Introduction

The article begins by examining two dichotomies used by the Organization for Economic Co-operation and Development (oecd) to define learning, showing the limitations of these dichotomies in understanding the phenomenon of learning. I then present approaches based on the analysis of social interaction and the notion of zone of proximal development which allows for articulating the analysis of social interaction and the psychological process of learning. The analytical part presents two extracts of interaction between a learner and a teacher. These extracts show that to understand these learning/teaching situations, an interactional analysis is necessary. The analysis of the interaction allows us to understand how the learner’s knowledge is taken into account by the teacher, and to what extent the learner can rely on it to develop a zone of proximal development, that is, an opportunity of learning. In conclusion, we return to the proposal to include the categories of leader and follower in the recommendations concerning learning of an organization like the oecd.

Why is an International Organization Interested in Learning?

The oecd aims at “reshaping learning to better match the needs of the 21st century knowledge economies and open societies” (oecd 2020). It is in this perspective that this organization advocates the recognition of non-formal and informal learning alongside formal learning. The goal of this recognition is to valorize the knowledge and experiences that individuals can construct outside the initial education system, and facilitate their integration in the labor market or access to the higher education system. The recognition of skills acquired outside the formal learning system might be crucial also for migrants arriving in Europe and needing to have their skills recognized, for instance in order to get a job. This policy is known as “Lifelong learning for all” (oecd 1996). At the core of this policy, we find therefore the notion of “learning”, under different categories: formal, non-formal and informal. However, I will argue that if international organizations wish to organize lifelong learning that is based on the personal experience of individuals, which, as we shall see, is a crucial condition for efficient learning, the current institutional categories are not sufficient. The implementation of a learning system adapted to the multiplicity of trajectories engendered by the mobilities of the 21st century implies that the interactional dimension of learning practices be valued, conceptualized and taken into account. This article provides a contribution to the conceptualization of the interactional dimension of learning situations and a simple tool to assess the taking into account of the learner’s perspective.

The Problem: Defining Learning Only Through Its Institutional Context

In an official document of the oecd, Werquin (2007) proposes to retain two dichotomies to define any type of learning:-Whether the learning is intentional; or whether it happens as a side effect as it were”; and – Whether the activity, whatever it may be, has [a] learning objective(s) or not. (Werquin 2007:5)

In Werquin’s definitions, learning is given as something that may happen, or may not. Learning is defined through two dichotomies (intentional/vs/ non intentional; presence/vs/absence of objectives) that are external and prior to the interactional process of learning itself. In these definitions, learning is explained by what precedes its accomplishment in social interaction, but learning per se, as something occurring between participants in an activity, is not specified. Little can be explained therefore if we ask the question of how learning happens when preliminary intentions and objective are absent, as in the case of informal learning (Werquin 2007:5), which is defined by the absence of intention and objective of learning. These definitions tell us little about how learning occurs. Another problem appears when we consider the fact that participants rely on learning objectives. Does it mean that if learners start with the same objectives and with the same intentions, they will learn the same things? Or is the learning process more complex, involving for instance various ways of implementing the preliminary objectives during the activity occurring between the learner(s) and a teacher or someone more experienced? Whenever we ask questions related to how learning is concretely organized to definitions of learning based on categories that precede its accomplishment in social interaction, we lack a tool to apprehend practices that work and practices that fail, and to identify the causes of success or failure. Indeed, in any situation, a moment of teaching-learning might be problematic and fail: the element to learn might be misunderstood or, on the other side, moments of teaching-learning might be very efficient. These problems are not automatically linked to the institutional context of learning, but rather to the organization of the interaction between the learner and the more experienced participant. While there are links between the institutional context and the organization of social interaction, these are not systematic, in that an institutional context is always renewed during social interaction (Mondada and Keel 2017). Furthermore, the notion of “legitimate peripheral participation” (Lave and Wenger 1991: 29) tells us that learning is a situated social practice during which social knowledge is renegotiated to be adapted to the present circumstances.

Toward an Ethnomethodological and Interactional Approach of Learning

If learning occurs during the interaction between a participant and someone more competent, (e.g. a teacher), then it should be examined as a process occurring in time. Indeed, time is constitutive of the object of learning, just as time is constitutive of “the matter talked about” (Garfinkel 1967:40) during a conversation. The object of learning as an interactional process can be conceived as “a developing and developed event over the course of action that produced it, as both the process and product were known from within this development by both parties, each for himself as well as on behalf of the other” (original emphasis, Garfinkel 1967:40). If we adapt what Garfinkel explains about the topic of a conversation to the construction of a learning object between a learner and another (more competent) participant, we must consider that this object is both a product and a process. This means on the one hand that the participants identify the learning object in a definite, fixed form, or in a form that can be modified, adapted to any new context or problem. In this perspective, understanding learning implies analyzing the ways participants organize their interaction during what they recognize and organize as learning activities. The analysis must take into account the participant’s perspective: how do the teacher and learner co-construct the element to be learnt from within? How do they establish and identify step-by-step what the object of learning is? One of the challenges of the analysis is to be able to account for the point of view from which the learning object is constructed. Which point of view is made public? that of the more competent participant? that of the learner? Does the more competent participant take into account the point of view of the less competent participant? As occurring in social interaction, is learning publicly available for the participants and for the researchers (Berducci 2011). Py states that learning “leaves observable traces in the discourse” (2000:6). Koshmann also states that “learning is an accountable, public, and locally occasioned process. By this view, interaction does not just provide evidence of learning; it is where learning itself is to be found. We might call this learning-in-and-as-interaction” (Koshmann 2013: 2).

