Analysing appropriations in a professional development program

2.1 Integrating support for students’ development of oral academic discourse skills into subject matter learning

The teacher PDP the present study is based on focused on how to facilitate productive classroom discussions (e. g., Michaels et al. 2013; Murphy et al. 2017; van der Veen et al. 2017) and foster students’ development of oral discourse skills through classroom talk (Heller and Morek 2015 a, b) at the same time. Empirical findings from research on classroom discourse and children’s acquisition of discourse skills as presented below formed the starting point of the PDP.

First, oral discursive practices such as explaining or debating represent a central medium of learning in classroom talk. Building on research in the sociology of knowledge (Keppler and Luckmann 1991), we use the notion of “oral discursive practices” to refer to sequentially organized, interactively accomplished, and culturally rooted procedures for developing, negotiating, and sharing different types of knowledge for different communicative purposes (Heller and Morek 2015 a; Quasthoff et al. 2017, 2022). Since such discourse practices typically require at least one participant to verbalize conceptual, causal, chronological, or other relations in a longer coherent turn (not just in a single word or a single sentence), discourse practices are particularly useful for joint knowledge construction in conversation and are, thus, an important component of academic language and academic discourse practices (Heller and Morek 2015 a; Bunch and Martin 2021; Schleppegrell 2010; Quasthoff et al. 2022; Snow and Uccelli 2009; Uccelli et al. 2014).

Regarding classroom interactions, various empirical studies have found that common discursive practices in classroom talk are describing (e. g., features of a phenomenon), reporting (e. g., observations), explaining (e. g., meanings, rationale of procedures), and arguing (e. g., evaluating hypotheses or interpretations) (Morek 2016; Heller et al. 2017; Erath et al. 2018; Uccelli et al. 2014; Vollmer and Thürmann 2010). Thus, students must have certain oral discursive skills to productively participate in classroom talk (Gillies and Khan 2008; Osborne et al. 2019; Quasthoff et al. 2022). However, language acquisition research indicates that students’ oral discursive skills cannot be presupposed at either the primary or secondary school levels. In fact, these skills continue to develop well into adolescence (for an overview: Nippold 2007; Stukker et al. 2024). Moreover, students’ opportunities to develop such skills outside of school – whether at home or among peers – vary greatly (e. g., Heller 2014; Morek 2015).

These insights have led to calls for incorporating support for students’ oral discourse skills into subject matter teaching (Heller and Morek 2015 b; Gillies and Khan 2008; Heppt et al. 2022; Kalinowski et al. 2019; Smit and van Eerde 2013; Uccelli and Phillips Galloway 2018). Key to such support are specific interaction strategies as identified by interactionist language acquisition research (cf. Quasthoff and Wild 2014), initially in adult-child interactions (e. g., Hausendorf and Quasthoff 1992) and later also in teacher-student interactions (Morek and Heller 2021). We can distinguish between interactional strategies to a) establish discursive tasks (e. g., requests for explanations, justifications etc.), and to b) help learners verbalize longer discursive contributions by means of listener activities (cf. Morek and Heller 2021; Quasthoff et al. 2022). Building on interactionist theories of language acquisition, such interactional strategies are a “necessary ingredient for successful discourse acquisition” (Quasthoff and Wild, 2014: 73). First, they repeatedly involve learners in discursive practices and have them actively contribute extended turns such as explanations, arguments, descriptions, and the like (Hausendorf and Quasthoff 1992; Quasthoff et al. 2022). Second, listener activities such as clarification requests or “candidate understandings” (Schegloff et al. 1977) help learners to form coherent contributions (cf. also Willemsen et al. 2020). When teachers regularly provide discursive tasks and listener strategies, students have been more likely to produce substantive contributions such as explanations and arguments (cf. Morek and Heller 2021; Quasthoff et al. 2022), thus expanding their discursive skills. In this article, we focus on interactional support strategies to establish discursive tasks (see Heller and Morek forthc. on teachers’ implementation of listener activities).

