From Life in a Group to Life in a Community

From Life in a Group to Life in a Community Kim Sterelny and Ronald J. Planer To appear in: Stephanie Collins, Brian Epstein, Sally Haslanger, and Hans Bernhard Schmid (eds) Oxford Handbook of Social Ontology. Abstract This chapter (i) begins by emphasising the differences between the social lives of animals (primarily chimps) and those of humans (including humans in deep time). Human social life depends on persistent, collective organisational features of communities (for example, their multi-level character and their economic inter-dependence). (ii) The chapter nevertheless argues that chimp social lives have robust supra-individual, collective features. (iii) These evolve as aggregate outcomes of individual-level selection, but once they emerge, they change the selective landscape. (iv) The chapter then briefly reviews ideas about the evolution of the much more diverse and complex collective features of even the simplest human communities, and (v) discusses ways of integrating methodological individualism with a recognition of the importance of these collective phenomena. (vi) The chapter concludes with a summary and some directions for future research in the area of social ontology and human evolution. Key Words Human behavioural ecology; methodological individualism; human niche construction; evolution of human social complexity; Dennettian cranes; source laws vs consequence laws 1. Two Ways of Living Socially Most humans now live in enormous aggregations with tens to hundreds of millions of fellow citizens. But these aggregations are not mere aggregations, like a vast herd of wildebeest or a cloud of locusts. These enormous human societies are structured by institutions (whatever these turn out to be, if this is a unitary category at all) and other social phenomena: collectively produced and distributed goods and services; legal systems and other less formal mechanisms of normative regulation; libraries, newspapers, and other collective mechanisms of the storage and flow of representations; collectively produced and inherited material products. However, these enormous, state-structured communities are very recent indeed. For the vast majority of our history, from approximately 300 kya to less than 10 kya (kya = thousands of years ago), all humans lived in small stateless communities (for an insightful discussion of just how profound this transformation was, see (Seabright 2010)). Until the Holocene (about 12 kya), all or almost all of these societies were some species of mobile forager communities (perhaps a few were sedentary, based on stored resources or regular harvesting of marine resources). But even in these mobile communities, life is structured by social phenomena. These include kinship systems, which are often elaborate, and depend both on language to provide an explicit and precise terminology, and on testimony, to map, for a given ego, kin terms onto specific individuals. But there are as well systems of norms, and often clans and moieties (moieties divide a community into sections constraining marriage: for example, in Australian forager societies, moieties divide the community into groups which in turn determine for an individual eligible and forbidden marriage partners). Moreover, while residential groups are typically small (perhaps 10 to 20 adults), each has a place within a larger but bounded community. While these small-scale social worlds did not have (and do not have) libraries or newspapers, they had (and do have) collectively produced and maintained lore, narrative, ritual and song, and their language itself; a paradigm collective product. This shared symbolic culture serves to bind smaller residential bands into a coherent community (typically, an “ethnolinguistic group”). (For a superb survey of forager cultures, emphasising their diversity, see (Kelly 2013).) Many animals live socially, but the contrast between the social lives of even mid Pleistocene humans (from about 800 kya) and those of other animals is profound. This is true even of those animals relatively closely related to us, the great apes (but also other mammals and birds). Some animals do indeed live in groups that are large, elaborate, and with cooperation extensive enough to generate interdependence. Most notably, that is true of the social insects and marine siphonophores (e.g., Portuguese man-o-war). It is true also of some mole-rats: in particular, the naked and Damaraland mole-rate, which are the only known eusocial mammals (eusocial animals have specialised reproductive castes: not all animals in a colony have a chance to breed). To a very much lesser extent, it is true of a few species of social carnivores like wolves and the African wild dog. But most animal associations are much looser and less organised, and that includes the lives of our closest relatives, the great apes. We will concentrate on great ape sociality, both because this comparison is especially striking given that they are our close kin, and because the comparison highlights the challenge of explaining the origins of the distinctively human forms of social phenomena. As we see it, the most salient contrasts between great ape lifeways and those of even the most simple mobile forager communities are as follows: 1. Humans live in multi-layered and residentially fluid groups. By comparison, great ape social groups are in general closed. The only regular form of movement into and out of great-ape residential groups is the one-time dispersal of juveniles (typically females) from their natal group. (This serves to reduce inbreeding.) Otherwise, great apes live out their days in the social confines of a single residential group (see, for example, (Stanford 2018)). 2. Humans do not just live in a community. They consciously identify as a member of that community (and are identified as such by others). That identity is often marked by material symbols (distinctive styles of dress and bodily markings), by linguistic codes, and by origin narratives (on some views, the origin of material symbols marking identity is a major upward step in human cognitive evolution; see (Henshilwood and d'Errico 2011), though we incline to the view that its origin marked greater social complexity, not greater intrinsic cognitive capacity). 3. Humans live in communities where social behaviour is regulated by norms of various kinds. For most agents, and many norms, these norms are internalised, with agents having some non-instrumental motivations to conform, though typically there are instrumental motivations as well. Moreover, humans exhibit an eagerness to punish third-parties for norm violation. Quite often, a particular course of punishment and its delivery is codified as part of the relevant norm (Boyd 2016). In contrast, there is very little evidence of third party punishment of anti-social behaviour in great apes, or in nonhuman primates more generally (Silk 2005). Without such punishment, many human cooperative practices would collapse under the pressure of free riding and/or bullying. 4. Humans live in economically inter-dependent communities. Even in the simplest forager communities, there is a sexual division of labour (and often an age-based one as well), which is usually critical to the resilience of their subsistence economy. As human communities become larger, specialisation, division of labour, and interdependence becomes much more marked (Ofek 2001). 5. In great ape communities, social interaction is mediated by individual personal histories of interaction, perhaps with a bit of social referencing, as offspring take cues from their mother’s attitudes. With the exception of that social refencing, there is little role for reputation. Thus, it is no surprise that the presence of a third party makes no difference to the willingness of a chimp to steal food from a conspecific or to help them (Engelmann et al. 2012). In forager communities, gossip massively amplifies the role of reputation in mediating interaction. And while individual history is of great importance, social role also mediates interaction. For example, in many Australian Aboriginal communities, young men behave very differently towards women who are eligible to be their mothers-in-law than they do towards other women. They must be treated very formally, and in some communities it is inappropriate to be alone with them at all. Likewise, the distinction between being an initiate and those yet to be fully initiated mediates important social and economic interactions (see for example (Meggitt 1965)). In the extreme: viewing objects or practices forbidden for one’s social role can result in the ultimate price: execution (see for example see Warner (1958)). 