Conversation Analysis as an Approach to Second Language Acquisition

Summary: This project uses the analytical tools of Conversation Analysis (CA) to examine the learning activities afforded by different activities in German as a foreign language (GFL) classrooms, as well as in out-of-classroom settings such as conversation partner sessions and proficiency testing interviews. A 2003 workshop provided hands-on training in CLAN software for CA, and a symposium in 2006 will bring together scholars working in various languages who use CA to understand the processes of second language acquisition.

Since the early 1990s, numerous theories of social practice have been imported to L2 studies from neighboring social science disciplines, including identity theory, community of practice, language socialization, and socio-historical/cultural/cognitive theory. Despite their different disciplinary origins, these all view L2 use and learning as contextualized, locally instantiated social practice rather than exclusively intrapsychological processes; consequently, they emphasize the interactive, collaborative, and socially situated nature of language use and learning. However, as Tarone has argued in a recent evaluation of the role of social context in SLA, most contemporary second language research under a social practice perspective suffers from two fundamental weaknesses: it focuses on SL use rather than acquisition, and it does not address the development or use of specific linguistic features, to which it may be added that in most studies - there are notable exceptions -, the inattention to the interlanguage itself extends to pragmatic, discourse, and sociolinguistic features. ResearchersÌ attention has shifted so strongly to the social context of SLA that the focal object of SLA has all but disappeared - or, as some will argue, has been fundamentally redefined. 

This has resulted in an undesirable split in the field, whereby SLAÌs traditional focal object has become the business of those favoring linguistic and cognitive-psychological approaches, whereas those pursuing sociolinguistic and sociocultural (neo-Vygotskyan) perspectives explore the social context of L2 learning but not the developing L2 knowledge - with notable exceptions, such as work by Hall, DuFon, or Ohta. 

Conversation Analysis (CA) shares with other social practice approaches its social-constructivist emphasis, yet of the various sociolinguistic and sociocultural approaches, CA is most able to address linguistic detail. This is because CA pays close attention to the details of interaction, especially the sequencing of action, the composition and construction of turns, and the temporal organization of interaction at the microlevel of verbal and nonverbal conduct. CA is equally concerned with the temporal, prosodic, and linguistic composition of interlocutorsÌ input and output. 

CA studies of interaction involving nonnative participants can provide several kinds of payoff. 

  • They can challenge assumptions prevalent in SLA about the defining characteristics of interaction involving NNS and about the relationship between interaction and L2 learning that is based on those assumed characteristics. For instance, SLA researchers working under the interaction hypothesis claim that there is more interactional modification in NNS than in NS discourse. However, microanalyses of NNS interaction outside language classrooms demonstrate a remarkable absence of the comprehension-securing strategies so frequently reported in the SLA literature. 
  • CA has the capacity to identify acquisitionally productive elements of interaction. Candidate interactional features include, on the one hand, the robust (universal) structural properties of any interaction, native or nonnative - such as the double contextuality of interaction, recipient design, turn-taking, and repair. For instance, Carroll (2000) found that linguistically low-proficient L2 learners were able to project accurately when the previous speaker finished their turn, as evident from their precision-timed turn start-ups. How precisely such properties of interactional competence are exploited as a resource for L2 learning largely remains to be explored. 
  • CA has the capacity to closely examine the opportunities for L2 learning in second and foreign language teaching. For example, as has frequently been noted, in the particular organization of classroom talk as a speech exchange system, notably its infamous Initiation-Response-Feedback/Follow-up (IRF) exchange structure, the teacher controls the interaction, allocates turns, takes most turns, takes longer turns, is in charge of topic management, and so forth. These institution-specific interactional characteristics of instructional discourse are commonly assessed as unproductive for language learning, from morphology to pragmatics. However, recent microanalyses of classroom interaction suggest that the IRF deserves another, more detailed look. Ohta demonstrates that material which recurrently appears in the teachersÌ follow-up turns eventually emerges in studentsÌ production. Translation is another classroom activity held in disregard by contemporary approaches to language teaching but alive and well in teaching practice around the globe. Through microanalysis of translation carried out in pairs, Ohta found that students orient to each othersÌ cues for assistance and through the task-focused interaction eventually learn a complex grammatical structure of Japanese. 

