EFFECTS OF CONTRALATERAL PRESENTATION AND OF INTERAURAL TIME DIFFERENCES IN SEGREGATING A HARMONIC FROM A VOWEL

The extent to which the 500-Hz component of a steady-state vowel contr ibuted to its phonemic category was measured by estimating the positio n of the /I/-/epsilon/ phoneme boundary along a first formant (F1) con tinuum. Shifts in the phoneme boundary were calibrated against shifts produced by physical changes in the level of the 500-Hz component. Whe n the 500-Hz component was presented contralateral to the rest of the vowel, the phoneme boundary changed by an amount equivalent to a physi cal reduction in the level of the component of about 5 dB (experiment 1). However, giving the 500-Hz component an interaural time difference (ITD) of -666 mu s and the remainder of the vowel an ITD of +666 mu s did not reduce the component's contribution to vowel quality (experim ent 2). Placing the 50O-Hz component of the vowel in a sequence of 500 -Hz tones substantially reduced the Contribution of the 500-Hz compone nt; the contribution was further reduced by giving both the tone seque nce and the component an ITD of -666 mu s and the rest of the vowel on e of +666 mu s (experiment 2). When these tone-sequence conditions wer e presented in the same experimental block as the ITD conditions witho ut a preceding tone sequence, these latter conditions did then show an effect of grouping by ITD. The results suggest that listeners can per ceptually segregate sound on the basis of different ITDs, but that thi s segregation is substantially enhanced if the direction has previousl y been cued. (C) 1995 Acoustical Society of America.