Abstract:

Many theories of speech processing propose the existence of motor control systems that utilize invariant vocal tract configurations to specify particular goals of speech production. One problem that challenges these theories is the fact that the associated structures involved with speech production go through a considerable amount of change during development. The same speech goals continue to be achieved during the course of development despite the changes that the vocal tract configuration undergoes. In this paper, speech production in the developing child is modeled by a dynamic neural network that incorporates value-dependent learning based on self-produced auditory stimulation across self-organizing sensorimotor maps. The conversion from articulatory configuration to acoustic signal is worked out by a modified version of the Maeda articulatory model that utilizes developmental parameters. The performance of the neural network was assessed at different points during development by determining the articulation-acoustic output needed to produce various segments of speech in relation to the corresponding trajectory of activation across the neural network. The results are discussed in relation to a dynamic theory of speech processing that can account for developmental change as well as some speech disorders. [Work supported by NIDCD.]