Training severely hearing-impaired children in the discrimination of the voiced-voiceless distinction

Six severely hearing-impaired children who were initially unable to set voiced-voiceless boundaries at voice onset times between 20 and 40 msec in the discrimination of speech were trained to do so by means of a structured auditory training program. An additional tactile cue, gradually withdrawn over four training steps, was used to signal presence of voicing (voice onset times 20 msec or less) during the initial phases of training. Half of the subjects were trained only with the bilabial stops /b-p/ and the remainder with the bilabial, alveolar, and velar stops /b-p/, /d-t/, and /g-k/. Although only syllables containing the vowel /a/ were used in training, all subjects generalized to syllables containing the vowels /u/ and /i/ and some subjects generalized to stop-initiated words. Implications for structured auditory training procedures are discussed.