Establishing discursive tasks in classroom talk is associated with two context-specific challenges: First, it is vital not to initiate random discursive practices during a lesson but to meaningfully and coherently interconnect discursive practices to subject matter learning (Heller and Morek 2015b). For instance, a request for an explanation should focus on key conceptual understandings of a lesson (see example 1). Secondly, it is important to ensure broad participation in discursive practices and to multiply discursive learning opportunities, e. g., by inviting other students to participate in discursive tasks (e. g., ‘who else....?’) or by calling on quiet students (Morek et al. 2022a). Example (1) illustrates how establishing discursive tasks and multiplying opportunities for participation may look like.

In example (1), taken from a geometry lesson in grade 6, the teacher establishes a meaningful discursive task by asking “which areas are larger” (line 1) and requesting the students to “give reasons” (line 2). Here, an argumentative contribution is elicited that addresses the core of what is to be learnt: the students are to understand the relationship between geometric shapes and their areas. Unlike short-answer questions, such discursive tasks cannot be answered by providing isolated statements or pieces of knowledge. Rather, they establish communicative expectations for the students to provide a coherently structured multi-unit turn as a next move (Blum-Kulka et al. 2010; Quasthoff et al. 2017). In example (1), Osama’s monosyllabic answer is treated as insufficient (line 5), and he is required to provide an argumentative multi-unit turn, combining his assertion (line 4) with a justification (line 6, 8–10) and a conclusion (line 11, 20–21).

The second interactional strategy to set up discursive tasks, ‘multiplying opportunities for participation’, refers to establishing participation structures (Goodwin 2017) that do not only generate discursive learning opportunities for one selected student, but for several students. One way to do so is using participatory follow-up moves (Morek et al. 2022 b; cf. Willemsen et al. 2020) as the teacher in example (1) deploys in lines 23 (“what do the others say”) and 35 (“what are we going to do with that [contradiction, M.M and V.H.] now”). “Revoicings” that pick up and/or summarize previous student contributions are another way (O’Connor and Michaels 1993; Berry 2006) (e. g., example 1, lines 33–34, “osama now says that the triangle is larger”, “özge now says that the arrow is larger”).

Altogether, the teacher in example (1) – who participated in our PDP on interactional support for students’ discursive skills – successfully elicits a series of discursive contributions from their students (Osama, Özge), and all the discursive learning opportunities are coherently embedded in the in-depth reflection on subject matter content (here: ‘shapes and areas’). Previous research on classroom talk shows, however, that teachers only rarely use such interactional strategies intuitively, without having learnt them in a PDP (e. g., Heller and Morek 2015b). We can assume that implementing these strategies into everyday classroom talk presents a major challenge for teachers, as it requires them to open discursive space for students while, at the same time, connecting discursive opportunities with content-learning in a meaningful and productive way.

2.2 Implementing novel interactional strategies for classroom discussions as a process of teachers’ appropriation

Research on teacher professionalization has highlighted that introducing novel interactional patterns into classroom discussions represents a major challenge for teachers (Borko et al. 2021; Gomez Zaccarelli et al. 2018; Hennessy and Davies 2020; Lefstein and Snell 2011; Kalinowski et al. 2019; O’Connor and Michaels 2019; Šeďová et al. 2020; Wilkinson et al. 2023). Several factors can hinder or complicate implementation, such as institutional constraints (e. g., curriculum, time limitations) or teachers’ professional knowledge and mindset (e. g., epistemological beliefs about students’ ability to engage in discussions) (cf. Gomez Zaccarelli et al. 2018; Resnick et al. 2018; Šeďová et al. 2020). Most importantly, it is very difficult to de-routinize long-standing interactional practices that have been used unconsciously for years (Lefstein and Snell 2011; Wilkinson et al. 2023).

