Rethinking language teacher training: steps for making talk-in-interaction research accessible to practitioners

ABSTRACT

The goal of this paper is to enhance the quality of language teaching and improve language teacher training by making spoken interaction research accessible to practitioners. Research on teacher cognition has shown that basic beliefs and assumptions about language affect language teacher training programs and language teachers’ priorities in the classroom. Such beliefs tend to reflect teachers’ own socialization and orient to current administrative guidelines in L2 teaching, often resulting in a focus on language production of individual speakers. In contrast, a social-interactionist perspective emphasizes the co-constructed nature of language and interaction. Unpacking teachers’ beliefs and their consequences for what is taught is necessary for implementing interactional competence-based instruction. This paper suggests concrete steps to facilitate teacher training, preparing language teachers for Conversation Analysis-based Interactional Competence instruction. Such training includes, (1) sustained critical reflection of teachers’ conceptions of what language is, (2) basic training of pre- and in-service teachers in micro-analytic procedures that enable the analysis of actual talk-in-interaction, and (3) models for translating and transferring research on spoken communication and interaction into pedagogical practice. These teacher training elements: advance an empirically informed, state-of-the art view on interactional competence (IC); provide teachers with the necessary tools for meaningful, reflexive work with IC materials; and can supplement current methodology textbooks.

KEYWORDS:

• Teacher training
• language ideologies
• conversation analysis
• interactional competence

I. Introduction

Research on language as social action has increasingly been applied to topics in second language teaching, including foundational SLA investigations into the emergence of language learners’ interactional competence over time (or ‘IC’ Hall 2018; Hall, Hellerman, and Pekarek Doehler 2011; Pekarek Doehler and Pochon-Berger 2015, 2016; Young 2011), and efforts by language teachers and teacher educators to include interactional structures and phenomena such as response formats, repair, or storytelling alongside words, grammar, and other functional skills underlying language proficiency in their courses (Barraja-Rohan 1997, 2011; Barraja-Rohan and Pritchard 1997; Huth and Taleghani-Nikazm 2006; Betz and Huth 2014). Thus, the teaching mandate in L2 classrooms now encompasses building learners’ knowledge and abilities that are situated in, and negotiate, the sequentiality of spoken interaction (Schegloff 2007). Insight into the sequentiality of interaction and its social-interactionist nature relies on research in Conversation Analysis (CA) (Sidnell and Stivers 2012).

Figure 1. Prompts for discussing language use in a common teaching methods textbook (Larsen-Freeman and Anderson 2011, 129).

Note the reliance on teacher intuition.

This view of language stresses the in-between space between interactants and actions, positing that mutual understanding, meaning, and cognition are distributed among interactants when they make sense of each other’s conduct in real time. Interactional abilities that facilitate this process include, for example: recognizing a turn containing a question as a request for information and responding to that question with an appropriate informative answer; realizing that a request for information such as ‘What are you doing tonight?’ may constitute a pre-invitation (Beach and Dunning 1982; Schegloff 2007, 29–34), and that an informative and appropriate response to such a pre-invitation is one that addresses availability for a common activity; or initiating and going through an appropriate closing-sequence before hanging up at the end of a telephone call (Harren and Raitaniemi 2008; Schegloff and Sacks 1973). Engaging in relevant, fitting, and recipient-designed social actions in specific interactional contexts is thus achieved collaboratively by two or more speakers who produce language and embodied (e.g. gesture, gaze) conduct across at least two turns, in the context of larger courses of action, and in a specific setting and within a history of interaction, thus relying on common ground (Clark 1996; Enfield 2006).

This paper discusses how teacher education may begin to design training materials that provide future language teachers with the knowledge, skills, and confidence to teach interaction competently in the classroom. Foundational education on what language is and how language works is part and parcel of introducing social-interactionist perspectives to second language teacher training. As we illustrate here, language teaching is at this point only beginning to initiate this process. To that end, we examine and see this reflected in basic conceptualizations of human language in methodology and pedagogy textbooks, in notions of proficiency in larger, transnational frameworks for curricula, and in approaches to assessment suggested in the ACTFL Standards (American Council on the Teaching of Foreign Languages 2012) and the CEFR (Council of Europe 2001, 2018).

The goal of this paper is twofold. First, we seek to illustrate that language teacher education needs to address this broad range of topics if current efforts to enrich language teaching with interactional perspectives are to be successful. Second, we suggest concrete initial training measures and discuss opportunities for how to implement these measures. In pursuit of these goals, we first consider the impact of teachers’ beliefs and attitudes towards language on classroom practice (‘language ideologies’). This is followed by a discussion of how such basic notions of language are reflected and put into action in central aspects of language teacher education, illustrating that a substantive engagement with social-interactionist views is largely absent in training materials (e.g. education textbooks). While there is a growing body of work that aims to sensitize teachers to the workings of classroom interaction and increase teachers’ interactional competence in the classroom (Seedhouse 2005; Sert and Seedhouse 2011; Walsh 2012; for the German context, see e.g. Heller and Morek 2015; Kotthoff 2013; Kupetz 2018; Paul 2010), there is little work addressing and guiding teachers’ engagement with their own views of language and interaction. Finally, we suggest three domains in which concrete steps to address this notable absence can be taken, namely sustained critical reflection of teachers’ conceptions of what language is, basic training of pre- and in-service teachers in micro-analytic procedures that enable the analysis of actual ‘talk-in-interaction’ (Schegloff 1987), and models for translating and transferring research on communication and interaction into pedagogical practice.

