The New York Review of Books

To the Editors:

In “Freud’s Permanent Revolution” [NYR, May 12], Thomas Nagel endorses Richard Wollheim’s and Paul Robinson’s 1993 critiques of my book The Foundations of Psychoanalysis: A Philosophical Critique, ¹ and he offers criticisms of his own. Alas, both Wollheim and Robinson demonstrably base their critiques on fundamental misportrayals of the avowed aims and contents of my Foundations. By relying on their strictures, Nagel builds much of his case on quicksand.

According to Wollheim, for example, my book concentrates entirely on the here-and-now clinical (“on the couch”) testability of Freudian theory.² On the contrary, only 30 pages of my 310-page critique focus on the issue of testability, and I explicitly went beyond present clinical tests to future epidemiologic and experimental ones. Thus, I proposed a specific extraclinical, epidemiologic test of Freud’s homosexual etiology of paranoia. Indeed, Wollheim entirely conceded my case (Foundations, p. 278) when he expressed “the hope…that eventually tests will be devised, presumably of an extra-clinical kind” (p. 109), which may supply the now conspicuously absent support for general Freudian psychological principles such as the crucial etiologies of symptom formation and the theory of psycho-sexual development. But in any event, both Wollheim and Nagel overlook the real core of my 1984 argument, which is that, testing aside, Freud’s presented arguments and actual evidence simply fail to sustain the major pillars of his theory of repression. And in my 1993 book Validation in the Clinical Theory of Psychoanalysis, which Nagel consigns to a footnote, I contended that Freud’s theory of dreams is false at its core.

As another straw man, Wollheim complains (p. 107) that my critique of Freud makes no allowance for the “infra-structure” of the theory, an animadversion endorsed by Nagel. Qua example of such infra-structure, Wollheim recalls Freud’s postulate that repressed homosexuality develops into paranoia via the operation of the defense mechanisms of reaction-formation and projection. But I myself gave a careful account of precisely that mediating dynamics to which I referred as Freud’s postulated “causal micro-structure” (Foundations, p. 76). Neither Freud nor any tests conducted after him furnished evidence pertaining to the mediating micro-structure. However, I offered refuting evidence against his hypothesis that repressed homosexuality is the causal sine qua non of paranoid delusions, a distinctive hypothesis not avowed by rival theories. Such contrary evidence is furnished by the existence of overtly practicing homosexuals who are paranoid, and by the fact that some paranoiacs feel persecuted by people of the opposite gender in the absence of a “primary” persecutor of the same gender, who is required by Freud’s etiology. I presented these refuting findings in publications of 1986 and 1993 cited by Nagel, who nonetheless simply ignored them.

Oddly, Nagel regards as “telling” against me Wollheim’s question “How is common sense psychology tested?” In the case of its ubiquitous causal hypotheses, such as that insults anger or humiliate people, good tidings create joy, or that people tend to put on protective clothing because they feel cold, I reply: To warrant that a factor of sort X (such as being insulted) is causally relevant to a kind of outcome Y (such as being angered or feeling humiliated) in a reference class C, evidence is required that the incidence of Y’s in the subclass of X’s is different from its incidence in the subclass of non-X’s. Nagel speciously pleads the idiocy or impossibility of statistical confirmation of such general hypotheses, irrelevantly lampooning statistical evidence for their individual application to particular instances in either common sense psychology or psychoanalytic theory.

Thus, he overlooks Freud’s enunciation of the general hypothesis that slips are caused by “motives of unpleasure,” which is statistically testable. Instead, Nagel points to Freud’s concrete, “circumstantially rich” explanation that a young man forgot the Latin word “aliquis” in a quotation from Virgil, because the man feared the pregnancy of his mistress. That fear is a particular instance of a motive of unpleasure. Yet Nagel’s unavailing comment is that, in this instance, “statistical confirmation is completely impossible.”

In lieu of such confirmation, he tells us,

we simply have to decide whether this is an intuitively credible extension of a general structure of explanation that we find well supported elsewhere, and whether it is more plausible than the alternatives…

Precisely this epistemological recipe of intuitive credibility fails completely even in deciding between Freud’s own theory and the major post-Freudian alternatives to some of his central tenets offered by Heinz Kohut’s school of “self-psychology” and the British school of “object relations” theorists, which include Melanie Klein, who is championed by Wollheim but was bitterly opposed by Anna Freud. In a powerful essay, Morris Eagle has pointed out that self-psychology has repudiated virtually every one of Freud’s major tenets.³ Thus, Kohut supplants Freud’s conflict-model of psychopathology, which is based on the repression of internal sexual and aggressive wishes, by a psychology of self-defects and faulty function caused by hypothesized environmental events going back to the first two years of infancy. Relatedly, Kohut denies, contra Freud, that insight is curative, designating instead the analyst’s empathic understanding as the operative therapeutic agent. Again, the object relations theorists deny that the etiology of pathology lies in Freudian (oedipal) conflicts and traumas involving sex and aggression, claiming instead that the quality of maternal caring is the crucial factor. How could Nagel’s intuition even get a handle on deciding between these contemporary alternatives and Freud’s classical doctrines? It cannot, as he admits re the Kleinian theory. And since the intuitions of different people collide, I ask: Whose intuition is to decide which of the rival explanations “makes sense” of the phenomena correctly? Nagel’s recipe degenerates into subjectivity.

