In response to:

American Prophet from the October 16, 1975 issue

To the Editors:

In his “American Prophet” [NYR, October 16], Frederick Crews criticizes Erik Erikson’s work on the ground that it is a personal world view masquerading as science. According to Crews, Erikson’s views are unscientific because they are based on psychoanalysis which itself fails to provide sufficient evidence for its claims. However, Crews presents his conclusions about both psychoanalysis and Erikson in a vacuum of supporting material.

First, he notes Erikson’s own contradictory attitudes toward psychoanalysis. But this is hardly adequate to deride psychoanalysis as a whole; in fact, psychoanalysts such as Edith Jacobson and Harry Guntrip have themselves pointed out these contradictions. More importantly, Crews makes critical generalizations about psychoanalysis: psychoanalysis uses “loose logic,” private data, and poor methodology; is permeated by ideology; postulates “crackpot categories,” and is dogmatic. Since Crews does not tell us to what aspects of psychoanalysis he is referring, we must, it would seem, accept that Crews knows what he is talking about. However, it might be worthwhile to take a quick look at what psychoanalysis holds itself to be. In a widely acclaimed text, The Technique and Practice of Psychoanalysis, Vol. 1 (New York: International University Press, 1967), Ralph R. Greenson describes “the minimum number of assumptions upon which the system of psychoanalytic theory is based” (p. 20ff). First, there is the topographic point of view which holds that conscious and unconscious phenomena function under different modes, the primary process and the secondary process. Is this one of the assumptions which Crews criticizes? Apparently not, for he says that the idea of “multiple significance,” which comes out of this assumption, is one of “the most promising aspects of psychoanalysis.”

In addition to the topographic, there is the dynamic point of view which assumes that mental phenomena are the result of the interaction of forces. Perhaps Crews finds fault with this aspect? Not possible, because he sees as “most promising” psychoanalysis’s “attentiveness to signs of conflict” and “its ideas of ambivalence, identification, repression, and projection”—all of which arise from this basic assumption. Another basic assumption of psychoanalysis is that psychic energy is distributed and transformed. Crews gives us no hint of his attitude toward this assumption, except in veiled remarks about sublimation. An assumption he does seem to be critical of is the point of view which concerns the origin and development of psychic phenomena. Crews apparently holds that human character is not demarcated by traumatic conflicts of the past. This is a remarkable position as Crews is seemingly denying the effect of the past on the present. As regards two other important assumptions, the structural (ego, superego, id) and the adaptive, Crews says little specifically.

Since in regard to the metapsychological assumptions just covered Crews seems either silent or in agreement (except in one surprising instance), he must have a different conception of psychoanalysis than psychoanalysts do. Therefore, we perhaps need to pay attention to what he specifically criticizes. For one, he seems to regard the “entire Freudian theory of childhood” as based on “remote and chancy inference,” and thus, it would follow, as false. Since Crews refuses to accept psychoanalytic methodology, it would be best to turn to a nonpsychoanalytic source for information regarding the validity of Freud’s theory of childhood. In a compilation of investigations into Freudian theory, Paul Kline in Fact and Fantasy in Freudian Theory (London: Methuen & Co., Ltd., 1972) notes that the Oedipus complex and castration complex have been supported by objective evidence. In fact, Kline finds that much of psychoanalytic theory has been supported by independent investigations.

Secondly, Crews seems to think that Freud’s views on female sexuality are inadequate. They may be. But to conclude that therefore psychoanalysis is dogmatic and inadequate is illogical. This view assumes that Freud’s own views are the sum and substance of psychoanalytic thought, and thus, if Freud is proved incorrect, psychoanalysis is also. However, this attitude ignores the expansion of knowledge within the field since Freud. I cite just one example from the foreword of Female Sexuality: New Psychoanalytic Views (Ann Arbor: University of Michigan, 1970), edited by Janine Chasseguet-Smirgel: “The authors agree on the overwhelming importance of the mother for the personality and later sexual adjustment of the little girl” (pp. vi-vii). Thus, for Crews to imply that the only thing psychoanalysis has to say about women is something about their “inceptiveness” and their dependency, is unjust.