6. Most of what great apes know, they know through their own experience, though sometimes because the activities of others have made an aspect of their environment salient to them. Humans live in an informational commons. Despite deception and unreliability, an enormous amount of information is shared, including, importantly, a shared language. This is most evident inter-generationally. Great apes learn socially, but for the most part, and perhaps entirely, social learning allows them to acquire more easily capacities they could and would acquire on their own (Henrich and Tennie 2017). Human children acquire socially many capacities they could never independently rediscover. For example, in the Marshall Islands of the remote Pacific, oceanic navigation is taught by a mixture of explicit teaching and an apprentice system (see Genz, Aucan et al. 2009). In addition, they learn socially the norms, narratives and customs of their community. These representational systems are not independently rediscoverable, for they do not track a reality independent of social existence. We see the establishment of such representational systems as plausibly being part of what many human evolutionary theorists have in mind by the notoriously slippery notion of “symbolic cognition,” though we leave it to another day to unpack this idea. 7. Great apes have kin, and their kin matter to them. But kin recognition and affiliation depend on their experiential history, depending mainly or entirely on observed patterns of association between known kin (e.g., one’s mother) and other individuals. As a consequence, their kinship networks are attenuated compared to humans; for example, it is not obvious that father-child relations are recognised in chimp communities, and it is hard to see how they could be, as fathers at most preferentially associate with their offspring for a very short period time after birth (likely providing protection against infanticide) (for discussion, see (Silk 2020); see also Murray, Stanton et al. 2016). Human kinship networks are recognised and sustained through experience, but also through testimony and language. So human kin networks are much larger, and kin affiliation and cooperation does not depend on co-residence. In turn, these are the foundations for other kinship-defined social phenomena: clans and moieties. Bernard Chapais has argued that these transformations in kinship were the most crucial changes in human social evolution (Chapais 2009). 8. Human lives are strongly decoupled from the present. Humans invest resources now for future benefit, and do so in highly flexible ways. This flexibility distinguishes the future-oriented behaviour of humans from, e.g., the caching behaviour of many animals. In contemporary industrial societies, they do so to an extraordinary degree: callow youths pay large sums to embark on training regimes from which they expect to profit years hence; young adults pay into superannuation schemes with the hope of enjoying a comfortable life 50 years later. There is a misleading terminology that suggests that this is a feature only of Holocene life; mobile foragers are sometimes described as having “immediate return economies”, as if they did not invest in the future (Woodburn 1982). But this only refers to the fact that many neotropical foragers do not store food. They still invest in the future in their production of artefacts, in the sustained practice through which they develop skills, and their investment in social capital. In mobile forager life, developing social capital — building one’s network of allies and supporters (for example, by exchanging gifts) — is a crucial risk management strategy. It is insurance. Investment in the future is often individually economically rational: agents investing in the future rationally expect the costs of current investment to be more than balanced by future rewards. But that rational expectation is only made possible by the remarkably stable and hence predictable features of human social organisation. It is an extraordinary fact that a New Guinea forager-horticulturalist can cut down a sago palm, and plant another, rationally expecting to profit from that labour in 15 years. Likewise, it is extraordinary that seven thousand years ago, Australian foragers could invest in building stone-reinforced kilometres-long water channels to preserve eel habitat, again expecting to profit from that very substantial labour, presumably for many years (McNiven, Crouch et al. 2012). Investment in the future depends on the stability and hence the predictability of crucial aspects of human social life. That stability depends on norms and institutions, including respect for individual property, or something like property. The archaeological record shows that these phenomena have a deep history in human life. Great apes do not literally live in the present. They make tools that are not used on the instant of their completion, and likely invest in social relations (e.g., by differentially grooming or sharing meat with other individuals). But they do not invest much in the future. Very likely, they do not have the impulse control that would readily allow them to sacrifice current benefit for greater future reward. Sally Boysen is responsible for famously amusing experiments showing the limits on chimp impulse control; see for example Boysen, S. (1996). "More is Less": The Eliciation of Rule-Goverened Resource Distribution in Chimpanzees. Reaching Into Thought: The Minds of the Great Apes. A. Russon, K. Bard and S. Taylor Parker. Cambridge, Cambridge University Press: 177-189. Human impulse control is of course far from perfect too, and we have and need environmental scaffolds to support impulse control. Indeed, Don Ross has argued that economic rational choice models work as first approximations only because we have those scaffolds (Ross 2006, Ross 2008). The point, though, is that selection has never favoured such impulse control in great ape lives. Their futures are not stable and transparent enough to make medium-run planning profitable on average. Human social life thus depends on persistent, collective, organisational features of human communities. In our view, the existence of these persisting collective features can and must be explained by appeal to patterns of decision and action by individual agents. Macro-phenomena have micro-foundations (Elster 2008). But that does not make the macro phenomena less real or important, and these contrasts set up two questions that will structure the rest of this chapter. Q1. In the social sciences, the project of methodological individualism aims to explain social life by explaining the behaviour of the individual agents who collectively constitute the community in question. In our view, the individualist project succeeds when applied to great ape social life. Given that fact, and given the relative simplicity of great ape social organisation, is there such a topic as the social ontology of chimp lives (for we will use the chimps as a model system for mammalian social life in general)? Supposing that there is, what is its theoretical significance? Q2. We have shown above that the social organisation (and hence the social ontology) of great apes is at most very much simpler than that of late Pleistocene foragers. Presumably, very early hominins had social lives falling within the great ape spectrum. If so, how did the much richer set of social phenomena emerge in human lives, as our ancestors evolved? At what stage through the process (if at all) do individualist explanations of human social life become inadequate? There is an enormous literature on the evolution of human social and cultural complexity. We will briefly review a few of the highlights, and focus more on the consequences for the individualist explanatory strategy that suffices for the great apes.  2. Individualism and the Apes Setting humans, social insects and a few other outliers aside, individualism has been fairly uncontroversial in evolutionary biology. Multi-level selection very likely has played a central role in a few key episodes in evolutionary history: the major transitions of Maynard Smith and Szathmary (Maynard Smith and Szathmary 1995). But in most lineages, selection acts on individual members of a population, selecting for individual traits that improve individual fitness and/or that of close kin such as offspring or siblings. We are part of that consensus, conjecturing that “broad-church” behavioural ecology suffices to explain mammalian agency and sociality in general, though we will mostly discuss the Pan species (chimpanzees and bonobos), our closest living relatives. We take broad-church behavioural ecology to include (a) response-independent behavioural ecology: that is, an analysis of those aspects of an agent’s behaviour whose costs and benefits are relatively independent of the decisions of their social partners. To use a common metaphor from game theory: here agents are playing games against nature, rather than against each other. “Nature” will often include the presence of other individuals, but the idea here is that those individuals, if present, are not playing the role of strategic agents. While chimps do travel in small “parties” while foraging in their territory, the act of foraging itself is usually