Clinical evaluation of mental patients is largely based on their verbalizations. Although some patients, while very sick, may develop a pseudoinsight or speak pseudorationally, we do not ordinarily see the converse of this: patients who are clinically well but cannot express themselves reasonably. Accordingly, if we were to set up a scheme for evaluating patients' statements with regard to increasing connectedness or relevance of ideas, we would expect a good, though not a perfect, correlation between improvement on our scale and actual clinical improvement.
When we try to make prognostic statements based on such a scale, however, we may run into difficulties, for we may find that patients who enter a hospital making no verbal sense at all go on to a speedy remission, while thoroughly rational-sounding patients may never budge from their admission-level plateau, or may actually seem to get worse. This is