UC Berkeley

Dating back to some of the earliest investigations of the use of text-based, online chat in foreign language instruction, researchers and instructors have been hypothesizing that and asking if there is some transfer between chatting and oral language development (e.g., Beauvois, 1992; Chun, 1994). The possibility of this sort of transfer is especially promising for the many students whose ability to speak their foreign language lags behind their ability to read and write. In these cases, the written nature of text-based chat might enable students to take advantage of their literacy skills, while the real-time interaction involved in chatting might support the acquisition of fluency and conversational genres associated with oral communication.

Research in this area has relied almost exclusively on experimental and quasi-experimental studies in which oral language development of students who engage in chat has been compared with the development of students who receive only classroom instruction (e.g., Abrams, 2003a; Beauvois, 1997b; Payne & Whitney, 2002). All published findings have been quite promising in that in all cases students who received chat-based instruction achieved gains in oral development that were equal to or greater than those of students in control conditions. Even so, while this body of research strongly supports the premise that transfer between chatting and speaking does occur, these studies do not adequately describe the phenomenon, much less explain it. As such, existing work in this area cannot provide any concrete suggestions for how to integrate text-based chat into foreign language instruction in order to target specific learning goals. At most, existing research can only vaguely suggest that in general chatting may be beneficial for speaking.

My dissertation attempts to provide a much greater level of specificity about the phenomenon of transfer between chatting and speaking than what currently exists in the research literature. Rather than adopting an experimental or comparative approach that would likely be limited to finding that transfer does or does not occur, I have used a qualitative, multi-case approach that has enabled me to construct detailed descriptions of the phenomenon, and to connect specific cases of this transfer to the instructional contexts in which they occurred. I refer to my investigation as "ecological" in part because I adopt many of the analytical constructs associated with ecological perspectives on second language acquisition (e.g., Kramsch, 2002b), and also because of my explicit consideration of and attempts to make connections between multiple scales, ranging from individual students transferring specific linguistic items between online and offline activities, to more general patterns that emerge throughout and across groups of students over multiple semesters.

Also unique to my investigation of transfer between chatting and speaking is the fact that my approach is multidisciplinary. Specifically, I draw heavily from three areas of research: second language acquisition (SLA), computer-mediated communication (CMC), and transfer of learning. My integration of research on transfer is especially valuable because it provides both methodological suggestions for investigating this phenomenon, as well as frameworks for analyzing more specifically exactly what transfers and under what circumstances (e.g., Barnett & Ceci, 2002).

I do not pretend to have captured the full range of possible transfer-related outcomes, but through my five rounds of data collection in university level classes of Spanish as a foreign language in which chat-based activities were either integrated into required classroom instruction, or offered as optional tutoring, two clear, but quite different patterns have emerged. On one hand, I have found that foreign language chat can be quite social, informal, and even playful. Under these circumstances, the benefits for oral language development may be primarily social and affective. Specifically, informal chat can provide students with opportunities to get to know one another better and to become more comfortable with each other. When this happens, students can become more willing to engage in oral communication with one another, and this increased engagement can support oral language development. On the other hand, chat-based instruction can also be highly structured in ways that provide students with opportunities to practice specific linguistic forms and communicative functions, including forms and functions associated with academic discourse. Under these circumstances, practice in chat can support subsequent oral use of these specific forms and functions.

My dissertation makes multiple contributions to the different fields from which it draws. For applied linguistics, the two different patterns described above support two pedagogical suggestions: (1) Unstructured or explicitly social chatting can facilitate subsequent oral communication by enabling students to become more comfortable with one another; (2) Highly structured chat-based instruction can target specific linguistic forms and communicative functions, providing practice for subsequent oral use of these forms and functions. For CMC, my range of outcomes points to the flexibility of chat as a medium that can support a wide range of interactions and genres, including those normally associated with academic discourse. Additionally, my data highlight the increasing fluidity with which students maintain relationships and identity performances across media (e.g., Baym, 2010; Leander, 2008). Finally, for research on transfer of learning, a sub-field of the learning sciences, my dissertation contributes to an emerging ecological perspective on transfer (e.g., Greeno, Smith, & Moore, 1993). This perspective views transfer as a relational phenomenon in which learners are active participants in creating and defining the contexts in which learning and transfer occur (Lave, 1988; Pea, 1987).