Evolutionary researchers have frequently explored the adaptive significance of religious and ritual practices, particularly those that entail significant costs to the participant. Some authors, using the logic of sexual selection and signalling theory (Zahavi, 1975, Grafen, 1990), have examined whether costly initiation rites could be a form of mating competition in which participants demonstrate their ability to survive the rites and thereby signal their overall genetic quality (Low, 1979; Ludvico & Kurland, 1995; Singh & Bronstad, 1997). The logic of costly signalling has been extended further to explore whether such signals support the establishment of co-operative groups thereby securing the benefits of their collective action. Ritual participation may be acting as a costly signal that conveys genuine commitment to the community and its moral and social codes (Brusse, 2020; Bulbulia & Sosis, 2011; Shaver & Bulbulia, 2016; Sosis, 2003). Costliness may be a particularly important feature of ritual practice because it increases the reliability of a signal by making its display hard to fake, discouraging dishonest commitment displays (Irons, 2001; Sosis, 2003). For costly signalling to be effective, high quality signallers should not just be able to bear higher costs but have a greater benefit compared to cost than low quality signallers (Johnstone, Rands, & Evans, 2009; Lang, Chvaja, Purzycki, Václavík, & Staněk, 2022; Lang, Chvaja, & Purzycki, 2024). Such a differential is often present in ritual; individuals able to incur high costs are willing to do so because this enables them to enter the group and participate in its successful collective actions, thus achieving a greater net benefit overall than non-participants. Participation in costly rituals has been positively associated with various prosocial outcomes such as reputation for trustworthiness (Hall, Cohen, Meyer, Varley, & Brewer, 2015; Purzycki & Arakchaa, 2013; Ruffle & Sosis, 2007; Shaver et al., 2018; Tan & Vogel, 2008), social network size (Power, 2017), longevity of communes (Sosis & Bressler, 2003) and participation in warfare (Henrich, Bauer, Cassar, Chytilová, & Purzycki, 2019; Sosis, Kress, & Boster, 2007). Warfare represents a particularly extreme form of group commitment as individuals may owe their lives to the sacrifice of other group members and they in turn may be required to give their lives for others. In a comparative study of 60 societies using data from the Human Relations Area Files, Sosis et al. (2007) found that males in societies that engage in high rates of inter-group warfare endure the costliest rites, but that there was no association between these rites and other collective actions including food sharing. They suggest that reciprocity can be employed in this activity which prevents free-riders from accruing long term benefits. In warfare, however, men cannot solely rely on expectations of future cooperation since they may not be alive to reciprocate.

In this study we test for a range of possible adaptive benefits of costly rituals across a large group of societies in Pama-Nyungan Australia. These communities have a number of characteristics that make them a relevant choice for examining the adaptive significance of costly ritual practice. Firstly, ritual, and in particular adolescent initiation for males, plays a central role in Australian life and is frequently costly to the individual. Rites include sub-incision (a deep cut along the underside of the penis), scarification, tooth extraction and fire ordeals (Berndt & Berndt, 1996; Maddock, 1974; Meggitt, 1974). Secondly, there is abundant ethno-historic evidence of conflict between groups in Australian aboriginal societies (reviewed by Allen, 2014), suggesting it is a relevant collective action problem. Thirdly, both the costliness of initiation and the intensity of conflict vary between societies, allowing meaningful comparisons to be made. Fourthly, all these societies are hunter-gatherers. Sosis et al. found that costly rites tended to be more associated with hunter-gatherer societies and they suggest this is because larger societies are able to coerce participation in warfare through punishment (e.g. imprisonment), so are less dependent on ritual signs of commitment. The finding suggests a detailed analysis of costly male rituals across a large group of hunter-gathers might yield additional insight into the evolution of these rites. Australian communities did not adopt agricultural methods to any great extent, despite proximity to New Guinea horticulturalists in the north (Lourandos, 1997, Barker, 2009: 70). The material is an important dataset for evolutionary studies because it is one of the largest hunter-gatherer groups in the world for which we have both comprehensive ethnography and a well-attested language phylogeny (Bouckaert, Bowern, & Atkinson, 2018).

