To the Editors:

Colin McGinn [“Freud Under Analysis,” NYR, November 4, 1999] fails to distinguish between exposition and demonstration in Freud’s writings. He assumes that the truths of human psychology are self-evident to the thoughtful intellect and that their pursuit requires no more than honest introspection.

“My point then is that I don’t know of any case in which such a lapse of memory on my part is attributable to repression.” He misses the point that in the presence of repression, even the amnestic hiatus is not evident. Parents of adult children often remember sentiments expressed by the children when they were very young but which the children no longer recall nor recognize when they are presented to them as adults. However these very same wishes may appear in the dreams or fantasies of these now adult children; such sentiments are often confirmable by checking with parents or siblings.

Freud’s propositions are derived from and based upon actual clinical experience. Slips of tongue and parapraxes are not important sources of information.

“We typically know the central part of the psychological explanation of our actions.” Do anti-Semites know why they hate Jews? Did Hitler know why he wanted to destroy all of Europe? Do masochists know why they want to be hurt and humiliated? And do sadists know why they wish to accommodate them? Does the depressed patient know why he wishes to die, and does the manic know that or why he wishes ultimately to be defeated? Do we understand the reasons for our moods? Does a man know why one woman appeals to him above all others? It is common to reply to these questions with rationalizations, but rationalization isn’t reason.

Although there is much to challenge in Mc Ginn’s essay, to be practical let me limit my comments to some of his remarks about dream psychology and dream analysis. McGinn questions the proposition that dreams serve to give expression to unfulfilled wishes.

Perhaps I can clarify the situation by taking advantage of information that was not available to Freud one hundred years ago. Mark Solms1 clearly demonstrates in his anatomical studies that a neurophysiologic system that Jaak Panksepp2 calls the seeking system is responsible for initiative and enterprise in the waking animal (and human), and for creating needs during sleep that find their way into dreams. This system (arising in the brain stem, traversing the lateral hypothalamus, and reaching the medial orbital surface of the frontal lobe) depends largely upon dopamine neurotransmission, one of the neurotransmitter systems that remain active even during sleep. Repeatedly during sleep, motivational impulses for the satisfaction of instinctual and other needs and for the resolution of troublesome memories and concerns strive for activation. The dream seems to attempt to attenuate the impulses by finding one or another way of representing them as fulfilled. The hallucinatory fulfillment of these impulses—or wishes, which is what they correspond to in ordinary thought—may result in temporary attenuation, or may elicit a signal that even a thought of fulfillment would elicit great danger so that anxiety appears, or the individual awakens.

The impulses that find their way into dreams are not exclusively unconscious wishes, nor did Freud say they were. If McGinn would reread the writings upon which he depends, he will find that Freud called attention to the fact that all kinds of unfulfilled needs may find their way into dreams. For example, needs for urination or defecation during sleep find their way into the dreams of many people at frequent intervals. The dreams often end with the introduction of some obstacle to the fulfillment of the need, for example one might dream of looking for a toilet but then be unable to find privacy. Characteristically the dreamer will awaken and take care of the need more realistically. Similarly, excessive heat or cold during the night, hunger or thirst, the discomfort of illness will achieve representation during the dream, and an effort will be made to resolve the problem with a dream hallucination. Of course it fails.

“What about common anxiety dreams that represent something feared, not wished for?” Surely it will have occurred to Professor McGinn that many of the things that we devoutly wish are to be feared, for example, to engage in an extramarital affair with the wife of one’s closest friend; to contrive the death of a rival; to tempt fate.

“We could certainly sleep without dreaming—we often do—so that dreams are hardly required to sustain sleep.” The fact is that about thirty minutes after we fall asleep we normally start to dream and continue dreaming to the accompaniment of rapid eye movements for fifteen to twenty minutes. As sleep continues, these rapid eye movement epochs become longer and occur more frequently. Commonly we awaken as the last of these episodes terminates, or shortly thereafter. Sleep does not last very long before an epoch of rapid eye movement accompanied by dreams intrudes.


