The production of Dutch word stress by Francophone learners. Preliminary results

Dutch is taught as a foreign language (DFL) in most schools of French-speaking Belgium. According to surveyed students and teachers pronunciation and prosody are however very often left aside in DFL courses, meaning that most learners would not be familiar with Dutch word stress. This projects aims at investigating the production (current stage of the project) and perception of Dutch stress by Francophone DFL learners. Dutch has a variable and contrastive word stress that is a lexical property of words [1]. It plays a major role in the process of word recognition (e.g. [2], [3]). French on the contrary does not have word stress, but a fixed ‘primary’ (compulsory) and demarcative accent located on the final syllable of a word group (sometimes called ‘rhythmic’ [4] or ‘clitic’ group [5]). In this production experiment DFL learners (n=20) and a native control group (n=10) produced 30 Dutch three-syllable words in carrier sentences. The stress position was determined perceptually by three independent labelers. We hypothesized that due to the lack of flexibility of the French ‘primary’ accent and learners’ lack of knowledge of stress, they would tend to stick to their L1 pattern regardless of the expected 1st-, 2nd- or 3rd-syllable stress. On the whole, the DFL participants showed a tendency to stress the 3rd syllable in every condition. However the percentage of final stress never exceeds 65% and the data show some variability, which moderates the hypothesis. Strikingly the percentage of correct stress adds up to 37% (97.9% for the control group), showing that those participants did not master Dutch word stress in this experiment. Moreover an interesting result is the lack of agreement among the labelers in 24% of the cases, seeming to mean that the production data were sometimes so ambiguous (e.g. several syllables made prominent, no consistent use of stress cues) that several perceptions were possible. In the near future acoustic measures should shed light on this phenomenon. References [1] Rietveld, T. & Heuven, V.J., van (2009). Algemene Fonetiek. Bussum: Coutinho. [2] Cutler, A., & Donselaar, W., van. (2001). Voornaam is not (really) a homophone: Lexical Prosod and Lexical Access in Dutch. Language and Speech, 44(2), 171-195. [3] Cutler, A. (2012). Native listening: Language experience and the recognition of spoken words. Cambridge, MA: MIT Press. [4] Schwab, S. & Llisterri, J. (2012). The role of acoustic correlates of stress in the perception of Spanish accentual contrasts by French speakers. In Qiuwu Ma, Hongwei Ding & Daniel Hirst (eds.): Proceedings of the 6th International Conference on Speech Prosody. Shanghai, 350-353. [5] Avanzi, M. (in prep.). Note de recherche sur l’accentuation et le phrasé prosodique à la lumière des corpus du français. Travaux neuchâtelois de linguistique.