23 Treatment of Anxiety Disorders: Implications for Psychopathology Edna B. Foa Michael J. Kozak Temple University Phobias, characterized by intense anxiety to circumscribed stimuli and a strong tendency to their avoidance, have been conceptualized by learning theorists as involving both classical and operant conditioning processes ( Mowrer, 1960). If phobias are thought to involve conditioned responses, then therapeutic procedures should resemble extinction paradigms. Behavior therapists have, indeed, tried to emulate experimental paradigms in developing clinical procedures. Despite the shortcomings of the two-factor theory of fear acquisition (e.g., Bolles, 1970; Herrnstein, 1969; Rachman, 1976), behavior therapy has been so successful with neurotic fears that it was termed by Marks ( 1978) "the behavioral revolution." Procedures such as desensitization and flooding have been developed to dis- sociate fear responses from the stimuli that evoke them by exposing the patient to these stimuli under "therapeutic" conditions. Social skills enhancement and assertion training were developed for individuals who suffer from social fears or inadequacies. Although they focus on competency in environmental inter- actions, these techniques also invariably include exposure to feared situations. Cognitive therapies directed at reducing fear by modifying thought habits usually encourage patients to expose themselves to feared situations using the new cog- nitive skills acquired in therapy. The last two decades have brought remarkable development in the treatment of anxiety disorders. Psychopathological conditions that were formerly untreat- able (e.g., obsessive-compulsive disorder, agoraphobia) have become amenable to therapy as more effective techniques have replaced their less effective pred- ecessors. Unfortunately, the advances in treating anxiety have been unparalleled by corresponding advances in understanding the mechanisms of treatment. This is perhaps because their theoretical exploration has been pursued less vigorously than the research on outcome variables. Nor have we advanced much in our -421- |