In response to:

Physician, Heal Thyself: Part II from the October 13, 2011 issue

To the Editors:

Frederick Crews, in the second part of his essay titled “Physician, Heal Thyself” [NYR, October 13] cites Freud’s letter to his fiancée: “You’ll be fed [gefüttert] with cocaine and will have to give me a kiss at every resting place.” Crews comments: “Freud’s use of the verb for giving fodder to an animal hinted at an insecure and coercive approach to his beloved. She would have to surrender to his will, but cocaine would be needed to make it happen.” Actually, the word fütternis also used for feeding children, ill people, and anybody who is helped eating by another person. An embarrassing misreading and overinterpretation that marks this article as the propagandistic polemic it is.

Dietrich Rueschemeyer
Professor of Sociology Emeritus
Brown University
Providence, Rhode Island

To the Editors:

It is a great sadness that Frederick Crews should expend his considerable writing talents on turning Freud’s life into a B-movie, lurid with sex, drugs, omnipotence, and paranoia. It is a great sadness, too, that The New York Reivew should persistently turn to him alone for reviews that have to do with Freud, a figure who so evidently clouds Crews’s intelligence that he ends up giving us pages worthy in their insinuations and partial facts only of tabloid journalism.

The story of Freud’s cocaine use is hardly new; nor are his flaws. As with every intellectual edifice, there is much to argue with in psychoanalysis. However, to find the first and only cause of Freud’s thought and writing in a cocaine-fueled dream is a little like saying that we owe all of Balzac’s Comédie Humaine to caffeine or Darwin’s evolution by natural selection to the repeated recourse to hydrotherapy—the mineral water cure.

Freud’s intuitions about the role in human behavior of instinct and of the unconscious have been largely confirmed by evolutionary psychology, cognitive science, and neuroimaging. Freud made two great contributions. First, he greatly expanded Darwin’s insight that human instincts grew out of primate instincts just as human morphology grew out of primate morphology. Second, he postulated that the way the neurons of the brain—then recently discovered—are wired could explain how unconscious drives influence so much of our behavior. Both were foundational to our current understanding.

Certainly, Freud didn’t get everything just right. He greatly overestimated the role of the sexual instinct in normal and pathological behavior and his guesses about neural networks were primitive. His epistemology often intertwined description and explanation. But even Freud’s worst mistakes were intricate, brilliant, and entirely plausible. Crews makes the simplest of logical errors by attributing them to cognitive distortions caused by cocaine intoxication: Freud used cocaine; Freud didn’t get everything just right; ergo Freud’s theories are all wrong and must have been cocaine-induced.

Crews’s potshots at Freud as treatment innovator are equally misguided. Before Freud, all psychiatrists worked only in hospitals: there was no outpatient psychotherapy. All the current forms of psychotherapy have evolved from his work and research shows them to be equal to medication in efficacy for the treatment of mild to moderate psychiatric problems. Hardly the work of a deranged cokehead…

Let us recognize Freud’s limitations, but give him his due and refrain from misinformed character assassination.

Lisa Appignanesi
Chair, Freud Museum London
London, England

John Forrester
Professor of History and Philosophy of the Sciences
University of Cambridge
Cambridge, England

Allen Frances, MD
Professor Emeritus and former Chair
Department of Psychiatry
Duke University
Durham, North Carolina
Chair, DSM-IV Task Force

To the Editors:

There is cause for wonder in Frederick Crews’s second installment of the article “Physician, Heal Thyself.” In accordance with his well-known opinion of psychoanalysis, the author invites us to accept what in my view is a faintly substantiated claim about the consequences of cocaine use on Freud’s intellectual production.

On the one hand, as in the following excerpt, Crews praises Freud’s works in neurology, which were written at a time when he was admittedly making a liberal use of cocaine:

No one in the immediate grip of cocaine could have written the impressive neurological tracts—notably on aphasia and cerebral palsy—that Freud was still publishing as late as 1897. They display a deft command of medical literature, of scientific reasoning, and of the conventions of collegial citation and measured disputation.

Yet Crews ascribes to the deleterious effect of cocaine Freud’s invention of psychoanalysis, whose clinical method and theoretical structure were elaborated during the forty-odd years after he became abstinent. There was a four-year overlap (1895–1899), but in that same period Freud was simultaneously laying the ground for psychoanalysis and producing some of the “impressive neurological tracts” praised by Crews. So one wonders what role cocaine did really play.


At one point, Crews introduces what at first seems a nuanced version of his idea:

Nor, for that matter, did Freud’s psychoanalytic prose ever suffer from the incoherence into which Halsted had briefly plunged…. Psychoanalysis, then, was not the product of a cocaine addict, but it emerged from what we might call Freud’s cocaine self—the side of his personality for which “everything became transparent” in bursts of illumination….

Then again, one wonders what difference could possibly run between the intellectual production of a “cocaine self” and that of a “cocaine addict.” Or, to put it differently, how could the so-called “cocaine self” affect Freud’s psychoanalytic—but not his neurological—thinking? This sounds rather like a conceptual hairsplitting aimed at glossing over the patent contradiction that undermines Crews’s thesis. Crews, indeed, cannot have it both ways: either Freud’s scientific judgment was disturbed by cocaine or it was not. If it was, whence the “impressive neurological tracts”? If it wasn’t, how then can one ascribe psychoanalysis to his “cocaine self”?

