There is uncertainty concerning the extent to which the auditory streaming effect is a function of attentive or preattentive mechanisms. The mismatch negativity~MMN!, which indexes preattentive acoustic processing, was used to probe whether the segregation associated with the streaming effect occurs preattentively. In Experiment 1, alternating high and low tones were presented at fast and slow paces while subjects ignored the stimuli. At the slow pace, tones were heard as alternating high and low pitches, and no MMN was elicited. At the fast pace a streaming effect was induced and an MMN was observed for the low stream, indicating a preattentive locus for the streaming effect. The high deviant did not elicit an MMN. MMNs were obtained to both the high and low deviants when the interval between the across-stream deviance was lengthened to more than 250 ms in Experiment 2, indicating that the MMN system is susceptible to processing constraints.Descriptors: Auditory stream segregation, Mismatch negativity, Event-related potentials, Streaming effect, Auditory sensory memory Multiple sources of acoustic energy can impinge on the ear constantly in our everyday experience. It is not uncommon for the sounds of voices and office equipment or the ringing of telephones to occur simultaneously. The task of the brain is to tease apart the cacophony of these sounds, forming meaningful representations of the incoming acoustic information. This process requires a mechanism for segregating the inputs into their original sources. A now classic example, the cocktail party phenomenon, illustrates the ease with which the brain is able to perform this task. Amid the steady din of party sounds~e.g., tinkling of glasses, multiple conversations, music!, the brain keeps the sources distinct. Auditory cues such as the location of the sound or the pitch of a speaker's voice help this process of segregating the total stream of sound, which has been called auditory stream segregation~Bregman, 1990!.When tones of a sufficient frequency separation are alternated continuously at a fast enough rate for a period of time, a streaming effect occurs~Bregman, 1978, 1990Bregman & Campbell, 1971!. The streaming effect, an aspect of stream segregation, is a perceptual phenomenon governed by both the rate of stimulation and the frequency relationship of a tonal sequence. The perception is that the sets of high and low tones split into separate streams of sound, one formed of the high tones and one formed of the low tones. It sounds somewhat like counterpoint in music, as though the two streams are occurring independently and simultaneously. The purpose of this perceptual segregation is presumably to sort the tones in terms of sound sources, thereby improving the ability to perceive patterns within them. Natural differences of acoustic properties emanating from different sound sources are often reflected in pitch~e.g., the voice of a man vs. the voice of a woman!. Therefore, within a mixture of sounds striking the ear, it is likely that sounds in a high fr...
