Contingency in fear conditioning: A reexamination

Abstract

The effects on classical fear conditioning of the rate of presentation of the unconditioned stimulus (US) and the contingency between the conditioned stimulus (CS) and the US were examined using the conditioned emotional response procedure with rats. Increases in US rate reduced suppression by the same amount whether the added USs were signaled by CS, thereby maintaining the CS-US contingency, or unsignaled, thereby weakening the CS-US contingency. Failure to control for the rate of US presentation in previous studies of the effect of CS-US contingency on fear conditioning has led to the unsubstantiated conclusion that CS-US contingency is fundamental to classical conditioning.It is now widely accepted that classical conditioning depends not only on the temporal contiguity of conditioned stimulus (CS) and unconditioned stimulus (US), but also on the CS-US correlation or contingency. With the contiguity, or joint occurrence, of CS and US held constant, CS-US contingency is weakened either by presentations of CS alone or by presentations of US alone. A commonly used index of contingency is given by the difference between conditional probabilities: the probability of US given CS, or P(US/CS), and the probability of US in the absence of CS, P(US/no CS). The contingency is positive when P(US/CS) is greater than P(US/no CS), negative when the reverse is true, and zero when the probabilities are equal. Mackintosh (1974) expressed the prevailing view of the importance of contingency for conditioning when he wrote: "conditioning experiments can be operationally defined as arrangements of correlations or contingencies between events ... . The most natural interpretation ... is that animals detect these contingencies: exposed to these particular relationships they learn to associate these correlated events" (p. 244). The results of the present experiment question this conclusion.The idea that contingency is fundamental to conditioning owes much to the work of Rescorla. In an influential paper on the proper control for classical conditioning, Rescorla (1967) argued that only those behavioral changes that depend on CS-US contingency should be regarded as conditioned. Accordingly, a noncontingent CS-US sequence is the proper control for classical conditioning, even though that sequence will generally include some contiguous presentations of (1968) reported an experiment on fear conditioning in which he found a close relation between the degree of CS-US contingency and the strength of conditioning to the CS, as measured by response suppression in the conditioned emotional response (CER) procedure. Subsequently, a theory of classical conditioning presented by Rescorla and Wagner (1972) provided an explanation for the dependence of conditioning on contingency. Although the theory did not assume that contingency was necessary for conditioning, it nevertheless showed how a trial-by-trial conditioning process could produce the close relation between CS-US contingency and the asymptotic strength of conditioning that was found in Re...

Autocontingencies: Factors underlying control of operant baselines by compound tone/shock/no-shock contingencies

“…This view, termed the relative waiting time (RWT) hypothesis by Jenkins, has recently been applied to the conditioned suppression procedure. In this study, Jenkins and Shattuck (1981) concluded that better conditioned suppression occurs with lower RWT ratios, independently of CS-US contingency.…”

Effect of signaling extra unconditioned stimuli on autoshaping

“…Because the average waiting time in the context and that in the target CS are not altered by the signaling of the extra USs, SET predicts that the effect of extra USs is no different whether signaled or unsignaled by a second CS. Several experiments have produced results supporting this conception (Balsam, 1982;Jenkins, 1984;Jenkins et al, 1981;Jenkins & Lambos, 1983;Jenkins & Shattuck, 1981;LoLordo & Randich, 1981). Jenkins et al (1981, Experiment 13) did not obtain significantly less acquisition to a target CS when extra USs were signaled as compared to when they were unsignaled, but such an effect was obtained by Durlach (1983), who used more prolonged target CS conditioning.…”

Escape From Serial Stimuli Leading to Food

“…In Pavlovian conditioning, to be sure, it is well known that the effectiveness of the conditional stimulus (CS) is a function of the length of time by which it has characteristically preceded the unconditional stimulus (US), but Balsam, 1984;Brown, Hemmes, Coleman, Hassin, & Goldhammer, 1982;Gibbon & Balsam, 1981;Jenkins, 1984;Jenkins, Barnes, & Barrera, 1981;Jenkins & Shattuck, 1981. ) As mentioned earlier, under standard Pavlovian procedures stimuli that are negatively correlated or that precede the US by longer than average times are found to be inhibitory (LoLordo & Fairless, 1985;Rescorla, 1969b).…”