But what does it mean that learning is accountable in social interaction? How to understand the notion of learning-in-and-as-interaction? One way of understanding this statement is proposed by Nishizaka (2006) who shows that learning implies the learner reconstructing a relevant environment during social interaction in which the object being learned can be manipulated or mobilized. This has been shown for example in relation to grammar rules (Lefebvre 2019).¹ This reconstruction of a new object implies that the learner relies in one way or another on known elements.

Construction of the Learning Object and Zone of Proximal Development

The conception of learning as the co-construction of an environment in which the learner and a partner will be able to elaborate or mobilize the object of learning presents many possible points of articulation within a socio-cultural theory of learning developed by Vygotsky (Lantolf and Thorne 2007) and in particular the notion of the zone of proximal development (zpd). Such approaches to learning are based on a situated approach to cognition and that they emphasize the centrality of social interaction in learning phenomena (Mondada and Pekarek Doehler, 2004).

Vygotsky’s contribution to the phenomenon of learning is to articulate its psychological and interactional dimensions. In a sociocultural approach, mental functions are first activated in the interaction between individuals before being internalized at the psychological level (Lantolf and Thorne 2007: 207). The phase of learning that occurs in social interaction is based on the possibility for the learners to build on what they already know how to do in order to amplify their skills through interaction with a partner who is more advanced in the accomplishment of a task. This is what the notion of the zone of proximal development refers to: “the distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers” (Vygotsky, 1978: 86).

The central contribution of the notion of zpd is to situate learning in the space between what the learners already can do and what they can only do with the help of others. This approach is therefore complementary to the ethnomethodological approach which assumes that any learning moment is reflexively made observable and public in the course of an activity where participants reconstruct their learning object. In their study on feedback as regulation, Aljaafreh and Lantolf (1994) showed that what the learners cannot do is made visible in the interaction and generates supportive practices from the teacher. The identification of the problem can also come from the learner him/herself who during the activity is confronted with a lack of some knowledge (e.g. a word search). In this article we follow the idea that the zone of proximal development leaves traces in social interaction and that it is by analyzing the turns of speech between participants in a learning activity that we can access the construction of this zone of proximal development.

To understand how participants reconstruct a relevant environment for learning new elements, analysis should then focus on how these participants organize their interaction. Concretely, analysis should address questions such as: who decides the relevant topic? Who decides when it is relevant to talk? Who asks questions and which type of questions? Who initiates corrections or new topics? Who leads the interaction? in relation with the activity of constructing the object of learning.

Conversation Analysis, zpd and the Data

Conversation Analysis (Sacks et al. 1974) is the study of how participants in social interaction coordinate their talk through the turn-constructional component and the turn-allocation component (Sacks et al. 1974: 702–703). The intersubjectivity and the rationality of any activity is accomplished in time through this interactional simplest systematics (Sacks et al. 1974). Previous studies have used conversation analysis to understand the organization of interaction in teaching/learning situations (see Gardner 2019 for an overview). Two seminal papers, McHoul (1978) and Mehan (1979), in conversation analysis investigate the interactional organization of learning, focusing on the organization of talk in classrooms. They showed that the rights to speak are not equally distributed in classroom, that the teacher continually evaluates the students’ talk and that this possibility of assessing the student’s answer depends upon the pervasive use of “known information questions” (Mehan 1979). More recently, studies in conversation analysis have reflected on how conversation analysis can provide resources for teachers to develop their awareness of the importance of interaction in their students’ learning (Walsh 2003). Following these previous studies, I argue that conversation analysis is a crucial approach for following how participants in a learning activity co-construct their object of learning. I will examine two interactional formats between a learner and a teacher that reveal two prototypes of co-construction of their learning object. In both extracts, the accountability of the object of learning is problematic for the participants themselves. Sharing the object of learning itself becomes a real challenge for both the teacher and the learner, and this has consequences for the zpd and learning opportunities.

The two extracts presented in this paper are parts of projects whose common goal was to better understand foreign language teaching and learning practices by analyzing naturally-occurring interactions in educational settings, from the perspective of conversation analysis. The first extract that will be examined is part of a project led by the French Institute of Yokohama, within the framework of in-service teacher training. The objective was to record courses and observe them with the teaching team in order to identify successful and problematic practices and to use these observations to improve their teaching practices. The second extract comes from a corpus recorded during the lali project (2020). One goal of lali was to create pedagogical activities during which participants could learn foreign languages through interacting in front of artwork in museums. Another goal was to improve the pedagogical practices occurring in museums through video recording and analysis (Lefebvre, Nyikos, Varhegyi, Dubrana, Mileti, Mutta, Peltonen, Hellström, Danzinger, Moosmüller, 2020).

In the two extracts we will examine, the teacher is completely fluent in the language being studied while the student has a very low-level in that language, and the common understanding of the object of learning between them is problematic. These extracts reveal two prototypical ways of organizing the interaction while reconstructing the object of learning and opening zpd s.

