Frank Guenther
Boston University
This talk describes a neural model of speech production and experiments designed to test and refine it. The model is implemented in computer simulations that control movements of an articulatory synthesizer. Model components correspond to regions of the cerebral cortex and cerebellum that become active during simple speech production tasks. A babbling cycle is used to train neural mappings between phonological, articulatory, auditory, and somatosensory representations. These learned mappings encode speaker-specific information regarding the relationships between the different representations. After learning, the model is capable of producing combinations of the sounds it has learned by commanding movements of the speech articulators in the articulatory synthesizer. Computer simulations verify the model's ability to account for a wide range of experimental findings concerning speech production, including data on acquisition of speaking skills, articulatory kinematics, and brain activity during speech.
Reception to follow in 1413 Marie Mount Hall.