On Caution and Courage in Psychoanalytic Epistemology

Michael Lacewing’s argument in this paper is impressive. His basic case is that research in social and clinical psychology threatens to undermine Hopkins’ (1988) well-known defense of psychoanalysis. This defense claims that psychoanalysis is an extension of, and as valid as, commonsense psychology. By questioning the reliability of commonsense psychological inferences, research in social and clinical psychology also challenges psychoanalytic validity. For, in extending commonsense psychology, psychoanalysis inherits its flaws. This is a fascinating contribution to arguments about psychoanalytic epistemology.

Lacewing deepens his contribution by attempting to respond to this challenge. He identifies clinical psychoanalysis as one of the contexts in which commonsense psychological inferences are less prone to error. In such contexts, thoughtful observation is not derailed by haste, selective access to information, or plausible but irrelevant distracting contextualizing factors. In other words, while an analyst contemplates a patient’s motives in the analytic context, his commonsense psychological inferences are likely to be accurate. As Lacewing (2012, 209) consoles us, “psychoanalytic methodology … removes the potential for many [cognitive] errors.”

In a particularly valuable part of the paper, however, Lacewing qualifies this statement. Psychoanalytic claims, he argues, are valid in only three forms: (1) As explanations of individual instances of behavior, (2) as generalizations about typically observed behaviors, and (3) as inferences about mechanisms that are related to these behaviors and that may have causal properties. He denies (4) the prima facie validity of psychoanalytic speculations concerning the origin of mental structures, and (5) theoretical extrapolations from these to a model of mind.¹

Lacewing also rightly cautions even in relation to (1)–(3) that a relatively sound epistemology should not be confused with the accuracy of an inference. He warns of biases within the analytic situation that lead both patient and analyst to privilege mental contents over situational factors when attempting to understand what determines the patient’s actions. Clinicians as well as philosophers should heed this caution; the fundamental attribution error remains alive and well in most analytic consulting rooms.

Despite this caution, Lacewing’s claims for the validity of analytic inferences I believe involve some idealization. The clinical encounter and clinical theory-building are perhaps a little messier than he makes out. Because of this, I think Lacewing’s restrictions on psychoanalytic validity should be tightened even further to include only certain explanations of individual behavior arising within the clinical encounter.

I also feel compelled to offer my own word of caution: Lacewing’s energetic defense of clinical psychoanalytic epistemology should not be taken to support those who wish to abandon cross-fertilization between psychoanalysis and neighboring disciplines. Most theoreticians of psychoanalytic practice assert that the patient’s free associations and the analyst’s free-floating attention are the hallmarks of psychoanalytic data gathering. Only facts obtained by these means are considered admissible, and only theories based on these facts are thought to contribute to the development of psychoanalysis. Unfortunately, this implies that psychoanalysis exists in isolation from all other disciplines, which do not use these methods. As Lacewing argues, however, psychoanalytic observations concerning human behavior are extensions of other forms of psychological observation. They are not, as some analysts would have it, incommensurate with any other form of observation. Furthermore, situational biases within the analytic setting need to be investigated and assessed in conjunction with other disciplines.

My concern is that Lacewing’s limited defense of psychoanalytic inference will be used to justify the accumulation of a uniquely psychoanalytic knowledge base. This would be illegitimate: The error-proneness of psychoanalytic inferences precludes them from independently validating psychological theories. Nevertheless, many psychoanalysts believe that the particular features of the analytic encounter—the quality and intensity of affective investment and involvement of both patient and analyst, the emphasis on the unconscious aspects of the therapist’s and patient’s functioning, and the area of unconscious communication thus inhabited by the therapeutic couple alone—make it a unique data source that cannot, despite Lacewing’s aspirations, be made “consilient” with convergent, perhaps experimental, data.² Whether it is on the couch or in the laboratory, however, the brain...