Since dreaming is activated by the arousal of instinctual needs under the influence of the seeking system, males almost always exhibit penile tumescence during dreaming. Since sleep usually terminates with a dream, whether or not the dream is recalled, they commonly awake with erections. If we do not recognize the dream as a response to instinctual activation, how can we explain the regular occurrence of these erections in the male dreamer (and engorgement in the female)?

“Dreams may be necessary for mental health, but not for the possibility of sleep itself.” It is possible to deprive an individual of dreaming by awakening him whenever the sleeping electroencephalogram demonstrates that an epoch of eye movements is commencing. After a day or two of such deprivation, it becomes impossible for the individual to sleep without immediately falling into a rapid eye movement dreaming sequence.

Let me illustrate by presenting briefly a dream that was reported during the last session before I read McGinn’s article. An aging business executive of some distinction dreamed that one of his younger protégées had deserted him. Then there was another fragment about holding a dog on a leash. And finally there was a fragment about a labyrinthine-like area of hills and ravines. The man had started the session by reporting that two of his close and important friends had recently died, a breach of attachment. I knew that he had been forced to leave his native country as an adolescent because of the Holocaust—another breach of attachment—and had found refuge here where he had achieved extraordinary success. He was seeing me because he was depressed (he was one of the few patients whose depression did not respond to antidepression medication) and there was a good deal of evidence that the main reason for his depression was his recognition that because of his age he was being forced out of the prestigious positions that he had obtained. This was an unwelcome thought to him and even though the evidence had forced him to acknowledge this a number of times, he managed to deny?—suppress?—repress? that realization. The dream too was about breaking attachments: the disappointment by the younger colleague whom he had intended to promote to succeed him, the physical attachment to the dog by the leash, and the fear of not being able to escape the comfort and tranquillity which retirement would have seductively offered to him (the attractive labyrinth). Although these considerations are repressed by day, they contribute heavily to his state of depression and seek representation at night.

Dreams do not deal exclusively with the sexual wishes repressed in early childhood. They may deal with real sources of stress: recent trauma; insistent physiologic homeostatic deviations, repressed, suppressed, or denied; current issues; and powerful but repudiated inclinations such as homosexual sentiments and murderous desires. Any of these may find their way into dreaming consciousness and the dream will attempt to dispose of them as expeditiously as possible. When it succeeds, the dream might or might not be recollected; when it fails the individual might awaken with anxiety.

Clearly, uninformed introspection does not suffice for scientific endeavor, certainly not for psychology.

Mortimer Ostow


Psychoanalytic Research and

Development Fund

New York City

To the Editors3 :

Colin McGinn’s piece on Freud is extraordinarily naive. It is as foolish to try to “analyze” Freud sixty years after his death without taking account of what has happened in psychoanalytic thinking in the meantime as it would be to appraise Darwin without locating him in post-Darwinian evolutionary biology. The measure of a thinker is not the correctness of his views but their fruitfulness for later research. Freud started people thinking about child development; he stirred us to look at children as having complex emotional lives. He was wrong about a great many things, for one, how little girls view their bodies, but he made us notice that children have a view of such things, and that it matters.

McGinn is skeptical of the Oedipal complex. Quoting a long passage in which Freud makes a number of claims, McGinn asks: “What is the evidence for this amazing suggestion?” Without noticing that these claims are separable, he then continues: “At the least one would want to see a careful empirical investigation into the desires and fears of male infants, which should not in principle be too hard to obtain” (p. 23). It isn’t hard to obtain. Infant and child studies have proliferated in the last seventy-five years, and they confirm a number of Freud’s observations as they disconfirm some others. Conflicts between autonomy and intimacy, the relations between a child’s self-representations and his representations of others, the different ways in which the need for psychological safety is experienced at different periods in a child’s life, are some of the dimensions of research in child development since Freud.


