The study of many sorts of psychological phenomena has been hindered by Behaviorism, the identification of psychological states with the behavior which results from these states. This is particularly true in the case of our topic today, pain, where we have an essentially inner, subjective phenomenon. The identification of pain with publicly observable pain behavior encourages lumping together importantly different sorts of behavior, for example, avoidance behavior, expressions of pain, verbal descriptions of various sorts, expressions of negative affect, and the like. These are all treated as behavior of one unified sort, "pain behavior," and therefore appropriate for the operational definition of pain. This behavioristic approach has led to ignoring important distinctions within the domain of pain phenomena and has prevented advances in our understanding of such phenomena. We are now beginning to investigate, without embarrassment, the more subtle features within the inner phenomena, even where we lack precise behavioral criteria which provide publicly observable evidence for the presence or absence of these inner features and the degree of which they obtain, although we still would not consider the job done until we had developed such criteria or at least come to understand why we could not.