In sum, microanalysis has the potential to uncover details of classroom interaction that give occasion to reassess how acquisitionally productive or unproductive different activities are. It also emphasizes that students and teachers as actors in classroom activities shape the specific format of their interaction through their mutual orientation to each otherÌs conduct. However, the research literature on CA-for-SLA in L2 classrooms is yet quite limited and requires expansion to different target languages and interactional practices. What CA on classroom discourse cannot do, of course, is tell us how the interactional practices in L2 instruction relate to ordinary discourse and interaction in institutional contexts outside classroom settings, and the opportunities for L2 learning that such noninstructional interactions may afford. To date, conversation-analytic studies involving nonnative speakers in noninstructional discourse have examined L2 use not development (cf. references above). There is thus an urgent need to develop a research program on CA-for-SLA in noninstructional L2 discourse.  

The research project to be carried out by the NFLRC has two goals: 

  • To examine the learning opportunities afforded through different activities in German as a foreign language (GFL) classrooms. This will be accomplished through the videotaping and analysis of at least one regular GFL class in the Department of Languages and Literatures of Europe and the Americas, UHM, over a two-year period. The classes will be German 101, 102, 201, 202 in consecutive semesters. Recent enrollments suggest that a minimum of seven and a maximum of twenty-five students will participate in the entire course sequence. Videotaped interactions will be transcribed in standard CA format using CLAN editor for CA, to allow combination of sound and videofiles and transcripts. Given the wealth of conversation-analytic literature, it would be possible to determine in advance a set of interactional phenomena for analysis, such as repetition, repair, receipt tokens, dispreference markers, and so forth. However, following CA practice, we will resist the temptation to start the analysis by looking for predetermined material. While our initial data analysis will obviously follow established principles and practices of CA, as outlined above, and draw on substantive findings from the CA literature, we will adopt one of the (few) research practices that phenomenology, ethnomethodology and ethnography converge on, i.e., look and listen for recurrent interactional patterns and examine the `workÌ they do in their local contexts. Other than in standard CA, however, we will focus our analytical lens on those practices that evidence potential for, and hopefully demonstrably result in, L2 learning. Once a small set of interactional practices has been identified for further analysis, we will build collections (data sets of contextualized occurrences) for each practice from the classroom and noninstructional discourse data. 
  • To examine the learning opportunities afforded through participation in non-instructional discourse, conducted with native speakers of German or in German as a lingua franca in out-of-classroom settings, for instance during study abroad or in interactions with German speakers in Hawai‘i. Because of accessibility and logistics, the classroom part will be the primary project component. For data on non-instructional discourse in L2 German, we will partly rely on the cooperation of Prof. Johannes Wagner, University of Southern Denmark; partly we will aim for making our own recordings on upcoming visits to Germany. 

Two workshops on CA for SLA will be offered during the grant period. The first, early in the first year (2002-03), will be on preparing digitized audio and video files combined with transcripts, using CLAN CA, and will be offered to researchers in the field of foreign language learning with any language that uses a roman or romanized script. The second, a summer institute offered in summer 2006, will be a workshop drawing on the results of the project for teachers of German as a foreign language. In addition to research reports, a report on recommendations for teachers of German as a foreign language will be disseminated, including video demonstrations, possibly in CD-ROM/DVD format. 


  • Conduct a national workshop for researchers in the field of foreign language learning and teaching on the uses of CLAN CA with digitized video files - CLAN for Conversation Analysis workshop
  • Videotape and transcribe German as a foreign language class. Select non-instructional settings involving nonnative speakers of German for analysis. 
  • Identify acquisitionally relevant interactional practices in foreign language classrooms. Start establishing collections. Carry out exploratory analysis of non-instructional discourse. 


  • Continue to videotape and transcribe GFL class and non-instructional discourse. 
  • Augment & analyze collections of classroom practices. Identify acquisitionally relevant interactional practices in noninstructional discourse. Start collections. 


  • Continue adding to collections and conduct analysis.


  • Continue analysis of all tapes and transcripts. 
  • Offer summer institute for language teachers on the pedagogical applications of conversation analysis (summer 2006).
  • Prepare research report for dissemination. 
  • Prepare a report on recommendations for teachers of German as a foreign language, including video demonstrations, for national dissemination.
Tagged as 2002-2006