II. Beliefs and attitudes toward human language

This section focuses on conceptualizations of language in the language teaching profession (‘teacher cognition’, see Borg 2003, 2009) and in the ways current curricular guidelines and assessment frameworks reflect basic ideas about language (such as the ACTFL Standards and the CEFR). We argue that, if innovation efforts towards curricula that include an orientation to IC are to be successfully implemented, addressing ideological matters concerning what language is and how it works in interaction is necessary. With ideological matters we do not mean political or ethical leanings of individuals, but rather language teachers’ notions and preconceptions about the nature of language which have their roots in a process of lifelong socialization rather than formal training or empirical research.

First, it is important to realize that language instruction that focuses on interactional matters approaches human language from a social-interactionist perspective. This perspective views language as social action (Maxwell Atkinson and Heritage 1984), highlights how social action is constructed across speakers and across turns in sequences (Schegloff 2007), and posits that sounds, words, morphosyntactic structures, and temporal (Deppermann and Günthner 2015) as well as embodied (Streeck, Goodwin, and LeBaron 2011) resources are part and parcel of human language. This understanding of language is usually not part of the classroom experience language teachers themselves had as learners, and it is therefore not available as a resource to draw from when formulating and putting into action instructional plans. On a basic level, language teaching entails negotiating and putting into practice one’s ideas, beliefs, and basic assumptions about what language is. Without an underlying idea about the object of learning and teaching in a language classroom, no instructor can set reasonable goals for their work. Not surprisingly, prior research demonstrates that teachers’ beliefs are deeply consequential for teachers’ ‘methodologies’ (Kumaravadivelu 2006, 84). This includes what is being taught and how teachers set priorities in their daily work (Andrews 1999, 2001; Borg 2003, 2009; Freeman and Richards 1996; Wood 1996).

Second, these ideas and attitudes often remain unexamined, forming an implicit ideological framework. Sociolinguists call such basic stances and attitudes towards language ‘language ideologies’ (Schieffelin, Woolard, and Kroskrity 1998; Garrett 2010). As Garrett notes, ‘people hold attitudes to language at all its levels: for example, spelling, and punctuation, words, grammar, accent and pronunciation, dialects and languages’ (Garrett 2010, 2). These abound and exist irrespective of empirical corroboration or validation and therefore include myths and misconceptions (Wardhaugh 1999). Common beliefs and misconceptions include prescriptive (hence moral) notions about correct and incorrect language use and its users (e.g. ‘Ain’t is not a word.’, or ‘People who say Nobody ain’t done nothing. aren’t thinking logically’), viewing written language as superior to spoken language, (e.g. ‘Correct language can be found in books, not in conversation.’), or viewing dialect as inferior to standard language (e.g. ‘I hide my southern accent at work.’). How talk-in-interaction and language use fare alongside all other levels of language would be subject to similar intuitions.

Third, and to confound matters, language teachers necessarily connect such basic assumptions with their ideas about language, language learning, and language teaching, all of which are empirical matters. Linguists study how human language works, cognitive research studies the general capacity of the human brain to learn (Brown, Roediger, and McDaniel 2014), and applied linguistics continue to (re-)conceptualize the process of language learning over time (Seedhouse, Walsh, and Jenks 2010; Wagner, Pekarek Doehler, and González-Martínez 2018). In applied research, emerging conceptualizations of language and language learning provide new knowledge for pedagogy and curriculum. The transfer of this knowledge into language education expresses itself as the emergence of new texts on pedagogy and methodology manuals, as well as in the process of updating specific modules, courses, or degree programs that serve language teacher training in educational institutions.

However, the success of this transfer is not guaranteed. Beyond research, another force affects how current and future teachers conceptualize all three domains of language education: the individual’s lifelong socialization by their surrounding linguistic, cultural, and educational context. Socialization (including language socialization, see Schieffelin and Ochs 1986) describes the molding and stepwise entrenchment of an individual’s knowledge and social practices over time by way of cultural transmission. Clearly, language teachers are not clean slates when entering teacher training programs. Rather, they start their formal teacher training already equipped with deeply held beliefs about language, learning, and teaching rooted in their experiences of what it means to learn, what it means to be taught, and how their teachers and their materials approached and handled the notion of ‘language’.

Trainee teachers are thus also prior language learners that have been shaped by the very educational systems they now seek to put into practice. In a study examining how teachers conceptualize and prioritize grammar teaching, Borg (2003) concludes that ‘studies of grammar teaching grounded in classroom practices and teachers’ rationales for these have elaborated a multifaceted conception of teacher thinking and teacher knowledge in which teachers’ understandings of language, pedagogy, and learners, shaped by the teachers’ educational and professional biographies, interact in complex ways to define instructional decisions and practices in formal instruction’ (105). Additionally, language teachers’ professional identity and professional ethics draw from often implicit and morally grounded definitions of what a good/bad teacher (or what good/bad teaching) may be and what specific practices, methods and techniques are connected to such definitions. In other words, when faced with their own classrooms, language teachers tend to reproduce roles and ideas that stem from their experience as learners (e.g. Borg 2003; Schmenk 2015). It is thus important that language teacher training includes an overt orientation to the ideas, beliefs, and assumptions trainee teachers bring to the training (Crane 2015; Crane et al. 2013).

This cycle of a system reproducing itself can stand in the way of innovation efforts. We emphasize that it is empirically difficult to obtain data that pinpoints with precision what exact views the majority of current and future second language teachers may hold towards language, learning, and teaching (see Section III for a discussion of one approach). However, it is possible to examine what ideas about language, learning, and teaching are currently reflected in the larger frameworks that language teachers will likely encounter once they enter teacher training and/or their own classrooms. We do so below by examining the topics, language, and priorities in two domains of language teaching, first in current teaching methodology textbooks, and second in in the larger transnational frameworks for curricular standards and proficiency such as the ACTFL Standards and the CEFR. Overall, this examination shows that substantive engagement with social-interactionist perspectives is largely absent in the materials that guide teacher training, curricular development, and proficiency or assessment standards.

