The Kin Selection of Religion

I describe an integrative hypothesis for the origin and evolution of human religious cognition and behavior. The hypothesis is based on kin selection, which leads to reduced competition and increased cooperation within family and small-scale groups, and generates novel forms of psychological kinship that foster cooperation in larger scale groups. The 'concept of God' is represented by, and originates in the context of, one's circle of kin and one's ancestors, such that serving God and serving these individuals become synonymous. Kinship pervades, motivates and supports all of the major manifestations of religiosity, including ancestor worship, animism, totemism, monotheism, and culturally based moral codes. The primary selective pressures favoring religious phenotypes are social and ecological selective pressures within and between groups, mainly centered on food acquisition and intraspecific competition. Religiosity, as combined forms of morality and supernatural beliefs founded in kinship, represents groupspecific adaptations that mediate cultural-group survival and reproduction as well as structuring opportunities and constraints for maximization of inclusive fitness. Religion originates in the setting of local, mainly biological kinship, but as groups increase in size, psychological kinship becomes more and more important. The kinship theory of religion is consistent with diverse data from anthropology, evolutionary biology, psychology, psychiatry, endocrinology, and genetics. It is not incompatible with most previous theories for aspects of the origins and evolution of religion, but it grounds them in basic, well-established evolutionary and behavioral-ecological theory. Keywords: religious behavior, inclusive fitness, kinship KIN SELECTION OF RELIGION 3 "There is something sacred about kinship, as most social anthropologists who have studied its operation in the field are prepared to admit" Myers (1975) The purpose of this chapter is to demonstrate that biological and cultural kinship form the core of human religious thought, behavior, and institutions. In doing so, I describe a simple model, centered on kinship, for how religion has evolved, step by step, under natural selection. This model can be broadly considered as a synthesis of Émile Durkheim's and Alfred Radcliffe-Brown's functionalist perspectives from anthropology, psychology, and sociology with empirical developments in evolutionary biology, behavioral ecology, endocrinology, and genetics over the past 50 years. It is founded predominantly on theory and ideas from Hamilton, Alexander, Trivers, Steadman, Coe, Palmer, and Lahti (Hamilton, 1964; Trivers, 1974; Alexander, 1974, 1987, 2013; Coe & Palmer, 2008, 2013; Palmer et al., 2008, 2010; Steadman & Palmer, 2008; Lahti, 2009). The ideas are straightforward and can be empirically tested. Their ramifications are important in that they conceptually unify the meanings of life in biology and spirituality, and provide an intuitive foundation for consilience between them. I begin by explaining the causes of cooperation among animals, including humans, from basic behavioral-ecological and evolutionary theory. Next, I describe genetic relatedness, kinship, and inclusive fitness, and how they are associated with cooperative, and competitive, behavior. Third, I explicate the basics of behavioral ecology, the discipline whereby variation in ecology can be used to explain and predict variation in behavior, and the social systems that result. In the context of this article, religion can, most generally, be regarded as a social system that derives from culture-specific ecological and social adaptation, through the effects of natural selection. Fourth, I discuss the roles of human cognition and affect (emotion and mood) in the origins and evolution of religious thought, behavior, and institutions. KIN SELECTION OF RELIGION 4 Fifth, in the spirits of Durkheim and Radcliffe-Brown, I describe a sequence of hypothesized stages in the evolution of religion, whereby this set of adaptations evolves, step by step, under selection, centrally mediated by effects of kinship. I conclude with a brief synthesis and suggestions for empirical tests. The evolution of cooperation As conceptualized here, religion involves cooperation (in a broad sense of the word) in the context of some combination of morality with the sacred or supernatural. Cooperation between two or more parties, whereby they provide benefits to one another, evolves under three possible conditions (Crespi & Choe, 1997; Lehmann & Keller, 2006). First, cooperation may be mutualistic. In a mutualistic system, both parties gain in the short or long terms from the cooperative interactions, such that eschewing them, or engaging in selfish or exploitative alternatives, are simply not favored. Plants and pollinators represent a simple example, as do humans engaged in trade. Mutualistic systems often evolve in situations where the two parties are phenotypically different in some way or ways that leads them to be able to provide