[Undated] [9]. Cogitations

The claim of psychoanalysis to be a scientific procedure has often been challenged by scientists of other disciplines. Stirred by a desire to repudiate what seems to be a denial of the analyst's aspiration to be honest and to state his findings truthfully, psychoanalysts tend either to assume that the criterion is based on a misunderstanding of their work, or else that there is such a thing as scientific method employed by physicists, chemists and others, but that it cannot be used in any valuable way in their own work. The explanation of this attitude probably lies in the nature of psychoanalytic training, which has to cover such an enormous area and is so closely associated with the cure of patients that the would-be analyst can hardly find time to concentrate on the essentials of psychoanalytic theory, let alone investigate metalogic and metatheory. It is likely that if he does so, he will rapidly conclude that they have no relevance to his work and so dismiss both as insignificant.

Yet he would be wrong, and it is one of my aims to show that metatheory and metalogic are of great importance in all scientific work, and of peculiar significance for the psychoanalyst – nowhere more than in the psychoanalysis of severely disturbed patients of the kind known to psychiatry as psychotic or borderline psychotic.

I shall approach the subject as if my aim were to see in what way the criticism that psychoanalysis is unscientific is justified. To do this it is necessary first to try to understand what constitutes scientific method and by what criteria it is possible to determine whether a given method is scientific or not. To this end I propose to consider the views of physicists and philosophers. I choose these two disciplines because it would generally be conceded today that it is in the field of physics that scientific method has achieved its greatest successes, and in the field of philosophy – particularly through the labours of the philosophers of science – that the most