School subjects are taught through academic language. Different studies have shown that students who are proficient AL users achieve better in school (Snow et al., 1989; Smit, 2013; Kleemans, 2013). AL is used at school to communicate efficiently about abstract, not directly visible content and it has specific features at the lexical, morpho-syntactic and textual level (Aarts et al., 2011, Henrichs, 2010). Not only are subjects taught by using AL, the students’ understanding and knowledge of the subject is also assessed in AL. In addition, knowledge about AL itself is part of the content of schooling (Schleppegrell, 2004). The language students use and need in the school setting however differs substantially from the language learned at home (Henrichs, 2010; Aarts et al., 2011). Teachers can stimulate AL learning of students by using AL themselves and by helping students understand and use AL (Zwiers, 2008).

AL is used in all school subjects. In mathematics AL is necessary because complex problems are placed in a contextual framework and to solve it students need to decontextualize it (Eerde et al., 2002; Mercer & Sams, 2006). Besides this teachers and students may use interactive mathematical conversation to learn and understand mathematical concepts. Students need to learn specific language features of mathematics before they can really participate in such discourse. This language is part of the AL register (Prenger, 2005; Sfard, 2001).

Instructional methods used during mathematics instruction offer different possibilities for AL stimulating behavior. The methods explanation and discussion offer possibilities for behavior aimed at understanding and at triggering AL by the students (Dokter et al, 2017). To stimulate students’ AL development teachers should use AL themselves and show AL stimulating strategies. There are six strategies aimed at students’ AL understanding (‘modeling with think-alouds’, ‘giving meaning’, ‘recasting own language’, ‘repeating own correct language’, ‘reformulating own language’, ‘visualizing’) and six strategies aimed at students’ AL production (‘asking to be more precise’, ‘giving directions’, ‘provocative statement’, ‘recasting language of the student’, ‘repeating language of the student’, ‘reformulating language of the student’). What is equally important, is that teachers connect the home language with the academic language. Strategies that change home language into language with more AL features are called power up, strategies where AL is unpacked back into home language are called power down (Harper & Parkin, 2017).

The goal of this research was to gain an insight in the AL stimulating behavior of teachers in grade 1 and 2 during mathematics instruction. The teacher plays an important role in stimulating students’ development of AL, but the extent in which they do this differs (Schleppegrell 2004; Elbers 2012; Tomasello 2000). The AL stimulating behavior that appears during explanation or discussion in mathematics instruction may differ, because the interaction during explanation is teacher lead while interaction during discussion also may be student lead (Nijland, 2011). This leads to the following questions:

1. What differences can be seen in the AL input of primary school teachers of grade 1/2 in mathematics instruction during the instructional methods explanation and discussion?
2. What AL stimulating behavior do primary school teachers of grade 1/2 show in mathematics instruction during the instructional methods explanation and discussion?

The instruction of 52 mathematic lessons of 27 teachers in grade 1/2 (age 6-7) was videotaped. To find an answer to RQ 1 the lessons were analyzed for the use of the instructional methods explanation and discussion. For each teacher 4 minutes of both instructional methods were transcribed. The transcriptions were coded for features of AL on different language levels, using a coding protocol based on DASH (Aarts et al., 2011). Paired T-tests were conducted to find significant results about the AL input at different language levels in the two instructional methods. The eleven features of AL, based on theoretical considerations and analyses using using a Pearsons'correlation matrix, could be reduced to five main features: ‘lexical diversity’, ‘lexical complexity’, ‘lexical specificity’, ‘grammatical complexity’ and ‘textual complexity’. These main features were used in the rest of this research. To answer RQ2 for each teacher the different kinds of AL stimulating behavior within the instructional methods was scored by looking at shown behavior in the relevant video fragments. The data of the observation study were analyzed by coding the teachers’ behavior as 0 (AL stimulating behavior did not occur) or 1 (AL stimulating behavior did occur) for each aspect of AL stimulating behavior during explanation and discussion. The total means and standard deviations were calculated for all types of AL stimulating behavior in the two instructional methods, aimed at students’ understanding of AL and production of AL. Students got significant more opportunity to talk during discussion than during explanation, which confirms a difference in interaction between explanation and discussion. It became clear that teachers of grade 2 used more features of AL than teachers of grade 1. Concerning RQ1 the input of the teachers during explanation consisted more AL features than their language input during discussion. The significant differences were found at the lexical level (lexical density and morfologically complex words) and at the morfo-syntactical level (clause combining). No significant difference was found at the textual level (level of abstraction). Concerning RQ2 more AL stimulating behavior was shown during the instructional method explanation than during discussion. About half of the teachers showed behavior aimed at stimulating understanding of AL by students. Less than a third of the teachers showed behavior aimed at triggering AL use. Some types of behavior, like modeling or making provocative statements, were hardly used by teachers.