If we follow the typology proposed by Werquin (2007), the first extract, recorded in the French Institute of Yokohama is typical of formal learning (i.e. as participants come to the classroom they might have the intention of learning, and have learning objectives) while the second, recorded during a lali workshop in the Kuntzforum of Vienna is typical of non-formal learning (i.e. participants have the intention to learn but without clearly established objectives). However, rather than examining these extracts through the presence or absence of learning objectives and intention to learn, we will examine their sequential and multimodal aspects, focusing on the interactional organization of the activity, examining how participants build their object of learning in this participation framework and which kind of zpd is available for the learner. By taking an interactional approach to these two extracts, I will highlight that the practical issues faced by the participants are not as different as their classification in the oecd categories might suggest. The differences come from the organization of the interaction while reconstructing the object of learning.

I follow Mondada (2016) for the transcript method. Wherever possible, translations are made word for word in order to remain as faithful as possible to the original resources used by the participants.

Leading the Activity Without Taking into Account the Learner’s Perspective

Extract 1A shows an interactional organization during which the teacher (pro in the transcripts) leads the construction of the object of learning and is followed by the student. This occurs in a situation during which the teacher, and the group of students are correcting a homework exercise. One by one, the students go to the board and write the answer for one question, so that the teacher can check the relevance of each answer and the correction of each question can be shared among the group (see Image Situation). At the beginning of Extract 1A, the student, (shi in the transcripts), just wrote a sentence “je n’ai pas d’origine espagnole” (I have no Spanish origin) while the expected answer for answering the textbook question “Est-ce que tu es d’origine espagnole?” (Are you of Spanish origin?) was “je ne suis pas d’origine espagnole” (I am not of Spanish origin). At this point of the activity, pro initiates a sequence of correction with the goal of leading the student to identify the reparable (the verb) and to correct her mistake. pro is elaborating a plan in situ for leading the correction step-by-step. However, as we will see, the fact that pro leads the construction of the object of learning without identifying explicitly the reparable with shi makes it hard for shi to answer pro’s questions and to enter in the shared understanding of the activity: shi does not enter in the construction of the object of learning initiated by pro.

We will see throughout the analysis of this excerpt that this organization of the interaction has implications for the student and teacher’s ability to implement a zpd, that is, an organization of interaction in which the student can accomplish a task with the help of the teacher.

pro Solicits an Affirmative Answer

As shi finishes writing the sentence “je n’ai pas d’origine espagnole” (I have no Spanish origin in answer to the question “vous êtes d’origine espagnole?” Are you of Spanish origin?), pro who is reading what shi is writing calls her attention (Line 1) and asks the question “vous êtes euh japonaise?” (You are uh Japanese? Line 3). pro’s question is contextualized by an activity in which the whole group is correcting a textbook exercise that consists of answering different questions with negative sentences. shi’s negative answer (Line 5) manifests an orientation toward this exercise context: she interprets pro’s question as an exercise question rather than as a question seeking an answer based on her actual nationality (she, similar to the whole group of students is of Japanese nationality). However, pro does not align with shi’s interpretation (Line 7) coordinating her “non” with an eyebrow flash expression (Images F,G, below), categorizing shi’s answer as non-relevant for the current activity.

In response to pro’s negative assessment, shi corrects her answer (Line 9) by saying “oui”, without extending her turn further. During the break (Line 10) opened after shi’s turn, pro accomplishes a gesture of both hands rolling forward, producing in this context the meaning of soliciting the extension of shi’s turn (Image B), which is indeed the way shi interprets it: she extends her turn by producing a complete and standard syntactical structure in answer to pro’s question (Line 11).

pro Produces a Grammatical Explanation Before Projecting shi’s Negative Answer

At this point of the interaction, pro produces a multimodal turn in which she mentions two conjugations relevant for answering the exercise: she makes explicit the connection between the exercise sequence of question/answer and the conjugation of the verb “être” (to be). In coordination with talking, she emphases this connection with two pointing gestures, one pointing toward shi, coordinated with “vous êtes” (you are, Line 13, Images C, D) and another one pointing back toward herself coordinated with “je suis” (I am, Line 14, Image E). Furthermore, pro stresses her voice on the relevant conjugations of the verb être “êtes” (are, Line 13); “suis” (am, Line 14), also adding a pause before “suis” (am, Line 14). Note however that pro does not mention the negative form during her explanation, while the negation is central to produce the correction and therefore the object of learning.

Through her multimodal turn, pro orients the activity toward the language in use at the time, as the object of the current activity, which is a typical practice of foreign language classroom interactions (Seedhouse 2013). Indeed, pro produces a grammatical model (Lines 13–14) after the student’s correct answer (Line 11) and she relies on this grammatical model to contextualize her next question (Line 17) which projects a negative answer. Sequentiality has to be taken into account to understand the projection of a negative answer at this moment. By asking shi her actual nationality, pro’s first question projected (after a correction initiated by pro, Line 7) a “yes” answer. pro’s second question (Line 17), by asking shi if she is of “French” nationality projects therefore a “no” answer. This is indeed what shi does (Line 19). We will examine further this part in the continuation of the extract, below.

