Perceptual recalibration of speech sounds following speech motor learning

Abstract

The functional sensorimotor nature of speech production has been demonstrated in studies examining speech adaptation to auditory and/or somatosensory feedback manipulations. These studies have focused primarily on flexible motor processes to explain their findings, without considering modifications to sensory representations resulting from the adaptation process. The present study explores whether the perceptual representation of the /s-b/ contrast may be adjusted following the alteration of auditory feedback during the production of /s/-initial words. Consistent with prior studies of speech adaptation, talkers exposed to the feedback manipulation were found to adapt their motor plans for /s/-production in order to compensate for the effects of the sensory perturbation. In addition, a shift in the /s-b/ category boundary was observed that reduced the functional impact of the auditory feedback manipulation by increasing the perceptual "distance" between the category boundary and subjects' altered /s/-stimuli-a pattern of perceptual adaptation that was not observed in two separate control groups. These results suggest that speech adaptation to altered auditory feedback is not limited to the motor domain, but rather involves changes in both motor output and auditory representations of speech sounds that together act to reduce the impact of the perturbation.

Sensorimotor adaptation affects perceptual compensation for coarticulation

“…Furthermore, the direction of the observed effects suggests that the change in perceptual function corresponds to motoric changes rather than auditory feedback. Previous experiments have found that passive exposure to the recorded speech of a participant compensating in response to AAF fails to change perceptual function, regardless of whether the recorded speech is made by an average compensating speaker (Shiller et al, 2009) or consists of a random selection of stimuli take from several compensating speakers (Lametti et al, 2014b). However, these experiments differed from the present study in that they did not examine how passive listening to such stimuli may affect CFC.…”

“…As in other experiments (Shiller et al, 2009;Lametti et al, 2014b), no within-session significant correlations were found between production (standardized F2) and perception (centered response). The results seem to suggest that the effect of the altered feedback on perception only emerges when considering the differences in production and identification across the two sessions.…”

“…The authors conclude from this that the shift in perception follows the direction of the articulation rather than the acoustic input. As in Shiller et al (2009), the possibility that this effect was due to simple auditory exposure to shifted feedback was ruled out by inclusion of a passive listening control group. These two studies, Shiller et al (2009) and Lametti et al (2014b), suggest that altering motor-to-auditory mappings can alter the perception of speech sounds.…”

“…As in Shiller et al (2009), the possibility that this effect was due to simple auditory exposure to shifted feedback was ruled out by inclusion of a passive listening control group. These two studies, Shiller et al (2009) and Lametti et al (2014b), suggest that altering motor-to-auditory mappings can alter the perception of speech sounds. In both studies, the authors suggest that their findings support the idea that production and perception are closely linked and that the motor system plays an active role during perception.…”

“…In a study by Shiller et al (2009), for example, participants were asked to produce [s]-initial CV or CVC words under conditions of altered (AF) or unaltered (UF) auditory feedback. Prior to and following speech training, both groups identified stimuli along a continuum between "a said" and "a shed."…”