Modeling Sensory Preference in Speech Motor Planning: A Bayesian Modeling Framework.

Jean-François Patri,Pascal Perrier,Julien Diard

doi:10.3389/fpsyg.2019.02339

Abstract

Experimental studies of speech production involving compensations for auditory and somatosensory perturbations and adaptation after training suggest that both types of sensory information are considered to plan and monitor speech production. Interestingly, individual sensory preferences have been observed in this context: subjects who compensate less for somatosensory perturbations compensate more for auditory perturbations, and vice versa. We propose to integrate this sensory preference phenomenon in a model of speech motor planning using a probabilistic model in which speech units are characterized both in auditory and somatosensory terms. Sensory preference is implemented in the model according to two approaches. In the first approach, which is often used in motor control models accounting for sensory integration, sensory preference is attributed to the relative precision (i.e., inverse of the variance) of the sensory characterization of the speech motor goals associated with phonological units (which are phonemes in the context of this paper). In the second, “more original” variant, sensory preference is implemented by modulating the sensitivity of the comparison between the predicted sensory consequences of motor commands and the sensory characterizations of the phonemes. We present simulation results using these two variants, in the context of the adaptation to an auditory perturbation, implemented in a 2-dimensional biomechanical model of the tongue. Simulation results show that both variants lead to qualitatively similar results. Distinguishing them experimentally would require precise analyses of partial compensation patterns. However, the second proposed variant implements sensory preference without changing the sensory characterizations of the phonemes. This dissociates sensory preference and sensory characterizations of the phonemes, and makes the account of sensory preference more flexible. Indeed, in the second variant the sensory characterizations of the phonemes can remain stable, when sensory preference varies as a response to cognitive or attentional control. This opens new perspectives for capturing speech production variability associated with aging, disorders and speaking conditions.

Highlights

The recent history of research that investigates the links between phonology, production and perception of speech has been marked by vigorous exchanges between proponents of purely acoustic/auditory theories (Stevens, 1972; Stevens and Blumstein, 1978; Blumstein and Stevens, 1979; Lindblom, 1990; Sussman et al, 1991) for whom the physical correlates of phonological units would be exclusively in the acoustic domain, and proponents of theories who rather saw these correlates primarily in the articulatory/somatosensory domain (Fowler, 1986; Saltzman, 1986)
Models that were designed to simulate and investigate the process of articulation and sound production from the specification of phonological sequences were split into two main categories: models in which the goals of the speech task were specified in the articulatory domain (Coker, 1976; The Task Dynamics Model: Kelso et al, 1986; Saltzman and Munhall, 1989; The DIVA Model Version 1: Guenther, 1995; Kröger et al, 1995; The C/D model: Fujimura, 2000), and models in which the goals were specified in the acoustic domain (The DIVA Model Version 2: Guenther et al, 1998; GEPPETO: Perrier et al, 2005)
The main contribution of our work is to present two different approaches implementing sensory preference in a speech production model that integrates both the auditory and the somatosensory modality

Summary

Introduction

The recent history of research that investigates the links between phonology, production and perception of speech has been marked by vigorous exchanges between proponents of purely acoustic/auditory theories (Stevens, 1972; Stevens and Blumstein, 1978; Blumstein and Stevens, 1979; Lindblom, 1990; Sussman et al, 1991) for whom the physical correlates of phonological units would be exclusively in the acoustic domain, and proponents of theories who rather saw these correlates primarily in the articulatory/somatosensory domain (Fowler, 1986; Saltzman, 1986). In the absence of any evidence supporting undeniably one of these theories, new theories emerged assuming that phonological units could be associated with both auditory and somatosensory goals (see for example the concept of “perceptuo-motor unit” in the Perception-for-Action-Control Theory of Schwartz et al (2012); or, for another perspective, the phonological processing of the HFSC model of Hickok (2012) distributed over an auditory-motor circuit for syllable and over a somatosensory-motor circuit for the phonemes)

Objectives

Methods

Results

Conclusion