Studies that examined teachers’ trajectories of change in the context of productive classroom discussions have shown that transformation involves fluctuating between habitualized and new interactional patterns (Billings and Fitzgerald 2002; Lefstein 2008) and going through phases of regression and stagnation (Šeďová et al. 2020). These findings point to the teachers’ active role in professional development (Clarke and Hollingsworth 2002; Schindler et al. 2021). Changing interactional strategies should, thus, be regarded as an active and constructive process which involves implementing new practices in their own classrooms, i. e., “field-testing”, according to Clarke and Hollingsworth (2002: 949), and reflecting on the outcomes of such field-testing.

From a teacher professionalization perspective, such professional learning processes can be conceptualized as “appropriation” (Segal et al. 2018). The notion of appropriation was introduced to teacher training research by Grossman et al. (1999). Originally, it was coined by Bakhtin (1981) in a much more general sense, i. e., to underline the fact that any speaking always involves adopting linguistic means previously used in other contexts for new contexts. This idea was subsequently taken up in sociocultural approaches to child development (e. g., Rogoff 1990). According to Wertsch (1998: 53), appropriation of new skills is the process “of taking something that belongs to others and making it one’s own.” In addition, Rogoff (1990: 195) pointed out that linguistic devices or communicative activities are never appropriated as a whole; rather, “individuals are seen as appropriating some aspects of an activity in which they are already engaged as participants and active observers” (emphasis added). Grossman et al. (1999) adopted the notion to conceptualize processes of (teacher) professional development. They used “appropriation” to refer to the incremental process of adopting new pedagogical tools and gradually becoming competent in using them and outlined five stages of appropriation from a rather abstract perspective (e. g., “appropriation of surface features”, “achieving mastery”). However, it can be assumed that stages of appropriation differ depending on the ‘tool’ or strategy to be appropriated.

Only a few studies have tried to explore how teachers’ appropriation of new forms of professional interaction takes place. Segal et al. (2018) did not focus on classroom talk but on peer-counseling among teachers. They observed that when trying to implement a guidance protocol for their meetings, the teachers initially used the protocol rather rigidly and formally before moving on to a more flexible and interactively embedded use. Furthermore, the teachers were found to adopt certain parts of the protocol while ignoring others. Regarding classroom talk, there are a few exploratory studies that focus on how teachers approached productive classroom discussions (Heyd-Metzuyanim et al. 2019; O’Connor and Michaels 2019; Šeďová et al. 2020). Quite like the observations by Segal et al. (2018), Heyd-Metzuyanim et al. (2019) demonstrate how two math teachers start out with a “rigid imitation” of interactional moves demonstrated by expert teachers in the PD sessions before they show a self-directed and flexible embedding of these moves in subject matter learning. In a similar vein, O’Connor and Michaels (2019) report on teachers’ “robotic usage” of particular talk moves in the course of a PDP on Accountable Talk. Šeďová et al. (2020) add a longitudinal perspective; they model teachers’ processes of change in relation to Dialogic Teaching as a continuous problem-solving process, in which attempts to implement new instructional practices temporarily result in unintended negative consequences, which in turn result in regressing to habitual behavioral patterns.

Although these studies provide initial insights into how teacher appropriation might unfold in the context of classroom discussions, no study to date has examined teachers’ implementation efforts for content-integrated support of students’ discourse skills. The study at hand examines how teachers appropriate interactional strategies to support students’ discursive skills in whole-class discussions. Assuming that appropriating such content-integrated support of discursive skills is a complex transformational process, we conceptualize seemingly ‘incomplete’ or idiosyncratic ways of integrating components into their own repertoire of conducting classroom discussions as a momentary state of appropriation or ‘appropriation phenomenon’. The aim of our exploratory study is to microanalytically investigate teachers’ ways of adopting interactional strategies to support students’ development of discourse skills. In this paper, we focus on interactional strategies to elicit extended discursive contributions from students. Our analyses address the following questions: How do teachers adopt interactional strategies to establish meaningful discursive tasks and multiply learning opportunities in classroom discussions? How can these in-situ adoptions be understood as instances of appropriation? Our aim is thus to provide a detailed sequential analysis of appropriation phenomena. An analysis of their frequency and distribution thus remains for further research.

