A Diagnostic Framework

The variety of people which we encounter in our psychiatric and psychoanalytic work seems to demand some sort of classification. As I noted in a previous paper (Bursten, 1973b), 'some ordering is necessary for us to cope with, or even to retain, the vast variety of personality data, and classification enables us to relate one set of observations to another, and to apply to a new situation the knowledge gained from a former one'.

While we generally take the act of classifying for granted, it may be helpful for us to consider briefly some of the issues which are involved when we separate data into categories. It is commonly assumed that this act proceeds from observation to abstraction. Stein (1969), for example, in his discussion of character theory, has stated that character is a 'thing or set of things which can be determined through observation of the subject, with a minimum number of inferential steps between sensory perception and conclusion, the latter including some type of systematic classification'. Now, one can classify on any level of abstraction or distance from observables. Like Klein (1969), I prefer a 'clinical theory'—one reasonably close to the observables (Bursten, 1977). Notice my use of the word 'prefer', for it points out a kind of rule of procedure which I have laid down even before observing the data. It is, if you will, a policy which I follow. In a similar vein, Stein has pointed out some less conscious and less controllable 'policies'—resistances in the psychoanalytic sense of the term—which influence the procedure of classification and theory building.

What I wish to point out is that observation is not the first step in classification. The 'first step', the one we are so often unaware of, is the formation of or the adherence to subtle rules of procedure that govern which data we observe and how we shall treat the data we gather. For example, the psychoanalyst treats 'introspection' as data; many behaviourists do not. This 'first