For now, let’s summarize what we have seen in this first part of Extract 1. pro’s first turn opens a sequence in which she leads each step of the activity while shi is in the position of follower. pro’s leading position is observable through the fact that each of her turns appears retrospectively as the projection of the previous one, showing that she is accomplishing in situ a pedagogical plan, in the following sequence: 1) pro first invites shi to produce a sentence at the affirmative tense, breaking with the expectancies of the exercise being corrected. Therefore, one method for controlling the activity is deciding what the relevant expectancies are for producing grammatical sentences. We saw that the expectancy of an affirmative answer appeared retrospectively through pro’s initiation of the correction of shi’s first answer in which the student followed the exercise expectancies of producing negative sentences. 2) pro relies on shi’s affirmative answer (she repeats the same verb, Line 14) to produce a grammatical explanation connected to the exercise structure, focusing on the conjugation of the verb être (to be, but not on the negation). 3) Within the context created by this grammatical explanation, pro asks a second question (Line 17) projecting this time a negative answer with the same verb, as in the exercise question being corrected. From the beginning to the end of this sequence, pro controls each step of the activity: she decides what is relevant to do (affirmative or negative answers), and when it is relevant to do so.

On her side, shi is in the position of following the different steps projected by pro. She answers pro’s question (Line 5) and corrects her first answer after pro’s initiation of a correction (Line 9). In both cases she answers minimally and completes her turn only after pro’s solicitation (Line 11). She also repeats the conjugation “suis” (am) after pro’s grammatical explanation (Line 16), manifesting that she is following pro’s explanation. shi participates in the activity by following and answering pro’s questions or instructions but she does not initiate actions, only producing minimal turn elements without extensions.

The asymmetry (Mondada and Keel, 2017) between the two participants is also observable through the ways they gesture. pro produces two gestures, one to give the instruction to shi to complete her turn (gesture of solicitation, Line 10) and one to create a visual dimension to her grammatical explanation Lines 13–14). pro also produces one facial expression (Line 7, Images F, G). On her side, shi remains almost immobile, only nodding after pro’s grammatical explanation (Line 16).

Asymmetry in the Shared Understanding of What is Relevant to Do

The asymmetry of the interaction between pro and shi appears even more clearly in the continuation of the extract (from Line 17, and Extract 1B, below). Indeed, we observed that the teacher projected through her second question (Line 17) a negative answer (with the verb être to be, in the negative form) offering an occasion to check if shi is able to produce a standard negative form with être (to be) and to correct the sentence she wrote on the board. However, pro’s exclusive control on deciding what is relevant to do during the activity produces a gap between what pro is targeting (i.e. having shi producing a standard answer with the verb être (to be) in the negative form) and what shi is doing (i.e. producing negative answers with avoir to have, or other affirmative answers). What appears in Extract 1b is that shi does not identify the specific point that pro is expecting her to correct, leading pro to point herself toward the reparable.

Failing to Construct the Common Understanding of the Reparable

In answer to pro’s second question (Line 17), shi produces a first segment in which she repeats the (non-relevant) segment she wrote on the board (Line 19). Just as shi finishes saying this first segment, pro produces a gestural negative assessment consisting of putting her right hand to her ear (Line 19, Image H), indicating a problem to shi and offering her a slot to correct her turn. However, shi repeats the same segment, extending it first with the nationality adjective (Line 22) and, after a pause repeats again the same first segment (“je n’ai pas” I have not, Line 22) without the nationality adjective. Note that shi does not repeat the nationality adjective proposed in pro’s question (French), but the previous one (Japanese), manifesting a gap between pro’s expectations and shi’s ways of following the activity proposed by pro. This gap is also manifested by the fact that shi repeats the problematic segment she wrote on the board. During shi’s turn after the first segment, pro manifests the non-relevance of what shi is saying by maintaining her hand-to-the-ear gesture and transforming it into a stop gesture (Images I, J, after shi’s repetition of the problematic segment, Line 22). At this moment, pro gives shi the instruction to take the time to think again (Line 24).

From this moment on, another element has to be taken into account to understand what happens. As pro initiates her stop gesture, the other student at the board finishes writing her sentence and goes back to her seat, opening a physical space between shi and pro. The second student leaves the place in front of the board at the moment when pro starts to repeat the question (Line 24). It is at this specific moment that pro interrupts the repetition of her question by giving an instruction to shi to come close to her (see the pointing gesture, Image K), attenuating this order by laughing (Line 27). This short side sequence, not directly linked to the accomplishment of the current action, however brings another element showing that pro is leading the organization of the interaction with shi, who again is in the position of following pro’s instructions.

As soon as shi comes closer, pro again asks the same question (Lines 27–28). shi initiates her answer with “no” but, this time, produces an affirmative sentence (Line 32), repeating her previous answer to pro’s first question (Extract 1A, Line 11). shi’s answer is grammatically correct, but not relevant from the exercise expectancy of producing negative answers. The double aspects of shi’s answer is first addressed by pro through (“hum hum”) indicating that shi’s answer does not match what she was expecting, but pro also recognizes the correct aspect of shi’s answer with a laugh and a positive assessment addressed to the rest of the group, raising her thumbs upwards (Image L). pro turns then directly to the identification of the reparable, pointing toward the two graphic elements marking the negation in French, ne… pas (not, Line 35, Image M).

We saw with extract 1B a gap between the expectation contained in pro’s question, in which she expects that shi produces a negative answer, and shi’s answers, in which shi produces non-expected linguistic forms (a negation with “avoir” to have, and an affirmative structure). We could also observe a misunderstanding in the use of the nationality adjective: pro uses the adjective “French” in her question, projecting a negative structure while shi keeps on using the adjective Japanese. These gaps in the shared understanding of what is relevant action at this moment of the interaction can be considered as a problem that pro attempts to solve by directly showing the item to be corrected on the board (Image M).

