Souhila Messaoud-Galusi Laboratoire de Psychologie Expérimentale, CNRS & Université René Descartes (Paris 5)

René Carré ENS des Télécommunications, Paris

Liliane Sprenger-Charolles Laboratoire de Psychologie Expérimentale, CNRS & Université René Descartes (Paris 5)

Willy Serniclaes Laboratoire de Psychologie Expérimentale, CNRS & Université René Descartes (Paris 5)

The identification of speech contrasts is known to depend on the integration of multiple acoustic cues. The DWS model (Nittrouer & Miller, 1997), based on previous results from Nittrouer et al (Nittrouer, 1992, 1996; Nittrouer & Studdert-Kennedy, 1987), predicts that children will weight dynamic acoustic information more while adults will weight static acoustic information more. The first goal of the present study was to evaluate the development of cue weighting strategies in French monolingual adults and children from 7 to 12 years of age. Listeners had to identify a [s-S] continuum followed by the vowel [a] or [u] whose formant transition configurations were appropriate for a preceding [s] or a preceding [S]. We observed that: 1) Adults weighted all the acoustic cues signaling the contrast more than children did, contrary to DWS predictions. 2) The weights of the transition and the vowel context increased at a younger age than did the weight of the frication cue, indicating that the weighting of different cues may not develop simultaneously. The same experiment was conducted with dyslexics, known to experience categorical perception deficit (Godfrey, Syrdal-Lasky, Millay, & Knox, 1981; Serniclaes, Sprenger-Charolles, Carré, & Demonet, 2001), and with the same chronological age (CAC) and reading level controls (RLC). 3) Dyslexics weighted frication similarly to the CAC but weighted transition and vowel less than did the CAC, indicating only some cues might be processed differently by dyslexic children. 4) Dyslexics weighted transitions and vowels more than the RLC did, indicating a developmental delay rather than a deviant developmental trajectory

Back to program