The voicing contrast between English /p/ and /b/ as a function of stress and position-in-utterance

Abstract

Normal adult female speakers of English produced nonsense utterances like “Papa papa” and “Baba baba baba” with a tube inserted in the corner of the mouth. Bilabial constriction of /p/ and /b/ was defined by means of variation in supraglottal air pressure; voicing (glottal pulsing) was monitored by means of a throat microphone. The percentage of bilabial closure intervals evidencing signs of glottal activity was calculated from these two signals. In addition, the frequency with which /p/ and /b/ showed a clear break in voicing during the closure interval was tabulated. The stop voicing pair /p-b/ was clearly distinguished by voicing in utterance-medial positions, but less so at the margins of utterances. In utterance-initial position only two of eight speakers produced /b/ with closure voicing (i.e., prevoiced), and in utterance-final words post-stressed /b/ was partially devoiced by half the speakers.