The major focus of this research proposal is on the early processing of acoustic information. The principal tool used in this endeavor is an event-related potential termed the mismatch negativity (MMN). Research with the MMN has uncovered a system located in auditory cortex that defects change in the acoustic environment on a automatic, pre-attentive basis. The proposal has four basic aims. One aim is to determine whether the system tracks sources of sounds, or objects rather than tracking acoustic features across objects. Another aim is to examine the system's ability to be set for certain stimuli prior to conscious detection, and in that sense be able to predict future stimuli. Three experiments will focus on this issue. One re-examines a previous study which concluded that the MNM system is not affected by the occurrence of predictable stimuli. A second will ascertain whether a change in one acoustic feature can set the MNM system for a change in a different feature. A third experiment will determine whether conscious awareness of a forthcoming acoustic change will affect the MMN system. Another aim is to characterize the memory associated with the MMN as being different from sensory memory. One experiment will test the hypothesis that the memory is not accessible to conscious processing. Another will examine whether it is common for these memories to become dormant and be reactivated by a single reminder stimulus. Finally, we will attempt to provide more evidence that the MMN system is pre-attentive and automatic by showing that features of acoustic stimuli are analyzed in parallel by the MMN system and in series by conscious processing.