III. Views of language in methods textbooks and proficiency frameworks

From the outset, we take it as given that educational frameworks, including those for language teaching and language teacher training, are not homogenous within and across national boundaries. That notwithstanding, this article applies a transnational perspective inasmuch as its suggestions are embedded in the relevant research discourse of European and Anglo-American language and language teacher education. As German language program directors/coordinators in Canada and the United States, our experience, considerations, and suggestions primarily reflect a North American educational context. However, we trust that many of our considerations are relevant across the boundaries of national educational contexts, such as the application of proficiency frameworks (such as ACTFL or the Council of European Framework provide them), the use of language teaching methodology textbooks in college-/university-level pedagogy courses, the usage of language textbooks and/or supplementary materials in second language classrooms, and a given measure of methodological pluralism as it is manifested in teacher education programs as well as in the day-to-day classroom realities in primary and secondary education.

Language teaching methodology textbooks are a central resource for trainee language teachers at colleges in North America, where language departments are typically led by a language program director. Consequently, the content of language teacher training programs is based on the language program directors’ beliefs and assumptions about the nature of language and the process and goals of language learning. Usually, language teacher training programs consist of a teaching methodology course and pre-and in-service activities such as workshops and regular staff and coordination meetings. One goal of such training is to help new language teachers develop a basic understanding of second language acquisition and language teaching methodologies.

We conducted a survey of a large number of language program directors at North American universities, and this survey revealed that a small number of teaching methodology textbooks (including Brandl 2008; Larsen-Freeman and Anderson 2011; Lee and VanPatten 2003; Loewen 2014; Kumaravadivelu 2006) are widely used in French, German, and Spanish language teaching methods courses. It is important to note that the view of language we find in these texts is grounded primarily in psycholinguistic research in second language acquisition. Hence, these language teaching methodology textbooks assume language to primarily comprise a code or linguistic knowledge (i.e. grammar and vocabulary) connecting form and meaning. This knowledge is viewed as complemented by language processing skills for efficient comprehension and production in real-time communication.

However, beyond psycholinguistic notions of processing, we can also find an orientation towards the social purposes of language. Viewing communication as the expression, interpretation, and negotiation of meaning (Brandl 2008; Lee and VanPatten 2003), some of these textbooks provide social-interactional dimensions for language teachers to consider when planning their language class. For example, Lee and VanPatten (2003) describe ‘psychosocial’ goals (for bonding with others) and ‘information-cognitive’ goals (for getting information) as the two goals for communication (150). However, when it comes to developing speaking skills and communication in the classroom, teaching methodology textbooks still primarily treat spoken interaction as a site for teaching and practicing new linguistic forms. Information- or opinion-gap activities, for example, focus on correct use of grammar, phonetics, and vocabulary rather than creating opportunities for negotiating social-interactional meaning.

One of the gaps we see in existing methods textbooks is that they do not offer examples and discussions of how speakers of the target language actually co-construct conversation and perform actions such as invitations or requests in real-time across two or more speakers and across two or more turns at talk. As the example below (Figure 1) illustrates, the discussion of language use and L2 interaction typically rely on the intuition of language learners and language teachers.

As a result, these teaching methodology textbooks do not address the structural systematicities of real-time communication/interaction and the highly context-dependent intricacies of communication: How speakers select linguistic forms for action, how they design their turns for a specific addressee, how they display and negotiate understanding in interaction, and how they repair locally emerging problems in intersubjectivity.

Language instructors who wish to teach with an orientation to IC in a second language therefore need to be exposed to a variety of recorded real-time interactions in the target language and to critically reflect on the structure and organization of conversations to discover that it is not just about what is said but also how and when it is said, that is, the details about composition and position of utterances. In light of the very limited intuition of native speakers about pragmatics and interaction structures (cf. Golato 2003), artificial and fabricated dialogues that are commonly presented in language textbooks can be misleading (Wong 2002). Thus, despite the widely recognized importance of equipping language learners for successful real-time interaction in the target language, instruction that targets pragmatics and interactional competence may be inadvertently encumbered by instructors and material designers.

In sum, even though action-based conceptualizations of language have been recommended in language learning/teaching (Bardovi-Harlig and Hartford 1997; Kasper 1997), see Figure 1, notions of language-in-interaction rooted in the sequentiality of real-time talk continue to be largely absent in methods books, cementing a conceptual blind-spot in the profession. We acknowledge that methodology textbooks are often not the sole resource informing language teacher education. In some programs, language teacher trainers who happen to have a background in social-interactionist research are supplementing methodological training with materials that are beginning to address the gap. However, within standard textbooks on language pedagogy, the gap persists and we posit that it would be desirable to move from supplementing to including materials on social-interactional perspectives on language and language learning into the conceptual canon of pedagogy training. If language teacher training is to include a stronger orientation towards insights on language as social interaction, improved, updated, and evidence-based textbook units on interaction as well as training and guidance for language program directors (who guide teachers) and teachers themselves are needed.

This notable absence is largely paralleled in the language, concepts, and goals of curricular guidelines and proficiency-oriented assessment frameworks offered by the ACTFL Standards and the CEFR, which are consequential for large parts of Europe and North America. Since these are consensus-based documents for second language teaching, they contain and reflect particular views of language, language learning, and language teaching (see Tschirner 2012 for comparisons, contrasts, and efforts for aligning both conceptually). Both frameworks are subject to regular updates (see the most recent addendum to the CEFR from 2018), including updates fueled by emerging insights, advances, and adjustments in theory and language research.

