legacy of the Aryan civilization (Khan, Svensson, Jogdand, & Liu, Reference Khan, Svensson, Jogdand and Liu2017) –, and yet present this shared history as a defining feature of nationhood. Hence, in the following, we focus on accounts that explicitly try to answer the main puzzle of historical myths, which is why a shared history is perceived as an essential condition for group cohesion in many societies.

2.2. Elite manipulation

By far, the most prevalent explanation for historical myths – but also for nationalism in general – is top-down elite manipulation. In this approach – sometimes called instrumentalism – elites produce historical myths to manipulate the masses for their self-interested purposes (Anderson, Reference Anderson1991; Gellner, Reference Gellner2006; Hobsbawm, Reference Hobsbawm2012; Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983). Case studies show that elites do indeed produce nationalist rhetoric, which in many circumstances seem to reflect self-serving motives (Gellner, Reference Gellner2006; Hechter, Reference Hechter1988, Reference Hechter2000; Hobsbawm, Reference Hobsbawm2012). More recent quantitative work supports such claims: For instance, governments invest more in public schools – which typically promote national ideology – following social unrest (Paglayan, Reference Paglayan2022; see also Solt, Reference Solt2011). Although the elite manipulation hypothesis is well documented and consensual, it does not actually explain why historical myths are so culturally successful. If anything, it reframes the puzzle in a more acute way. Assuming that powerful elites want citizens to commit to the nation and self-sacrifice in wars, and that they are willing to manipulate information to do so, why emphasize the past? Elites may boast the country's current military power, public services, prestige, or make appealing promises for the future. Why then do they also celebrate the long history of the group? Why do self-interested individuals advertise information about the shared past – and not just more directly palatable arguments – to mobilize the masses?

Instrumentalism also has a second limitation. By definition, this approach is focused on the producer's side – the elites – but fails to provide a comprehensive account of the reception of historical myths. In particular, a common assumption of instrumentalist accounts is that the masses are actually indoctrinated by the elites (Boyer, Reference Boyer2018; Gat, Reference Gat2012). For instance, Eugen Weber's famous study of French nation-building in the nineteenth century suggests that the national historical narrative transmitted in French public schools successfully inculcated patriotism in the rural masses (Weber, Reference Weber1976). Although case studies frequently report an apparent correlation between history curricula and the development of a national consciousness, it is not clear that this relationship is causal (Mylonas & Tudor, Reference Mylonas and Tudor2021). Indeed, research on the psychology of human communication shows that humans do not passively absorb whatever cultural norm is in their environment. Rather, they are equipped with cognitive mechanisms for epistemic vigilance, by which they are able to track false information (Mercier, Reference Mercier2017, Reference Mercier2020; Sperber et al., Reference Sperber, Clément, Heintz, Mascaro, Mercier, Origgi and Wilson2010). Moreover, the way humans transmit and receive cultural traits is heavily dependent upon pre-existing cognitive mechanisms (Boyer, Reference Boyer2000, Reference Boyer2007; Mesoudi & Whiten, Reference Mesoudi and Whiten2008; Morin, Reference Morin2016; Sperber, Reference Sperber1996; Sperber & Hirschfeld, Reference Sperber and Hirschfeld2004). Put differently, humans are not so easy to manipulate, and it is not clear that historical myths put forth by self-interested elites should be endorsed by the masses without question (Hirst & Manier, Reference Hirst and Manier2008; Mercier, Reference Mercier2017, Reference Mercier2020). In fact, examples of nationalist propaganda failures abound (Mercier, Reference Mercier2020, pp. 128–141). Supporters of secessionist movements that contest existing political boundaries typically also contest mainstream historical myths and rely on narratives that highlight their historical distinctiveness (for a review, see Coakley, Reference Coakley2012, pp. 94–111; Catalan and Basque nationalism in Spain: Boyd, Reference Boyd1997; Kanak secessionism in New Caledonia: Chappell, Reference Chappell1993).

In sum, the instrumentalist approach, by itself, is insufficient to explain the cultural success of historical myths. First, it fails to explain why producers find it so intuitive to produce narratives of the past when trying to mobilize the masses. Second, it does not explain why people would endorse these myths. We argue that these gaps should be filled by investigating the human cognitive mechanisms that underlie the production and reception of historical myths (Hirst & Manier, Reference Hirst and Manier2008). In short, we need to understand how specific features of human psychology work in order to understand what makes historical myths so appealing for producers and under what conditions they are endorsed by recipients (André et al., Reference André, Baumard and Boyer2023).

2.3. Social identity theory

To understand the psychological roots underlying individuals' belief in the continuity and longevity of their group, it might be useful to turn to one of the most influential accounts of group psychology in the social psychological literature: Social identity theory (SIT; Tajfel, Turner, Austin, & Worchel, Reference Tajfel, Turner, Austin and Worchel1979). The central tenet of SIT is that humans have a spontaneous disposition to classify the world into “in-groups” – groups to which the individual belongs and become part of their identity – and “out-groups.” Because group membership can be a source of pride and self-esteem, individuals thus have a tendency to behave in ways that increase the prestige of their in-groups and, under some circumstances, degrade the reputation of the out-groups (Tajfel, Reference Tajfel1982; Tajfel & Turner, Reference Tajfel, Turner, Worchel and Austin1986; Tajfel et al., Reference Tajfel, Turner, Austin and Worchel1979). Importantly, this psychological disposition may have implications for the spread of cultural information. In particular, SIT predicts that individuals would be more likely to endorse any type of cultural item (stories, myths, songs, etc.) that increase the prestige of the in-group (Tajfel, Reference Tajfel1984). Hence, the SIT explanation for the cultural success of historical myths would be that people endorse these myths because they increase the prestige of their group – and, eventually, in-group members' own self-esteem.

However, this account does not address the fundamental characteristic of historical myths, which is that they are about the distant past. In particular, SIT cannot explain why notions of “ancestral origins” and a sense of shared history can play such an important role in nationalist discourse. Indeed, historical myths are more than just catalogues of past glories: Their essential characteristic is that they present the group as a perennial entity, rather than a recent construction (Anderson, Reference Anderson1991; Berger & Lorenz, Reference Berger and Lorenz2016; Coakley, Reference Coakley2004; Smith, Reference Smith1999; Thiesse, Reference Thiesse2021). One possibility, consistent with SIT, would be that perceiving one's group as ancient and continuous in time somehow increases one's self esteem. But, just like the instrumentalist account, this explanation raises more questions than it answers: Why is it in the first place that people take pride in the deep history of their group? Why would information about the ancient history of the group matter for group members' sense of identity and solidarity?

