The neural mechanisms underlying the processing of speech beyond the cochlear nerve are virtually unknown. At the level of the cochlear nucleus, to which the nerve projects, there occurs a physiological diversification of potential significance to the encoding of the speech spectrum. This proposal concerns the response of two broad response classes, the onset units and the choppers, in the posteroventral portion of this nuclear complex. The onset units appear to act as """"""""coincidence detectors,"""""""" responding to stimulation with a precisely timed discharge, which could be useful for encoding information pertaining to the fundamental frequency (pitch) and first formant of voiced speech sounds. Choppers may serve, through their regularity of discharge, to temporally encode low-frequency information and provide a rate/place representation of the formant pattern. The proposed project will study in detail, the temporal, rate and latency properties of the onset and chopper units and attempt to correlate them with the responses of single cochlear-nerve fibers to the same stop-consonant+vowel stimuli. The ensemble response, derived from the activity of hundreds of units distributed over a broad range of characteristic frequencies, will be constructed for both the cochlear nerve and the posteroventral cochlear nucleus. Masking noise will be used to infer the tonotopic afilliation of inputs to individual chopper and onset units and to delineate the nature of the stimulus representation under conditions of low signal-to-noise ratio. The results of this project may be germane to the loss of speech intelligibility suffered by the hearing impaired.
