Glowacki's detailed account of small-scale societies’ endogenously emerging tendencies to oscillate between phases of peace and war highlights a need for understanding better the incentives governing “internal” policing for “external” peacekeeping. Here, I sketch some of these incentives and point out a resulting dilemma which Glowacki's account leaves unresolved for the time being.

We take issue with Glowacki's assumption that intergroup relations are characterized by positive-sum interactions and suggest to include negative-sum interactions, and between-group independence. As such, peace may be better defined as the absence of negative-sum interactions. Rather than being a consequence of cooperation, peace emerges as a necessary but not sufficient prerequisite for positive (in)direct reciprocity between groups that, in turn, is key to social identities and cultural complexity.

Peace is a hallmark of human societies. However, certain ant species engage in long-term intergroup resource sharing, which is remarkably similar to peace among human groups. We discuss how individual and group payoff distributions are affected by kinship, dispersal, and age structure; the challenges of diagnosing peace; and the benefits of comparing convergent complex behaviours in disparate taxa.

The emergence of social norms would have been dependent on the evolution of the cognitive capacity for mentalising to multiple orders of intentionality. Common knowledge is a related phenomenon that can solve coordination problems. That the same cognitive and social mechanisms should facilitate both peace and war is resonant with Girard's scapegoat hypothesis on the relationship between violence and religion.

Glowacki defines peace as harmonious relationships between groups maintained without the threat of violence, where groups can be anything from families to nation states. However, defining such contentious concepts like “peace” and “groups” is a difficult task, and we discuss the implications of Glowacki's definitions for understanding intergroup relationships and their evolutionary history.

Although Glowacki proposed that peace developed from the relatively recent advent of intergroup norms and tolerance for out-group members, we submit that (a) positive intergroup relations developed from a psychology grounded in the regulation of intragroup relations, (b) the “default” intergroup orientation is uncertainty, and (c) positive intergroup relations likely existed early in our evolutionary history.

Enhanced cooperation increases the capacity for humans to engage in large-scale warfare. This ability provides the foundation for male coalitionary behavior, leaving open the question of whether cooperation evolved in the same way, or for the same purpose, in females. Such coalitionary behavior entrains hierarchical forms of leadership that remain inherently unstable, providing a spark for conflict to emerge.

If group norms and decisions foster peace, then understanding how norms and decisions arise becomes important. Here, we suggest that neither norms nor other forms of group-based decision making (such as offering restitution) can be adequately understood without simultaneously considering (i) what individual psychologies are doing and (ii) the dynamics these psychologies produce when interacting with each other.

Glowacki recognizes the importance of norms in enabling war and peace, but does not focus on the cultural evolutionary mechanisms by which these norms are maintained. We highlight how group-structured cultural selection shapes the scale and nature of peaceful intergroup interactions. The mechanistic perspective reveals that there are many more cases of peaceful intergroup relations than the current account implies.

Glowacki discusses how humans regularly face collective action problems that may result in either peaceful or aggressive between-group interactions. Peace and war probably coevolved in humans. Using a gene–culture evolutionary framework is a powerful way to analyse why, when, and how humans have the capacity to build and maintain long-term peaceful interactions between groups and also to wage deadly wars.

While necessary parts of the puzzle, cultural technologies are insufficient to explain peace. They are a form of second-order cooperation – a cooperative interaction designed to incentivize first-order cooperation. We propose an explanation for peacemaking cultural technologies, and therefore peace, based on the reputational incentives for second-order cooperation.

Glowacki's account overlooks the role of religion in the regulation of cooperation, tolerance, and peace values. We interrogate three premises of Glowacki's argument and suggest that approaching religion as an adaptive system reveals how religious commitments and practices likely had a more substantial impact on the evolution of peace and conflict than currently presumed.

In this commentary, we set out the specifics of how Glowacki's game theoretical framework for the evolution of peace could be incorporated within broader cultural evolutionary approaches. We outline a formal proposal for prisoner's dilemma games investigating raid-based conflict. We also centre an ethnographic lens to understand the norms surrounding war and peace in intergroup interactions in small-scale communities.

The target article offers a game-theoretical analysis of primitive intergroup aggression (i.e., raiding) and discusses difficulties in achieving peace. We argue the analysis does not capture the actual strategy space, missing out “do-nothing.” Experimental evidence robustly shows people prefer doing nothing against out-group members over cooperating with/attacking them. Thus, the target article overestimates the likelihood of intergroup aggression.

I challenge the idea by Glowacki that “strong sanctions” such as fines, physical punishment, or execution are more effective in promoting peace than “weak punishments” like social rejection. Reviewing evidence that social rejection can have significant social and psychological costs for norm violators, I propose that social rejection can serve as a powerful reputational sanction in fostering peace in society.

While effective institutional practices are critical for the evolution of peace certain factors deter their effectiveness. In-group and out-group dynamics may make peace difficult between culturally distinct groups. Critical ecological conditions often lead to intractable conflict over resources. And within group conflicts of interest most prominently between generations may inhibit effective peace making.

We elaborate on Glowacki's claim that humans are more capable of establishing peace than other mammals. We present three aspects suggesting caution. First, the social capabilities of nonhuman primates should not be underestimated. Second, the effect of these capabilities on peace establishment is nonmonotonous. Third, defining peace by human-centered values introduces a fallacy.

Glowacki asserts that interdependent relationships beyond group boundaries are exceptionally rare among nonhuman mammals. However, rudimentary forms of interdependence can be seen in primate species that form multilevel societies, that is, core social units embedded within higher-level grouping categories. Studies of primate multilevel societies can enrich discussions about the evolutionary origins of peaceful between-group interactions in humans.

The capacities required for both peace and war predate 100,000 years ago in the genus Homo are deeply entangled in the modes by which humans physically and perceptually construct their worlds and communities, and may not be sufficiently captured by economic models.

Here we revise Glowacki's model by proposing a simple and empirically tested mechanism that is applicable to a comprehensive set of social interactions. This parsimonious mechanism accounts for the choice of both cooperative and peaceful alternatives and explains when each choice benefits the interacting parties. It is proposed that this mechanism is key to the evolution of both peace and conflict.