A close reading reveals that, on the one hand, some the programmatic language in the ACTFL Standards and in CEFR reflects overt acknowledgements of advances in research that casts human language as a variable and diverse cultural and social resource. However, this acknowledgement is not (yet) used as the basis for further action, as it has been neither translated into tangible learning targets, nor reflected in the proficiency and assessment standards of the ACTFL Standards and CEFR (which otherwise present nuanced descriptors for specific proficiency levels). Above all, it is notable that the proficiency descriptors focus on individual language learners and individual language production rather than on how two or more interlocutors negotiate meaning in the back and forth of talk. Thus, from the narrow context of proficiency descriptors alone, a view of language emerges that is both steeped in, and indeed primarily shaped by, the institutional constraints in which it operates. Written and oral functionality in an L2 continues to be ascertained for the individual language learner primarily in terms of individual language production through presentation or interaction, and it is done with a strong focus on a language learner’s relative access to L2 lexical and morphosyntactic material (Huth forthcoming).

This view of human language drives second language teaching across a large number of educational systems, as it relevantly aligns with the institutional contexts and goals that produce it. It is useful to realize that, beyond building target language proficiency in learners through teaching, one inevitable goal of educational institutions is to process, sample, and assess language production of individual language learners as they progress through the institution over time. This goal drives much of how language educators must view learner language production – namely, as a ratable, assessable quantity to be ascertained for individuals (Roever 2018; Roever and Kasper 2018). We posit, however, that the expansion of this view of language in light of insights from social-interactional research on language is both possible and desirable.

In sum, efforts to innovate curricular frameworks should address language teachers’ prior socialization as language learners and the resulting ideological stances towards language, language learning, and language teaching that teachers bring to their formal training. In order to conceptualize and eventually implement an IC-oriented and IC-research-informed instructional model in second language teaching, teacher education courses and teaching methodology manuals stand to benefit from a bottom-up approach to educating teachers-in-training by including materials suitable to present a view of language that includes basic insights from social-interactionist research. Furthermore, top-down efforts that aim at updating and thus facilitating knowledge transfer from the empirical world of research to the world of curricular models, proficiency descriptors, and assessment frameworks are needed. As the profession continues to balance its orientation to the contingencies of educational systems on the one hand and the advances of empirical research on human language on the other hand, we propose some concrete initial measures below that can be undertaken in this pursuit.

IV. Practical suggestions for teacher training

We now outline some elements of teacher training that can facilitate effective CA-based IC instruction. We propose that certain aspects of teacher training need to be reconceptualized and complemented with critical reflection on the impact of language attitudes and teacher cognition, especially on the importance of teachers’ conceptions of what language is, and what should thus be included in teaching. Additionally, pre- and in-service teachers need basic training in analyzing interaction (micro-analytic principles) to practice seeing patterns and understanding that language is for social action. Lastly, teachers need models for ‘translating’ linguistic and communication research into pedagogical practice. We focus on teacher training in language programs at North American universities, but what we propose also connects with goals in teacher training in Europe, for instance the current discussions regarding professionalization of teachers in educational sciences.¹

A central part of teacher training is to help teachers develop reflectivity and reflexivity (Kumaravadivelu 2006; see also the ‘Leibniz principle of reflected action’²), which includes teachers’ ability to reflect on their beliefs and actions in and outside of the classroom. Teachers’ beliefs about language and language use are rooted in both, in discourses about language as well as in teachers’ experience with language and language teaching (see, for example, Allen 2002; Flores 2001). Beliefs about language are rarely systematically reflected in teacher education, and this includes the training of language teachers. Research on teacher cognition (for example Li and Walsh 2011; Schmenk 2015; see Borg 2003 for an overview) suggests that teachers’ experiences, beliefs, and assumptions strongly influence their practical decisions as teachers, and this research specifically shows that different cognitions about language and how languages are learned shape teaching decisions and practice (e.g. Almarza 1996; Williams and Burden 1997, 57). Such cognitions can play a crucial role in teachers’ decisions to accept or reject a new curricular approach (Donaghue 2003).

We suggest that reflection on one’s own view of language can be instrumental in raising teachers’ language awareness and may affect not only behaviour but cognition, thus potentially enhancing willingness to cooperate in the curricular implementation of an IC approach. In line with research positing that teachers should engage with their prior experience, pre-existing assumptions, and beliefs early in their training (e.g. Cabaroglu and Roberts 2000), we begin with a specific exercise that is suitable for the beginning of a teacher training session, course coordination meeting, or professional workshop series.

(a) Reflection about beliefs

One useful method for accessing implicit beliefs can be found in metaphor-focused research (Barcelos and Kalaja 2011; Cortazzi and Jin 1999). We suggest the sentence completion exercise shown in Figure 2 as an entry into exploring teachers’ beliefs about talk-in-interaction (Schegloff 1987), thus unpacking beliefs about what aspects of language should be taught in a language class and exploring the potential connection between beliefs and pedagogical decisions. Participants are invited to complete the prompt presented, and to do this, they have to find metaphors (or rather ‘metaphorical similes’, Cameron 1999, 131) for conversation. In order to allow them to use their full range of linguistic expression, we present the prompt in at least two languages shared by workshop participants; they complete the prompt in any language they choose. The following prompts and the responses shown in Figure 3 are taken from an authentic workshop conducted at the University of Waterloo, Canada.³

Figure 2. Example prompt in English and German to elicit beliefs about language-in-interaction.

Figure 3. Sample metaphors collected in response to elicitation prompt (Sept. 2018, translated).

After giving participants time to consider possible sentence completions, they are asked to collect them on a physical or shared online space. This is followed by a group discussion in which each participant shares their metaphor and explicates what exactly conversation and, e.g. ballroom dance, table tennis, an onion, a labyrinth, and pulling teeth (see Figure 3) could be said to have in common, and also how they differ. Participants are encouraged to probe others’ explanations, expand them, and connect them to their own. In this process, they are encouraged to articulate what crucial elements of conversation beyond talk are, thus building a multimodal view of language-in-interaction (Mondada and Traverso 2015; Streeck, Goodwin, and LeBaron 2011).

