Avoidance learning to stimulus objects presented following shock

Abstract

An earlier informal experiment by Hudson (1950) in which rats learned to avoid a bundle of pipe cleaners presented only following shock is replica.ted and extended. Five groups of 20 Ss each received a single shock each while taking a sucrose pellet from a novel striped panel, A black-out period ranging from 1 to 40 sec. began with the onset of the 3/4 sec. shock. During the black-out the striped panel (forward-order CS) was removed; immediately following the black-out, a rubber toy hedgehog descended into the apparatus, Following a short exposure to the toy hedgehog and an intervening 24 hr. in the home cage, S was observed in the apparatus with the toy hedgehog at one end and the striped panel at the other. Control groups received either shock without the toy hedgehog or the toy hedgehog without the shock. All behavior was video recorded. Significant differential avoidance of the toy hedgehog occurred in the short inter stimulus interval groups (1, 5, and 10 sec.), but not in the 40 seCc group or in the control groups. In further analyses, individual’s were classified as differentially avoiding either the toy hedgehog, the striped panel, the shock location, the opposite end of the apparatus or no identifiable stimulus, according to two schemes. In the first, the basis of classification was differences in time spent in a normal posture at the two ends of the apparatus relative to a distribution of such differences in the unshocked control group. In the other, a combined score derived from differences in four other classes of behavior was the basis of classification. In both analyses, significant numbers of Ss from the 1, 5, and 10 sec. groups were identified as avoiding the toy hedgehog, while insignificant numbers of Ss from the 40 sec. and control groups did so. Only insignificant numbers of Ss avoided the striped panel. The results demonstrate that the "backward" association of the toy hedgehog with the shock is a reliable and robust phenomenon that can occur despite a 10 sec. UCS-CS delay, a single trial procedure, a 24-hr. delay between shock and testing, and the availability of a potential forward - order CS. The results cannot readily be explained either in terms of an unconditioned response to the toy hedgehog or simple sensitization. Both logical considerations and experimental results in backward conditioning preclude describing these results in terms of stimulus cuing. The results are interpreted as a. demonstration of the ability of rats to perceive causal agent-effect relationships in certain specific situations. Support for conclusions drawn from the inference that rats can make causal agent-effect connections is taken from the areas of belongingness, stimulus selection in avoidance learning, delayed taste -avoidance learning, novelty, reflexive aggression, and species-specific defense reactions. Theoretical literature relevant to this inference and the broader question of what is learned is discussed.