Co-adaptation processes in plenary teacher-student talk and the development of L2 interactional competence

ABSTRACT

This paper shows how a language teacher and her students observably recalibrate their interactional behaviour over time. With respect to the students, this recalibration can be interpreted as increased L2 interactional competence. The participants’ turn designs indicate that they are contingent upon: 1) the local interactional context; 2) the students’ changing linguistic and interactional competence; and 3) the teacher’s evolving pedagogical goals. The phenomenon is examined in oral collective reconstructions of reading texts in plenary teacher-student talk. The data stems from a longitudinal corpus of video-recorded classroom interaction, distributed over 5 schoolyears of instructed L3 French. The analysis focuses on the way the teacher’s questions, the students’ responses and the teacher’s subsequent third turns are formatted with regard to the use of linguistic resources afforded by the reading texts. At the pre-intermediate level, the teacher invites and reinforces the use of afforded linguistic elements, while the students’ answer turns are characterized by appropriations of linguistic resources from the teacher’s talk or the reading text. Towards the intermediate level, the teacher’s turns increasingly reflect a changing pedagogical agenda allowing for more learner agency, while the students’ turns exhibit a more fine-tuned recipient design despite fewer afforded elements.

KEYWORDS:

• Interactional competence
• conversation analysis
• affordances
• co-adaptation

1. Introduction

This paper¹ aims to contribute to a growing body of research exploring the development of L2 interactional competence from a conversation analytic research perspective (Cekaite 2017; Cekaite 2007; Dings 2014; Farina, Pochon-Berger, and Pekarek Doehler 2012; Hall, Hellermann and Pekarek Doehler 2011; Hellermann 2007; Hellermann and Lee 2014; Nguyen 2012; Pekarek Doehler and Berger 2018; Pekarek Doehler and Pochon-Berger 2015; Watanabe 2017; Young and Miller 2004; Hall 2018; Pekarek Doehler 2018). L2 interactional competence is defined as the ability to participate in [L2] discursive practices (cf. Young and Miller 2004), which comprises a set of skills or methods for designing and accomplishing actions within such discursive practices (cf. Cekaite 2007; Cekaite 2017; Hall, Hellermann and Pekarek Doehler 2011; Mondada and Pekarek Doehler 2004; Pekarek Doehler and Pochon-Berger 2015). Some researchers in the field prefer to conceptualise the development of L2 interactional competence as the broadening of an individual’s interactional repertoire (cf. Cekaite 2017; Hall 2018), while using the term ‘interactional competence’ to refer to a more basic concept acquired through socialization in childhood (Hall 2018).

In this paper, ‘L2 interactional competence’ is used to refer to the result of L2 learning rather than to an underlying resource needed to achieve it. Previous research has shown that increasing interactional competence is characterized by a diversification of linguistic resources and interactional techniques to accomplish actions and by a heightened ability to precision-time and recipient-design turns (Pekarek Doehler and Pochon-Berger 2015). Importantly, interactional competence in this sense is a dynamic, co-constructed and processual notion, tying in with a view of competence not as essentially inherent to an individual, but rather thought of as action-based and socially situated (cf. Mondada and Pekarek Doehler 2006; Pekarek Doehler 2018). In other words, interactional competence is seen as the ‘ability of joint action’ (Pekarek Doehler 2018) that is observable in participants’ conduct. As the term ‘joint action’ implies, such conduct is not a stand-alone enterprise but is necessarily connected with that of other participants.

As a consequence, the development of L2 interactional competence over time is equally inextricably linked with the co-participants’ behaviour in the interaction (Dings 2014; He and Young 1998; Hellermann and Lee 2014; Pekarek Doehler 2018; Seedhouse 2005). In fact, several studies have found co-adaptation and socialization processes to be constitutive for the development of interactional competence of language learners, both in dyadic student talk (Hellermann 2007) and in teacher-student talk (Cekaite 2007; Young and Miller 2004). The present study adds to these findings by focusing on all participants’ (i.e. teachers’ and learners’) co-adaptive conduct in teacher-student talk, by showing how their mutually dependent actions are linked to linguistic competence as displayed by students and oriented to by teachers, and how they change over time. In this way, the paper attempts to ‘document the systematicity of the change’ (Wagner et al. 2018: 21) in the accomplishment of a particular interactional practice from one schoolyear to the next.

The mutual adaptation of students and teachers to each other’s conduct is observably informed by at least three factors: from the learners’ perspective, the shapes of their turns-at-talk depend on both the linguistic resources that are available to them at the moment of turn production and on the contingencies of the local interactional environment. In the data studied for this paper, this local interactional environment is heavily influenced and structured by the teacher. The discursive activity under investigation is teacher-student talk situated within the ‘IRF continuum’ (Van Lier 2000a, 94) and therefore characterized by an inequality in discursive power in which ‘teachers […] own the question and comment turns’ (Markee and Kasper 2004, 492). In other words, it is the teacher’s privilege and responsibility to set the agenda for any stretch of discourse, whereby setting the agenda is a crucial part of the work of teaching. Part of this work consists of assessing the students’ competence and of adapting the teacher’s conduct accordingly, e.g. by designing turns in a particular way, especially questions (cf. Koskela and Arminen 2012), but also third turns (Margutti and Drew 2014; Daşkın 2015). Teachers ‘do’ teaching by designing their turns in order to open up and to allow for varying sizes of interactional space in which the students can produce their own turns, a phenomenon which has been studied extensively from the perspective of classroom interactional competence (= CIC, e.g. Walsh 2012; Sert 2015; Sert 2017). For the analysis of my data this means that a teacher’s turn shapes are not only informed by their own available linguistic resources and by local interactional contingencies but also by their teaching goals. The longitudinal data explored in this study allows the addition of a further analytic layer to studies of interactional competence by investigating changes in their co-adaptive conduct over time. It enables a description of the systematic differences in the ways in which the teacher shapes her turns in the same type of classroom interaction over the course of several years, teaching the same class of students at increasing levels of competence.

