In response to:
Putting Freud to the Test from the January 31, 1985 issue
To the Editors:
I agree with much of Jonathan Lieberson’s very interesting review [NYR, January 31] of Adolf Grünbaum’s book on psychoanalysis, but I would like to stress two points, even if neither contradicts any statement in the review.
First, although Lieberson praises the book, I think he underestimates Grünbaum’s achievement. In the article of mine that he quotes in his review, I examine the experimental evidence and conclude that it fails to provide strong support for any part of Freudian theory. (I also argue that there is no acceptable evidence that Freudian therapy is more effective than a credible placebo in producing any sort of therapeutic benefit.) If I am right, and in the absence of firm evidence from epidemiological or anthropological studies, Freudians must rely primarily on clinical evidence. What Grünbaum has done is to provide the most detailed and powerful arguments developed so far that the clinical evidence is too defective to provide adequate support for Freudian theory. Other critics, and even some Freudian sympathizers, have reached the same conclusion, but without providing an argument rigorous enough to establish that conclusion.
If Grünbaum is right, he has undermined the clinical evidence for other Freudian causal hypotheses beside those he discusses and for newer psychoanalytic theories that are based on the same type of evidence. This achievement is very important, given the influence that Freud’s ideas continue to exert in clinical psychology, psychohistory, literary criticism and the culture at large.
My second point concerns the ultimate fate of psychoanalytic theory and therapy. Dr. Lieberson is quite right to stress that failures of confirmation, so long as they can be explained in terms of defects in experimental designs, do not constitute strong evidence against Freudian hypotheses. It is possible that some hypotheses that are distinctively Freudian, and not merely suggested in some vague way by Freud’s ideas, will be confirmed by future experimental work. Why should a Freudian think, however, that this is likely? It is not encouraging to be told, as B.A. Farrell tells us in his book on psychoanalysis, that the theories of Newton and Harvey also remained unconfirmed for a long time. So many unconfirmed scientific theories have turned out to be wrong, that an inductive argument based on the history of science would, despite the exceptional successes, tell against rather than for Freudian theory. My point, then, is this. To anyone who believes that a non-trivial Freudian causal hypothesis is true, the work of Grünbaum and others raises the following challenge: What is the hypothesis and what are the grounds for thinking it to be true?
University of Miami
Coral Gables, Florida
Jonathan Lieberson replies:
Dr. Erwin claims that Grünbaum has provided “the most detailed and powerful arguments developed so far that the clinical evidence is too defective to provide adequate support for Freudian theory.” This achievement is important, according to Erwin, because it complements his own claim that the most significant nonclinical evidence—experimental evidence—also fails to provide such support for Freudian theory. If he and Grünbaum are both correct, the question arises how Freudians who regard their hypotheses as open to empirical test are to support their views.
The challenge strengthened by such work as Erwin’s and Grünbaum’s is indeed an important one. However, as I remarked in my review, it is limited in certain significant respects. For one thing, many of Freud’s hypotheses are at present insufficiently refined to be open to direct experimental test. Many tests whose results have been claimed to provide “no evidence” for these hypotheses did not succeed in testing them, but only other hypotheses suggested by them or otherwise indirectly related to them. Furthermore, neither Erwin nor Grünbaum (as they both acknowledge) has shown that inquirers will not develop experimental tests in the future that will support Freudian theory. With respect to “clinical evidence,” finally, psychoanalysts may discover ways of “purifying” clinical observations to escape Grünbaum’s strictures; and whether or not they will be successful, a number of contemporary psychoanalysts are trying to find ways of ensuring that clinical data (such as a patient’s acceptance of his therapist’s interpretation of a dream) are free from such “contaminations” as suggestion.
To turn to Erwin’s second point: my own position on the future of psychoanalytic theory or therapy was that we should exercise a “benevolent skepticism” toward it. I took this attitude to contrast with what I regarded as Grünbaum’s insufficiently imaginative approach, which concentrates excessively on Freud’s own view of how his hypotheses could be tested; I urged that “Grünbaum might profitably have examined in greater detail…the full range of efforts that have been made to test the main hypotheses of psychoanalysis outside the clinic, and to specify what obstacles must be surmounted by similar tests in the future if they are to yield strong confirmations or disconfirmations of that theory.” I called this approach “prudent” because psychoanalytic theory—for all the enormities that have been committed in its name—does not (at any rate to me) seem to be simply a bundle of self-evident falsehoods or juvenile speculations, but to suggest hypotheses and guides to inquiry that may indeed help us in discovering truth, even if its hypotheses as currently stated are open to objection.
I am aware that such a claim must seem elusive to detractors of psychoanalysis—as indeed must any claim about the “promise” of a theory to those who have no sympathy for it. But I did not in any case claim that Freudian theory will be confirmed in the future. I described in detail the possibility that psychoanalysts will never succeed in formulating their ideas clearly, so that they can be tested; once tested, I added, these ideas might well be disconfirmed so decisively that scientific researchers will cease to pay attention to psychoanalysis altogether. I referred to Farrell’s point that several notable scientific theories took a long time—in one case, two centuries—before they could be adequately tested in order to support my view that the current lack of confirmation of Freudian theory is not sufficient grounds for deeming it unworthy of critical attention.
Erwin’s claim that “so many unconfirmed theories have turned out to be wrong, that an inductive argument based on the history of science would, despite the exceptional successes, tell against rather than for Freudian theory” does not seem to me convincing. Since most “confirmed” theories have later turned out to be wrong, an “inductive argument” of this kind would seem to tell against any theory at all. My central point was to encourage the elucidation and development of Freudian theory so that it might be tested. Otherwise we should never be in a position to tell whether this theory, if it can be refined and developed, is mistaken (as some seem entirely too willing to believe already) or whether (as others suppose) it will have what Einstein called, in a remark made in an entirely different context, “the fairest fate” of any theory, “that it should point the way to a more comprehensive theory in which it lives on, as a limiting case.”