Material Use in Collaborative Dialogue by Japanese University Students Learning Future Tenses in French as a Foreign Language: A Discourse and Interaction Analysis

This article reports on the use of an experimental material to enable discrimination between the periphrastic future (PF) and simple future (SF) intended for initial level university students in Japan studying French as a Foreign Language. A dyad was filmed using the material in a task on tense choice in short dialogues modelled on Galperin’s procedure. A discourse and interaction analysis were then performed on excerpts of the videoed session where the participants interacted with the material. It was found that the material enabled the dyad to complete the task. Moreover, if the pragmatic effect of the SF was not fully grasped, the analysis nonetheless showed that the participants transformed the linguistic knowledge from the material into metalinguistic resources of their own


Introduction
The present article reports on the analysis of the use by first-year university students in Japan of original material for Future tenses in French. This study seeks to answer two needs: reconciling a crucial domain of grammar-tenses-with language use and documenting how the material designed for this very purpose was used.
For students studying French at university in Japan, grammar is, indeed, essentially a set of explicit rules unconnected with any situation of language use. In the first-year students' schedule, grammar and communication are, significantly, split into two courses without any coordination most of the time: a grammar course and a conversation course, of 90 minutes each, over 15 weeks, resulting in a total course time of 90 hours over the year. In the textbooks for the grammar course, contents are presented from simple to complex, in isolation, and through guidance of metalinguistic descriptions in Japanese based on grammatical terminology from first-language education in France (Delbarre, 2014). Next to the grammatical presentations, the exercises focus exclusively on morphological features. The two main Future tenses are no exception. For instance, students will not learn about the rationale of choice between "votre valise (your suitcase) va tomber" (go-PRS.3SG fall) and "votre valise tombera" (fall-FUT.3SG). Let us think of a simple task involving this choice, as in the following: Vous êtes monté(e) dans le train: -Attention, votre valise va tomber ! tombera ! You have boarded the train: -Watch out, your suitcase go-PRS.3SG fall! fall-FUT.3SG! In the example above, the so-called Go-Future "va tomber", known as periphrastic future, is the only possible choice, because, as we will see, only this tense fits with the understanding the speaker has of the situation. This is an example of what the material for this study was designed for: enabling learners to choose the appropriate tense in various contexts.
The second need we seek to answer is to report on the actual use of such material during an experimental task, which is made of four others of these dialogues with tense choice. Indeed, as Tomlinson and Masuhara (2018) note, "we can still find very little literature on what students actually do with materials" (p. 360). In any case, no example was found of any qualitative analysis of experimental materials that assessed the materials' efficiency within a task in real time. The goal of the present study is therefore twofold: Determining whether our experimental material enables students to discriminate between the periphrastic future (PF) and simple future (SF) and documenting how learners actually use it during the task.
The conception of the material, the implementation of the task-a "collaborative dialogue"-as well as the analysis of the discourse arising when the learners use the material during the task draw on a Vygotskian approach to second language learning. Moreover, for documenting how learners adapt to their continuously evolving environment during the task, we will further support our analysis with elements from the sociocognitive approach (inter alia Atkinson, 2011). This framework is introduced in the next section before presenting the study itself.
SCT considers learning from the perspective of Vygotsky's theory of mental development (2012 [1934]). Two key elements of Vygotsky's theory are of interest in our case: 1) the internalization of concepts -which, in speech, are the meaning of words -and their conscious use by the individual to control her mental activity (concepts "mediate" thinking), 2) the development of concepts in the course of an activity with a more experienced individual, this development covering a zone ("zone of proximal development" in Vygotsky's terms) which varies, in particular, according to the age of the individual and to the domain of the activity.
Some scholars in SCT have brought up to date Galperin's teaching procedure, which specifically draws on these two key elements. In Vygotsky's footsteps, himself also a specialist in mental development, Galperin was specifically interested in improving teaching methods and materials (Haenen, 2001).
