Clinical Interpretation: Some Comments on the Nature of the Evidence

As we listen to the stream of associations from an analytic patient, we hear an almost continuous flow of linked ideas. They seem to resolve themselves into a number of strands—sometimes few, sometimes many. But ask a friendly critic to listen to the same material: he hears a mixture of unrelated words and has the impression that the patient simply skips from one topic to the next. Who is right?

To put the problem in these terms helps to show how quickly we lose sight of the fact that the links between ideas in clinical material may be imposed from outside and are not always inherent in its structure. Since we find it so easy to see connections between disparate ideas, we assume that the connections are contained in the structure, which is to say that other judges would hear the material in much the same way. But would they? Would 12 friendly critics independently arrive at the same underlying structure, or would each hear his own favorite melody?

A suspicion that the latter is often the case comes from a recent study by Hartvig Dahl (personal communication) on hypothesis formation. A group of analysts listened to the early hours of a case and interrupted the reading whenever any one of them decided that enough evidence had accumulated to support one among a set of possible hypotheses. Each of these interruptions was defined as a “call,” and many single calls were heard—in fact, it was rare for the group to decide in unison that such and such a hypothesis could be supported at such and such a point. Dahl found that a judge's confidence was significantly higher on his own calls than on those