Flexible signalling strategies by victims mediate post-conflict interactions in bonobos

Abstract

Compared to other animals, humans supposedly excel at voluntarily controlling and strategically displaying emotional signals. Yet, new data shows that nonhuman great apes' emotion expressions may also be subject to voluntary control. A key context to further explore this is during post-conflict (PC) periods, where signalling by distressed victims may influence bystander responses, including the offering of consolation. To address this, our study investigates the signalling behaviour of sanctuary-living bonobo victims following aggression and its relation to audience composition and PC interactions. Results show that the production of paedomorphic signals by victims (regardless of age) increased their chances of receiving consolation. In adults, the production of such signals additionally reduced the risk of renewed aggression from opponents. Signal production also increased with audience size, yet strategies differed by age: while immatures reduced signalling in proximity of close-social partners, adults did so especially after receiving consolation. These results suggest that bonobos can flexibly adjust their emotion signalling to influence the outcome of PC events, and that this tendency has a developmental trajectory. Overall, these findings highlight the potential role that flexible emotion communication played in the sociality of our last common ancestor with Pan. This article is part of the theme issue 'Cognition, communication and social bonds in primates'.

Keywords: Pan paniscus; consolation; emotion expressions; empathy; reconciliation; social interaction.

Figure 1.

Photographs depicting multimodal and multicomponent emotion expressions of bonobo victims following social conflicts, taken at Lola ya Bonobo Sanctuary. (a) Adult female victim presenting her rump with scream face expression and victim scream vocalization, being consoled by an adult female; (b) example of bared-teeth facial expression, (c) example of pout face expression, (d) example of victim scream and scream face expression; (d) example of victim with pout face being consoled by a juvenile bystander © Zanna Clay/Lola ya Bonobo Sanctuary.

Figure 2.

Uncertainty intervals from MCMC draws with all chains merged for models 1.1 (a, dependent variable: consolation), 1.2 (c, dependent variable: reconciliation) and 1.3 (e, dependent variable: renewed aggression). Points in (a,c,e) denote posterior means, inner dark grey bands correspond to the 80% CrIs, and the outer fine-lined bright grey bands correspond to the 95% CrIs. Below are plots showing a summary of the raw data on the relationship between the proportion of having received consolation (b), reconciliation (d) or renewed aggression (f) in relation to paedomorphic signal use. Points denote proportions of victim consolation, reconciliation or renewed aggression among all observations of the victim, depending on whether these victims produced paedomorphic signals. Size of the points indicates the number of observations per victim. Diamonds depict mean proportion and upper and lower whiskers denote standard error of the mean. Bold print in (a,c,e) indicates substantial effects.

Figure 3.

Uncertainty intervals from MCMC draws with all chains merged for model 1.4 (a, dependent variable: signalling persistence). Points denote posterior means, inner dark grey bands correspond to the 80% CrIs, and the outer fine-lined bright grey bands correspond to the 95% CrIs. (b) shows a summary of the raw data on the relationship between the proportion of signalling persistence in relation to consolation. The size of the points indicates the number of observed bouts per victim. Diamonds depict mean proportion and upper and lower whiskers denote standard error of the mean. Bold print in graph A indicates substantial effects.

Figure 4.

Uncertainty intervals from MCMC draws with all chains merged for model 1.5 (a, dependent variable: signal number) and 1.6 (d, dependent variable: signal display duration). Points in (a,d) denote posterior means, inner dark grey bands correspond to the 80% CrIs, and the outer fine-lined bright grey bands correspond to the 95% CrIs. (b, c, e and f) represent a summary of the raw data combined with model results on the relationships between signal number (b,c) and signal display duration (e,f) and audience members/friends within 10 m. Points denote single conflicts (i.e. one point represents signal number produced and presence of an audience in one distinct conflict). Shaded upper and lower ribbon edges depict 95% CrIs, and the mid-ribbon-line represents estimated posterior means. Bold print in (a and d) indicates substantial effects.