Pratiques et représentations de l’orthographe parmi les étudiants de Sciences du Langage

This paper deals with a set of paradoxical questions: how French students of linguistics are able to fathom their own command of French spelling rules and their metalinguistic ability in analyzing their own mistakes produced when writing graduate examination tests? How linguists teaching these students can handle discrepancies between the linguistic performance they might expect from students and actual results they get from an inquiry on students' habits in writing French? How methods initially contrived for the analysis of speech variation, such as co-variationism (the labovian paradigm.) could be applied to written forms of language in order to observe linguistic change in other fields than spoken usage? A survey of students' usage of written French, and attitudes towards written norm carried out at the University of Montpellier shows that students tend to have a fuzzy knowledge of their own command of written French and of French spelling rules, but still cling to a conservative approach to French orthography, rejecting claims for reform (réforme de l'orthographe). We also address the question of what kind of linguistic change is implied in spelling mistakes, as many trends for structural change can be observed through this kind of corpora, which nevertheless have few chances to trigger real changes in the written norm of French, because of wide-spread support to maintain French spelling as much difficult as it is, regardless of how it could be made more easy and functionnal.