In a new articulation of the Social Brain Hypothesis, Dunbar argues that increasing investment in the connectivity and alignment of a species’ social relationships—made possible by dietary shifts— underpinned the evolution of “neuronally expensive” socio-cognitive skills that, in turn, favoured step-changes in primate sociality. For example, Dunbar calls attention to two species, Theropithecus gelada and Papio hamadryas, to argue, tacitly, for evolutionary convergence toward “super” group sizes on the basis of exceptional levels of social grooming and smaller-than-expected brain sizes. Such reasoning is rooted in the comparative method in biology, and it is undeniably compelling.

But there is a problem. The approach relies on summary values of a given trait, such as mean social group size, to identify commonalities between living species—the tips of phylogenetic branches—and infer processes that operated in the past. Dunbar uses it to impute selection on the primate lineages that gave rise to the variable expression of social brains today. Testing this hypothesis is difficult given the spotty nature of the fossil record, but there is a potential path forward: we could reorient our attention toward species-level plasticity, the “noise” that is more typically treated as an inconvenience.

The present version of the Social Brain Hypothesis is closely aligned with the evolutionary concept known as Behavioural Drive (Lister, Reference Lister2014; Fannin et al., Reference Fannin, Seyoum, Venkataraman, Yeakel, Janis, Cerling and Dominy2025). It holds that behavioural shifts must precede the evolution of corresponding morphological adaptations, such as expensive neuronal tissue. An essential precondition of behavioural drive is a high level of behavioural variability in the ancestral lineage, a “flexible stem” (Antón & Kuzawa, Reference Antón and Kuzawa2017). It follows that flexible stems—with exceptional levels of behavioural plasticity—predated the origin of every exemplary social brain, the final step-change in Dunbar’s model. Such a view inevitably raises questions concerning the causes of such plasticity and the amount that matters, topics best assessed with modern species. Mounting evidence of extraordinary variation within primate species—between individuals, groups, populations—could offer an unrecognised opportunity to detect the socio-ecological mechanisms underpinning Dunbar’s hypothesis.

Many primate species are highly social and long-lived. Describing their activity budgets and social relationships takes years of careful observation, often in challenging field conditions. As a result, a few well-studied populations have shaped our understanding of species-typical behaviours (e.g., Kappeler & Watts, Reference Kappeler and Watts2012). But this perspective has shifted in recent years; indeed, the behavioural plasticity of some populations has blurred even definitive species differences. For example, recent observations of wild chimpanzees (Giuliano et al., Reference Giuliano, Stewart and Piel2022; Koops et al., Reference Koops, Akankwasa, Camara, Fitzgerald, Keir, Mamy and Hobaiter2024) highlight the biases of our textbooks toward seminal studies of Tanzanian populations, in which groups of ∼60 individuals live in high-degree fission-fusion societies (Nishida, Reference Nishida1968; van Lawick-Goodall, Reference van Lawick-Goodall1968). Yet, chimpanzee group sizes can vary from 10 to 200 individuals, with individual encounters between some community members occurring daily, monthly, or only yearly (Lehmann & Boesch, Reference Lehmann and Boesch2004; Wilson et al., Reference Wilson, Boesch, Fruth, Furuichi, Gilby, Hashimoto and Wrangham2014).

Such plasticity is conducive to testing the putative causal mechanisms underpinning Dunbar’s hypothesis. If a given community is 2x larger than the species mean, then it predicts: i) greater investment in grooming time and longer grooming chains; ii) greater group stability in high-quality territories that minimise travel time between foraging patches; and iii) greater investment in cognitive solutions. These predictions invite systematic testing, but there is some tantalising support: the Ngogo super-community fractured along lines of increasingly entrenched, spatially-separated social “cliques” (Sandel & Watts, Reference Sandel and Watts2021), disrupting Dunbar’s suggested web of grooming chains, argued to foster adult unity. In contrast, the Waibira super-community appears stable—but employs a rotating core-periphery social structure in which independent males cycle through socio-spatial associations (Badihi et al., Reference Badihi, Bodden, Zuberbühler, Samuni and Hobaiter2022) and show greater investment in social signalling than smaller groups (Safryghin, Reference Safryghin2024).

Understudied species (or unstudied populations outside of well-studied habitats) offer further opportunities. We recently completed preliminary fieldwork in the Ennedi massif, a remote corner of north-eastern Chad with a relict population of patas monkeys (Erythrocebus patas) in an extraordinarily arid rocky Saharan habitat (Figure 1). Encounters during the day reflected “species typical” group sizes of 20-30 individuals (Hall, Reference Hall1966). In contrast, evening encounters at sparse water-holes could surpass 60-80 individuals—most likely the result of multiple groups converging on an essential resource during a limited daily access window. Some populations of other primate species, including baboons (Loftus et al., Reference Loftus, Harel, Ashbury, Núñez, Omondi, Muttinda and Crofoot2024) and bonobos (Samuni & Surbeck, Reference Samuni and Surbeck2023), form short-term (hours or days) “super-aggregations” across otherwise clearly separated group boundaries. Where populations show additional dynamic plasticity in social association, Dunbar’s hypothesis predicts that the costs of doing so should be offset by matching population-specific fluctuation in other behaviours. For example, where competition for limited resources is offset through the increased use of mechanisms for social bonding—such as grooming or vocal exchanges—social “rules” about whether these activities only occur within, rather than across, group boundaries, may fluctuate depending on local resource availability.

Figure 1.

Patas monkeys (Erythrocebus patas) negotiating steep cliffs in the Ennedi Natural and Cultural Reserve, Chad. Photo credit: John Hall.

Within-species behavioural plasticity not only provided the fuel for behavioural drive across primate evolution, but—where species retain plasticity—should continue to drive morphological differences. Where that plasticity relates to their capacity to form sophisticated social relationships, behavioural drive predicts a faster rate of change in the evolution of neurophysiologically expensive neocortical tissues, and increasing divergence/spread across the tips of the primate family tree. Crucially—while behavioural plasticity may be expressed as dynamic patterns of (super)grouping—concomitant impacts on expensive brain tissues would be stable. If populations, like the relict patas monkey population of the Ennedi plateau, habitually experience intermittent super-sociality as a behavioural response to an unusual local ecological niche—there should be observable differences in both their social behaviour and the underlying cognitive tissues that support it.