To solve this puzzle, scholars have suggested the existence of a deeply entrenched “need for continuity” or “need for meaning” in human psychology. For instance, one of the leading scholars of nationalism – Anthony D. Smith – argued that “[b]y placing the present in the context of the past and of the community, the myth of descent interprets present social changes and collective endeavors in a manner that satisfies the drive for meaning by providing new identities that seem to be also very old” (Smith, Reference Smith1999, p. 62). More recently, social psychologists have proposed a “need for self-continuity” in humans to explain the appeal of historical myths (Sani, Reference Sani2010; Sani et al., Reference Sani, Bowe, Herrera, Manna, Cossa, Miao and Zhou2007, Reference Sani, Herrera and Bowe2009; Siromahov et al., Reference Siromahov, Buhrmester and McKay2020; Smeekes, McKeown, & Psaltis, Reference Smeekes, McKeown and Psaltis2017; Smeekes & Verkuyten, Reference Smeekes and Verkuyten2014a, Reference Smeekes and Verkuyten2014b, Reference Smeekes and Verkuyten2017). In this account, individuals develop a belief in the continuity of their group as a way to compensate for their own finitude. However, it remains unclear why exactly humans have such psychological needs in the first place – which is what we want to explain. Second, these theories connect historical myths with general existential needs for meaning making or self-continuity but do not explain why they are also related to more specific concerns about group membership – as suggested by the literature in political science. For instance, it does not explain why many governments design history curricula for compulsory mass schooling – not with a general aim to alleviate children's existential anxiety but to inculcate patriotism (Weber, Reference Weber1976).

2.4. Kin altruism and imaginary kinship

The most influential psychological hypothesis that explicitly considers how historical myths may relate to group solidarity is based on the observation that they are also typically myths of common descent. The “family resemblance” between human groups and kinship networks has been extensively noted and is manifest through the use of kin terms – “brotherhood” or “motherland” – to refer to group members (Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019; Horowitz, Reference Horowitz1985; Van den Berghe, Reference Van den Berghe1987). Kin altruism theory indeed predicts that relatedness should inspire strong feelings of solidarity and cooperative behaviors (Hamilton, Reference Hamilton1964). In line with this idea, scholars often propose that human societies develop cultural techniques to instill a sense of imagined kinship and thus promote costly acts of cooperation (Atran, Reference Atran2016; Whitehouse, Reference Whitehouse2018; Whitehouse & Lanman, Reference Whitehouse and Lanman2014).

Relatedness, though, is unlikely to be able to explain cooperation at the scale of ethnic groups or nations. Indeed, the presumed genetic relatedness of members of the same national community is arguably so low that it should have no impact on cooperative decision making (Jones, Reference Jones2018). Perceptions of relatedness cannot easily misfire either. Because the adaptive level of cooperation with close kin is very high, organisms are incentivized to pose as close relatives in order to attract resources. For this reason, research on kin recognition in human and nonhuman animals suggests that kin altruism is systematically coupled with kin recognition mechanisms that are particularly accurate at detecting fake relatives – individuals that pass as relatives to attract resources from another one (Lieberman, Tooby, & Cosmides, Reference Lieberman, Tooby and Cosmides2007; Park, Schaller, & Van Vugt, Reference Park, Schaller and Van Vugt2008). Kin recognition mechanisms can certainly misfire, but evidence suggests that this only happens in highly specific circumstances that very closely mimic typical kin relations – as in the case of adopted babies or children raised together in Israeli Kibbutzim (Lieberman & Lobel, Reference Lieberman and Lobel2012; Lieberman et al., Reference Lieberman, Tooby and Cosmides2007). Given the high costs associated with national commitment (tax compliance, civic duties, military service, etc.), it is unlikely that the mere evocation of presumed kinship ties in the form of myths of descent is sufficiently credible to substantially alter cooperative behavior at the scale of nations.

In a nutshell, the existing literature agrees that historical myths are somehow important for mass mobilization and group solidarity but fails to explain why. The exact psychological mechanisms by which information about the remote collective past becomes relevant to humans remains mysterious. In the following sections, we propose that the cultural success of historical myths is driven by the adaptive challenge of competitive coalitional recruitment. Historical myths are cultural artifacts designed to attract coalitional support, not because they suggest genetic relatedness, but because displays of a long shared history convey compelling cues of the high degree of fitness interdependence that binds group members.

3.1. Fission and fusion dynamics in human coalition formation

Whether in the ancestral environment or modern times, the process of joining and forming coalitions involves high-stake decisions. Indeed, coalitions are highly beneficial: Organisms that pool resources for a common goal generate fitness benefits that could not be achieved individually (Cosmides & Tooby, Reference Cosmides, Tooby, Hogh-Olesen, Boesch and Cosmides2010; Tooby et al., Reference Tooby, Cosmides and Price2006). Individuals benefit from coalitional support whenever they compete for limited resources: Social status, political power, food, or mates (Redhead & von Rueden, Reference Redhead and von Rueden2021; Cosmides & Tooby, Reference Cosmides, Tooby, Hogh-Olesen, Boesch and Cosmides2010). When competing for scarce resources, more numerous and organized coalitions are more likely to succeed than less efficient groups or isolated individuals (Cosmides & Tooby, Reference Cosmides, Tooby, Hogh-Olesen, Boesch and Cosmides2010). Accordingly, coalitions are widely observed in human societies but also in other cooperative species like dolphins or dogs (Seyfarth & Cheney, Reference Seyfarth and Cheney2012; Wiszniewski, Brown, & Möller, Reference Wiszniewski, Brown and Möller2012). One comparatively unique characteristic of humans is their ability to form large-scale coalitions with unrelated strangers in the form of tribes, ethnic groups, or nations-states. Such large-scale groups can indeed be considered as coalitions, in which individual contribution mainly takes the form of paying taxes, engaging in warfare, and acting civically (Boyer, Reference Boyer2018; Hechter, Reference Hechter2000; Kroneberg & Wimmer, Reference Kroneberg and Wimmer2012; Wimmer, Reference Wimmer2008, Reference Wimmer2018).

Critically, human coalitions are not fixed: Coalitional boundaries frequently change with contextual incentives, and people may belong to more than one coalition at one time. These changes can take multiple forms, but they typically manifest in fission–fusion dynamics, with either subgroups seceding or multiple groups joining forces (De Dreu, Gross, & Romano, Reference De Dreu, Gross and Romano2024; Horowitz, Reference Horowitz1985, pp. 64–74). In large industrialized societies, political scientists have documented multiple cases of such large-scale ethnic boundary change. For instance, in a classic paper, David Posner observed that two ethnic groups, the Chewa and the Tumbuka, were political adversaries in Malawi, but became allies in Zambia to compete with larger ethnic competitors (Posner, Reference Posner2004). Likewise, in many ethnically diverse countries, people must balance their affiliation with their ethnic group (like the Kikuyu) and with a larger national group (like Kenya). The tense relationship between concentrically organized group affiliations has been widely documented by political scientists (Fukuyama, Reference Fukuyama2018; Horowitz, Reference Horowitz1985; Robinson, Reference Robinson2016). This tension is best illustrated in nation-building policies, whereby national elites attempt to secure the social support of citizens to the detriment of more local loyalties (Weber, Reference Weber1976; Wimmer, Reference Wimmer2018). For instance, when the Zionist movement first emerged in the nineteenth century, it was not successful in the diaspora. It took decades of convincing and dramatic historical events to make Jews from initially distinct communities want to engage in a joint political coalition within the Israeli state (Saadoun, Reference Saadoun2012; Traverso, Reference Traverso2016).