Both teacher and student turns are subject to co-adaptation processes emerging from the interplay of the three factors ‘interactional contingencies’, ‘available linguistic resources’ and ‘doing teaching’. But – in a perspective slightly different from the above-mentioned research on CIC – this paper analyses the teacher’s changing conduct not as a change of her classroom interactional competence (i.e. not as a process from less competent to more competent) but as the manifestation of her co-adaptive behaviour contingent upon the students’ development and her teaching goals. However, the students’ involvement in this mutually adaptive conduct is described as a manifestation of their changing interactional competence (cf. Wagner et al. 2018: 18 on the different interpretations of changes in practices in longitudinal studies).

2. Data

The data for this study stems from a longitudinal video corpus of French as a foreign language classes in an Austrian secondary school, where a class of 25 students attended 6 years of instruction in French. From this corpus, 24 different occasions of one recurring classroom activity in one class were selected, spanning a period of 3,5 years and covering 5 schoolyears, stretching from the second half of schoolyear 2 (occurrences in May and June) to the first half of schoolyear 6 (occurrences in November) of the students’ learning trajectory:

Table

	Autumn	Winter	Spring
2nd Year of French			4 occurrences
3rd Year of French	3 occurrences	3 occurrences	1 occurrence
4th Year of French	4 occurrences	4 occurrences
5th Year of French	2 occurrences
6th Year of French	3 occurrences

During their six years of instruction, Austrian secondary school students of French typically progress to approximately level B1 of the CEFR scales of foreign language proficiency (Trim, North, and Coste 2001) regarding the ‘Speaking’ skill. The present data set starts at the end of Year 2 of this timespan, when students have usually reached roughly level A2. This study investigates one particular type of interaction which recurs throughout the data at all levels of proficiency. The type of interaction under investigation comes about as part of a discursive teaching/learning activity in which the students are led to talk about a text they had previously read in class, ranging from short texts in the course book in lower level lessons to whole books in higher level classes. The teacher herself uses the phrase ‘talking about the text’ to refer to this activity; however, this label may be misleading for the reader, since ‘talking’ does not mean spontaneous conversation here. Rather, the interaction is teacher-led, highly structured and largely organized in IRF-sequences (Sinclair and Coulthard 1975; Mehan 1979), topically driven forward and linguistically scaffolded (Koole and Elbers 2014; Wood, Bruner, and Ross 1976) by the teacher, who elicits answers from the students. In a tightly-knit sequence of questions, answers and feedback turns, the teacher guides her students to jointly build a near-exact reconstruction of the text’s content or plot (depending on the nature of the text) by using language from the text and maintaining the order in which the content or plot elements are presented in the text.

3. Findings

Looking at the classroom activity ‘talking about a text’ through the lens of the ‘task-in-process’ (Seedhouse 2005), the teacher appears to be pursuing multiple pedagogical goals. The most obvious aim is arguably to check reading comprehension, since the teacher’s questions elicit information and knowledge about the text from the students. Another goal seems to be language production practice. Besides lauding the students’ speaking competence and thus expressing an orientation towards considering the activity as a speaking exercise, the pedagogic goal of practising spoken language also manifests itself in the way the teacher formats her questions and her third turns. In both kinds of turns, two elements point to this orientation: first, a clear focus on the use of specific language material, especially such as provided by the text itself (e.g. specific words, collocations or particular grammatical constructions) and second, specific repair work by the teacher regarding the linguistic correctness or appropriateness of student utterances. Finally, ‘talking about a text’ is among those activities in my corpus where students most often produce relatively extended turns, particularly in the early years.

One factor that observably influences participants’ turn designs in my data is the amount and kind of linguistic resources available to them. Availability here refers to two different aspects: 1) participants’ linguistic knowledge, i.e. which words or grammatical forms of the target language they know and can use, and 2) what level of access participants have to the linguistic material of the text at the time of producing a turn (e.g. if the course book is open or closed, how much text there was to read, etc.). Over the course of their five years of learning French, as the students’ linguistic competence changes (and generally increases), so does the teacher’s turn design in both questions and third turns. This change in the teacher’s recipient design can be described as adaptive behaviour not only to react to local interactional or linguistic needs the student may have but also as a strategy in order to achieve her pedagogical efforts, i.e. the institutional goal of teaching the students L2 French (cf. Seedhouse 2005, 184–85). On the students’ side, adaptive behaviour can also be observed; in their case, this is considered as development of their interactional competence which is linked to increased linguistic competence.

One observable yardstick to trace and describe the development of this linguistic competence seems to be the changing degree of what I call the ‘affordedness’ of both students’ and teacher’s turns-at-talk, i.e. to what degree the language of the turn is afforded by previous talk or teaching material (cf. Eskildsen 2008, 57). Affordances in this sense are opportunities for incorporating contextual features into one’s own talk. The term itself is rooted in psychology but was adapted by Van Lier (2000b, 2004) for an ecological perspective on classroom interaction. Affordances are conceptualized as emergent features of (interactional) environments that create spaces for participation and language use (cf. Thoms 2014). Importantly, in this perspective, the learners themselves determine if and how they want to make use of an affordance available to them, so it aligns well with the CA view that learners are considered agents rather than passive recipients in the learning process (cf. Firth and Wagner 1997). In my data, one striking feature of the talk is the remarkable amount of language taken verbatim from the textbook in the early years of learning – both by the students and the teacher – and the noticeable decline of this phenomenon over the course of the students’ learning trajectory. This general observation about the affordedness of participants’ talk forms the background against which the analysis of development is set.