carried out independently, hence most of their subsistence behaviour is response-independent. (b) It includes life history decisions (to the extent that life history is not fixed by inherited biology): the decisions great apes make about birth spacing; their patterns of investment in their offspring, and whether that pattern varies by sex or birth order; the degree and character of their investment in other relatives. (c) It includes strategic interactions of the kinds modelled by game theory. Here the costs and benefits of particular decisions are strongly conditioned by the response of others. Game theory is probably especially relevant to male chimp decision making, and in bonobos to both sexes. For rank in the local dominance hierarchy is of great importance to their life prospects, and strongly influenced by their success or failure in forming coalitions. We think it is likely that great apes can be understood as fitness maximisers. That is, when they have had a chance to learn about the physical and social contingencies of their local environments, and where those environments resemble environments typical of their evolutionary history, great ape decision-making approximates those decisions that would maximise their fitness (and deviation from optimal decision making is noise, not an unexplained bias). Of course, this is a conjecture: there is no way we (or anyone) can establish the truth of this supposition. But nor is there anything in great ape life that looks mysterious or opaque from an individualist perspective. For example, Richard Wrangham has provided a plausible individualist explanation of the suspicious and typically violent relations between chimp residential groups, based on the consequences of unbalanced male numbers (Wrangham 1999). Parties of patrolling males from one group will attack and kill lone males (especially) from neighbouring groups. Wrangham’s imbalance of power hypothesis depends on two ideas. First: the attacking males risk virtually nothing, as they attack only if the odds are massively in their favour. Second, reproductive and subsistence resources are limited, so it would always be in the interests of the males in one community to expand into the territory of a neighbouring community, expropriating those resources. Given that, it is always in their interests to kill a neighbouring male if it is safe to do so: if the neighbouring community has more males, that reduces the power imbalance, making their grip on their own country more secure; if they have fewer males, killing another helps to produce an imbalance sufficient to allow take-over. However, the fact (if indeed it is one) that behavioural ecology offers a fully satisfactory account of the social lives of chimps, does not imply that there are no collective structures in chimp lives. Even if the causal mechanism is individual agency, that mechanism can cause collective outcomes; organisational or structural features of the chimp social world. Those may then become a causally relevant feature of individual decision-making environments. The hostility between male chimps in adjoining residential groups is a structural feature of chimp life, and it has important consequences. Chimps tend to avoid boundary areas, especially if alone or with only one companion. The hostility and the existence of relatively distinct territories forms a context which gives boundary patrols of a group of males their function; and it makes the diffusion of subsistence innovations from one chimp residential group to its neighbours less likely. Nut cracking, and the use of simple stick spears, are both confined to limited numbers of groups. The chimp lineage in an African landscape is divided into a metapopulation of mutually suspicious groups connected only by adolescent female dispersal. This matters, and it is clearly a collective, population-level property. So there is a social ontology of great ape lives: the pattern of their division into residential groups, and the kinds of connections and interactions between those groups. Inter-group connection is a collective property: the fact that only adolescent females can shift between a group and one its neighbours is not a fact about any individual chimp. These ideas about the role of collective properties in chimp social lives require some defence, for there is a delicate issue about what counts as a collective property. Our rough proposal is as follows: a feature of a social world (say, that of chimps) is a collective feature of that world if it stands in no transparent relationship to the causal property or properties of the individuals on which that feature depends. Thus, to take a classic example, G.C. Williams pointed out that a fast herd of horses is just a herd of fast horses. Speed is not a structural or organisational feature of the herd, even though its pace is important for its interactions with predators (Williams 1966). Contrast this with a social insect nest, say, an ant nest. Arguably, the fitness of the nest is biologically important: that is, the nest’s prospects for founding colony nests that themselves establish and reproduce. There is some relation between nest fitness and the fitness of the individual ants, but it is complex and indirect. Most of the ants, the sterile workers, have zero direct fitness. The male drones have some prospects, but almost all will fail (the nest’s prospects are much better than any of their individual prospects). The same is true of the relationship between nest and queen fitness, in those many species where the nests have more than one queen at, or over, time. Colony fitness of course supervenes on the traits of the individual ants and their context. But, again, the relationship is complex and indirect. Colony fitness will be invariant over many different distributions of ants and ant traits. The collective property of chimp social organisation is its division into mutually suspicious groups, allowing only a one-way migration of adolescent females to neighbouring groups. That collective property supervenes on (at least) the following individual traits: (i) basic chimp resource needs; (ii) the fact that chimp mothers feed themselves and their unweaned young without male help (so for males, access to females is always a fitness resource); (iii) the ability of male chimps to form coalitions; (iv) the limited ability of fathers to recognise their offspring (see below). The relationship between the collective property and its individual enablers is not transparent, and so in our view then, chimp residential groups have genuine collective features or properties. We have a somewhat similar view about what makes an aggregate a group. Chimp groups are not mere aggregates. There is a difference between being co-present and being part of a group. One prominent strand in ecological thought is sceptical about the very idea of an ecological community. The advocates of this line of thought suspect that plants (especially) do not care who their neighbours are, and that the distribution of plants across a landscape is largely determined by their individual tolerances of the abiotic aspects of that landscape; rainfall, temperature, soil profile and the like, together with a fair degree of contingency (for a vigorous defense of this view, see (Colinvaux 2017)). If a hillside is populated by pine trees because its aspect and soil suits each of them, those pines are not where they are, acting as they do, because of the presence and behaviour of the others. In contrast, the members of a forager band at a campsite are there because of others, and their actions are sensitive to the actions of the others, often in ways that are congruent rather than merely mirroring others’ actions. This is true of the chimp case too, though their repertoire of mutually sensitive actions is smaller. Genuine groups, as opposed to mere aggregates, exhibit interactional complexity among parts, in turn generating genuine structural or organisational features. On this view, whether a collection of individuals counts as a genuine group, with genuine structural or organisational features will be a matter of degree, but we think that is a good consequence: evolutionary accounts should expect and accept intermediate and hard to classify cases. A flock of pied stilts taking off more or less simultaneously in response to a raptor is a marginal case. A residential group of chimps is a clear case: they recognise one another as individuals, remember one another’s allies, and tune their mutual interactions in the light of this knowledge. While these are collective, organisational properties of chimp social life, they have intelligible micro-foundations. Comparing chimp social life with bonobo social life helps throw these micro-foundations into relief. . Bonobo and chimp communities both have collective properties, but contrasting ones, with bonobo group-group interactions being much less hostile. While there is no uncontroversial explanation of the collective-level difference, there