complementary resources or services to one another. However, they may also evolve when two or more parties can jointly benefit from collective action, performing tasks with higher efficacy than could be otherwise be achieved. Mutualisms are ecological, but also social, in that they evolve and are maintained only if one party cannot or does not 'cheat', exploiting the other by receiving benefits without providing gains in return. This situation is epitomized in the 'Prisoner's Dilemma' game, whereby two parties acquire more benefits by both cooperating with one another than by both cheating, but one party would gain even more (than by joint cooperation) by cheating while the other cooperated (Axelrod & Hamilton, 1981). Strategies in the game can evolve towards mutual cooperation, rather than joint cheating (a lack of cooperation), but only if KIN SELECTION OF RELIGION 5 some means or mechanism exists to avoid or suppress cheating. This evolutionary game generalizes across many individuals to a 'tragedy of the commons', whereby a shared resource can be depleted if individuals each act in their self-interest rather than for the good of the group as a whole. In the study of religion, prisoner's dilemma and tragedy of the commons situations are usually represented in terms of the 'free rider problem', referring to individuals who gain the benefits of religious cooperation without paying their fair share of the costs (e.g., Wallis, 1991). Hunter-gatherer groups, within which humans have spent the great majority of their selective, evolutionary history, are typically highly cooperative in mutualistic and egalitarian ways (e.g., Richerson & Boyd, 2001; Whiten & Erdal, 2012). Such cooperation is exemplified by extensive food sharing and cooperative, mutualistic divisions of labor, which have apparently evolved, at least in large part, in the ecological-economic contexts of selective pressures from starvation and malnutrition increasing the risks and effects of disease. As such, from an evolutionary-ecological perspective, human hunter-gatherers should be considered as a species of great ape that has evolved complex, cooperative social-ecological adaptations, specific to their local environment, and supported by large brains and cultural transmission of social and ecological adaptations (Crespi, 2014; van Schaik & Burkart, 2011). The more or less unique culture of each group can be conceptualized as an integrated set of 'cultural survival vehicles', adapted to local ecological and social conditions (Pagel, 2012; Palmer, 2010). These vehicles include, of course, different forms and manifestations of religion. More so than any other primate species, human adaptation is based around cooperating social groups, with most interactions being mutualistic. Second, parties may cooperate due to manipulation or repression (Alexander, 1974; Frank, 2003). Under this model, one party exerts partial or full control over the behavior of the other, but also depends upon it, to some degree, for benefits. Examples include (1) lichens, comprised of fungi and algae, with the former in control of resources; (2) dominance in paper-wasp nests, where the queen behaviorally KIN SELECTION OF RELIGION 6 suppresses worker reproduction and coerces individuals to work; (3) 'policing' in honey bee workers, whereby worker-laid eggs are eaten by other workers; and (4) humans engaged in jointly-beneficial ventures where one party is physically or socially dominant over the other. By this model, the parties are cooperating, though not as equals or free agents, and they are forced (with control taken away) or coerced (with costs actually or potentially imposed on non-cooperators) in some way to act more cooperatively. Humans, like other primates, exhibit asymmetries in physical power, information, abilities related to food acquisition and group defense, social status, and skills, including social leverage through friendship and kinship ties (e.g., Lewis, 2002; Smuts, 1985); these and other variables can, in principle, mediate cooperation due to manipulation or repression. Most commonly, however, the repression of competition and coerced manipulation towards cooperation are some function of the moral codes specific to each culture. Such codes are enforced, preor post-emptively, by the group itself, or some subset of it, usually through aspects of religious cognition, behavior, and institutions. Generally, their role is to preclude or suppress behavior that is personally self-serving (with costs to others, or to the group more widely) or nepotistic (serving kin with costs to others, as described in more detail below) in ways that are detrimental to the interests of the group members, or relatively influential group members (Hughes, 1988). The supernatural or sacred nature of religious moral codes means that they can, in principle, be psychologically and culturally 'enforced' by agents beyond human control, which instantiates fairness and equality in ways that are not subject to dispute or human dominance, with beneficial sequelae to most individuals, most of the