"Social system” is an umbrella term that encapsulates the social organisation (size and demographic composition of a social group), social structure (content, quality, and patterning of social relationships among group members), mating system and care system of and among the social units of a population or species. Primate social systems show considerable diversity.

Multilevel societies represent a particular primate social system that was first described for hamadryas baboons by Hans Kummer (1968); they consist of stable core units that form increasingly higher levels of grouping. Among primates, multilevel societies are best known from a few papionin (e.g., gelada, Guinea baboons, hamadryas baboons) and colobine (snub-nosed monkeys and Angolan colobus) species as well as humans. Recently, researchers have suggested that some of the gregarious apes could have multilevel social societies (see Grueter & Wilson 2021). Are such claims warranted?

The traditional view of ape social systems can be summarised as follows: orangutans are semisolitary; gibbons are pair-living; gorillas live in groups with one male and multiple females; and chimpanzees and bonobos live in communities with multiple males and females, but community members forage in compositionally fluctuating subgroups (fission-fusion dynamics). More recent studies showed that orangutans have differentiated social interactions, and in some populations gather in larger subgroups (Singleton & van Schaik 2002).

Recent studies have led to a reassessment of the canonical view of ape social systems (e.g., Pisor & Surbeck 2019, Furuichi 2020), particularly those intergroup encounters which are tolerant rather than aggressive. Instances of temporary intermingling, simultaneous exploitation of resources, and friendly relations between groups have been reported in some ape taxa. Some of these observations have been used as a basis to support the existence of multilevel societies in apes.

Western gorillas represent a curious yet unresolved case. They have been known to often interact non-aggressively, feeding on the same resource, engaging in social play with out-group members, and occasionally they spend the night nesting together in close proximity. The results of Forcina and colleagues (2019) support a multilevel gorilla society, with several groups forming larger entities. Morrison and colleagues (2019), using data on co-visitation of forest clearings by groups or solitary gorillas on the same day detected two hierarchically nested tiers of social organisation. However, western gorilla groups typically spend only a tiny fraction of their time in clearings, and whether "associations" seen in those clearings are durable and extend into the gorillas' forest habitat is not well understood. It therefore seems premature to declare western gorilla societies as multilevel. These findings, in conjunction with research showing that peaceful coexistence among neighbouring groups may be mediated by a dispersed network of related males (Bradley et al. 2004), hint at the possibility of a "community"-level organisation in at least some populations.

A recent study of mountain gorillas in the Virunga Volcanoes (Mirville et al. 2018) found that roughly one-fifth of intergroup encounters were peaceful. The familiarity of interacting groups (i.e., whether they had split from a single group in the past) was the main determinant of peacefulness. In a follow-up study, Morrison and colleagues (2020) showed that the familiarity effect applies only when interacting groups are within the periphery of their home ranges; encounters within core areas were consistently aggressive. In the Bwindi population, the most common behaviour shown by the study group in the context of between-group encounters was mild to moderate aggression, followed by tolerance (Robbins & Sawyer 2017). Pacific intergroup encounters occur only occasionally and do not constitute a major part of the gorillas' daily social context, conditions that are incompatible with a multilevel social system.

In bonobos, encounters of communities can last for several days. The transfer of females between communities may influence the affiliative relationships subsequently seen when two communities meet, so females may be important mediators of peace between communities. Lucchesi and colleagues (2020) as well as Pisor and Surbeck (2019) state that the tolerance seen in bonobo intergroup interactions could be the basis for the formation of multilevel societies. Given the non-permanence of these associations it may be assumed that bonobos do not exhibit multilevel societies in the strict sense.

