John Gibbon N.Y.S. Psychiatric Institute and Columbia University Russell M. Church Brown University
When a signal of some duration is presented to an animal, what does the animal perceive as its duration? It is common in sensory psychophysics to postulate a subjective scale with scale values and variance properties so that, in some sense, what the animal perceives is a position on the y axis of the graph relating subjective time to physical time. This is a value that may be temporarily stored in a working memory. If food is delivered for a response after a signal of one duration, and not others, what does the animal remember? Presumably, some representation of the value stored in working memory is transferred to a more permanent reference memory. After the presentation of a signal of some duration, how does the animal decide whether or not to respond? Presumably, a comparison is made between a current experience of a duration in working memory and a remembered value in reference memory. The purpose of this chapter is to develop an information processing model of timing along these lines that is compatible with experimental results. The development focuses particularly upon where, in the course of perceiving and discriminating time, variability arises.
The key features of our view of the processing of time are exemplified in a simple experimental procedure that will be the reference experiment for our analysis of sources of variance in the perception and memory of time. This reference experiment is a straightforward temporal generalization task with rats reported earlier ( Church & Gibbon, 1982). The procedure is simple. The houselight in an operant chamber is turned off for a duration of time, T. Then a retractable lover is inserted into the chamber. After a 5 sec opportunity to respond, the lover is withdrawn and a 30 sec intertrial interval begins. If the light-off stimulus lasted the "correct" duration, say, 4 sec, a response is reinforced with a pellet of food. If it was not 4 sec, no reinforcement is forthcoming.
The typical result is that rats come to respond with high probability when the duration is 4 sec, and with decreasing probability for stimuli either longer or shorter than 4 sec. Figure 26.1 shows how this decline changes with changes in the size of the reinforced duration (S+). The