How do L2 learners and L1 writers differ in their reliance on working memory during the formulation subprocess?

Abstract

L1 and L2 writers attend to different aspects of the formulation subprocess of writing. L2 writers devote more time and attention to low-level aspects such as grammar correction and spelling (Barbier 1998; Fagan and Hayden 1988; Whalen and Menard 1995), leading to better spelling performances than L1 writers (Gunnarsson-Largy 2013). In deep-orthography languages such as French or English, L1 writers retrieve a phonological form of the word and then tend to automatically transcribe the most frequent corresponding orthographic form, whereas L2 writers seem to directly retrieve the exact orthographic form. For L2 writers, the visuo-orthographic form of the word therefore seems to prevail over the phonological one. Accordingly, we hypothesized that L1 and L2 writers rely differently on working memory (WM). To test this hypothesis, we designed an experiment where two groups (Levels B1 and C1) of instructed L2 French learners and an L1 French control group wrote dictated sentences, with compulsory negation marking in an ambiguous phonological context. While writing, they performed a concurrent task that induced a cognitive load on either phonological or visual WM, in order to identify the nature of the form maintained in WM during semantic checking. Results indicated that L2 French learners gradually move from a visual to a more phonological form of retrieval.