We greatly value Dunbar’s emphasis on comparative research into the causes and consequences of variation in group size and structure, and also present several critical challenges to his approach and conclusions.

Dunbar’s analysis of group cohesion in primates highlights two core challenges of group-living: the desynchronisation of activity cycles and the stress associated with close social proximity, both of which are expected to intensify as group size grows. These centrifugal forces, he suggests, are mitigated by evolved structural, behavioural, and cognitive mechanisms. However, this analysis lacks attention to the socioecological variations in the costs and benefits of group-living, focusing only on what he implies are universal costs of increased group size and explicitly denying attention to potential social benefits. Dunbar portrays groups solely as networks built up from dyadic social relationships, placing excessive emphasis on cognitive processes while neglecting the crucial roles of group-level dynamics and external forces, which can shape sociality and cohesion independently of higher cognition.

Group processes reducible to the sum of dyads have been described as bottom-up, in contrast to top-down processes that inherently function at the group level (Brooks & Yamamoto, Reference Brooks and Yamamoto2022). Dunbar’s framework implicitly assumes that dyadic interactions simply scale up to account for group cohesion (i.e., bottom-up), but a more encompassing perspective requires acknowledging the essential role of top-down pressures in shaping group cohesion. While Dunbar states that “we are here only interested in the constraints on evolving large groups, and not the reasons why large groups evolve” (p. 22), these are not independent questions. Group-level solutions are shaped by the group-level problems they solve, which, in turn, define the capacities needed for their maintenance. They may incur some group-level costs, but these costs vary as a product of the species’ socioecology, as opposed to being shaped solely by forces of desynchronization and stress that are implied by Dunbar to affect all species relatively equally. Therefore, to fully understand what keeps groups together, we must move beyond a perspective focused solely on dyadic interactions and focus on the causes and consequences of species’ top-down group cooperation problems.

Dunbar’s emphasis on relationship-tracking through large brains overlooks alternative routes to group stability, bonding, and cooperation. We suggest that the sense of commitment may be a fundamental driver of social cohesion in species facing top-down group cooperation challenges. The sense of commitment has been conceptualised as a cognitive mechanism that arises when individuals rely on each other to achieve a shared goal, and as such, does not assume or depend on higher-order cognition (Michael, Reference Michael2022). This raises the possibility that a sense of commitment may be at play in species where shared goals – such as territory defence, predator avoidance, cooperative breeding, or information transfer – generate interdependence. Interdependent goals, that is, goals where individuals depend on one another, may therefore generate commitment dynamics sustaining group cohesion in the absence of high-order cognition. Commitment, in this sense, serves as an alternative mechanism for group cohesion, challenging the assumption that strong social bonds enabled by rich cognitive capacities are the only solution to counter fragmentary forces, as Dunbar suggests. Group processes such as intergroup competition and cooperative breeding can create evolutionary conditions that foster a sense of commitment, reinforcing social cohesion without the need for sophisticated cognition (Silk, Reference Silk and Nesse2002; Silk, Kaldor, & Boyd, Reference Silk, Kaldor and Boyd2000).

Relatedly, Dubar’s narrow focus on desynchronisation and stress-induced decreases in fertility is reductive and may misrepresent the broad range of collective costs that groups must overcome. As such, costs of group-living do not neatly track group size equally across species, but crucially depend on the species’ socioecology. For example, among the largest reported non-human primate groups are the Angolan black-and-white colobus monkey (Colobus angolensis) and mandrill (Mandrillus sphinx), both living in groups that can reach more than 300 individuals, much larger than those of the big-brained great apes (Fimbel et al., Reference Fimbel, Vedder, Dierenfeld and Mulindahabi2001; Hongo, Reference Hongo, Reyna-Hurtado, Chapman and Melletti2023). In such species, it may be a reduction to the costs of group-living, specifically in the absence of intense feeding competition among group members in these cases, that enables large groups to persist over time. When collective costs of sociality sufficiently decrease, large groups can emerge. This runs counter to the proposal of Dunbar that desynchronisation and reduced fertility are the main constraints on group size and can only be solved by large brains.

To guide future research on how the sense of commitment stabilises group cohesion and interacts with diverse collective costs of group-living in non-human primates, we propose to adapt a framework designed to spell out the minimal structure of situations which may give rise to a sense of commitment (Michael, Sebanz, & Knoblich, Reference Michael, Sebanz and Knoblich2016) through a comparative approach across species inhabiting varied ecological niches. If indicators of a sense of commitment in this minimal sense vary across species in accordance with top-down group cooperation challenges (such as cooperative breeding), rather than group size and number of grooming partners, it would suggest that these capacities can provide an alternate pathway for group cohesion, while comparison of species according to their feeding competition within the same clade would help distinguish systems that accommodate costly group-living through cohesive forces from systems where group-living is simply less costly. We do not challenge the fact that rich cognitive abilities are crucial to life in large groups for some species, but our emphasis is that this narrow focus on an association between brain size and group size overlooks other social mechanisms of cohesion, mechanisms that are tuned to the group-level costs and benefits they address.

While large brains can indeed facilitate the tracking of numerous dyadic relationships, which are undeniably vital for many complex social phenomena, they are not necessarily what brings and keeps groups together. Instead, factors like group-level interdependence and the entailed commitment dynamics play crucial roles. Future research should work towards integrating these perspectives to better understand how and why species have evolved, and continue to maintain their distinctive group structures.