Further studies of the influence of sentence rate on the closure duration cue for voicing in medial stop consonants

Abstract

Previous research has shown that the closure duration cue for a medial voicing distinction, as in the contrast rabid versus rapid, is perceived in relation to the tempo of a carrier sentence: at faster rates of speech, the voicing boundary moves toward a shorter closure duration [R. Port, J. Acoust. Soc. Am. 59, S41–S42 (A) (1976)]. In this experiment, we asked whether the voicing judgment is influenced by the overall rate of the sentence, which depends both on the articulation rate (the rate at which the speech itself is produced) and pause time, or whether it is influenced primarily by only one of those variables. By independently manipulating articulation rate and pause time, we found that articulation rate is a significantly better predictor of voicing judgments than either overall rate or pause time. Furthermore, when we asked subjects to provide subjective estimates of the rate of our experimental sentences, we found that those judgements were also based almost entirely on articulation rate. These results suggest first, that the interpretation of the closure duration cue is influenced by the articulation rate of the sentence, with pause time having little effect and, second, that listeners use the same rate information, namely, articulation rate, both when estimating the rate of the sentence, per se, and when adjusting for rate in processing the closure duration cue. [Work supported by NIH.]