Key points are not available for this paper at this time.
In a corpus of group discussions recorded for a social science research project, participants frequently use representations of discourse. Many of these uses represent words that were not, might not be, or could not be spoken or thought. This article explores the ways these shifts are marked in relation to the represented context and the ways they function in the reporting context. The hypothetical status of these representations is often indicated, not by explicit markers, but by their sequential position. I propose three main functions: to model possible responses, propose counter arguments, or enact thought experiments. Hypothetical represented discourse enables participants to enact tensions in their own thinking and to deal with opposition between possible views. I argue that it is useful to see represented discourse in conversation not only in terms of narrative, but also in terms of rhetoric
Greg Myers (Fri,) studied this question.