Control of voice-onset time in the absence of hearing: a review.

The relation between partial or absent hearing and control of the voicing contrast has long been of interest to investigators, in part because speakers who are born deaf characteristically have great difficulty mastering the contrast and in part for the light it can cast on the role of hearing in the acquisition and maintenance of phonological contrasts in general. One of the phonetic characteristics that distinguish voiced from voiceless plosives in English (p/b, t/d, k/g) is voice onset time (VOT): the interval from plosive release to the onset of voicing of the following vowel. This article first reviews research on VOT anomalies in the speech production of prelingually and postlingually deaf speakers. Then it turns to studies of the mechanisms in speech breathing, phonation and articulation that underlie those anomalies. In both populations of speakers, there is a tendency for the difference between voiced and voiceless VOT to be reduced, to the point for many speakers that there is in effect a substitution of the voiced for the voiceless cognate. The separation of the cognate VOTs can be enhanced when some hearing is restored with a cochlear implant. Both populations also present anomalies in speech breathing that can hinder the development of intraoral pressures and transglottal pressure drops that are required for the production of the VOT contrast. Its successful management further requires critical timing among phonatory and articulatory gestures, most of which are not visible, rendering the VOT contrast a particular challenge in the absence of hearing.