We examine whether costly ritual practices serve to signal group commitment by testing whether they are associated with two variables associated with group collective action: warfare and beyond-household food sharing, and two further variables: group size and kinship. We also examine whether male rites serve to signal mate quality by testing whether they are associated with more polygynous societies. To control for the possibility that trait similarities might be the result of common ancestry not adaptation, we test associations with a comparative phylogenetic approach using the Pama-Nyungan language phylogeny (Bouckaert et al., 2018) as the control for non-independence (Mace & Holden, 2005).

Whilst ethnographic accounts of inter-group conflict in Australia may relate to pressures brought about by colonial expansion, and may suffer from ethnocentric bias, they demonstrate some consistency in their descriptions of two main types of combat (Allen, 2014; Basedow, 1925; Warner, 1937; Wheeler, 1910). Firstly, regulated settling of disputes that normally involves the exchange of spears and sometimes individual hand to hand fighting. This only rarely led to serious injuries or fatalities (as the combat was structured in a way to avoid them). Secondly, active raiding of other groups that involved expeditions to other residential camps for the purposes of killing one or more of their party. Reasons for these raids included revenge for murders or unexplained deaths and disputes over women or territory. Such attacks nearly always resulted in fatal encounters and regularly escalated into repeated cycles of violence. Based on these descriptions, we consider that it is the presence of active raiding (not regulated settling of disputes) that equates closest to the presence of ‘warfare’ as commonly defined i.e. as armed combat between two political communities (Otterbein, 1994). We would therefore expect that the presence of costly rites would be expected to be associated with the presence of active raiding if the costly signalling hypothesis is supported.

The practice of sharing food within hunter gatherer groups is widespread and well documented, however there are few quantitative comparisons and none, to our knowledge, which relate specifically to Australia. In cross cultural studies, the presence of resource stress, including chronic food scarcity, low plant and animal richness and low and unpredictable precipitation, was predictive of more customary beyond-household sharing of resources (Ember, Skoggard, Ringen, & Farrer, 2018; Skoggard, Ember, Pitek, Jackson, & Carolus, 2020). In this study we have therefore used environmental ecology, specifically desert location, as a proxy for a high level of beyond-household food sharing. Australian deserts are one of the world's most hostile inhabited places (Gould, 1980: 61). Rainfall is both low and highly variable and they contain a much lower range of edible species than, for example, the African Kalahari desert (Gould, 1969). As well as the cross-cultural data, there is strong theoretical support to suggest that food sharing intensity is associated with high daily variances in food acquisition (Kaplan et al., 1985) and diet composition (Gurven, 2004). Ethnographies report the use of extensive food sharing in the Australian desert. Even small game is redistributed such that ‘everyone gets a share’… with people compelled to share food ‘in order to assure that, when an emergency arises… the relationships that require sharing between kin are strong’ (Gould, 1981: 432–5 quoted in Gurven, 2004). As well as food, sharing access to water such as creeks and waterholes is an important characteristic of Australian desert societies (Meggitt, 1974: 67), where rainfall is not only low but localised and highly variable (Gould, 1969). Furthermore, authors specifically examining ritual variation have suggested costly rites may be particular important in marginal habitats such as deserts, with the strong bonds forged through these rites enabling groups to be better adapted to compete for limited resources (Hayden, 1993; Whitehouse, 2004; Whitehouse, 2022).

If male rites signal group commitment we should also expect that larger groups, which are likely to face greater free-rider problems (Olson, 1965; Sosis et al., 2007), will experience selective pressures favouring the development of high cost rites that deter these free-riders. Independently, we should also expect societies composed of unrelated males to have higher cost rites than those that are patrilineal since the latter will have expectations of solidarity based on their kin ties, leading to less reliance on costly signals of commitment (Sosis et al., 2007). In Australia, societies with patrimoieties are composed of patrilineal clans each connected to a clan estate (Dousset, Koch, & McConvell, 2015). Related males are more likely to live together in these societies and therefore would be expected to have less costly rites.

If male rites signal mate quality we would expect the costliness of these rites to increase as competition for mates increases. Polygynous societies offer males greater potential reproductive success than monogamous societies and consequently there is greater competition for female parental investment. In stratified polygynous societies, this competition may be manifested in resource displays with wealthy males likely to out-compete poorer males (Irons, 1979). In non-stratified polygynous societies, such as Australian hunter-gatherers, ritual signalling is likely to be a more important mechanism to communicate mate quality. If the hypothesis is valid, we would expect the most extensively polygynous societies to have the costliest male rites.