In the following three sections (3.1–3.3), four representative excerpts from the second (1 extract), fourth (2 extracts) and fifth year of learning (1 extract) will be discussed to illustrate the students’ and teacher’s evolving co-adaptive behaviour.

3.1. Years 2&3: targets, triggers and building bricks

The first data extract shown below is representative for the activity type ‘talking about a text’ in the lower level data (pre-intermediate, approximately A2 according to the CEFR), featuring abundantly in Year 2 and 3. The extract is taken from the end of the second year of French (= spring of Year 2). Just before the sequence in Extract 1, the class had read a text in the schoolbook A plus! (Bächle et al. 2013) called ‘Une catastrophe’ (‘A catastrophe’), which tells the story of a family who are watching TV and see the news of an oil spill that has happened off the coast of Brittany. After reading the text, the teacher asks the class a question about the text:

Extract 1. (2nd Year) (book open) ‘sa profession’.

46 T   (...)       quelle était sa profession. maintenant

            what was his profession now

47    il est trop vieux je crois, (.) mais quelle était sa

     he is too old I believe,     but what was his

48    profession noelle.

    profession noelle

49 NOE *sa profession a été pê**cheur.

    his profession was fisherman

noe  *gaze into book    **gaze to T ––>

50 T    oui. ou il a été pêcheur¿ trè:s bien. mm↑hm.*

   yes. or he was a fisherman, very good.

noe                          – –>*

The teacher’s first turn (ll. 46–48) is made up of two questions that are lexically, syntactically and prosodically identical, separated only by an aside which provides additional contextual information to the question, but is not taken up in any way by the student. This pattern appears frequently in the 2nd and 3rd year data, as does the fact that the production of the second question is not due to a lack of uptake. In fact, regularly in the data, the second question follows immediately upon the first. It is in that sense not a ‘response pursuit’ (Pomerantz 1984; Hosoda and Aline 2013). Rather, its perspective is forward-oriented and proactive. This practice has been labelled in other studies as ‘same-turn multiple question’ (Kasper 2006; Kasper and Ross 2007) and identified as typical for institutional contexts (e.g. doctor-patient interactions and interviews), especially if one participant is (seen as) less proficient in the language (Gardner 2004) and ‘in fragile environments where relevant answers may be difficult to generate’ (Kasper and Ross 2007). Same-turn multiple questions also seem to be more frequent in topic-shift environments (cf. Kasper and Ross 2007). Indeed, the interaction from which Extract 1 has been taken is characterized by all of these features: less proficient speakers are the recipients of the question, the topic development trajectory is disjunctive and it is a ‘fragile’ environment in the sense that it is by no means certain that the students will be able to produce an answer to the teacher’s question.

The consequences of the teacher’s same turn repeat of her question are two-fold: first, this practice extends the teacher’s question turn and so gives the students more time to understand its propositional content, to plan their answer and to decide if they want to bid for a turn. So, the teacher’s strategy is a pre-emptive one (cf. Svennevig 2013), anticipating potential non-uptake. In fact, my data shows that students frequently start raising their hands during or towards the end of the repeat question. In this way, the gap between the teacher’s and the student’s turn is kept very short. Second, the teacher’s same turn repeat constitutes reinforced language input and in that sense is doing language teaching work – by providing an opportunity for the students to listen to and observe appropriate chunks of the target language being used, chunks that the teacher has selected and that she reinforces by repeating them.

A further aspect that is typical of the 2nd and 3rd year data is the narrow scope of the questions, especially those initiating a new sequence. Typically, the teacher designs her first question turn in such a way that it targets a very specific element in the reading text, in terms of syntax as well as content. Here, the format is that of a WH-question with an extremely narrow focus: quelle était sa profession (‘what was his profession’), while the corresponding sentence in the book reads ‘Yvon Leroy a été pêcheur’ (‘Yvon Leroy was a fisherman’).

The basic pattern throughout my 2nd year data is as follows: a concrete sentence or part of a sentence from the text gets turned into an interrogative, often using exactly the same lexical material. This particular manner of designing the question turn has two immediate consequences for the interaction: 1) it serves as a trigger for the students who can use it to identify the passage or sentence that it targets. They can then lift linguistic building bricks for the construction of their answer turn from the text in the vicinity of the trigger; 2) the turn can itself provide the necessary structure for the students to design their turn by using a ‘format-tying technique’ (Harness Goodwin 1983). In this way, the way the question is phrased provides affordances for the learner.

Looking now at the student’s answer turn (l. 49), we can see that it is in fact doubly afforded: first, Noelle uses a format-tying technique by turning the teacher’s interrogative into a declarative utterance starting with sa profession (‘his profession’). Second, she adds a prefabricated chunk consisting of copula verb + complement which she has taken verbatim from the reading text: a été pêcheur (‘was a fisherman’).

