The American literary critic Frederick Crews explains why he has rejected Freud

How is it possible to pass so quickly from being an advocate of applied psychoanalysis to being an antagonist of the entire Freudian movement?

I wish that process had happened more quickly in my case. It actually look about twelve years – time enough to become disillusioned by misgivings I had always harboured, but had preferred not to weigh very heavily. I found myself gradually obliged, by critiques I couldn’t ignore, to shift the balance between the enticing features of psychoanalysis and unanswerable doubts about it.

What doubts?

Even my earliest published remarks on the subject granted that the empirical credentials of psychoanalysis are drastically weak. Well, that turns out to be something more than a passing concession or an inconvenience! If a purported science, or indeed any secular discipline, can’t satisfactorily lay out the grounds on which outsiders should accept its propositions, that should prove by itself to be an eventually terminal defect. Its most serious consequences will be felt within the discipline, for theorists will not be able to settle their differences on a reasoned basis. Hence the telltale history of the psychoanalytic movement thus far – a history of conceptual muddle, patchwork reform, excommunications and schisms, with an unending proliferation of competing chieftains, each one saying in effect: après le déluge, moi!

But surely psychoanalysts would dispute your statement that their claims are not demonstrable?

Many would dispute it. The question is: could they refute it? Let’s separately review the claims made for psychoanalytic therapy and those made for the Freudian theory of mind. Comparisons between the ‘success rates’ of various therapies can never be entirely accurate, but the best recent studies strongly suggest that psychoanalysis does no better – and probably no worse – than any of the other 200-plus competing psychotherapies. Some earlier studies indicate a much less favourable comparison, but they are open to challenge. And to avoid needless controversy, let’s also set aside the studies indicating that behavioural therapy is clearly superior to any psychotherapy, or ‘talking cure’, in treating certain disorders. Accepting only the least damning results, we find that they flatly contradict the unique therapeutic claims to which psychoanalysis has always pretended – claims of superior efficacy based on the rooting out of pathogenic causes, not of mere symptoms, and on the re-ordering of the patient’s whole personality.

Those claims are essential to psychoanalysis for two reasons. First, if they are false, analysts lack the only imaginable justification for steering patients into the most time-consuming and expensive of all therapies. And second, Freudians have generally maintained that their patients’ recoveries validate the accuracy of the causal reconstructions that those patients were persuaded to accept in the course of their treatment. Such validation may or may not be of interest to the cured patient, but it is vital to analysts themselves in providing evidence that their concepts fit well with psychic reality. Thus the supposedly unmatchable success of the therapy has been crucially involved in supporting the theory and its clinical applications against charges of arbitrariness.

But if the patients of some two hundred other psychotherapies recover at approximately the same rate of success as the Freudian ones do, that source of support counts for nothing. The conclusion seems inescapable that all those talking cures, including psychoanalysis, are in the same boat. Their unique and mutually exclusive causal claims are at once factually dubious and placebogenically useful, in that it is fortifying, after all, for a patient to have something to believe in. That initiation into a system of thought, however fanciful, is probably one among several factors that make any psychotherapy superior to no treatment at all.

Whether it is altogether ethical for the therapists to go on maintaining their unique curative pretensions is another matter. In the case of psychoanalysis, it seems a rather questionable practice to charge, say, $80,000 over a five-year period for the purpose of unearthing ‘repressed memories’ and grievances from early childhood, when the patient’s chances of recovery at a much earlier date would apparently be about the same with any simpler regimen that made no mention of those alleged pathogens.

Now let me address the claims of Freudian theory, which, as I explained, can’t rest on the uniquely successful results of the therapy, for no such results have been sustained. Where else might those claims lie? Analytic theorists have adduced three other sources of validation – each of them patently vulnerable to a neutralising criticism. First, there is ‘clinical insight’. Whether or not a patient improves, analysts maintain that they can confirm or disconfirm their immediate interpretations – thus vindicating the applicability of their general theory – by observing how the patient responds to those interpretations through emotionally-tinged language, gestures, ‘free associations’, slips of the tongue and subsequent dreams. Unfortunately, the constructions placed on all those phenomena suffer from a dubious reliance on the very theory that is supposedly being validated. The analysts’ rules for confirming their hunches, furthermore, are so loose that the very same utterance or reported dream element can be taken to mean one thing, its opposite, or something quite different. Nor, as Freud’s early friend Wilhelm Fliess bluntly told him, can analysts show that their sources of evidence are undistorted by suggestion. Even dreams, as is well-known, take on a ‘Freudian’, ‘Jungian’ or other thematic character according to the therapeutic doctrine to which the dreamer has been recently introduced. In sum, clinical insight is just as unavailing as therapeutic success for the justification of Freudian psychology.

