This study investigates the role of social interaction in language gain among study abroad students in France. Using the ACTFL Oral Proficiency Interview (OPI), the Can‐Do self‐assessment scale (Clark, 1981), a revised version of the Language Contact Profile (LCP; Freed, Dewey, Segalowitz, & Halter, 2001), and preand postdeparture questionnaires, we examine gains in oral proficiency as related to language contact in the study abroad environment. This research shows that language gain is possible during a semester‐long study abroad program. It does not uphold the common belief that living situation and contact with authentic media differentiate students who improve from those who do not. Looking at the background of students (age, gender, grade point average, etc.), it reveals that only prior coursework in French correlates strongly with gains in proficiency once abroad. In its most surprising finding, this study suggests that speaking French with Americans may impede proficiency development.

Although study abroad is viewed as an ideal environment for interaction in the target language, research in this area has relied mostly upon self‐reported data, which pose challenges regarding recall bias and participant commitment. This article shows how Facebook data can be used to analyze naturally occurring learner interactions during study abroad. Drawing from a 10‐month, Web‐based ethnography of three learners, findings indicate an overall increase in Portuguese use during study abroad, as well as acquisition of computer‐mediated communication–relevant terms in Portuguese. Comparisons to Language Contact Profile data show the advantages of using social media data to gain a detailed view of learner interactions while abroad. Implications for language instructors and the use of social media in the classroom setting are also discussed.

Using data from a multisited ethnography of Ecuadorian transnational musicians, I applied Lave and Wenger's (1991) concept of legitimate peripheral participation and Jacoby and Ochs's (1995) notion of co-construction to examine two musicians' attempts to learn Quichua, an Ecuadorian indigenous language. Through an analysis aided by constructivist grounded theory (Charmaz, 2006), I found that co-constructed notions of language beliefs, ethnic identity, and community perspectives led to perceptions of these individuals as legitimate or nonlegitimate learners of Quichua. These co-constructions played a role in a successful acquisition of Quichua for one individual, for whom Quichua was a heritage language, and a failure to succeed by the other. The results of this article enrich discussions on maintenance of threatened languages and add to the ongoing debate in second language acquisition on the importance of cognitive/individual versus social/external factors in language learning.