Speech perception training with orofacial somatosensory stimulation affects speech production - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Speech perception training with orofacial somatosensory stimulation affects speech production

Résumé

Sensorimotor adaptation to speech motor training results in changes in speech perception (Nasir & Ostry, 2009; Shiller et al., 2009). Given the interactive nature of speech perception and production mechanisms, the reverse can also be true, that is, speech perceptual training may change speech production. Importantly, somatosensory inputs during speech motor learning appear to intervene in recalibration effects in speech perception (Ohashi & Ito, 2019). Hence, receiving a specific pair of auditory-somatosensory inputs may be a key to formulate or calibrate representations in both directions: production-to-perception and perception-to-production. In this study, we examined whether speech perception training with somatosensory stimulation can modify speech production.We carried out a perception training paradigm based on a vowel identification task and compared the production performance of corresponding speech sounds before and after the perception training phase. Fourteen native speakers of French participated in this study. In the training, the participants were asked to identify if the French vowel they heard was /e/ (lip-spreading) or /ø/ (lip-rounding). The auditory stimuli consisted of a synthesised eight-member /e/-/ø/ continuum. Each stimulus was presented one at a time in a pseudo-random order. Along with the presentation of the vowel sounds, somatosensory stimulation associated with facial skin deformation was applied in the backward direction, which corresponds to the articulatory movement for the production of the vowel /e/. In the pre- and post-test, we recorded the corresponding vowels /e/ and /ø/ in the form of the French words ‘dé’ [/de/, “dice” in English] and ‘deux’ [/dø/, “two” in English]. The participants were asked to produce these vowels using a picture-naming task. The first, second and third formants in the recorded vowel sounds were extracted and were compared to assess a possible difference between pre- and post-tests. We found that the third formant of the post-test was significantly increased in the vowel /e/, but not in the vowel /ø/. The first and second formants were not changed between the pre- and post-test for both vowels. Considering that lip spreading results in an increase of the third formant due to an increase of the front cavity resonance (Schwartz et al., 1993), we postulate that receiving somatosensory stimulation related to lip-spreading action for the production of /e/ along the perception training phase may induce a change in production after the training phase, resulting in an increase in the third formant value for /e/. Since orofacial somatosensory inputs associated with facial skin deformation can provide information of articulatory movement, the repetitive exposure to a pair of auditory-somatosensory stimulation related to a specific speech sound can modify the production of the corresponding speech sound. This could result from speech perceptual recalibration produced by the association of auditory and somatosensory inputs. These findings shed insights on the role of the somatosensory system in speech learning.
Fichier principal
Vignette du fichier
Ashokumar_SMC.pdf (612.01 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03813257 , version 1 (13-10-2022)

Identifiants

  • HAL Id : hal-03813257 , version 1

Citer

Monica Ashokumar, Jean-Luc Schwartz, Takayuki Ito. Speech perception training with orofacial somatosensory stimulation affects speech production. SMC 2022 - 8th International Conference on Speech Motor Control, Aug 2022, Groningen, Netherlands. ⟨hal-03813257⟩
73 Consultations
17 Téléchargements

Partager

Gmail Facebook X LinkedIn More