Second, many believers cite as corroboration of psychoanalytic theory its power to illuminate extraclinical phenomena, from myths and poems to religious systems and the conduct of historical figures. I myself was once impressed by this line of argument, but I am afraid it leads nowhere. The illumination at issue is perceptible only to Freudian loyalists, who are determined from the outset to ‘apply the theory’ here, there and everywhere. Advocates of palm-reading and numerology can make entirely comparable claims about the ample scope of phenomena about which they can say something wise-sounding and self-consistent. The relevant question should be, not whether the favoured theory can ‘cover’ a given phenomenon, but whether it provides us with the most plausible of competing explanations.

To make such a case, it is of course necessary to weigh one’s own hypothesis against others – especially against simpler ones that purport to account for the same data. But in my experience, pro-Freudian investigators invariably have such a strong prior commitment to their own style of reasoning that those other hypotheses simply don’t come into view. Thus the vast body of ‘applied analysis’, though it certainly yields some useful reflections, can’t be said to have any bearing on the validation of psychoanalytic theory.

The same objection eliminates the final potential source of validation – namely, the experimental study of Freudian concepts and hypotheses. It is true that some investigators (all prior Freudians) think that they or others have achieved some experimental confirmations. But in every case the ‘confirmation’ consists of showing – usually with a scandalously weak experimental design – that a certain predicted consequence does obtain. Even in the best of cases, that is only a start toward validation. The Freudian investigator never thinks to ask the essential next question: whether a rival hypothesis or concept could explain the same results as well or better, importing fewer gratuitous assumptions. A memorable example that I cited in my essay is the experiment that purported to show, from the fact that some female college students dreamed of penises, that they yearned to affix those appendages permanently to the exterior of their crotches. Who but a Freudian would fail to think of a more parsimonious explanation of the data in that instance?

The picture you paint is very bleak, but is it necessarily ‘terminal’? Why can’t psychoanalysts reduce their claims and bring their theory into conformity with what they actually know?

First of all, what do they know? Ideas that can’t be confirmed to some minimum degree simply don’t count as knowledge. But let me grant that there are analysts at work who appreciate the problem of unsupported claims and are trying to simplify the theory. Their efforts are commendable as far as they go. I submit, however, that any truly searching reappraisal of psychoanalysis would leave the analysts without any rationale for their circuitous and exceptionally protracted therapy. Thus, if a theorist were to admit (as several have) that he doesn’t know why Freudian therapy sometimes works and sometimes doesn’t, he would be bringing into question the value of laboriously ‘undoing repression’ and ‘analysing the transference’ – activities that cost dear in time and money without yielding any specific demonstrable benefit. And if another theorist were to strip Freudian concepts down to what is perceptible in the therapeutic encounter itself, without reference to infantile stages of libido, oedipal fixations, repressed traumas and the like, he too would be undermining the raison d’être of all that excavating into the forgotten past. I think, by the way, that quite a few analysts who were trained in an orthodox spirit have retreated from full-scale practice of the therapy. All I ask from them is some candour about why they have made a de facto break with Freudianism.

If as you say, the claims of psychoanalysis remain wholly unvalidated, how do you account for the influence psychoanalysis has exerted on American psychiatry?

That is a complex matter that could be approached in a variety of complementary ways. For one thing, psychoanalysis rode into authority on a wave of sexual ‘enlightenment’ that had less to do with authenticated discoveries than with changed social and moral norms. Second, when psychoanalysis came to America it closed down Freud’s hospitality to lay candidates and became a tightly-controlled medical guild, thus blurring the line between orthodox psychiatry and a movement that had hitherto been on the borderline of respectability. Even though the training that leads to an MD or even onward to a psychiatric licence has little bearing on the practice of psychoanalysis, analysts have been able to share in the august aura that surrounds doctors in this society. And third, many psychiatrists are no less susceptible to the promise of a comprehensive explanatory system than literary intellectuals are – especially when the system is couched in rigorous-looking scientific terminology. Even so, I’m sure you are aware that the prestige of psychoanalysis has been waning in recent decades.

Is it your position that psychoanalytic theory has been completely disproven?

No. Actual disproof is possible only when a given idea has been formulated clearly enough to be tested, and some segments of Freudian theory are as safe from refutation as they are lacking in any practical implications. My position is simply that psychoanalysis shows every sign of being a fatally troubled explanatory paradigm. There is no need for opponents to disprove its postulates one by one. Psychoanalysis will wither away if discerning people lose interest in it – cease to find it helpful or plausible as a total outlook. Such a loss of interest usually occurs when people notice that the doctrine has generated more anomalies than solutions and that, after a reasonable period, it has still failed to correct its ambiguities or to justify its pretensions. Those are signs that the doctrine probably is erroneous in more ways than anyone has established, but a flawed paradigm typically stumbles into oblivion without having been decisively refuted. I expect that to be the fate of psychoanalysis.

