Introduction

One of Harold Garfinkel's legacies is his sociological experiments, which are commonly referred to as “breaching experiments.” Garfinkel created and developed these experiments to discover new phenomena that his contemporary social scientists had not found, adequately discussed or properly analyzed. The experiments allowed him to adapt the properties of praxeologically and conceptually unrelated scientific rationalities to the interactions in the lifeworld (e.g., Garfinkel 1952, 1963, 1967). That is, Garfinkel experimentally made trouble in situ by breaching people's normative expectations of various kinds, but paradoxically, he described, discovered and characterized the rational properties of mundane routine activities. We refer to this interest as a spirit of experimentation.

This chapter investigates the historical development of Garfinkel's ideas through experiments and examines the ways in which the participants deal with interactional, technological trouble in human–robot interactions (see, e.g., Ruiter and Albert 2017, for an overview of conversation analysis and experimental settings). In the following, we discuss Garfinkel's orientation to people's methodical accomplishments in practical activities that since his doctoral research he explored through experimental methods. We show that the sociological experimental attitude has influenced and shaped the development of ethnomethodological and conversation analytic studies, including current research on human–robot interaction. By examining experimentally conducted human–robot interaction, we argue that the analysis of human–robot interaction can also reveal aspects of routine grounds of everyday activities.

Our chapter is structured as follows. First, we consider Garfinkel's breaching experiments through the lens of his overall research trajectory from his doctoral dissertation (Garfinkel 1952) through to his Studies in Ethnomethodology (Garfinkel 1967). Second, we clarify two important foundations of background expectancies and everyday life that Garfinkel's breaching experiment revealed: (1) the temporal organization of future action, which includes projection and the anticipation of forthcoming actions, and (2) knowledge based on accountability and membership categories. Third, we describe how we applied Garfinkel's spirit of experimentation to develop and test a service robot that can interact with multiple people simultaneously (i.e., a multiplex care robot system). Fourth, we demonstrate how the two key elements of the breaching experiments work in human–robot interactions in a spice shop where a parent and a child interact with and through a robot acting as a shopkeeper. In conclusion, we provide a possible description of the reason for why Garfinkel emphasized the spirit of experimentation and discuss how it is connected to human–robot interaction research.