Extract 1 shows an IRF-sequence whose structure is markedly characterised by the teacher’s question design but also by the way she formats her third turns. The latter constitutes a crucial interactional moment that brings to the fore how the teacher has interpreted the sequence so far and how it fits into the larger context of the activity (cf. Lee 2007; Margutti and Drew 2014; Park 2014). The privilege of occupying the third turn by default also puts the teacher in a position from which they can determine the interactional space given to the students, both in terms of actual talking time and options for continuing or re-entering the talk (Walsh 2012). In this extract, the teacher’s 3rd turn produces a positive assessment of the student’s answer, evidenced by the turn-initial receipt token oui with falling intonation and the explicit evaluation très bien. (l. 50). Positive assessments of student answers are a typical format for closing a sequence and thereby limit the interactional options of the student who had just been speaking (cf. Waring 2008). In this case, for instance, the teacher’s evaluation is followed by a next question (l. 51, not printed here), which is not directed at Noelle anymore but at the whole class again.

However, the teacher’s evaluative comment does not repeat the student’s turn verbatim, as is often the case in positive assessment third turns of IRF sequences in my data (cf. also Hellermann 2003, 88). Instead, she presents an alternative version of the answer which she explicitly offers as being a candidate (rather than replacement) by introducing it with or at l. 50. She shapes Noelle’s turn by using only slightly different linguistic material, whereby the pitch contour contrasts with that of the student’s turn and prosodically highlights the alternative element in her version. In this way, the teacher’s turn not only closes the sequence but also does ‘language teaching work’ (Lee 2007) and more specifically, vocabulary teaching work, by ‘highlighting what is acceptable and what needs to be learned’ (Waring 2008, 589), i.e. by evaluating the student’s version and offering her own alternative, given from her position of epistemic authority as the teacher but also as the more competent French speaker.

We can also see in Extract 1 how the teacher is doing teaching because she orients to the fact that this is not genuinely dyadic but actually multi-party talk (cf. Schwab 2011), indexicalised by her presenting an alternative version (l. 50), hearably produced for ‘ratified overhearers’ (Goffman 1981). By doing so immediately in the ongoing interaction (rather than, for instance, taking notes and talking to the student in question later), she makes this new version available not only to the dyadic interaction participant, but to the whole class.

To summarise, in the data of Year 2 and 3, as illustrated here by an excerpt from year 2, teacher-student interactions are characterised by the teacher designing her first turns typically in the shape of same turn multiple questions and providing multi-layered affordances through targeting specific elements of the text and indicating them through linguistic triggers, while the targeted textual elements in turn offer building bricks for the students’ own turn design. The students respond to this conduct in kind, by frequently visually scanning the text while the teacher is formulating her question and, before they begin their turn production, by making heavy use of format-tying strategies and lifting whole chunks (‘building bricks’) from the reading texts. In her third turns, the teacher treats the activity as largely form-focused by producing assessments of the students’ speaking performance and by repeating linguistic structures for the benefit of the whole class. The teacher’s turns in this position are usually sequence-closing thirds followed by a re-orientation away from current and towards potential next speakers and do not allow for further sequential expansions by the students.

3.2. Year 4: changing question types, ‘débrouillardise’ and turn-shaping

Studying the data from the middle of the six-year-learning-trajectory (autumn of Year 4), we can see a change in terms of the affordances provided by the teacher’s questions and by the reading text, which has consequences for the turn design work the students have to accomplish. Two excerpts are shown to illustrate the nature of this change.

Prior to the sequence shown in Extract 4 below (from the autumn of the student’s 4th year), the class had read a text with the title ‘Changer la vie’ in their coursebook A plus! (Gregor, Jorißen, and Schenk 2015). Compared with the text from the second year of French, this one is only marginally longer. However, its language is considerably more complex, in particular regarding sentence structure. In addition, the text type is different from texts in Year 2 and 3: it takes the form of a transcript of an interview with a group of teenagers who report on their attitudes towards a topic which touches on the philosophical: what they would like to change about their way of life. The facts that the text is written in direct speech, that its parts are not linked by chronology but by ideas and that its content veers towards the abstract are likely to present a challenge to the students when it comes to formulating their turns since they might need to invest more effort to process and to reproduce the content. Moreover, the way the teacher asks her questions is noticeably different from her practices in the second year.

Extract 2. (4th Year) (book open) ‘clémentine’.

01 T  ok. ↑Alors mira. [qu’est]-ce que ces jeunes-là aimeraient changer.

      so mira  what would these young people like to change

02 MIR         [oui]

           yes

03 T  *tu parles [peut-être de clémentine.]°en[premier lieu.°]

   you talk maybe about clementine first of all

04 MIR   [ahm (xxx)] [clémentine ]veut changer

                           clementine wants to change

mir   *gaze into book – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – >

05 MIR ehm (0,7) que: elle veut que (.) ses parents étaient >un peu*

       that she wants that her parents were a bit

mir – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – >*

06 MIR *plus sévères avec elle,<

    stricter   with her

mir  *gaze to T

07   (0,5)

[...((13 lines omitted))]

21 T  alors tu continues. on corrige après. m↑hm

   so you continue we’ll correct later

22 MIR *ahm

mir  *gaze into book – – >

23 T  [qu’est-ce qu’elle veut encore]

   what else does she want

24 MIR [e:t que]ses pare:nts travaillent *(.) MOINS

   and that   her parents work   less

mir – – – – – – – – – – – – –book – – – – – – – – – – – – – – – – – – -> *

25    (1,0) (yes.) (0,4)