Galperin was the promoter of a teaching procedure that aims to guide learners in their problem-solving activity by providing them with a schema detailing the concept that would be useful for carrying out the activity. This orienting schema can be a chart, a diagram, or a visual where the action that the concept achieves is broken down into its parts. Then, the action represented on the card generates oral exchanges during practicethe verbalization phase -and is eventually abbreviated into an action on the mental plane (Galperin,1967(Galperin, , 1989(Galperin, , 1992; see also Arievitch & Haenen, 2005;Haenen, 2001). In Galperin's procedure, conditions for internalization to occur are specifically recreated for educational purposes. According to Gaperin (1989), an advantage of his procedure is that learners avoid the "trial and error" phase, which is costly in time. Another advantage is that the orientating schema avoids the effort of initial memorization: learners unintentionally memorize concepts using the schema in the course of the activity (Galperin, 1989: 70-71).
In L2 acquisition, Galperin's procedure is known as Concept-Based Approach (Lantolf, 2011) or Concept-Based Instruction. The target concepts of the L2-for example, tense-aspect (Gánem-Gutiérrez & Harun, 2011;Negueruela and Lantolf, 2006), phrasal verbs (Lee, 2016), irony and sarcasm (Kim, 2013), tu vs. vous in French (van Compernolle, 2011)-are represented in diagrams or schemas, then explained and used in communicative activities. The experimental task for testing our material is based on these principles, although Lantolf (2006Lantolf ( , 2011's proposal to rely on Cognitive Linguistics will not be followed. Our analysis will focus on the verbalization phase, i.e. the immediate circumstances of the use of the pedagogical material and on the discourse and actions it arouses. Swain (2000) rightly approaches the verbalization phase of Galperin's procedure with her own notion of "collaborative dialogue" (pp. 104 et sq.), which she defines as "linguistic problem solving through social interaction" (p. 104). In collaborative dialogue, language is a communicative tool at the service of achieving a goal. However, when language takes as its object the language used for a communicative purpose, it also becomes a cognitive tool (Swain et al., 2002). Later, Swain (2010) will call "languaging" the activity in the course of problem-solving which consists in discovering elements of understanding by the mere fact of putting the problem into words. Finally, Swain and Lapkin (2013) report several studies on how L1 mediates L2 tasks. When the L1 serves to focus attention when the task becomes cognitively challenging, these authors envisage the L1 as scaffolding.
In Vygotskian research in L2 acquisition, the analysis of verbalizations follows Vygotsky's genetic method. This method aims to document the integration of new concepts into the individual's thinking process (Lantolf & Thorne, 2006). Since the analyst seeks to describe changes on short time scales and which occur in real time, this method is also referred to as microgenetic (Ortega, 2008: 224). However, in the literature on Concept-Based Instruction, collaborative dialogue or L2 gesture research (McCafferty & Stam, 2008 for instance), emphasis is not put on local interactions with the material environment of the task, particularly with the pedagogical material. While pedagogical materials are certainly not the main focus of collaborative dialogue analysis, it is of decisive importance in Galperin's procedure as well as in the Concept-Based Instruction. A discourse analysis of the verbalization phase -a collaborative dialogue in Swain's sense -must account for any phenomena relating to the use of the concepts explained in the material (their integration, in the best-case scenario). In L2 acquisition, Atkinson (2011Atkinson ( , 2013Atkinson ( , 2019 and colleagues (Atkinson et al., 2007;Churchill et al., 2010) propose a holistic framework, named "sociocognitive", which aims to account for the creation of meaning through the course of action.
The participation framework (Goodwin, 2003) is a tool in this approach which we believe is essential to complete the analysis of a collaborative dialogue. Goodwin defines "participation framework" as the participants' joint engagement in an action. When they focus their attention to build an action, the participants mobilize in addition to speech different semiotic resources such as gesture, gaze, and bodily orientation, in combination, in some cases, with material artifacts in their environment. Atkinson (2011) further argues that, during its use, pedagogical material functions as "extended cognition", as defined by Clark and Chalmers (1998). Clark and Chalmers' thesis is that most of the time cognitive processes take place by delegating tasks to artifacts specially designed for this in our environment.
Since they focus on the moment-to-moment organization of interaction, as does the genetic method, these elements from the sociocognitive approach will be integrated to account for the use of the material and the analysis of the discourse it triggers.

Objectives
In light of these theoretical and methodological principles, an experimental task consisting of five short dialogues was carried out by two dyads of learners at the initial level. In addition to the dialogues, each pair had the material (a worksheet, presented below) that we wanted to test, developed according to the principle of the "orientating schema" in Galperin's procedure and in the Concept-Based Instruction. The sessions were videoed for analysis. Our goal is to answer the following two questions: 1. Does our experimental material enable students to discriminate between the Periphrastic Future (PF) and Simple Future (SF) in those dialogues? 2. How do learners use it during the task?
This section will provide the details of the study, starting with the linguistic aspect of the material.
The Contrasting use of the periphrastic future and the simple future First, three theoretical accounts of the contrasting use of the PF and SF are provided below. These descriptions were selected primarily for the value they have in terms of their adaptation to the Foreign Language classroom. Gosselin (2005) suggests that a verb in the indicative mood draws a division between the irrevocable and the possible, which he terms "modal cut" ("coupe modale"). Gosselin explains that the living present is characterized by a conversion of the possible into the irrevocable, and the main property of the tense system is to sustain the simulation of this conversion at another point in time (2005: 102). In Gosselin's terminology, this modal cut corresponds to a reference interval I-II and is more precisely operated at bound II (2005: 91). Table 1 below represents the PF and SF in the Reischenbachian stylization used by this author.

Table 1
Reischenbachian models for the periphrastic future and simple future, adapted from Gosselin (2005) PF (periphrastic future) "je vais attendre" SF (simple future) "j'attendrai" Note: "01-02" refers to the speech interval, when the speaker uses the language; "B1-B2" is the process interval, which refers to the time of the situation being evoked; the reference interval "I-II" is the moment at which the process is seen/imagined.
The distinction between the PF and SF becomes clear: with the PF, the process B1-B2 is in the possible, to the right of bound II, but with the SF, the process B1-B2 remains in the irrevocable, to the left of bound II. 1

Related to (PF) vs. detached from (SF) the situation of uttering
The analysis provided by Camussi-Ni (2018) is slightly different, as it is based on the situation of uttering. 2 First, this author observes that in French, both the non-finite verb and the SF refer to a "potential" process (as opposed to a "confirmed" process). The potential value of the non-finite verb is due to its absence of actual reference to the situation of uttering. In contrast, the SF draws this value from its reference to this very situation of uttering. Hence, both the PF and SF carry on a potential value in the non-finite verb and in the verb predicated on a grammatical subject, respectively (Camussi-Ni, 2018: para. 7-17). However, the distinction between the two is based on the following: PF: The auxiliary verb is in the present tense. For Camussi-Ni, the present tense construes a coincidence between the process and the situation of uttering. The basic meaning of the PF in, for example, "ils vont partir [they go-PRS.3PL leave]", is that "a potential situation (partir [leave]) is related to a situation that coincides with the situation of uttering (ils vont [they go-PRS.3PL]), itself constructed by the speaker." (2018, para. 22, our translation) 3 SF: Camussi-Ni argues that the SF has undergone a repositioning within the tense system. Considering that it shares with the simple past ("passé simple") the same morpheme [a], Camussi-Ni claims that similar to the simple past, it is also detached from the situation of uttering (2018, para. 23-30).

Predicting from a shared ground (PF) or authoritatively (SF)
Bottineau (2014) focuses more directly on the material scene of linguistic interaction and aims at accounting for how participants coordinate their thoughts through their speaking. Bottineau argues that the relevance of the so-called enactive approach in foreign grammar instruction is to familiarize learners with the fact that grammatical choices shift the participants' position within the environment of the interaction. Hence, the contrast between the PF and SF touches on the attitude of the speaker toward the other participants in relation to the event that is being evoked. With the PF, the speaker invites the listener to share the ground from which the future event is contemplated. On the contrary, the SF excludes the imaginable futures of the other speakers, by authoritatively and performatively deciding on one unique future for all (Bottineau, 2014: 197).

Material
Our material is designed to function as does the "orienting schema" in Galperin's procedure. It must not be based on translation in equivalent verb forms in the L1. On the contrary, it must provide a class-friendly and theoretically coherent instruction of use, so as to promote the internalization of the rationale of choice between the two tenses.
The oppositions between possible/irrevocable (Gosselin) and related to / detached from the situation of uttering (Camussi-Ni) appear transposable in an opposition between the internal and external space. The internal space would be the speaker's bodily space (or of the subject to which the verb is predicated), which would correspond to the speech interval and the situation of uttering, respectively. Bottineau's approach adds an essential dimension to these distinctions by considering how the choice affects the interactional configuration. This differentiation would be mapped onto the opposition between the speaker's bodily space and an external space.
The material-a worksheet-created out of these accounts is reproduced in Figure 1 below.

Figure 1
Material for discriminating the periphrastic future and simple future.
The title of this worksheet, "近接未来形 / 単純未来形" "kinsetsu miraikei / tanjun miraikei" ("close future form / simple future form"), uses the conventional terms found in grammar textbooks. 4 Then, both tenses are illustrated with the visual of a woman standing within a circular space that represents her perceptual present. For the SF, the word "figé" ("frozen") is superimposed, as if it were stamped, onto the space of the present, and an arrow points toward a zone marked as the future. The space ahead is indeed a conventional representation of the future in French speakers' gestures (Calbris, 1990: 88).
The visual elements and the explanations are the results of our effort in tailoring the theoretical accounts to the context of use. The circle that stakes out the present around the woman is meant to match the notions of speech interval in Gosselin's (2005) analysis and that of the situation of uttering in Camussi-Ni's (2018) work. The notion that with the PF, the potential action occurs within the situation of uttering is taken over nearly word-forword ("L'action future est dans le présent.").
To highlight that the tenses differ in the way they sustain the emerging stance of the speaker and listeners, we opted to emphasize the peculiar pragmatic effect of the SF with the metaphor of freezing. This metaphor is meant to suggest that the authoritative choice the speaker is making in predicting one unique future, as Bottineau (2014) argues, is a constraint which limits every participant's movements. In our view, it seemed important to stress the fact that this constraint is felt in the present. The present is frozen, as it must conform to a unilateral future. Finally, the notion that "L'action future est après" ("The future action is after") refers to the detachment of the event from the situation of uttering, as described by Camussi-Ni (2018).

Participants
At the start of 2020, the four participants of this study were recruited in the author's institution via announcements in classes. They were explained the goal and design of the experiment. They provided their written and informed consent for the use of their anonymized audio-video recordings and written data, collected during the experiment, for research purposes. In full accordance with the standards of my institution, they received financial compensation for their participation in the three experimental testing sessions. Among the participants, three were Japanese, and at the end of their first university year. They have been named Mika, Hiroki, and Ryōta for the purpose of reporting in this paper. The fourth participant, who was Burmese, and at the end of her second university year in a four-year bachelor's program in English, has been named Kyaw. Table 2 below provides information on the participants' background in French language study, including the estimated rate of grammar courses taken by them. At the beginning of the first session, we conducted individual interviews with each participant to assess their acquisition stage (Bartning & Schlyter, 2004) in spontaneous oral production. I asked them to introduce themselves, to report on what they had done the day before, what they had planned for the day after, etc. A short picture-based narrative task was also used to elicit the use of past tenses. Measurements indicated that all participants were at stage 2 (post-initial), marked by the emergence of grammatical phenomena and an increasing number of finite forms (Bartning & Schlyter, 2004: 295). Individual differences seemed to be influenced by the number of study hours. Notably, Ryōta demonstrated lesser control over utterance structure and genre agreement. The four participants used the PF for basic reference to the future.

Corpus
The experimentation was implemented in four stages (approximate time in minutes in parentheses: 1) presentation (5), 2) pair-work (12), 3) sharing with the other pair (7), and 4) synthesis with the instructor (the author) (5). The four stages were videoed with two cameras. For constituting our corpus, the following inductive procedure was applied.
The videoed sessions were first browsed through, without preconception. At stage 2, the dyad with Ryōta and Hiroki proved to have a larger volume of exchanges and interaction with the material than the other dyad and was then selected. For each of the five dialogues, we further selected every sequence where the dyad interacted with the material. For the 10 excerpts constituted this way, speech, gesture and interaction with the material will be described in line with our objectives. "Interaction with the material" includes pointing or gazing at the worksheet (represented above in Figure 1), and every manipulation of it. "Gesture" is defined following Kendon (2004) as any visible bodily action that has "the features of manifest deliberate expressiveness" (p. 15).
The transcription conventions are as follows: \: falling intonation; /: rising intonation; ____: overlapped words; wo-: unfinished word; (()): transcriber's description; (n s): measured pause of n seconds; xxx: incomprehensible ("x" the number of syllables). In the body of the transcription, the numbers inserted refer to the screen captures-shown below the transcription-taken from the video footage at this very moment in the discourse. A translation of each excerpt is added in the right-hand column. We will first comment on the dialogue above to explain the choice of tense, so we can better judge the work achieved by the dyad. In situation 1, the speaker is spurred on by spontaneous empathy and warns the passenger. The falling of the suitcase is certainly not irrevocable (Gosselin, 2005); the potential action (Camussi-Ni, 2018) is in the speaker's present, and the suitcase's instability becomes an observation shared (in the sense of Bottineau's (2014) characterization) with its owner. Consequently, "tomber" in the PF, "va tomber", is the correct choice. Ryōta and Hiroki have put the worksheet 7 with the dialogues between them within their immediate field of vision. The worksheet of the material is placed just above it (Photos above). These two documents are the two foci of the "participation framework" (Goodwin, 2003). Ryōta and Hiroki's engagement in this participation framework can vary but will remain stable until the end of the task.

Situation
Here, Hiroki immediately chooses "tombera" and translates it with "ochiru darō" (l. 6). At this moment, Ryōta raises his head towards the material (Photos 2 to 3). In line 7, he engages again in the reading of the dialogue (Photos 3 to 4), then contests his partner's choice.
Hiroki's preference for direct translation illustrates well what can be termed the "think in L1 to speak in L2" effect, characteristic of students at initial levels (Pavlenko, 2014: 303). In Japanese, the future is the default reading of the -RU form of eventive predicates (Jacobsen, 2018); for example, "ochiru" ("to fall"). Hiroki adds the epistemic modal "darō", which can be loosely translated here as "will". However, Hiroki has obviously misunderstood the situation of the dialogue. Indeed, the speaker cannot establish the certainty of the fall of the suitcase.
In any case, the sequence of actions shows that Ryōta relied on the material to make his choice. In Excerpt 2 (immediately following), Ryōta repeats his claim, pointing to "va tomber" (Photo 1). Hiroki acknowledges his mistake, but with the slightest concession, observing that "it's not a divination" (l. 4)-he refers to one of the examples of use for the SF on the material. Sensing that his partner does not understand, from line 5 onward Ryōta further explains his choice: In line 5 (Photo 2), Ryōta reorients the worksheet of the material in his direction: he seems to adjust the means that his explanation requires, but these means are that which need to be explained, hence the hesitations. These hesitations from lines 5 to 10 show that he is looking for ways to reformulate the problem for the benefit of Hiroki. Ryōta first considered the use of the material and pointed to the SF column (Photo 3). However, he changes his mind and points again to the dialogue (Photo 4, l. 9). In line 11, he literally proceeds to an analysis of the utterance, resorting to reported speech, with the Japanese marker "tte iu", and noticing the punctuation. By focusing on the situation of uttering, he is able to-correctly-demonstrate how the utterance is anchored in the present.
However, in line 13, Ryōta goes further and impersonates the speaker, using again a marker for quoting, "mitai na" ("like"). The segment from Photo 5 to 6 corresponds to what Kendon (2004) calls the "preparation", preceding the stroke of a gesture. The stroke achieves the gesture's expressiveness (p. 112). Kendon's thesis (2004) is that gesture is an integral component in the construction of the utterance. Here, the stroke coincides with the verb "ochiru", and determines its spatial reference: the place of the object about to fall. Ryōta's gesture then accomplishes an "abstract deixis" (McNeil, 2005: 40) as the gesture builds up the target and surrounding environment in the absence of tangible objects. For McNeill (2005), gesture and speech develop the same underlying idea in two different semiotic modes: continuous and synthetic, discontinuous and analytic, respectively. This author envisages this underlying idea as a "growth point", the minimal psychological unit simultaneously unfolding its content in speech and gesture. However, beyond the complementarity between gesture and speech, a more holistic analysis of interactions must emphasize that Ryōta and Hiroki's environment has changed as the suitcase is imagined in it. For instance, Nevile et al. (2014) counts in the "interactional ecology of objects", the set of affordances-in the Gibsonian sense-of the objects in the environment in which one engages, the abstract objects created as supports of thought in discourse. Here, the suitcase is now present and mobilizable in thought, and its observation is shareable-the very value of the PF that Ryōta strives to illustrate.
Ryōta's next gesture makes sense in the "ecology" of this environment: He does not rest his hand but goes on with another gesture composed of two rapid strokes of the hand towards his bust (Photo 7). The sequence of the two gestures (Photos 6 to 7) demonstrates that the suitcase is in the same time as that of the body: this gesture towards the body intervenes on "ima" ("now"). Interestingly, only then does Hiroki turn his head towards his partner (Photo 7). Finally, Ryōta concludes his explanation with his hand now shaping into a pointing gesture on the column of the PF on the worksheet (Photo 8).
This rapid succession of pointing gestures in this excerpt performs a specific function. Indeed, the expressiveness of a pointing gesture is specific. Pointing is "a deictic gesture used to reorient the attention of another person so that an object becomes the shared focus for attention" (Butterworth, 2003: 9). Here, in the same participation framework-in Goodwin (2003)'s sense-, the consequentiality of Ryōta's gestures concretely builds up relationships between the words of the dialogue, the concepts from the material, the imaginary suitcase, the body of the speaker, and the tense of the Foreign Language, relationships that would otherwise remain abstract.
Situation 1 represents the first concrete use of the material. In the four next excerpts, Ryōta and Hiroki turn their attention to it, temporarily setting the dialogues aside.

9
(2,5s) (3) The key passage of this excerpt is the paraphrase of the PF by Ryōta in line 6. Although Ryōta just illustrated the specific anchoring in the present of this tense by impersonating the speaker, Hiroki points to some terms of the material still problematic for him and read them (Photo 1). After Hiroki, Ryōta points to the same terms, reads, and comments on them (l. 4 to 6. Photos 2 to 3). Ryōta's paraphrase uses the grammatical means of the Japanese language to render the aspectual value, first with the adverbial "masani" ("right now") and then with the complex form "-yō to shite iru", which expresses imminence (as in the periphrastic phrase "to be about to").
What occurs here is, we argue, what Swain (2010) calls "languaging", the fact to bring about new understanding by talking through a problem. Indeed, talking entails the externalization of ideas into words, which in the process brings up associations that one did not think of before. In our case, the material does not provide a translation. Then, when Hiroki asks for an explanation, Ryōta answers but his search for equivalence in L1 is new to him as well (l. 6 is a question). The exploration of the material continues in Excerpt 4. In excerpt 4, Hiroki asks the same question, this time about the SF (Photos 1 to 2, l. 1 to 5). Photo 2 shows him reorienting the worksheet. In lines 9 and 11, Ryōta also reads the description of the SF. The metaphor of freezing stands for a present limited by the speaker's unilateral decision. In Lines 12 to 14, in "languaging", Ryōta paraphrases the explanation written on the worksheet. As he does so, he reuses a kanji that first was used in the material, "kata-" in "katamaraseta" ("hardened", see section 3.3.), and incorporates it with its Chinese reading "ko" into the term "kotei" ("fixation", l. 14). This kanji is framed in green in the transcription and in Photo 5, which is the worksheet that was on the table. However, it is the metalinguistic commentary line 12 that inspires Hiroki. Indeed, a few seconds later, Hiroki exclaims, "Ah, I see, something that's been decided, sort of", using the term "kakutei" ("decision, determination") provided by his partner. At line 17, Hiroki starts writing down "hobo kakutei shita koto" ("the fact of being nearly decided") onto the material next to the grammatical description (Photo 5, in red). The last link in this spontaneous web of signs is "tei", for the idea of "regularity", which appears both in "kakutei" and "kotei" (in yellow). This sequence will prove crucial for the remainder of the task: through interaction with the material, new signs-which, in Vygotskian terminology, mediate concepts-emerge, are recombined in discourse, and, in the case of "kakutei", hand-written back onto the material.
However, if this term, "kakutei", Hiroki's choice for "decision, determination", as a characterization for the SF is compatible with Gosselin (2005)'s value of irrevocability, it is less so with the negation of the choice of others, as in Bottineau (2014) This description, "L'action future est après" ("The future action is after"), is correct from a formal point of view, but puzzles the dyad, hence the laughter (l. 8). From Photo 2 to 3, Ryōta wonders about "after", and then he points to the words on the worksheet with a comment whose meaning, unfortunately, escapes us. It is true that the effect of the SF is more abstract than that of the PF, which is anchored in the present experienced by the speaker. However, "the fact that it runs out" (l. 11) and "a sensation" (l. 13) remain vague from the analyst's point of view. The characteristics Ryōta sought to describe cannot be known.
Hiroki cuts short these speculations and concludes the first dialogue:  (3) There's still a way to stop it.

