Language is both a medium to reach subject-specific learning goals and it is also a learning goal within the subject itself. Learners need to use language to participate in school communication, to understand relevant content or to work on given tasks. Thus, they are required to learn school- and subject-specific language to participate in institutionalized learning environments. The dual role that language plays in subject learning might pose challenges for all learners. However, emerging multilingual learners (EML), whose skills in everyday and academic language are yet to develop, are likely to encounter even greater difficulties (Warren & Miller, 2015). Regarding these impediments in linguistically diverse settings, teachers face the need to respond to learners´ challenges or learning needs.

Although there is a growing body of research on language-inclusive approaches in multilingual mathematics lessons, most of these studies focus on the design-level of teaching (Götze & Baiker, 2021; Prediger, 2019; Wilkinson, 2018). This means that during lesson planning teachers identify language elements, which might impede learners´ understanding of the subject, and seek for ways to make them comprehensible for learners. However, little is known about how teachers adapt language requirements during classroom interaction, when learners cannot comply with linguistic expectations or when teachers realize a language learning opportunity and use it. Erath et al. (2021) differentiate between design level and teaching practices level for instruction that creates language learning opportunities in mathematics in multilingual contexts. What we know about language adaptations on the teaching practices level mainly comes from two areas: 1- studies conducted by Heller and Morek (Morek & Heller, 2020; Quasthoff et al., 2022), which mainly focus on fostering discursive skills such as arguing and explaining 2- scaffolding approaches (Gibbons, 2015).

This study investigates micro-interactional adaptations on the teaching practices level with a specific focus on so-called divergences in interaction. Throughout this paper, the term microinteractional adaptations will refer to teachers´ spontaneous adjustments in the learning environment in order to help learners meet linguistic demands in content or to challenge them by increasing the demands. The definition of a divergence is twofold in this study: first, divergences are defined as discrepancies between learners´ actions and teacher´s expectations. Second, they are defined as teachers´ actions which deviate from their initial aim and evolve into a new learning goal. In this regard, divergences can be initiated by learners as well as teachers. The reason why the study focuses on divergences is that teachers´ practices to resolve divergences enable researchers “to decipher certain implicit norms that unfold in the teacher’s and the students’ joint actions” (Ligozat et al., 2018). Although the main research interest is on teachers´ language-related adaptations in classroom interaction, the study takes a holistic approach and investigate learner´s actions and the evolvement of the content as well. Gruson and Sensevy (2013) state that in order to understand teachers´ actions, learners´ actions and “knowledge structure and function” need to be taken into account. Another reason for adopting a holistic approach is the assumption that language adaptations in subject teaching might not necessarily appear in the form of explicit language feedbacks but instead in actions through little changes with respect to the arrangement of tasks or classroom interaction (e.g giving the right to speak to another student). The following research questions will guide the study:

1)What kinds of divergences can be observed in classroom communication in the moments of new knowledge construction by learners and teacher?

2a) Which micro-interactional adaptations do teachers make to navigate learners towards the learning goal in the case of divergences?

2b) To what extent do these practices foster language-inclusiveness?

The study utilizes an exploratory qualitative design. In order to answer the research questions, classroom observations will be conducted with voice recordings and observation protocols. Sampling Procedure This PhD project is conducted within the larger study INTERFACH (https://interfach.de/en/), which is conducted in primary school contexts and financed by DFG. Since pre-observations conducted by the researcher in 1st and 2nd grade multilingual classes showed that language demands are often reduced to a large extent and discursive activities such as explaining or arguing take place rarely in these levels, it was decided to collect data in 3rd and 4th grade classes. Eligibility criteria required for classrooms to be included in the study is that they are attended by EMLs. In order to understand this, classroom teachers are contacted and asked about the linguistic profile of the classroom. Additionally, a language background questionnaire is conducted prior to data collection. The questionnaire consists of one question that asks students which language(s) they use at home. Data Collection and Analysis It is currently planned to visit six primary schools and 10 classrooms in Hessen, Germany. Each classroom is observed for a week and observations take place during the introduction of a new learning theme. Data collection is conducted in four steps: 1-conducting the language background teaching plans prior to the new learning theme, 3-asking teachers about their main lesson goals before each lesson to be observed, 4-classroom observations by voice recording and creating observation protocols. Voice records are transcribed and the following time frames are chosen for the data analysis: the moments when classroom discussions take place during new knowledge construction or when learners and the teacher work collectively on a task. Data analysis is accomplished with turn-by-turn conversation analysis (Krummheuer & Fetzer, 2005) and by using the analytical tools of the joint action theory in didactics (JATD) framework. JATD enables to identify the divergences and to analyse how the teacher and learners try to adjust their actions to be able to reach the learning goal with regard to changes in the task, adjustment in positions (e.g a teacher´s adopting a higher position than the learners during instruction), or “evolution of the content” over time (Ligozat et al., 2018). Turn-by-turn analysis helps to identify communicative and linguistic aspects in adaptation practices.