are relevant individual bonobo level differences. Bonobos do not share their habitat with gorillas, and so are less resource stressed. Female bonobos, like female chimps are less aggressive, but unlike female chimps, female bonobos are good at coalition formation. Male bonobos are not (except with their mothers). As a consequence of these individual differences, inter-group interaction is less fraught, and bonobos have no reason to avoid, and hence under-exploit, boundary regions, and males do not patrol in coalition. But reduction is not elimination. The fact that organisational features of the chimp and bonobo lineages can be explained as aggregate outcomes of individual attitudes and behaviours does not imply that those features are not real, nor does it imply that they are causally epiphenomenal. To the contrary: these structural features of (for example) chimp life support robust explanations of its dynamics. It is in virtue of these structural features that male innovation, for example, will very rarely diffuse to adjoining groups. In one region, chimps use very simple spears to extract small animals from tree hollows (Stanford 2018). This is probably not a male innovation, but imagine that it was, and consider the prospects of that innovation spreading and stabilising. They are low, in ways that do not depend on the idiosyncrasies of particular males, or on the specific, individual history of the innovation and innovator. For the only regular channel of information flow between groups is the information and skills a subadult female takes to a neighbouring group when she leaves her natal group. As a young female typically forages with her mother, rather than in mixed sex groups of age-mates or older individuals, she will not routinely see distinctive male skills in action, so is unlikely to acquire and export one. Likewise, as males avoid foraging in border zones, they rarely have the chance to peacefully observe the foraging skills of their neighbours. So male innovations have limited dispersal potential, and this fact is counterfactually resilient. It is dependent on the gross organisation of chimp life, and independent of the specific history of any given innovation. Dan Dennett coined the term “crane” for phenomena that are stable products of routine evolutionary processes that once built, help to structure the environment in ways that make a difference to future evolutionary trajectories (Dennett 1995). Others have used the idea of niche construction to make a similar point: as a lineage evolves it often reshapes its selective environment (and sometimes that of other species (Odling-Smee, Laland et al. 2003), (Sterelny 2003)). The emergence of language in human evolution is one such crane: once humans were language-equipped, their developmental and selective environment was transformed. Chimp and bonobo social organisation is the result of selection acting on individuals in these lineages. But once established, these organisational features shape options and payoffs. Thus male chimps then come to be under positive selection for the social ability to act in coalition with others. The evolution of hominin sociality sees this in spades. Indeed it is arguable that cumulative niche construction — an incremental increase in material and informational resources over the generations — is the single most distinctive feature of our lineage (Sterelny 2012). 3. The Road to Hominin Social Complexity We began this chapter with a list of contrasts between the social lives of humans, even those living in small scale societies, and those of great apes (and other mammals). This contrast fuels the major challenge to the human evolutionary sciences: how did those very special forms of social complexity evolve? In this section, we will briefly review some of the most important suggestions. In the next, we take up a methodological challenge, the implications of this complexity for evolutionary individualism. Multi-Level Society. As already mentioned, one foundational contrast between human social worlds and those of great apes is that between closed and open residential groups. Great apes differ in the degree to which intergroup relations are hostile. But once subadults disperse, their entire social world is their residential group. That is not true of foragers, who live in multi-layered communities. In most forager cultures, at favourable moments, residential groups aggregate, often ostensibly for ritual purposes, but often for sexual, social and information exchange as well. Moreover, residential groups are porous; families move in and out; individuals visit for varying periods. On the parsimonious assumption that early hominins were great ape-like in this respect, this represents a fundamental and deeply consequential change in social organisation, arguably foundational to our capacity to accumulate knowledge, skills, lore and custom over generations. It is difficult to date. Hrdy (2011) argues that, judging by (inferred) erectine interbirth intervals and other life history traits, Homo erectus was an obligate cooperative breeder. Erectine mothers could not have birthed erectine babies so close in age without help. In turn, Hrdy takes this to strongly suggest that erectine mothers had female consanguineal kin (their own mothers, maternal aunts) in their residential groups to care for and provision offspring. If, as is the case with chimps and bonobos, female dispersal was the norm for early hominins, then this would signal the evolution of a more open social world, one in which female kin and possibly others were permitted to move amongst residential groups. On this view, the transition in question evolved (or began to evolve) early in Homo: around 2 mya or so. Better archaeological signatures, in our view, are long-distance movement of exotic materials like obsidian, and human penetration of very demanding environments. This would place the evolution of extended social networks and multi-level community organisation considerably later, in the 800 kya to 100 kya zone (Layton, O'Hara et al. 2012). One reason we find the latter scenario more plausible is that there is an increasing signal of innovation and cumulative technological evolution during this period, as well as of accumulating natural history knowledge. That, as just noted, is precisely what one would expect as human social worlds became more open. In any case: how did this happen? The change was very likely multi-causal. Social expansion probably depended in part on changes in individual cognitive capacity: improved impulse control, better memory, and better theory of mind to manage potentially hostile interaction, and to respond to incipient peaceful interactions. As we have seen, Bernard Chapais has argued that enhanced kin recognition was fundamental. This he takes to be a consequence of pair-bonding in our line. Once fathers routinely recognised and cared about their children, and vice versa, adult males in one residential group could recognise their dispersed children in adjoining groups. That is a potential pathway to peaceful interaction. In addition, a male with dependent young may well welcome assistance from his partner’s female kin, which would help facilitate female migration into the group. We suspect early forms of language were relevant to this process (Planer 2020), for example, in helping to maintain kin affiliation even when kin are no longer co-resident, and making that affiliation less dependent on long histories of intimate interaction. Robert Layton complements this idea with his proposal that the change in hominin trophic level (i.e., position in the food web) leveraged this change in social organisation. As our ancestors became top predators, as with other apex predators, they required large home ranges, and so with an expanding use of space, and a fission-fusion foraging economy, the original residential group camped as a unit less frequently (in fission-fusion socal organisation, a residential group disperses to forage as individuals or small parties, re-uniting at night, though not always every night). Consequently, previously ephemeral foraging associations became the stable unit of daily life. The community, now consisting of a network of bands, is the descendant of the early Pleistocene residential group, and forager bands of ethnographic report are stabilised, enhanced descendants of once ephemeral foraging associations (Layton and O' Hara 2010, Layton, O'Hara et al. 2012). Norms. Human social life is structured by norms, and by religion and ritual. There is nothing similar in the lives of great apes. Capuchin monkeys have been shown to avoid certain acts that generate unequal outcomes for themselves and a partner (Fletcher 2008; de Waal et al. 2008, Brosnan et al. 2010), a tendency which has been described as an evolutionary precursor of a sense of fairness. However, this characterization is contentious not least because the tendency is one-sided: the acts capuchins avoid are ones that provide a larger benefit