“I'd warned her that intellectuals didn't count for much in the scheme of things: but academics are like the clergy in their blind conviction that their vocation has relevance to the rest of the world.” (McCauley, 1990).

In the beginning, the founders of behaviour therapy were quite explicit about the link between basic laboratory research and their new therapies (Eysenck, 1960; Wolpe, 1958). Their claim that behaviour therapy was based on “modern learning theory” was not, however, without controversy (e.g., Breger & McGaugh, 1965; London, 1972); but while the logical necessity for appeal to laboratory research on learning might have been questioned (Breger & McGaugh, 1965), the historical contingencies linking Pavlov's dogs, Masserman's cats, and Wolpe's first systematic desensitisation patients can scarcely be questioned (American Psychological Association, 1980; Kazdin, 1978).

It is equally clear that there was a similar close historical link between laboratory studies of operant behaviour, especially by Skinner and his students, and the development of behaviour modification applications such as reinforcement therapy and the token economy (Kazdin, 1978; Krasner 1971; Skinner, 1953). Again, the continuing necessity for behaviour modification to be linked to basic research in operant behaviour has been challenged (Azrin, 1977; cf. Deitz, 1978) but the historic fact of the relationship is well established (Krasner, 1971).

When behaviour therapy was developed in the 1950s and 1960s, it was based firmly on the experimental psychology of the time, principally learning theory. Since that time effective verbal therapies have been developed, and clinicians are faced with the difficult task of selecting among so-called behavioural and cognitive interventions based on radically different philosophies. This paper reviews developments in learning theory, particularly human learning, which suggest that conditioning is a complex cognitive process giving rise to verbally accessible expectancies and beliefs. Such a perspective provides a common theoretical framework (cf. Bandura, 1977) that allows behavioural and cognitive strategies to be coordinated to modify a particular cognitive target, such as biased expectancy of harm in anxiety. It is argued that the terms behavioural and cognitive carry unwanted connotations, and that it is preferable to specify independently the content of information to be conveyed in therapy (e.g., reduction in expectancy), and the mode by which it is conveyed — experience, observation, or language. Learning theory, in conjunction with other areas of experimental psychology, continues to provide a conceptual basis for the understanding of psychopathology and the development of new psychological interventions.

This paper describes a two-component conditioning model of anxiety disorders whose two main components are an associative process and a process which determines the evaluation of the unconditioned stimulus (UCS). It is argued that the inclusion of UCS revaluation processes enables conditioning models to explain a variety of anxiety-related phenomena that hitherto had been difficult to explain solely in associative terms. In particular, (a) experiments are described which suggest that incubation of the conditioned response (CR) can be explained in terms of processes contributing to UCS inflation, (b) the experimental evidence on UCS revaluation is assessed in relation to cognitive processes that might lead to trauma revaluation within specific anxiety disorders, and (c) the role of personality factors in effecting UCS and trauma revaluation is discussed.

This paper falls into three distinct sections. In the first, the attempt is made to discuss those aspects of emotion theory, particularly with respect to the links between emotion and cognition, that have been considered by cognitive-behaviour therapists. In the second, the aim is to describe recent emotion theory which has, in part, been directed towards cognitive-behaviour therapists and others who are concerned with emotional dysfunction, but which has not yet been taken up. In the final section, the goal is more speculative, arrived at through a description of some of the less obvious excursions made by emotion theorists. Although they mostly do not have emotional dysfunction in mind when addressing their theories, their views are, in the belief of the present author, particularly pertinent to the future directions of cognitive-behaviour therapy.

The purpose of this article is to emphasise the importance of basic research in the advancement of theory and clinical technique. To illustrate the power of such a strategy the theory and imagery technique of Implosive (Flooding) Therapy will be briefly described along with a recent modification that has resulted in the development of a new technique which is capable of reactivating all past traumatic memories responsible for the development and maintenance of psychopathology. This new approach has resulted in a number of important discoveries that challenge many currently held beliefs.

Behavioural momentum refers to the tendency for reinforced behaviour to persist when conditions are altered. Research on persistence with pigeons as subjects has suggested that response rate under steady-state conditions and its persistence when conditions are altered are independent aspects of behaviour, with response rate determined by response-reinforcer contingencies and persistence determined by stimulus-reinforcer contingencies. These results have been replicated with humans in a natural setting, and may therefore be relevant to clinical practice. The following prescriptions are suggested by the research results: (a) To eliminate undesirable behaviour, it is not sufficient to reinforce alternative behaviour in the same setting because this may actually enhance persistence of the undesired response; change to a new setting may also be necessary; (b) to establish desirable behaviour, the behaviour should be reinforced frequently, and to make it persistent, both the therapeutic and natural environments should be highly correlated with reinforcement.

This paper critically examines the recent impact of cognitivism upon the field of clinical psychology, and concludes that certain criteria of scientific adequacy have been compromised. The argument is developed that the introduction of information processing constructs to theoretical models of psychopathology has made a potentially valuable contribution to the discipline; but that the acceptance of mental events as dependent measures has severely undermined the scientific credibility of experimental attempts to evaluate such models. It is proposed that future progress will require the adoption of a particular methodological constraint. Specifically, it is suggested that cognitive explanations of psychopathology can only be tested adequately by evaluating the validity of the behavioural predictions that they generate. Using examples of recent research that has investigated the cognitive characteristics of vulnerability to anxiety and depression, an attempt is made to demonstrate that adherence to this proposed constraint not only is possible, but actually provides a far greater degree of understanding than could be attained through the use of alternative methodologies. It is postulated that the future scientific status of clinical psychology may depend upon our collective response to the issues that are raised in this paper.