ORIGINAL PAPER Reputation-like inference in domestic dogs (Canis familiaris) Shannon M. A. Kundey • Andres De Los Reyes • Erica Royer • Sabrina Molina • Brittany Monnier • Rebecca German • Ariel Coshun Received: 6 August 2010 / Revised: 17 November 2010 / Accepted: 19 November 2010 / Published online: 8 December 2010 Ó Springer-Verlag 2010 Abstract Humans frequently interact with strangers absent prior direct experience with their behavior. Some conjecture that this may have favored evolution of a cog- nitive system within the hominoid clade or perhaps the primate order to assign reputations based on third-party exchanges. However, non-primate species’ acquisition of skills from experienced individuals, attention to communi- cative cues, and propensity to infer social rules suggests reputation inference may be more widespread. We utilized dogs’ sensitivity to humans’ social and communicative cues to explore whether dogs evidenced reputation-like inference for strangers through third-party interactions. Results indi- cated dogs spontaneously show reputation-like inference for strangers from indirect exchanges. Further manipula- tions revealed that dogs continued to evidence this ability despite reduction of specific components of the observed interactions, including reduction of visual social cues (i.e., face-to-face contact between the participants in the inter- action) and the nature of the recipient (i.e., living, animate agent versus living, inanimate self-propelled agent). Dogs also continued to demonstrate reputation-like inference when local enhancement was controlled and in a begging paradigm. However, dogs did not evidence reputation-like inference when the observed interaction was inadvertent. Keywords Dog Á Canine Á Cognition Á Vicarious reinforcement Introduction Humans regularly interact with strangers, often determin- ing how to act toward them with little direct experience with the individual’s behavior (e.g., Seabright 2005). This feature may have favored evolution of a cognitive system that assigns reputations (e.g., Subiaul et al. 2008). Such judgments involve ascribing enduring behavioral disposi- tions by conjecturing that behaviors are predictable and consistent over time and contexts. Assessment might occur directly through first-party interactions or indirectly through social learning by viewing others’ exchanges (McGregor and Dabelsteen 1996; Parejo and Aviles 2007; Subiaul et al. 2008). While direct reputation judgments likely provide more accurate information (Axelrod 1984; Alexander 1987; Hauser et al. 2003), indirect reputation judgments allow behavior prediction when direct interac- tions might be costly (e.g., assessing strength or domi- nance; Axelrod 1984, 1987). Recently, some have posited that non-human species might also make indirect reputation judgments based on their performance in a variety of social learning studies. For example, chimpanzees respond differently to humans behaving as unwilling versus unable helpers and to humans that intentionally rather than accidentally fail to give food, as well as recruit the best collaborators (Call and Tomasello 1998; Povinelli et al. 1998; Call et al. 2004; Melis et al. 2006). Such studies suggest chimpanzees infer a variety of social information through observation. How- ever, these studies often involved subjects receiving direct experience with humans or conspecific individuals, as S. M. A. Kundey (&) Á E. Royer Á S. Molina Á B. Monnier Á R. German Á A. Coshun Department of Psychology, Hood College, 401 Rosemont Avenue, Room ROS 27, Frederick, MD 21701, USA e-mail: kundey@hood.edu A. De Los Reyes Department of Psychology, University of Maryland at College Park, College Park, MD 20742, USA 123 Anim Cogn (2011) 14:291–302 DOI 10.1007/s10071-010-0362-5