Operant conditioning and slow potential changes from monkey cortex

Walter et al. (1964) reported that a slow negative variation in the EEG could be recorded from the human scalp during the inter stimulus interval of a constant foreperiod reaction time experiment. This negative potential, called Contingent Negative Variation (CNV), has since been correlated with such variables as expectancy (Walter et al., 1964), motivation (Irwin et al. , 1966), and attention (Tecce and Scheff, 1969). The demonstration that an analogous slow potential (SP) can be recorded in rhesus monkeys (Low et al., 1966a) provided an opportunity to investigate the physiological origins of CNV. In order to delineate the source and functional significance of CNV, the present study attempted to obtain information concerning the distribution of SP changes over monkey cortex. Two modes of reinforcement were employed in an attempt to determine the relationship of SP with behavior. Transcortical platinized-platinum electrodes were stereotaxic- ally implanted in five rhesus monkeys. Subjects were trained to respond in order either to receive food reinforcement or avoid shock. A warning stimulus (click) was followed one second later by the imperative stimulus (1500 Hz tone of 500 msec, duration); subjects were required to respond during the tone period by pressing a lever. Approximately 100 trials were given a day, five days a week, for a two-month period. Brain electrical activity was suitably amplified and recorded on magnetic tape for subsequent computer analysis. Negative SP changes were recordable from many portions of frontal cortex. The most marked finding in the present study was the demonstration of the existence of two independent SPs related to the conditioning task. A frontally dominant potential developed gradually with training while the centrally dominant SP was of significant magnitude from the onset of training. Each potential was of equal magnitude bilaterally and persisted for the duration of the two-month testing period. Type of reinforcement was a significant variable in determining the distribution of cortical SPs. Shock reinforcement caused an increase in premotor regions that was not manifested .in frontal areas. It was suggested that the frontal potential was analogous to CNV recorded in humans and reflected the association of the two stimuli while the centrally dominant SP was an indication of motor inhibition and readiness to respond. The possible relationship between these potentials was discussed in terms of triggering mechanisms for the initiation of timed movements.