Social induction and the developmental trajectory of participation in intergroup conflict by vervet monkeys

Abstract

We assess the proposition that intergroup conflict (IGC) in non-human primates offers a useful comparison for studies of human IGC and its links to parochial altruism and prosociality. That is, for non-linguistic animals, social network integration and maternal influence promote juvenile engagement in IGC and can serve as the initial grounding for sociocultural processes that drive human cooperation. Using longitudinal data from three cohorts of non-adult vervet monkeys (Chlorocebus pygerythrus), we show that non-adults are sensitive to personal (age) and situational risk (participant numbers). The frequency and intensity of participation, although modulated by rank and temperament, both mirrors maternal participation and reflects non-adult centrality in the grooming network. The possibility of social induction is corroborated by the distribution of grooming during IGC, with non-adults being more likely to be groomed if they were female, higher-ranking and participants themselves. Mothers were more likely to groom younger offspring participants of either sex, whereas other adults targeted higher-ranking female participants. Although we caution against a facile alignment of these outcomes to human culturally mediated induction, there is merit in considering how the embodied act of participation and the resultant social give-and-take might serve as the basis for a unified comparative investigation of prosociality.

Keywords: aggression; collective action; cooperation; development; non-adults; social networks.

Graphical abstract

Figure 1.

Posterior density estimates of the probability of participation (Y/N) in intergroup conflict in relation to age, sex (ref: female), maternal participation, the number of individuals in the troop, rank, grooming eigenvector centrality (EC), spatial EC, neophilia, the number of participants from the focal and opposing groups, together with their interaction. The blue fill is truncated to indicate the 95% credible intervals.

Figure 2.

Posterior density estimates of changes in the level of aggressive intensity in relation to age, sex (ref: female), the number of individuals in the focal group, rank, grooming eigenvector centrality (EC), spatial eigenvector centrality, neophilia, the number of participants from the focal and opposing groups (and their interaction), and maternal aggressive intensity. The blue fill is truncated to indicate the 95% credible intervals.

Figure 3.

Posterior density estimates of the probability of grooming (Y/N) in relation to participation (Y/N), age, rank, neophilia, spatial eigenvector centrality, grooming eigenvector centrality, sex (ref: female), the number of individuals in the focal group, and the number of participants from the focal and opposing groups. The blue fill is truncated to indicate the 95% credible intervals.

Figure 4.

Posterior density estimates of the effects of sex, age and dominance rank on (a) the probability that non-adult participants would be groomed, and (b), if they were groomed, that it would be by their mothers. The blue fill is truncated to indicate the 95% credible intervals.