Over the past 25 years, cognitive–behavioural therapy (CBT) and cognitive–analytic therapy (CAT) have discretely jostled for position in the competition for scientific respectability and, perhaps more important, funding. In this sibling rivalry, the slightly younger brother (CAT) has, so far, been less effective, at least in securing funding. There are concerns about its evidence base and this may underlie the reluctance of clinicians and managers to expand its availability. Ryle (2000) has described how he has recently unsuccessfully applied for research and development funding for a large-scale 24-session randomised controlled trial (RCT) of CAT with a group of patients with borderline personality disorder. There are numerous small-scale studies of CAT where successful outcomes have been achieved, but this seems to be the first attempt to substantially evaluate it. Is it unfair to ask why this has not been done before? Psychodynamic psychotherapy has eschewed such forms of scientific evaluation in the past, although some practitioners are now accepting the need for them. CAT practitioners certainly seem to accept that need. As Denman (2001, this issue) states, one of the two main theoretical structures forming the basis of CAT deals with “aim-directed action” including “evaluation of consequences and […] remedial procedural revision”, subjecting the therapy itself to the same procedures seems to follow. Research funding is an issue, but small pilots can be run using individuals' research or personal time. These make the case for definitive research (e.g. Turkington & Kingdon, 2000) and there are now many published controlled studies of CBT.

How similar is the practice of CAT and CBT and how do they differ? Both are used as mediumterm therapies and are rarely long term. For both modalities, any long-term treatment is mainly for support, with relatively infrequent sessions offered. CBT is being developed increasingly for very short-term use, e.g. six-session interventions. Both CBT and CAT are structured interventions, although CAT is arguably less so. Its practitioners do not use treatment manuals, although valuable texts are available, and fidelity questionnaires (e.g. the CBT scale of Young and Beck) are not widely used as in CBT training and research. There are similarities in form and content of therapy: both emphasise homework, diary-keeping and agenda-setting. Collaborative development of formulations is central to both, based on understanding predisposing, precipitating and perpetuating factors. The regular use of letters early in and at the end of therapy is specified in CAT and has much to commend it.

Cognitive–behavioural therapy has demonstrated effectiveness in anxiety, depression, bulimia, chronic fatigue and psychosis. CAT has identified complex and very important problem areas where it could potentially have an impact. Personality disorder is certainly such an area, especially the borderline group as currently identified by Ryle. Could there also be a place for it in work with dependent, obsessional and even dyssocial types? On the basis of one small RCT (Reference LinehanLinehan, 1993), CBT practitioners claim some success in borderline personality disorder using dialectic behaviour therapy; success is also claimed using schema-based therapies (Young, 1980; Reference Beck, Freeman and EmeryBeck et al, 1990), although no RCT evidence currently exists. CAT has much in common with schema-focused CBT, and both types of therapy need to be evaluated for their effectiveness. CBT should work with formulating current feelings, thoughts and behaviour in terms of past and present experiences. The methods used in CAT may also be effective ways of doing this, but comparison is vital to finding out who benefits from the differing approaches and what the essential ingredients are.

Both CAT and CBT use unhelpful jargon. Are terms such as ‘procedural sequence model’, ‘dilemma’ (described by Denman as the “presentation of false choices or of unduly narrowed options”), ‘snags’ (subtle negative aspect of goals) and ‘placation trap’ clarifying or confusing? The first sounds remarkably like problem-solving. Redefining a commonly used term such as dilemma (defined in the Concise Oxford Dictionary of Current English as “an argument forcing an opponent to choose two alternatives both unfavourable to him” (Reference Watson, Fowler and SykesWatson et al, 1976)) may not be helpful. And is not falling into a placation trap simply being unassertive? CBT also has its confusing examples, such as ‘arbitrary inference’ and ‘selective abstraction’ – more simply, getting things out of proportion and getting them out of context. ‘Schema’ has also been appropriated and narrowed in meaning compared to the way it is used generally in psychology.

Is CAT a way of developing psychoanalytical concepts for shorter-term therapy? Can it be used as an effective treatment for complex problems? It promises much but the evidence is currently lacking. CAT certainly has advantages over CBT practised, inappropriately, in a rigidly technical manner neglecting attention to emotions and relationships. But when CBT is practised in the holistic manner developed by its founder, Aaron Beck (Reference Beck, Rush and ShawBeck et al, 1979) and subsequently developed by others, as a cognitive behaviour therapist, I have to ask what added benefit can CAT offer my patients?