One of the key challenges for animals living in large, stable groups is overcoming the inherent pressures that pull individuals apart – what Dunbar has termed centrifugal forces (Dunbar, Reference Dunbar2025). Among these are the internal psycho-physiological costs of close proximity, such as chronic social stress that can lead to reproductive suppression in females, known as the “infertility trap.” Dunbar proposes that in primates, this trap is overcome by the formation of enduring female–female coalitions, which buffer individuals from social stress and enhance reproductive success. These coalitions are built through selective, high-quality social bonds – primarily maintained through grooming – that function not only to mitigate stress but also to enhance predictability and trust within the social network.

Dunbar mainly phrases selective social bonding as a solution to the infertility trap. We conjecture that social bonds may go much further in addressing the broader problem of group fragmentation, which Dunbar solves by advancing structural, behavioral, and cognitive strategies (Dunbar, Reference Dunbar2025). Strong, reliable bonds between individuals can actually serve as the scaffolding for structural cohesion within a group (Grueter et al., Reference Grueter, Qi, Zinner, Bergman, Li, Xiang, Zhu, Migliano, Miller, Krützen, Fischer, Rubenstein, Vidya, Li, Cantor and Swedell2020; Silk, Reference Silk2025). This ultimately increases resilience to socio-ecological disturbances and reduces the cognitive load associated with maintaining “weak ties.” In this sense, bondedness acts as both a behavioral and structural solution (sensu Dunbar) to the problem of maintaining group cohesion at scale. This might be of particular importance in species that rely on personalized relationships rather than anonymous group membership, like primates.

Besides close bonds, another key factor in supporting cohesion may be social tolerance. In primatology, social tolerance is increasingly viewed as a behavioral trait rather than an emergent phenomenon following from (unrelated) social patterning (DeTroy, Haun, & van Leeuwen, Reference DeTroy, Haun and van Leeuwen2022). Tolerance involves allowing or condoning others to benefit when something is at stake (DeTroy et al., Reference DeTroy, Haun and van Leeuwen2022), and thus facilitates behaviors such as cooperation and social learning (van Schaik, Reference van Schaik2003; Harrison, van Leeuwen, & Whiten, Reference Harrison, van Leeuwen and Whiten2021; van Leeuwen et al., Reference van Leeuwen, DeTroy, Kaufhold, Dubois, Schütte, Call and Haun2021). Levels of tolerance also consistently differ from one another within groups and individuals of the same species (Cronin et al., Reference Cronin, van Leeuwen, Vreeman and Haun2014; Fichtel, Schnoell, & Kappeler, Reference Fichtel, Schnoell and Kappeler2018; van Leeuwen et al., Reference van Leeuwen, Staes, Brooker, Kordon, Nolte, Clay, Eens and Stevens2023). This pattern points to tolerance being an active rather than passive vector in the equation of social bonding, both at the dyadic and group level. Unlike strong bonds, which are selective and intensive, social tolerance operates as a low-cost, extensive mechanism, allowing for smoother interactions and reducing the risk of fission. Consequently, it may enable groups to scale up in size by reducing the need for cognitive tracking of relationships across the entire group.

The question then arises: what underlies stable interspecific and intraspecific variation in social tolerance? One possibility is that social tolerance is itself grounded in cognitive capacities – particularly inhibitory control (a cognitive solution sensu Dunbar (Dunbar, Reference Dunbar2025)). The ability to inhibit impulsive or aggressive responses allows individuals to tolerate proximity with a wider range of group members. In this view, tolerance is not merely an emotional or temperamental trait, but a cognitive one. This would suggest that social tolerance, like bondedness, may ultimately be a cognitively mediated solution to group cohesion, albeit one that operates through generalized, low-intensity mechanisms rather than selective, emotionally rich ties. Indeed, this framework opens up a testable hypothesis: that species with greater inhibitory control should also exhibit higher baseline levels of social tolerance and greater group-level cohesion.

