R. W. Hukin
C. J. Darwin
Lab. of Exp. Psychol., Univ. of Sussex, Falmer, Brighton BN1 9QG, UK
The effective contribution of a 500-Hz component of a steady-state vowel was measured by the position of the vowel boundary on a /I/--/e/ continuum varying only in first formant (F1) frequency. Shifts in the vowel boundary were calibrated against those produced by level changes in the 500-Hz component. When the 500-Hz component was given an ITD of 666 (mu)s and the remaining components an ITD of -666 (mu)s, there was a small but significant reduction in the 500-Hz component's contribution to phonetic quality. When the vowel was embedded in a sequence of six 500-Hz tones, each having the same ITD as the 500-Hz component of the vowel, there was a significant reduction its contribution compared to when all components had the same ITD. When the sequence conditions were presented in the same experimental block as the previous ITD conditions, then the latter conditions did show an effect of grouping by common ITD. These results show that listeners can group sounds on the basis of ITD information, but their ability to do so is greatly enhanced if the position and frequency have been previously cued. [Work supported by UK MRC.]