Anthony Bushell
Today German’s own position as a second language represents something of an enigma. As the largest l...
Download (PDF - 617 kb)   Motivation and Multiple Languages in Learners of German ...
Download (PDF - 2751 kb)   Investigating Language Learner Beliefs Using the Lego...
    My guiding rules when I think about creatin...
Gender and Media: Course, Website, and Research Report Katrina Sark, University of Victor...


Language Choice in Small Group Activities

Teyana Reichert

Second language acquisition (SLA) research has been involved in an ongoing discussion and controversy about the role of code-switching, i.e., the use of two or more languages within the same conversation in educational settings (Gafarange, 2007). Until relatively recently code-switching to first language (L1) has often been associated with insufficient knowledge in the second language (L2), e.g., with limited vocabulary. Most recent research in SLA offers another perspective on code-switching practices in classrooms, however. Studies on language alternation informed by conversation analysis (CA) describe code-switching as an important interactional device learners employ to manage the task, interpret the situation, organize and structure the discourse (e.g., Unamuno, 2008, Halmari, 2004, Chaves, 2003). Nonetheless, most published research on language alternation does not provide sufficient accounts about its functions in learner interactions in situations when the teacher is not present.

This paper contributes to the current discussion about CA-informed research in code-switching practices in an educational context. It analyses interactions of students who were given a speaking task and engaged in collaborative learning in small groups, which took place outside of the classroom but in a learning-related context. Using qualitative interactional analysis and a sociolinguistic perspective (Gumperz, 1982), I examine interactions between adult learners at a major Canadian university as they engaged in a task to prepare a role-play to be presented in front of a class. Applying the methodological technique of CA, I analyse the use of more than one language in the beginners’ repertoire of conversations, i.e., code-switching in learner-learner interactions. In particular, my focus is on ways learners switch between languages as meaningful strategies and mechanisms while working on the task. The present paper argues that code-switching can flag a sequence of the social and individual processes in which learners are producing, allocating and consuming in-group relationships. My central question for the analysis is: “What role does code-switching play in language learning in the particular situations that I examine?”

Auer (1998) developed a theoretical postulate on code-switching based on its local production. Traditionally, research in this theoretical framework sought to see code-switching as a resource that participants in a conversation have at their disposal in order to understand verbal actions in its context. The code-switching practices in learner interactions are also a means for both discourse-related and participant-related uses. Auer (1998) differentiates between both uses of code-switching, defining discourse-related code-switching as a way to generate meaning about the organisation of talk and participant-related code-switching as a way to generate accounts regarding participants. Auer uses the term preference-related code-switching to refer to participant-related switching: “By preference-related switching, a speaker may simply want to avoid the language in which he or she feels insecure and speaks the one in which he or she has greater competence. Yet preference-related switching may also be due to a deliberate decision based on political considerations” (1995, p. 125).

I will first analyse sequential code-switching within the context of the task required as part of the course curriculum; then I shall draw connections to the wider context in which code-switching is situated paying attention to external factors found, e.g., in questionnaires and interviews, as well as the overall order of interaction. One way to link the conversational function of code-switching with the wider context is to consider the question of language choice by participants in group interaction. For this purpose, I will adopt the concepts of “we-code” and “they-code” as differentiated by Sebba and Wootton (1998), for whom the opposition of these codes “presupposes a particular relationship between monolingual and bilingual communities, as well as particular types of social relationship within the minority group” (1998, p. 262). The aspects which inform notions of “we” and “they” can be found both inside and outside the conversation itself. “We” and “they” codes cannot be taken as given in any particular conversation. I will consider both notions based on the analysis of conversation-internal criteria and see how the “we-they-code” within the sequential structure interacts with episode-external facts.

Data and Task
The data under analysis are extracted from students’ interactions from beginner level German Foreign Language (GFL) classes. The students were working on a task, which involved the preparation of a role-play in small groups (two to three students) outside of the classroom. The role-play, to be presented in front of the class, was based on aspects of students’ lives or similar topics of their choice related to the course. The main data consist of questionnaires, interviews, recordings of the group interactions and the presentations gathered during the Fall 2007 Term. Learners’ conversations were video-taped and transcribed. Approximately 50 hours of these data were collected. Instances of code-switching were identified in the transcripts and then analysed applying CA procedures.