Far from being an exclusive property of modern politics, exposure to multiple coalitional arrangements – and thus, the need to navigate multiple coalitional memberships – is an ancient feature of human social life. Humans with highly diverse subsistence modes – including hunter–gatherers – have been able to form various coalitions, including at very large scales, since the Pleistocene (Bird, Bird, Codding, & Zeanah, Reference Bird, Bird, Codding and Zeanah2019; Boyd & Richerson, Reference Boyd and Richerson2022; Singh & Glowacki, Reference Singh and Glowacki2022). Ethnographic data from the Yanömamo, the Waorani, or the Inuit confirm these findings: Individuals in small-scale societies routinely shift their alliances – sometimes fragmenting and other times merging – in reaction to surrounding incentives (Burch, Reference Burch2005; Chagnon, Reference Chagnon2013; Macfarlan et al., Reference Macfarlan, Erickson, Yost, Regalado, Jaramillo and Beckerman2018).

In sum, humans are routinely exposed to social contexts in which they must navigate highly complex coalitional landscapes that require them to optimally invest their limited resources in available coalitions – a process we call coalitional choice (Pietraszewski, Reference Pietraszewski2020). Naturally, this choice is heavily constrained, especially in large-scale settings. A French citizen cannot easily choose to defect from their nation and become a dedicated Chinese citizen. Moving across group boundaries frequently involves prohibitively high transaction and coordination costs – because of linguistic barriers, cultural distance, and moving costs (Amundsen, Reference Amundsen1985; McElreath, Boyd, & Richerson, Reference McElreath, Boyd and Richerson2003; Oh, Selmier, & Lien, Reference Oh, Selmier and Lien2011). Consequently, we expect the mechanisms involved in coalitional choice to be particularly active when deciding whether to splinter or to fusion into a larger group – as these processes generate much less friction and are routinely observed.

Although the human psychological mechanisms underlying coalitional behavior are probably universal, people typically vary in the type of coalition they want to see prosper. This variation has two main sources. First, individuals vary in how they assess a coalition's ability to provide fitness benefits. For instance, Europeans vary in their approval of European integration: Some suggest that building a European coalition is the only way to remain competitive in a globalized world, whereas others do not trust people beyond their national boundaries (Foster & Frieden, Reference Foster and Frieden2021; Kentmen, Reference Kentmen2008; Tanasoiu & Colonescu, Reference Tanasoiu and Colonescu2008). Second, variation can reflect the contextual selfish interests of “political entrepreneurs” who seek leadership positions (Kroneberg & Wimmer, Reference Kroneberg and Wimmer2012; Wimmer, Reference Wimmer2008). For instance, marginalized local elites in multiethnic states can have an incentive to lead a secessionist or regionalist movement in order to increase their relative status (Brass, Reference Brass1991; Hechter, Reference Hechter2000; Schneider & Teske, Reference Schneider and Teske1992). Although individuals differ in their view on such debates, they would all benefit from additional social support to their preferred coalition. To take an example that illustrates a recurrent pattern in modern societies, nation-builders typically need citizens to be loyal to the state, secessionists want to reawaken ethnic mobilization and, on the contrary, supranationalists want to temper down national identity (Horowitz, Reference Horowitz1985; Wimmer, Reference Wimmer2018). Hence, because social support is a limited rival good, people with diverse coalitional interests compete to attract social support in their most profitable coalition. This creates two complementary adaptive challenges: Individuals must not only identify which coalition is most beneficial to them, but also convince others to invest in this particular coalition (e.g., committing to the French nation) rather than to invest in a different coalition (e.g., cooperate at the scale of the European Union) (Lopez, Reference Lopez2020; Pietraszewski, Reference Pietraszewski2020).

3.2. The psychology of coalitional choice: The decisive role of fitness interdependence

How do individuals determine whether they should invest their limited resources in a given coalition? Following a growing line of research, we argue that one of the most important factors that determine the fitness benefits of joining a coalition is the degree of fitness interdependence among group members. Positive fitness interdependence describes the degree to which the fitness of an organism is directly impacted by the fitness of other organisms (Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; Ayers et al., Reference Ayers, Sznycer, Sullivan, Guevara Beltrán, van den Akker, Muñoz and Aktipis2023; Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019; Roberts, Reference Roberts2005; Tomasello et al., Reference Tomasello, Melis, Tennie, Wyman and Herrmann2012). This configuration occurs whenever individuals derive a direct fitness benefit from the continued existence and welfare of others. For instance, the fitness of meerkats is closely tied with the fate of other group members because the size of meerkat groups plays a crucial role in deterring predators – resulting in a strong incentive for meerkats to protect other group members (Clutton-Brock, Reference Clutton-Brock2002; Clutton-Brock et al., Reference Clutton-Brock, Gaynor, Mcilrath, Maccoll, Kansky, Chadwick and Brotherton1999; Kokko, Johnstone, & Clutton-Brock, Reference Kokko, Johnstone and Clutton-Brock2001; Roberts, Reference Roberts2005).

In humans, fitness interdependence has been especially useful to explain cooperation between unrelated individuals in the context of dyadic relationships or small groups. For instance, in human mating relationships, partners are often interdependent in terms of their welfare as well as their reproductive success if they have offspring together. Similarly, in times of war, soldiers in the same unit are highly interdependent, relying on one another for protection and survival. Also, in mutual help systems, individuals share resources with their partners in times of need; this need-based transfer system makes it more likely that both partners will survive and support the other partner (for a review of relationship types that can involve a high degree of fitness interdependence, see Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019, p. 284). In line with this idea, humans appear to have psychological mechanisms that allows them to detect the level of fitness interdependence they have with other individuals in their environment, and adjust their cooperative decisions accordingly (Ayers et al., Reference Ayers, Sznycer, Sullivan, Guevara Beltrán, van den Akker, Muñoz and Aktipis2023; Balliet & Lindström, Reference Balliet and Lindström2023; Colnaghi et al., Reference Colnaghi, Santos, Van Lange and Balliet2023; Columbus & Molho, Reference Columbus and Molho2022; Jin et al., Reference Jin, Columbus, van Lange and Balliet2024; Pleasant, Reference Pleasant2021).

Fitness interdependence has been mostly investigated in small-scale settings (Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; Balliet & Lindström, Reference Balliet and Lindström2023; Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019; Roberts, Reference Roberts2005). When coalitions are sufficiently small, the impact of fitness interdependence on the process of coalitional choice is straightforward: Individuals should be more willing to invest in a coalition if their own welfare is positively correlated with that of other members – as it increases the net fitness benefit of their cooperative action (Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; Colnaghi et al., Reference Colnaghi, Santos, Van Lange and Balliet2023; Jin et al., Reference Jin, Columbus, van Lange and Balliet2024; Roberts, Reference Roberts2005). Yet, configurations of fitness interdependence can also emerge at much larger scales: Members of a very large group can become fitness interdependent to the extent that each individual benefits from the general welfare of other group members (Baldassarri & Abascal, Reference Baldassarri and Abascal2020; Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019; De Dreu et al., Reference De Dreu, Gross and Romano2024). Accordingly, humans can perceive the degree of fitness interdependence that they have not just with other individuals, but with entire groups – including large-scale ones. One important line of evidence in support of this idea comes from the literature on identity fusion. Indeed, identity fusion captures the extent to which individuals perceive their fate to be inseparable from the fate of other individuals – friends, coreligionists, war brethren, and so on – and the extent to which they feel connected by intense kin-like bonds (Swann & Buhrmester, Reference Swann and Buhrmester2015; Whitehouse, Reference Whitehouse2018). As such, identity fusion can be considered as a proximate measure for perceived fitness interdependence (Cronk & Aktipis, Reference Cronk and Aktipis2018). Tellingly, lab experiments and surveys from countries as diverse as Spain, Indonesia, and China show that individuals experience a substantial degree of identity fusion with large-scale entities such as their country or their religious community (Swann et al., Reference Swann, Buhrmester, Gómez, Jetten, Bastian, Vázquez and Zhang2014; see also Swann & Buhrmester, Reference Swann and Buhrmester2015). Hence, just like people can attend to individual-level cues to assess the degree of interdependence they have with individual partners in dyadic relationships, they also seem equipped with cognitive mechanisms that are able to detect coalition-level cues to assess their degree of fitness interdependence with entire groups – including large and abstract ones.