to the recipient than the actor (them). Similarly, a form of chimpanzee rock-throwing behavior has been dubbed by some as “ritualistic,” if not “proto-religious.” The behavior in question concerns the apparently functionless throwing of rocks at the base of trees (and in one case, into a hole at the base of a tree), producing large stone accumulations (Kühl et al. 2016). This finding was sensationalized to a ridiculous extent in popular media coverage (see Stanford 2018, p. 171-2 for a discussion of the behavior). There is a rough consensus that these features of human life contribute to making uniquely human forms of cooperation possible, and that this helps explain their evolution (see for example (Kitcher 2011, Henrich 2016, Tomasello 2016, Boyd 2018). But there is no consensus on finer-grain detail. Consider first a few of the main ideas about norms. Michael Tomasello (2016) has argued that norms evolved to support high stakes cooperation. Such cooperation requires trust, for one partner stands to suffer serious costs if the other (or others) drops the ball. So they must trust one another, and normative commitment, as shown in their personal history, engenders trust, as it shows that one’s partners’ choices are relatively insensitive to instrumental considerations and immediate temptations. Philip Kitcher (2011) has argued that normative commitment evolved to support cooperative interactions dispersed in space and time. In face to face interactions, social emotions are often enough to motivate good-faith cooperation. But imagining others’ disappointment and disapproval is much less vivid and powerful than seeing it. So in settings without face to face contact, commitment to norms partially controls temptations to free-ride. Peter Richerson and Rob Boyd (2018)have argued that norms make meso-scale cooperation possible. Once cooperative partnerships hit a threshold scale (about 4 agents), cooperation cannot be stabilised by the contingent willingness to cooperate if but only if others do. For withdrawing cooperation in response to free-riding imposes a cost on those others who did not free-ride, triggering a general retreat from cooperation. For cooperation to be stable, costs must be imposed specifically on the free-rider. But for any given individual imposing that cost on his or her own will often be prohibitively expensive. Indeed, in some cases it could be potentially lethally dangerous. So norms are needed to motivate collective response to free-riding through normative outrage, and to make what counts as free-riding less ambiguous. Real world interactions are complex and noisy, and without fairly explicit rules, it will often not be obvious what counts as free-riding. We have a dog in this fight too. Sterelny has argued that norms are a response not to the scale of cooperation but to its economic complexity (Sterelny 2014, Sterelny 2021). We will not adjudicate these debates here, but instead confine ourselves to three comments. First: this is only a sample of the theoretical options, and obviously more complex hybrid scenarios are also possible. Second, these are all explanations of the stability of norms as a central plank of human social lives once they have emerged. They are not origin explanations. None explain the evolution of a norm-sensitive mind from a mind innocent of them (for once such proposal, see (Birch forthcoming)). Third: these rival proposals are to some extent empirically testable against the ethnographic record, as they make different predictions about the pattern of variation in the elaboration of norms, and in the extent of social investment in their teaching. For example, Tomasello’s proposal predicts that we should see greater elaboration of, and investment in, norms in communities with more high-stakes cooperation. Religion and Ritual. There is even less consensus on religion and ritual, in part because it is likely that its role has changed over time, and in part because religion and ritual pose an extra puzzle, that of explaining beliefs that seem patently irrational. So one strand of the evolution of religion literature develops proposals about its social functions. These vary considerably. Some emphasise the experiential aspects of ritual and religion, identifying its main effect as that of building community cohesion through shared experience of song, dance and narrative (for example, see (Lewis 2015, Lewis 2016)). Another approach treats religion and ritual as a signal one community sends to others, signalling strength and commitment, with the costs of those practices increasing signal credibility (Sosis and Alcorta 2003). On these views, religion and ritual is one of the mechanisms through which human communities come to be aware of themselves as distinct communities, and one of the main mechanisms through which individuals come to identify with and as a member of their own community. In contrast, the ideological aspects of religion are central to theories that propose that religion has a virtual policing function. The gods see and punish deviations from norms, and this enhances compliance even when no-one is looking (Wilson 2002). The “big gods” hypothesis, proposing a link between large scale social worlds and extra-powerful, moralising gods is one version of this idea (Norenzayan 2013). Timing is a key constraint on hypotheses in this area. Ethnography suggests that ritual will leave material traces: evidence of the use of space for non-utilitarian purposes; the production of material symbols, with the use of ochre, shells, beads and the like (for example to decorate masks); figurines and cave art; special ways of disposing of the dead. Rituals, and the songs and material symbols they feature, likely played a crucial role in consolidating religious ideas, as well as transmitting them down the generations; accordingly, they were probably an evolutionary pre-condition for the emergence of stable, cohesive bodies of supernatural belief (Mithen 2001). Indeed, on some views of the evolution of religion (e.g., Sterelny 2018), early religion consisted primarily or entirely in such ritual practices, with religious ideologies emerging only later. But signs of ritual practice are all relatively recent. There is one impressive and enigmatic cave site at about 175 kya (a Neanderthal site). There are earlier uses of ochre (the first uses are earlier than 200kya (McBrearty and Brooks 2000), but ochre has utilitarian uses too). But the other traces are from about 100kya. So a defender of any of these functionalist proposals about religion needs to explain why robust evidence of ritual behaviour only appears in the archaeological record from around 100 kya. To the extent that the function in question (e.g., building group cohesion) was important to the lifeways of earlier hominins, then the late appearance of ritual traces is a challenge for the theory. That is clearly the case for some psychological functionalist theories, such as those that claim religion functions to reduce existential angst (see for example Landau et al. 2007). For considerably more ancient humans plausibly had an awareness of death and dying, and anxiety surrounding it (Stiner 2017). There is no fully satisfactory account of the credibility of religious belief (i.e., of why individuals are apt to find the contents of such beliefs believable). Joseph Henrich and others have shown that once these beliefs establish in a community, their persistence is readily understood (Henrich 2016). The commitment of the adult generation enhances their credibility to their children. But this does not explain initial establishment. The most popular suggestion is that religious belief is a side-effect of our acute sensitivity to the presence of other, potentially hidden, agents. The idea is that there is a cost asymmetry between falsely believing someone is there, when no-one is; and believing that no-one is there, when an agent is indeed present. Imagine, for example, you are making your way through a forest. It is much worse to falsely believe that all is clear than it is to falsely believe that a predator is lurking. Since no detection system can be perfect, we are tuned to err on the side of believing that there are agents present where there are none, and religious belief piggybacks on this (Atran 2005). On analysis though, the supposed cost asymmetry is bogus. Late Pleistocene humans were top predators; animals were in danger from them, rather than vice versa. Moreover, false positives (belief an agent is present; no agent) are expensive: in natural environments, they would sharply degrade foraging efficiency. In social environments, they become paranoia. Social Roles. The most fundamental difference between humans and other animals is our dependence on cooperation. We are obligate co-operators, and cooperate on a temporal and spatial scale with non-relatives in ways that no other animals do. The literature on this is vast beyond the possibility of brief review (but see (Sterelny 2021)), so our aim here is