Scaffolding language in multilingual mathematics classrooms

Against the background of increasing linguistic diversity in classrooms worldwide, we aimed to provide scientifically grounded insight into how language-oriented mathematics education can be designed, enacted and evaluated. We addressed the following main research question: How can teachers in multilingual primary classrooms scaffold pupils’ language required for mathematical learning? The first study addressed the question of how a teacher participating in dual design research learned to scaffold language required for mathematical learning. Dual design research is a special type of design research that aims to promote and trace the development of the pupils’ and the teacher’s learning. A case study involving one teacher and a multilingual upper primary classroom pointed at the teacher’s changes in knowledge and beliefs (e.g., concerning pupils’ hidden linguistic incompetence), changes in practice, and intentions for practice. We concluded that this learning process could be attributed to the characteristics of dual design research, that is its cyclic and interventionist character, the continuous process of prediction and reflection that lies at its heart, and the teacher’s co-designer’s role complemented with stimulated recall interviews. The second study addressed the language that is required for mathematical learning in the mathematical domain studied (line graphs). Drawing on genre pedagogy, we introduced the notion of pedagogical genre to refer to a genre that is deliberately designed for pedagogical purposes – in our case to support pupils’ reasoning about line graphs. The aim of the study was to summarize the development of the pedagogical genre and to evaluate what pupils learned from our approach that focused on promoting students’ proficiency in this pedagogical genre. We concluded that pupils progressed in deploying the pedagogical genre to a large extent. A multiple case study involving four pupils showed pupils’ gradual language development with some falling back over the course of the lessons. Cases where hypothetical and observed learning diverged reinforced the idea of a linguistic turn that we propose for educational design. The third and the fourth study respectively addressed the conceptual and methodological challenges of whole-class scaffolding. Referring to temporary and adaptive support, scaffolding has become widely used, also in whole-class settings. However, it has often been used in loose ways: key characteristics of scaffolding are often not clearly distinguished or scaffolding refers to any help. In response to these conceptual challenges we propose three key characteristics of whole-class scaffolding based on the scaffolding literature: diagnosis, responsiveness and handover to independence. These can be enacted both during and outside whole-class interaction, and they are distributed and cumulative over time. This conceptualisation brought along the methodological challenge of how to measure such long-term whole-class scaffolding. Therefore we investigated whether the enactment of strategies intended to scaffold the development of subject-specific language in a multilingual upper primary mathematics classroom did lead to whole-class scaffolding as identifiable by its key characteristics, taking into account the long-term nature. The empirical analyses indicated that this was indeed the case: The teacher diagnosed and responsively performed strategies over time, and she significantly handed over to linguistic independence.