From an interactional perspective, we provide a fine-grained qualitative analysis of Data-Driven Learning interactions between French as Foreign Language students and trainers, during the exploration of the FLEURON database. Conversational and multimodal resources play a pivotal role in the selection of concordance-based instances and the identification of recurrent linguistic patterns.
Data-Driven Learning studies have shown the effectiveness of this approach not only for English, but also for other languages (e.g. French, Boulton & Tyne 2014). The exploitation of spoken corpora and concordancers in the classroom have therefore been the subject of growing research interest in recent years, but rarely it has been analysed in detail from a qualitative viewpoint. The description of dedicated concordancers (often with examples chosen by the teacher) and the solutions found by students in describing linguistic patterns represent the main focuses of DDL studies. In this paper, from an interactional perspective, we provide a conversational multimodal analysis (Mondada 2018) of interactions between FLE (French as a Foreign Language) students and trainers, who act as supervisors during the exploration of concordance-based instances. Through a fine-grained qualitative analysis, we show the pivotal role of conversational and multimodal resources (i.e. gaze, gestures, body postures) in the selection of the different types of instances and the identification of recurrent linguistic patterns. We analyze video-recorded interactions involving intermediate FLE students, i.e. B1 level according to the Common European Framework of Reference. Learners works on concordancers and audiovisual corpora from the FLEURON database, whose sections were initially presented by trainers. Language learning through corpora is not merely seen as a pre-defined pedagogical strategy, but above all as a collaborative process in which participants' contributions are realized in various modalities and negotiated over time in interaction.
