Reassessing shelter dogs' use of human communicative cues in the standard object-choice task.

Unlike other animal species, domesticated pet dogs reliably use a range of human communicative cues to find a hidden reward in the object-choice task. One explanation for this finding is that dogs evolved skills for understanding human communicative behaviour during and as a result of human domestication. However, contrary to this domestication hypothesis, Udell et al. found domesticated shelter dogs failed to locate a hidden reward using a human's distal point cue, a cue pet dogs easily use. Hare et al., however, suggested the unorthodox methods used in Udell et al.'s object-choice task resulted in the shelter dogs failing to use human cues. In support of this, Hare et al. found that shelter dogs could use a human communicative pointing cue when tested with a standard object-choice task method. Yet in contrast to Udell et al., Hare at al. used a much simpler proximal cue that cannot exclude success based on stimulus enhancement rather than an understanding of the cue's communicative nature. We therefore addressed this issue by testing shelter dogs' abilities to use a range of proximal and distal human communicative cues in a standard object-choice task. We found shelter dogs could use proximal cues that may involve stimulus enhancement, but they continuously failed to use distal cues that excluded this possibility. Object-choice tasks with dogs typically involve non-vocalised human cues. We tested if vocalising would help shelter dogs to use distal cues. We found shelter dogs could use a vocalised distal continuous cue when the subject's name was called during cue presentation. It is therefore possible that vocalised cues help domesticated dogs learn about non-vocalised human communicative cues. Overall our results do not support that domesticated dogs' understanding of human communicative cues is a direct result of the domestication process.