In the steady flow of articles defending the scientific status of psychoanalysis it is often asserted that clinical psychoanalyses constitute a series of “natural experiments” under conditions of controlled observation. Because psychoanalysis is a therapeutic as well as an investigative procedure, the distinction between discovery and verification tends to become lost or blurred. Clinical case reports often slide back and forth from observation to inference or slip from assumption to fact without hesitation or warning. Academically oriented investigators readily acknowledge that the psychoanalytic situation provides a matchless opportunity for generating insights, but stress its limitations as a setting for rigorously testing psychoanalytic hypotheses. Therefore they turn to the laboratory. Most often, however, the concepts studied have eluded translations precise enough to serve the requirements of fine-grained investigation in the laboratory without sacrificing the meanings involved in the broad-gauged use of theoretical concepts