Following from the documented group-level variation in primates’ behavioral phenotypes (e.g., see (van Leeuwen, Cronin, & Haun, Reference van Leeuwen, Cronin and Haun2018; Kerjean, van de Waal, & Canteloup, Reference Kerjean, van de Waal and Canteloup2024; Koops et al., Reference Koops, Akankwasa, Camara, Fitzgerald, Keir, Mamy, Matsuzawa, Péter, Vicent, Zuberbühler and Hobaiter2024)), we posit that Dunbar’s analysis falls short in acknowledging variability. Selective pressures on group size vary across environments (Siepielski et al., Reference Siepielski, Gotanda, Morrissey, Diamond, DiBattista and Carlson2013), shaping social complexity in ways that static models may overlook. Factors such as predation, resource availability (Foster et al., Reference Foster, Franks, Morrell, Balcomb, Parsons, van Ginneken and Croft2012), and competition fluctuate across regions, meaning the benefits of group living are context-dependent. In areas with low predation and abundant resources, large groups offer little advantage, reducing the need for strong social cohesion. This challenges the idea that social group size alone drives cognitive evolution, as many cognitively advanced species – like orangutans or corvids – live in small groups or alone. A testable prediction from this view is that the link between sociality and cognitive demands is strongest where socio-ecological pressures make group coordination essential, such as in high-predation or inter-group competition environments (Radford, Majolo, & Aureli, Reference Radford, Majolo and Aureli2016; Samuni et al., Reference Samuni, Mielke, Preis, Crockford and Wittig2020).

Building on Dunbar’s proposed solutions to the fragmentation problem, we propose two additional mechanisms that are essential for maintaining group cohesion in primates: cultural transmission and dominance dynamics. In humans, culture is a fundamental mechanism for social cohesion through creating a sense of belonging and group identity (Tomasello, Reference Tomasello1999; Bell, Richerson, & McElreath, Reference Bell, Richerson and McElreath2009; Chudek & Henrich, Reference Chudek and Henrich2011). Cultural norms help regulate social behavior, reducing conflict, and promoting cooperation in societies of all sizes (Chudek & Henrich, Reference Chudek and Henrich2011). Cultural behaviors in animals have the potential to enhance social cohesion in a similar way (Cantor & Whitehead, Reference Cantor and Whitehead2013). When individuals within a group learn and adopt specific behaviors from one another – such as vocalizations, foraging strategies, or customs such as handclasp grooming or social games (McGrew et al., Reference McGrew, Marchant, Scott and Tutin2001; Rendell & Whitehead, Reference Rendell and Whitehead2001; Perry, Reference Perry2011, p. 2001) – these practices help define group membership (van de Waal, Borgeaud, & Whiten, Reference van de Waal, Borgeaud and Whiten2013; Berdahl et al., Reference Berdahl, Kao, Flack, Westley, Codling, Couzin, Dell and Biro2018; O’Corry-Crowe et al., Reference O’Corry-Crowe, Suydam, Quakenbush, Potgieter, Harwood, Litovka, Ferrer, Citta, Burkanov, Frost and Mahoney2018). Furthermore, by aligning behavior within a group, culture may reduce social uncertainty, minimize conflicts, and foster cooperation. All of these contribute to greater group stability. As such, culture may function as a short-cut bonding principle, leveraged by relatively low-cost mechanisms of social learning (Hoppitt & Laland, Reference Hoppitt and Laland2008). In this sense, cultural transmission offers a complementary route to Dunbar’s bonding mechanisms, allowing species to mitigate the fragmentation problem and sustain complex social structures.

Another way in which group cohesion can be secured is by dominance structures. Dominance dynamics can streamline groups by formalizing behavior and enabling dominant individuals to police others (Flack et al., Reference Flack, Girvan, de Waal and Krakauer2006; Beisner et al., Reference Beisner, Hannibal, Finn, Fushing and McCowan2016). However, Dunbar does not elaborate this aspect of living in social groups. In many species, hierarchical social structures provide a framework for maintaining order and minimizing conflict as group size increases. Despotic systems, in which a clear dominance hierarchy regulates access to resources and mating opportunities, may allow species to sustain larger groups without requiring extensive cognitive investment in relationship management (Kaburu & Newton-Fisher, Reference Kaburu and Newton-Fisher2015; Bhattacharjee et al., Reference Bhattacharjee, Cousin, Pflüger and Massen2023). In contrast, more egalitarian societies, where individuals must engage in frequent social negotiation to maintain group cohesion, may demand greater cognitive complexity (Aureli, Schaffner, & Schino, Reference Aureli, Schaffner and Schino2022; Rincon et al., Reference Rincon, Waller, Duboscq, Mielke, Pérez, Clark and Micheletta2023). This suggests that dominance style plays a key role in determining the maximum stable group size a species can maintain (Tibbetts, Pardo-Sanchez, & Weise, Reference Tibbetts, Pardo-Sanchez and Weise2022). A possible testable prediction is that species with strict dominance hierarchies, such as baboons and macaques, should be able to sustain larger groups with lower social instability compared to more egalitarian species like bonobos, where group cohesion relies more heavily on intricate social negotiations.

Taken together, building on Dunbar’s foundation (Dunbar, Reference Dunbar2025), we propose that social tolerance, cultural transmission, and dominance dynamics offer additional, complementary mechanisms that help overcome the challenges of group living.