Psychoanalysis was given its name in 1896, but 1895 is generally considered the transitional year between Freud’s neurological and psychoanalytic periods. His writing of a “psychology for neurologists,” better known as the “Project for a Scientific Psychology,” is considered by many as the major marker of that transition. Freud himself obviously considered the piece a “premature attempt”: he never tried to publish it and was ready to have it burned with his letters to Fliess when Marie Bonaparte informed him that she had them in her possession. So the Project, by way of its “hybrid” nature and its rejected status, looks like a good candidate, as Crews indirectly suggests, for illustrating what Freud’s “cocaine self” was capable of.

Surprisingly, however, the Project is a sober, rational, and outstandingly rich essay. One of its central concepts, neural “facilitation” (among many other fascinating ideas), is clearly redolent of today’s widely accepted “Hebb’s law,”1 only fifty years ahead of time. In spite of the limitations imposed by the then nascent neurology, the Project was thought to be worth a fresh scientific scrutiny in the modern era, following which it was dubbed a “preface to modern cognitive theory and neuropsychology” by the eminent psychologist Karl Pribram (in no way a Freudian).2 So, even this “hybrid” essay, this “kind of madness” (Freud’s words cited by Crews), far from supporting the theory about Freud’s “cocaine self,” actually runs against it.

Cocaine may—or may not—have given Freud the feeling that “the barriers suddenly lifted, the veils dropped, and everything became transparent,” but it did not seem to hamper the intellectual achievement. As Freud’s correspondence shows, the paper was the result of many months of laborious theoretical strivings, hardly the product of acute intoxication. Many authors, by the way, report euphoric states similar to Freud’s when experiencing a breakthrough in their work and are unable to fully reconstruct the state of mind they were in at the time of their epiphany.

Leaving aside the question of the intrinsic value of psychoanalysis and the complexities of how cocaine influences intellectual production (it clearly does, though probably not in the way Crews thinks it did in Freud’s case), I only wish to stress that, in view of the preceding remarks, one can hardly espouse Crews’s theory. If his thesis indeed carried any truth, then one could just as well conclude, ironically, that cocaine’s influence was a rather positive one. Or else, professor Crews should explain how it is possible for cocaine to have had opposite consequences on Freud’s neurological and psychological thinking respectively.

Dominique Scarfone, MD
Professor of Psychology
Université de Montréal
Montreal, Canada

Frederick Crews replies:

1. Permit me to doubt that sadness on my behalf was the feeling experienced by the Freudians Lisa Appignanesi and John Forrester upon reading my two-part review essay. Their letter, an exercise in faux imperturbability, is the dignified alternative to many apoplectic blog posts by members of the dwindling psychoanalytic community. And the dignity of the strenuously maintained Freud legend is just what they are hoping to preserve. Move along, everybody, they are urging at the scene of trouble; there’s nothing unusual for you to see here.

Appignanesi, Forrester, and their co-signer Allen Frances tell us that “the story of Freud’s cocaine use is hardly new.” They must be referring to the story told by Ernest Jones, who touched only glancingly on Freud’s obtuseness, malpractice, and persistent untruthfulness regarding Ernst Fleischl’s addiction, and who withheld his own privately confided opinion that Freud “took more cocaine than he should” in a fifteen-year period stretching from 1884 to 1899. The “tabloid” facts in my essay are mostly Freud’s own statements in long-suppressed letters. As biographical scholars, Appignanesi and Forrester might be expected to welcome such a correction of the deliberately falsified record, but as loyal Freudians they prefer the status quo.


These writers pretend that I treated cocaine as “the first and only cause of Freud’s thought and writing.” Actually, I made it clear that psychoanalysis had many intellectual sources; in one sentence I cited eight of them without mentioning cocaine. The drug, I proposed, wrought its chief effect in lowering Freud’s resistance to acting out the dream of heroic conquest he had nursed since childhood. As he encountered frustrations in a medical career for which he was ill suited, and as he found himself incapable of devising purposeful experiments, cocaine euphoria encouraged him to believe that he could take shortcuts to fame.

The first such shortcut, tellingly, was his public advocacy of cocaine itself in papers whose impressionistic manner, carelessness, boasting, and, finally, sneering sarcasm toward opponents had no precedent in his earlier work. The second and more successful bid for glory was psychoanalysis. Its false therapeutic claims, its indefiniteness, its circular demonstrations, its revolutionary tone, its pretension to limitless insight, and its recourse to bluffing and slander in place of evidence all pointed to a break with the scientific ethos and a surrender to grandiosity on Freud’s part. Do Appignanesi, Forrester, and Frances seriously believe that cocaine played no part in such a radical transformation?

