The effect of conditioned stimulus intensity in classical conditioning of the common carp

Abstract

The purpose of this study was to determine whether intensity of the conditioned stimulus (CS) in classical conditioning exerts its effect on associative processes ("learning") or on nonassociative processes ("performance"). An attempt was made to answer this question by using a factorial design in which two major experimental groups were trained under either high or low CS intensity conditions. Halfway through training (after five trials) one half of each major group was shifted to the opposite CS intensity condition and training was continued. In order to assess the effects resulting from the shift itself, four control groups, each group trained under CS intensity conditions comparable to those of each of the experimental groups, were given pseudoconditioning trials on which the CS and unconditioned stimulus (US) were not paired. A delayed conditioning situation was used in which the CS was the onset of light, the US was an electrical shock, the conditioned response (CR) was a suppression of respiratory activity, the CS-US interval was four seconds, and the intertrial interval was variable with a mean of four minutes. A total of 127 fish (Cyprinus carpio) were tested, there having been approximately 20 fish in each experimental subgroup and 12 fish in each control subgroup. In order to maintain the regularity of respiratory activity, fish were trained in a constant direct current electric field at an intensity 2 of 0.07 ma./cm.^2 and in an anodic orientation. The results of this study indicated a reliable CS intensity effect during conditioning: fish trained at a high CS intensity performed at a higher level than fish trained at a low CS intensity. Substantial effects resulting from the CS intensity shift itself were found and were partially removed from the experimental group data by an adjustment procedure involving the subtraction of control group scores. These shift effects were interpreted as consisting mainly of unconditioned-response-to-light effects. contrast effects, and generalization effects. After taking into consideration the shift effects thought to remain in the final experimental group data. the following conclusions were made. In the classical conditioning of carp. and under the conditions imposed by this study, (a) there was evidence that CS intensity exerted an effect mainly on associative processes. and (b) there was inconclusive evidence with respect to the question of whether CS intensity exerted any effect on nonassociative processes (other than shift effects), although there was evidence suggesting that if there were nonassociative effects they must have been quite small as compared to associative effects. Some secondary findings of a tentative nature concerning electric field effects on fish were also discussed.