to connect cooperation to social role, and its importance in mediating our social life. One of the main ways social role does this is by conferring upon our social lives predictability over time, as it shapes both the actions of role occupants—of mothers-in-law, initiates, unmarried women and the like—and the actions of those with whom they interact, the behavior of the latter being shaped by their own social roles. We begin with the point that some forms of cooperation are not organisationally complex: think of a mob of hominins chasing a predator from a kill by shouting, waving sticks, throwing stones. We think these were probably foundational for other forms of collective action that involve teamwork, with coordination and role division. In making a safe camp, one agent constructs a hearth, another collects fuel, a third uses a fire drill to ignite tinder; others erect a rough shelter and a windbreak. Teamwork is much more flexible and efficient if, on necessity, roles can be swapped. Michael Tomasello has argued that in our lineage this flexibility is supported through the evolution of a distinctive form of action representation, “bird’s-eye view” representation (Tomasello 2014). This kind of representation encodes the structure of a joint or collective action independently of the specific agents involved in the action This makes role reversal much more straightforward (I now collect fuel, you help with the windbreak). We are good at this, but great apes find it difficult. They do not naturally excel at joint action tasks. With chimps, for example, monkey hunting is mob-work, not teamwork (though see Boehm (2012) for a dissenting view), and male patrols do not involve distinct roles. When in experimental settings they have learned to play their part in a joint action, they do not routinely transfer with ease to the complementary role (see, e.g., Melis et al. 2005). This has been seen as evidence that the chimps have an impoverished understanding of the nature of the cooperative task and/or the effect of their partner’s actions. We find Tomasello’s suggestion that bird’s-eye view representation gives us the capacity for fluent role reversal persuasive, and with such representation comes the distinction between role and occupant. For this schema specifies distinct roles in joint or collective action independently of representing their occupants. This is the conceptual machinery needed to identify and tune one’s behaviour towards an individual qua her social role as potential mother-in-law, rather than qua her specific personality and history. On this picture, the evolution of the capacity for flexible teamwork bought with it a distinctive cognitive capacity to conceptualise distinct social roles. With language, our lineages as well acquired the ability to name those roles. Once those social and cognitive capacities were in place, roles multiplied and their importance increased with changes in the scale and complexity of human communities. In our view, multi-level sociality; social norms and their individual internalisation; practices of religion and ritual; social roles and the role-occupant distinction are all Dennettian cranes. They evolve as consequences of natural selection on individual agents. But once they evolve, they become important structures in the theatre of action, often making possible more powerful and complex forms of cooperation. 4. Individualism and Social Complexity. “men make their own history, but they do not make it just as they please; they do not make it under circumstances chosen by themselves, but under circumstances directly encountered, given and transmitted from the past” (Marx, 18th Brumaire of Louis Bonaparte, 1852) Evolutionary and economic explanations are a species of population-based explanation. In such explanations, an outcome is the result of iterated, aggregated and somewhat stereotyped interactions amongst a fairly small number of types of agents, responding to one another and to features of their common environment. A classic example in the human sciences is one standard explanation of the great demographic transition in Europe, the dramatic fall in average fertility over the eighteenth, nineteenth and early twentieth century. The fall was at different times in different countries, but it was very widespread, and often in the teeth of pro-natalist efforts by church and state. Urbanisation and perhaps improvements in public health is one powerful explanation of this trend, as urbanisation reduced the benefits of a child labour force, and increased the costs of child support, while improved public health increased the expected survival value of each child (this explanation was originally developed by Gary Becker; see for example his (Becker 1991)). In this explanation of the transition, the exact number of agents, their specific locations, their idiosyncratic characteristics and decisions make no difference to the overall outcome. This is supposed to be typical of population-based explanations in general, and evolutionary explanations in particular. There is supposedly a problem for giving evolutionary explanations of the historical trajectory of a population once that population becomes structured by institutions and similar social phenomena. First, institutions organised through role division multiply the number of causally distinct types of individuals within a group. Even in relatively small and simple pre-state societies, clan membership, moiety, initiate status, marriage and marriage-ability, and kin affiliations all condition the ways individuals interact, and hence contribute causally to the future state of the group. Second, hierarchical institutions amplify the importance of particular individuals within the community or polity. In his work discussing the prospects of models of cultural evolution, Tim Lewens makes this point about political power, reflecting on arguments of Richard Lewontin (Lewens 2015). The idea has recently been reinforced by Nathan Cofnas with the help of a vivid example (Cofnas 2018). He points to the 1492 decree of Isabella I and Ferdinand II of Spain, requiring Spain’s Jews to emigrate, convert, or risk execution if they attempted to practice their religion in secret. The dramatic fall in the proportion of individuals of that faith in Spain after 1492 could of course be modelled as a particularly extreme example of prestige-bias transmission, in which the individuals of generation N+1 take their cue from a particularly salient individual in generation N. But, the argument runs, this would be a historically uninformative description of the change in Spain, which depended instead on the specific beliefs and actions of a couple of individuals, backed by their control of institutionalized coercion. So perhaps the emergence of institutionally structured social life de-Darwinises social life. As hominin social life became more influenced by institutions, evolutionary and other individualist models became less applicable. A community structured by institutions is more like a mechanism than a population. Its dynamics depend on the organised interplay between its institutions, rather than aggregate outcomes of individual decision. At this point methodological individualism breaks down. As Tim Lewens notes, one response to this challenge is to divide the cake. Over the last few centuries, social life in Europe was profoundly influenced by many institutions, and yet the urbanisation explanation of the great demographic transition is credible. It is a good candidate explanation. So population-based explanations can be given of some important aspects of the social life even of institution-structured societies. Moreover, earlier in the human career, institutions, and especially hierarchical institutions, played much less of a role. Even after they emerged, as the demographic transition shows, individuals retained reasonably unfettered agency over important aspects of those lives, and population-based explanations are potential explanations of trajectories through those aspects of social existence. For this drop in birth rates took place in the face of natalist attempts by the church and state to prevent and reverse it. A more ambitious idea is to recognize the importance of structural features of human social life (even hierarchical institutions) and combine them with individualist, population-based explanations from economics and behavioural ecology. These structural features are cranes: they are central aspects of the environment, and as with other aspects of the environment, they help explain the options agents have, and the probable costs and benefits that attach to each option. They are aspects of the causal background, and as such, they are part of a collectively constructed social niche. Moreover, in general, as Elliot Sober has pointed out, in the formal models of evolution, cultural evolution, and economics, the determinants of the option pool, and the sources of costs and benefits are always in the background. In his Nature