These writers are disingenuous once again when they caricature my essay as having asserted that psychoanalytic theory is wrong because Freud ingested cocaine. Appignanesi and Forrester have long been familiar with my actual position: psychoanalysis went astray from the outset by exempting itself from consensual standards of validation. Thus “psychoanalytic knowledge” immediately became an oxymoron; and it remains such today, because post-Freudian claims are as anecdotal and unconstrained as their predecessors.

The standard contemporary means of shielding Freud from fundamental criticism is to declare that he was wrong about some matters but prescient about others; the critic who isn’t satisfied with that bland generality is then labeled a fanatic. Appignanesi, Forrester, and Frances play this familiar tune: Freud “didn’t get everything just right,” but many of his “intuitions” have been “confirmed by evolutionary psychology, cognitive science, and neuroimaging.” Need I point out that only propositions, not intuitions, are scientifically testable? Some of Freud’s misty pronouncements about “instinct” and “the unconscious” can be analogized to actual findings, but that is unsurprising, as he speculated broadly and took no pains to avoid self-contradiction.

Readers shouldn’t rely on either Freudians or me to assess the status of psychoanalytic ideas. The question is whether those ideas have found acceptance among researchers who honor the rules of science. Out of 1,314 items referenced in the standard work on cognition and emotion, only one is by Freud.3 And twelve years ago, a comprehensive citation study in the flagship journal of American psychology reported that “psychoanalytic research has been virtually ignored by mainstream scientific psychology over the past several decades.”4

This collapse could prompt historians to return to Freud’s career and ask what the empirical basis of his tenets was, how he checked for error and excluded rival explanations, and how he dealt with objections by skeptical colleagues. But no satisfactory answers would be found. The defense mechanisms, “dream work,” and symbol code that Freud claimed to have unearthed were his own devices for twisting every patient’s words into support for arbitrary postulates and interpretations. As for his wondrous breakthroughs, we have only his self-dramatizing word for them; and for corroboration we have only his question-begging assurances that “psychoanalytic experience” had proved him right. Freud imagined himself a second Darwin, but he had more in common with Walter Mitty.

Appignanesi, Forrester, and Frances credit Freud for the success of recent psychotherapies that came into existence precisely because of dissatisfaction with the weak results obtained by psychoanalysis. (Thank you, Freud!) He was indeed very influential, but his vogue retarded both empirical research and the treatment options that could have flowed from it.

2. Dominique Scarfone characterizes psychoanalysis as having taken shape over forty-plus years, during most of which time Freud wasn’t using cocaine.5 Nice try. But the “overlap,” 1895–1899, when Freud increasingly showed symptoms of cocaine intoxication, was the very period in which he irreversibly committed himself to ungrounded psychological guesswork, insulated from disproof.

The central issue raised by Scarfone is how Freud could have been affected by cocaine in his psychoanalytic cogitations but not in his neurological ones. My essay addressed that question and granted that the author of On Aphasia couldn’t have been a drug addict like William Halsted in his most abject phase. Even Halsted, however, later managed to quarantine his cocaine habit from his work. I suggest that Freud did the same until about 1897, but only when holding himself accountable to collegial norms. Scarfone’s challenge ought to be turned on its head: What besides cocaine could account for the fact that Freud, for a few years, simultaneously followed a scientific agenda and a pseudoscientific one?

Scarfone would like Freud’s 1895 “Project for a Scientific Psychology” to stand as a test of my thesis about the effect of cocaine on Freud’s mind. That essay, Scarfone writes, marks the transition “between Freud’s neurological and psychoanalytic periods,” and it is a triumph—“sober, rational, and outstandingly rich,” as well as being anticipatory of modern neuroscience. Surely, then, it couldn’t have been composed under the influence of cocaine.

Scarfone, however, misrepresents the Project’s place in Freud’s career, its intention, and its degree of success. The work didn’t lead Freud to psychoanalysis; his central doctrine of repression had already become a dogma, and he was freely theorizing about the psychoneuroses. But suddenly he became obsessed with devising an enormously ambitious model of the mind, reducing thought, sleep, memory, sexuality, symptom formation, and much else to what he called “quantitatively determinate states of specifiable material particles.”

Scarfone calls the Project the fruit of “many months of laborious theoretical strivings,” but the editor of Freud’s Standard Edition, James Strachey, found that it had been “dashed off” and left unfinished after two or three weeks of actual work. Strachey characterized the document as a torso without a head and wisely counseled us not to “read into [Freud’s] sometimes obscure remarks modern interpretations that they will not bear.”

There is no doubt that the undertaking was a fiasco. Freud realized with mortification that he had somehow forgotten to consider the crucial feature of mental activity, consciousness. And there were further reasons why the Project was doomed from the start. Why, when Freud’s clinical notions were still in flux, had he allowed himself to spend frenzied nights on such an overreaching quest? Scarfone has chosen a strange example to illustrate Freud’s sober rationality in 1895.6

3. Dietrich Rueschemeyer is right: I failed to recall the gentler meaning of füttern. Fortunately, the mistake was immaterial to my general argument. Even in the passage at hand, we still see Freud, in his sexual insecurity, fancying that he will extort kisses by serially feeding cocaine to his passive bride.