How do listeners perceptually organize complex auditory sequences composed of two or more simultaneous sub -sequences? We demonstrate that previous results obtained with two-sub-sequence mixtures generalize to more comple x ones, and describe four new organizing principles. A new paradigm identified whether listeners perceive the mixt ure as a single unit (integrative listening) or segregate it into two (or more) perceptual units (stream segregati on). Listeners heard two complex sequences, each composed of 1, 2, 3, or 4 isochronous sub-sequences (each define d by a specific frequency/tempo combination). Their task was to say whether a small temporal irregularity created in one of the sub-sequences was situated in the first or second complex sequence. Attention was focused on one par ticular sub-sequence by preceding the complex sequences by a single sub-sequence. Two control experiments confirm ed that detection was similar in all sub-sequences and was not possible if listeners did not perceptually organize the complex sequences into streams. Then, Experiment 1 showed that the smallest frequency separation under which listeners were able to focus on one sub-sequence was unaffected by the number of co-occurring sub-sequences, sugge sting that the non-focused sounds were not perceptually organized into streams. Experiment 2 showed that temporal detection improved progressively (not abruptly as has been assumed previously) as the frequency separation between the sub-sequences was increased from .25 to 6 auditory filters. Finally, we propose a model of the perceptual org anization of complex auditory sequences.
Article paru dans : Journal of Experimental Psychology: Human Perception and Performance vol. 25