A growing number of CA-based studies have documented the development of L2 speakers’ interactional competences in various settings such as L2 classrooms (e.g., Hellermann, 2008) and beyond (e.g., Hall, Hellermann, & Pekarek Doehler, 2011; Pekarek Doehler & Berger, 2016). However, given that interactional competences are co-constructed, CA-SLA researchers need to overcome the conundrum of how to identify development. With the aim of tackling this conundrum, the present CA study analyzes different sets of video-recorded conversations in which Steve, an intermediate L2 speaker of Japanese, participated with different interlocutors during a period of 16 months. One set of conversations, collected at his university in the U.S. before and after he studied abroad in Japan, are both first-encounter conversations in which he interacted with an L1 Japanese speaker who he had never met before. Another set of data, collected during his study abroad, consists of nine conversations in which he interacted with his L1-Japanese friends. They are casual conversations and tutoring sessions.

By focusing on Steve’s actions in topic shifts (Jefferson, 1984; Maynard, 1980; Svennevig, 1999), the analysis documents his interactional competence in contributing to topic development and topic changes. The comparison of the two first-encounter conversations shows that, while there were frequent disjunctive topic transitions (Holt & Drew, 2005) in his pre-study-abroad conversation, the post-study-abroad conversation had only occasional disjunctive ones. Steve contributed greatly to stepwise topic transitions (Holt & Drew, 2005) in the latter and his use of linguistic resources for navigating the topics had greater variations. Such higher competences are also found in conversational encounters recorded during his study abroad, and this finding suggests the transportability of interactional competence. The present study also identifies practices that are found only in a particular conversational encounter, namely tutoring sessions, and thus points to the co-constructed nature of interactional competence that are highly local. This paper contributes to the research on development of interactional competences by exemplifying identification of what is transportable across situations and time, and what may not.