3.1 Design of the professional development program

Our PDP combined practical implementation in teachers’ classrooms with video-based reflection (Chen et al. 2020; Kalinowski et al. 2019; Lipowsky and Rzejak, 2015; Šeďová et al. 2020). It covered a 12-month period and included a sequence of five thematic workshops that addressed different components of integrating support of academic discourse skills into subject matter learning (see Figure 1). After each workshop, the practical use of an interactional strategy in the teachers’ lessons was videotaped. These lesson recordings formed the basis for reflections within teacher-scientist tandems that focused on reinforcing successful implementation and developing approaches for overcoming difficulties.

Figure 1

Design of the Professional Development Program

In terms of methods, the workshops included short theoretical inputs based on transcripts that made both supportive and unproductive interactional strategies accessible to reflection, and practical simulation of strategies. In addition, planning grids aimed at facilitating teachers’ integration of discursive tasks into classroom discussions.

3.2 Setting, participants and data

A total of 14 teachers participated in the PDP. They came from three secondary schools in low-income urban communities in Germany all of which provide intermediate education that qualifies for vocational training (‘Realschule’, ‘Hauptschule’, ‘Sekundarschule’). We conducted our study in grade 6, with monolingual and multilingual students aged 12 to 13 years.

Our study followed ethical guidelines to protect participants’ rights and privacy and was approved by the Ministry of Education of North Rhine-Westphalia. Informed consent was obtained from teachers, students, and their parents.

We videotaped several lessons of each of the 14 teachers during the PDP. To be able to conduct fine-grained analyses of their classroom discussions, we created 12-minute samples per lesson which contained a) a whole-class discussion and b) represented the opening phase, the discussion phase, and the results phase of the lessons (4 minutes each). These sequences were transcribed and mutually verified by two research assistants according to the conventions of GAT 2 (Selting et al. 2011)^[2].

To explore teachers’ in-situ adoptions, we selected a subsample of teachers of German (Table 1) for two reasons. First, their professional focus on language support seemed a proper starting point. Second, this choice allowed us to maintain some degree of comparability across curricular topics (content areas comprised fairy tales, legends, fables, personification in poems, reading strategies, argumentation, and direct/reported speech).

3.3 Microanalysis of teacher adoptions of elicitation strategies

In a preparatory step, drawing on lessons during the PDP, we created collections of sequences in which the five teachers observably attempted to elicit explanations, arguments, or other discursive practices from students (e. g., “Who can explain....?”; “... and give reasons for your assumption.”). These sequences were then analyzed independently by each author in a two-step-procedure:

In a first step, we analyzed the sequential unfolding of the excerpts following interactional discourse analysis (IDA) (cf. Quasthoff et al. 2017, 2022). This approach draws on conversation analysis (CA) (Sacks 1995) and extends the examination of how participants accomplish actions in talk to larger activities such as explaining, arguing, and other discourse practices. For this, we first examined all the questions and prompts of the teachers regarding whether they required complex turns such as explanations, justifications, etc. (cf. Quasthoff et al. 2017) as reactions or whether they only called for short answers (e. g., name a technical term). The first case sets a discursive task (e. g., to explain something), the second does not. Our analyses also acknowledged linguistic, prosodic, and embodied resources to capture the ways in which discursive tasks were framed. By means of interactional discourse analysis, we thus examined if and how teachers elicited discursive contributions and if and how teachers multiplied opportunities so that more students could contribute to discursive tasks (see 2.1).

In a second step, we looked at the discursive tasks regarding their relevance and functionality for subject-related learning in each lesson (Pauli et al. 2022). We interpreted a discursive task as meaningful if it coherently contributed to achieving the learning objective of the lesson (e. g., justify an estimate of the size of an area when the concept of ‘area’ is to be learnt). On the other hand, discursive tasks are without functionality for content learning if they are not coherently connected to the learning objectives of a lesson but rather lead away from the topic, or if they do not allow for building up a deeper understanding of the subject matter.

