The role of perceived duration in the identification of vowels

Abstract

The role of perceived duration in discriminating between tense and lax vowel quality was explored. Two target vowels, having identical formant structures but different durations, were employed as stimuli in a backward recognition masking task. One of the two target vowels was presented on each trial, followed after a variable silent intervowel interval by a masking vowel. Six masking vowels were utilized, differing in both formant structure and duration. The subjects’ task was to identify the target vowel as being tense or lax. In Experiment 1, identification of the long target improved monotonically with increases in the intervowel interval, while identification of the short target decreased slightly in accuracy with increases in the intervowel interval. Identification of the long target vowel was most accurate when followed by a long masking vowel and least accurate when followed by a short masking vowel. While the opposite results were obtained for the short target vowel. In Experiment 2, subjects rated the quality of the target vowel along a continuous scale between the two vowel alternatives. The target vowel was rated as being progressively more tense with increases in the durations of the target vowel, the masking vowel, and the intervowel interval. The results were interpreted within an information processing model describing the perception of duration. Both experiments demonstrated that perceived duration rather than actual stimulus duration is critical to the tense-lax distinction.