Sequential stream segregation based on spatial cues: Behavioral and neural measurements

Abstract

Differences in simulated spatial cues in the horizontal plane have been shown to enhance both voluntary and obligatory stream segregation of sounds with realistic spectro-temporal variations, such as sequences of syllables. In this experiment, listeners were presented with sequences of speech tokens, each consisting of a fricative consonant and a voiced vowel. The CV tokens were concatenated into interleaved sequences that alternated in simulated spatial positions. The interleaved sequences lasted 1 min. The listeners had to press a button each time they heard a repeated token. In the selective attention task, the listeners were asked to attend only one of the two interleaved sequences; in the global attention task, the listeners had to perceive the interleaved sequences as single stream to detect a repetition between the sequences. Simultaneous EEG measurements were made. The behavioral results confirmed that listeners were able to either attend selectively or globally, depending on the task requirements. The EEG waveforms differed between the two tasks, despite identical physical stimuli, reflecting the difference between global and selective attention. Both behavioral and EEG results reflected the effects of increasing spatial separation in enhancing selective attention and making global attention to the sequences more difficult.