26 T   pour faire quoi,

      to do what

27 MIR  *pour ahm

     to uhm

mir   *gaze into book – – – – –>

28    (1,4)

mir – – – >

31 T   elle veut que ses parents travaillent moins pour pouvoir

   she wants her parents to work less to be able to

32 MIR  pour pouvoir *euh temps (0,4) pour elle (0,8)

    to be able to uh time for her

mir – – –book –– > *gaze to T

33 T   pour pouvoir passer

    to be able to spend

34 MIR  passe:r *du temps [(xx)

     spend time

mir      *leans back

35 T            [avec elle, voilà, elisabeth.

          with her, that’s it, elisabeth

In the above example from the students’ fourth year of French, the scope of the teacher’s question is much broader than the scope of the questions we have seen in the second year. In fact, the interrogative in the teacher’s first turn (l. 1) basically targets the reading text as a whole. It is followed (in l. 3) by a specific suggestion, hearable as a request, concerning which particular interviewee from the text Mira should choose to talk about. Although the teacher’s specification narrows down the number of relevant reading passages, these still need to be found in the text, and there is no obvious thematic order such as there would be in a narrative text. It is therefore the student’s task to select an appropriate aspect from memory or by looking at the text again. This presents a new challenge in the fourth-year data while also constituting a new level of agency the students are permitted – and expected – to perform.

The second aspect observable in the teacher’s turn is that it does not offer much in terms of a useful affordance. Nevertheless, Mira goes down this familiar route in her first attempt. Her method for managing the turn design task at hand is to begin with a format tying strategy: she turns the teacher’s interrogative into a corresponding declarative utterance: clémentine veut changer (‘clémentine wants to change’) (l. 4). This part of Mira’s turn is produced in overlap with the second part of the teacher’s turn, while she is still searching the text for the answer. But after this smooth start, there is a gap (0,7 sec, l. 5) while Mira scans the text. She then produces a restart and reformulates her turn in a way that is a) grammatically easier to construct (elle veut changer ‘she wants to change’, l. 5, projects a potentially complex noun phrase and a need to nominalise verbal elements from the text whereas elle veut que ‘she wants that’ requires a clause) and b) better suited to match the language material she can lift directly from the text. In order to illustrate this, here is the relevant passage from the reading text:

Clémentine: Bien sûr, il y a plein de choses que je voudrais changer. D’accord, nos relations avec les adultes sont assez bonnes, mais si mes parents étaient un peu plus sévères, je n’aurais rien contre. Et si je ne devais pas les rassurer tout le temps, ce ne serait pas plus mal. Ce serait mieux s’ils avaient plus de temps pour moi, s’ils travaillaient moins. (Gregor, Jorißen, and Schenk 2015)

The combination ‘structure of the reading text’ + ‘scope and syntactical format of the teacher’s question’ means that it is not possible for the student to produce an adequate and correct answer purely with material directly lifted from the text. In particular, reporting talk is a new requirement that demands a more advanced linguistic competence than being able to identify appropriate chunks from the text. But Mira meets the challenge. First, she changes her turn construction strategy mid-turn and recalibrates her projected sentence structure. Next, while still taking some of the material for her turn from the reading text (the whole chunk parents étaient un peu plus sévères ‘parents were a bit stricter’, ll. 5–6), she also produces several adaptations: she introduces a reporting verb (elle veut que ‘she wants that’, l. 5), she changes the possessive pronoun mes ‘my’ into ses ‘her’ (l. 5), and she adds her own incremental element at the end of the clause: the prepositional phrase avec elle ‘with her’ (l. 6). inally, in terms of prosody, Mira combines the afforded elements relatively smoothly with her own material and her adaptations, so that the bricolage character of her turn is barely noticeable.

In line 24, Mira continues her response to the teacher’s question after a side-sequence (not printed) in which the teacher unsuccessfully tries to other-initiate self-repair and to incite Mira to correct the ungrammatical verb form she used (the indicative instead of the subjunctive form). The participants’ re-orientation to the main sequence causes a partial overlap of Mira’s continuation with the teacher’s encouragement to carry on talking about Clémentine (ll. 23–24). Interestingly, the teacher’s reformulating of her initial question (qu’est-ce qu’elle veut encore ‘what else does she want’, l. 25) does more than simply repeating her own talk; it also takes into account how the student has transformed the original question in the meantime: by using the same verb as the student had used (veut ‘want’ as a full verb) and by shaping her turn as the request for a continuation of an enumeration (que…encore ‘what else’). This stretch of talk thus displays further evidence for the participants’ ongoing co-adaptation, this time on the teacher’s side.²

While the teacher’s very first question (l. 1) provided something that evidently looked like a trigger to Mira but was not initially helpful for her turn production, the teacher’s next question does not contain any such hints that might be linked to the text: it is a follow-up question that links back to Mira’s own turn and asks for further information related to what she just said. Syntactically and lexically it is closely connected to Mira’s turn but provides no clues as to what element from the text might fit in response nor a framework in which to insert such an element. Still, Mira seems to rely on that being the case: she attempts to format-tie her turn (pour ahm ‘to ahm’, l. 27) while she looks at the text to find an appropriate element to fill the syntactic slot she has created for herself. This slot, however, cannot be filled so easily, since the order of the argumentative structure of the text is diametrically opposed to the order in which the respective elements have been introduced in the interaction. s’ils travaillaient moins (‘if they worked less’) is the last part of the paragraph about Clémentine in the reading text, so Mira has to go further back in the text to find out why Clémentine would like her parents to work less. However, the link is not made explicit grammatically, so Mira has to infer. This explains the 1,4 seconds silence in line 28. The teacher orients to this silence as displaying a potential problem by offering a candidate turn beginning in the shape of a designedly incomplete utterance (DIU, cf. Koshik 2002) (l., 31). Although this gives Mira more time during which to find a solution to complete the teacher’s turn, it also greatly limits her options for turn construction. The difficulty for Mira is increased by the teacher’s addition of a modal verb to her suggested turn beginning. In order to be grammatically fitting, Mira would have to complete this construction with a verb in the infinitive. But she ignores this projection in her completion of the DIU and (incorrectly) uses the modal verb as if it were a full verb.

