Abstract
Demonstrated that the interpretation of projective test data is semantic, not probabilistic. The clinician does not employ the model of statistical inference in evaluating the meaning of test responses, although he may employ probabilistic rules as guides to interpretation. Procedural rules for interpreting meaning, the nature of clinical diagnosis from psychological tests, and the meaning of prediction as a clinical activity are discussed. It was concluded that clinicians do not make inferences, in the mathematical sense of this term.