A critique and empirical assessment of Alexandra Horowitz and Julie Hecht’s “Examining dog–human play: the characteristics, affect, and vocalizations of a unique interspecific interaction”

Robert W. Mitchell

Horowitz and Hecht (Anim Cog 19:779-788, 2016) presented data about activities and vocalizations during brief videotaped dog-owner play provided by owners, examined these in relation to human affect during play, and made comparisons from their results to other research on activities and vocalizations during dog-human play. In this critique, I describe problems with Horowitz and Hecht's methodology, analyses, and evidence; in their interpretations of the data, evidence, and categorizations provided in other research, particularly my own studies of dog-human play; and in their claims of novelty for their findings. I argue that, to support their ideas about vocalizations and play types during dog-human play and their comparisons to other studies, their study requires fuller descriptions and reliability for their coding of vocalizations and play types, appropriate statistical analyses, and accurate descriptions of prior research. I also argue that their methodology provides results strikingly similar in many aspects to those of other researchers studying dog-human play, contrary to their claims of novel findings. Finally, I examine their suggestions about relationships between human affect and types of play activities and vocalizations using the videos of dog-human play I discussed in earlier publications, discovering minimal, if any, relationship.