time, from moral adherence. The third condition under which cooperation evolves is kinship (Hamilton, 1964). To a biologist, kinship is the sharing of alleles between two individuals due to descent of the alleles from a shared ancestor, such as their father or grandmother. Biological kinship is based on genetic relatedness; it is the probability that a specific allele in one individual (color it blue, conceptually) is also present in another KIN SELECTION OF RELIGION 7 individual. Genetic relatedness values depend on the nature of inheritance regarding the (blue) allele concerned. If the allele is on an autosome, then relatedness from parent to offspring, or vice versa, is onehalf, since meiosis leads to the random transmission of one allele or the other, at a locus, from one individual to the other (see Figure 1). Each additional meiotic link halves relatedness, to one-fourth for aunts and uncles with nieces or nephews, one-eight for first cousins, and so on. We usually consider 'distant' relatedness links, such as those beyond first cousins, as weak; however, from the standpoint of natural selection, links of even a few percent, corresponding to selection coefficients of this magnitude, can generate substantial effects across evolutionary scales of time. In 1964, the eminent evolutionary biologist William D. Hamilton first demonstrated how genetic relatedness is expected to influence the evolution of behavioral interactions between relatives. Hamilton showed that individuals are expected to behave so as to maximize their 'inclusive fitness', which represents one's own reproduction plus one's effects on the reproduction of other individuals devalued by one's genetic relatedness to them. Individuals should thus commonly provide benefits to non-descendent (as well as descendent) kin, because doing so increases their inclusive fitness. This mathematical logic corresponds to a gene's eye view of natural selection; an allele may increase in frequency by having relatively beneficial effects on its bearer (compared to alternative alleles at a locus), and by having relatively-beneficial effects (through the behavior of its bearer) on kin, who probablistically also bear copies of the allele identical by inheritance. Such natural selection involving kinship is sometimes called kin selection. The validity and universal applicability of inclusive fitness theory has been demonstrated in hundreds of studies (Abbot et al., 2011; Bourke, 2011); it is not controversial, and represents a simple extension of Darwinian evolutionary theory to the level of genes. For humans, the centrality of kinship systems to social and cultural behavior (e.g., Déchaux, 2008) attests to the long-term evolutionary KIN SELECTION OF RELIGION 8 importance of inclusive fitness effects in our species, despite post-modernist attempts to discredit kinship studies through political criticisms of the biological and anthropological sciences more generally. Kinship can favor cooperation via altruism, mutualism, or manipulation among kin. Altruistic behavior evolves under kin selection whenever an individual bears a cost in reduced personal reproduction that is outweighed by benefits to non-descendent kin. Mutualistic behavior is favored by kinship because it increases the fitness benefits from such cooperation, since aided individuals who benefit are kin. Additionally, kinship mediates the effects of cooperation by manipulation because, for example, subordinate individuals will be more likely to cooperate with, and help, a dominant individual who is kin than a dominant non-relative (Trubenová & Hager, 2012). The strongest, most general prediction that one can make about behavior is that individuals are expected to (unconsciously) behave so as to maximize their inclusive fitness, barring mistakes and effects from evolutionary novelties (Alexander, 1979, 1987, 1989). Among humans, however, doing so most effectively need not always involve favoritism towards kin, or closer kin, at some cost to more distant kin and non-relatives. Indeed, if every individual in a large social group sought unilaterally and unswervingly to maximize their own inclusive fitness, the result would be pervasively expressed conflicts of interest, and likely social/ecological disaster. What is important to bear in mind here is that although kinship usually engenders some higher level of cooperation in any specific instance, variation in kinship represents one of the strongest factors potentially generating social/behavioral divisiveness within groups, unless mechanisms are in place to ameliorate its disruptive effects. Variation in kinship is important because when variation is greater, individuals are more favored to provide benefits to, and cooperate with, closer kin, and are less favored to help others (more distant kin and non-relatives). For example, at one extreme, we might have two sets of brothers in a group, with cooperative behavior restricted to withinKIN SELECTION OF RELIGION 9 brother dyads. Towards another extreme, we might have a set of four cousins, who should cooperate