Teachers use more AL features during explanation, the instructional method that is more teacher lead than during discussion, where students also lead the interaction. The results of the AL input analyses showed that teachers varied less in lexical features than in features at the grammatical and textual level. All AL features were used by the teachers and a large variety was found. Overall, teachers more often used AL aimed at content (lexical diversity and lexical specificity) than at complexity (lexical complexity and grammatical complexity). In the textual complexity the variation between the teachers was large. All strategies were used during the eight minutes that were analyzed, although individual teachers used a limited set of strategies. The AL stimulating behavior corresponds with the AL input teachers use themselves; the most teachers show stimulating behavior during explanation, the method in where also the most AL features were shown. Teachers show during explanation more behavior aimed at understanding. The AL behavior that is aimed at triggering AL use of students is shown less, even during discussion. Power down strategies were used the most and this is in accordance with the AL use of the teachers: they simplify their language to make sure students understand them. Although teachers in general use less power up than power down strategies, all teachers also used strategies aiming at their students’ AL production. Especially during the instructional method discussion they used significantly more power up strategies and students were stimulated to produce more language. In order to stimulate students’ AL development, teachers could use the instructional method discussion more often during their mathematics instruction.

Aarts, R., S. Demir & T. Vallen (2011). Characteristics of academic language register occurring in caretaker-child interaction: Development and validation of a coding scheme. Language Learning, 61(4), 1173-1221. Dokter, N., R. Aarts, J. Kurvers, A. Ros & S. Kroon (2017). Stimulating students’ academic language: Opportunities in instructional methods in elementary school mathematics. L1-Educational Studies in Language and Literature, 17 1-21. Eerde, D. van, M. Hajer, T. Koole & J. Prenger (2002). Betekenisconstructie in de wiskundeles. De samenhang tussen interactief wiskunde- en taalonderwijs. Pedagogiek, 22(2), 134-147. Elbers, E. (2012). Iedere les een taalles? Taalvaardigheid en vakonderwijs in het (v)mbo. De stand van zaken in theorie en onderzoek. Utrecht/Den Haag: Universiteit Utrecht en PROO Harper, H. & B. Parkin (2017). Scaffolding academic language with educationally marginalised students. Report of research project funded by the Primary English Teachers’ Association of Australia (PETAA), Research Grant 2016-2017. Henrichs, L. (2010). Academic language in early childhood interactions: A longitudinal study of 3- to 6-year-old Dutch monolingual children (diss. Universiteit van Amsterdam, Amsterdam Center for Language and Communication (ACLC)). Kleemans, T. (2013). Individual variation in early numerical development: Impact of linguistic diversity and home environment (diss. Radboud University, Behavioural Science Institute). Mercer, N. & C. Sams (2006). Teaching children how to use language to solve maths problems. Language and Education, 20(6), 507-528. Nijland, F.J. (2011). Mirroring interaction: An exploratory study into student interaction in independent working (diss. Tilburg University). Prenger, J. (2005). Taal telt! Een onderzoek naar de rol van taalvaardigheid en tekstbegrip in het realistisch wiskundeonderwijs (diss. Groningen University). Schleppegrell, M. (2013). The role of metalanguage in supporting academic language development. Language Learning, 63(1), 153-170. Sfard, A. (2001). There is more to discourse than meets the ears: Looking at thinking as communicating to learn more about mathematical learning. Educational Studies in Mathematics, 46, 13-57. Smit, J. (2013). Scaffolding language in multilingual mathematics classrooms (diss. Utrecht University). Snow, C., H. Cancini, P. Gonzalez & E. Shriberg (1989). Giving formal definitions: An oral language correlate of school literacy. In D. Bloome (ed.), Classrooms and literacy. Norwood, NJ: Ablex, 233-249. Tomasello, M. (2000). Do young children have adult syntactic competence? Cognition, 74(3), 209-253. Zwiers, J. (2008). Building academic language: Essential practices for content classrooms, grades 5-12. San Francisco: Jossey-Bass Teacher.