Recovery from conditioned suppression to a partially overlapping compound stimulus

Abstract

In a conditioned suppression paradigm, a partially overlapping compound stimulus signaled occurrences of electric shock. That compound CS consisted of 3 min of continuous illumination of the houselight with three discrete 5-sec presentations of a tone superimposed. A .5-sec electric shock was coterminous with each tone presentation. Rats that received this treatment in early conditioning sessions showed considerable recovery from conditioned suppression to the houselight as the experiment progressed. However, the effect was not reversible, and it could not be demonstrated in rats that experienced extensive prior conditioning to the houselight alone. These results are discussed in relation to a hypothesis concerning the modulation of behavioral control exerted by elements of compound stimuli and as they relate to a recent theoretical model for Pavlovian fear conditioning.The hypothesis that contextual stimulus control in certain Pavlovian conditioning situations is modulated by the existence of special CSs that reliably signal US events is suggested by results from two lines of research. One of these lines of research is Seligman's (1968) study of contextual cues as reliable predictors of the absence of electric shock (i.e., as "safety signals"). That research demonstrated that when rats (already trained to press a lever to obtain food) experience signaled electric shock they show reduced baseline response rates in early sessions of the experiment, gradually recover preshock baseline rates, and finally show reduced rates only during CS presentations. Rats for whom electric shocks and CSs are uncorrelated also show reduced baseline response rates in early sessions, but in later sessions of the experiment they evidence no recovery. They remain suppressed in the presence of CSs and contextual stimuli alike. These results indicate that, in a situation where electric shocks occur, the existence of CSs signaling those shocks is a necessary condition for recovery of appetitively reinforced baseline performance in the presence of contextual stimuli alone. When CSs predict shocks, contextual stimuli (in the absence of CSs) come to serve as cues for performance of the baseline response.The second line of research which has produced results suggesting the above hypothesis is that of Hershiser and Trapold (1971). Those investigators examined the preferences of rats given a choice between environments where appetitive reinforcements were either signaled or unsignaled. Since reinforcement consisted of the injection of a sucrose solution directly into the rat's mouth by means of a surgically implanted tube, reinforcers could be delivered completely unsignaled. In those experiments rats came to prefer the environment in which reinforcements were unsignaled.

Extending conditioned stimuli before versus after unconditioned stimuli: Implications for real-time models of conditioning.

“…We therefore decided to use the intermittent noise CS of Experiments 1 and 2 as our contextual cue and to use a flashing light CS paired with the shock US of those experiments as our embedded CS-US pairing (cf. Brodigan & Trapold, 1974).…”