All in all, the second part of this sequence, after the repair-initiation attempt, is very much steered by the teacher who, through her follow-up questions and DIUs, narrows down the syntactic and topical possibilities for turn construction. Her turns clearly direct the students towards particular linguistic options, and they implicitly and explicitly suggest specific chunks of language. In this way, the teacher, in a step-by-step fashion, provides almost all of the language the student uses. Yet, the task is still complex for the student: the multilevel activity of reading and understanding the text and producing contingent turns at the same time is observably difficult.

The next excerpt, Extract 3, is taken from a little later in the same interaction as Extract 2 and begins with the teacher selecting a new student – Carolina – and requesting her to carry on talking about the text:

Extract 3. (4th Year) (book open) ‘qu’est-ce que cela apporterait?’.

01 L   ↑super. ↑très ↑bien. très bien. ↑carolina. tu continues?

         very good very good are you continuing

02 CAR  äh oui. (0,3) (goui:)/gʊjom/

      yes

03 L   m↑hm, [/gijom//gijom/

04 CAR      [ähm äh guillaume

05 L   des u (x) (0,4) sag’ ma net guillaume.

    that u we don’t say

06 CAR guillaume äh (.) dit (0,2) qu’il (.) veut vivre un peu dans

    guillaume says that he wants to live a bit in

07   un(e) autre pays.

    another ((DET-f)) country

08 T  oui:,

     yes

09 CAR °un autre pays°

    another ((DET-m)) country

10 T  un peu de temps dans un autre pays¿

    a bit of time in another country

11 CAR °oui°

    yes

12 T   ↑pourquoi, qu’est-ce que cela apporterait?* s’il faisait cela,*

    why    what would this bring if he did that

         ‘what would the benefit of this be’

 car               *moves hand slightly*

13 CAR e*hm [parce que:* le:s cultures différents, (.) s’intéressent, =

       because the different cultures REF interest

 car   *raises hand *

14 T      [mhm

15 CAR =(0,4) °beaucoup°?

        a lot

16 T   oui: l’intéressent [beaucoup,

    yes interest him a lot

17 CAR         [ah l’intéressent [beaucoup.

                  interest him a lot

18 T                   [et on pourrait peut-ê:tre,

                    and you could perhaps

19    (0,2) regarder les choses d’une autre (.)

   see DET things from another

20    perspective. [(.)]si on vit dans un autre pays. oui:. très bien.

   perspective if you live in another country yes very good

21 CAR      *[°oui°]*

      yes

car      *nods *

Carolina duly chooses a character from the reading text and at first struggles to pronounce his name correctly (l. 2), so the teacher assists her (l. 4–5). She then makes a statement about the character’s desires for change, quoting word-for-word from the text. Similar to previous extracts, the teacher’s response is to produce a third-turn repeat. Then, however, she does something quite different: she asks a question whose scope turns out to be considerably larger than the previous one. Her question pourquoi, qu’est-ce que cela apporterait? s’il faisait cela, ‘why, what would the benefit of this be? if he did that’ (l. 12) follows on from Carolina’s turn but goes beyond the information given in the text. Rather than asking for a straightforward reproduction of content, it requires Carolina to add her own thoughts. It is also a multi-unit question, whereby the second part narrows the scope of the first question, or rather specifies it, in a forward-oriented and preemptive technique to increase the chances of an adequate answer (Svennevig 2013) while additionally projecting a preferred syntactic format for the answer.

Questions whose scope goes beyond the text are a new feature in the 4th year data and do not appear in earlier ‘talking about the text’ activities. In these types of questions, as in the case of Extract 3, the propositional content for the answer cannot be found directly in the text and has to be inferred, interpreted or even thought up by the students themselves; at the same time, there is no or hardly any relevant language material readily available in the reading text to formulate that content. In the case of Extract 3 the teacher’s turn additionally projects a particular syntactical format for the answer, namely a hypothetical clause, and thus puts constraints on the possible linguistic choices for the answer.

This is clearly a challenge for the student, but Carolina readily accepts it: in fact, she alerts the teacher to her wish to keep the floor quite early on, well before the teacher’s turn is completed (hand movement in l. 12). However, in her turn design she does not align with the normative projection of the second part of the teacher’s turn; that is, she does not produce a conditional clause, but instead selectively responds to the broader pourquoi that the teacher had first uttered. In response to that pourquoi, Carolina produces a type-conforming answer starting with parce que (l. 13), ignoring the teacher’s narrowing down of the focus of her question. One could interpret this as a lack of sensitivity towards the grammatical projections of a turn, but one could also say that this is an example of the student’s growing ‘débrouillardise’ (Farina, Pochon-Berger, and Pekarek Doehler 2012), i.e. their coping ability, where they resort to spontaneous ‘make-do solutions’ (Larsen-Freeman 2010): rather than being thrown or even silenced by the teacher’s demanding question, Carolina does produce an answer, but only to that part of the teacher’s question and in those words that she is capable of. Evidently, this strategy is successful, since the teacher ratifies the answer (l. 18). Rather than sanctioning it as inappropriate or lacking, she grants Carolina some interactional space here. She does then, however, also use the opportunity provided by her third turn to do some language teaching work by reformulating (or ‘shaping’, cf. Walsh 2012) Carolina’s contribution (ll. 18–20). In this way, the teacher also orients to one aspect of her initial pedagogical agenda as displayed in the formatting of her question turn, namely, that the answer should include a hypothetical clause. But, importantly, she casts her own version of the answer not as preferable over Carolina’s but as an additional option (et on pourrait peut-être), thus constructing an epistemic stance that grants the student more agency and more extensive epistemic rights than observable in data from lower level learners.

