Situations of Group-Robot Interaction: The Collaborative Practice of “Robot Speak”

Abstract

Social robots are designed to mimic embodied human interaction capabilities and are envisioned as social interaction partners either for individuals or groups of people. To interact with such robots, human sensemaking and active practical effort is required. In this microanalytic study, we examine video-recorded multi-party interactions with the social robot Pepper to illustrate some interactional and collaborative practices that humans engage in to achieve interactions with the robot. We examine situations where a group of university students engages in embodied practices trying to get responses from Pepper. We coined two concepts to describe these encounters: “robot speak” refers to embodied, spoken utterances directed at the robot with the aim of prompting a response. It is a specific, situated way of speaking shaped by assumptions about the robot’s interactional competence. “Framing talk” describes the participants’ collaborative commentary used to make sense of the situation, co-constructing the human-robot interaction as a meaningful social event. Additionally, in reference to our previous work, we illustrate a practical-ethical dimension related to social robots by examining how even a minor gaze shift of a robot can become immediately recognized as contextually significant for a participant, even when such are not explicitly designed as invitations to interact.