Nagel’s brand of “making sense” provides him with an “obvious,” oedipally derived, explanation of why an elderly man reputedly fell asleep whenever he began to hear stock market reports on the radio. Indeed, Nagel asserts that this case leaves “no credible alternative” to the conclusion that the old man’s unconscious was still troubled by a paternal injunction, issued over fifty years earlier, to listen to such reports. But the argument is basically flawed, if only because Freud never gave good evidence that when painful experiences are forgotten, the forgetting (construed as repression) is due to their painfulness, and thereby produces symptoms. As Freud had to acknowledge even from a childhood humiliation by his father, there are numerous instances of painful experiences that are persistently remembered vividly and even obsessively.

Thus, as we learn from a review of Darwin’s Autobiography [NYR, October 10, 1991, p. 31], “Darwin (presumably Charles Darwin’s father] had developed a remarkably retentive memory for painful experiences….” Yet Freud declares peremptorily without any statistical evidence:⁴ “The tendency to forget what is disagreeable seems to me to be a quite universal one.” Indeed just that ill-supported assumption is crucial to his entire theory of repression, which holds axiomatically that negative affect of various sorts (trauma, anxiety, motives of unpleasure) actuate forgetting to the point of repression and symptom production. Apparently Freud did not recognize that factors other than the painfulness of an experience determine whether it is remembered or forgotten. But the discernment of these factors is a key to understanding the as yet unknown statistics of the ratio of forgetting to recall of negatively charged experiences. And absent such statistics, there is clearly insufficient grounds for attributing the forgetting of negative experiences to their affective displeasure, let alone for ascribing neurotic symptoms to the repression of such experiences.

Nagel uncritically does just that when he takes the elderly man’s failing asleep during the stock market reports as a neurotic symptom, rather than, say, a case of boredom, and attributes to negative affect the forgetting of the paternal injunction. Besides, Nagel does not tell us whether the gentleman ceased falling asleep once he had lifted his own presumed repression of that injunction, as expected by Freudian therapeutic theory. Nor does he allow for the psychiatrist’s rash creation of a mind-set in the gentleman, when he told him, without any additional information, that his falling asleep during the stock market news “probably expressed difficult feelings about his father.” Thus, far from being “obvious,” Nagel’s Freudian explanation of this sleep pattern is simply baseless and contrived.

Nagel also falters when he joins Wollheim in objecting that, unless I can produce a theory of just how suggestion operates, I am not entitled to invoke suggestion as a cause of patient compliance with the therapist’s ideas. Why not? Surely pharmacologists concerned with assessing the therapeutic efficacy of the chemical content of drugs were entitled to assert that the suggested expectation of improvement makes for placebogenic gain, even though they had as yet no knowledge of the mediating dynamics of placebogenic therapeutic efficacy.⁵ Clearly, such entitlement did not have to await the discovery that placebos activate the psychogenic secretion of therapeutic endorphins, interferon, and steroids. The same point applies to suggestion operating in the course of psychoanalytic treatment, as shown by studies I cited in Foundations (pp. 211-212) and by recent research on induced false memories (New York Times, May 31, 1994).

Nagel asserts falsely with David Sachs and Paul Robinson that, in my view, “therapeutic success…[is] the empirical ground on which Freud’s theories must stand or fall.” This account is wrong, if only because I stressed the need for extra-clinical evidence, and even proposed the aforementioned epidemiologic test of Freud’s etiology of paranois. Nagel carries coals to Newcastle by citing against me (footnote 10) a statement by Freud that I myself had both quoted and explained in Foundations (p. 141). It is incontestable that Freud offered a therapeutic justification for the etiologic probativeness he attributed to his method of free association.⁶ But, as I pointed out in publications mentioned by Nagel,⁷ once Freud had convinced himself, on therapeutic grounds, that free association is etiologically probative in some cases, he also felt justified in deeming the method reliable as a means of uncovering the etiologies of the narcissistic neuroses—such as paranoia—in which he avowedly expected no therapeutic success. True, in his account of the Irma specimen dream, Freud offered a non-therapeutic argument for the use of free association as a method of dream interpretation. But, as I have shown (Foundations, ch. 5), there his case is transparently one of mere salesmanship.

Nagel believes that “Freud would have been delighted to tangle with Grünbaum.” But very disappointingly, he claims that the sociological fact of Freud’s supposed “pervasive” influence in our culture is “evidence for the validity” of the psychoanalytic enterprise. Yet this is a very dubious argument, if only because of the prevalence of vulgarized pseudo-Freudian concepts. Such vulgarization alone besets any attempt to give an objective answer to the question as to the extent and nature of the cultural influence of psychoanalysis proper. To illustrate this phenomenon, I gave an example of a pseudo-Freudian slip that had been provided by a psychoanalyst of long experience (Foundations, pp. 199-201). Yet my purpose was lost altogether on Nagel. Indeed, if pervasive cultural influence were evidence of validity, then religious superstitions and ethnic canards of stereotypes, which are far more prevalent than Freud’s ideas, as well as earlier witchcraft and slavery ought to possess a high degree of validity. Does Nagel apply his cultural criterion to them as well? If not, why not?