Retour vers liste

Détail de la contribution

Auteur: Violaine MICHEL LANGE

Co-Auteur(s): Marina Laganaro

The span of phonological ahead planning in the production of adjectival noun phrases in French

Abstract/Résumé: Research on ahead planning at the phonological level includes diverging results with studies favouring a span of encoding limited to the first word (Meyer, 1996) and studies claiming that the entire utterance is planned before articulation (Schnur, 2011). Interestingly, cross-linguistic studies showed that the span of encoding for noun phrases (NPs) can be determined by the position of the head-noun. Post-nominal adjectival NPs (N+A) present therefore a span of encoding limited to the first word while the span of encoding for NPs with pre-nominal adjectives (A+N) comprise the entire NP (Schriefers & Teruels, 1999; Damian et al, submitted). We investigated this hypothesis within the same language by testing French noun phrases (NPs) in a picture word priming paradigm with auditory distractors. Adjectives were either prenominal (le vieux balai) or postnominal (le balai rouge). In Experiment 1, distractors were phonologically related or unrelated to the noun. Phonological distractors led to a reliable facilitation effect with shorter reaction times only when the noun was in initial position (N+A) but with a marginal effect in second position (A+N). Before drawing any conclusion, we primed both the noun and the adjective in Experiment 2. A facilitation effect of the first element only was reported suggesting that phonological planning does not extend the first word. However, a marginal facilitation effect for the noun in A+N was observed. The non significant facilitation effect in A+N is difficult to account for in the light of specific syntactico-phonological phenomena such as liaison in French. Liaison is obligatory in A+N sequences and require ahead planning at least up to the initial phoneme of the N in A+N. We decided therefore to create Experiment 3 with A+N sequences and include ¼ of liaison sequences. Moreover, we tested a larger group of participants to verify whether naming latencies could affect encoding processes. A facilitation effect was reported for the A for all speakers. However, we report an interaction between the group of participants and their naming latencies for the N. These results suggest that the span of encoding is not blind to syntactic structure of the NP and can be modulated by the speed of naming latencies.