with each other more or less equally compared to non-relatives. To understand kinship effects, it is thus essential to take into account both cooperation and conflict. Consider two specific cases. First, a child is related to its mother, and its full siblings, by one-half. Our focal child is, however, related to their own (eventual) offspring by one-half, and to the offspring of their siblings (their eventual nieces and nephews) by only one-quarter. This means that the child will be favored to value itself, genetically, twice as highly as it values each sibling. Children in sibships are competing for limited resources from their parents. In this situation, each child 'wants' (i.e., has been subject to selection to take, or solicit) more resources from its mother (or father) for itself than for any sibling. Siblings thus compete with one another for parental resources, in a manner only partially constrained by their relatedness to one another (Bossan et al., 2013; Trivers, 1974). By contrast, the mother is equally related to each offspring, so her inclusive fitness is maximized, all else equal, by the same allocation to each. We have conflicting optima, and expect sib-sib and mother-offspring conflict. Who wins, if anyone, in this situation? It depends on who is in control of what aspects of the interactions, and on the strength of kin selection on each party. If only the mother could, in some ways, use her privileged position as a primary source of childhood enculturation, and morals, to discourage or limit sibling competition, thereby reducing the degree to which copies of her genes interfere with their own transmission. This example illustrates how close kin are often also close competitors, and how all parties concerned can, in principle, benefit in inclusive fitness from reducing the level of wasteful competition between them. As a second case, consider a hunter returning to his village with a large prey item. He might either distribute it among his kin, according to degrees of genetic relatedness combined with benefits gained by each relative, or distribute it equitably among all village members. The former distribution would KIN SELECTION OF RELIGION 10 maximize his inclusive fitness, but only in the absence of a broader social network that provides mutualistic support by culturally-motivated egalitarian food sharing. The hunter may, indeed, suffer a run of bad luck in hunting later on; if he had been relatively selfish and nepotistic, he would be less likely to gain help from others in this situation, and if one's culture did not dictate sharing food more or less equally, the group as a whole would also be less likely to survive and prosper, relative to more-egalitarian groups. This example applies to 'immediate return' food economies, as found in almost all hunter-gatherer groups. Such benefits from the 'moral' sharing of food are ecologically-based, and should be less pronounced in cultures with a higher predictability of food supplies. To a biologist, these applications of inclusive fitness theory, in the context of genetically-based biological kinship, are straightforward. But anthropologists are well aware that biological genetic relatedness, and the actual kinship terms used by individuals in small-scale societies, are by no means always the same (see Figure 1). Why should this be so? A primary reason is that humans learn who their 'kin' are from being told by others, usually in childhood. This information need not be biologically accurate, and should indeed reflect the inclusive fitness interests of the controlling party or parties, and/or the summed interests of the group as a whole. Referring to mother's brother (or all father-aged males in a group) as 'father' may be genetically false, but it may, in some social and ecological contexts, be beneficial to both 'fathers' and 'offspring'. Similarly, calling some categories of cousins 'siblings' may usually be incorrect genetically, but doing so can serve to prevent inadvertent incest, as well as fostering closer kinship bonds between them. At an extreme, systems of 'universal kinship' have developed in many cultures, whereby everyone in a large cultural group has a designated kin relationship to every other member, and everyone outside of the group is non-kin (Barnard, 1978). KIN SELECTION OF RELIGION 11 The key point here is that kinship is cultural and psychological, as well as biological (Bailey & Wood, 1998; Jones, 2000). Genetic relatedness, especially between parents and offspring, was the original context for the generation and maintenance of altruistic and mutualistic social bonds; it creates 'bonded' relationships, motivated by neurological and endocrinological reward systems, whereby individuals provide for one another. But because humans learn (and are told) who their 'kin' are, and are taught proper behavior towards kin and others in the group, kinship in our species is fundamentally psychological (Haig, 2011). Psychological kinship differs notably from biological kinship in that it is culturally malleable (e.g., Qirko, 2004) and