Control Gestures of pro and zpd

pro develops a whole range of gestures through which she acts on shi: to show that what shi has just said is not relevant, to interrupt her and reformulate the instruction, to invite her to change location, to share with the rest of the group the relevance of what shi has just said or to identify relevant linguistic elements on the board. If we think about the possible construction of a zpd at this moment of the interaction, we can notice that this gestural profusion of pro compensates for the fact that shi does not have the opportunity to propose elements from which the exchange could take place with pro. We can put forward the hypothesis that the possibility for shi to propose elements to accomplish the activity is inversely proportional to the quantity of gestures produced by pro. In other words, this organization of the interaction does not allow shi to enter a zpd, the interaction with pro does not allow her to connect her knowledge of French with new linguistic elements to accomplish the activity. This remark is valid for the moment we have examined and does not mean that in the continuation of the interaction this connection cannot or is not made.

An Interactional Organization of Learning: the Teacher Leads the Activity and is Followed by the Learner

In Extract 1 as a whole (1A and 1B), pro and shi accomplish a participation framework in which pro is in the position of deciding what is relevant action in order to construct the object of learning (i.e. the sentence “je ne suis pas d’origine espagnole”, I am not of Spanish origin). pro initiates a sequence of correction without explaining the nature of the current activity to shi, without sharing the reparable. She formulates a model to produce the sentence: a grammatical model to follow. We saw that pro produces a correction plan, in which she first has shi conjugating the relevant verb (être, to be, instead of avoir, to have), then she produces a grammatical explanation connected to the instruction of the exercise, after which she asks a question projecting the type of answer expected in the exercise being corrected. In this organization, shi occupies the interactional position of the follower in the construction of the object of learning. However, she does not follow pro’s plan without problem. On the contrary, until the moment pro identifies the reparable herself (at the end of Extract 1B) shi does not produce the expected sentence (“je ne suis pas d’origine espagnole”, I am not of Spanish origin), which is finally produced after one minute more of interaction between shi and pro after the end of extract 1B. shi has actually, in the extract we have scrutinized, no access to the criteria for constructing the object of learning in collaboration with pro.

By connecting this analysis to the notion of zpd, we can observe that in this interactional format, the formation of a zpd between the learner and the teacher is made problematic by the fact that the knowledge mobilized at this moment by the teacher does not offer any resources to the learner to contribute to the activity. As a result, throughout this excerpt, the learner spends time interpreting what the teacher expects of her, which finally leads the teacher to explicitly identify, through pointing gestures, the linguistic elements that the learner should mobilize to produce an expected response. This learning problem is not directly linked to the institutional or formal aspect of the situation but to the organization of the interaction at that moment.

The Learner’s Lead of an Activity of Word-search in Museum

Extract 2A, recorded at the Kunstforum in Vienna (January 2019), shows how participants follow an instruction that suggested they freely observe and comment on the paintings in the museum hall. The extract shows lil, a migrant woman newly arrived in Austria, with a low proficiency in German (level A1) becoming the leader of a word-search sequence: a word in German that she needs for describing a painting. The extract also shows how ste, the facilitator, follows the learner’s methods to find the relevant lexical item. The specificity of this situation lies in the fact that the learner knows what she wants to say but lacks the lexical resource to accomplish her description while, on the other side, the facilitator is fully competent in the language in use, German, but at the moment when the lexical problem emerges, does not understand exactly what the participant is attempting to say. The whole extract shows various methods initiated by the learner and followed by the facilitator in order to adjust the respective relevant knowledge to find the word that is being searched for.

Extract 2A

The Facilitator’s Manifestation of Sensitiveness to the Learner’s Talk

During the verbal pause (Line 1) three participants are gazing at a Japanese print in front of which they have just arrived. As soon as lil initiates her turn, ste manifests sensitiveness toward lil’s talk through multimodal resources: she gazes at lil as soon as she starts talking (Line 2) and encourages her to pursue her talk by a re-launch turn (Line 4) – i.e. she takes a turn to give the floor back to lil. While lil is talking, ste also produces a facial expression with the eyes opened and smiling (Image 2, Line 6), manifesting that she is focusing her attention on what lil is saying.

lil produces a turn in non-standard German (Line 6) proposing a description of the depicted character (see Image 3). In response, ste (Line 8) interprets and reformulates lil’s sentence in a standard form of German: she repeats the same verb (gehen, go) in another form (geht, goes) with the same subject (familie, family) to which she adds the standard article (die, the) and complement (hause, home), introducing a more complex morpho-syntactical form (nachhause, to home), manifesting again her stance toward lil of affording her with linguistic resources.

ste’s turn provides an interesting instance of when and how a facilitator can correct a participant’s talk by reformulating their talk with more standardized linguistic resources after the participant’s turn, but without interrupting the activity, here of describing the painting. This method of accompanying the learner’s talk and correcting it en passant is quite different in comparison with the method of asking a question to initiate a correction without mentioning the reparable (Extract 1).

This is an example of how the facilitator lets the learner initiate a topic, a topic that is therefore embedded in what the learner already knows, with varying degrees of confidence. The facilitator then builds on this participant-proposed item to offer a more advanced or standardized version. In terms of zpd, we see the facilitator accompanying the participant in the formulation of a more complex sentence than she could produce on her own.

