ON THE MECHANISM OF THE POST-ASYMPTOTIC CR DECREMENT PHENOMENON

Three experiments with 49 dogs explored the decrease in CR magnitude that sometimes occurs when CS-US pairings are continued beyond those needed to reach maximum CR magnitude. The first experi- ment confinmed the existence of the phenomenon, obtaining less condi- tioned excitation after 300 CS-US tfials than after 18 CS-US trials. The second experiment demonstrated that the phenomenon is not dependent upon paired CS-US presentations because 18 CS-US pairings yielded little excitation if preceded by, followed by, or intermixed with 282 US presentations. The third experiment indicated that in contrast to the decremental effects on excitatory conditioning, inhibitory conditioning was faciliated by large numbers of prior US exposures, suggesting expla- nation of the post-asymptotic decrement phenomenon by opponent-pro- cess theory. Pavlov (29, lecture 14) argued that the ultimate destiny of conditioned stimuli (CSs) - reinforced or not - was to become "inh3bitory". In sup- port of this, he described experiments by Shishlo, Petrova, and Speranksy in which a stimulus (CS) was reliably reinfonced and after the first few trials elicited a large stable conditioned reflex (CR); but with continued reinforced trials the CR diminished and even disappeared. The pheno- menon, later reported in the American literature as well (11, 12 and others), may be described as post-asymptotic CR decrement and has been