“Granted that the child obtains pleasure from its mouth and anus, during sucking and defecation; but why insist on calling this sexual pleasure?… These pleasures are not in adulthood sexual in nature, so why are they in the infant?” McGinn asks (p. 23). He seems unaware of the so-called sexual “perversions,” not to mention fore-play. How can we account for the fact that activities like sucking, shitting, and pissing on another, taking in and extruding, along with mastering and being mastered, seeing and being seen, hurting and being hurt, so readily acquire a sexual meaning? Freud’s answer to this quasi-philosophical question is that sexuality begins much earlier than we had thought, that it is not to be identified exclusively with genital sex and procreation, and that it is structured by those bodily activities which bring the infant into daily, intimate, pleasurable, highly charged, and often conflicted contact with its caretakers.

One of Freud’s major lessons is that the world does not come already interpreted but is given the meaning it has for any person in part by what has happened to him, which is itself construed through the lens of his passions, interests, and internal conflicts. Another lesson is that the mind is not the master in its own house. McGinn’s essay reminds us just how hard that lesson is to take in. I am not suggesting that in this very instance McGinn is guided by unconscious biases, but that he writes as if he never were. He reports the “impartiality” of his reading of Freud as confidently as one might report the color of his hair; he says “I don’t know of a single desire or memory I have ever repressed, in Freud’s sense” without a wink.

“Children are supposed by Freud to be particularly adroit at this operation of repression. But why can they achieve easily what I am unable to bring about as an adult?” McGinn asks. “Am I not usually more in control of my mental life than a child? I am sure that I have forgotten much that I once knew, but I have to report that I don’t know of a single desire or memory I have ever repressed, in Freud’s sense” (p. 22). McGinn anticipates the objection that repression, by definition, is a phenomenon of which one is not self-aware. He answers that one can have indirect evidence of a memory lapse. Yes, and a “lapse” of memory or perception of something which is conflicted is indirect evidence for repression. There is a good deal of such evidence even from nonpsychoanalytic quarters. For example, experiments with a tachistoscope show subjects to be registering subliminal perceptions that are connected with anxieties they have previously reported, perceptions of which they claim ignorance. So far, the best explanation is that something inhibits these stimuli from being dealt with in consciousness. This is the phenomenon Freud called “repression.”

In Freud’s later theory, repression is bound to anxiety (the key text is Inhibitions, Symptoms, and Anxiety), a concept McGinn does not mention. The literature on repression, on “splitting” of consciousness, on “dissociation,” terms for overlapping areas of phenomena, is enormous. Yet armed with nothing more (if his footnotes are a clue) than Grünbaum and Sartre, McGinn thinks he is in a position to suspect “that the whole idea of Freudian repression is a myth…” (p. 22).

As a reader of Freud, McGinn is careless. He criticizes Freud’s theory in The Interpretation of Dreams that every dream is a wish fulfillment; yet it was Freud who noticed that some dreams seem to repeat a past trauma and so do not fit this theory. This led him to the revisions in Beyond the Pleasure Principle in which he posited the death instinct, one of McGinn’s principal targets. Most psychoanalysts repudiated this piece of theory from the beginning, even as they realized that Freud was pointing to the puzzling and central clinical fact that people often “repeat” scenarios which were, and are, deeply painful.

“What of the id, the ego, and the superego—weren’t these important discoveries?” McGinn asks (p. 24). Neither Freud nor any psychoanalyst after him thinks of them as “discoveries.” Freud was trying to give us tools with which to study the various interdependent functions of the mind and the complex phenomena of internal conflict.

McGinn asks: “Is [Freud] a bold and brilliant scientist…or an opportunistic founder of a crack-brained cult with little or no intellectual credibility?” (p. 20). The last paragraph answers: “It may be that… Freud’s theories are just one more myth about human nature…” (p. 24). The scope of McGinn’s essay is clearly not just Freud as writer or individual, but the Freudian legacy. No data come free of theory, but to appraise the theory one has to know the data it tries to comprehend. It’s a two-way street. This doesn’t require “having psychoanalysis” as McGinn puts it. It does require reading some of the contemporary psychoanalytic literature. Thinking about child development, mental functioning and conflict, the relationship between analyst and patient, the use of dreams in therapy, the sources of anxiety have all changed since Freud got us started.