This sequence is closed by the mutual positive assessment of what has been said: lil confirms that ste’s reformulation corresponds to what she said previously (Line 11), while ste validates lil’s description (Lines 13, 15). However, while the description could be closed by ste’s and lil’s positive assessments, lil opens a new sequence by self-selecting to add another element to the description she just proposed. She faces a lexical problem.

The Emergence of a Lexical Problem During the Learner’s Description

Extract 2b starts with lil proposing several versions of a word she needs in order to extend her description of the painting. After three repetitions of the word she is looking for, she gazes toward ste selecting her as the next speaker to provide the standard form of the word. The extract shows how both participants try to find the word lil is looking for.

Extract 2B

The Learner Initiates the Word-search Through Acoustic Approximations, the Facilitator Makes Propositions

lil initiates her new segment of the description of the painting by proposing several acoustic versions of a word she needs and is searching for (Lines 1–4), selecting through gaze (Line 3) the facilitator as the next speaker to provide her the relevant standard word, projecting that she, the facilitator, could be able to deduce it from these acoustic approximations. By testing different acoustic versions of the word, lil also manifests that she knows an approximate version of that word and is able to identify the standard version if ste proposes it.

Entering the task of searching for the word lil is attempting to say, ste makes two hypotheses (Line 7) based on the available graphic elements on the painting (Image 4) among which lil selects one element (Line 9) without however pursuing the activity of describing the painting by extending her turn (see short pause Line 10). ste asks a confirmation (Line 11) and proposes then to lil two new items to describe frau (woman, Line 15). Here again, lil selects one (Line 17) without pursuing the activity, manifesting that the resources proposed by ste do not match with the lexical resource she needs in order to pursue the description. However, on her side ste manifests that it is lil’s turn to provide more information concerning what she wants to say. This reciprocal attempt of getting more information from the other is shown between Lines 18–23, during which each gives the floor back to the other. In the last of these four turns, ste gives the floor back to lil (Line 23) who produces a turn describing the women (Lines 26–28) trying once again to introduce the word she is searching for when reaching the syntactical point in which the place category could be uttered, after gehe (go). lil repeats at this point the same acoustic approximations as at the beginning of the sequence (Line 1). ste therefore provides a new proposition, possibly based on the phonological resemblance between “gran² haus” (gran house) and “krankenhaus” (hospital, Line 30). However, ste’s new proposition does not match, leading lil to mobilize other resources by moving to another place of the museum room. Note that lil initiates the movement (Line 33) and that ste follows her and that during their interaction, the painting affords them a common ground to search for the word and to produce sentences.

The Construction of the Object of Learning in Front of a Painting

If we consider that this interaction is, for lil, the unplanned occasion of learning a new word, we can consider that her attempt to make ste understand the word she is searching for manifests an active stance toward her object of learning (i.e. she herself selects the object of learning, the context for using this object and the method to find it). Furthermore, her active stance toward her object of learning is made possible thanks to the availability of a common graphic ground all participants have visibly available in front of them, and the open instruction to interpret and/or describe any of its elements. On her side, the facilitator manifests a sensitivity to lil’s active learning stance (i.e. her attempt to find a word). She responds to her by providing possibly relevant words and initiating a syntactical construction that lil could complete with her word. ste also manifests her sensitivity through embodied behaviour (gaze, gesture, etc.). In this sense, through her practice of following what lil says, ste aligns herself with lil’s contributions, letting her conduct her own word search and organize her learning moment.

In the situation of a conversation with an aphasic person, a similar type of practice has been observed: family members of the person affected by aphasia follow the latters gestures and the two words that the aphasic person can still say to suggest words (yes, no). The aphasic person accepts or rejects their propositions (see Goodwin 2004). As with ste and lil, the participant with the fewest resources at their disposal leads the conversation by initiating topics, making devisions about what is appropriate action, and evaluating the discourse of others. ste ad lil’s mutual positions during the construction of the object of learning (i.e. the learner leads the construction of the object of learning while the facilitator follows her choice) and their ways of organizing it step-by-step (i.e. the learner initiates the turns for searching of words and the facilitator responds to these turns) is radically different from the previous organization observed in Extract 1 that I have categorized as a teacher-lead relationship. In Extract 2, participants build a relation during the construction of the object of learning which is structured around lil’s initiations of word searching turns and ste’s contributions through which she provides a possibly relevant lexical resource to lil. This co-construction of the object of learning is maintained throughout the extract until the end, with the mobilization of multimodal resources.

Using Gestures to Depict Meaning but Failing to Find the Searched for Word

At the end of the previous extract, in order to mobilize new resources for completing the word-search, lil initiates a move toward a different place in the room, and is followed by ste. lil sits on a bench and initiates a new turn to find her word with ste. In this sequence as the participants are no longer facing the painting, they rely on various gestures in order to depict the searched for word.

In response to the word lil formulates (Line 55) in order to re-initiate her word search, ste proposes a full construction (Line 60) in which she reemploys the word proposed by lil, in coordination with a gesture of miming the action of drinking tea (Image 6). In order to interpret the scene depicted on the painting and possibly relevant to complete lil’s word-search, the facilitator provides a multimodal turn which is formatted to be, thanks to her miming gesture, visibly intelligible (Kendon 2004). The learner might understand it even if she ignores some linguistic resources used by ste. In this sense, ste’s turn manifests her attempt to understand what lil is trying to say: she co-constructs with lil her object of learning.

