Voicing of obstruents: a study of deaf and hearing

Abstract

This paper analyses voice properties in obstruents in Brazilian Portuguese, and considers the oral production of severe deaf speakers who use the Brazilian Sign Language. Five experiments were carried out with six deaf participants and one experiment with six hearing participants. The acoustic analysis showed that deaf speakers produce only voiceless obstruents at the beginning of words, regardless of whether the voice property of the word initial obstruent was voiced or voiceless. The notion of covert contrast was taken into consideration to posit the hypothesis that a vowel which follows a voiced obstruent would be longer than a vowel which follows a voiceless obstruent. This hypothesis was proven to be correct offering, thus, evidence that covert contrast operates in a deaf person’s speech and enables the subject to discriminate between voiceless and voiced obstruents.