You begin your recent article ‘Analysis Terminable’, though, by conceding that Freudian discourse has never been more popular in certain departments of the universities.

Yes, but that fact tells us more about anti-empirical tendencies in the academy than it does about the health of psychoanalysis. The Freudian thinkers who are most admired in ‘advanced’ literary-critical circles – writers like Jacques Lacan – are obscurantists who openly scorn validation, and the concepts most in vogue, such as the death instinct, tend to be those most distant from any possibility of confirmation.

Do you still admire Freud in spile of everything?

I admire Freud in about the same way I admire George Bernard Shaw: he dominated an age, and he had extraordinary rhetorical skills. Both of them promoted a good many half-baked ideas. Unfortunately, in Freud’s case the ideas were primary: after we have seen through his pretensions to certain knowledge, there is no independent literary legacy to cherish. When Shaw declared himself superior to Shakespeare, it was nothing more consequential than a marvellous piece of offstage theatre, but when Freud likened himself to Copernicus and Darwin, he was laying serious claim to scientific discoveries that were in fact arbitrary pronouncements on his part. Until recently, most of the commentary surrounding Freud has been devoted to interpreting him, as if we could arrive at exact knowledge of the human mind by sucking dry those ambiguous texts of his. But Freud is now seen increasingly as a fallible would-be dictator and as a prisoner of outmoded scientific notions. Think, for example, of Paul Roazen’s two books about Freud’s less than saintly relations with his followers, or of Frank Sulloway’s exhaustive book about Freud’s reliance on the vulnerable biology and ‘sexology’ of the late 19th century. The more we learn about Freud’s actual grounding in his age, the less possible it will be to sustain the Promethean myth of his having seized essential, forbidden knowledge that liberated mankind from the thrall of repression.

The main reason I feel confident in predicting that Freud’s reputation will dwindle is that, unlike the great pioneers of science, he never presented any evidential grounds for his theory. That may sound astonishing, but I mean it in a precise sense. Freud merely told us, for example, that he had successfully analysed himself and seen that his self-diagnosis applied to everyone in every past or present society. He told us that every neurotic symptom and every mental or behavioural slip has an exact meaning that condenses conflicting desires; he told us that dream elements can be resolved into those that express contemporary wishes and those that convey the persistent wishes of infancy; he told us that his therapy had produced singularly effective results; and so forth. Beyond all those ex cathedra assurances, which look less satisfying with every passing year, all we get from Freud are interpretations. The interpretations are spectacularly complex and inventive; they can seduce us by their combination of boldness and subtlety. Yet they merely apply the still untested theory, showing that it can be made to ‘cover’ just about anything. Coverage alone, as I have explained, is not corroboration. Once people begin calling for the real evidential goods to be delivered, they must see that they have been asked to admire a master, not of scientific scruple, but of intellectual self-will.

I do think, though, that Freud will remain an intriguing figure to investigate. I predict, for instance, that scholars will increasingly bring his irrationalist and even spiritualist side into focus. Although Freud talked the language of scientific scepticism, the truth is that he was credulous in ways that are connected to Faustian currents in late 19th-century thought. For example: in full awareness that Lamarckianism was in disrepute, he insisted on Lamarckian ideas about the effect of certain imagined prehistoric events on the modern sense of guilt; he flirted with telepathy; he cast his lot with those who thought that Moses was an Egyptian and that the Earl of Oxford had written Shakespeare’s plays; and he thought he could remount the stream of history so far that he could even reveal to the world how weaving had originated – namely, with the coquettish braiding of adolescent pubic hair among our earliest female ancestors. Such indulgences on the part of an alleged scientific pioneer should be fascinating to cultural historians.

Let me ask a question you must have heard many times. In rejecting psychoanalysis, aren’t you turning your back on the intricacy of human motivation?

You are right about the frequency of the question. It was most recently raised by several psychoanalysts in the October 1980 issue of Commentary, in protest against the implications of ‘Analysis Terminable’. They said that in challenging their discipline, I was denying the very complexity of the mind – reducing my horizon, in effect, to the level of stimulus and response. Very well, let’s consider that position. Notice, first of all, a confusion here between a subject-matter – the subtleties of human motivation – and specific, highly debatable explanations of that subject-matter. It doesn’t follow that someone who doubts a certain theory is thereby denying the whole domain of evidence that the theory seeks to cover. Furthermore, the plea to retain psychoanalysis because it allegedly does justice to the mind’s complexity begs the question of whether psychoanalysis is conceptually and empirically adequate. For if Freudian ideas are judged to be radically faulty, they surely have to be set aside whether or not some comparably sweeping theory is waiting in the wings. The shortest answer to your question is that, in order to debunk psychoanalysis, I am under no obligation whatever to replace it. That answer, however, is cold comfort to people who want a psychological theory to do double service as a comprehensive world view. In my opinion, it is the need for such a world view – in effect, for a surrogate religion – that makes many adherents of psychoanalysis indifferent to empirically-based critiques.

