Abstract
A number of research projects in mathematics education have highlighted the potential impact of variations in grammatical structures on mathematical learning. However, these grammatical variations are often not taken into account in assessments, whether local, national or international. It can be assumed that these difficulties are exacerbated in bi-plurilingual contexts, because of the interaction with translation issues. This may be the case in contexts where the language of instruction is not the vernacular language of the pupils, or in the case of international assessments such as PISA. Dans une première partie nous montrons en appui sur la littérature pourquoi il est nécessaire de prendre en compte les questions de grammaire et leurs variations suivant les langues dans l’enseignement et l’apprentissage des mathématiques, notamment dans le cas d’énoncés comportant une négation. Dans une seconde partie nous analysons les recommandations du guide Adaptation et traduction du PISA du point de vue des questions de grammaire en les mettant en perspective avec les travaux présentés dans la première partie. In the third part, we analyse an example from the PISA 2012 released items for mathematics, highlighting grammatical difficulties likely to raise non-trivial questions of interpretation and translation. To do this, we will use the logical language analysis tools we have developed. We conclude with the prospects opened up by this work as part of a multidisciplinary project involving researchers in mathematics didactics, language sciences and language didactics. One of the challenges of this project is to identify possible biases in the interpretation of the results of both national and international assessments resulting from the failure to take account of grammar and translation issues in the a priori analyses of the items submitted as part of PISA.