One-trial memory and habit contribute independently to matching-to-sample performance in rhesus monkeys (Macaca mulatta).

Abstract

Multiple memory systems often act together to generate behavior, preventing a simple one-to-one mapping between cognitive processes and performance in specific tests. Process dissociation procedures (PDPs) have been adopted in both humans and monkeys to quantify one-trial memory and habit, with the assumption that these two processes make independent contributions to performance. Violations of this independence assumption could produce artificial dissociations. Evidence for independence has been reported in humans, but similar tests have not been conducted with monkeys until now. In a within-subjects design using a matching-to-sample task, we manipulated one-trial memory strength and habit strength simultaneously. Memory delay intervals and encoding conditions affected one-trial memory scores without affecting habit scores. In contrast, biased reinforcement selectively changed habit scores but not one-trial memory scores. This behavioral double dissociation clearly shows that one-trial memory and habit can be manipulated independently, validating PDP as a valuable tool for cross-species studies of learning and memory and reinforcing the view that one-trial memory and habits are served by distinct brain systems.