Exercises such as this, which elicit metaphors or metaphorical similes, offer ‘a way to expose thinking that may otherwise be difficult to articulate’ (Fisher 2017, 329). They have been used to gather data on how teachers conceptualize themselves and understand their role as language teachers, and how learners and students conceptualize the experience of learning a language (see Fisher 2017 for an overview). There is little work, however, on beliefs regarding language or talk-in-interaction as it relates to teachers(-in-training). We have found this task to be useful in teacher training, because it facilitates articulation and discussion of different concepts of language and can lead to deeper or new insights about teachers’ own underlying assumptions and their connection to knowledge, experiences, and socialization processes. Moreover, it enables participants, with guidance from a workshop facilitator, to collaboratively collect core concepts of talk-in-interaction and thus establish the foundation for analyzing transcripts of actual interaction and for discussing interaction research in a next step.

Our workshop participants’ metaphors (a sample of which we depict in Figure 3) constitute genuine attempts at expressing beliefs and conceptualizing something highly complex creatively. For example, one attempt by a recent workshop participant visualizes conversation as a carpet. Her explanation focused on how contributions to conversation, just like threads in a carpet, hang together and over time amount to a larger structure co-created by more than one person. In group discussion, this comparison was expanded, and participants were able to collect and unpack basic concepts in understanding interaction, including conversation as intersubjective meaning-making, the importance of sequentiality and multimodal resources, and the connection between interactional history, shared knowledge, and relationship building.

It should be noted that the goal of this exercise is not to elicit ‘stable cognitions capturing how people conceive of the world’ (cf. Fisher 2017, 331) but rather to open up a space for reflection, for the expression of differing, often contradictory, views of language, and for teachers to negotiate shared understanding of the complexity and importance of language in social interaction.

(b) Understanding real-time interaction

The exercises that build on this initial reflection expand teachers’ knowledge of how interaction works through guided data analyses. Unpacking how interaction works entails tracing and seeing speakers’ systematic methods for meaning-making. The goal of learning to analyze actual data is to enable teachers to see and understand the connection between language choices and social action, to further their understanding of potential benefits of IC-based materials to learners and to the overall curriculum, and to help teachers develop views of competence and proficiency that are interactionally grounded.

Why work with transcripts rather than speakers’ (teachers’) intuitions about talk-in-interaction? Members of a language community cannot normally describe in detail what they systematically do in real-time interaction. Intuition is thus not a reliable source for understanding members’ interactional practices (Golato 2003). Working with transcripts of recorded interaction (along with the interactions themselves) can provide basic training in micro-analytic principles to teachers, because it allows them to inspect closely what happens in interaction. We argue that teachers need to engage consciously and systematically with basic structures of interaction in order to understand the difference between language intuition and actual use, to see core aspects of interaction such as sequentiality and co-construction in action, to realize that linguistic items are ‘fundamentally embodied’ in nature (Eskildsen and Wagner 2015, 442; see Langacker 1987) and to thus come to understand how ‘language is linguistically realized social action’ (Eskildsen 2017).

The following illustrates a step-by-step analysis of short transcripts, specifically of the actions done in each turn. Each example is taken from real-life conversations and shows how language is actually used in interaction, and each can be used to illustrate concepts of language as action, its sequentiality, and its progressivity.⁴ During the exercise, each data example is revealed turn by turn, and teachers are asked to describe first what each speaker does in each turn and, second, how the specific words and syntax (and possible non-verbal elements such as gaze) fit together. The discussion introduces the notions of syntactic and action projection and the basic concept of a sequence, and it emphasizes the importance of next-turn-position as the crucial place for locating displays of understanding. During this task (that is, after each extract), we ask teachers to use the specific evidence just gathered from the transcript to answer broader questions such as the ones listed in Table 2.

In this exercise, the examples themselves guide the ‘discovery’ of core aspects of what determines linguistic choices and the resulting recurrent patterns in language-in-interaction. The aim is to enable teachers to see that language primarily serves social-interactional purposes (e.g. the negotiation of aligning evaluative positions when assessing a shared experience, as in Extract 1 ‘fun’ in Table 1) rather than, say, efficient information transfer, and that real-time-interaction is essentially temporal, sequential, and co-constructed. The analysis of actual recorded and transcribed data in teacher training also allows for a refocusing away from intuitions about language use. Analyzing recordings and transcriptions of interaction can thus serve for critical discussion of and reflection on the notions and conceptions of language that teachers hold; it also emphasizes the need for basic linguistic and communication research in teaching interaction patterns.

Table 1. Data extracts for guided analysis of talk-in-interaction and discussion points for teachers.

Data extracts (Sidnell 2010, 2; Sacks 1995, 665; Schegloff 2007, 109; cf. Wong and Waring 2010, 60):	Discussion points:
(1) ‘fun’ 01 Ann: that was fun, 02 (0.4) 03 Jeff: mm 04 Ann: ish. 05 Jeff: yea:, I guess (2) ‘roger’ 01 Dan: well, roger uh 02 Roger: hm? 03 Dan: introduced a kind of topic 04 when he uh ... (3) ‘bud’ 01 Cus: May I have a Budweiser? 02 Ser: Are you twenty-one? 03 Cus: No. 04 Ser: No.	for extract (1): units of language v. units of action in interaction; agreement and affiliation as essentially co-constructed for extract (2): actions; recipient design; importance of non-verbal aspects of interaction (e.g. gaze); multimodality for extract (3): sequentiality; two-part sequences; the concept of projection; co-construction of identities in interaction

The materials and steps we suggest here, notably the detailed analysis – with teachers – of spontaneous interaction, are not to train future conversation analysts. They are instead an approach to professional training that is based in conversation analysis. In using interactional linguistic and conversation analytic insights and evidence about recurrent patterns and practices, it takes seriously the connection of research/theory and praxis.⁵ In using transcripts and recordings (see footnote 5 for sources) of actual encounters rather than intuition, role-play, or invented examples, this approach exposes teachers to evidence of how real people manage real interaction and how meaning is negotiated in the minute details of interaction. This ‘examin[ation of] communicative practices in forensic detail .. to understand what works from a rigorous empirical basis’ (Stokoe 2014, 257) includes not only interaction between persons considered fully competent speakers of a language, but also language learner interactions. We will see this illustrated in section (c) below.