The analytical observations of both steps were intersubjectively verified in data sessions within the research group. In a cyclical process of comparing different sequences across teachers, we ultimately identified four ways that teachers in our data enacted to elicit discourse contributions from their students within whole-class discussions. We interpret these as appropriation phenomena. To check whether the observed phenomena represented adoptions of new strategies from our PDP (and not ‘business as usual’), we made sure that comparable phenomena were not observable in these baseline recordings collected before the start of the workshop series. We present the four appropriation phenomena in the results section through prototypical excerpts from different lessons of German.

4.1 Undoing and blurring a discursive task

The first attempt to enact an interactional strategy from the PDP we identified was characterized by teachers first establishing a discursive task but then promptly revoking it. It, therefore, appears as if a meaningful discursive task was established when, in fact, it was not: Before calling on a student to answer, the teachers replaced the communicative expectation for a multi-unit turn with one for a short answer (“undoing”, Example 2), thereby minimizing the subject-related task.

Example (2) is a prototypical case of ‘undoing the discursive task’. In the context of a lesson on reading comprehension, the class reads a text about a girl fighting for children’s rights. While talking about the text and some unfamiliar words in it, the teacher (Mr. Schiller) establishes a discursive task, which he then immediately revokes, however.

Referring to the girl’s appearance on a radio show (375–377), Mr. Schiller asks “why does she actually want to do that;” (378). With this why-question, he establishes the expectation to provide an explanation (here: to explain a protagonist’s cause), i. e., he seems to establish a meaningful discursive task. However, the expectation is immediately followed by a chain of questions which ask for only short and simple answers: a rhetorical question (379: “is she bored;”) that provides an implicit hint where not to look for an answer, and a question that invites the formulation of a simple assumption (381: “what does she probably want to talk about?”). Furthermore, Mr. Schiller gives a hint as to what kind of assumption is unlikely (382) and then preformulates an answer with a placeholder “something else” (383–387) that has to be filled.

The chain of short-answer questions minimizes both the subject-related task and the discursive task for the students: from a subject perspective (here: developing reading comprehension skills), the students no longer have to make a global coherence inference (e. g., inferring a character’s motives from their actions reported in the text), but merely have to identify and reproduce a fact stated in the text. From the perspective of discursive learning, they do not have to provide a multi-unit turn to explain how the character’s motives relate to her actions, but simply have to state a piece of textual information. In fact, Torben merely states what the character will say on the radio (389), which the teacher confirms (392–394).

Some teachers also decrease the initial discursive demand through linguistic mitigations, thereby additionally blurring the discursive expectation, as in example (3).^[3]

While the teacher initially establishes a discursive task (taking a position on the summer camp; giving reasons 9–11), the communicative expectation to justify one’s position is decreased twice through epistemic hedges (“maybe”), thereby blurring the communicative expectation. Subsequently, as in example (2), the expectation for a multi-unit turn is revoked by a short-answer question (19) that only expects the stating of a position but not its justification.

Cases like the above are characterized by teachers starting to use a new practice – establishing meaningful discursive tasks –, but not allowing it to become interactively effective. Instead, they return to their habitualized pattern (Korthagen and Kessels 1999; Šeďová et al. 2020) of small-step questioning, which allows them to navigate the interaction in a way that is familiar to them. Through their fallback to short-answer questions, the teachers – unintentionally and unconsciously – also bring about or confirm low expectations regarding the cognitive and discursive competence of their students.

4.2 Limiting the interactive consequences of a discursive task

In contrast to undoing the discursive task of an initiating move (section 4.1), the second phenomenon is when teachers succeed in establishing an expectation for a multi-unit turn, but resort to various maneuvers to limit the interactive consequences – which are still unfamiliar to them. The maneuvers observed were discursive acting in the as-if mode (Example 4) and taking over the fulfillment of a discursive task (Example 5).

