We are grateful for the opportunity to respond to Dr Calton. He challenges claims for objectivity in the diagnosis of any disorder that has an interpersonal component, taking as his example some research into the diagnosis of personality disorder. He queries why we do not consider the role of the therapist's personality in our paper, ‘Objectivity in psychoanalytic assessment of couple relationships' (Reference Lanman, Grier and EvansLanman et al, 2003). In that paper, where we show evidence of a good degree of objectivity (based on interrater reliability) for the diagnoses we discuss, we specifically refer to the fact that those making the judgements need to have had a psychoanalytically based training in order to develop their ability to make use of their emotional reactions to the patient.

Our paper deals with psychotherapeutic diagnoses, rather than with psychiatric ones, but on the basis of our work we would like to comment on Dr Calton's position. First, there are likely to be very significant differences in what is judged to be a helpful ‘fit’ between therapist and patient, between the two different domains of general psychiatry and psychoanalytic psychotherapy. In the former, a friendly and sympathetic stance may be the crucial therapeutic vehicle for providing medication and other treatments. But in the domain of psychotherapy it is not necessarily a good thing to ‘match’ therapist to patient, if by this one means attempting to avoid prejudices or sensitive areas, because this is likely to lead to a serious evasion of the darker areas of interaction, conscious and unconscious, where the significant problems will tend to lie. If the study of the interaction ‘in the room’, between therapist and patient, is itself the treatment, then the therapist's best equipment for this is self-knowledge, including knowledge of the darker areas of his or her own personality and knowledge of how to recognise and use the ways in which these affect him or her.

While personal psychotherapy together with detailed supervision by no means guarantee the development of such knowledge - and there will be practitioners who are unable to respond, as well as therapies which do not go far enough - these remain the best available means of acquiring the skills necessary to work with unconscious processes, enabling a therapist to understand a patient's personality difficulties of and the way these interact with their own. Outside this particular field, it may not be widely recognised that one of the principal tools of contemporary psychoanalytic therapy is the constant monitoring by practitioners of their own emotional responses to patients, not simply in order to suppress or redirect them, but in order to gain information that the therapist will then be able to employ in clinical diagnosis and engagement with patients. This is not to be confused with the self-disclosure advocated by some therapies. In our view there is no substitute for a rigorous psychotherapeutic training in this area, which includes selection of trainees, personal psychotherapy and detailed supervision.