Example 1 below shows an exchange among three students: Rut, Lea and Sam. Rut and Lea are native speakers of Spanish. The exchange takes place while they are engaged in group work for the assigned task. The students are alone in a room, discussing the content of the dialogue they will act out in class. Example 1 illustrates discourse-related and at the same time participant-related language switches¬. In this example we observe the use of code-switching to mark boundaries between various discursive and practical activities that merge in the course of collaborative learning in small groups1. This particular exchange illustrates the ways students deploy code-switching for self-presentation, task management and fostering interpersonal relationships.

Example 1 2

After an extensive search for a name Lea would like to adopt for this role-play, she suggests that there is no need for her to tell the name. Rut responds to Lea’s words with a contradiction in line 2. The verbal act directed at opposing the perspectives of the other speaker or addressee, e.g., through contradiction, relates to what Brown and Levinson (1987) define as a face-threatening act. The term “face” is taken from Goffman (1967) who describes it as the public self-image of a person during a particular contact with others. When a speaker commits a face-threatening act, s/he risks establishing a negative face or losing face, which would impact the efficiency of communication (Su, 2009). In such circumstances, speakers often employ various resources in order to save face. In example 1 (line 2), Rut’s act of contradiction simultaneously creates a face-threatening situation. Rut switches from English into Spanish in line 2, which can be considered a face-saving strategy. One possible interpretation of the pause in line 2 is that the speaker is aware of the consequences a face-threatening act can have. Another is that Rut needs some time to find a proper response to save face. On the one hand, this code-switch relates to the opening of the conversational sequence to reinforce her argument. The switch to the native language (Spanish), on the other hand, also stands for a stepping stone to entering common ground with somebody who speaks the same L1, i.e., possibly linking to the Spanish-Canadian background of Rut and Lea. Both women mentioned in interviews that they used their mother tongue as a “secret language” while working on the task together with Sam (the third participant), who does not speak Spanish. This could also be confirmed through the analysis of the transcripts of the conversation where Spanish recurrently serves for Rut and Lea to discuss private matters. Spanish functions as a “we-code” or in-group minority language for both women. Rut uses code-switching from English to the “we-code” to release the possible tension between her and Lea due to her opposition (yes you do) to another speaker’s (Lea) argument (well i don’t have to say my name so that’s fine). The switch to Spanish to reinforce the argument functions as a “buffer” to mitigate the given face-threatening situation.

At the same time, through their language choice of Spanish, Rut and Lea create “otherness” on the part of their friend Sam who does not speak any Spanish and cannot understand them. But perhaps because of the nature of the collaborative learning activity, wherein each learner depends on and is accountable to the other, Rut switches back to English (line 2) in order to maintain work with Sam. In other words, the choice of language plays a role among all participants, be it in collaborative work on task or in interpersonal work. The move back to Spanish in line 3 confirms that the “we-code” has become important during problem solving in this conversational episode, especially for the organisation of the group work. Lea agrees to introduce the name of the role-play character and her positive response in the next turn (line 4) shows her willingness to commence to work on the assigned task. Therefore, the group can progress to the next step towards the completion of the task. The subsequent lines of the transcript provide evidence for collaborative work on the next scene of the role-play.