If the editors of the NYR wanted yet another essay on Freud, why didn’t they ask someone who knows what he’s talking about?

Marcia Cavell

Berkeley, California

To the Editors:

In his generally perspicacious article, Colin McGinn quotes Freud’s account of the events which activate the male infant’s castration complex, and asks: “What is the evidence for this amazing suggestion?… It is not enough to erect this hypothesis merely upon the basis of what neurotic adult Viennese patients at the turn of the century suggestibly say to their theoretically committed analyst.” By writing in these terms McGinn inadvertently perpetuates the familiar notion that Freud’s psychosexual developmental theories were based on material obtained from his neurotic Viennese patients. In reality the material adduced to corroborate the seduction theory, which Freud later claimed had led to his momentous discovery of infantile sexual phantasies and thereby to the uncovering of “the whole range of a child’s sexual life,”4 was essentially a product of his own preconceptions, and was largely rejected by the patients themselves.5 (A similar conclusion may be drawn from the case history of “Dora.”) Freud himself wrote in 1935, “The information about infantile sexuality was obtained from the study of men…,”6 yet few have queried why a medical practice consisting predominantly of women should have led Freud to a theory of male psychosexual development.

Notwithstanding Freud’s description of them as analytic “findings,” most of the psychosexual developmental claims consisted of his own imaginative conjectures, purportedly validated by tendentious interpretations of patients’ dreams, symptoms, and associations. This is nowhere more evident than in the late expositions of female sexuality, as can be seen by reading them without preconceptions: at one point Freud even claims that “girls regularly [sic] accuse their mother of seducing them,”7 something he had never reported previously, but suspiciously in accord with his belated “discovery” that female infants’ first libidinal attachment is to their mother. More generally, what never seems to have been asked is why, after three decades of probing the Unconscious of his female patients, the numerous “findings” reported in these papers had only just come to light.

McGinn’s critical comments in regard to the notion of “penis envy,” while well founded, fail to address the central issue. Freud apologists are only too happy to acknowledge that his prejudices led him astray and that such culture-bound ideas have long been relegated to the rubbish bin of psychoanalytic history. What they fail to provide is an explanation of the fact that, for several decades prior to its becoming politically unacceptable, psychoanalysts consistently claimed to have demonstrated the ubiquitous occurrence of penis envy.

Allen Esterson

London, England

Colin McGinn replies:

When I accepted the commission to write about Freud for this journal I took myself to be vaguely Freudian. It was only reading and thinking about Freud with a more critical and rigorous eye, as I might any other theorist of the mind, that caused me to question his central doctrines. My aim was to provide a fresh, clear analysis of Freud, and to provoke a less entrenched debate about the truth of his theories. Among the dozens of critical letters received in response to my essay, many of them vituperatively ad hominem, those by Mortimer Ostow and Marcia Cavell seem to be about the best of what can only be described as a bad bunch. However, I find that most of their criticisms are already dealt with in my original essay, had they bothered to read it with any care and attention; and the heavy hand of dogma and special pleading is evident throughout. Let me reply to their points seriatim.

Dr. Ostow seems to have understood hardly anything of what I wrote. Of course, I do not “assume that the truths of human psychology are self-evident to the thoughtful intellect,” as I make clear at a number of points. But I do think—and Freud would not disagree—that we very often have conscious motives for our actions, and hence know an important part of the psychological truth about why we act as we do—i.e., we have conscious minds. Nor do I “miss the point” that in repression “the amnesic hiatus is not evident”; I argue that, despite this, there ought to be evidence of repression in memory gaps that are best explained in terms of this mechanism, if repression ever really occurs. The fact that people forget childhood “sentiments” is by itself no evidence of such repression, since there may well be other explanations of such forgetting, e.g., the lapse of time (do I need to point out that memories tend to fade with time?).

