This proposal requests continued support for a broad program of empirical research dealing with speech perception, analysis and synthesis. The major goal of the proposed work is to seek a better understanding of the earliest stages of speech processing, particularly in terms of how the initial sensory-based acoustic-phonetic information in the speech signal interacts with other sources of knowledge to support spoken language understanding. The project is entering its eighth year. The proposed work will be carried out with both natural and synthetic speech stimuli in the hope of gaining a better understanding of the perceptual and cognitive processes used by human listeners in perceiving phonemes, words, sentences and passages of connected fluent speech. The proposed work is divided into four major projects that include: (1) acoustic analysis and perception of speech sounds; (2) spoken word recognition and lexical access; (3) processing fluent speech, and (4) perception of synthetic speech. From the projects we hope to learn more about how speech signals are processed by the nervous system and how the earliest stages of speech perception are linked to some of the more abstract linguistic and cognitive processes involved in spoken language understanding. The findings have implications for problems concerning aids for the handicapped, computer-assisted instruction and voice-response devices used for man-machine communication.
Showing the most recent 10 out of 22 publications