Review of “Exploring Theory Integration in Psychoanalysis”: Theory and Testing Need One Another

We have come a long way since the 1954 Conference on Psychoanalysis and the Philosophy of Science, and the follow-up 1958 New York University Institute of Philosophy Symposium in which philosophers, other scientists, and analysts examined theory-building and -testing in psychoanalysis. It was widely recognized at the time that psychoanalytic theory and hypotheses present great difficulties in regards to study and confirmation (Carnap, 1955; Ellis, 1955; Nagle, 1959). Such impediments include vaguely or metaphorically described constructs with a lack of clear relationships to observable phenomena, and the problem of devising explicit tests to confirm or disconfirm such relationships (Nagle, 1959). These concerns have been reconsidered and refined by subsequent authors (Leahey, 1980). One early proponent of psychoanalytic research noted that the study of lower-order hypothetical constructs would be possible as long as there were clear “correspondence rules” from hypothetical construct to observable data (Ellis, 1955) overlapping with what Bridgman (1959) dubbed “operational definitions.” Aside from concerns about the appropriateness of applying operationalism, common in the physical sciences, to the psychological sciences (Feest, 2005), many agree that psychoanalysis needs more research in the form of testable hypotheses and controlled observations (Kernberg, 2006; ).

Yet the conclusions of these two conferences were not pessimistic; rather, more instructive, if a bit scolding. Generally it was agreed that clear definitions were lacking at almost every level; there was a reliance on theory, theoretical entities, and processes that were unsupported; inference