TIMING PROCESSES IN THE REINFORCEMENT-OMISSION EFFECT

A comparison was made of the effects of long-term exposure to fixed-interval reinforcement with unsignaled and with signaled reinforcement omissions. When successive fixed-interval cycles ended in reinforcement, subjects showed a clear ''scalloped'' response pattern. When reinforcement was omitted, but a brief signal was given in lieu of reinforcement, responding in the next cycle started earlier than it did after reinforcement. When reinforcement was omitted without exteroceptive cues, response rates peaked near the time of reinforcement and then declined to a flat but substantial level. The classic reinforcement omission effect was observed, in that total responding was greatest after uncued omission, somewhat less after cued omission, and least after reinforcement. However, response rate plotted as a function of time showed that the uncued-omission condition had a very different function from that of the cued-omission or reinforcement conditions. A failed-discrimination account of the omission effect might accommodate the three functions if the discrimination is considered to be a temporal one. The temporal-discrimination account argues that high-rate responding reflects the accumulation of subjective time proximal to the memory of the time of reinforcement. The accumulation resets completely with food reinforcement, incompletely with cues in lieu of reinforcement, and not at all in the absence of cues.