Vocal fry: A cue for voicing in postvocalic stops

Abstract

The hypothesis that vowel‐termination characteristics signal the voicing distinction of post‐vocalic stops in English [F. Parker, J. Phonet. 2, 211–221 (1974)] was investigated by recording CVC(C) monosyllables produced by six speakers and analyzing the utterances spectrographically. Although the vowel terminations before voiced stops were asystematic, the voiceless stops typically displayed a coarticulated glottal stop (evidenced by vocal fry). More importantly, the glottal stop appeared only with labial and alveolar voiceless stops, thus accounting for two phenomena noticed (but not explained) in the literature. (1) Vowel durations preceding voiced/voiceless nonvelar stops overlap, whereas those preceding velars do not [L. Raphael, J. Acoust. Soc. Am. 51, 1296–1303 (1972)]. (Vocal fry is not available as a voicing cue in velars, thus making vowel‐length crucial.) (2) Final voiced/voiceless velar stops are confused more often than voiced/voiceless labials and alveolars [M. Wang and R. Bilger, J. Acoust. Soc. Am. 54, 1248–1266 (1973)]. (The latter had vocal fry as a secondary cue, thus enhancing their perceptibility.) In short, a glottal stop coarticulated with a following nonvelar voiceless stop produces vocal fry, which not only serves as a secondary cue for voicelessness, but also helps explain the differences in vowel length and perceptibility exhibited by velar and nonvelar postvocalic stops.