Example (4) is from a lesson on argumentation; the concepts of ‘questioning’ and ‘contradicting’ were introduced, and the students read a text about a girl who tried to change the world with a small charity project (see Example 2). Using a hand puppet (“Paul”), the teacher reproaches the fictional character Anne for not being able to really change the world. He then asks the students to prepare and present a written statement on Paul’s position and address it to the hand puppet.

By installing the hand puppet as the addressee, Mr. Schiller establishes an as-if mode of arguing and frames the argumentation as a kind of performance. This is underlined by the fact that he gives ‘stage directions’ as to how Jonas should stand (866–687). A framing reminiscent of a theatre play is created: those present are not supposed to act as discussants, but as actors (880: “really loud and clear”) and spectators (878-879: “look to the front; and listen carefully;”). Characteristically, the stage directions also relate to the communicative intentions behind the discursive practice to be performed: as there is no real, but a fictitious reason for argumentation, a desire for argumentation must be introduced from the outside. Mr. Schiller does this by using a declarative statement instead of a question (864-865: “you want to [...] give him your opinion on his statement”).

This way of establishing a discursive task allows Mr. Schiller to maintain a high degree of control over the interaction: The very fact that the arguments were prepared beforehand results in the students not acting in a conversational, but in an isolated manner. Likewise, establishing a fictitious addressee prevents a spontaneous conversation. Instead, the contribution is followed by a cesura, and despite argumentative responses, there are thanks (893), praise (895), clapping (896), and a correction (898), none of which address the discursive validity of Jonas’ argument, but the performance as such.

In a second variant of this appropriation phenomenon, the teacher retroactively controls the interactive consequences of a discursive task by taking over its fulfillment himself. Example 5 is from a lesson aimed at critically reflecting on social media. Mr. Fontane initiates a discussion about friendship by talking about a girl who claims to have more than 600 friends.

After Chantal has contradicted the claim of the fictional girl to have 600 friends, Mr. Fontane establishes a discursive expectation to give reasons (122). He therefore appears to implement the interactional strategy ‘establishing meaningful discursive tasks’. As expected, Chantal produces a justification (123), which, however, spontaneously – and not brought about by the teacher – provokes dissent among some classmates (125–126). The teacher’s control of the conversation is thus briefly suspended. At this point, Mr. Fontane regains control of the talk by taking over the task of giving reasons: First, instead of prompting the students to give reasons, he objects to Chantal’s justification himself and insinuates that she meant something else (127–128). Second, he asks two short-answer questions (129: “where do you have your friends PTCL.” and 132: “where else,”). This shifts Chantal’s role to a “small piece supplier” (Quasthoff et al. 2022) who is no longer able to contribute an argument on her own, but only has to complement the teacher’s argument. Interrupting her (136), Mr. Fontane eventually takes over the fulfillment of the discursive task and produces an extended argument himself (136–154). Finally, he credits Chantal with having meant what he has just said (155).

In summary, both variants of this appropriation phenomenon are characterized by the teachers establishing a discursive task, but either confining its interactive consequences from the outset (by setting up an as-if mode of argumentation) or retroactively mitigating the interactive consequences (by taking over its fulfillment). Both cases limit the opportunity for spontaneous argumentative contributions by the students and a dynamic development of the discussion. The teachers, thus, dare to try something new, but adopt it in a way that ensures the predictability and controllability of the ‘discussion’ familiar from low-level interaction patterns. Both variants of this appropriation phenomenon were found solely in the context of argumentative discourse practices. The difficulty to anticipate and steer trajectories of argumentative talk might render argumentation particularly susceptible to this appropriation phenomenon.

4.3 Initiating discursive practices without subject-related integration

The third type of eliciting discursive contributions refers to instances in which teachers integrate discursive tasks into classroom discourse but fail to connect them to relevant subject-related learning goals. This lack of integration results in the disconnectedness of the discursive tasks and content learning. As this phenomenon is related to the underlying didactic structuring of a lesson, it can occur in combination with phenomena 1 and 2. Example (6) showcases a particularly obvious case of removing content relations by foregrounding the formal aspect of discursive practices. Example (7) demonstrates how topically disconnected discursive tasks can occur in a series, resulting in an incoherent chain of discursive contributions.

