Introduction

Across languages, we find minimal vocalizations, such as “mhm” in English and “un” in Japanese, that are used by participants in interaction to respond to co-participants' prior or ongoing talk. These “response tokens” have drawn a sustained interest in the study of language and social interaction over the past several decades (see Sorjonen 2001 for a useful review). One major line of research in this area has been concerned with distributional analysis, exploring statistical correlations between types of response tokens and turn-taking (e.g., Duncan 1972, 1974; Duncan and Fiske 1977) or correlations between the frequency of response tokens and such parameters as culture and gender (e.g., Maynard 1986, 1990; White 1989; Fishman 1983). Because their primary interest is in discerning patterns for the aggregate occurrences of response tokens across situations, these quantitative studies tend to isolate response tokens from their local contexts of use, glossing the interactional contingencies to which participants orient when producing response tokens in any given context. By contrast, a line of research on response tokens that emerged within conversation analysis (CA) pays more serious attention to the sequential/interactional contexts for the deployment of response tokens (e.g., Jefferson 1984, 1993; Schegloff 1982; Heritage 1984, 1998, 2002; Goodwin 1986; Mazeland 1990; Beach 1993; Sorjonen 1996, 2001, 2002; Gardner 2001; Mori 2006; among others). These studies investigate the kinds of interactional activities the participants engage in that give rise to the deployment of particular response tokens, how the co-participants orient to them, and what consequences their deployment has for the organization of subsequent talk.