Defining what constitutes a society is instrumental to understanding how social entities form, how they endure, and the mechanisms that allow them to thrive. With increasing research attention to collectives of all kinds, it is crucial to refine our language and develop specific constructs that can be shared across disciplines. Moffett introduces a welcome and thorough framework for the study of societies, calling for increased definitional clarity in building a taxonomy of social organizations. We thoroughly embrace many of Moffett's arguments, but raise two critical questions: (1) Can the spaces controlled by societies go beyond physical territory? (2) Can we really ignore cooperation as a foundational component of societies?

Moffett highlights the importance of spaces to societies, arguing that one distinguishing attribute of a society is that it regulates access to the physical space(s) it ultimately controls. Incorporating spaces as a definitional component of societies highlights the functional benefit that separates societies from other forms of social aggregations, but must these spaces necessarily be physical territory? We bring forward the idea that societies in humans and other species can also maintain control of social, reproductive, or even conceptual spaces. For example, although bonobo (Pan paniscus) groups are discussed as a society in the target article, there is little evidence that they maintain exclusive access or compete over territory (Furuichi, Reference Furuichi2011; Samuni & Surbeck, Reference Samuni and Surbeck2023). Bonobo groups do have distinct home ranges, but these overlap extensively with neighbouring groups and there is no evidence for defence or monopolizability of any patch of territory from neighbours. Instead, bonobo societies are better characterized through control over access to social spaces manifested through distinct group membership and ingroup/outgroup identities (Samuni, Langergraber, & Surbeck, Reference Samuni, Langergraber and Surbeck2022). Similarly, groups of western lowland gorillas (Gorilla gorilla gorilla) often range in the same physical spaces, and can have intergroup interactions ranging from hostile to tolerant and even affiliative (Cooksey et al., Reference Cooksey, Sanz, Ebombi, Massamba, Teberd, Magema and Morgan2020; Forcina et al., Reference Forcina, Vallet, Le Gouar, Bernardo-Madrid, Illera, Molina-Vacas and Vilà2019). Although Moffett touches on intersecting home ranges by introducing the notion of “mobile territory” in which a group “attempts to monopolize whatever site it occupies at a given time by defending that space and its resources when necessary,” this does not seem applicable to associations where members of different gorilla groups may meet, feed, and even play together. Specifically, bais, where “groups commingle while feeding on grasses rich in salts” (Forcina et al., Reference Forcina, Vallet, Le Gouar, Bernardo-Madrid, Illera, Molina-Vacas and Vilà2019) are difficult to understand as any group's mobile territory. In this species, we may consider whether control over access to reproductive spaces regardless of temporary spatial association and even affiliation between groups may be a more accurate description of their societal organization. Finally, in our own species, there are social entities such as academic societies and online communities that are disparately spread across the globe and have no consistent or permanent physical spaces, yet otherwise resemble, and even call themselves, societies. These types of human societies are instead structured around conceptual and digital spaces where, for example, members of academic societies may inhabit a theoretical niche and forage for research topics within their society's conceptual territory. We suggest that maintaining the centrality of societal space in Moffett's definition, while allowing for the relevant societal space to shift from physical to social, reproductive, or otherwise, provides a more holistic framework for the study of diverse societies.

Relatedly, maintaining control over a space of any kind is itself a group cooperation challenge. In arguing that “cooperation can be so varied and shifting…that it is judicious to define societies in a way that is neutral to its existence,” Moffett lumps together the diversity of cooperation and rejects their importance in one motion. Although we agree that networks of positive interactions and reciprocal pairwise cooperation are neither necessary nor sufficient to define societies, we draw particular attention to the plurality of cooperation and emphasize the distinction between bottom-up and top-down group cooperation (Brooks & Yamamoto, Reference Brooks and Yamamoto2022). Bottom-up group cooperation refers to the apparent group cooperation arising from several cooperating pairs, and is the target of much of Moffett's discussion around cooperation, but top-down group cooperation refers to forms of cooperation that are irreducible to the sum of dyadic interactions. Without a space there is no society, and we therefore highlight the role of collective cooperation and coordination in maintaining spaces that are indispensable to any society. Control over access to a space necessarily implies that societal outsiders could face conflict or consequence for unsanctioned use. However, any consequences must come from the society, which inherently requires risk or effort. Maintenance and defence of the space is therefore a costly endeavour, and each member must decide whether or not they themselves are willing to pay such a cost for the society or instead attempt to freeride on the efforts of other members. This is a top-down group cooperation challenge that must be solved for any societal space, and therefore society, to endure over time. Chimpanzees go on risky border patrols to protect their territory, bonobos sustain social group (but not territorial) boundaries despite incredible tolerance and bottom-up cooperation between members of different societies, gorillas fight violently against younger males attempting to invade their reproductive space, and human academics write cutting takedowns against outsiders breaching their conceptual domain.

The perspective argued here gives rise to a more plural and intersectional notion of societies than that defended by Moffett. It explicitly endorses the potential for multiple societal memberships by individuals exclusively controlling and identifying with multiple spaces and allows for the inclusion of displaced and dispersed societies as true societies. We believe that broadening the definition of societal spaces in this way offers a more parsimonious framing of our own and other species' societal cognition, including non-territorial species and those with multi-level organizations. Further, by explicitly focusing on the top-down group cooperation challenge of defending and coordinating these diverse spaces, the biological and evolutionary importance of societies comes into clearer focus. We believe these points can help form a more robust foundation for organizing and understanding the varied forms of social collectives.