American Journal of Primatology 71:1–10 (2009) RESEARCH ARTICLE Post-Conflict Third-Party Affiliation in Chimpanzees: What’s in it for the Third Party? SONJA E. KOSKI 1– 3Ã AND ELISABETH H. M. STERCK 3,4 1 Leverhulme Centre for Human Evolutionary Studies, Department of Biological Anthropology, University of Cambridge, Cambridge, United Kingdom 2 Department of Biological and Environmental Sciences, Centre of Excellence in Evolutionary Research, UniversityofJyva¨skyla¨,Finland 3 Department of Behavioural Biology, Utrecht University, Utrecht, The Netherlands 4 Ethology Research, Animal Science Department, Biomedical Primate Research Centre, The Netherlands Affiliative behavior after conflicts between conflict participants and other group members is common in many primate species. The proposed functions for such triadic interactions are numerous, mostly concerning the benefit for the former conflict opponents. We investigated post-conflict third-party affiliation (TPA) in captive chimpanzees (Pan troglodytes) with the aim of assessing what the affiliating third parties may gain from affiliation. Specifically, we tested whether third-party-initiated affiliation protects the third parties from further aggression by conflict opponents. We found support for this ‘‘self-protection hypothesis,’’ in that third parties selectively directed affiliation to those opponents who more often gave further aggression to them, and affiliation effectively decreased their chance of receiving aggression from these opponents. However, a subset of affiliation, provided to conflict victims by their own kin, appeared to not be self-protective and the function of it remained open. We conclude that chimpanzee third-party-initiated affiliation is a more heterogeneous behavior than thus far assumed. Am. J. Primatol. 71:1–10, 2009. r 2009 Wiley-Liss, Inc. Key words: chimpanzee; Pan troglodytes; third-party affiliation; self-protection; empathy; conflict management INTRODUCTION Many social animals exhibit affiliative behavior after aggressive conflicts [Aureli & de Waal, 2000]. Post-conflict (PC) affiliation between the former opponents, or reconciliation, is widespread across primates and described in some non-primate mam- mals also [reviews by Arnold & Aureli, 2007; Aureli et al., 2002; Schino, 2000]. These contacts restore the relationship between the former opponents, reduce the anxiety caused by a conflict in their participants, and reduce the likelihood of further aggression between the opponents and between the opponents and other group members [Arnold & Aureli, 2007]. Affiliative PC contacts between conflict opponents and uninvolved bystanders are also common in many species. Many mutually non-exclusive functions for such third-party affiliation (TPA) have been pro- posed (Table I), often differentiated by kinship between the affiliating dyad, whether TPA involves the former aggressor or victim, and whether the initiator of affiliation is an opponent or a third party [Das, 2000; Watts et al., 2000]. Usually these functions concern benefits for the conflict partici- pants, although they probably provide benefits for the third party also and possibly for other group members. TPA between a conflict participant and its former opponent’s kin is described in many species [e.g. Call et al., 2002; Das et al., 1997; Judge, 1991], and hypothesized to be an alternative way to reconcile a conflict when contact between the opponents is too risky [Aureli & van Schaik, 1991; Judge, 1991; Wittig et al., 2007]. Support for this substitute, or triadic, reconciliation [Judge, 1991] in terms of relationship repair was recently found in wild chacma baboons, where TPA between victims and kin of the aggressor improved the relationship of the conflict dyad [Wittig et al., 2007]. Suggestive evidence for reconciliation via TPA was also found in wild chimpanzees, although TPA was not given by the former opponent’s kin, but by a close associate of Published online in Wiley InterScience (www.interscience. wiley.com). DOI 10.1002/ajp.20668 Received 2 October 2008; revised 8 January 2009; revision accepted 9 January 2009 Contract grant sponsor: Lucie Burgers Foundation for Com- parative Behaviour Research. Ã Correspondence to: Sonja E. Koski, Leverhulme Centre for Human Evolutionary Studies, Dept. of Biological Anthropology, University of Cambridge, Fitzwilliam Street, Cambridge CB2 1QH, UK. E-mail: sek39@cam.ac.uk r r 2009 Wiley-Liss, Inc.