can, in principle, promote close kin-like cooperation between any set of individuals in a group. Especially in matrilineal and patrilineal societies, links to shared common ancestors can also reinforce psychological kinship, because all individuals can trace themselves to the same 'parent' or 'grand-parent' (or more distant) kin, thus making the group a single, expanded, “nuclear family” of a sort, with everyone exhibiting the same level of psychological genetic relatedness to an ancestor and to one another (see, e.g., Shiels, 1975) (see Figure 1). Indeed, as noted by Wallwork, (1984), referring to the work of Durkheim: The clan is a kinship group with self-sufficient religious, political and economic functions. Within it, solidarity is maintained by real or fictitious consanguinity. Actual blood relatedness lies at the basis of these ties, but real consanguinity is less important ... than shared beliefs about a common origin. (p. 5) How does inclusive fitness maximization work, then, under both biological and psychological-cultural kinship? Its operation should depend on the social ecology more or less specific to each cultural group. At one extreme, where ecological selection (mainly from food and disease) and social selection (mainly from competing groups, as well as from competition within groups) for group-wide cooperation are weakest, individuals should be most likely to strive to maximize their (biological) inclusive fitness, KIN SELECTION OF RELIGION 12 subject to the effects of competition with other individuals also trying to do so. At the opposite extreme, where ecological and social selection make extensive within-group cooperation essential to survival, individuals should be selected to generally treat all others within the group as close kin in order to reduce the socially-divisive effects of genetic relatedness variation. In this general regard, perceptions and applications of kinship can be considered as culture-specific adaptations, as many anthropologists have described in their discipline-specific ways. However, despite such variability in kinship term usage across cultures, one still expects strong selective pressure for treating kin according to their genetic relatedness, so much as possible in any situation or culture; all individuals will indeed benefit from doing so, so any cultural norm that supports the general and successful ability to maximize inclusive fitness should also spread and be maintained if it does not also generate overly socially-divisive effects. What does psychological kinship have to do with religion? As described in more detail below, it is my thesis that religion represents a form of kinship at its psychological and conceptual-institutional core, but a special kind—one that is grounded in the ideas of the sacred and supernatural, and thus is more effective than biological kinship at promoting cooperative acts. Behavioral ecology of human hunter-gatherers We have discussed kinship, and religion, in terms of adaptation. How, then, do adaptations evolve and generate diversity within and between species in traits such as social behavior, cognition, and aspects of culture? Behavioral ecology is a scientific discipline that seeks to explain adaptive variation in behavior from variation in ecology, within and across taxa. For example, most passerine birds defend territories to ensure access to sufficient food for rearing offspring, and their mating systems are socially monogamous KIN SELECTION OF RELIGION 13 because males benefit from helping to feed and defend the territory and offspring. The ecological situation (food for babies, and a territory that is defensible and contains such food) selects for the behavior (territoriality) and the social system (monogamy). Where food is not defensible, as in many seabirds, territoriality does not evolve. What is the behavioral ecology of humans? We are great apes with especially large brains that live in large cooperative and competitive social groups, who use tools extensively to procure food and transmit ecological and social information culturally. Despite these commonalities, and likely because of them, humans inhabit a vast diversity of ecological habitats, from rain forest to arctic, desert to high mountains. This ecological diversity is expected to generate behavioral and cultural diversity, as each group has evolved, under gene-based natural selection and cultural selection, a set of ecological adaptations that promote survival and reproduction under their particular conditions. Most of these ecological adaptations relate to procuring food, ameliorating abiotic forces, avoiding and surviving disease, and dealing with predators and competitors. Their diversity attests to the flexibility and effectiveness of cultural adaptations, and their key roles in producing the uniquely pronounced behavioral and social diversity of our species. So humans have myriad ecologies and consequent diversity in behavior and systems of sociality. Humans also compete with one another, both within and between groups. We have discussed competition