Pet dogs' behavior when the owner and an unfamiliar person attend to a faux rival.

While dog owners ascribe different emotions to their pets, including jealousy, research on secondary emotions in nonhuman animals is very limited and, so far, only one study has investigated jealousy in dogs (Canis familiaris). This work explores jealousy in dogs one step further. We conducted two studies adapting a procedure devised to assess jealousy in human infants. In each study 36 adult dogs were exposed to a situation in which their owner and a stranger ignored them while directing positive attention towards three different objects: a book, a puppet and a fake dog (Study 1: furry; Study 2: plastic). Overall, the results of both studies do not provide evidence that the behavioral responses of our dogs were triggered by jealousy: we did not find a clear indication that the fake dogs were perceived as real social rivals, neither the furry nor the plastic one. Indeed, dogs exhibited a higher interest (i.e. look at, interact with) towards the fake dogs, but differences in the behavior towards the fake dog and the puppet only emerged in Study 2. In addition, many of the behaviors (protest, stress, attention seeking, aggression) that are considered distinctive features of jealousy were not expressed or were expressed to a limited extent, revealing that dogs did not actively try to regain their owner's attention or interfere with the interaction between the owner and the faux rival. Finally, a differentiated response towards the attachment figure (the owner) and the unfamiliar person (the stranger) did not emerge. Differently from what reported in human infants, dogs' behavior towards the attachment figure and the stranger interacting with the potential competitor (in this case, the fake dog) did not significantly differ: in both studies dogs paid attention to the owner and the stranger manipulating the fake dog to the same extent. In conclusion, we do not exclude that dogs could possess a rudimentary form of jealousy, but we suggest that research on this topic should require the use of a real social interloper (conspecific or human) and more naturalistic procedures.