Occasion setting of timing behavior

Abstract

Rats were trained on a temporal switching discrimination, with 2 features, A and B, signaling 2 target conditioned stimuli, x and y; feature offset and target onset were separated by a 5-s feature-target interval, and all target stimulus presentations terminated in a food pellet. The target conditioned stimuli were either short or long (6 or 30 s): Specifically, when signaled by A, x was 6 s and y 30 s, but when x and y were signaled by B, x was 30 s and y 6 s. Trials with 6-s and 30-s targets were termed short and long trials, respectively. Probe tests indicated that the animals correctly anticipated when food was to be delivered on these 2 types of trials. In further testing, the interval between feature offset and target onset was lengthened, to investigate the precise mechanism underlying this behavior. This manipulation did not have a substantial effect on discrimination performance. These results are discussed with reference to theories of occasion setting, timing, and configural learning.