In robot-assisted language learning (RALL), language learners interact with an embodied and multimodal agent in the foreign language (L2) in different types of conversational settings. However, learners often encounter troubles that disrupt the progressivity of interaction with the robot (Jakonen et al. forthcoming, Veivo & Mutta, in review). One strategy to cope with these problems includes shifting the focus from the robot to other human participants present in this situation and getting advice from them (Honkalammi et al. 2022).
In this paper, our main goal is to study epistemic repair between learners who talk with the robot in small groups of 2-4 learners for the first time. By epistemic repair, we refer to repair that concerns the learners' orientations toward knowledge and information (cf. Bolden 2013). Especially other-initiated repair is one crucial way to maintain progressivity and achieve joint goals in social interactions (Dingemanse et al. 2015). In L2 interactions, other-initiated repairs make it evident that participants monitor and orient to ongoing sequences of interaction (Hellermann 2011: 150). This is important for the development of interactional competencies that are situation-based and context-bound (Pekarek Doehler 2019: 30). In RALL, this means that peer collaboration is necessary to establish progressivity. Previous studies on digital learning environments suggest that some confusion is typical in this context and can be productive for learning, as it induces problem-solving (cf. Arguel et al. 2019). Thus, according to our interpretation, frequent troubles in RALL and subsequent peer collaborations are essential instances for collaborative language learning.
Our corpus consists of 15 video-recorded situations, totaling 5 hours and 13 minutes, from English as a L2 in a Swedish-speaking school in Finland. There were 34 learners, aged 9 to 13, accompanied with their teacher. In this data, we identified c. 160 repair sequences between learners, of which over 60 were epistemic. In our analysis, we focus on the repair sequence – how repair gets initiated and how it is solved.
References
Arguel, A. & al. (2019) Seeking optimal confusion: a review on epistemic emotion management in interactive digital learning environments, Interactive Learning Environments, 27:2, 200-210.
Bolden, G. (2013). Unpacking "Self": Repair and Epistemics in Conversation. Social Psychology Quarterly, 76(4), 314–342.
Dingemanse, M. et al. (2015). Universal Principles in the Repair of Communication Problems. PloS One, 10(9),
Honkalammi, H.-M., Veivo, O., & Johansson, M. (2022). Advice-giving between young learners in robot-assisted language learning. Proceedings of the Conference Human Perspectives on Spoken Human-Machine Interaction, 46–51.
Hellermann, J. (2011). Members' methods, members' competencies: Looking for evidence of language learning in longitudinal investigations of other-initiated repair. In: Hall, K. & al. (eds.). L2 Interactional Competence and Development. Multilingual Matters.
Jakonen et al (forthcoming). 'Am I saying it wrong?' Managing progressivity-related troubles in child-robot L2 interaction.
Pekarek Doehler S. (2019). On the nature and development of L2 interactional competence. In: Salaberry, M. & Kunitz (eds.) Searching and Testing L2 Interactional Competence. Routledge.
Veivo, O. & Mutta, M. (in review). Dialogue breakdowns in robot-assisted L2 learning.