Around the middle of the students’ learning trajectory, in the first half of Year 4, teachers’ questions have changed considerably from Year 2 and 3. First, their turn design no longer targets particular syntactic structures or even specific locations in the reading text. Sometimes, questions even go beyond what is explicitly mentioned in the text. This requires the students to do more work to find the answer, both in terms of necessary re-readings of a text passage and of cognitive efforts related to reading comprehension. Second, their questions furnish fewer affordances for students’ turn production; students have to make greater efforts to produce a turn, by using format-tying techniques, through the on-line recalibration of thought processes and linguistic resources, visible in restarts and reformulations. Students still lift building bricks from the text but with observable adaptations to incorporate them in their own turn designs. The teacher’s third turns reflect the students’ growing capacity to deal with increased interactional and textual complexity by being more selective when it comes to repair initiation of linguistic errors and by proffering linguistic material as suggestions to students for their own use. Moreover, the third turns are no longer dominated by a purely evaluative aspect and include follow-up questions, thereby frequently creating longer sequences with the same student speaker. In sequences with questions whose scope goes beyond the text itself, the teacher may add her own interpretation in a third turn, presenting it as a further, or alternative, view, not necessarily preferential to the student’s.

3.3. Years 5 & 6: towards a focus on meaning

The last example is from the 5th year, showing an interaction where the students are being asked questions about a whole book they are reading in class, the 228-page novel Le voyage d’Hector ou la recherche du Bonheur by François Lelord. At the time at which the interaction takes place, they had already read the first five chapters. In Extract 4 below, the teacher asks a general question about the main character.

Extract 4. (5th Year) ‘personnage central’.

40 T   ahmmm alors le perschonna- personnage central noelle?

       so the main character

41   (0,4)

42 NOE ahm le *personnage central**c’est hector,**e:t ahm =

     the main character that’s hector and

noe      *gazes at notes **gazes at T **gazes at notes->

43 NOE = dans le (.) dans le premier =

     in the in the first

44    = chapitre ahm(.) on peut lire que (0,3) ahm (0,3)* =

    chapter   you can read that

 noe                         – – >*

45     =*comment, * *(0,5) (son) (apparence) °physique°

    how   (his) (appearance) physical

noe  *gaze to side* *gazes at notes – >

46 T   ah pardon. [comment?]

      sorry how

47 NOE       [(°xxx°)]on peut ahm lire ahm *comment il:

            you can read how he

 noe *gazes up left->

48    *(2,9).* *on p-

          you c-

noe    ->* *gazes at notes-->>

49 T   comment il vit, peut-être?

    how he lives perhaps

50 NOE  °oui.° e:t ahm il a une moustache, et des lunettes, ahm

    yes and he has a moustache and glasses

51    (.) et il a l’air ahm pensif et intellectuel,

      and he looks thoughtful and intellectual

52 T   exactement.*ahm ou est-ce qu’il vit *qu’est-ce qu’on a dit?

    exactly

noe          *gazes at T *

Similar to what we saw in Extract 2, here the teacher produces an invitation to speak and nominates the subject to talk about. This kind of teacher turn requires even more summarizing and selecting work from the student than in Year 4 and as such is typical for the later learner data. The reading text is much too long for the teacher to be able to target one particular sentence or even an individual passage; however, the students have also taken notes on particular topics of the text, for example on descriptions of the main characters. They can refer to these notes during the activity, and Noelle does so too. But first of all, we can observe that she starts her response turn without delay using refined techniques (l. 42): Noelle uses a left-dislocated construction as a turn-entry and tying device. This kind of construction – a noun phrase + co-indexical clitic pronoun, in this case used in the common construction c’est X – is not only a highly effective but also a target-language-specific device to link her answer to the teacher’s turn while at the same time projecting an expansion (cf. Pekarek Doehler and Stoenica 2012), together with the slightly rising intonation at the end of the TCU and the elongated e:t (‘and’) that follows. The student clearly has the means and competence to project an extended turn but, importantly, she is also given the opportunity to do so by the teacher, who does not intervene – a further indication of the co-adaptive development that is going on. Some elements in Noelle’s turn appear to be relatively routinized chunks (dans le premier chapitre, on peut lire que ‘in the first chapter you can read that’) since they are produced faster and more connected relative to the surrounding talk, but they are not taken from the reading text; rather, they are typical of the language used to talk about a book and seem to be linked to that particular discursive practice. These elements are also indicative of the reporting and summarising work higher level learners in my data regularly do in ‘talking about a text’ activities. In ll. 43–44 Noelle actually produces a metalinguistic frame for her answer before she starts properly. It is at the end of a series of routinized chunks, when she starts to formulate bottom-up, that she runs into a word-search problem, starting with more hesitant talk (pauses in ll. 44–45). The trouble becomes so evident (2,9 sec pause in l. 48) that the teacher offers her help (l. 49) by providing a candidate solution, formulated as a suggestion, thus not insisting on the epistemic gradient between them (Heritage 2012). Interactionally speaking, Noelle’s next move (l. 50) is efficient, since she uses the interactional slot to acknowledge the teacher’s suggestion which closes down her word search, but without uptake of the teacher’s suggested language material. Instead, after her initial oui, which confirms the teacher’s candidate turn, she produces the discourse marker et as if she were continuing a previously started utterance (which, syntactically or propositionally, is not actually the case) and then goes on to enumerate a list of traits of Hector’s physical appearance. Despite the non-contingency between the teacher’s and the student’s turns in ll. 49–50, there is no repair-initiation from the teacher. Her turn in line 49 was evidently indeed only a candidate and is not being insisted upon. Instead, the teacher simply closes the sequence by evaluating Noelle’s answer as correct. Noticeably, there is no third-turn repeat nor any ‘heavy’ vocabulary teaching work here, despite the student’s struggle when formulating her turn. In terms of the teacher’s turn construction, both the question design and the third turns allow for a larger interactional space than in lower level classes: the questions are open, extended turns by the students are permitted and even encouraged by giving (linguistic) assistance and by refraining from explicit error correction. Altogether, this is illustrative of a development appearing in the 5th and 6th year data, namely that the teacher’s pedagogical focus has shifted from mainly concentrating on form towards mainly concentrating on meaning.