within groups, in the context of inclusive fitness maximizing; competition between human groups appears to have been at least as potent a selective factor in human evolution, although the prehistorical evidence for such competition is indirect (Alexander, 1989; Bowles, 2009). What is important about between-group competition in humans is that (1) it should strongly select for within-group unity and cooperation, and cultural adaptations for defense of the group as well as success in prevailing over other groups (Alexander, 1979; Lahti & Weinstein, 2005); (2) it should be at least as efficacious a selective KIN SELECTION OF RELIGION 14 pressure in ecologically-favorable habitats as in harsh ones, since better environments can support higher population densities and generate stronger demographically-based group competition for the best lands; and (3) it may often involve human groups that treat one another as different species (as indeed they can be, culturally and with regard to interbreeding) and have the capacity to eliminate one another completely. Many small-scale groups thus refer to themselves in some way as 'The People', indicating that other groups are not considered to be human at all. The upshot of these behavioral-ecological considerations is that the primary selective pressures affecting humans throughout most of their evolutionary history are expected to have been both ecological (especially food and disease) and social/ecological (other humans). The origins, evolution, and diversity of religious beliefs, behaviors, and institutions must be set in these selective contexts in order to understand their variable application and relevance for different cultural groups. By what selective, evolutionary mechanisms would religious behavioral/ecological adaptations come about? Within the ecological and social contexts described above, religious adaptations would have evolved under both genetic and cultural selective processes, where 'culture' is conceptualized as humangenerated cognitive information and material objects that are transmitted through social learning and physical inheritance. Most broadly, human 'culture' thus represents aspects of the environment—including other humans, and their ideas—that can be perpetuated vertically (i.e., across generations) and horizontally within groups, as well as horizontally between groups (Alexander, 1979). Like language, religious beliefs and behaviors thus evolve genetically as well as culturally; humans have evolved the genetic capacities for language (for example) and religiosity, but their expression depends on interactions between the effects of genes and the effects of environments. Genes 'for' religion, which are evidenced by statistical associations between allelic variation and variation in measures of religiosity, must therefore exist (e.g., Bradshaw & Ellison, 2008; Kandler & Riemann, 2013) since the cognitive capacity for religion KIN SELECTION OF RELIGION 15 has evolved, but their effects always depend on the environment in which they are expressed. Genes 'for' religion should also, by the thesis described here, represent genes for perception, cognition, and behavior related to kinship and bonding. How, then, have they evolved? Human cognitive ecology The starting point for discussing the origin of religion must, of course, be what it is conceived to be, and what phenotypes were present just prior to its inception. As noted above, we are considering religion as some combination of morality with the sacred or supernatural. Morality here is indirect reciprocity, which evolved from mutualism and altruism among kin (Alexander, 1987) (Table 1), and the sacred or supernatural refer in some manner to gods and other immaterial presences, faith, magic, worship, and veneration. “Supernatural” and “religious” are usually contrasted with “natural” and “secular”; these distinctions need not exist as cultural divisions (and indeed do not exist in most small-scale societies), but phenomena must still vary, somehow, in religious relevance or import. I consider some degree of morality to be present initially, prior to its combining with the supernatural or sacred, in part because conceptions of 'fairness' have been demonstrated across multiple species of primate other than our own (e.g., de Waal, 2013). Cognitive and emotional traits that can be considered as necessary though not sufficient for the origin of religious thought, some of which are unique to humans, include the following: (1) sharing of attention, since religion represents a group-level phenotype in most of its effects; (2) empathic connections or bonds with others, whereby feelings or thoughts link one individual with another in a fitness-salient way, or link two or more individuals together in shared fitness-salient experience; KIN SELECTION OF RELIGION 16 (3) capacity to establish, maintain, and remember a social relationship with specific others even in their physical absence, or after their death; (4) theory of mind, the belief that other individuals or entities have minds, mental