Preliminary Results The most common types of divergences occur when learners deviate from the main subject in their contributions during discussions or when learners make several attempts to give the expected answer by the teacher. When divergences occur, teachers generally adopt a higher position than learners in classroom discussions and consequently less space is left for learners to engage with “oral academic discourse practices” (Heller & Morek, 2015). Adaptation practices during divergences will be classified according to their focus on subject-specificty or language in the final results and will be shown in a four-field array adapted from Prediger et al. (2022). Microinteractonal adaptations of teachers appeared in the following forms: 1- Reducing language demands. Interestingly, teachers tend to reduce demands even before learners show any struggles (e.g “How can we calculate this problem? Name the numbers to me.”) subject focus: strong language focus: ambigious 2-Shifting registers. Students´ contributions in classroom discussions are reformulated by the teacher with switching from everday language to subject-specific language. subject focus: weak language focus: strong 3-As-if treatment. (adapted from Quasthoff 1997) Students´ incomplete or wrongly-formulated utterances are treated “as-if” they fulfilled the linguistic demands in a task. Wrong or incomplete utterances are generally not addressed. subject focus: strong language focus: weak 4-Reformulating. Teachers tend to formulate their questions in different ways to make their demands more comprehensible for learners. In this way, several forms of the same question are directed in a row. subject focus: ambigious language focus: strong 5-Using multiple modalities. Teachers use different forms while explaining a mathematical concept. A common practice in the primary school context is to use three different forms sequentially: to draw or write on the board, to show with real objects, to explain orally. subject focus: strong language focus: strong

Erath, K., Ingram, J., Moschkovich, J., & Prediger, S. (2021). Designing and enacting instruction that enhances language for mathematics learning: a review of the state of development and research. ZDM – Mathematics Education, 53, 245–262. https://doi.org/10.1007/s11858-020-01213-2 Gibbons, P. (2015). Scaffolding Language Scaffolding Learning: Teaching English Language Learners in the Mainstream Classroom (2nd). Heinemann. Götze, D., & Baiker, A. (2021). Language-responsive support for multiplicative thinking as unitizing: results of an intervention study in the second grade. ZDM, 53, 263–275. https://doi.org/10.1007/s11858-020-01206-1 Gruson, B., & Sensevy, G. (Eds.) (2013). The Joint Action Theory in Didactics: A case study in videoconferencing at primary school. CSCL Proceedings: 1: Full Papers & Symposia. Heller, V., & Morek, M. (2015). Academic discourse as situated practice: An introduction. Linguistics and Education, 31, 174–186. https://doi.org/10.1016/j.linged.2014.01.008 Krummheuer, G., & Fetzer, M. (2005). Der Alltag im Mathematikunterricht: Beobachten - Verstehen - Gestalten. Spektrum Akademischer Verlag. Ligozat, F., Lundqvist, E., & Amade-Escot, C. (2018). Analysing the continuity of teaching and learning in classroom actions: When the joint action framework in didactics meets the pragmatist approach to classroom discourses. European Educational Research Journal, 17(1), 147–169. https://doi.org/10.1177/1474904117701923 Morek, M., & Heller, V. (2020). Individualisierter Zuschnitt diskursiver Anforderung und Unterstützung: Finetuning diskurserwerbsförderlichen Lehrerhandelns in der Unterrichtsinteraktion. In U. Quasthoff, V. Heller, & M. Morek (Eds.), Reihe germanistische linguistik: Vol. 324. Diskurserwerb in familie, peergroup und unterricht: Passungen und teilhabechancen (1st ed., pp. 381–424). De Gruyter. Prediger, S. (2019). Mathematische und sprachliche Lernschwierigkeiten: Empirische Befunde und Förderansätze am Beispiel des Multiplikationskonzepts. Lernen Und Lernstörungen, 8(4), 247–260. https://doi.org/10.1024/2235-0977/a000268 Prediger, S., Quabeck, K., & Erath, K. (2022). Conceptualizing micro-adaptive teaching practices in content-specific ways: Case study on fractions. Journal on Mathematics Education, 13(1), 1–30. https://doi.org/10.22342/jme.v13i1.pp1-30 Quasthoff, U. (1997). An Interactive Approach to Narrative Development. In M. Bamberg (Ed.), Narrative Development (pp. 51–83). Routledge. Quasthoff, U., Heller, V., Prediger, S., & Erath, K. (2022). Learning in and through classroom interaction: On the convergence of language and content learning opportunities in subject-matter learning. European Journal of Applied Linguistics, 10(1), 57–85. https://doi.org/10.1515/eujal-2020-0015 Warren, E., & Miller, J. (2015). Supporting English second-language learners in disadvantaged contexts: learning approaches that promote success in mathematics. International Journal of Early Years Education, 23(2), 192–208. https://doi.org/10.1080/09669760.2014.969200 Wilkinson, L. C. (2018). Learning language and mathematics: A perspective from Linguistics and Education. Linguistics and Education. Advance online publication. https://doi.org/10.1016/j.linged.2018.03.005