The epistemological problems of the clinical research in psychoanalysis after Grünbaum: some directions and resolutions

Abstract

Since the advent of the clinical-psychoanalytic method, philosophers and scientists have identified problems in its epistemological pretensions or at least in the degree of disposition of psychoanalytic communities to try to overcome its epistemological shortcomings. The aim of this doctoral dissertation is to investigate to which degree and by which rationale the data ensued by the clinical-psychoanalytic method would be able to cogently support hypotheses about the behaviour of individuals and classes of individuals as well as about some operations of the human mind in general. It is thus an elaboration on both the actual and the potential logic of both the generation and the justification of clinical-psychoanalytic hypotheses. For its enduring influence, I choose Adolf Grünbaum’s critique on the foundations of that method to frame such an elaboration. With his Millian orientation, Grünbaum concludes that Freud and his followers are unable to defend the probative character of their clinical method. He argues that Freud’s attempt to link therapeutic outcome with truthfulness, his “Tally Argument”, is unsound; that the clinical data from which clinical-psychoanalytic inferences are made are probably contaminated by the analysts’ suggestions; and that, in any case, many of such inferences instantiate causal fallacies and biases. I take these three classes of arguments in Grünbaum as three classes of epistemological problems to be redirected and solved rather than just evaluated in their fairness with the psychoanalytic corpus and practice. I counterargue that clinical-psychoanalytic inferences are or can be essentially Millian as long as they are or can be almost identical to the Explanationist-Bayesian inferences in the historical sciences. I discuss the concept of suggestion and the theories about the conditions to its occurrence, thus deflating Grünbaum’s contamination argument but also revealing that analysts should be able to include heterosuggestions in their database; I formalize a technical device to help them in this task. Faced with the problem of how truthful enunciations during treatment are related to therapeutic effects in the patient, I argue that the former can be considered INUS causes of the latter, which implies that therapeutic effects cannot be considered unequivocal evidence, but only complementary evidence, for antecedent truthfulness; that, therefore, Grünbaum is right, but shallow, about Freud’s Tally Argument. I conclude, in contrast with Grünbaum, that research from clinical-psychoanalytic data is feasible, as the method involved in this research is already cogent in many respects and can become more and more cogent through the stabilization or implementation of certain principles and guidelines; and that, moreover, it is an irreplaceable and extremely relevant kind of research. I also conclude, however, that a question of how consistent the results of this method can be across its applications is still barely raised by analysts and by those committed to advancing this kind of research. To make clinicalpsychoanalytic research widely recognized as scientific, a combination of philosophical, technological, and institutional efforts is needed