Our approach is similar to a current, widely known approach to applying conversation analytic evidence to professional training, the Conversation Analytic Role-Play Method (CARM, Stokoe 2011, 2014). In police work and community mediation services, CARM training has proven to be effective in removing roadblocks to institutional interaction by helping ‘people to understand the landscape of their particular workplace’ (Stokoe 2014, 258). One of the most recent expansions of this work is to classroom interactions between students and teachers (CAiTE project; see Skovholt 2018; Skovholt et al. 2018). At the core of the CARM approach is ‘an analysis of people actually doing their job’ (Stokoe 2014,, 258) rather than of intuitions about what is or should be done. Similarly, in our approach, we guide teachers’ engagement with and discussion of what people actually do in interaction and thus problematize reliance on intuition about language and interaction practices. Different from recent CARM integrations in the training of teachers, we are not focusing on the analysis of teachers’ interactions with students or of classroom interaction, but on the analysis of everyday interaction. This is because our goal is different. We aim to raise teachers’ awareness of their own beliefs and assumptions, and of the ‘machinery of interaction’ (Sacks 1995; cf. Levinson’s (2006) ‘interaction engine’); we do not aim to intervene in teacher-student communication practices. However, what we suggest here can be followed and complemented with classroom interaction analysis, either the analysis of teachers’ own interactions with students (cf. Sert 2015; Skovholt et al. 2018) or of peer interactions among learners.

The success of CARM training in other areas is suggestive of the success we have observed in our teacher training interventions. Critical reflection on the impact of language attitudes and teacher cognition, and especially basic training in CA theory and methods, improves professional practice.

(c) Understanding learner interaction

In addition to promoting reflection about one’s own concept of talk-in-interaction, guided analysis of interaction data can also help teachers differentiate notions of proficiency and competence from notions of correctness and target-likeness, and thus better understand the complexities of language learners’ ‘evolving conversational practices in educational settings’ (Wagner, Pekarek Doehler, and González-Martínez 2018, 16). Reconceptualizing proficiency as diversification and fine-tuning of speakers’ methods for action (Pekarek Doehler and Pochon-Berger 2011, 2015) helps us account for some of the complexities of L2 development, as illustrated in findings on L2 precision timing (Cekaite 2007), sequence organization and topic management (Pekarek Doehler and Berger 2016), and repair practices (Hellermann 2009).

Below is a sample task that focuses on conversational repair. We chose repair as an example in the materials, because it lets us highlight the important distinction between universal mechanisms and language-specific resources: Repair is a generic organizational principle in interaction, that is, it is a fundamental part of the machinery of interaction across languages (on other-initiation of repair, see Dingemanse and Enfield 2015; Dingemanse et al. 2015; on self-repair, see Fox, Maschler, and Uhmann 2010; Fox et al. 2017). The methods or resources for accomplishing repair, however, are language-specific (ibid.). Focusing on repair patterns can enhance teachers’ understanding of both generic principles in interaction (e.g. that a common way to initiate repair across languages is with open-class repair initiators such as English Huh? or What?; Drew 1997) and language-specific connections between form and function (e.g. what the choice between different available repair initiators communicates).

The first three examples (Table 3 and Figure 4) show self-repair in English and other-repair in French L1 every-day interaction. They serve to prompt the exploration of social functions of repair: Repairs do not just address errors in speaking and hearing, but are systematically connected to social-interactional considerations. Furthermore, repairs can be a window onto online cognitive processes.

Table 2. Broad questions for discussion and guiding suggestions for teachers.

Broad questions for discussion:

Specific guiding points for discussion:

What are basic units of language? What is the relationship between specific language forms and practical action? What is context? How does context determine how we understand what somebody else just said (did) and what is relevant in response?

for extract (1): ‘What counts as a recognizable word?’, ‘As a recognizable action?’; ‘What do we do with language?’; We see that assessments are co-constructed in and through interaction: Jeff and Ann negotiate a shared perspective on a past event. for extract (2): To whom a turn is addressed is essential in understanding the action it does, for example as addressing a person/summoning their attention or as referring to someone. for extract (3): Institutional identities such as customer and salesperson in an American liquor store and visible identities such as age are essential dimensions for making sense of another’s actions.

Table 3. Data extracts for guided analysis of self-repair practices in English.

Data extracts (Goodwin 1987, 119; Jefferson 1974, 193):

Notes for workshop facilitators:

(1) ‘to bed’ 01 Jane: we went t- I went to bed 02 really early. (2) ‘held up’ 01 Jane: well, she said that there was 02 some woman that- tha-that  03 they were whh- had held up 04 in the front there, that 05 they were pointing the gun at, 06 n everything, (0.4) a k- negro 07 woman.

step 1: Ask participants to articulate what they see in the transcripts; no specific further instructions. step 2: Revisit each example; ask about possible social functions of self-repairs. Repairs may be connected to recipient design: to what is considered socially (in)appropriate word choice or to what is in the common ground. Note that Jane in Extract 1 is speaking to her parents, who do not know that she is living with her boyfriend (we, line 01).

