The perception of intervocalic /t/ and /d/: A comparison of child and adult listeners

Abstract

Child and adult listeners were presented with two sets of stimuli to be judged as either petal or pedal. One set was constructed by editing a naturally produced token of petal, and the other, a token of pedal. Construction of each set involved systematically varying three acoustic cues important for the voicing distinction between intervocalic stops: (1) preceding vowel duration, (2) stop closure duration, (3) voice onset time. Besides the silent closures used in each set, some tokens with glottal buzz were used in the pedal set. Preliminary analysis shows that, for silent closures, VOT had the strongest effect on the subjects' responses, except for the adults' responses to stimuli created from petal. There closure duration had a slightly greater influence. Closure duration otherwise had a strong effect just secondary to VOT. The effect of preceding vowel length, though observable, was weak. Stimuli with buzz-filled closures were heard almost entirely as pedal, regardless of other cue values. Adults and children showed similar overall patterns of responses, but with the adults being more sensitive to changes in vowel and closure duration.