However, in contrast to its application to small-scale settings, fitness interdependence is unlikely to directly incentivize cooperation at the scale of entire nations or ethnic groups. In both cases, fitness interdependence entails that each group member has an interest in the overall welfare of other group members; but in very large groups, each individual's contribution to this general welfare is likely to be negligible. For instance, the war effort or tax contribution of one single citizen is unlikely to significantly affect the plight of other citizens in the country. In this situation as in other social dilemmas where individual contribution is diluted, group members may not have a direct interest to cooperate – despite sharing a perception of fitness interdependence (Olson, Reference Olson1965; Ostrom, Reference Ostrom1990). For this reason, predominant theories of large-scale cooperation typically emphasize cultural systems of monitoring, rewards, and punishment – usually in the form of state institutions – that incentivize cooperation with group members (Acemoglu & Robinson, Reference Acemoglu and Robinson2012; Jin et al., Reference Jin, Columbus, van Lange and Balliet2024; Lienard, Reference Lienard, Van Lange, Rockenbach and Yamagishi2014; Ostrom, Reference Ostrom1990; Powers & Lehmann, Reference Powers and Lehmann2013; Powers, Van Schaik, & Lehmann, Reference Powers, Van Schaik and Lehmann2016). Yet, explanations that center exclusively on the role of such institutions fail to consider one fundamental point: To be efficient and maintain themselves in the long run, they must be perceived by participants as mutually beneficial arrangements, in other words they must appear as morally legitimate (André, Fitouchi, Debove, & Baumard, Reference André, Fitouchi, Debove and Baumard2022). When institutions are perceived as enforcing a coalitional arrangement that does not optimally benefit all citizens, they risk being perceived as irrelevant – at best – or as unfair, extractive, or oppressive (Ostrom, Reference Ostrom1990). This happens, for instance, when secessionist movements or colonized social groups reject existing state institutions, hoping to establish new institutions encompassing different coalitional boundaries that are perceived to enforce a more mutually beneficial coalitional arrangement – typically, in the form of a new independent sovereign state (Horowitz, Reference Horowitz1985; Wimmer, Reference Wimmer2018). Thus, when institutions are widely viewed as morally illegitimate, large-scale cooperation may fail to materialize, because individuals resist these institutions and support alternative forms of coalitional arrangements.

As a result, coalitional recruiters are strongly incentivized not just to create systems of monitoring, rewards, and punishments, but to convince their audience that the cooperative interaction these institutions seek to enforce will prove exceptionally beneficial to them. They must demonstrate to all group members that cooperation within their coalition will bring them substantial fitness benefits – provided that other members do cooperate as well. In particular, in the context of coalitional choice and competition for coalitional support, they need citizens to perceive cooperation in their coalition as the most mutually beneficial coalitional arrangement available to them – in order to increase their commitment to the coalition and reduce their temptation to engage in alternative forms of coalitional arrangement. For instance, French nation-builders in the nineteenth century relied heavily on state power to incentivize the masses to pay taxes and engage in war effort, but they also needed to convince their audience that organizing state institutions at the level of France rather than at a subregional level like Britanny, Corsica, or Provence is more mutually beneficial for the parties involved – and therefore that all French people have a personal interest to commit to this new, emerging large-scale coalitional arrangement (Weber, Reference Weber1976).

We argue that one of the most important pieces of information that strategic recruiters may advertise to increase the perception that cooperation in their coalition constitutes an especially productive coalitional investment – and therefore, to achieve the commitment of citizens to their large-scale coalition – is the degree of fitness interdependence that binds all group members. Indeed, fitness interdependence entails that each individual in the coalition derives a fitness interest from the general welfare of other group members (Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; De Dreu et al., Reference De Dreu, Gross and Romano2024). As a result, the best mutual interest for individuals in this situation is one where everyone invests significant resources to maintain the welfare of other group members. Under these conditions, individuals who perceive themselves as being fitness interdependent should be more supportive of institutional arrangements that organize an extensive cooperative interaction among themselves – and prefer it to alternative arrangements where they are required to help people with whom they are not (or simply less) fitness interdependent.

Additionally, the mental representation that cooperation within a given coalition constitutes a mutually beneficial cooperative arrangement – one that would bring an optimal amount of fitness benefits to all the parties involved – should also manifest as a feeling of moral duty toward the coalition. Indeed, a growing body of literature in moral cognition suggests that individuals who perceive themselves as having especially strong fitness incentives to cooperate together should perceive themselves as having special moral obligations toward each other (André et al., Reference André, Fitouchi, Debove and Baumard2022; McManus, Mason, & Young, Reference McManus, Mason and Young2021; Tomasello, Reference Tomasello2020). This probably contributes to a recurrent finding in moral psychology, which is that individuals who belong to the same coalition feel that they have special obligations toward other group members, even in large-scale coalitions where people don't know each other (Baron, Ritov, & Greene, Reference Baron, Ritov and Greene2013; Cappelen, Enke, & Tungodden, Reference Cappelen, Enke and Tungodden2022). For instance, most American participants expressed strong moral preferences for policies that increased the overall welfare of Americans, even if it came at the expense of global welfare (Baron et al., Reference Baron, Ritov and Greene2013). As a result of this psychological mechanism, perceived fitness interdependence can significantly increase the shared perception that group members – even in a very large groups – have special moral obligations toward each other. This, in turn, increases the reputational benefits that people may reap when acting for the sake of the group as well as the reputational costs of not doing so – which further contributes to the stabilization of large-scale coalitions (André et al., Reference André, Fitouchi, Debove and Baumard2022; Baumard, André, & Sperber, Reference Baumard, André and Sperber2013; Everett et al., Reference Everett, Colombatto, Awad, Boggio, Bos, Brady and Crockett2021; Everett, Pizarro, & Crockett, Reference Everett, Pizarro and Crockett2016).

In sum, perceived fitness interdependence plays a structural role in the process of coalitional choice and in the emergence and stability of large-scale coalitions. When members of a large coalition perceive that they are bound by a significant degree of positive fitness interdependence, they should be more willing to invest their limited resources for the sake of other group members. First, perceived fitness interdependence increases support for a given coalitional arrangement – thus reducing temptations to secede and form alternative coalitions; second, it also increases support for the institutions of monitoring, rewards, and punishments that stabilize this coalitional arrangement, instead of rebelling against them as unfair or irrelevant; and third, by suggesting that group cooperation is a moral duty, it increases the reputational cost of not cooperating. Overall, perceived fitness interdependence is not a substitute to more traditional accounts of large-scale cooperation – which often emphasize institutional constraints and reputational pressures – but a crucial complement to some of their limitations (Acemoglu & Robinson, Reference Acemoglu and Robinson2012; Barclay & Barker, Reference Barclay and Barker2020; Jin et al., Reference Jin, Columbus, van Lange and Balliet2024; Lienard, Reference Lienard, Van Lange, Rockenbach and Yamagishi2014; Ostrom, Reference Ostrom1990; Powers & Lehmann, Reference Powers and Lehmann2013; Powers et al., Reference Powers, Van Schaik and Lehmann2016; Számadó, Balliet, Giardini, Power, & Takács, Reference Számadó, Balliet, Giardini, Power and Takács2021). Institutions do matter, but perceived fitness interdependence explains when people support them. Similarly, reputational incentives do play a decisive role in motivating cooperation in large groups, but perceived fitness interdependence explains why commitment to large and abstract groups – like ethnic groups and nations – come to emerge as a psychologically compelling moral duty in the first place.

