Crosslanguage Study of Voicing in Initial Stops

Abstract

In describing the phonemes of a language, phoneticians often find it useful to divide them into “voiced” and “voiceless” categories. For stop consonants, some languages are said to utilize aspiration in conjunction with voicing to yield two, three, or four categories, while in some few languages categories are said to be distinguished solely by differences in aspiration. Some phoneticians, moreover, speak of strongly articulated (fortis) and weakly articulated (lenis) categories. Earlier work has shown that the two English categories /ptk/ and /bdg/ can, in initial positions, be distinguished acoustically by a difference in the time interval between the burst that marks stop release and the onset of periodicity that reflects laryngeal vibration. This measure of voice-onset time was applied to eleven languages that vary in the number and phonetic description of their initial-stop categories. In all these languages, this measure was found to be highly effective as a means of separating phoneme categories; the languages vary in where along the continuum of voice-onset time they put the boundaries between stops. Speech synthesis is being used to test the perceptual relevance of this dimension, and physiological data are being collected on underlying articulatory mechanisms. [This work was supported in part by the National Science Foundation under grant G-23633.]