Résumé
Two animals are said to have reconciled if, soon after an aggressive interaction between them, they engage in friendly contact (de Waal & van Roosmalen, 1979). Several nonexclusive hypotheses have been proposed to predict the distribution of reconciliation across opponent dyads (Arnold, Fraser, & Aureli, 2010), with proximate or ultimate explanations emphasizing its conciliatory nature (see Table 1). Friendly reunions following conflict may serve to repair mutually beneficial (i.e. fitness-enhancing) relationships (the ‘valuable relationship’ hypothesis: Kappeler & van Schaik, 1992; de Waal & Aureli, 1997), attenuate uncertainty or anxiety (the ‘uncertainty reduction’ hypothesis: Aureli, 1997; Aureli & van Schaik, 1991), or communicate peaceful intentions between former opponents (the ‘benign intent’ hypothesis: Silk, 1996). The ‘integrated’ hypothesis combines the uncertainty reduction and valuable relationship hypotheses by proposing that disruptions to valuable bonds are particularly stressful for opponents, and therefore warrant stronger resolution attempts (Aureli, 1997). Although they occasionally generate different predictions, these hypotheses are generally treated as complementary, with expressions of benign intent serving to reduce uncertainty about the conflict and thereby restore valuable relationships (Aureli, Cords, & Van Schaik, 2002; Cords & Aureli, 1996; Schaffner & Aureli, 2004). Tests of these hypotheses have commonly focused on primates, where the frequency and patterning of reconciliation vary according to the nature of the social systems and relationships under study (Arnold et al., 2010).