However, once again, ste’s proposition is not relevant for lil who, in response to ste’s multimodal turn repeats her previous word by adding another one (Line 60) which is not an action (as ste proposed the action of drinking) but a place. In so doing she manifests that she is continuing to search for the relevant word and that ste’s previous turn did not afford this targeted word. What lil does therefore in Line 60 is to manifest that she is not trying to describe the character’s action but the place in which the character is sitting. That is the way ste understands lil’s turn in Line 63. She repeats the place category proposed by lil (also proposing a more standard pronunciation of lil’s version) while in overlap, lil draws through a gesture of both hands the spatial limits of an area corresponding to the place where the character in the painting is drinking tea (Line 63, Image 7). As both partners finish their turn in overlap, lil manifests through a facial expression (Line 64) the non-relevance of ste’s last proposition: Teehaus is not the word lil is searching for. She restarts the word search by proposing a place related to the previous one (Lines 67) and depicting this place through a gesture pointing ahead (Image 8). This gesture emphases the word “tür” (door) and shows the action of going outside, introducing the meaning of what one can find when getting out of one’s house (i.e. the word that lil is trying to find). As ste does not answer, lil stands up to depict again through a gesture the word of the space category she is searching for, in relation with tür (door, Image 9). This new attempt fails again: ste remains silent. lil makes a comment in Arabic to which ste responds by smiling (Image 10). As the method of depicting meaning through gesture does not allow lil and ste to find the searched for word, lil initiates again a move towards a new place in the room, and is again followed by the facilitator (Image 11).

In Extract 2C, the facilitator and the learner are no longer in front of the painting. As the painting is no longer in front of them and in order to find the searched for word, they depict actions and objects through gestures, knowing that these gestures refer to the painting, as well as to the word targeted by lil. ste first makes visible the action of drinking tea, and in answer, lil draws a space relying on its position in front of a house. Previous research in foreign language interaction shows that in some contexts, a gesture might take the place of a word during the construction of a turn to make visible the meaning of a missing word (Olsher, 2004). Participants here might have this experience. However, in this specific situation, using gestures does not allow them to find the word, or deduce a relevant meaning. lil is attempting to connect the meaning of her gesture with the name teahouse by pointing to a place in front of a teahouse, but this connection is not intelligible for ste. As in the first parts of Extract 2, the word search is lead by the learner, who can validate or not the propositions made by the facilitator.

The Learner Initiates a List of Elements to Deduce the Generic Word with the Facilitator

In Extract 2D (below), lil adopts a new strategy to make the facilitator understand the word she is searching for. Through pointing gestures toward paintings, she establishes a list of elements that can be found in the place she is trying to designate (Images 12–14). lil leads a movement towards two paintings and then points to a graphic element within them, getting ste to say baum (tree, Line 80). ste also represents the word with the same gesture as lil (Image 10). From this point in the interaction, lil leads the collaborative establishment of a list of two other words (Lines 84, 87), proposing a version of the word and getting a confirmation from ste (Lines 86 and 88).

Both ste and lil understand that this list of words should help ste to find the word lil is searching for. This is manifested by the fact that ste proposes “draussen” (outdoors, Line 92), which can be understood at this moment as a generic word encompassing the previous words introduced by lil. ste pursues her turn by proposing a sentence articulating the word she just proposed (draussen, outdoors) with another word selected previously by lil (baum, tree), Line 95.

However, this proposition does not match lil’s search, who moves towards her smart-phone (between Lines 96 and 97), opening a side sequence (not transcribed here) during which she tries to find the word on her mobile phone – an attempt which fails. After this side sequence, ste relaunches the word search (Lines 97–98) by proposing another generic word (Line 100), which is still not relevant for lil. After the proposition of other words, including another co-participant (P2) ver, the other facilitator, summarizes the previous attempts and proposes “garten” (Lines 119), which is positively assessed by lil (Lines 124–127).

During this interaction, the method lil uses to find the word she is looking for is to establish a list of other words that relate to a garden. She leads the word-search, she assesses ste’s various propositions and decides when the word-search can be closed. The facilitator organizes her participation by positioning her contributions as responses to the learner’s propositions of words and pointing gestures toward different paintings. The object of learning, the searched-word, is constructed in an interaction lead by the learner, who proposes a series of approximate pronunciations of the word thanks to which the facilitator could find the word/place she is searching for; lil moves across space to select relevant resources (paintings and graphic elements in them), and describes the word/place through multimodal turns (i.e. gestures depicting a space in coordination with specific places: hause, home, tür, door); lil establishes a list of words to reconstruct the generic meaning corresponding to the word she is searching for. We notice that her way of leading this word-search was possible thanks to the facilitator’s sensitivity and alignment of her turns to the learner’s initiations, and proposing a large amount of words.

The construction of the learning object that can be observed in Extract 2 is not that the most competent speaker leads the activity but rather that the speaker who has something to say leads the activity. This organization of the learning activity is made possible in the frame of an activity organized on the basis of a non-expected-answer instruction: it can be categorized as learner-led and teacher-followed format. Finally, we can also notice that the material world in which the interaction occurs has consequences on the organization of the construction of the object of learning. The availability of a series of paintings provides the participants with resources to build multimodal descriptions and share word meanings through pointing gestures and sentences. The paintings afford a graphic background to produce and interpret sentences. The possibility to move in the museum room is also an important resource to open new opportunities of constructing the object of learning and creating a zpd. In Extract 2, the learner can open a zpd by connecting elements of her knowledge (i.e. the word she is looking for) with graphic elements available on the paintings thanks to the interaction with the facilitator.

