CS-Termination as a Factor in the Emergence of Anticipatory Avoidance

Abstract

The role of response termination of the CS in the reinforcement of shuttlebox avoidance responding is well established (Mowrer & Lamoreaux, 1942; Kamin, 1957b). Previous study of this effect suggested the incidental finding (Kamin, 1957a) that delay of CS-termination on the initial escape trials retarded the of avoidance responding. The running response, though instrumental in terminating the US, did not readily become anticipatory unless, on escape trials, it terminated both CS and US. This observation, confirmed in a subsequent study (Kamin, Campbell, Judd, Ryan, & Walker, 1959), does not make clear whether response termination of the CS positively contributes to the emergence of avoidance, or whether continuing the CS after the escape response inhibits avoidance responding. These alternatives can be de-confounded by use of an experimental group for whom, on escape trials, the CS terminates before the escape response is made. The present note contrasts performance under such conditions with performance under prompt, and delayed, response termination of the CS. Ss were 24 experimentally naive hooded rats, about three months of age. The apparatus was a modified Mowrer-Miller shuttlebox, more fully described in an earlier paper (Kamin, 1957a). The CS was a 74 db buzzer, and the US 1.1 ma. electric shock. Interstimulus intervals were controlled by a Hunter timer. There were, for all groups, 100 trials in a single session. The CS-US interval was 5 sec., and the intertrial interval, 1 min. Whenever, in any group, S responded to the CS by running before the scheduled delivery of the US (avoidance trial), the CS was at once terminated and the US withheld for that trial. Whenever, in any group, S ran in response to the US (escape trial), the US was at once terminated. The sole differential treatment of groups occurred on escape trials, and only with respect to termination of the CS. There were three experimental groups, with eight Ss randomly assigned to each. With Group I, the CS on escape trials terminated simultaneously with delivery of the US; since escape latency approximates .75 sec., this was the average interval by which CS-termination preceded the escape response. (Termination of the CS after some briefer interval would have involved trace, as opposed to delayed, conditioning; this would in turn involve differential treatment of groups on avoidance trials.) With Group 11, the CS on escape trials terminated immediately with S's running response. With Group 111, the CS continued for 5 sec. after S's response. The basic procedure, as well as data for Groups I1 and 111, are given in detail elsewhere 'This research was supported by a grant from the Associate Committee on Applied Psychology of the National Research Council of Canada, and facilitated by a University College grant-in-aid from McMaster University. Thanks are due to F. Van Fleet and J. Walker for experimental assistance.