Abstract
The syntactic and semantic splicing of utterances by two speakers to form a single prepositional unit known as joint productions has received attention from Sacks (1965–1971), Ochs, Schieffelin, and Platt (1979), Falk (1980), Schegloff (1984), and Lerner (1987, 1991). The phenomenon supports claims about the role of the audience in shaping talk and the growing tradition of work on cospeakers (Duranti & Brenneis, 1986; Tannen, 1987, 1989). Data from tape‐recorded individual psychotherapy sessions help answer questions of role and gender asymmetry, the social risk of rejection, and whether or not the practice is a bid to seize the turn. Four categories of joint productions are described. Motivations range from desire to insure verity, to help in word searches, and to covertly elicit information. The study finds that both males and females complete each other's utterances, and that, contrary to Wardhaugh (1985), joint productions are not interruptions or attempts to wrest the floor. The study demonstrates the use of computerized timing, and claims that joint productions are significant because they challenge our notion of how discourse is produced and interpreted, and what constitutes an appropriate unit. Joint productions indicate that the sentence itself may be perceived as a discourse unit under construction, an opportunity for interlocutors to contribute, clause by clause, to discourse.

This publication has 23 references indexed in Scilit: