MULTILINGUAL INTERACTIONS IN A COLLABORATIVE SITUATION : THE EXAMPLE OF HIGH SCHOOL STUDENTS IN ST . MARTIN

This article aims to explore the bilingual practices (English and French) used by students from St Martin in a collaborative situation, in order to enrich the debate on languages and knowledge. This research is based on two corpora consisting of interactions between high school pupils during their educational support sessions, and which are studied using conversational analysis; the study uses this combination of collaborative methods and bilingual education to model the management of linguistic resources with regard to the factual knowledge and practical skills brought into play during the interaction. The challenge of this study is therefore to establish what is involved, and under which conditions and through which mechanisms the interactions between two languages (English and French) may contribute to the construction of cognitive learning skills.


INTRODUCTION
The aim of this study is to discover how the pupils use their languages, and to evaluate the impact of this usage on collaborative tasks.In the context of this research, we present extracts from the interactions of pupils in the top two years of high school as they carry out collaborative tasks in an upper secondary school on the island of Saint-Martin, a dependency of Guadeloupe, during which they resort to the resources of their bilingual repertoire.This research aims to establish what is involved, and under which conditions and through which mechanisms the interactions between two languages (English and French) may contribute to the construction of cognitive learning skills.
Collaborative activity, which is a specific realization of collective action, is part of a sociocultural environment common to all the partners of the group.The importance of collaborative learning is particularly evident in the characteristics of the self-regulation of the group and the empowering of each of the engaged parties (Mattar & Blondin, 2006).In this study we wish to jointly address two approaches: collaborative activity (Johnson & Johnson, 1999) as a realization of collective action (Sensevy, 2011) and research on language teaching methods (Anciaux, Forissier, & Prudent, 2013).If the effectiveness of collaborative learning is of particular importance in fostering interactions within these groups, it seems to us that it is of interest to use this as a basis for a focused observation of bilingual language practices.
With the goal of presenting as explicitly as possible the implementation of the strategies and the resources used by the learners, we have endeavored to break down the different competences employed by the pupils to co-construct collaborative tasks.With this in mind we have identified three types of competence: encyclopaedic, heuristic and linguistic.Encyclopaedic competence necessitates referential understanding and inferential skills related to the disciplinary field.Referential understanding stems from the knowledge utilized by the pupil in order to realize a task, whereas inferential skills comprise all the inferences implied by the instruction and reconstructed through the pupils' interpretation.Heuristic competence describes the totality of the research procedures used to resolve a difficulty, or a disagreement over interpretation.Finally, linguistic competence addresses the degree of expertise of the interlocutor and is manifested in the realization of the utterance in whichever language the interlocutor chooses to use.
We seek here to understand, through the discursive interactions between the pupils, how the tasks through which the instructions could be carried out are divided.Our study can therefore be broken down into two specific sub-aims.The first consists of observing the methods of discursive management from the point of view of a coherent approach to the task; and the second of identifying the questions of the individuals involved, during their participation in the execution of an instruction.The division of tasks (Pochon-Berger, 2010) is twofold.The first theme is concerned with the attribution of the task's sub-themes, while the second is procedural and concerns the allocation of the executive aspects of the task in hand.
We will formulate two hypotheses relating to the role of code switching in the discursive interaction, which takes place during collaborative activities: 1.They reveal in particular the asymmetry of encyclopaedic, heuristic and linguistic competences between interlocutors; 2. They allow learners to integrate themselves into the gradual construction of the disciplinary content, by taking the opportunity to control the language of the exchange to the detriment of that hitherto favored by group members.

METHOD
Our corpora are based on exchanges between pupils during their educational support sessions and were collected between January and May 2014.From the original recordings we have selected specific sequences, which illustrate certain aspects of the interaction (ICOR, 2006).We have specified the languages spoken and understood by the interlocutors, as well as the level of expertise of each pupil in French (as a discipline) according to a simplified rating scale supplied by the class teachers.
Corpus 1 is 3 minutes and 5 seconds in duration.The teacher asks the group in seconde class [normally 16-17-year-olds] to compare an extract from a novel with a film adaptation.The triad creates an asymmetrical distribution of expertise: linguistic (in English and French for Sasha and Tina) and disciplinary (particularly for Mathias).The specific nature of this corpus relates to the pivotal positioning of Tina, who, by means of her superior expertise, both linguistic and disciplinary, is capable of ensuring collaboration between the two other interlocutors.
Corpus 2 lasts for 8 minutes and 17 seconds.Two pupils, a girl and a boy, are participating in a première class [final year of high school, normally 17-18-year-olds] in a bilingual writing project (French and English) involving the production of a narrative illustrating the changes linked to St-Martin's transition into modernity.The level of linguistic expertise (English and French) is very similar for the two pupils.However, Charles demonstrates greater disciplinary expertise than does Natasha.

RESULTS AND DISCUSSION
The study of the functional distribution of languages (Causa, 2011) with regard to prescribed tasks allows us to understand the ways in which linguistic and disciplinary skills are constructed in interaction, with regard to the three competences: encyclopaedic, heuristic and linguistic.We will first approach the corpora separately, and will then go on to propose a common context leading to a possible modeling of the interactions.

Corpus 1 1st extract
The pupils, whose first names are here reduced to abbreviations, have just sat down.The teacher gives the instruction:

Analysis
Corpus 1 captures a relatively even distribution of languages in the tasks engaged in by the interlocutors.The clarification of the task takes place in two steps.At TR8, Tina has solicited Mathias' expertise and it is Sasha's turn to ask in an almost concomitant way about the passage outlined by the teacher.Here the allocation of languages is clear.One can first observe a general formulation of the task in French (by the teacher) then a delineation of the material in English (by Sasha).Tina then takes on an expert role when she expresses herself to Sasha in English as indicated at TR10 where she hetero-reformulates the sequence selfinitiated by Tina.Tina's ratification of English as the exchange language has allowed Sasha to take part in the activity.The brief controversial exchange on the passage to be read in English (TR13 to 14) has allowed Sasha to be fully integrated into the activity by authorizing her to take charge of the

Analysis
The episode with the stick at TR87 to 104 demonstrates the full extent of the efforts deployed by Sasha.She is so eager to fill the lexical gaps (she does not know the French term) that she feels a failure in terms of heuristic competence (she does not understand why it has not been mentioned in the film).At that point in the interaction Tina is able to restore the roles of Mathias and Sasha, acting as a passeur, a linkperson who encourages the transition between the stages of the interaction, which can itself be perceived as dangerous for the interaction.
An in-depth reading of the passage shows that she is not content with simply translating Sasha's questions to Mathias and putting the responses to Sasha into English as one might think: Even if the transition from one language to another carries a certain level of risk-taking on the level of identity, it can assume take this on board as it encourages a fairer redistribution of tasks.Tina takes the role of passeur, in that she spontaneously reformulates the instructions to Sasha, in order to give her permission and to motivate her heuristic competence and to call into question Mathais' supremacy during several TR, even though Mathias's encyclopaedic and linguistic competences (in French) are well in evidence.

Analysis
The recourse to the extra-discursive object (the writing produced by the pupils) is an opportunity to alter the situation, in particular in the attribution of encyclopaedic, heuristic and linguistic competences.At TR100, we witness what we will term a coup d'état, comparable to a seizing of power (taking the initiative of the response to write) to compensate for a momentary lack of expertise.The transition to writing has appeared at a crucial moment, allowing the establishment of the legitimacy of the instigator of the paper trail or of those associated with it.The presence of the triad has allowed us to observe a specific organization of the allotment of tasks.Mathias' expertise in French (in an activity which combines discipline and language) and Tina's offer to Sasha to be involved in the activity on an equal footing with the others, thanks to her use of English, has allowed us to show that mastery of the language may have been an asset in the struggle to gain a voice in the exchange.
In the next part of the extract (from TR109 to 118) Sasha has been allowed to become more directly involved in the activity.Tina resorts to her linguistic competence to bring Sasha into the exchange and to advance the interaction in the sense of achieving the task.At that point, one could say that the triad finds an equilibrium, albeit fragile, in which each of the interlocutors occupies a specific place.Sasha initiates an interrogation in which a difference between the book and the film is noted.Tina reformulates her request to Mathias who responds negatively.The best proof of the acquisition remains Sasha's re-use of the term baguette at TR101.From this point on, Mathias recognizes the expertise of Sasha, to whom he henceforth concedes a legitimate place within the activity.The exchange can be considered closed with TR109 to 118 in which Mathias completes Sasha's utterance: Tina has activated a bilingual competence, which has respectively allowed Sasha to deploy a heuristic procedure (of research) and Mathias to deploy an encyclopaedic procedure (of knowledge).In Corpus 1, we have therefore been able to see that the distribution of languages, if it is a function of the tasks with which it is associated, is also likely to exhibit the encyclopaedic, heuristic and linguistic competences of each of the group members.

Corpus 2 1st extract
The pupils are evoking the trials and tribulations of their hero who has gone in search of his sheep, which has become separated from the flock:

Analysis
This whole exchange is constructed around a disagreement on how to say in French Tasha's utterance: he went up the hill (TR1).The opposition takes place first of all in French.It is based on the restitution of two strong linguistic elements: the tense (how to evaluate the preterite?)and the import of the preposition up.The extract clearly shows the traces of this translational negotiation, without the two interlocutors describing all the stages involved.TR3 clearly demonstrates that Tasha has ratified the periphrasal proposition est allé to form the preterite.In contrast, the disagreement persists with the restitution of the movement implied by up.The disaffiliation with Charles' utterance is shown at TR5: [est allé GRIM PER/ dans la colline.Tasha takes the initiative to reformulate Charles' proposition by completely reconstituting it.Charles then repeats the utterance in English at TR8 and thus ratifies his first proposition.Analysis of the distribution of languages with regard to the tasks with which the interlocutors are engaged brings to light a set of particular phenomena.Here we are able to outline two distinct language situations, which we will distinguish as a surface distribution and a second deep distribution.The first covers the usages, which can be attributed to languages, and is subject to a separation between TR1, 7 and 8 which highlight an explicit use of English, and TR 2 to 6 in which French is used.The second situation involves a language which is less clear, in which we find a hybrid language form (TR2), and an autonomic use of French (GRIM PER at TR5).We argue that the choice of language forms influences the interaction.The confrontation of the two distributions allows the restitution of a greater intelligibility of contact languages than we had suggested at the beginning of the analysis of this corpus.According to the surface distribution, we can note a clear separation in the use of languages.TR1 to 8 therefore obey the following communicative logic: Using this interactional schema we can therefore distinguish two levels of language use: the elaboration of a proposition of translation and the proposition of the response in French, then the explanation of the processes at work in English.English thus intervenes to suspend the progress of the interaction, and its use provides an opportunity for a reflective return to a time preceding the opening utterance.Indeed, at TR7, when Tasha mimes climbing, she evokes the verbal image which has led her to propose the utterance at TR1.
Nevertheless, an in-depth analysis of the language distribution allows us to refine this reading.TR2 and 6 effectively invite us to apply more nuance to our perspectives.From TR2 onwards, we can identify the case of the elaboration of a translation into French being accompanied by an unexpected recourse to English: climbé.We notice here that the recourse to English clearly allows the establishment of a French form.From this point of view, TR3 and 4 are interesting: the forms used in French are a restitution of different forms, which are problematic in English, firstly, that of the preterite, admitted by Tasha, and then that of the preposition.In other words, even if the interlocutors express themselves in French, the progression of the interaction in the two TR proves that they are also reliant on the interaction in English.Finally, TR5 is a distancing of Charles' utterance: the accentual rise on GRIM PER and that of the voice which follows it exhibit the doubt of the female interlocutor.The utterance in French is not ratification but the expression of a first hypothesis of translational resolution, although it is not recognized as such.We note here that the recourse to French is only one stage in the construction of the interaction.This distancing of the utterance itself involves the immediate recourse to English as evidenced in the following TR.It is thus clear that French and English do not enjoy a complementary distribution in the corpus, as we had supposed at the beginning.
The model, according to which English should be reserved for the elaboration of the utterance and French for implementing the text, is not in operation here.An utterance in French is constructed from the difficulties encountered in English (TR2-3) and may assume the status of first proposition before ratification.TR6 involves the transition into English before the final ratification.In consequence, we can confirm that the reworking of forms (from the code switching which is sometimes difficult to codify, as at TR2) reflects an interactional process of reworking (in which TR5 and TR6 in French lead to the following TR in English).The hetero-reformulation produced by the teacher at TR14 is clearly not retained as such by the two interlocutors.Indeed, they return to Charles' first utterance without even having recognized the transformation made by the teacher.The teacher's reprise of Charles' proposition has not resulted in a metalinguistic comment on the actual purpose of the reparation.Neither the disappearance of the periphrasal formula est allé in favour of the standard form a(voir) grimpé, nor the lexical permutation (monté instead of the grimpé proposed initially) have been used for an evaluation by the teacher.It will be no surprise that the utterance ratified by the two non-expert interlocutors does not incorporate the modification of the auxiliary: the formula est allé grimper at TR20 has been definitively endorsed.

Analysis
The positioning of Tasha as the expert in terms of linguistic and heuristic expertise is affirmed.She initiates the majority of the TR in French, while Charles is forced to continue or reformulate them in English.Tasha directs efforts towards the elaboration of the narrative in French, culminating in her reprise at TR82 in which she is forced to repeat all the elements in French.Once again, it is the expertise of the one who dares to express themselves in this language which is revealed.Just as in Corpus 1, in which the process of writing had allowed Sasha to enter fully into the collaborative process, this extract from Corpus 2 suggests that the written record reveals the expertise of those with whom it is associated.The length of the TR82 started by Tasha and the chronological repetition of the discursive stages mean that the written record takes the form of a meaningful whole capable of being organized into a completed text.
Corpus 2 shows that the language distribution in a set of utterances is all the more likely to disrupt the expected allocation of tasks.

CONCLUSION
We have therefore been able to assess the functional distribution of languages (Causa, 2011), and the interaction between language expertise and disciplinary expertise, in terms of encylopaedic, heuristic and linguistic competences.
The distribution of languages covers diverse functions.In Corpus 1, we were able to observe that the configuration of the group and the asymmetry of their expertise had had a significant impact on the distribution of tasks and the realization of the written production.The transition to the written form proves to be a crucial moment, which allows the legitimacy of the individual who created the written record, or those associated with it, to be established.Corpus 2 allows us to show that the expected distribution of languages (the elaboration of the utterance in English and committing it to paper in French) did not apply.
The corpora have also allowed an assessment of the encylopaedic, heuristic and linguistic competences in play.In Corpus 1, Tina acted as a passeur, mediating through recourse to linguistic competence and thus permitting Sasha to bring into play her heuristic competence and to question the supremacy of Mathias, whose encyclopaedic and linguistic competences are evident.In Corpus 2, Tasha's linguistic and heuristic competences are affirmed, and she has directed her efforts towards an elaboration of the narrative in French.Just as in Corpus 1, in which the process of writing had allowed Sasha to enter fully into the collaborative process, Corpus 2 has shown that the one tasked with writing finds that this act intensifies their heuristic competence.
With this in mind, we ask whether our two hypotheses been confirmed that: • the language alternation clearly shows the asymmetry of encyclopaedic, heuristic and linguistic competences.However, and the nuance is significant, it also allows their scope to be reconfigured.By resorting to English, the passeurs offer their peers the chance to make particular use of their heuristic competence; • the language alternation enables those who take responsibility for its use, once again allowing the weakest pupils to be involved in the activity and sometimes even to orchestrate the coup d'état, by taking control of the exchange.Investigations on collaborative work and the educational performance of the interactants would benefit from further reflection on the interplay of encyclopaedic, heuristic and linguistic competences, notably by interrogating the criteria by which the expertise of pupils considered to be strong or weak is assessed.One could ask whether the similar degree of competences between peers is inclined to favour better exchanges within the group, especially through the more equal distribution of tasks in collaborative activity.It is likely that the dedicated implementation of operative educational measures would benefit from being more firmly based on a thorough reading of the complex interactions involved during collaborative activities.
• TR1 : uttered in English by Tasha • TR2 : first suggestion of translating into French by Charles (restitution of the preterite by periphrasal formula and restitution of the preposition by an action verb) • TR3 : recognition of the first part of the translation into French by Tasha • TR4 : start of an utterance completed in French by Charles in an echo of TR1 • TR5 : distancing from the utterance in French by Tasha • TR6 : hetero-initiated auto-reformulation of the utterance in French by Charles • TR7 : explanation of the presupposition in English by Tasha and metalinguistic support through gesture • TR8 : opposition and reformulation of the presupposition in English by Charles.