Oddly, Ostow says that parapraxes are “not important sources of information”—which was clearly not Freud’s view, since he emphasized them as keys to the unconscious.

Ostow’s response to my point that not all dreams can plausibly be regarded as wish fulfillments is thoroughly confused. Of course, many dreams do involve wish fulfillment, conscious or unconscious; my point was that not all do. Ostow thinks it is a satisfactory reply to this to observe that we are sometimes anxious about what we wish for. But that does not show that all anxieties are about things that we also desire (as with my example of dreaming that I fail an exam). I did not say that Freud takes all dream-producing wishes to be unconscious, which they plainly are not (and Freud is quite clear about this). On the question of whether dreaming that a wish is fulfilled is a way of fulfilling it, Ostow seems to vacillate: on the one hand, he says that a dream can “attenuate” a wish; on the other, he remarks (quite correctly) that a dream hallucination obviously fails to remove actual thirst. He does not come to grips with my obvious point that you cannot satisfy a desire simply by hallucinating or imagining or dreaming that you have satisfied it.

Ostow cites evidence that we dream during most of our sleeping time. That may be so, but it does not contradict my claim that we can sleep without dreaming, so that dreams are not required to keep us asleep. It is not that we would spend our whole lives wide awake if we did not dream. Doesn’t fatigue have something to do with sleep?

Ostow asks why we have erections during and after dreams if Freud is wrong about dreams. Again, he totally misses the point: it is not that dreams are never sexual—they clearly often are—but they are not exclusively sexual. And it is surely bizarre to claim, as Ostow does (but not Freud, so far as I know), that all instinctual impulses give rise to erections during dreaming sleep: dreaming I am eating french fries when I go to bed hungry is hardly likely to cause an erection. I may still have an erection but Ostow hasn’t explained why.

At the end of his letter Ostow lists some of the different emotions that can show up in dreams, as if this goes against something I said. But my point was precisely that dreams are of many kinds, contrary to Freud’s monolithic wish-fulfillment theory; Ostow is agreeing with me here, not Freud.

Ostow portentously concludes that “uninformed introspection does not suffice for scientific endeavour,” as if anyone might disagree with this. Part of my point, of course, was that Freud’s theories do not meet the evidential standards we require for scientific theories of the mind (hence my remarks about the Oedipus complex). But it is also obvious that clarity of thought is a necessary condition for any kind of theoretical discussion, and Ostow’s letter is woefully lacking in that regard.

I wish I could say that Marcia Cavell’s letter is more rigorous, but it too seems to have been composed without any serious attention to my actual arguments. After describing my essay as “extraordinarily naive” and “foolish,” she goes on, in the same paragraph, effectively to concede that Freud’s views are largely wrong. We should measure a thinker not by the correctness of his views but by his fruitfulness for later research, she assures us. So was I right to claim that almost all of Freud’s distinctive doctrines are groundless, false, and confused, according to Cavell? I also fail to see why one has to consider all of subsequent psychoanalysis in order to evaluate Freud’s own theoretical pronouncements, any more than one would have to do this in order to evaluate Jung or Adler.Darwin has been triumphantly vindicated by later biology, while Cavell seems to agree that Freud’s ideas have been rejected. Nor did it take Freud to make us think about child development and the emotions of children; and is this the best thing she can think of to say about Freud’s contributions to psychology? Faint praise indeed.

Cavell goes on to say that later developmental work has confirmed some parts of Freud and disconfirmed others, without being very specific. But nothing she mentions bears upon the claim of mine she is trying to criticize—namely, that there is no empirical evidence that male children have the sexual emotions and fears contained in Freud’s theory of the Oedipus complex. Does she agree with me on this or not? We are not told.

