An Application of Wagner's Standard Operating Procedures or Sometimes Opponent Processes (SOP) Model to Experimental Extinction

The present study used simulations to examine whether Wagner's Standard Operating Procedures or Sometimes Opponent Processes (SOP) model explains various extinction phenomena. These included the so-called signature characteristics of extinction-renewal, reinstatement, and spontaneous recovery-as well as the effects on extinction of manipulations such as preexposure, the interval between extinction trials, the rate at which reinforcement ceases, and the presence of other stimuli. The simulations showed that SOP accounts for the effects of each of these manipulations. It does so for 2 reasons. First, the form of stimulus representation and rules for generating associative change mean that SOP can explain conditioning phenomena by appeal to changes in processing of both conditioned (CS) and unconditioned (US) stimuli, in contrast to other theories which (=fine changes in processing to either the CS (e.g., attentional theories) or the US (e.g., the Rescorla-Wagner model). Second, the processes that generate associative change in SOP are at least partially independent of those that generate performance. Hence, stimuli that differ in associative strength can extinguish at the same rate, and stimuli with equal associative strength can undergo different amounts of renewal, reinstatement or recovery.