Dogs responding to human utterances in embodied ways

Abstract

This article introduces the notion of interspecies pragmatics and suggests that embodied CA studies can inform such pragmatics. The aim of the study was to investigate how dogs provide embodied responses to human utterances when a human is about to throw a ball for them to catch. Studying dogs' recipiency in interaction requires that multimodal approaches be taken into consideration. The data were collected with GoPro cameras attached to two dogs: each dog brought a first-person camera perspective to the analysis and, when they looked at each other, they also provided a third-person camera angle. The findings based on this multi-camera CA approach demonstrate that when dogs produce embodied responses to human utterances, they can distinguish Goffman's addressed recipient from the unaddressed recipient. Compared to human response production planning time (600–1200 ms), dog response planning time proved much faster (a maximum of 200 ms in this context). Another interesting methodological finding was that the distance between the speaker and recipient influenced the transcription of reaction times in turns-of-talk, as sound travels at only 0.3 m/ms. By contrast, embodied actions observed from video material contained no such transcription-related limitations due to the far higher speed at which light travels.