Conversely, the decisive role of perceived fitness interdependence in large-scale coalition formation suggests that advertising high levels of fitness interdependence among group members should be an important component of large-scale coalitional recruitment strategies. By displaying cues of positive fitness interdependence among group members, strategic agents may be able to motivate their targets to invest more resources for the sake of the coalition. To do so, coalitional recruiters must come up with a wide range of strategies that aim to efficiently target their audience's social cognition. Historical myths, we argue, are an important component of the arsenal.

Our hypothesis is that historical myths are designed by strategic agents to secure the coalitional support of others by conveying compelling cues of fitness interdependence with other group members. We show that humans attend to specific types of information about the shared history of their group because they can signal information about the fitness interdependence of its members. In turn, this incentivizes individuals to produce historical myths with highly specific features to convey such cues. The specific features of historical myths – their insistence on the ancient and shared past of the nation, and on the collective experiences that group members have gone through – are thus designed to activate specific features of the human cognitive mechanisms that detect fitness interdependence.

4.1. Intuitive beliefs about group continuity

Because of their extensive reliance on social interactions for survival and reproduction, humans are equipped with highly specific cognitive mechanisms for reasoning about social categories (Hirschfeld, Reference Hirschfeld2001, Reference Hirschfeld, Banaji and Gelman2013; Liberman, Woodward, & Kinzler, Reference Liberman, Woodward and Kinzler2017; Rhodes, Reference Rhodes2013; Shutts & Kalish, Reference Shutts, Kalish and Lockman2021). One remarkable feature of human social categorization is the intuitive belief in the temporal continuity of groups. Indeed, humans find it intuitive to speak of collective entities as having a past and a future – that is, an existence outside of the population composing the group at a particular time (Gil-White, Reference Gil-White2001; Sani et al., Reference Sani, Bowe, Herrera, Manna, Cossa, Miao and Zhou2007; Tooby et al., Reference Tooby, Cosmides and Price2006). Importantly, people do not merely believe that groups have a continuous existence in time, but that this is also the case for group-level traits – for instance, that the prevailing values and customs in a given country are part of its temporally stable properties (Gil-White, Reference Gil-White2001; Obradović & Howarth, Reference Obradović and Howarth2018; Roth, Huber, Juenger, & Liu, Reference Roth, Huber, Juenger and Liu2017; Sani et al., Reference Sani, Bowe, Herrera, Manna, Cossa, Miao and Zhou2007, Reference Sani, Herrera and Bowe2009; Siromahov et al., Reference Siromahov, Buhrmester and McKay2020; Smeekes et al., Reference Smeekes, Jetten, Verkuyten, Wohl, Jasinskaja-Lahti, Ariyanto and Van Der Bles2018; Smeekes & Verkuyten, Reference Smeekes and Verkuyten2014a, Reference Smeekes and Verkuyten2014b, Reference Smeekes and Verkuyten2017). In line with this idea, psychometric studies on a wide range of samples consistently find that most individuals indeed view their nation as a permanent entity with transcendent group-level characteristics (Jetten & Wohl, Reference Jetten and Wohl2012; Sani et al., Reference Sani, Bowe, Herrera, Manna, Cossa, Miao and Zhou2007, Reference Sani, Herrera and Bowe2009; Siromahov et al., Reference Siromahov, Buhrmester and McKay2020; Smeekes et al., Reference Smeekes, Jetten, Verkuyten, Wohl, Jasinskaja-Lahti, Ariyanto and Van Der Bles2018; Smeekes & Verkuyten, Reference Smeekes and Verkuyten2014a, Reference Smeekes and Verkuyten2014b, Reference Smeekes and Verkuyten2017; Warner et al., Reference Warner, Kent and Kiddoo2016).

To some extent, this cognitive disposition certainly has an adaptive value. Indeed, many group-level traits tend to persist over substantial periods of time, including in large-scale groups. As documented by a growing body of literature in economic history, many historical contingencies generate path dependencies that can durably affect the social norms, political institutions, and cultural traits of a society for centuries or more (Giuliano & Nunn, Reference Giuliano and Nunn2021; Nunn, Reference Nunn2020; Robinson & Acemoglu, Reference Robinson and Acemoglu2012). Remarkably, such path dependencies can also affect patterns of cooperation within groups. For instance, econometric studies show that countries with an older history of state centralization tend to have higher levels of public goods provision and more inclusive political institutions (Wimmer, Reference Wimmer2018). Likewise, ethnic groups in Africa that have experienced more raids during the transatlantic slave trade centuries ago still report lower levels of social trust to this day (Nunn & Wantchekon, Reference Nunn and Wantchekon2011). These results are consistent with theoretical work in the field of cultural evolution. Indeed, the persistence of cultural traits within a group over a very long time can result from many aspects of cultural evolutionary dynamics, such as environmental stability, conformist bias in cultural transmission, and technological accumulation (Comin, Easterly, & Gong, Reference Comin, Easterly and Gong2010; Gil-White, Reference Gil-White2001; Giuliano & Nunn, Reference Giuliano and Nunn2021; Nunn, Reference Nunn2020; Spolaore & Wacziarg, Reference Spolaore and Wacziarg2013). Just like any adaptive psychological mechanism, intuitions about group continuity can occasionally generate inaccurate beliefs. However, in many cases, paying attention to the deep history of a group can actually provide important information about its characteristics in the present – which probably explains why this intuition appears so psychologically compelling (Smeekes et al., Reference Smeekes, McKeown and Psaltis2017).

This cognitive disposition explains why beliefs about group continuity – for instance, that the French population in 2022 and the population living in the same territory in 500 AD share similar attributes because of some invisible permanence – can appear intuitive instead of being rejected as preposterous (Smeekes et al., Reference Smeekes, McKeown and Psaltis2017). Because humans intuitively think of groups as having time-enduring properties, historical events that affected some group-level trait in the past – for instance, an important war that has brought country members together – can be perceived as having an enduring impact in the present. In fact, the typical structure of historical myths reflects this intuition, as they portray some historical events in the remote past as defining the trajectory of a group for a very long time (Liu & Hilton, Reference Liu and Hilton2005; Smith, Reference Smith1999). For instance, Serbian nationalists portray the battle of Kosovo as a foundational moment in the development of a Serbian national identity, not just in the Middle Ages but to this day (Lomonosov, Reference Lomonosov2021; Malešević, Reference Malešević2022). As a result of this intuitive disposition, information about historical events that could have increased fitness interdependence among group members in the past can be perceived as having durable consequences for a very long time, especially if such events are not isolated but are frequent.