Example (6) is taken from a literature lesson that introduces the literary genre of fables. The class has gathered some preliminary knowledge on fables and looked at a textbook picture showing different animals (e. g., a lion, a fox). The textbook page entitled “We speak of animals and refer to ourselves” implies that a central literary learning objective for the students is to recognize that fables attribute human characteristics to animals in a stereotypical manner to illustrate typical human problem situations. However, the discursive activity the teacher initiates does not lead in this direction at all: He first goes through a list of different adjectives and their meanings (e. g., “vain”, “sly”, “strong”) and then establishes a discursive task that asks the students to draw connections between animals and their traits (101–105).

This projected discursive activity, which is referred to as ‘connecting the adjectives with certain animals’, is not established as topically coherent with the preceding contextual focus. Instead, it is framed as a completely new focus (e. g., by transitional markers such as “okay”, pauses, “now”) with only a loose connection to the overall topic ‘fables’. Furthermore, the projected discourse activity is framed as a form-oriented exercise, while any underlying communicative or cognitive functions (e. g., “explain why a certain trait is attributed to an animal”) remain unaddressed. The fact that the teacher places particular emphasis on the formal structure of the multi-unit turns to be delivered is also reflected in his modelling of two examples. Drawing on the lion (108–112 and 120–121) and the donkey (116), he highlights the formal structure (e. g., prosodic foregrounding of the connective “WEIL” (because), 110, 116) while simultaneously downgrading the relevance of the content (e. g., “like this and this and that”, “I don’t know”).

Being only asked to copy this model (114–115, 118–119), the students are not positioned as participants in a meaningful discursive exchange, but rather as replicators of a linguistic format to be rehearsed. When one of the students actually tries to contribute as requested (“a dog is fast”, 130), the teacher steps in in a preventive manner, and emphasizes the need to continue the utterance with “because” (133). At the end, he reformulates the student’s contribution to fit the predefined format (137).

Initiating discursive practices without a subject-related embedding is, thus, characterized by the fact that on the interactional surface, the teacher appears to establish a discursive task but from a language and literacy education perspective, the discursive demand is not meaningfully connected to the learning objectives of the lesson. The slightly mechanistic impression of such attempts also becomes apparent when teachers elicit unconnected multi-unit turns in a series. This is particularly common when it comes to explaining meanings of words, as shown in Example (7). It is from a lesson in which the class reads and discusses an informational text about teenagers’ use of the internet. Our didactic reconstruction suggests that the major learning objective in this lesson is that the students can reflect critically on their personal use of social media. The teacher, Mr. Fontane, uses the read-aloud situation to establish a series of discursive tasks, all leading the students to explain the meanings of words. He does so by continuously interrupting the reading whenever he anticipates that a word might be unfamiliar to the students.

In the above extract, the teacher elicits multi-unit turns three times (241, 253, 289). In all cases, he takes up a word occurring in the text and asks for its meaning. Note that the linguistic format of all three teacher initiations is very similar (“what does ‘on average’ mean”; “what are virtual friends”; “what does ‘fake’ mean”). By serially drawing on one and the same genre type of discursive activity (word meaning explanations) and using a constant format, he manages to establish a number of discursive tasks in quick succession. As providing discursive learning opportunities was one of the central goals of the PDP, this strategy is very functional for field-testing this. However, while each of the discursive tasks might be meaningful, in the above example and comparable sequences from our corpus, they lack connection to each other. For instance, in excerpt 6, the question of whether ‘fake’ is the same as ‘virtual’ remains unaddressed.

While a series of unconnected elicitations of multi-unit turns can be seen to facilitate students’ production of extended contributions, these successively produced discursive contributions remain isolated, which means they do not contribute to advancing the joint construction of knowledge.

