Dogs tuned to conspecific vocalizations: behavioral evidence for a voice processing preference

In many species, conspecific vocalizations are processed preferentially over other sounds. However, it is unknown whether this reflects sensitivity to voiceness (acoustic relevance), conspecificity (biological relevance), or socially significant species' vocalizations (social relevance). Dogs provide a unique model to disentangle biological, social, and acoustic relevance on mammalian voice perception, as both conspecific and human vocalizations are highly relevant for them. Previous behavioral research relied on highly salient distress calls, potentially masking finer perceptual differences. Here, we used a habituation-dishabituation paradigm to test dogs' sensitivity to different test sounds, including socially relevant vocalizations (dogs and humans), socially irrelevant vocalizations (chimpanzees), and non-vocal sounds (violin music), with pink noise as the habituating sound. Dogs' behavioral responses were analyzed using a Bayesian approach. Dogs showed dishabituation to all test sounds. Dogs' reactions to dog, human, and music sounds were above chance, whereas reactions to chimpanzee vocalizations were not. Additionally, dog vocalizations elicited stronger dishabituation responses than either human or chimpanzee vocalizations. In contrast, there was no difference in dishabituation responses to human vs. chimpanzee vocalizations. These results provide behavioral evidence in dogs of a processing advantage for conspecific over heterospecific vocalizations, but not for socially relevant over socially irrelevant heterospecific ones, or for vocal over non-vocal sounds in general. The findings reinforce the distinct role of conspecific vocalizations in perceptual sensitivity. Moving beyond highly salient distress calls allows the fine characterization of subtle relevance effects, highlighting the importance of biological over social and acoustic relevance in a domesticated species.