Voicing cues in English final stops

Abstract

The relation between the acoustic characteristics of final stop syllables and the perception of the voicing distinction was investigated. Stimuli which had been truncated at various points before the end of a vowel-consonant syllable were presented to subjects for identification. The results indicated that the formant transitions, closure, burst, and vowel duration are important in determining whether a stimulus is heard as voiced or voiceless. In addition, there are differences in the formant frequencies of the preceding vowel present within the first 50 ms of the syllable which appear to affect the pattern of voiced-voiceless judgments. On the basis of the results, two classes of voicing cues for final stops were proposed: offset cues, which occur in the vicinity of closure, and vowel cues. It was suggested that offset cues in final stops may be similar to onset cues to voicing in initial stops, and that the amount of low frequency energy in the vicinity of onset or offset may serve as a common basis for the perception of the voicing distinction in initial and final position.