of Selection, he distinguished between “source laws” and “consequence laws”, pointing out that the formal apparatus of evolutionary theory (in those days, population genetic models) says nothing about the causal grounding of fitness difference, population size and structure, mating patterns, migration or heritability (Sober 1984). Rather, they show the predicted historical trajectories of a population (or family of populations) given certain assumptions about the size and structure of the starting population, the fitness differences between the variants within that population, and so forth. In many cases these predicted trajectories are far from obvious, and so these consequence models are informative. In one famous example, Fisher showed that in a large population, over long time spans, variants with even a very slight fitness advantage will come to dominate the population. In principle, consequence laws can be multi-level, modelling the interactions between various forms of group selection and selection on individuals. In practice, most of them are individualist. Source laws are tied to the ecology and reproductive biology of the organisms, and these ground the parameters assumed in the consequence laws. They explain why the variants differ in fitness, why the initial population has a specific size and structure, and so forth. These can include collective phenomena, like the multi-level social organisation discussed above. The historical trajectory is explained only by the combination of source and consequence laws. Consequence laws are typically informative. But there are no doubt degenerate cases where the trajectory is obvious: lethal mutations will disappear; so too will a population if the fitness of every variant is below replacement rate. Conversely, no population expands indefinitely. One might treat the 1492 Spanish case as a degenerate outlier. But even in this case, treating executive power backed by coercion as a special case of prestige-bias transmission is probably less trivializing than it looks. Coercive power is important but rarely absolute. Lewontin’s example is very similar to that of Cofnas: the 1555 settlement of religious conflict in Germany that gave each prince the power to determine whether his territory was catholic or protestant. However, as Lewens points out, in many cases princes were strongly incentivised to either choose the majority religion of their territory, and/or turn somewhat of a blind eye to failures to conform. In most realistic cases, power manifested by social institutions changes the cost/benefit profile agents face, without literally foreclosing every option but the one demanded by authority. Consequence modelling of these more realistic options can then be genuinely informative. What is the threshold of the perceived benefits of proscribed behaviours that allow them to survive or even spread, at some given level of authoritarian coercion? Conversely, at what punishment threshold do they disappear? To what extent is coercive authority impended by population structure and other barriers to information flow between elites and others? So one option is to aim at a more ambitious synthesis, simultaneously recognizing the causal salience of social institutions (especially those that concentrate power in a few hands) while still using evolutionary models to track the aggregated and iterated consequences of individual agent responses to these structures. Individualist explanations of human social life retain their importance, but are part of a larger explanatory package. Indeed, echoing our earlier claim about the conditions that allow individuals to invest in their futures, it is arguable that many explanations of human social change are only robust because they have this hybrid character. They model and aggregate individual decision making, but those aggregate outcomes are predictable only because background organisational features like institutions and customs restrict the range of options, and make the costs and benefits of each option stable, common, and predictable over time. In the presence of the institutions that organise their world, large numbers of agents will have relatively similar options, with similar pay-offs attached to each. Moreover, this tends to be common knowledge. That makes the social world more transparent. Relatively persistent collective facts—norms, customs, habits, institutions—play a stabilising role. They make it possible for agents to predict and hence to plan. The emergence of a recognizably human social life depends on these cranes, these organizational features that make it possible to predict the actions of others, even when they are strangers. The success and failure of one agent’s actions typically depends on what others do. Customs, norms, social roles and institutions help make the social world transparent enough, and stable enough over time, to make this decision making tractable. On this view, individuals make decisions that direct their lives, and social outcomes are collective consequences of these decisions. But those decisions are not made in contexts that individuals choose themselves, but in contexts which are the accumulated, filtered and stabilised outcomes of the summed choices of their predecessors. 5. Summary and Future Directions Human social life is pervasively shaped by collective features of groups that bear a complex and indirect relationship to the properties of individuals. Very likely, this has been true of humans since at least the Middle Pleistocene. While the lives of some other animals, including, notably, our closest great ape relatives, are similarly shaped by collective-level features, the evolutionary history of our species has seen a dramatic increase in both the number and complexity of such features. In the process, humans gradually came to live in what is more accurately described as a society than a mere group. In particular, as we evolved collective features such as social roles, norms, and institutions, human social life became more organized and hence more (locally) predictable at and over time. These collective features are multi-functional: they serve not only to regulate behavior, but also to help us in calculating the outcomes of our own and others’ behavior, and in settling conflicts when they arise. As these features established, humans experienced individual selection for greater self-control, as conforming to others’ expectations in general promoted fitness; most obviously, by helping one avoid costly sanctioning, but also by making one a more attractive cooperative partner. With the transition away from life in a great ape-like group to life in a bona fide society, a purely individual-focused framework for explaining human social life and its evolution has become inadequate. To fully understand and explain our evolutionary trajectory, an appreciation of how these complex collective-level features constrain, stabilize, and determine the costs and benefits of human social behavior is crucial. This leaves much work to do. As the last section served to bring out, the question of how to best utilise the individualist explanatory package (e.g., standard evolutionary and economic models) in the face of such collective-level complexity, especially hierarchical institutions, is an issue in need of further attention; here, we have simply sketched some reason for thinking that these tools may still have an important role to play. In addition, better and richer accounts of the origins, not just of the stability, of critical cranes such as religion and ritual are needed. This will involve, among other things, dovetailing such accounts with our growing knowledge of other crucial drivers of human social and cognitive evolution, for example, cooperative breeding, collaborative foraging, and the control of fire. Language, too, is obviously critical, for without it, the vast majority of these collective-level features could neither be constructed nor maintained. To the extent that there is a single, most pressing challenge at the intersection of social ontology and human evolution, it is perhaps to develop a deeper understanding of the evolution of multi-level human society. This topic has not received the same degree of attention as many of the other topics discussed above. Yet it is no less crucial. For, as discussed above, the evolution of more peaceful and open groups, permitting individuals from different groups to socially interact, and to even flexibly transfer from one group to another, was very likely foundational to our capacity to accumulate cultural information over generations, and to maintain that information in the face of threats posed by dramatic changes in climatic conditions. This of course includes the very cultural information upon which collective features of human groups such as norms, kinship systems, religion, and ritual depend. It is encouraging that the evolution of multi-level society lends itself to empirical testing in ways that the evolution of some of the