11
(2,3s) Slightly replacing the worksheet, Hiroki redirects the discussion to the solution that was chosen, the PF (Photo 1). In line 6, Ryōta reuses the terms he previously used in excerpt 3, "it's about to fall". Similarly, as he points to the column of the SF (Photo 2), Hiroki reuses the metalinguistic term he appropriated: The fall of the suitcase is not confirmed or decided indeed ("kakutei shitenai", l. 9). The following comment is indicative of a good understanding of the value of the tense: "There's still a way to stop it" corresponds to the modal value of the PF-process envisaged as possible-as described by Gosselin (2005, see section 3.2.).
Taking Excerpt 1 as a point of comparison, the understanding achieved by Hiroki is noteworthy: He now has metalinguistic terms adapted from the material and provided by Ryōta, sufficient-from his point of view-to discriminate between the two tenses. We argue that the dyad is indeed building up "cognitive tools", that is, means to orient the thinking process involved in L2 use (see Swain, 2000;Swain et al. 2002;Swain and Lapkin, 2013). Here, the moment of the appointment is in no way related to the situation of uttering: I am already with my friend. Moreover, the irrevocable quality of the future situation comes from the fact that my information limits to one place the possibility for my friend to find me. The SF, "serai", is the most appropriate tense.

H 3'20 mmh Mmh
The dyad quickly finds out the solution. The sequence of the three photos shows how Hiroki proceeds after Ryōta poses the alternative between "serai" and "vais être" (Photo 1). Hiroki glances at the column of the SF (Photo 2) and of the PF (Photo3) and then offers the correct answer. As a justification, Ryōta puts forward a comparison with English. This point will be discussed later because comparison with English is not an adequate means. In any case, Ryōta refers to the assumed epistemic value of "will" to choose the SF, using the term "kakushin" ("conviction, certainty") close to "kakutei" (l. 8) The dyad then engages in an exchange which we have not been able to transcribe in full. The comparison with a "memory" (l. 9) can only be understood from Hiroki's point of view. However, there is indeed with the SF a detachment, in the sense of Camussi-Ni (2018), of the future situation from the situation of uttering, as there is in the direction of the past with a memory. When one consults the calendar, it is not its cultural artifactuality which is looked at. The time that is measured seems that of a natural order. Here, the person who asks the question puts himself or herself in a position to be reminded of this order, by definition non-negotiable: the SF, "sera", is the most appropriate choice.
The sequence of Photos 1 to 2 shows what Hiroki accomplishes: In line 4, he first responds with the term he gave himself to characterize the SF, "kakutei", and only then does he look on the worksheet at the column which matches with the tense (Photo 2). This means that the concept, in the Vygotskyan sense, becomes functional. The concept of "kakutei"-idiosyncratic, which raises other questions-is being internalized: Hiroki is making of it a mental tool for choosing the tense.
The dyad continues along the path of the SF. Photo 3 shows Hiroki reorienting the worksheet, readjusting the participation framework by doing so. In Photos 5 and 6, both have a hand on the worksheet of the material, which becomes a shared focus of attention. Ryōta exploits an analogy-which is correct-between the calendar and the example of the weather forecasts on the material, while also showing its limit ("if you can say that", l. 14). In line 15, he uses, "kakutei", the term used by his partner, but that he created (see Excerpt 4). While ticking "sera" (Photo 6), Hiroki adds a comment: "something that can't be changed". In the next dialogue, he will make the distinction of "can / can not be changed" a new cognitive tool. Finally, Ryōta uses the original term for the SF from the worksheet. Contrary to the relentless order of the days on a calendar, the future presence of the two persons on the train depends on their present behaviour. Since missing the train is only a possibility, only the PF is suitable here: "on va rater". The dyad reads the dialogue by translating it (l. 1 to 10). Here, a mistake is made: 17:15 is understood as the time of arrival of the train. This does not change the stakes: The two persons must hurry to be on time.
Hiroki immediately proposes the correct solution, but Ryōta opts for the SF. The whole sequence shows a series of five almost identical visual interactions during which Hiroki is turned towards the worksheet of the material (Photos 2 to 3, 5 to 6, 8 to 9, 11 to 12 and 14 to 15), whilst arguing for the PF. If it is not easy to determine which part of the material he is looking at, the sequence is nonetheless evidence that Hiroki relies on the material to produce his speech. Galperin (1989) writes about the advantage of using "orienting schemas" (or just "cards" in the following quote) to lessen the burden of prior memorization: "The cumbersome content of the cards is learned unexpectedly easily in action in the process problem-solving. The capacity of involuntary memory is much greater than the capacity of voluntary memory [...]. Furthermore, [...] this involuntary memory can be guided quite reliably and even accelerated considerably" (p. 70). We argue that this is what we are witnessing here, at least in the case of Hiroki. The worksheet of the material becomes a landmark for regulating his thinking process and speech (for instance l. 16) and functions, we argue, as an external memory, drawing on Galperin's (1967Galperin's ( , 1989Galperin's ( , 1992 own accounts of his procedure, but also as a "cognitive extension" (Clark & Chalmer, 1998: 12). Indeed, the ambition of developing material in Concept-Based Instruction is, in our view, that the design is based on an assumed continuity of cognition within the environment, as argued by the proponents of the extended mind thesis. 8 Finally, Hiroki concretely illustrates the modal value of possibility of the PF with a further comment in the form of an if-clause, "if you run, you can make it" (l. 19).
A fundamental point is the absence of a pointing gesture-but not of interactions. We interpret it as a tendency to "the abbreviation of action" in the sense of Galperin (1967Galperin ( , 1989Galperin ( , 1989. Hiroki is, it seems, able to recognize in the different configurations (settling an appointment, checking the date, deciding to hurry...) attributes pertaining to the relation with the present particular to the PF or the SF. 9 The notion of a situation of uttering detached from the present captures well the stakes of the choice here. The little girl and I both know that, for a child, growing is an implacable destiny: "seras" should be preferably used. As in the previous excerpt, the dyad no longer points to the material and just checks the analogy with the situation by looking at the worksheet. First, translation leads Ryōta and Hiroki towards the right solution, the SF (l. 4). The first visual interaction with the worksheet occurs a few seconds later (Photos 2 (Hiroki) and 3 (Ryōta)). In line 9, Ryōta takes a look at the material (Photo 5) and then returns to the dialogue (Photo 6). At this moment, Hiroki also turns to the material and verbalizes the link between the content of the dialogue and the value of the tense by resorting to his metalinguistic tool, "kakutei" (l. 14), which allows the dyad to conclude (Photo 8).