Our hypothesis is that strategic individuals can take advantage of such beliefs to produce and transmit historical myths for self-interested coalitional recruitment purposes. In particular, intuitive beliefs about group continuity make historical myths especially well-suited to activate specific cognitive mechanisms for detecting fitness interdependence among group members.

4.2. Cognitive mechanisms for detecting fitness interdependence

One of the most important drivers of fitness interdependence among nonkin is a history of repeated interactions. Recent evolutionary models show that recurrent and positive interactions between social partners are not only stabilized by the reciprocity mechanism, but also by an additional mechanism: Fitness interdependence (Barclay, Reference Barclay2020; Tooby & Cosmides, Reference Tooby and Cosmides1996). Individuals are highly incentivized to help a frequent reciprocator, such as a friend or recurrent business partner, to maintain their ability to pursue the reciprocal relationship in the future (Barclay, Reference Barclay2020).

Many parameters can influence the level of interdependence that binds individuals in such recurrent reciprocal exchanges. Of particular importance is the parameter of irreplaceability: Individuals have a greater stake in the welfare of a recurrent cooperation partner when this person is harder to replace by another equivalent partner (Barclay, Reference Barclay2020; Tooby & Cosmides, Reference Tooby and Cosmides1996). Conversely, if an individual can easily replace a recurrent cooperation partner by an equivalent one, it may become less costly to shift partners than to help the endangered one. This explains why individuals attend to cues that signal a partner's irreplaceability and typically engage in various strategies to make themselves irreplaceable to their cooperation partners (Tooby & Cosmides, Reference Tooby and Cosmides1996).

The irreplaceability of partners engaged in reciprocal relationships over time can result from a variety of mechanisms. Most importantly, the duration and frequency of the reciprocal relationship itself can make partners harder to replace. Indeed, as individuals repeatedly cooperate over time, they increase their ability to coordinate efficiently – a phenomenon that has been observed in a variety of contexts. In particular, studies in organizational psychology show that teams with members who have more experience working together are more performant (Dubnicki & Limburg, Reference Dubnicki and Limburg1991; Goodman & Leyden, Reference Goodman and Leyden1991; Guzzo & Dickson, Reference Guzzo and Dickson1996; Jehn & Shah, Reference Jehn and Shah1997; Shah & Jehn, Reference Shah and Jehn1993; Watson, Michaelsen, & Sharp, Reference Watson, Michaelsen and Sharp1991). For instance, Watson et al. (Reference Watson, Michaelsen and Sharp1991) showed that the extent to which groups of students were more performant on a standardized assignment than individual members in isolation increased with the time spent in this group. Similarly, participants who reported being close friends performed better in collective tasks than mere acquaintances (Jehn & Shah, Reference Jehn and Shah1997; Shah & Jehn, Reference Shah and Jehn1993). Interestingly, the better performance of close friends was in substantial part mediated by increased interpersonal communication – which suggests an important role of coordination gains. As a result, individuals with a longer history of cooperation are more irreplaceable to one another, because finding a new partner would require to build up all of the acquired coordination gains from zero, which arguably increases the cost of this strategy (Shah & Jehn, Reference Shah and Jehn1993).

In addition, the duration and frequency of a reciprocal relationship also allows for the development of a “raise-the-stake” cooperation strategy. Indeed, although classical evolutionary models for the evolution of cooperation only considered two behavioral options – cooperation and defection (Axelrod & Hamilton, Reference Axelrod and Hamilton1984), more recent scholarship rely on more realistic models, where individuals do not merely choose whether to cooperate or not, but how much they want to cooperate (Roberts & Sherratt, Reference Roberts and Sherratt1998). In such configurations, one especially efficient strategy to maximize the fitness benefits from cooperation without risking too much exploitation is to start cooperating with small amounts of resources and gradually increase them as long as the partner reciprocates the offers (Roberts & Sherratt, Reference Roberts and Sherratt1998). This strategy allows individuals to “test the waters” before engaging in costlier forms of cooperation. Lab experiments confirm the overall logic of this model: When individuals are given the option of choosing the amount of resources they want to engage in a repeated cooperative game, their typical strategy is to gradually increase their contributions over rounds (Kurzban, McCabe, Smith, & Wilson, Reference Kurzban, McCabe, Smith and Wilson2001; Kurzban, Rigdon, & Wilson, Reference Kurzban, Rigdon and Wilson2008; Majolo et al., Reference Majolo, Ames, Brumpton, Garratt, Hall and Wilson2006; Roberts & Renwick, Reference Roberts and Renwick2003; Van den Bergh & Dewitte, Reference Van den Bergh and Dewitte2006). Just like coordination gains, the ability of such “raise-the-stake” strategies to increase the level of cooperation that individuals can achieve makes partners with a longer history of cooperation more irreplaceable to one another – and, consequently, more fitness interdependent. Shifting partners would require to build up the gains from the “raise-the-stake” strategy all over again – at a considerable cost.

Evolutionary models of fitness interdependence usually focus on dyadic relationships, but the same dynamics can occur at much larger scales (Aktipis et al., Reference Aktipis, Cronk, Alcock, Ayers, Baciu, Balliet and Winfrey2018; Barclay, Reference Barclay2020; Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019; De Dreu et al., Reference De Dreu, Gross and Romano2024; Gross et al., Reference Gross, Méder, De Dreu, Romano, Molenmaker and Hoenig2023). Just like in dyadic relationships, large groups where individuals have a long history of fruitful social exchanges develop significant gains in coordination, efficiency, and trust that make group members more irreplaceable to each other – increasing their fitness interdependence. The perceived fitness interdependence that emerges from learning about one's group shared history signals to recipients that their group constitutes the best possible coalitional arrangement available to them (André et al., Reference André, Fitouchi, Debove and Baumard2022). As a result, information about shared history should promote group members' sense of moral duty toward the group, and their commitment to coalitional boundaries and the institutions that establish them (see sect. 3.2).

In line with this idea, repeated interactions increase willingness to cooperate even when they occur at the scale of large groups. For instance, a behavioral experiment shows that as individuals cooperate across group boundaries, they become less prejudiced against out-groups and more likely to cooperate with them – even when the groups are quite large (n = 128 per group) (Gross et al., Reference Gross, Méder, De Dreu, Romano, Molenmaker and Hoenig2023). Hence, even when individuals do not all interact face-to-face, the mere fact of repeatedly engaging in a group-wide collective action establishes strong ties among them (De Dreu et al., Reference De Dreu, Gross and Romano2024). More generally, an important literature on intergroup contact suggests that individuals who are made to interact across group boundaries are less prejudiced toward out-groups in general – and not just toward the particular out-group individuals they interacted with (for a review, see Paluck, Green, & Green, Reference Paluck, Green and Green2019; Paluck, Porat, Clark, & Green, Reference Paluck, Porat, Clark and Green2021). Even at the very large scale of entire nation-states, policies that stimulate intense interactions between citizens from all over the country were found to significantly increase the national commitment of targets (Bazzi, Gaduh, Rothenberg, & Wong, Reference Bazzi, Gaduh, Rothenberg and Wong2019; Cáceres-Delpiano, De Moragas, Facchini, & González, Reference Cáceres-Delpiano, De Moragas, Facchini and González2021; Okunogbe, Reference Okunogbe2018; for a review, see Rohner & Zhuravskaya, Reference Rohner and Zhuravskaya2023). For instance, Spanish individuals from regions with weak Spanish identity (e.g., Basques, Catalans, etc.) who were randomly assigned to perform their military service outside of their region – and thus, had the opportunity to interact with Spanish people from all over the country – increased their self-reported identity as Spanish (Cáceres-Delpiano et al., Reference Cáceres-Delpiano, De Moragas, Facchini and González2021).

