Altered auditory feedback experiments exploring the grain-size of speech production representations

Abstract

Speaking consists in producing sequences of linguistic units. Various models have proposed words, syllables or phonemes as (the main) speech production units – often in an exclusive view. In this talk I will present two studies investigating the grain-size of speech production units in adults and children using an altered auditory feedback paradigm. This paradigm consists in altering in real time the auditory feedback of a speaker by shifting formants to transform one vowel into another one. Speakers usually change their pronunciation in response to the perturbation. Measuring to what extent this change (1) persists after the perturbation is stopped –the aftereffect- and (2) transfers to other words containing the same vowel or the same syllable –transfer at phoneme or syllable level - allows us to unravel the respective contribution of different linguistic units to speech production. Our first study revealed a multi-level pattern in adults. Our second study indicated that preliterate 4 year-old children already use phonemic representations in speech production but rely more on holistic word representations than 8 year-old children.

About Tiphaine Caudrelie

Tiphaine Caudrelier is a postdoctoral researcher at the Basque Center on Cognition, Brain and Language (BCBL), in San Sebastian, Spain. She received her PhD from the University of Grenoble, France. Her research focuses on transfer of learning as a way to probe speech representations, in production and perception, adults and children, monolinguals and bilinguals. She also currently develops a line of research on the acquisition of foreign language pronunciation.