Factors affecting key pecking in response-independent variable-time schedules: implications for theories of the conditioning of this response

Abstract

A series of five experiments examined factors affecting key pecking in variable-time schedules, where response-independent reinforcement is presented at varying intervals of time. In the first experiment, rates of responding to each color in a multiple variable-time, extinction schedule were proportional to the frequency of reinforcement in the presence of each component, but response fell off altogether in a multiple variable-time, variable- time schedule. In the second experiment, response in a multiple variable-time, extinction schedule was, shown to be affected by the length of the extinction segment but not by the length of the variable-time segment. In the third experiment, extensive exposure to a multiple variable-time, variable-time schedule resulted in a powerful suppression of responding in a subsequent variable-time, extinction schedule, relative to the effect of an equal amount of exposure to an extinction, extinction schedule. In Experiment 4, variable-time key pecking was maintained by presenting extinction segments on alternate days, but only in a group of pigeons with no prior experience with consecutive sessions on a multiple variable-time, variable-time schedule. Such experience, in another group, resulted in minimal responding. Experiment 5 showed that variable-time response is higher if the variable-time segment is temporally contiguous with the extinction segment than if the two are separated by 24 hours, and this effect appeared to be nonassociative in nature. These results were dealt with in terms of a contiguity-contrast theory, according to which the contiguity of stimulus and reinforcer produces stimulus-reinforcer associations which are activated by extinction experience, and in terms of the idea that an active suppression of response in variable-time schedules, which can be overcome by conditions which maximize activation effects. This suppression appears to be the result of the conditioning of responses which actively compete with key pecking.