Using the first language (L1) to solve problems in understanding the second language (L2) may be beneficial for L2 learning. However, the overuse of L1 may deprive L2 learners of exposure to the L2. It appears that the question is not whether to use L1 or L2; it is when and how each language can be used to support L2 learning. This study focuses on situations in L2 education in which L1 is not used in an attempt to solve the L2 learner’s problems of understanding the L2 and the current assignment. The aim is to investigate what role participants’ management of each other’s L2 knowledge plays in maintaining intersubjectivity and doing L2 learning in these situations. The results show that participants may act on diverse L2 knowledge and diverse possible understandings regarding the assignments and that the use of L2 risks expanding these epistemic discrepancies. Using L2 also appears to restrict doing L2 learning on conceptual knowledge, for example, word-meanings. The participants agree on a minimally required epistemic balance as a way to move on. The use of L1 as needed in these situations, as a resource for meaning-making, may be beneficial for doing L2 learning.

Video is currently used in many studies to document the interaction in conversation analytical (CA) studies on learning. The discussion on the method used in these studies has primarily focused on the analysis or the data construction, whereas the relation between data construction and analysis is rarely brought to attention. The aim of this article is to discuss different approaches to data construction that CA studies in and on learning utilize, and how these approaches facilitate different analysis and understandings of learning and cognition from emic, participants', points of view. Three, partly overlapping, thematic approaches can be discerned: (1) setting-centred, (2) participant-centred and (3) content-centred. The underlying interest of the study seems to influence the data construction, which in turn affects what kind of analysis that can be done. There is a considerable variation in which aspects data sets focus on, where an emphasis in data construction on setting, participant or content also seems to project the subsequent analytic emphasis. This relation between data construction and analysis is important to be aware of and to address.

This study investigates the video game play of a multiplayer first-person shooter, Counter-Strike: Global Offensive, as part of an esports programme at a vocational school. The game environment is multilingual, and the focal participants are all Finnish-Swedish bilinguals who are proficient in English. The study focuses on the action of providing callouts, which are coordinated English words that refer to specific in-game locations and, when provided, point to opponents’ locations. The aim is to investigate how participants employ callouts as part of their in-game interaction and teamplay, and what they orient to as ‘callout competence’. With a greater understanding of the social organisation of the multilingual game environment and actions, such as callouts, we can better understand the affordances for collaborative and multilingual learning that games can provide for education. Callout competence appears to align with skills and knowledge that may be transferrable into the educational setting; that is, the components that are part of callout competence require collaborative skills and multilingual competence. These skills are part of what makes the teamwork work, as well as an inherent part of activities in an esports education programme that has broadened the classroom to encompass esports game play outside of the classrooms.