Operant punishment of eye elevation in the green crab, Carcinus maenas.

Extension of the eye after reflex withdrawal was suppressed by punishing each extension with a brief puff of air. Experimental animals showed a decrease in the rate of responding, and an increase in the latency to the next response during 30-min sessions. The effect of punishment per se was controlled for by the use of yoked animals that received punishments whenever the experimental (master) animals did. This control group did not show the increased latency, and kept the eye erect for most of the session. Experiments were performed with pairs of animals, one eye of each used as master or control, or, alternatively, with single animals in which one eye served as the yoked control for the other. This latter group gave more reliable and reproducible differences between master and yoke than the pairs of animals. Retention was tested by subjecting animals to three sessions separated by a 12-hr rest. The results indicated some savings but this was not a dramatic effect. To demonstrate that the learning was operant in nature, that is, that it depended on the contingency between the act of eye extension and punishment, experiments were performed in which a delay was introduced between the response and the onset of punishments. A delay of 20 s was found to completely eliminate the learned suppression: animals showed latencies close to that of naive animals and responded at a constant high rate throughout the session. Delays of 10, 5, and 2.5 s were found to have a decreasing effect on the learning, and a delay of 1.25 s produced behavior that was within experimental error of that of animals subjected to immediate punishment.