Date of Award

Spring 1983

Project Type


Program or Major


Degree Name

Doctor of Philosophy


The control of behavior by prior responses and stimuli was studied using pigeons in a signalled reversal learning procedure. During regular trials, an intertrial interval was followed by illumination of two keys. Initially one key was designated correct. A peck to either key turned off both keys and, if correct, produced food. After a criterion number of consecutive correct responses occurred, a reversal/nonreversal sequence was initiated. Either the reversal or nonreversal signal was presented briefly during the next intertrial interval. Following a reversal signal only, the correct and incorrect keys were reversed. A correct response on the next trial resulted in a larger food reinforcer. Responding on this trial type provided the major dependent measures of behavior. Following this, the regular sequence was reinstated.

The duration of the intertrial interval was varied from 0 to 30 sec on reversal/nonreversal trials, and the reversal/nonreversal signal could occur either at the beginning or end of the interval. Accuracy was inversely related to delay and was greater on late-signal trials. Finally, during baseline, the houselight was on for some birds and off for others during the interval. In a later phase, all birds were exposed to both conditions. Baseline illumination did not affect accuracy, but accuracy was lowered by changes in illumination.

The results were examined with respect to two cognitive accounts of memory coding. However, a behavioral account based on the effects of delays on stimulus control was adequate, and entailed fewer theoretical assumptions.