Self- and other-repairs can make visible participant’s methods for indexing and negotiating membership, for example in linguistic communities. The participants in the face-to-face interaction in Figure 4 negotiate a lexical item, and ‘multiple instances of repair on the lexical items bleuets “blueberries” (used in Canadian French) and myrtilles “blueberries” (used in “Metropolitan French”) [aid] in the establishing, confirming, asserting, or negotiating of speakers’ membership in one category rather than another’ (Huensch 2017, 359; cf. Maheux-Pelletier and Golato 2008).

Figure 4. Data extract for guided analysis of other-repair practices in L1 French.

Data extract ‘bleuets’ (Maheux-Pelletier and Golato 2008, 694–695)

The target of repair in the next example (Figure 5 ‘windsurfing’) is also a lexical item, but this time the source of trouble is knowledge or retrieval of the item, resulting in a search for a word. The extract is from a L2 French classroom in a Swiss German high school. In our workshops, we typically present ‘bleuets’ and ‘windsurfing’ initially without ethnographic information and invite participants to make guesses about the interactants’ identities (and about each interaction’s setting). We also ask participants to anchor their guesses in details of the interactants’ observable conduct. Participants quickly classify windsurfing as a classroom/instructional interaction, but they also tend to do this with bleuets (which is from an interaction in which friends cook together), since trouble with meaning of every-day words is commonly associated with learner interaction. This can be another entry point for reflecting on talk-in-interaction, specifically on the connection between repair and errors and on the processes by which we use language for membership categorization.

Figure 5. Data extract for guided analysis of L2 repair practices.

Data extract ‘windsurfing’ (Fasel Lauzon and Pekarek Doehler 2013, 329)

In Figure 5, we can see that André is searching for a word, which is visible in hesitations and pauses in line 01. He also makes the unavailability of a word explicit with je ne sais pas ‘I don’t know’ (line 01) and then offers a tentative solution, which his teacher translates into French (line 03). André repeats the solution, thus redoing the completion of the first item in his projected list of activities and resuming his telling (line 04). We can also note that the action of searching for a word is organized turn-by-turn, and that – just as in word searches that competent speakers carry out – hesitation and pauses are integral parts of the action, and of making the action visible and understandable for others. This, in turn, enables the teacher to react appropriately and the student to resume what was put on hold by the trouble.

Fasel Lauzon and Pekarek Doehler note that ‘[t]he teacher’s alternative wording (l.3) is interactionally occasioned by the very way André delivers his preceding turn: the teacher does not merely react to an error; rather, he reacts to a student’s display of trouble with a linguistic issue, which he appears to interpret as a call for help.’ They emphasize that ‘[c]onsidering the correction here as a teacher’s reaction to a student’s error would render an oversimplified picture of the socio-cognitive processes at work’ (2013, 330) which becomes visible in the turn-by-turn unfolding of the interaction. Extracts such as ‘windsurfing’ underscore that repair is an important part of interaction and thus of interactional competence. Methods for repair are diversified, and the types and social functions of repair are systematically connected.

Understanding these foundational processes should be part of the training in linguistics for language teachers. They should also be taught and practiced in L2 classrooms, since languages provide different lexical, syntactic, and sequential resources for accomplishing repair (Betz and Huth 2014; Waring 2018). As a next step after analyzing L1 and L2 data on repair, we typically work through actual classroom teaching materials with teachers, for example a unit on German obwohl ‘except/ although/ come to think of it’ as a repair initiator (Bendig, Betz, and Huth 2016). A combination of analysis of interaction and work with teaching materials models for teachers how findings about language-in-interaction can be applied to the classroom. A combination of both analytic formats enable teachers to better understand language empirically and understand how they can teach interaction.

Thus, beyond having to grapple with an understanding of socially distributed cognition and of meaning as essentially co-constructed, work with recorded and transcribed language-learner data can help balance a view of L2 language production as essentially deficient, that is, it can counter a focus on what learners cannot do. It shows teachers that not all trouble in interaction is due to limitations in L2 knowledge or to processing problems. Similarly, pauses, hesitations, and other apparent disfluencies do not necessarily indicate issues with speech production but are systematic features of talk-in-interaction and thus need to be understood as resources for speakers. Work with actual interaction data also allows teachers to observe the use of non-vocal resources, to realize how interactants use gaze, gesture, movement in space, and objects to accomplish actions (e.g. Mondada 2014; Rossi 2014; for L2 interaction, see Eskildsen and Wagner 2013, 2015; Taleghani-Nikazm 2015), and to revise the ‘widely held view that embodiment in L2 talk has a largely “compensatory” function, making up for deficient language knowledge’ (Roever and Kasper 2018, 349).

In the materials we present here as a first step in language teacher training, we focus on work with transcripts. An important next step would be to develop additional materials that offer both transcripts and original recordings. In footnote 5, we point to a variety of sources that provide suitable video and audio data for this purpose. Such insights foster a differentiated and complex understanding of what successful real-time communication entails. They are crucial for developing a proficiency, can-do-view of learners’ engagement in interaction. Thus, engagement with recordings and transcripts of interaction enables both beginning and experienced teachers, including language program directors, to contextualize and particularize institutional curriculum guidelines and assessment frameworks (such as the ACTFL Standards and the CEFR) in teaching, materials development (e.g. textbook writing), and testing.⁶

V. Conclusion

The language teacher training materials presented here aim to raise teachers’ awareness of their own beliefs about language and to enhance their knowledge of the machinery of interaction. The explicit goal is to enable teachers to look at language and meaning as essentially co-constructed in interaction. This would enable them to reconceptualize language learning, and specifically learners’ development of spoken interactional competence, in ways that shift away from a deficiency-focused, native-speaker-oriented model and foster a view of learners’ development as a ‘re-calibration [that] entails an increased ability for context-sensitive conduct based on speakers’ progressive diversification of methods for action’ (Wagner, Pekarek Doehler, and González-Martínez 2018, 69). The necessary rethinking on the part of language teachers in an educational setting is difficult to achieve without a sustained engagement with the precise machinery of social interaction through recorded and transcribed data. We argue that such an engagement, and particularly a better understanding of the connection between language and social action, enables teachers to understand how ‘speakers’ progressive diversification of methods for action .. is inextricably intertwined with their becoming more central participants in the communities in which they interact’ (ibid). This understanding, in turn, may impact teacher cognition and teachers’ classroom practices, and it can thus be instrumental in preparing language teachers for, and implementing, interactional competence-based instruction.