In summary, the attempts to establish discursive tasks described in this section – and their often serial occurrence – are characterized by the teacher establishing one or more discursive tasks, but without connecting them coherently or functionally to the subject matter learning. The students are requested to provide multi-unit turns and they actually do, but their discursive moves are aimless, as their contribution to the construction of knowledge or thematic progression remains unclear. It can be assumed that this way of initiating a discursive practice reflects an appropriation phenomenon of teachers who make a particularly strong effort to provide discursive learning opportunities for their students. However, it is only the performative side of discourse practices that the students are entrusted with, while the cognitive and communicative side of discourse practices, their use as a tool of joint thinking about a topic, is disregarded.

4.4 Schematic follow-up moves for multiplying discursive learning opportunities

Apart from the cases described above, all of which involved initiating discursive activities, we also identified attempts to deploy follow-up moves geared to multiply discursive learning opportunities. In one of the workshops, the teachers learned that it is usually the discursively proficient students who volunteer to answer cognitively and discursively challenging questions, which makes it essential for teachers to involve other students in the discursive activity with follow-up moves. The teachers saw different examples of how such follow-up moves engage students in rephrasing a classmate’s contribution in their own words – expanding on a previous turn, introducing alternative perspectives, evaluating suggested ideas, and so on. The example for ‘expanding on a previous contribution’ that we presented on a chart reads “Now we’ve heard from Amir that.... Who else has something to add to this explanation?”

In our data, we identified several instances of teachers trying to incorporate such follow-up moves into their classroom talk. However, they did so in a schematic way, disregarding the actual sequential context in which they placed the follow-up moves. This resulted in rather dysfunctional and incongruous applications of follow-up moves, as illustrated in example (8). This excerpt is from a literary lesson on fables. The aim was to understand the plot of the fable “The fox and the stork” that was presented through a sequence of pictures. The picture in question shows a stork trying unsuccessfully to eat from a plate with its beak while the fox devours everything on his own plate without problem. The teacher sets up the discursive task to explain the kind of problem shown in the picture (167).

In the first explanation (180–182), the student, Laura, names the shape of the stork’s mouth as an anatomical obstacle to the stork’s ability to eat, which she contrasts with the fox. In this way, she provides a condensed, abstract explanation (Heller, 2021) that succinctly explains the fable’s central problem. Despite this, the teacher employs a follow-up move, asking for additions (183). In this way, she implements one of the strategies proposed in the PDP for encouraging multiple students to engage in explanatory and argumentative discourse. Note that her wording (“can someone add to that?”) is very similar to that on the chart (“Who else has something to add to this explanation?”).

The teacher’s subsequent self-corrections of the follow-up move (184–188) reveal that a request for an addition does not make sense at this point in the discussion: First, the teacher reformulates her follow-up move and asks for a “better” version of the given explanation (184) (presumably noticing that nothing more could be added to it). She then corrects herself again, asking for an expansion of the explanation (“a little further”) and explicitly dispels any implications that the first explanation was not good enough. While Lucas tries to add a new aspect (the ‘teasing’, 190), he comes back to Laura’s original version (‘fox can eat’ – ‘stork cannot eat’ – ‘long beak’, 197–199). The only detail that can be added in the end is a description of the shape of the plate (‘flat’). Falling back on the initiation-response-feedback pattern, the teacher finally extracts this detail from the students and closes the sequence with a complete explanation that she provides (210–212).

Altogether, the teacher’s enactment of a follow-up move does lead to a second student engaging in explanatory discourse. In terms of content, however, this does not move the conversation forward, but rather results in a tenacious search for a small detail of information which might satisfy the teacher. The actual communicative goal of a discourse activity (here: understanding a problem) recedes behind the surface form of engaging several students into producing multi-unit turns. Other instances in our data show that this also occurs for schematic follow-up moves during an argumentative activity (e. g., ‘Who thinks otherwise?’). Such contextually incongruous, communicatively dysfunctional uses of follow-up moves resemble what other scholars have described as “robotic usage” (O’Connor and Michaels 2019) and “rigid imitation” (Heyd-Metzuyanim et al. 2019).