Articulatory effects on speech perception: a second report

Abstract: The effects of repeated articulation on speech perception were assessed to determine whether the production and perception of speech rely on the same processing mechanism. Subjects were asked to whisper an alternating sequence of syllables repetitively, in the absence of auditory feedback, and were then presented syllables to identify auditorily. In the first experiment, subjects uttered a sequence of the syllables [si]–[su] or [sti]–[stu] and then identified a series of seven synthetic speech syllables spanning the distinction between [si] and [sti]. These syllables varied from one another in the duration of silence preceding the vowel. A significant effect of perceptuo-motor adaptation was obtained for [sti]-[stu] but not [si]–[su] articulation. In Experiment II, the same test procedure was applied to another series of synthetic syllables, [ba]–[wa], differing from one another in the duration and rate of the first three formant transitions. A significant effect was obtained after repetitively uttering a sequence of [wa]–[ya] syllables but not after uttering a sequence of [ba]–[da]. In Experiment III, a category judgement response was required for the [ba]–[wa] stimuli, using the integers 1 through 7. A signal detectability analysis was applied to the results of this experiment and showed a significant perceptuo-motor effect after uttering a sequence of [ba]–[da] but not [wa]–[ya]. Finally, the category judgement task was used in Experiment IV with another series of syllables, [ba]–[pha], which differed from one another in voice-onset time. No significant effects of perceptuo-motor adaptation were obtained with the [ba]–[pha] series, and their absence was attributed to the absence of voice-onset time as a cue for distinguishing such syllables in whispered speech. The perceptuo-motor results are compared with the results obtained with a strictly perceptual adaptation paradigm.