The student’s code-switch to Spanish in line 5 can hardly be attributed to the linguistic incompetence in L2 because in line 7 she reformulates the same utterance (What is your name?) using German. This code-switch has participant-related function and Spanish is used again to form the 'we'-group. Here, the question in Spanish is not simply directed to the interlocutor as it is immediately followed by a prompt in English (same line). The switch from Spanish to English rather relates to the organisation of discourse and task management from an interactional perspective. The student orients to building an argument by providing more information, rather than treating the question (What is your name?) as a request for a response. The code-switch from Spanish to English has both participant-related and discourse-related functions, since the student expands on her own turn and adds an emphasis to her argument. The pause in line 6 signals the participants, with the help of Spanish and English, mutually agree on this part and are ready to move on to the next stage, formulating sentences in German (line 7). An agreement on mutual understanding of the content of the role-play is of a special importance for learners; it needs to be reached before they may proceed to the next step, i.e., to practice the role-play in the target language. We see in this example ways alternations between L1 and L2 turn into an elaborate discourse competence. The learners’ practice of code-switching functions not only as a device to handle the task but simultaneously and perhaps as importantly as a strategy to create the learners’ public self-image and to define roles and relationships within the group.

Episode 2 below represents another example of language use in learner interactions. Participants step out from the task modus and enter an activity in which the parameters involving the school task are suspended. In such situations, one would expect learners to interact in a language more familiar to them, i.e., L1. However, in this non-task related conversation, students’ language of preference becomes the target language. Language alternations between L1 and L2, in fact, were observed frequently in off-task exchanges in most of the groups of the collected data.

From lines 1 to 4, learners search for a language to discuss the private issue of Rut’s date. The code-switch from English to Spanish (line 3), which is the “we-code”, is participant-related and again serves as a “secret language” to share personal information.

Example 2

In line 4, Sam requests information in German, where German functions as “camouflage” for his curiosity about his interlocutor’s friend. This language negotiation sequence reveals that the use of both Spanish and German (lines 3 and 4) functions as indirect strategy to enquire about something which may be followed by a decline of the request. The initial request from Lea in line 3 is still awaiting a response, i.e., the Spanish adjacency pair remains open, and the second request from Sam has been just uttered. Rut is now under sequential pressure to provide a response, which she does. She replies in German, and with it chooses L2 to continue the conversation. The rest of the talk in this episode displays language of preference. Although English is a language in which Rut, Lea and Sam have greater proficiency, German (not English) becomes the common symbolic ground participants have found. Learners adopt the target language as the medium for the ensuing conversation about private affairs and plans. German becomes the language in which participants feel secure to express affection, thoughts and personality, even if it is not a language in which they have higher proficiency. The function of L2 in this example contradicts Auer's framework in its features for language preference because here the language of preference is not the language in which speakers have greater competence. This use of German similarly challenges studies that report learners most frequently fall back into their native language to speak about intimate and sensitive issues (e.g., Simon, 2001).

During data analysis, I found several episodes, where learners use German in non-task related talks. I would like to argue that in such off-task conversations, German also functions to create linguistic solidarity among participants in order to try to speak in L2 about their lives (e.g., romantic affairs as in example 2, lines 5 and 7). Such cases of L2 use are remarkable because learners exhibit involvement and willingness to participate in an interaction and use the target language in situations outside of the class where restrictive code rules are suspended, and the teacher is absent.

Overall, we observed that learners treat code-switching as a meaning making strategy at the early stages of L2 development. The data provide examples of learners’ strategic use of language alternation to serve functions such as interpretation of the situation, discourse organisation, interpersonal work and expression of public self-image. In this collaborative learning context, therefore, code-switching practices have both discourse and participant-related functions that are related to foreign language learning. In a non-classroom but learning-related context, group work on task merits space for spontaneous use of L2 structuring specific opportunities for learning.

The role played by languages other than L2 in language learning has been discussed in a number of studies (e.g., Liebscher & Dailey-O'Cain, 2005 and Simon, 2001). However, no consensus on the value of code-switching in learning has yet been attained. Previous studies including those of Donato (1994), Di Camila & Anton (1997), and Ohta (2001), among others, have shown how learners can provide each other with mutual peer support in collaborative work. The current data contribute to this research but indicate an even more complex picture. Learners can intertwine the linguistic resources available to them in order to carry out specific activities, and therefore participate in multilingual interactive practices. By code-switching, learners orient to specified participants and to different social situations. Therefore, alternation between languages has social functions specifically in group work activities and cannot be simply associated with learners’ deficient in L2 language abilities; in some situations L1 use may represent mediating means for L2 learning.