Discussion
Ryōta and Hiroki found the 5 answers. For situations 1 and 4, the anchoring in the present of the future event-the imminent fall of the suitcase, and the departure of the train-was associated with the PF. An intrinsic sense of limitation in situations 2, 3 and 5, where reference is made to an appointment, the calendar, and the child's age, oriented the dyad towards the SF. It should be noted that here the PF can also possibly be used 10 even if the SF is preferred. However, the positive outcome of the whole task is obviously limited in its generalizability, since it is valid only for these dialogues and under the conditions of this experience. For instance, the use of the SF for attenuated order was not tested. Above all, learners with another L1 and in regions of the world where the culture of grammatical instruction is different would have carried out the same task in a completely different way.
Also important, perhaps more so, is the answer to our second question: How was the material used? Two answers emerge. First, the analysis shows that Ryōta and Hiroki developed their own resources in interacting with the material, stimulated by the goal of completing the task. Ryōta, in "languaging" (Swain, 2010), uses "kakutei" while paraphrasing the explanation of the SF from the material (Excerpt 4). Hiroki takes up the word as a "cognitive tool". This does not contradict the claim that L1 happens to mediate the learning of the L2 (see Swain & Lapkin, 2013), quite the contrary. However, to promote real multilingualism right from the beginner level, displaying the target forms in a way that they remain irreducible to translation should be emphasized. The second point is that, with hindsight, a movement of internalization becomes visible through the task, in accordance with the design of Galperin's procedure. Clear evidence is the regularly decreasing volume of gesture: from Ryōta's impersonation in the first dialogue and the many pointing gestures on the material to the simple glances to it for the last two dialogues. Further studies would be useful to report on subsequent states of internalization. The use of the material, however, has not always been smooth. The main problem is the limited understanding of the value of the SF as described by Bottineau (2014), while this description is the most stimulating in the context of a Communicative Approach. The solution would be to represent, by a means that suits the teacher, the inclusion of the other participants in the choice of the perspective on the event with the PF, and their exclusion in the case of the SF. The choice of tense could also be addressed in a situation where the speakers' communicative goals are opposed in relation to the same event. For example: Votre ami(e) : On va rater notre train ! Your friend: We'll miss our train! Vous : Eh bien, on le ratera ! You: Fine, we'll miss it! 11 Another problematic point was the implicit use of the contrast between "be going to" and "will" in Excerpt 7. For learners, the temptation of the analogy is great as both English and French have a Go-Future. As a result, the SF is considered close to "will". The fact is, without going into details, that the PF is mostly translated as "will" (Celle, 1997), as in the example above. Mika and Kyaw, in the other dyad, wasted considerable time due to this erroneous analogy with English.
Regarding our theoretical and methodological options, it seems to us that a holistic perspective for the analysis is fully justified to integrate discourse and interactions with the material, unlike a sole focus on gestures (as in McCafferty & Stam (2008)) which would not have been compatible with our goal to account for the use of our material. An example that illustrates this point is the case study reported on in Lantolf (2010), based on McNeill's "growth point" model. An English L1 student is interviewed by the investigator on her choice of tenses in a narration task in French L2. The author assumes that the student's gesture is a depiction of the schema provided in class four months earlier (Galperin's procedure was followed). This intuition is not without interest, but it seems necessary to us to document how such schemas are used in the first place, as we tried to do. The analysis of gestures in the SCT for L2 adopts McNeill's "growth point" model. However, accounting for the interaction with objects goes beyond the relationship between speech and gesture. Manipulations of objects further change the conditions in which subsequent gestures occur by modifying the affordances in the environment (Nevil et al., 2014). About McNeill (1992)'s work, Goodwin wrote: "However, in that its analytic point of departure is processed inside the mind of the individual speaker/gesturer, this approach does not provide the resources necessary for investigating how phenomena outside the speaker, for example, a consequential physical environment, contribute to the organization of gesture" (2003: 23). This observation is still relevant in our opinion and justifies documenting the learning process in Galperin's procedure in a holistic theoretical framework given the importance of the material in this procedure. For this reason, the methodology of Atkinson and colleagues in their sociocognitive approach was used as a model in this study.
Beyond the inherent variety of levels, domains and designs, our study also converges with others on the fact that Concept-Based Instruction can contribute to a better understanding of L2 concepts and promote their conscious use in context (Gánem-Gutiérrez & Harun, 2011;Kim, 2013;Lee, 2016;Negueruela and Lantolf, 2006;van Compernolle, 2011). The present study, however, addressed the use of the material in real time, and the data obtained distinguishes it from these. A manifestation of the internalization process could then directly be inferred from the observation of a reduction in the volume of gesture and manual interactions with the material.
Another level of comparison is the theoretical perspective chosen to describe the target concepts. For example, Negueruela and Lantolf (2006), Gánem-Gutiérrez & Harun (2011) and Lee (2016) rely on internal representations of the event from the speaker's point of view regardless of the situation of uttering, much in the spirit of Cognitive Linguistics-an option suggested by Lantolf (2006Lantolf ( , 2011. In our study, however, and with the reservations expressed above, it is from the angle of its pragmatic effect that we tried to approach the contrast between the PF and the SF, a choice in this respect similar to Kim (2013) and van Compernolle (2011).

Conclusion
The initial idea of the present study was to reconcile grammatical instruction and language use in French at initial levels of university in Japan. Galperin's procedure seems to us to be well suited to this challenge in this specific context. According to his inventor, this procedure has, as mentioned, the advantage of avoiding the trial-and-error phase. Even if imperfect-there is not enough room here to imagine all the possible improvementsthe material tested helped learners, even beginners, to take, at least to imagine, the perspectives on the situation that are specific to the two Future tenses in French. Grammatical instruction should be carried out, obviously, in parallel with sessions focused on oral communication, where for fluidity dialogues are written and role-played. We also believe in the virtue of collaborative dialogue that promotes autonomy.
Moreover, Galperin's procedure also seems to us to be suitable for the testing of material, and the work, perhaps still not well known, of this continuator of Vygotsky can give rise to an important reflection on teaching materials. Indeed, it is rare to see learners perform a learning task without consulting resources. On the contrary, in the course of a task, simple or complex, learners will use all means available to achieve their goalcompleting the task, sometimes misusing these resources. It is the role of the teacher to provide resources that help learners to complete the task. From a Vygotskyan perspective, however, it is the internalization of the resources that is, for the teacher, the real goal.