This study investigates the way in which participants use echo questions to initiate repair of other participants' immediately preceding utterances in Japanese conversations. More specifically, it examines how participants understand and respond to these allo-repetitive (Tannen 1987) questions in terms of form, intonation, body movements, and sequence organization. Although echo questions have been studied extensively at the syntactic level (Adachi 1989, Inada 2005, Minami 1985, Nitta 1987) and the interactional level (Ozaki 1989, 1992, 1993; Schegloff 1984, 1997; Schegloff, et al. 1977; Weber 1993), none of these studies have fully discussed the intonations and body movements accompanying echo questions.
The data for this study come from videotapes of natural and spontaneous face-to-face interactions between native speakers of Japanese. I found that echo questions are not always pronounced with a rising intonation on the end of the repeated words. Based on the body movements of the repeater such as leaning toward the interlocutor or showing a blank face, I demonstrate that when the repeated words are not marked as questions syntactically or intonationally, the participants interpret the echo question as a request for repair, not as an aizuchi 'minimal response'. I also observed that even when participants are talking in formal speech, they sometimes do not use formal expressions such as desu ka? 'is (it)?' with the repeated words of the echo question. I attribute this lack of formality to the fact that the echo question is in fact reporting the speech of others.
This research sheds light on how allo-repetition is used by elucidating how participants in conversations use echo questions to initiate repair from a multiple of perspectives. It also addresses the issue of what constitutes a question. Finally, this research has pedagogical applications, because Japanese language learners frequently use echo questions to initiate repair in conversations.