Claims of chimpanzees featuring multilevel societies exist in the literature. However, multilevel societies require core units to exhibit a high degree of spatio-temporal stability in composition. In chimpanzees, mothers and young offspring do represent stable units, but parties are not stable. That is, the frequent separating and coming together of individuals, or fission-fusion nature of chimpanzee social dynamics, are irreconcilable with multilevel societies. At Ngogo, male subgroup members tended to remain in spatial proximity to each other and engage in joint territorial boundary patrols (Mitani et al. 2003). Females formed distinct association clusters termed "cliques" within which affiliative interactions occurred more than expected by chance. The modularity among males seems to have been a precursor towards a split of the community into two distinct 'daughter' communities. The modular social configuration among chimpanzee males has not led to frequent or permanent association among subunits but instead to a complete fission or split of the community.

Classifying a species as having a multilevel social structure or not also affects the predictions we develop regarding various evolutionary processes and phenomena such as disease and information transmission, cognitive ability, and sexual selection. For example, the busy and competitive environment that multilevel societies entail can be a strong selective force for the evolution of secondary sexual traits in both sexes. Multilevel societies can also reduce mating conflict in the form of infanticide, structure the flow of information/culture and possibly channel the transmission of pathogens and have been proposed to lower the cognitive load of its adherents. While the prerequisites for multilevel systems - sub-structuring and tolerant inter-group encounters - are indeed in place in some of the ape taxa, these are insufficient for a reclassification.

Grueter and Wilson (2021) stress that the term "multilevel society" should be used exclusively to refer to cohesive and compositionally consistent core units within a relatively stable larger society, as found in some papionins, colobines, and in humans. Expanding the term to include any sort of population-level substructuring makes the term vaguer and undermines its utility. They suggest that although the criteria for multilevel societies per se do not appear to be met in gregarious apes, the evidence accumulated thus far points to a possible higher-level organisation. Grueter and Wilson (2021) advocate the use of the term supra-group organisation to denote the existence of tolerant relationships between social groups that becomes manifest during intergroup encounters. Bonobos, mountain gorillas, western gorillas and gibbons may exhibit such a supra-group organisation. Tolerant intergroup relationships can be underpinned by a kinship network, as seen in western gorillas.

Summary of
Grueter, C. C. & Wilson, M. L (2021) Do we need to reclassify the social systems of gregarious apes? Evolutionary Anthropology 30 (5), 316-326

References
Bradley, B. J. et al. (2004): Dispersed male networks in western gorillas. Current Biology 14, 510-513
Forcina, G. et al. (2019): From groups to communities in western lowland gorillas. Proc R Soc B. 286, 20182019
Furuichi, T. (2020): Variation in intergroup relationships among species and among and within local populations of African apes. International Journal of Primatology 41, 203-223
Kummer, H. (1968): Social organization of Hamadryas Baboons: A field Study. The University of Chicago Press
Lucchesi, S. et al. (2020): Beyond the group: how food, mates, and group size influence intergroup encounters in wild bonobos. Behavioral Ecology and Sociobiology 31, 519-532
Mirville, M. O. et al. (2018): Low familiarity and similar 'group strength' between opponents increase the intensity of intergroup interactions in mountain gorillas (Gorilla beringei beringei). Behavioral Ecology and Sociobiology 72, 178
Mitani, J. C. & Amsler, S. J. (2003): Social and spatial aspects of male subgrouping in a community of wild chimpanzees. Behaviour 140, 869-884
Morrison, R. E. et al. (2019): Hierarchical social modularity in gorillas. Proceedings of the Royal Society B 286, 20190681
Morrison, R. E. et al. (2020): Inter-group relationships influence territorial defence in mountain gorillas. Journal of Animal Ecology 89, 2852-2862
Pisor, A. C. & Surbeck, M. (2019): The evolution of intergroup tolerance in nonhuman primates and humans. Evolutionary Anthropology 28, 210-223
Robbins, M. M. & Sawyer, S. C. (2007): Intergroup encounters in mountain gorillas of Bwindi impenetrable National Park, Uganda. Behaviour 144, 1497-1519
Singleton, I. & van Schaik, C. P. (2002):The social organisation of a population of Sumatran orang-utans. Folia Primatologica 73, 1-20