Abstract
In this report we review the vowel and consonant recognition ability of patients who use a multichannel cochlear implant and who achieve relatively good word identification scores. The results suggest that vowel recognition is accomplished by good resolution of the frequency of the first formant (F1) combined with poor resolution of the frequency of the second formant (F2). The results also suggest that consonant recognition is accomplished (1) by using information from the amplitude envelope, including periodicity/aperiodicity, as cues to manner and voicing, (2) by using F1 as an aid to the identification of manner and voicing, and (3) by using information from cochlear place of stimulation to provide a very crude indication of the shape of the frequency spectrum above 1kHz.