4. Discussion

The excerpts discussed above have illustrated what my data shows, namely that observable changes in language learners’ interactional behaviour in a specific, recurring classroom activity (i.e. ‘talking about a text’) are closely interrelated with a number of aspects. Among the factors that play a role are, first, the learners’ growing linguistic competence, visible in the progressively reduced necessity to rely on chunks from the text and the increased ability to adapt such chunks for their own turn-design purposes, both prosodically and syntactically. A second factor is the sequential environment in which the students act, and which is determined by the scope and formatting of the teacher’s questions as well as the agenda displayed in the teacher’s third turns. The third factor is the structure, content and length of the reading text which entails varying degrees of reading comprehension work and varying amounts of affordances over the course of the instructional trajectory.

In the beginner/pre-intermediate data, the teacher frequently invites the use of prefabricated elements from the text 1) by organizing her questions in a highly structured way that closely follows the order of the reading/listening text and 2) by using utterance formats that induce very specific conditional relevances such as wh-questions or designedly incomplete utterances (Koshik 2002). Expectably, the students’ answer turns in these sequences are characterised by frequent afforded repetitions (cf. Eskildsen 2008), appropriating the teacher’s utterance and/or making use of chunks of language taken from the reading text. Students’ efforts to attend to local sequential contingencies and render their turns more ‘context-sensitive’ (Pekarek Doehler and Pochon-Berger 2015) are achieved mostly by shaping the prosody of their utterances appropriately (Hellermann 2003). In her third turns, the teacher continues to firmly guide the students in closely following the content and sequential order of the reading/listening text while reinforcing the use of and focus on lexico-grammatical chunks. She also limits their interactional space by intervening without delay and preferring sequence-closing types of turns.

Towards the middle of the students’ learning trajectory, the teacher’s pedagogical agenda evolves to invite more global reconstructions of the reading/listening texts. Her question design tends to allow for increased learner space and learner agency (Waring 2009, 2011; Young and Miller 2004) within the interactions, regarding both the content and formatting of the students’ turns. As the more competent speaker of French, the teacher still has primary epistemic access to the linguistic accuracy and appropriateness of the students’ contributions; however, as far as the propositional content of their turns is concerned, the epistemic imbalance between teacher and student is noticeably reduced. Concurrently, the prefabricated elements in the students’ turns become less evidently afforded, as they increasingly adapt them in view of the local sequential requirements of the interaction. Meanwhile, the teacher’s third turns exhibit a growing emphasis on meaning-oriented turn shapings, suggestions and the provision of alternative language material. Sequentially, her third turns increasingly serve to open up further spaces for the students to expand, rather than closing them down.

In the penultimate year of instruction, the texts providing the basis of ‘talking about the text’ activities change considerably. The class now read whole novels or plays and the teacher’s questions no longer focus on particular paragraphs, let alone individual lines in the text. Instead, her first turns invite the students to summarise parts of the plot, to describe characters and to offer interpretations and analyses of literary themes. Neither her turn design nor the scope of her question generally allow for much direct lifting of material from the text. However, the students do often refer to their notes and retrospectively to class discussions. In their responses, they make use of routinized phrases to introduce and structure their extended turns. They also revert to creative ‘make-do’ solutions (Larsen-Freeman 2011) in situations that demand skills beyond their current linguistic competence, thus indexing their increased interactional repertoire. The teacher provides linguistic assistance when help is requested by the students, for instance in word searches, but in her third turns she generally focuses considerably more on the negotiation of meaning than on linguistic accuracy or on particular linguistic chunks. In terms of sequential management, the teacher refrains from immediate interventions and allows for extended student turns even if there are production problems.

The analysis of the changing conduct of the students and the teacher over time has highlighted a change that is essentially co-constructed and co-dependent. While the students show increased flexibility in adapting their available linguistic resources to the local interactional needs, the teacher provides a kind of scaffolded training ground for them by using different kinds of texts to talk about and by modifying the kinds of questions she asks and the kinds of feedback she gives. The development of students’ interactional competence within these teacher-student interactions is thus at least partly a function of the teacher’s conduct and cannot be analysed independently. This may be a specificity of instructed learning environments where one of the aims of the interaction is to learn to interact. More studies would be needed to investigate how individual teachers and their classes develop together in interaction over time.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Carmen Konzett-Firth

Carmen Konzett-Firth works as a researcher at the University of Innsbruck in the Department of Romance Philology and in the Department of Subject-Specific Education. Her main research interests are classroom interaction in foreign language teaching and the development of the interactional competence of language learners in the classroom.

Notes

1. I would like to thank three anonymous reviewers for their insightful comments on this paper that have greatly helped to improve the original manuscript.

2. I would like to thank one anonymous reviewer for bringing this aspect to my attention.