states, awareness, motivations, and agency comparable in some way to our own; (5) ability to infer meaning and purpose from events in the world; (6) imagination, the ability to ability to generate or form a mental concept or image of someone or something that is not real or present; (7) social learning, which forms the primary basis for cultural transmission of narratives, cultural norms, and ideas; (8) language, which, like shared attention and social learning, is required for the sharing and transmission of cultural phenotypes with abstract content; and (9) self-awareness, which is required for conception of the 'self' as a 'mind' or entity separate in some way from the 'body'. Such lists of hypothesized preconditions for religiosity have been discussed before (e.g., Wade, 2009), but they have seldom been considered together in the context of a stepwise evolutionary process at the origin of religious phenotypes. It is essential to recognize, initially, that all of the nine phenotypes above evolved in selective situations entirely separate from religious ones. As in the evolution of other qualitatively 'new' traits, the novelty must arise from some process of change or divergence in function (such as fins to legs at the origin of amphibians), or some combination of traits with new, emergent properties (such as the tendency to combine two actions, or ideas, with synergistic effects)—and in both cases, the initial, small, novel change must be selectively favored. Of the phenotypes above, imagination, theory of mind, and selfawareness appear most relevant to the origin of supernatural aspects of religious cognition, because they KIN SELECTION OF RELIGION 17 require something that is imagined and not material, comprising some element of 'mind' or 'agency' that is comparable to that of the self. The closest such entity may well be the 'spirit' of a close relative who has recently died, because they indeed still exist in an immaterial way as thought-patterns and memories in the minds of the living (Lahti, 2009), with an emotional bond, forged by close kinship, connecting the survivor to the deceased (Bowlby, 1999). There is no good reason to believe that supernatural thinking and morality were connected with one another at the onset of supernatural thought; they presumably evolved in parallel at least initially (with morality present beforehand, as noted above), but then combined under some cultural and cognitive circumstances. Indeed, it is this combination that, under our definition, produced the first religious phenomena per se. What selective circumstances, then, might lead to these two phenotypes coming together? At this nexus in our scientific narrative of religious creation, it is important to address—and dispel—a biologically-based conception of the origin of religion that I see as antithetical to evolutionary biology itself. Religion is thus seen by some researchers as arising as a spontaneous byproduct of other traits, as driven by the 'hyperactivity' of agency-detecting mental modules, as 'canalized' by evolved constraints on brain functions, or as maladaptive, non-adaptive, or delusional side effects of selection in other contexts (e.g., Atran, 2002; Atran & Henrich, 2010; Dawkins, 2009; Powell & Clarke, 2012). The primary difficulty with these arguments is that religiosity and religion represent central features of the cognitive and cultural landscapes of virtually all human societies, including hunter gatherers who are believed to resemble our ancestors over tens of thousands to hundreds of thousands or even millions of years. If religious cognition, rituals, and institutions were maladaptive, or neutral and under the radar of selection, their aspects could not have increased in frequency and it would certainly not be universal; this is precisely opposite to the pattern expected under basic evolutionary theory (Bulbulia, 2004). There is no KIN SELECTION OF RELIGION 18 question, at least in my view, that humans have genetically evolved, adaptively, the capacity to conceive, enact, and experience religiosity and culturally-evolved forms of religion that serve core functions in human societies. The questions are how this evolutionary process occurred, and most importantly how to find out. The origins and early evolution of religious thought and behavior We have discussed 'religion' as an entity, but we must remember that it involves a mosaic of cognitive, behavioral and cultural elements, connected by commonalities that include the sacred, supernatural, magical and spiritual. It is the linking of these elements with cooperation and conceptions of morality or 'proper behavior' that constitute the origins of religion. As described above, the general, proximate mechanisms whereby cooperation can evolve include mutualism, manipulation or repression, and kinship. Our goal thus becomes, in part, generating plausible and testable scenarios for how these three processes can join the supernatural or sacred with morality, in the frameworks of typical human social interactions structured by maximizing of inclusive fitness.