| Description |
Three experiments were conducted in an attempt to establish exteroceptive stimulus control over pigeons' responding using a correlated block counter. In both Exps. I and II the counter did not acquire control over responding. Experiment I utilized a singles-key procedure, a block counter which apparently lacked salience, and there were no consequences for not attending to the stimulus. In Exp. II consequences were imposed for failing to attend to the stimulus and a S now was required to peck a second key for food at the end of a chain. The time out consequences of Exp. II proved to disrupt responding. Experiment III utilized a distinct, red and white block counter stimulus, fully illuminated, behind a transparent key (key I). A subject's erroneous response on the second key (key II) before completing the ratio on key I had the following consequences: (a) The error reset the component requirement, wiping out the accumulated sum of correct responses for that component, and (b) also set the block counter back one step. This arrangement allowed an increase in a S's work and an increase in the interreinforcement interval, without inducing long pauses. The Ss learned to peck key I (FR 60 plus conunter), making few errors on key II, the last link (FR 1) in the chain. Weak stimulus control was established rapidly in Exp. III and extensive training substantially strengthened control for most Ss. All four Ss gave evidence of at least moderate control as measured by component latency, duration, key II activity, and key II errors, all showing a clear and marked gradient. Moderate control was shown in all Ss' resistance to disruption during manipulation of component sequences. Strong control was evident in two Ss throughout much of the testing, particularly in a probe in which a component was locked in position for an arbitrary duration. One S demonstrated strong started strong control as well as conditioned reinforcing effects of the counter components. In a final test, this bird learned a treadle-pressing response with only changes of the counter components as reinforcement (it.e., Sr s). The strong stimulus control and the conditioned reinforcing effects shown in one bird were credited to respondent as well as operant conditioning. While no measures of the interval were taken, it appeared that the time between the last response on key I and the key II food-getting response was critical, perhaps having to be mad in the period of from 0.0 to 1.0 sec. It is likely that the one bird giving evidence of conditioned reinforcing effects met the requirement of responding within 1.0 sec. The discriminative and reinforcing effects of the block counter components in Exp. III aided in achieving and maintaining stimulus control. |
