This conversation analytic study examines occasions when tutors provide tutees with ambiguous or markedly generic advice in English as a second language (ESL) writing conferences. On such occasions, tutors seem to be unable to give advice on matters that tutees seek assistance with, which typically leads to advice pursuits for a sought-for response. Data come from video-recorded interactions between two L1 English tutors and eleven L2 English tutees, enrolled in ESL writing workshops at an urban community college in the United States. Findings reveal that the mismatch between tutees’ gesture and speech, the incremental build of advice-seeking sequences and elements that lack clarity in those sequences lead to unsuccessful advice giving. Subsequently, tutees tend to engage in post-advice interrogations, which is treated as either clarification requests for promoting tutee’s agenda or resistance for ambiguous advice. The study has implications for understanding the potential trouble sources in the construction of advice-seeking sequences, and thus may increase the awareness of participants regarding the nature of tutees’ advice requests in ESL writing conferences.

The role of social interaction in second language (L2) learning has by now gained widespread recognition in Second Language Acquisition (SLA) research. Conversation-analytic SLA has started to uncover how L2 speakers develop their interactional competence – the ability to co-construct and coordinate social actions in context-sensitive ways (Hellermann, 2008; Pekarek Doehler, 2010; Skogmyr Marian, 2022). Recent studies focus on how L2 speakers develop linguistic resources (e.g., discourse markers) for managing structural features of interaction, contributing to their ‘L2 grammar-for-interaction’ (Pekarek Doehler, 2018).

This study examines the development of one such resource in L2 French: 'tu vois' ("you see"). While traditionally a complement-taking predicate with the literal meaning of visual perception, 'tu vois' has grammaticalized into a discourse marker in spoken French (Bolly, 2012), often used turn-finally to elicit a response (Stoenica & Fiedler, 2021). But do L2 speakers adopt this usage?

Using longitudinal conversation analysis and usage-based SLA, the study investigates 'tu vois' in 80h of video-recorded interactions among 23 adult L2 French speakers (A1–C1 levels) in a French-speaking Swiss region. Participants met bi-monthly for 3–21 months in a conversation circle. Analyzing 298 instances of 'tu vois', the study reveals both common patterns and individual differences in the data. Two case studies show that while initially deploying 'tu vois' in its literal sense, both speakers eventually routinize it as a discourse marker used to engage recipients, mirroring L1 French usage. These findings suggest parallels between L2 acquisition and L1 grammaticalization and have implications for L2 learning and teaching practices.

Previous studies on digital game-based language learning (DGBLL) have shown that digital games can have a positive effect on vocabulary learning, for example. Most research, however, has been done in L2 English contexts with adult learners. Drawing on sociocultural theory and an ecological approach to language learning, we investigate Finnish pupils’ (aged 14–16) interaction that occurs during a digital language-learning game aimed at improving learners’ lexical, grammar and pragmatic skills in L2 Swedish. Specifically, we analyse the types of language-related episodes (LRE) and metatalk that learners engage in while playing a digital game with multiple choice answers in pairs. The results show that pupils produce a wide range of LREs during the game. Typically, the episodes are negotiations about correct answers with a focus on vocabulary. Pupils also engage in metatalk relatively often, justifying their answers in the game through meaning-related and grammatical argumentation. In addition, they rely on pragmatic reasoning and their own intuition in deciding over language items. The study shows that playing digital games together can provide learners of L2 Swedish with opportunities to discuss and learn lexical, grammatical and pragmatic aspects of language. DGBLL can thus offer effective language-learning methods also outside of the L2 English context.

In L2 learning contexts, participants with epistemic authority in the second language (e.g., language teachers) are typically regarded as experts in the target language. However, sometimes experts’ assumed epistemic authority is at stake, e.g., when they encounter difficulties explaining the appropriate use of linguistic constructions. While prior research focused on self-talk in classroom interaction (Hall & Looney, 2021), we still know little about other practices that participants employ in such moments, especially in less institutionalized encounters. This paper aims to fill this gap by investigating how language tutors and L1 speakers navigate observable trouble responding to questions related to language matters. The study is based on a corpus of 28 hours of video-mediated L2-German lessons involving a tutor and two beginning-level adult learners and 40 hours of video-recorded tandem meetings in German-Spanish, German-Italian, and German-English. It analyzes language experts’ responses to learners’ requests for information, actions that question the language expert’s prior language-related informing, and requests for confirmation. The responsive practices include dispreferred design features, ranging from producing epistemic hedges and embodied displays of uncertainty to engaging in self-talk and introducing alternatives. However, instead of foreshadowing a normatively dispreferred response, they allow participants to gain time to formulate an answer and restore the momentary imbalance in the epistemic ecology. By not providing a straightforward response, language experts treat the learner’s question as valid and position themselves as responsible for providing the most helpful answer. The study discusses implications for our conceptualization of preference organization in informal L2 instructional settings.