other cranes we have discussed do not. In particular, as inter-group relations became more peaceful and open, we would expect people to move over larger geographical areas. In turn, we would expect to see greater variation in diet, with information about diet coming to us via chemical signatures in human fossils (Sterelny 2021b). In addition, it is plausible that anatomical markers of reduced reactive aggression and hence heightened social tolerance (e.g., craniofacial feminization See, for example, Cieri, Churchilll, et al. (2014); Hare (2017). ) can help us date this critical evolutionary transition. References Atran, S. (2005). In Gods We Trust: the evolutionary landscape of religion. Oxford, Oxford University Press. Becker, G. (1991). A treatise on the Family. Cambridge, Harvard University Press. Birch, J. (forthcoming). "Toolmaking and the Origin of Normative Cognition." Biology & Philosophy. Boehm, C. (2012). Moral origins: The evolution of virtue, altruism, and shame. New York, Soft Skull Press. Boyd, R. (2016). A different kind of animal: How culture made humans exceptionally adaptable and cooperative. Princeton, Princeton University Press. Boyd, R. (2018). A Different Kind of Animal. Princeton, Princeton University Press. Boysen, S. (1996). "More is Less": The Eliciation of Rule-Goverened Resource Distribution in Chimpanzees. Reaching Into Thought: The Minds of the Great Apes. A. Russon, K. Bard and S. Taylor Parker. Cambridge, Cambridge University Press: 177-189. Brosnan, S. F., Houser, D., Leimgruber, K., Xiao, E., Chen, T., & de Waal, F. B. (2010). "Competing demands of prosociality and equity in monkeys." Evolution and Human Behavior, 31(4), 279-288. Chapais, B. (2009). Primeval kinship: how pair-bonding gave birth to human society. . Cambridge, Harvard University Press. Cieri, R. L., Churchill, S. E., Franciscus, R. G., Tan, J., & Hare, B. (2014). Craniofacial feminization, social tolerance, and the origins of behavioral modernity. Current Anthropology, 55(4), 419-443. Cofnas, N. (2018). "Power in Cultural Evolution and the Spread of Prosocial Norms." Quarterly Review of Biology 93(4): 297-318. Colinvaux, P. (2017). Amazon Expeditions: My Quest for the Ice-age Equator. New Haven, Yale University Press. Dennett, D. C. (1995). Darwin’s Dangerous Idea. New York, Simon and Shuster. De Waal, F. B. (2008). "Putting the altruism back into altruism: the evolution of empathy." Annu. Rev. Psychol., 59, 279-300. Elster, J. (2008). Exploring Social Behaviour: More Nuts And Bolts For The Social Sciences. Cambridge, Cambridge University Press. Engelmann, J. M., Herrmann, E., & Tomasello, M. (2012). "Five-year olds, but not chimpanzees, attempt to manage their reputations." PLoS One, 7(10), e48433. Fletcher, G. E. (2008). "Attending to the outcome of others: Disadvantageous inequity aversion in male capuchin monkeys (Cebus apella)." American Journal of Primatology: Official Journal of the American Society of Primatologists, 70(9), 901-905. Hare, B. (2017). Survival of the friendliest: Homo sapiens evolved via selection for prosociality. Annual review of psychology, 68, 155-186. Henrich, J. (2016). The Secret of Our Success. Princeton, Princeton University Presss. Henrich, J. and C. Tennie (2017). Cultural evolution in chimpanzees and humans. Hrdy, S. B. (2011). Mothers and Others: The Evolutionary Origins of Mutual Understanding. Cambridge, Harvard University Press. Chimpanzees and human evolution. M. Muller, R. Wrangham and D. Pilbeam. Cambridge, Harvard University Press: 645-688. Henshilwood, C. and F. d'Errico, Eds. (2011). Homo Symbolicus. The Dawn of Language, Imagination and Spirituality. Amsterdam, John Benjamins. Kelly, R. (2013). The Lifeways of Hunter-Gatherers: The Foraging Spectrum Cambridge, Cambridge University Press. Kitcher, P. (2011). The Ethical Project. Cambridge, Harvard University Press. Kühl, H. S., Kalan, A. K., Arandjelovic, M., Aubert, F., D’Auvergne, L., Goedmakers, A., ... & Boesch, C. (2016). "Chimpanzee accumulative stone throwing." Scientific Reports, 6(1), 1-8. Landau, M. J., Solomon, S., Pyszczynski, T., & Greenberg, J. (2007)." On the compatibility of terror management theory and perspectives on human evolution." Evolutionary Psychology, 5(3), 147470490700500303. Layton, R. and S. O' Hara (2010). Human social evolution: a comparison of hunter-gatherer and chimpanzee social organization. Social brain, distributed mind. R. Dunbar, C. Gamble and J. Gowett. Oxford, Oxford University Press: 83–113. Layton, R., S. O'Hara and A. Bilsborough (2012). " Antiquity and Social Functions of Multilevel Social Organization Among Human Hunter-Gatherers." International Journal of Primatology 33: 1215–1245. Lewens, T. (2015). Cultural Evolution: Conceptual Challenges. Oxford, Oxford University Press. Lewis, J. (2015). "Where goods are free but knowledge costs: Hunter-gatherer ritual economics in Western Central Africa." Hunter-Gathere Research 1(1): 1-27. Lewis, J. (2016). Play, Music, and Taboo in the Reproduction of an Egalitarian Society. Social Learning and Innovation in Contemporary Hunter-Gatherers,. H. Terashima and B. Hewlett. Dordrecht, Springer: 147-158. Maynard Smith, J. and E. Szathmary (1995). The Major Transitions in Evolution. New York, Freeman. McBrearty, S. and A. Brooks (2000). "The Revolution That Wasn't: A New Interpretation of The Origin of Modern Human Behavior." Journal of Human Evolution 39(5): 453-563. McNiven, I., J. Crouch, T. Richards, N. Dolby, G. Jacobsen and G. M. T. O. A. Corporation (2012). "Dating Aboriginal stone-walled fishtraps at Lake Condah, southeast Australia." Journal of Anthropological Sciences 39(2): 268-286. Meggitt, M. (1965). Desert people: a study of the Walbiri Aborigines of Central Australia. Sydney, Angus and Robinson. Mithen, S. J. (2001). "The evolution of imagination: An archaeological perspective." SubStance, 30(1), 28-54. Murray, C. M., Stanton, M. A., Lonsdorf, E. V., Wroblewski, E. E., & Pusey, A. E. (2016). "Chimpanzee fathers bias their behaviour towards their offspring." Royal Society Open Science, 3(11), 160441.Norenzayan, A. (2013). Big Gods: How Religion Transformed Cooperation and Conflict. Princeton, Princeton University Press. Odling-Smee, J., K. Laland and M. Feldman (2003). Niche Construction: The Neglected Process in Evolution. Princeton, Princeton University Press. Planer, R. J. (2020). "Towards an evolutionary account of human kinship systems." Biological Theory, 1-14. Planer, R. J., Sterelny K. (2021). From Signal to Symbol: The Evolution of Language. Cambridge, MIT Press. Ofek, H. (2001). Second Nature: Economic Origins of Human Evolution. Cambridge, Cambridge University Press. Ross, D. (2006). "The Economic and Evolutionary Basis of Selves." Cognitive Systems Research 7: 246-258. Ross, D. (2008). "Economics, Cognitive Science and Social Cognition." Cognitive Systems Research 9: 125-135. Seabright, P. (2010). The company of strangers: A natural history of economic life. Princeton, Princeton University Press. Silk, J. B. (2005). The Evolution of Cooperation in Primate Groups. In Gintis, H., Bowles, S., Boyd, R. T., & Fehr, E. (Eds.), Moral sentiments and material interests: The foundations of cooperation in economic life (Vol. 6). Cambrdige, MIT press. Silk, J. (2020). "The Phylogenetic Roots of Human Kinship Systems." Biological Theory. Sober, E. (1984). The Nature of Selection: Evolutionary Theory in Philosophical Focus. Cambridge, MIT Press. Sosis, R. and C. Alcorta (2003). "Signaling, Solidarity, and the Sacred: The Evolution of Religious Behavior." EVOLUTIONARY ANTHROPOLOGY 12: 264-274. Stanford, C. (2018). The New Chimpanzee. Cambridge, Harvard University Press. Sterelny, K. (2003). Thought in a Hostile World. New York, Blackwell. Sterelny, K. (2012). The Evolved Apprentice. Cambridge, MIT Press. Sterelny, K. (2014). "A Paleolithic Reciprocation Crisis: Symbols, Signals, and Norms." Biological Theory 91(1): 65-77. Sterelny, K. (2018). Religion re-explained. Religion, Brain & Behavior, 8(4), 406-425. Sterelny, K. (2021a). The Pleistocene Social Contract. New York, Oxford University Press. Sterelny, K. (2021b). The origins of multi-level society. Topoi, 40(1), 207-220. Stiner, M. (2017). "Love and Death in the Stone Age: What Constitutes First Evidence of Mortuary Treatment of the Human Body?" Biological Theory 12(4): 248–261. Tomasello, M. (2014). Natural History of Human Thinking. Cambridge, Harvard University Press. Tomasello, M. (2016). A Natural History of Human Morality. Harvard, Harvard University Press. Williams, G. C. (1966). Adaptation and Natural Selection. Princeton, Princeton University Press. Wilson, D. S. (2002). Darwin's Cathedral: Evolution, Religion and the Nature of Society. Chicago, University of Chicago Press. Warner, W. L. (1958). A Black Civilization: A Social Study of an Australian Tribe. Revised ed., New York: Harper. Woodburn, J. (1982). "Egalitarian Societies." Man 17(3): 431-451. Wrangham, R. (1999). "Evolution of Coalitionary Killing." Yearbook of Physical Anthropology(42): 1-30.