The present study provided experimental support for Mowrer's conditioning theory of neurosis. Naive hooded rats were given fear-conditioning (CS-shock) trials in one side of a two-compartment apparatus. Then, in the absence of shock, they were allowed to escape fear by jumping a hurdle to a safe compartment. The hurdle-jumping response was considered to be analogous to anxiety-based symptomatic behavior indicative of psychopathology in that the performance of the response was unrelated to the receipt of further shock. All subjects learned the hurdle-jumping response, and then performance gradually declined until an extinction criterion was reached. Thus, a response instrumental in reducing fear was learned and was maintained over many trials. A subsequent single fear-conditioning trial led immediately to a high level of performance followed by gradual extinction. These findings imply that, when fear is extinguished, instrumental responding can cease even though some response strength remains; they suggest the importance for therapy of focusing on the extinction of both instrumental (symptomatic) behavior and fear. Evidence was also provided for the spontaneous recovery of fear. Some human behavior is paradoxical in the sense that it persists even though, from the viewpoint of an observer, the net effect of the behavior appears to be maladaptive because it does not protect the individual from any objective danger. However, the persistence of such behavior suggests that it has some functional significance for the individual, and its explanation is clearly demanded of any theory that purports to have general relevance. One of the historical attempts at such an explanation is based primarily upon research on learning conducted in the laboratory with infrahuman subjects. Eysenck (1979) calls it the "conditioning theory of neurosis," and its basic tenets can be traced from Mowrer (1939) and Miller (1948,1951, through Wynne (1954) andBrown (1961), to Stampfl and Levis (1967) and Levis and Boyd (1979).According to this theory, when a noxious or aversive event occurs, the neutral stimuli accompanying it acquire the capacity to elicit an emotional response, typically called fear or anxiety. These events constitute a classical conditioning procedure. The neutral stimuli, both discrete and situational stimuli, constitute the conditioned stimulus complex (CS); the noxious This research was supported in part by Grants MH-29232 and MH-36610 from the National Institute of Mental Health. Portions of these data were reported at the 1980 meeting of the Midwestern Psychological Association.The authors are indebted to Edward J. Callen for aid in the analysis of the data and for constructive comments on an earlier draft of this article.

Experiment 1 demonstrated that hurdle-jumping, escape-from-fear (EFF) performance was affected deleteriously when the handling of the rats was by the tail rather than the body and was facilitated by the administration, just prior to EFF training, of one additional CS-shock pairing along with the opportunity to jump the hurdle. Additional experiments elucidated the basis for the effect of these two variables. Experiment 2 indicated that tail handling degraded performance by decreasing reinforcement. In Experiment 3, the facilitatory effect of the extra CS-shock treatment was found to result from the occurrence of a response at the termination of the pairing and not from the CS-shock pairing per se.In a typical escape-from-fear (EFF) task, classical fear-conditioning trials (CS-shock) are administered in one side of a two-compartment apparatus. Later, in the absence of shock, subjects are allowed to jump a hurdle to an adjacent safe compartment and, thereby, to escape the fear-eliciting CS and situational cues in the shock compartment. Theoretically, the learning of the instrumental hurdle-jumping response is motivated by these two sources of fear and is reinforced by the reduction of this fear following the response.Recently, Crawford, Masterson, and Wilson (1977) have reported difficulty in consistently obtaining such instrumental learning based either on fear of the CS alone or on fear of both the CS and situational cues. In order to obtain learning, they found it necessary to administer not only the classicalfear-conditioning trials but also an additional such trial, along with the opportunity to jump the hurdle, just prior to regular hurdle-jumping training. The necessity for the inclusion of such an additional step in the normal procedures is surprising because there is a relatively large literature reporting successful escape-from-fear learning without such a requirement. These studies have been conducted in a number of laboratories with different experimenters, with different strains and sex of rats, and with differences in a number of specific details of apparatus and procedure (e.g., Brown & Jacobs,

No abstract