Thirdly, another objection that Crews seems to have is against the concepts of resistance and denial—“With the concepts of resistance and denial ready at hand, the psychohistorian can place his central dramatic emphasis wherever he pleases.” Since Crews has been careful to explore Erikson’s contradictions, this is a curious statement. For, in fact, early in his article Crews has accepted defense mechanisms as some of the “most promising aspects of psychoanalysis”—“identification, repression, and projection”—yet now Crews objects to the concept of resistance which is the operation of the defense mechanisms of the patient within the therapeutic situation. It is unclear how clear Crews’ understanding of psychoanalysis is. In sum, Crews has presented nothing to indicate that “Psychoanalysis finds itself conceding one failing after another in the hope that some shrunken core of method or insight will be considered salvageable.”

Let us now turn to Crews’ conclusions about Erikson. First, he lauds Erikson as the “most resourceful,” “most humane,” and “most popular” of Freud’s heirs. To take nothing away from Erikson and his work, which I admire, how has Crews drawn these conclusions? As a practicing psychoanalytic psychotherapist I think this broad generalization of Erikson as the leading thinker in the field of psychoanalysis can be debated. In fact, many have concluded otherwise. Harry Guntrip in Psychoanalytic Theory, Therapy, and the Self (New York: Basic Books, 1971) gives Erikson credit, among others, in formulating new psychoanalytic views but finds the key focus elsewhere: “The work of Melanie Klein is the real turning point in psychoanalytic theory and therapy within the Freudian movement itself” (p. 47). He is not referring to her views on the death instinct but rather her focus on object relations.

In addition, there are other psychoanalytic theorists besides Erikson who must be given their due. W.R.D. Fairbairn has developed the concept of the internalized objects; Heinz Hartmann has focused on autonomous ego processes; D.W. Winnicott has developed the concept of “ego-relatedness”; and Edith Jacobson has explored ego-identity. Just as Crews seems to downgrade these analysts in comparison to Erikson, he also seems to ignore Margaret Mahler’s work on separation-individuation (see her new book The Psychological Birth of the Human Infant, New York, Basic Books, 1975).

Let me be clear: I do not mean to devalue Erikson. I do mean to point out that Crews has attempted to make Erikson seem more important than he is. Why? By making Erikson the most resourceful psychoanalyst and then knocking him down, he thereby believes he has knocked psychoanalysis further down. And, of course, this is what Crews attempts to do. He “explains” Erikson’s motivation, as the editors realize on the cover page—“Erik Erikson Analyzed.” Erikson “always has at least one eye on his reputation.” After all, he wonders “How independent of popular opinion has Erikson ever been?” About seven paragraphs later, Crews gives us the answer: Erikson is called the “supremely adaptable Erikson.” Crews has complained of the poor methodology of psychoanalysis, he has complained of Erikson’s poor methodology in book-length efforts, yet in seven paragraphs he believes he has presented enough data for this conclusion.

In conclusion, Crews has presented a personal view masquerading as objective. He speaks as if psychoanalysis has been refuted and is conceding point after point, when, in fact, point after point is being supported and there is still no better comprehensive theory of the human personality. He criticizes both psychoanalysis and Erikson for insufficient data, but he feels secure in presenting vague generalization after generalization.

Robert Mollinger

New York City

Frederick Crews replies:

Readers who are familiar with my numerous published attempts to sort out what is usable in applied psychoanalysis will know that Dr. Mollinger is incorrect in considering me a simple foe of the movement. Others may be puzzled by his unawareness of what Dr. Leo Stone, in a recent number of Psychoanalytic Quarterly, designates as “the currently mounting tide of criticism of psychoanalysis, both as to its scientific status and its therapeutic usefulness.” How high must the tide rise before Dr. Mollinger notices it? Classical psychoanalysis, Stone observes, “has been invested with a group narcissism of an inflexible and grandiose sort.” No deviation from the norm can be found in Dr. Mollinger’s Panglossian apologetics.

It would take a letter as long as Dr. Mollinger’s simply to repair his misrepresentations of my words and motives. I doubt whether any other readers of my essay imagined that I was “denying the effect of the past on the present” or claiming that “the only thing psychoanalysis has to say about women is something about their ‘inceptiveness’ and their dependency.”