Learning about code-switching practices and their functions in interaction allows us to understand the organisation of learning activities and more specifically of L2 learning in pedagogy related contexts. The interactional approach to examine ways learners deploy their linguistic resources helps to develop a situated sense of language practices in the course of group work.


1. Such strategic code-switching was observed in most groups of the beginner level courses.
2. Refer to the Appendix for transcription conventions used in the examples.


Auer, P. (1998). Introduction: Bilingual conversation revised. In P. Auer (Ed.), Code-switching in conversation: Language, interaction and identity (pp. 1-24). London: Routledge.
Auer, P. (1995). The pragmatics of code-switching: A sequential approach. In L. Milroy & P. Muysken (Eds.), One speaker, two languages: Cross-disciplinary perspectives on code-switching (pp. 115-135). Cambridge: UP.
Brown, P. & Levinson, S. C. (1987). Politeness: Some Universals in Language Usage. Cambridge: UP.
Chaves, M. (2003). The diglossic foreign language classroom: Learners' views on L1 and L2 functions. In C. S. Blyth (Ed.), The sociolinguistics of foreign-language classrooms (pp. 163-208). Heinle: Thomson.
Gumperz, J.J. (1982). Duscourse Strategies. Cambridge: UP.
Di Camila, F. & Anton, M. (1997). Repetition in the collaborative discourse of L2 learners: A Vygotskian perspective. Canadian Modern Language Review, 53, 609-633.
Donato, R. (1994). Collective scaffolding. In J. P. Lantolf & G. Appel (Eds.), Vygotskian approaches to second language research (pp. 33-56). Norwood: Ablex.
Gafarange, J. (2007). Code-switching as a conversational strategy. In P.Auer & L. Wei (Eds.), Handbook of multilingualism and multilingual communication (pp. 279-313). Mouton de Gruyter: Berlin.
Goffman E. (1999). (1967). On face-work: an analysis of ritual elements in social interaction. In A. Jaworski & N. Couplan (Eds.), The discourse reader. (pp. 306-320). Routledge: London.
Halmari, H. (2004). Codeswitching patterns and developing discourse competence in L2. In D. Boxer & A. D. Cohen (Eds.), Studying speaking to inform second language learning. (pp. 115-146). Clevedon: Multilingual Matters LTD.
Heller, M. (1995). Code-switching and the politics of language. In L. Milroy & P. Muysken (Eds.), One speaker, two languages: Cross-disciplinary perspectives on code-switching (pp. 158-174). Cambridge: UP.
Liebscher, G., & Dailey-O'Cain, J. (2005). Learner code-switching in the content-based foreign language classroom. The Modern Language Journal 89(ii), 234-247.
Ohta, A. (2001). Second language acquisition process in the classroom: Learning Japanese. Mahwah, NJ: Erlbaum.
Sebba M. and Wootton T. (1998) We, they and identity: Sequential versus identity-related explanation in code-switching. In P. Auer (Ed.), Code-Switching in Conversation: Language, Interaction and Identity (pp. 262-290). London: Routledge.
Simon D.-L. (2001). Towards a new understanding of codeswitching in the foreign language classroom. In R. Jacobson (Ed.), Codeswitching worldwide II (pp. 311- 349). Mouton de Gruyter: Berlin.
Su, H.-Y. (2009). Code-switching in managing a face-threatening communicative task: Footing and ambiguity in conversational interaction in Taiwan. Journal of Pragmatics 41, 372–392.
Unamuno, V. (2008). Multilingual switch in peer classroom interaction. Linguistics and Education 19, 1-19.


The following transcription conventions are used in the examples:

  • (1.0) length of silence in seconds
  • HI increased volume in relation to surrounding talk
  • : lengthened pronunciation, the more colons the more lengthened
  • * error
  • = latching of speakers’ utterances
  • ( ) unsure hearing
  • ? final rising intonation
  • << > > extratextual information on utterance
  • (( )) transcriber’s comments