In sum, the teacher training elements we suggest are designed to make visible teachers’ own views of language, to broaden their understanding of interactional practices, and to shift their conceptualization of students’ interactional development towards an understanding of it as a diversification and fine-tuning of methods for action. The materials we suggest thus advance an empirically informed view on interaction, connect research and praxis, provide teachers with the necessary tools to do reflexive work with IC materials, and supplement current methodology textbooks in teacher training.

Acknowledgements

We would like to thank our anonymous reviewers and Samuel Schirm for their insightful and valuable commentary on earlier drafts of this manuscript. All remaining errors are our own.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Thorsten Huth

Thorsten Huth (PhD, University of Kansas) is Assistant Professor of German & Linguistics at the University of Tennessee. His research focuses on pragmatics, interaction, language learning, and language teacher education. He co-authored Deutsch heute and is co-editor of ‘Beyond Grammar: Teaching Interaction in the German Language Classroom’ (2014–2016).

Emma Betz

Emma Betz (PhD, University of Illinois) is Associate Professor of German at the University of Waterloo. Her interests are in conversation analysis and language teaching. She is the author of ‘Grammar and Interaction: Pivots in German Conversation’ (2008) and co-editor of ‘Beyond Grammar: Teaching Interaction in the German Language Classroom’ (2014–2016).

Carmen Taleghani-Nikazm

Carmen Taleghani-Nikazm (PhD, University of Texas-Austin) is Associate Professor of Applied Linguistics at Ohio State University. Her research focuses on language use in social interaction and on cross-linguistic and cross-cultural issues, and language teacher training. She is the author of ‘Request sequences: Intersection of grammar, interaction, and social context’ (2006).

Notes

1. For example the goals laid out in the framework of the German ‘Quality offensive (in) teacher training’ (‘Qualitätsoffensive Lehrerbildung’): https://www.qualitaetsoffensive-lehrerbildung.de/.

2. ‘The Leibniz-principle of reflected action is characterized by three main aspects: Practice-relevant theorizing, cooperative professionalization, and reflexivity’ [‘Das Leibniz-Prinzip der Reflektierten Handlungsfähigkeit lässt sich durch drei wesentliche Merkmale charakterisieren: praxisbezogene Theoriebildung, kooperative Professionalisierung und Reflexivität’]. It was developed within the project Theoria cum praxi at the Leibniz Universität Hannover, Germany. See www.leibniz-prinzip.uni-hannover.de/leibniz-prinzip_massnahmen.html.

3. Participants were mainly graduate students (MA and PhD) in languages/linguistics, but the workshop also included advanced undergraduates (BA) interested in language teaching. Participants’ practical classroom experience varied: Some had no prior teaching experience at all, but most of the graduate students had completed or were in the process of completing a cross-departmental Certificate in University Language Teaching at the University of Waterloo (https://uwaterloo.ca/centre-for-teaching-excellence/support-graduate-students/certificate-university-language-teaching) and were themselves teachers or teaching assistants of language courses.

4. Sources of data for such exercises can be CA and IL textbooks, e.g. Couper-Kuhlen and Selting (2018), Sidnell (2010), Wong and Waring (2010), and online corpora that make available recordings of telephone and face-to-face interaction to researchers and/or teachers. Examples of the latter are the publicly accessible FOLK corpus (http://dgd.ids-mannheim.de) housed at the Institut für Deutsche Sprache/Institute for the German Language in Mannheim, Germany; the CLAPI corpus of French interaction at the ICAR Research Lab of the Université de Lyon 2 (http://clapi.ish-lyon.cnrs.fr/V3_Accueil.php?interface_langue = EN); and the CallHome and CallFriend telephone corpora at TalkBank (Carnegie Mellon University, The University of Pennsylvania, coordinated by Brian MacWhinney), which provide English, French, German, Japanese, Spanish, Arabic, and Chinese conversation data for use in research (https://ca.talkbank.org/). Corpora of short written interactions (e.g. text messaging, WhatsApp messages) for research and teaching are part of the Mobile Communication Database 2 (https://db.mocoda2.de/#/c/home#/c/home), which is housed at the Universität Duisburg-Essen in Germany. There are also useful repositories of materials and suggestions for teaching (the analysis of) conversation: ROLSI’s blog on CA Teaching (https://rolsi.net/teaching-2/) includes clips from TV shows which are suitable for introducing and illustrating basic talk-in-interaction concepts; the Universität Münster in Germany houses Gesprochenes Deutsch für die Auslandsgermanistik (Spoken German for German Studies Abroad) (http://audiolabor.uni-muenster.de/daf/), which offers recordings and transcripts of German interaction for use in teaching abroad and collects materials for teaching interaction in German as a Foreign Language (DaF) classes.

5. This also addresses one important component within the framework of the German ‘Quality offensive (in) teacher training’: It heeds the call for a better integration of research and subject knowledge with pedagogical knowledge and teaching praxis in training teachers.

6. Roever and Kasper (2018) specifically discuss the relevance of such insights in testing speaking.