Comparison of the Two Interactional Organizations

In the first extract, the learner cannot at any time rely directly on personal elements to participate in the construction of the learning object relevant to the activity in progress. She certainly mobilizes knowledge related to the way of organizing participation in a classroom, but as far as the construction of the learning object is concerned – i.e. the formulation of a sentence in French – her interaction with the teacher does not allow her to identify the relevant criterion to accomplish the expected task. In the terms of sociocultural perspectives, the learner cannot connect any personal element to participate in the construction of her learning object or lay the foundations of a zone of proximal development. Symmetrically, the teacher directs the activity unilaterally, only gradually taking into account the learner’s interpretation problems. We are dealing here with an organization of the interaction directed by the teacher and followed by the learner in which the learner cannot open a zpd to build new linguistic knowledge.

In the second excerpt, on the other hand, the learner can rely on personal elements to guide the activity and open a zpd by connecting elements of her knowledge (i.e. the word she is searching for) with new elements (e.g. graphic elements observable on the paintings, descriptive gestures), producing an active stance towards the activity and her learning object which does not only concern linguistic elements, but also cultural elements related to the paintings in the exhibition. She can mobilize a whole set of resources showing her appropriation of certain paintings on which she can invoke elements to serve her word search. In this situation, the learner is in control of the activity. She directs the construction of an experience of the world and of a learning object based on criteria that she chooses herself. Symmetrically, the teacher follows the learner’s initiatives and tries to respond to them. In the terms of sociocultural perspectives, the learner and the teacher lay the foundations of a proximal development zone from which the learner can adopt an active stance in the construction of her learning object, including the accomplishment of multimodal actions.

These analyses provide an empirical basis for the assertion that, in order to organize a learning situation that is beneficial to the learner, it is necessary for the learner to be able to actively participate in the construction of the learning object. If we are now considering an international organization that wishes to promote lifelong learning for individuals with various sociocultural profiles, one strategy may be to mention the opposition between the categories of leader and follower and to encourage learning situations in which the learner takes the position of leader.

Conclusion

The article provides elements for understanding how a zpd is observable in social interaction by analyzing how information is shared in the interaction between learner and more experienced others, for instance, a teacher. To understand whether a zpd is being implemented, it is crucial to know the extent to which the learner can rely on a piece of knowledge as a basis for interacting with their more competent partner. The reverse is also true: it is also necessary to analyze to what extent the more competent partner takes into account what the learner knows in order to accompany him/her in the accomplishment of an activity that is more complex than his/her current knowledge or experience would suggest. The analysis presented has also highlighted how the establishment of a zpd is linked to bodily actions and the possibility of mobilizing gestures, movement in space, graphic elements, etc.

What these situations have in common is that the reconstruction of the learning object is led by a participant. In both cases it is the participant who expects a specific item that directs the interaction but in the first case it is the participant who teaches that leads the activity while in the second case it is the participant who learns who leads the activity.

What distinguishes these two activities is not so much the pre-existing categories found in the opposition between formal and non-formal learning but rather the organization of the interaction between the learner and the teacher and the roles of leader and follower in the interaction. What the study presented in the article shows is that there is no automatic relationship between the categories learner/teacher vs leader/follower. A learner can be the leader of a learning activity, and the teacher the follower. I have shown that if the learner is in the position of leader in the activity, he/she can connect his/her knowledge with new elements in the interaction with the teacher and progress in a zpd. In the case where the learner is in a follower position, this connection is problematic, as is the opening of a zpd. However, the study presented here does not assert that this dialectic is systematic. Moreover, I have not consider situations in which the participants were peers, nor have I considered situations in which the participants were in groups larger than two. That said, the question of who leads the activity and who follows it remains relevant in these situations as well. Finally, it should be considered that being the leader or follower is not a position that is defined once and for all. These roles can be reversed.

We saw that a defective interactional organization in a learning situation can prevent the learner’s progress. It therefore seems crucial to take into consideration the interactional aspect of learning in order to qualitatively orient educational policies, one of the objectives of the oecd (oecd 2021). Consequently, it is important that the description of learning that underpins public policies takes into account its interactional dimension and highlights the categories “learner or teacher as leader or follower of the learning activity”.

Understanding these details of social interaction can have an impact on teacher training, allowing for instance teachers to identify ways to involve learners during the construction of the learning object. The teacher’s sensitivity to the learner’s initiation of actions during the learning activity can also be a way to avoid hegemonic processes (Bagga-Gupta 2018) which typically may result in the exclusion of specific points of views during the construction of the object of learning. Recent projects funded under the Erasmus program of the European Union follow an approach taking into account social interaction, (e.g. cooba project 2020) developing ways of conceiving pedagogical activities and teacher training focusing on the sensitivity of who leads or follow learning activities.

This article addresses the problem of the in situ organization of participation in learning situations. By placing the analytical cursor at the level of social interaction, it is possible to identify how inclusion/exclusion dynamics can be played out in teaching/learning situations. In an era where many issues are at stake on the scale of global mobilities, it is by paying attention to the way individuals mobilize their experience to organize their social interactions and learning that international institutions will be able to encourage inclusive societies and lifelong learning.