In sum, all these findings converge to paint a consistent picture of an important aspect of the human social cognition. Evolutionary models suggest that individuals become more fitness interdependent when they have a history of positive reciprocal relationships, and this effect is amplified when this history is such that it makes it costlier to shift partners. This mechanism can explain the development of fitness interdependence in a simple dyadic relationship, but also in much wider social networks. As a result, individuals are endowed with psychological mechanisms that track ecologically relevant cues of relationship duration, frequency, and intensity. Consequently, individuals exposed to cues that they have been engaged in repeated, long, and intense forms of cooperative exchanges should perceive themselves as highly fitness interdependent. As a result, perceived fitness interdependence in large coalitions should make people more willing to engage in high-stake cooperation in the future – by increasing the stability of institutions and the moral reputation that people can derive from helping group members (see sect. 3.2). In turn, these cognitive mechanisms can interact with humans' intuitive beliefs about group continuity. Indeed, because humans intuitively conceive their group as having time-enduring properties, any fitness interdependence gain acquired from an episode of interactions is assumed to persist over the next generations. As a result, information about the shared history of a group – even when it involves generations of group members over centuries – can be perceived as a reliable cue to infer the degree of fitness interdependence that binds group members.

4.3. The design of historical myths

In light of these findings, historical myths are particularly well-designed to convey cues of fitness interdependence through repeated interaction. By definition, historical myths present the shared and distinct historical experience of a group: Its ancient origins and the succession of important events it has been through (Berger, Reference Berger2009; Berger & Lorenz, Reference Berger and Lorenz2016; Coakley, Reference Coakley2004; Smith, Reference Smith1984, Reference Smith1999). Importantly, this shared history is assumed to be ancient and continuous (Coakley, Reference Coakley2004; Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983; Smith, Reference Smith1999; Thiesse, Reference Thiesse2021). These features are considered fundamental to national cohesion by the elites that typically contribute to spread them. Historians frequently note that many apparently immemorial national traditions were in fact recently “invented” with the clear aim of “establishing or symbolizing social cohesion or the membership of groups, real or artificial communities” (Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983, p. 9).

As an example, the consolidation of nineteenth-century European nation-states required elites to convince the masses to become committed national citizens – and made extensive use of historical myths to do so (Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983). State officials and elites devised new historical narratives that “placed great emphasis on the origins of the nation” and were “intended to retrace the continuity of a collective body through the ages, from its ancient founders to the present” (Thiesse, Reference Thiesse and Demossier2007, p. 20). Importantly, the production of historical myths was driven by the widespread perception that “[what] made the nation was a sense of sharing the same collective history” (Thiesse, Reference Thiesse and Demossier2007, p. 16). Similarly, historical myths were a fundamental part of the nationalist movements in post-Soviet states (Kuzio, Reference Kuzio2002a, pp. 251–254). In these countries, “historians [were] tasked by the ruling elites to claim the right of the indigenous population the privilege of possessing a separate history” (Kuzio, Reference Kuzio2002a, p. 247). In particular, Ukrainian nationalist rhetoric relied heavily on the claim that post-Soviet Ukraine was not a recent invention but rather the continuation of a “1000-year tradition” of statehood (Kuzio, Reference Kuzio2002a). The ancient origins of the Ukrainian state was emphasized to advertise the deep roots of Ukrainian cohesion: “Ukrainians were never an inert mass – but ‘always striving towards liberation and independence’” (Kuzio, Reference Kuzio2002b, p. 209).

The perceived importance of historical myths in fostering group cohesion is not confined to the European continent (for an overview, see Fig. 1). For instance, in Syria, a government decree in 1947 defined the role of history as being “to strengthen the nationalist and patriotic sentiments in the hearts of the people … because the knowledge of the nation's past is one of the most important incentives to patriotic behaviour” (Lewis, Reference Lewis1975, p. 65). Similarly, southeast Asian nationalist movements relied heavily on historical myths for nation-building purposes in the second half of the twentieth century (Salem-Gervais & Metro, Reference Salem-Gervais and Metro2012; Suryadinata, Reference Suryadinata2014). For instance, Burmese nationalist leaders “[attempted] to project ‘Myanmar’ identity backward into ancient history” (Salem-Gervais & Metro, Reference Salem-Gervais and Metro2012, p. 47). Specifically, they celebrated the fact that Burmese people from all ethnic groups had been unified and had lived together since the Pagan Empire – which was founded in the ninth century – and downplayed all historical events that may have signaled disunity among Burmese people. As one official textbook emphasized: “In Pagan era, all the indigenous groups/national races Pyus, Mons, Palaungs, Karens, Taugthus, Thets, Chins, Arakanese, Burmans, Shans etc., united with solidarity to build a Myanmar nation. They lived in harmony” (Salem-Gervais & Metro, Reference Salem-Gervais and Metro2012). Historical myths are also prevalent in many ethnic groups across the world (Clark & Coe, Reference Clark and Coe2021; Horowitz, Reference Horowitz1977, Reference Horowitz1985; Wiessner, Reference Wiessner2018). For instance, among the Yoruba – one of the most numerous ethnic groups in Nigeria –, “[group] consciousness … was mainly created by invoking historical links” among group members (Ajala, Reference Ajala2009, p. 1) and Yoruba leaders explicitly used a “mythological history of origin … to establish a pan-Yoruba identity” (Ajala, Reference Ajala2009, p. 10).

In all these situations, coalitional recruiters insist not only on the fact that their audience would stand to gain from joining forces, but also on the fact that the group has existed as a cohesive entity for a very long time. Why such an insistence on the collective past? Our account suggests an answer: Historical myths show that group members have a remarkably long history of cooperation. The ancestral origins of the group demonstrate that the cohesion that coalitional recruiters ask from their audience is not an ex-nihilo creation, but dates back to centuries or even millennia. And its continuous history, marked by a succession of major collective experiences, shows that group members have been interacting together, solving problems and overcoming challenges for a very long time. As a result of these features, historical myths can compellingly activate the human cognitive mechanisms for detecting fitness interdependence from repeated interactions. Just like two friends with a long history of cooperative interactions become irreplaceable to each other, members of a group with a 1,000-year-old history of cooperative interactions can be perceived as highly interdependent, increasing one's motivation to invest in this particular coalition (Barclay, Reference Barclay2020).

