Soutenance de Thèse – Elodie RONAYETTE

This work investigates how physical effort is taken into account in speech adaptation to altered auditory feedback. It tests the hypothesis that adaptation relies not only on the correction of acoustic errors, but also incorporates the motor cost associated with the compensatory gesture. It is proposed that interindividual differences in the self-evaluation of this effort contribute to adaptive variability. Three behavioral experiments were conducted in 36 typical adults.

The first experiment aimed to determine whether the saturation of acoustic compensation in response to increasing auditory perturbation reflected the attainment of a neuromuscular limit of the compensatory gesture. During the production of the vowel [ø], acoustic compensation saturated for perturbations exceeding approximately 40% of the second formant (F2), in parallel with a saturation of the electromyographic activity of the orbicularis oris superior (OOS). However, the level of activity reached remained far below a physiological maximum, representing only ~24% of maximal voluntary contraction, while approaching a « functional » maximum observed during hyperarticulated productions of [ø].

The second experiment investigated whether this saturation could correspond to the attainment of a maximum level of perceived effort. Participants produced lip-rounding gestures at different levels of muscular contraction and subsequently rated their perceived articulatory effort. Although perceived effort increased quadratically with muscular contraction, the level of muscular activity observed at the saturation point in the first experiment did not correspond to effort levels close to the maximum perceived effort. In contrast, substantial inter-individual variability in effort perception was observed, and several indices of effort perception were correlated with the adaptive behavior observed in the first experiment.

The third experiment assessed the impact of increasing the motor cost of the compensatory gesture by means of lip tubes of different stiffness. In participants who compensated through a lip-rounding gesture, greater mechanical resistance resulted in reduced compensation and earlier saturation. As in the first experiment, this saturation coincided with the saturation of lip muscle activity. Moreover, under these more effortful production conditions, participants modified their articulatory strategy beyond a certain level of auditory perturbation, and this reorganization occurred earlier with the stiffer tube.

These findings support the idea that the motor cost of the compensatory gesture is taken into account during speech auditory-motor adaptation, influencing both the degree of compensation and the selection of articulatory strategies. However, compensatory saturation and reorganization do not appear to result from the attainment of an absolute limit of neuromuscular activity or perceived effort. Rather, effort-related constraints seem to be relative to the task being performed. In the specific context of [ø] production, compensatory saturation may therefore correspond to reaching a level of muscular contraction close to the « functional » maximum observed in hyperarticulated speech, or to a threshold of gesture modification exceeding the habitual variability of conversational productions of this vowel.

These findings provide new theoretical perspectives on the mechanisms underlying speech adaptation and their modelling, suggesting that they could be viewed as decision-making processes involving tradeoffs between different constraints (linguistic, physiological, situational, etc.). Finally, by suggesting that perceived effort may contribute to inter-individual variability in adaptation, this work offers clinical perspectives for understanding and improving speech-language therapy interventions for disorders such as stuttering, dysphonia, and Parkinsonian dysarthria.

Membres du jury :

• Cécile FOUGERON, Directrice de Recherches, Laboratoire de Phonétique et Phonologie, CNRS Paris (rapporteur)
• Rudolph SOCK, Professeur des Universités, Institut de Phonétique de Strasbourg, Université de Strasbourg (rapporteur)
• Serge PINTO, Directeur de Recherche, Laboratoire Parole et Langage UMR 7309, CNRS Aix-Marseille Université (examinateur)