Cavell misunderstands my point about oral, anal, and genital pleasure, which was that the pleasures of eating and defecating are not essentially sexual, though of course they may be co-opted into sexual contexts. Granted that the infant derives pleasure from sucking at the breast, why is this not just an instance of the pleasure animals generally derive from being fed, instead of being somehow covertly sexual? When adults eat and smoke and chew gum they derive pleasure from these oral activities, but what is the point of calling this pleasure sexual? The obvious point here is not all pleasure is sexual pleasure. I here refer Cavell to Malcolm Macmillan’s discussion of this and other matters in his magisterial Freud Evaluated, cited in my original essay.

The point I made about repression in my own case was intended as an invitation to readers to ask themselves how much Freudian repression they have observed in themselves and their acquaintances—as opposed to conscious suppression, denial, self-deception, simple ignorance of one’s character, and other sources of self-blindness. The alleged phenomenon of repression always seems to be about other people. Freud suggests that repression occurs in response to painful memories and desires, and that it is remarkably common; but can’t we all see that this simply isn’t so? I may try my hardest to push an unwelcome desire out my mind, or rid myself of a painful memory, but the fact is that I have no power to do this—I cannot intentionally make a conscious mental state unconscious. Try it and see: it just won’t work.

Cavell’s citationless remarks about subliminal perception do nothing to establish the reality of Freudian repression. First, the anxiety in question is conscious anxiety. Second, tachistoscopes present stimuli so quickly that conscious perception is impossible; let’s see whether a five-second presentation of an anxiety-provoking stimulus still results in unconscious perception. The general point I would make here is that we have come to use the concept of repression far too promiscuously and uncritically; it needs to be asked exactly what it means, how it is supposed to work, and whether it ever really happens. And I gave a number of arguments, which Cavell does not address, that show severe problems with the notion—particularly the problem that repression involves both knowing and not knowing that one has a certain mental state.

Cavell says that anxiety is a concept I do not mention. Actually, I mention it several times, in arguing that anxiety dreams are not necessarily disguised wish fulfillments.

Of course, Freud noticed that not all dreams appear to fit his wish fulfillment theory—how could he not?—which is why he had recourse to such implausible constructs as the death instinct: this provides the missing wish when a person dreams of trauma or failure or her own death or some such. But Cavell seems not to notice that this is exactly what I said about Freud: when a dream appears not to fit the wish fulfillment theory he introduces an ad hoc unconscious wish so that his general theory can be preserved. And again, Cavell appears to agree with me that the death instinct was a bad idea to begin with (or is she being “extraordinarily naive”?). Nothing she says here shows me to be a “careless” reader of Freud, though it does raise doubts about her reading of me.

I have no clear idea what Cavell means by saying that no one regards the id, ego, and superego as “discoveries” but only as “tools.” Does she think that the dynamic unconscious is not, for Freud, a discovery but only a “tool”—or the idea of ego defense, or the Oedipus complex, or the death instinct? Clearly, Freud regarded these ideas as important scientific discoveries about the human mind, not merely useful tools. Or does Cavell believe that even the atom or evolution by natural selection are not discoveries but “tools”?

Insofar as later psychoanalysis shares any of the Freudian doctrines I criticized—such as repression and the dynamic unconscious—then it falls under the strictures I directed against specifically Freudian theory. Insofar as it does not, it would have to be evaluated separately. I did not undertake to discuss later theorists in the psychoanalytic tradition, and I see no objection to limiting my attention to Freud himself.

Do I know what I am talking about? The fact is that it is not too difficult to find out what Freud believes, which is why he has always been such a popular success: he is a perfectly clear writer for the most part, and a rhetorically persuasive one. There is no need to be an encyclopedic Freud scholar in order to understand his theories—any more than one needs to be a Darwin scholar to understand his theories. The hard thing is to arrive at a critical assessment of Freud’s theories that is not clouded by ideology, osmosis, and inertia.

Allen Esterson’s letter strikes me as entirely sensible and a useful reminder of Freud’s dubious methods. I certainly did not intend to give the impression that Freud was free of preconception, bias, and opportunism—quite the contrary. Freud’s patients were indeed not so much the source for his theories as their unwitting victims.

This Issue

February 24, 2000