This perspective on historical myths explains their core features, but can also shed light on some puzzling aspects of their manifestation. First, our account can explain why the notion of shared ancestry is considered as such an important component of group cohesion – especially in societies where ethnicity constitutes an important social divide (Clark & Coe, Reference Clark and Coe2021; Gil-White, Reference Gil-White2005). As many anthropologists have noted, myths of the past are often myths of ancestry – sometimes even genealogies (Van den Berghe, Reference Van den Berghe1987; Wiessner, Reference Wiessner and Eibl-Eibesfeldt1998). One possibility is that the rhetoric of shared ancestry is produced to evoke actual genetic relatedness. However, again, it is unlikely that human kin detectors could be easily fooled by cheap verbal relabeling. A more parsimonious explanation is that shared ancestry is but a narrative device that roots the group in a deep past and starts a chain of repeated interactions. Emphasizing a comprehensive shared historical narrative rather than shared ancestry alone can be particularly useful to instill a sense of fictive kinship in populations that already believe in distinct sets of ancestors. For instance, over the twentieth century, Chinese state propaganda has rewritten the official history of non-Han ethnic groups in favor of a narrative “in which the Uyghurs had been a member of the great family of the Chinese nation, and Xinjiang had been party of China since ancient times” (Bovingdon, Reference Bovingdon2001, p. 97). Similarly, early Mexican nationalists who sought to rally indigenous populations to a nation dominated by Europeans crafted a narrative of a “Mestizo country” founded upon the shared history of both people (Gutiärrez, Reference Gutiärrez1999; see also similar historical myths in Peru: Molinié, Reference Molinié2004; and in Ecuador: Foote, Reference Foote2010). In such cases, the use of kin terms most likely reflects computations of fitness interdependence than actual genetic relatedness (Cronk et al., Reference Cronk, Steklis, Steklis, van den Akker and Aktipis2019).

Second, our account explains why historical myths are so often inaccurate and contested by profession historians (Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983). Indeed, our theoretical framework suggests that veracity is not the main force driving the cultural evolution of historical myths. People do not spread historical myths because they are true, but because they perceive that spreading them would benefit them, as they would motivate others to be more committed to the coalition. This explains why historical myths are often wrong, but in a predictable way: They will tend to exaggerate the ancestry and historical continuities of the group even when these claims are not warranted. For instance, although most historians argue that current nation-states are recent political constructions, dating back from the eighteenth, nineteenth, or twentieth century, nationalist ideology tends to reject this idea (Hobsbawm, Reference Hobsbawm1992, Reference Hobsbawm2021; Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983; for experimental evidence, see Smeekes & Verkuyten, Reference Smeekes and Verkuyten2017). In fact, ancestry and continuity appear to be one of the most important grounds on which historians typically contest historical myths (Hobsbawm & Ranger, Reference Hobsbawm and Ranger1983).

However, this does not mean that historical truth plays no role in the elaboration of historical myths. Indeed, coalitional recruiters must craft historical myths that are sufficiently credible to bypass the epistemic vigilance of their audience while still achieving their coalitional objective – and must therefore exploit available historical information (Mercier, Reference Mercier2020). The need for credibility explains why elites typically emphasize visible historical and archeological evidence – ruins, ancient monuments, and so on – in support of historical myths (for instance, see Athanassopoulos, Reference Athanassopoulos2002; Bernhardsson, Reference Bernhardsson2006). For instance, French nationalists are probably more credible when they claim a filiation with the Gauls – a people that did exist and whose existence can be easily verified – than if they claimed descent from an imaginary people (Dietler, Reference Dietler1994; Thiesse, Reference Thiesse2021). In sum, the combination of inaccuracies and kernels of truth in historical myths reflects the tension between the strategic intentions of producers and the epistemic vigilance of receivers (for a similar point made on the cultural evolution of religion, see Fitouchi & Singh, Reference Fitouchi and Singh2022).

5.1. Variability in the prevalence of historical myths

5.2. Variability in the content of historical myths: Why narrative wars?

Our framework also predicts consistent and predictable variability in the content of historical myths. Indeed, because individuals do not share the same coalitional incentives, they support different coalitional boundaries. One major source of variability in boundary preferences is likely to be the perception of unequal treatment of group members. Members of a social group who share a belief that they are treated unfairly may support secession from the majority group, whereas the latter would benefit from maintaining the union. This situation describes most anticolonial conflicts and the claims of many secessionist movements (see, for instance, Elias & Franco-Guillén, Reference Elias and Franco-Guillén2021; Giuliano, Reference Giuliano2015) –, and is supported by quantitative evidence. Using a dataset including representative samples from 123 countries, one study found that members of ethnic groups that were less represented in political institutions expressed less pride in their nation – a common measure of national affiliation (Wimmer, Reference Wimmer2017; see also Wimmer, Reference Wimmer2018). Likewise, other econometric studies show that discriminated groups identify less with their nation and sometimes more with their subgroup (Abdelgadir & Fouka, Reference Abdelgadir and Fouka2020; Dehdari & Gehring, Reference Dehdari and Gehring2022; Fouka, Reference Fouka2020; Green, Reference Green2020). These results show that coalitional preferences indeed vary and that a major driver of variability is the perception of social and political exclusion. This fits with our view of coalitional choice psychology: Excluded members observe that they have little to gain from investing in their current coalition and may find it more profitable to create a coalition of their own.

Consequently, we expect variations in the particular content of historical myths across countries, social groups, and historical periods and is not random but reflects pre-existing salient group affiliations or divides. In a New Zealand survey that asked participants which elements of history should be taught in schools, Māori participants gave more weight to their distinct ethnic past and to Polynesian compared to European history than White participants; and they were more likely to recall historical events predating the arrival of Europeans – thus reflecting the tense relationship between these two communities (Liu, Wilson, McClure, & Higgins, Reference Liu, Wilson, McClure and Higgins1999). Interestingly, historical myths vary across ethnic groups only in contexts where ethnic divides are more salient than national affiliation. In line with this idea, in both Singapore and Malaysia, national identity is very high among all ethnic subgroups; and ethnic and national identity measures are positively correlated – suggesting that nationality, not ethnicity, may be the relevant coalitional boundary in this context (Liu, Lawrence, Ward, & Abraham, Reference Liu, Lawrence, Ward and Abraham2002). Consequently, survey data in these two countries find no significant difference in social representations of history across ethnicities (Liu et al., Reference Liu, Lawrence, Ward and Abraham2002). Hence, historical myth endorsement is not predicted by a blind ethnic preference or a mere preference for cultural familiarity, but by the most salient coalitional divides in one's environment.

One major way in which historical myths can establish their distinctiveness and clarify the type of boundaries that they advertise is by emphasizing some distinctive cultural marker. For instance, in many Western countries, xenophobic individuals tend to endorse historical myths that emphasize the whiteness and Christian roots of their nation whereas individuals who are willing to accept culturally diverse immigrants tend to highlight the historical contribution of immigrants and foreigners to their nation (Moran, Reference Moran2011; Schildkraut, Reference Schildkraut2007; Smith, Reference Smith2012). This idea is supported by quantitative evidence showing that religious Americans are more likely to situate the foundation of America in early religious settlements – emphasizing the religious roots of America; but secular Americans tend to situate it at independence – reflecting deep differences in the perception of the cultural markers that define Americanness (Yamashiro, Van Engen, & Roediger, Reference Yamashiro, Van Engen and Roediger2022). Similarly, reflecting the fact that conservative Americans report narrower moral circles (Waytz, Iver, Young, Haidt, & Graham, Reference Waytz, Iyer, Young, Haidt and Graham2019), results from the quantitative content analysis of Texan history textbooks show that more conservative counties tend to purchase textbooks with less representation of women and Black people (Lucy, Demszky, Bromley, & Jurafsky, Reference Lucy, Demszky, Bromley and Jurafsky2020).