Speaking of setting things aside, what is your present view of your book on Hawthorne? Many readers, including non-Freudians, have found your psychoanalytic account of Hawthorne convincing. Do you?

If my reading is convincing, that’s because it adduces evidence beyond the usual Freudian range: for example, explicit statements by Hawthorne about hidden motives and the quirks of his own temperament, such as his furtiveness and his preoccupation with ancestral guilt. Much of what I said about Hawthorne’s texts is demonstrably ‘there’. The Freudian method, however, predisposes any investigator to assign certain kinds of evidence to ‘the unconscious’ – even when, as in Hawthorne’s case, it should be apparent that the writer consciously grasped his own themes. That flaw shows up at several points in my argument. Thus, if I were to rewrite The Sins of the Fathers today, I would be more inclined to depict Hawthorne as someone who understood himself remarkably well and who used ingenious means to hint at ‘sensitive’ material without alarming an audience accustomed to hypocrisy and euphemism. The coincidence of emphasis between Hawthorne and Freud remains impressive, but it needn’t be taken as support for the universal applicability of Freud’s theory. The fact is that, in their very different idioms. Hawthorne and Freud were both Faustian Romantics who wanted to break through ‘Victorian’ restraints on an intellectual or imaginative plane while continuing to affirm the pieties of bourgeois life. My book doesn’t deny that parallel, but it does err in treating Freud’s ideas as transhistorical, immutable laws. Actually, I remain rather fond of The Sins of the Fathers, simply as a polemic. I think I helped to rescue Hawthorne from a generation of goody-goody allegorical critics who had turned him into a crashing bore.

Your changed view of psychoanalysis must have lost you some old friends as well as earned you some new ones. What has been the response of students and one-time allies who still subscribe to Freudianism?

The response has varied according to the degree of their remaining commitment to psychoanalysis. Quite a few of my former dissertation advisees, I am happy to say, have abandoned Freud’s system without any prompting from me. I have tried to show others why I no longer consider myself a Freudian. Real dialogue with committed Freudians has been hard to establish, however. One of the less appealing features of psychoanalysis is that it encourages the believer to regard an opponent’s ideas, not as so many arguments to be answered in kind, but as symptoms of neurotic flight from reality. Thus some people with whom I would like to exchange views simply feel sorry for me, as they would for a friend who has been seized by delusions. One even wrote publicly of the ‘developmentally early dynamic’ that must have left me incapable of remaining content with the correct theory of mind. And several of the correspondents in the October Commentary indulged in similar hints about my compulsions. All they required for a jiffy diagnosis, apparently, was the fact that I don’t care for psychoanalysis any more.

To play the devil’s advocate, let me ask if they could conceivably be right about your ‘case’.

It’s always possible. But in reasoning in that style, they commit the fallacy of psychologism, whereby attention is diverted from the substance of a person’s argument to the supposed mental state that led him to propose it. The psychologistic ploy is a means of evading an issue that you can’t successfully address on its merits.

Can you sum up for us how you feel now that ‘Analysis Terminable’ has seen the light of day?

On the one hand, it is disagreeable to have to repudiate my own former ideas and to admit that a decade’s worth of intellectual effort was largely misplaced. On the other hand, I know that I have always been sure of my commitment to rationally-based knowledge, and I think I have demonstrated that commitment in the process of detaching myself from psychoanalysis. This may sound platitudinous to you. In my field of literary studies, however, the most dazzling current fashion is to deny that people can ever get outside their presuppositions. Thus an appeal to independent evidence is considered to be only a ‘reinscribing’ of one’s subjectivity, since evidence itself is thought to be an artifact of the theory it is meant to corroborate. Of course this is an unnecessarily bleak position: the undeniable achievements of modern science rest on the possibility of establishing sufficiently uncontaminatcd grounds to support or cast doubt on a given theory. But by choosing to ignore that fact, literary theorists can promote a dandyish pyrrhonism that reduces all critical discourse to play, without regard for the relative correspondence of statements to discernible realities. In recent published statements, I have tried to resist that trivialisation of literary criticism. I believe that the difference between empirically adequate and inadequate statements ought to count as much in my discipline as in any other. If the main objection to that programme is that people never escape from what they already believe, I can put my own case forward as contrary testimony. I know it is possible to be weaned away from a defective theory by unanswerable objections and a preponderance of evidence. Though the process wasn’t exactly pleasant in my case, it has given me hope that I can go on learning, and not be merely a prisoner of my current mind-set. That hope is what academic life is or ought to be all about.