Of more interest, I hope, is the general issue of psychoanalysis as a science. My position is Karl Popper’s: a body of theory cannot be called scientific unless its hypotheses can be exposed to tests for falsification. By this criterion psychoanalysis, though it states laws of development and behavior and may be right on many points, is no science. Its root assumptions, such as (to quote Dr. Mollinger) “that conscious and unconscious phenomena function under different modes” and “that psychic energy is distributed and transformed,” are unamenable to empirical testing, yet they exert a controlling influence over the whole of metapsychology.

The notion of “transformed psychic energy,” for example, leads some theorists (including Heinz Hartmann) to posit a congeries of aggressivized, deaggressivized, and neutralized energies, pressed into service not by any observation, but by the need to make a theory that starts from sexual energy cover the myriad human achievements it is clearly unable to explain. These conceptual novelties in their turn are beyond testing; some analysts believe in them and others don’t, according to taste. And the assortment of theoretical prejudices held by analysts inevitably reenters the technical literature of psychoanalysis, biasing the inferences drawn from clinical results which are themselves largely anecdotal. Furthermore, as I emphasized in my essay, concepts such as resistance, denial, and displacement encourage a Freudian to see confirming evidence where anyone else would see disproof. To believe a given hypothesis requires blind faith that its inventor—Freud, let us say, or Dr. Mollinger—is a person whose objectivity, skepticism, and good judgment can be trusted. No such faith is exacted by a genuine science.

Dr. Mollinger takes comfort in the thought that psychoanalysis continues to come up with vital new thinkers. Except for Heinz Hartmann, who published his chief contribution in 1939, his examples are Melanie Klein, whose major work is dated 1932, and several followers who have modified some of her more fanciful ideas about the “normal” sadism and paranoia of preverbal infants. “The work of Melanie Klein,” Dr. Mollinger reads in a book by another descendant of her school, “is the real turning point in psychoanalytic theory….” And so he believes it, though perhaps 95 percent of his American colleagues do not. Such impulse shopping for “real turning points” and definitive authorities is symptomatic of the discipline’s underlying want of empirical rigor. What a true believer like Dr. Mollinger takes for progress may be merely a change of intellectual fashion or a subtilizing of postulates that had little merit to begin with.

In Dr. Mollinger’s view, I am guilty of inconsistency for finding value in certain psychoanalytic ideas, for those ideas “come out of” broader notions that I unaccountably failed to dispute or even mention in my review of Erik Erikson’s latest book. The logic here is that I have inadvertently endorsed the general picture by praising some of its details. What this reasoning in fact shows is that Dr. Mollinger regards psychoanalysis as a deductive system, one whose particular concepts entail assent to the whole of metapsychology.

This differs strikingly from the view of Freud, who wrote (in “On Narcissism”) that “the whole structure” of psychoanalysis stands apart from metapsychological propositions, which “can be replaced and discarded without damaging” that structure. Freud knew that high-level theory was an after-thought to the relatively direct inferences of the consulting room. But for Dr. Mollinger and other Freudian scholastics, the broadest vistas inspire the most profound reverence.

Fortunately, quite a few others have come to see that this attitude is pointing their movement toward oblivion, or at best toward what Leo Stone calls a “genteel quasi retirement.” Language-conscious analysts like Roy Schafer are trying to see whether the theory cannot be trimmed to explain only what is known to take place between therapist and client, and academic researchers like Lloyd Silverman and Joseph Reyher have been trying to devise adequate experimental procedures for testing low-level hypotheses about the arousing and abatement of conflict. Though these procedures cannot reach to the questionable etiological principles on which psychoanalysis continues to rely, they can probably show that unconscious conflict exists and is associated with various types of symptomatic behavior. Analysts would do well to attend to such necessarily limited tests instead of concluding from one another’s loyalist writings that “point after point is being supported.”

I persist in thinking that psychoanalysis, whether or not it is an efficient therapy, has things to teach us—and that its most dogmatic partisans are in truth its most damaging enemies.

This Issue

February 5, 1976