Abstract
The perceptual load model in selective attention (Lavie & Tsal, 1994) suggests that selection locus depends on perceptual load. We extend this idea to cross-modal processing and investigate whether vision and audition share the same limited source of attention by using the paradigm of repetition blindness (RB; Kanwisher, 1987) with task-irrelevant sounds. RB is a phenomenon that the observer fails to detect the second occurrence of a repeated item in a rapid serial visual presentation (RSVP). We manipulated perceptual load in the temporal dimension and examined whether the effect of irrelevant sounds on RB varies with perceptual load. The two critical items (C1 and C2) embedded in each RSVP were either accompanied by two sounds with the same pitch or with different pitches, and a no-sound control condition was used as the baseline. The stimulus onset asynchrony and lag between C1 and C2 were manipulated to lead to a high-load or a low-load condition. Results show that the facilitaroty effect of sounds on character identification varies with perceptual load. However, only those who are aware of the difference in sounds are affected by them, and this participant group also performs more accurately for the repeated control condition than the other group. Together, these results suggest that visual-auditory processing share a common limited resource to execute selection.
Supported by the National Science Council of Taiwan, NSC95-2752-H-002-007-PAE and 95-2413-H-002-020