In this paper I give an overview of utterance expansions, and demonstrate how the overall structure of invitations, casual conversation, and animation retellings relates to the goal of the conversation, e.g., whether the goal is to get the other to respond (accept, refuse, etc.) or to receive (understand, confirm). Building on research on invitations (Szatrowski 1991, 1993), casual conversation (Suzuki 1994, 1995; Kato 1999), and animation storytellings (Watanabe 1998, 1999, 2001; Szatrowski 2002), I demonstrate that invitations develop in relation to "what participants are trying to do," i.e., with respect to the goal of the inviter to invite and the invitee to respond to the invitation, casual conversation develops according to changes in topic, i.e., "what participants are talking about," and sections in animation storytellings coincide with scene changes in answer to the question "what happened."
I propose the unit of the wadan, a section which consists of the utterances of both participants, as a unifying concept for stages, topics, and episodes in Japanese conversations. Based on an analysis of the wadan used in 13 invitations, 4 casual conversations, and 4 animation retellings determined by survey, I demonstrate how the way that participants use linguistic and nonverbal resources to achieve wadan in these conversations can vary. While participants alternate in their roles as information-presenters and supporters by choosing from different sets of utterance functions to create wadan in invitation conversations, the roles of information-presenting participant and supporting participant tend to be set throughout animation retellings. The wadan in casual conversation relate more to changes in topic, although they may coincide with changes in the distribution of utterance functions when the supporting participant introduces a new topic.
Results of this research suggest ways to teach Japanese learners how to interactively construct wadan in Japanese conversations.