À la recherche de la morphologie silencieuse : sur le développement du pluriel en francais L2 écrit

Sammanfattning: This dissertation deals with the morphological development in written L2 French by Swedish learners. The thesis presents a detailed analysis of number marking and agreement in NPs and VPs, focusing mainly on the production of the morphemes –s (in NP) and the third person plural –nt (in VP). These agreements are particularly interesting in French where number morphology is often silent in the oral language and thus potentially difficult to produce in writing. The aim of the thesis is to describe the developmental sequence of number morphology in written L2 French, from an initial to a lower advanced level, and to discuss underlying factors that influence the morphological development in this domain. The results are interpreted through two different theoretical models: Pienemann’s Processability Theory (1998) and Goldschneider’s and DeKeyser’s multiple factors approach (2001). The empirical part of the thesis is based on the CEFLE corpus (Corpus Écrit de Français Langue Étrangère) which includes approximately 400 texts written in L2 French by instructed Swedish learners and by a French control group. A cross-sectional and a longitudinal study of this material are presented. The cross-sectional study of four groups of learners (N=105) and the French controls (N=30) demonstrates a clear and gradual development in the L2 production of number morphology. The following sequence is observed: 1) plural marking on nouns/pronouns and quantifiers, 2) determiner-noun agreement, 3) subject-verb agreement, and, last of all, 4) noun-adjective agreement. It is argued that the lack of phonological saliency has a minor influence on the acquisition of plural morphology in written L2 French in the instructional setting as compared to L1 acquisition. It is also shown that the semantically motivated plural markers are used initially and that the high morphological regularity in the plural, and possibly transfer, has an impact on the acquisition process at initial levels. A multiple factors approach, as proposed by Goldschneider and DeKeyser, is necessary to understand the very late noun-adjective agreement in written L2 French. The longitudinal study of fifteen individual learners, framed within the Processability Theory, shows a similar developmental pattern to that of the cross-sectional study. In general terms, the morphological development observed in the data can be accounted for by Pienemann’s processing hierarchy. However, the analysis calls for the notion of intra-stage sequencing in order to explain the differences within developmental stages, especially at the phrasal level (NP). Other factors than processing constraints, such as morphological regularity, syntactic status and frequency seem important to understand the details of the observed development. In conclusion, the learners show a gradual and rather early morphological development of plural marking in written L2 French. The two theoretical approaches meet different problems when applied to the written French L2 data. This observation raises the question of a possible synthesis of the two models discussed in this thesis.