There is a lot to commend in this piece. Moffett offers a broad, interdisciplinary perspective on human societies that spans not only the various human social sciences, but is coherent with research from evolutionary biology and the study of non-human animals. Such perspectives are sorely needed in the social sciences (see also Nettle, Reference Nettle2018). Moffett defines a society as a form of identity group. This strikes me as an acceptable definition, and potentially a useful one. However, I worry that, on its own, it is not a definition with any special explanatory or predictive power. Rather, it is probably most useful rhetorically when combined with the instrumental functions of identity groupings. I argue these take two key forms, which have very different scientific uses.

First, identity groupings are useful for individual decision making. They form the social categories with which individuals parse salience or trustworthiness of information or individuals, and their related propensities for cooperation, exploitation, or violence. As Moffett notes, the transition to anonymous societies in our evolutionary past allowed us to make decisions based on signals or markers of identity rather than on specific knowledge of individuals, which dramatically increased the potential scope of our cultural products. Identity information can be signaled overtly or covertly, allowing strangers with shared commitments, norms, and backgrounds to preferentially assort for coordination and cooperation (Barth, Reference Barth and Barth1969; McElreath, Boyd, & Richerson, Reference McElreath, Boyd and Richerson2003; Smaldino, Reference Smaldino2019; Smaldino, Flamson, & McElreath, Reference Smaldino, Flamson and McElreath2018). Identity categories shape our perceptions of the likely relation between ourselves and others, and thus are targets for strategic manipulation (Donath, Reference Donath, Kollock and Smith1999; Goffman, Reference Goffman1956; Smaldino & Turner, Reference Smaldino and Turner2022). As Vonnegut (Reference Vonnegut1962) noted in Mother Night, “We are what we pretend to be, so we must be careful about what we pretend to be.” Identity signals also allow individuals to target specific individuals for social learning, while ignoring or discounting information obtained from those belongs to other groups (Smaldino & Pérez Velilla, Reference Smaldino and Pérez Velillain press; Steiglechner, Smaldino, Moser, & Merico, Reference Steiglechner, Smaldino, Moser and Merico2023). The tagging of information sources as stemming from ingroup or outgroup sources may be adaptive under many conditions, but is also associated with the polarization and dehumanization that is becoming increasingly common in our world (Dias & Lelkes, Reference Dias and Lelkes2022).

All of these facets occur at the level of individual decision making. The aggregate of many people behaving thusly shapes the nature of our societies, while the cultural and structural groupings of that society shape the categories and decisions being made. As such, it is unclear to what degree a society is a causal force for explanation and to what extent it is the thing being explained. Moreover, it is unclear to what extent societies should play a special role for the social science of individual behavior, without a functional theory of how societies and individuals influence one another.

Second, societies can be seen as units for collective behavior and information processing. That is, the society as an emergent entity (or “crude superorganism” in the words of Richerson & Boyd, Reference Richerson and Boyd1999) may be usefully studied in terms of how it adapts to its own needs and competition with other societies (Galesic et al., Reference Galesic, Barkoczi, Berdahl, Biro, Carbone, Giannoccaro and Stein2023; Smaldino, Reference Smaldino2014). In this sense, understanding the nature of a society and its boundaries probably can help us to better understand the behavior of the collective. However, it also seems critical to delineate the characteristics of a society that would allow us to distinguish one society from another, and in this way predict their behavior as collective entities both independently and in interaction with other societies of similar or different character. Such a delineation is largely absent from the target article, and is surely beyond the scope of a short commentary. Nevertheless, it is hard to resist the temptation to offer suggestions. Characterization of societal traits may include coarse grain features such as population size and geographical distribution of constituents, structural features such as inequalities of wealth and decision-making power, institutional features such as higher education and electoral policies, economic features such as urbanization and sectoral diversity, and normative features such as cultural tightness and parochialism. A more thorough consideration of these traits is surely warranted.

Separating out these two ways of seeing societies is important. Moreover, the complexity of identity topologies makes delineating the boundaries of a society difficult in the modern world, highlighting another limitation of Moffett's approach. Individuals belong to multiple groups, which can be nested and are usually context-dependent (Roccas & Brewer, Reference Roccas and Brewer2002; Smaldino, Reference Smaldino2019). Moffett notes this, but seems to imply that these identity categories are simply subordinate to one's societal identity. This does not acknowledge the fuzzy nature of a society in a cosmopolitan world. For example, the nation of France is nested within the European Union, and while the French have a distinct national identity and language, they also often speak other languages with non-French inhabitants and visitors to their nation, and regularly cross borders into neighboring nations with relative impunity. Is France a society? Is the EU? Do the inhabitants of Tarascon-sur-Ariège, a small French town at the base of the Pyrenees, constitute a society? If I make an extended visit to these places and learn the language, do I become a member of their society despite retaining my US citizenship and home residence?

What sorts of societies are there, and how do their differences matter? I sympathize with the difficulty of answering this question, having faced similar challenges to my own target article (Smaldino, Reference Smaldino2014) in which I attempted to define the term “group.” Despite my best efforts, that piece suffered from an imprecision of scope, helpfully discussed by Gerkey and Cronk (Reference Gerkey and Cronk2014). I believe that Moffett is charting important and underexplored territory in his article, but more work – and particularly more attention to the problems for which his perspective adds value – is similarly needed to elevate society as a useful category in the way proposed.