SPECIAL REPORT Two cognitive systems simultaneously prepared for opposite events WALTER RITTER, a ELYSE SUSSMAN, a DIANA DEACON, b NELSON COWAN, c and HERBERT G. VAUGHAN, JR. a a Department of Neuroscience, Albert Einstein College of Medicine, Bronx, New York, USA b Department of Psychology, City College of New York, New York, USA c Department of Psychology, University of Missouri, Columbia, USA Abstract Event-related potentials ~ ERPs! were recorded during a go 0no-go reaction time ~ RT! task in which subjects responded to rare target tones and withheld response to frequent tones. In a predictable condition, a rare visual stimulus signalled the impending occurrence of a rare tone, whereas a frequent visual stimulus signalled that a frequent tone would be presented. The rare visual stimuli elicited P3, associated with violations of expectations, whereas the rare tones, being predictable, did not. The rare tones elicited the mismatch negativity ~ MMN!, a component associated with preattentive deviance detection, despite the fact that they were expected. RT was faster in the predictable condition than in another condition in which the tones were not predictable. The P3 and RT data indicate, respectively, that subjects anticipated and were ready to respond to the target tones. The MMN result indicates that immediately before target tones, the preattentive system underlying the MMN was set for frequent tones, being unaffected by the information available to the higher order system. Thus, the higher order cognitive system associated with P3 and the lower order cognitive system associated with the MMN were prepared simultaneously for opposite events. Descriptors: ERPs, Mismatch negativity, P3, Preparation, Modularity Two event-related potential ~ ERP! components reflect preparation for forthcoming stimuli associated with higher and lower order cognitive systems. The P3 component, which is positive in polarity and largest at midline parietal sites, is elicited when subjects dis- criminate stimuli on a physical or semantic basis ~ Pritchard, 1981!. The occurrence of P3 has been hypothesized to reflect the updating of working memory such as might result from violations of ex- pectation ~ Donchin & Coles, 1988!. The mismatch negativity ~ MMN! is elicited by changes in acoustic stimuli, such as infre- quent deviant tones interspersed among a train of identical stan- dard tones ~ Näätänen, 1992!. In contrast to P3, the MMN can be elicited whether or not subjects pay attention to the tones. The MMN is negative in polarity at midline frontal recordings and often inverts in polarity at the mastoids. The MMN is considered to be an output of a preattentive, automatic deviance detection system, located in auditory cortex ~Alho, 1995!, that serves to alert higher systems to changes in the acoustic environment ~ Näätänen, 1992!. It has been hypothesized that the deviance detection system generates and maintains representations of invariant aspects of the recent acoustic past ~ Ritter, Gomes, Cowan, Sussman, & Vaughan, 1998!. In this sense, the deviance detection system is prepared for subsequent stimuli that specifically match these representations of invariances. All tones that fully match these representations do not elicit the MMN, whereas stimuli that fail to match these represen- tations in any discriminable way do elicit the MMN. The purpose of the present study was to determine whether the deviance detection system has access to information about forth- coming stimuli available to a higher cognitive system. Subjects responded differentially to standard and deviant tones and were informed prior to each tone which stimulus was about to occur. The critical issue was whether expectation for the stimulus about to be delivered would affect the setting of the deviance detection system. Specifically, would knowledge that a deviant was about to occur preclude automatic elicitation of the MMN? It has been shown that if subjects are informed prior to each stimulus exactly which stimulus is to be delivered, the P3 is small or absent ~Sutton, Braren, Zubin, & John, 1965!. In the present study, stimuli were presented as pairs, each pair consisting of a visual stimulus followed by one of two tones. In the predictable condition, a visual stimulus informed subjects that an infrequent deviant tone would follow, requiring a response. An- other visual stimulus informed subjects that a standard tone would follow that did not require a response. In the unpredictable condi- tion, in which subjects were also required to respond to the infre- quent deviant tone, the visual stimulus was identical on all trials and therefore did not inform subjects which tone would follow. This research was supported by USPHS grants NS30029, HD21338, HD01799, and DC00223. We thank Drs. Robert Efron, Steven Hackley, and Emanuel Donchin for helpful comments. Reprint requests to: Walter Ritter, Dept. of Neuroscience, Albert Ein- stein College of Medicine, 1410 Pelham Parkway South, Bronx, NY 10461, USA. ~ E-mail: monterey@bcn.net. Psychophysiology, 36 ~1999!, 835–838. Cambridge University Press. Printed in the USA. Copyright © 1999 Society for Psychophysiological Research 835