Revisiting Freud’s “Discrediting”

Frederick Crews has died. Let’s lay his views on psychoanalysis to rest.

By Lily Meyersohn

On June 22, literary critic Frederick Crews died. In a dutiful obituary, the New York Times spent nearly 1,500 words extolling Crews and his legacy and uncritically platforming many of his ideas, including the oft-repeated assertion that psychoanalysis is a pseudoscience. Crews was not the sole author of this myth about psychoanalysis, but he had enormous success in getting others to accept it. So far, he seems to be finding similar success in death, too. The Times obit notably failed to include pushback from any opposing voices from the psychoanalytic community.

Once a lively proponent of psychoanalysis and of psychoanalytic literary theory, Crews devoted most of his professional life to discrediting the field. He began sharing his skepticism in the 1970s, and his criticism grew near-constant by the 1990s. His most recent book, Freud: The Making of an Illusion, published in 2017, was a 746-page biography of the father of psychoanalysis. In the past several years, Crews spent much of his energy falsely equating psychoanalysis with the “recovered memory movement.”

Illustration by Mike Turoff

Though Crews’s death is being used, rather predictably, as an opportunity to recycle his views without questioning, the psychoanalytic community might consider using this moment to do something different: to engage with the Freud bashers, rather than roll over in the face of conflict. As interesting a man as Crews was, many of his critiques of Freud and of psychoanalysis writ large have already been put to bed. It is far past time to let go of the dominant narrative that Freud—and, by extension, all of psychoanalysis—has been discredited.

Psychoanalysis may be experiencing a popular resurgence. But this resurgence is still small in comparison with the need presented by patients. As patients go out into the world—and onto online platforms—and seek out new methods to alleviate their suffering en masse, they should be able to turn confidently toward psychoanalysis as a tried-and-true tool.

Like many young clinicians, I grew up steeped in the notion that Sigmund Freud was, essentially, a fraud. Many of us learned, perhaps osmotically more than anything else, that Freud created a nonscientific discipline that has little to no evidence supporting its methods or theories. This belief is not unique to early-career clinicians. Many established psychoanalysts remain ignorant about all the research undergirding their life’s work. They may know, intuitively, that psychoanalysis works. They may have seen it with their own eyes year in and year out. But the lore is very powerful. Jonathan Shedler, whose 2010 review in The American Psychologist lays out the extensive evidence in support of psychoanalytic psychotherapy, has termed this lore the “master narrative” of psychoanalytic psychotherapy.

In truth, the central theoretical pillars of Freudian psychology—including unconscious emotion and motivation, the importance of childhood experiences in later life, and the value of long-term talk therapy in effecting lasting psychological change—have been well validated and remain essential for understanding and addressing human behavior and suffering. Neuroscience has now shown us how Freudian theories play out in biology. For a long time, “neuro folks wanted to think that the brain is a supercomputer,” says Kevin Meehan, associate director of the PhD program in clinical psychology at Long Island University-Brooklyn. But they, too, have come to the conclusion that we are constantly making complex motivational decisions outside awareness. “People make conflicting decisions” to reduce their suffering, Austen Riggs Center researcher Katie Lewis told me. Psychoanalysis has always recognized that. “That’s where things get interesting.” 

Stephanie Kors, a clinician, assistant professor of psychology at Loyola University Maryland, and lecturer at Harvard Medical School, has designed studies investigating how graduate schools can combat biases against psychoanalytic approaches in the next generation of clinicians. She says she kept hearing people make comments about the definitive discrediting of Freud—and of all the later psychoanalytic and psychodynamic therapies. “Saying ‘Freud’ or ‘psychoanalysis’ is like saying ‘Voldemort,’” Kors told me. “People have a volatile reaction to it.”

Jonathan Shedler goes a bit further, as is his wont. Psychoanalysis has become “the whipping boy” of psychotherapy, Shedler told me on the phone this spring. “Everyone wants to make it the bad guy.”

Neither Kors nor Shedler would be shocked to read the material assigned to me as a first-year masters in social work student in the spring of 2024. Professors take for granted that Freudian psychoanalysis, and the therapies that evolved out of it, are, in the words of an unassuming textbook like Essentials of Lifespan Development, “unpopular,” “not widely accepted,” and impossible to test empirically. Freud, meanwhile, has been presented as a “patriarchal, sexist idiot,” says Nancy McWilliams. But McWilliams sees the psychoanalytic world today as effervescent, and constantly generating new paradigms to reflect advances in research. Elizabeth Danto, a professor of social welfare at Hunter-CUNY, says she is bored of the “old arcane tropes [used] to dismiss Freud.”

Meehan, who teaches a course on the Rorschach test, noted that graduate students are often quite skeptical of it—though there’s been considerable empirical support for its use. “There’s an expectation that it’s all tea leaves,” Meehan said. Even doctoral students, Shedler writes, have “highly inaccurate preconceptions.” Many of these misconceptions come from media, undergrad professors, and textbooks like the one I read this semester, which present what Shedler sees as “caricatures” of Freud’s theories. Other scholars suggest that students have been understandably uncomfortable with some of the complexity presented by psychoanalytic theory. In Kors’s pilot study results, for example, students expressed that psychoanalytic terminology is unnecessarily complicated. “People hate what they don’t understand,” Kors added.

At times, even professors teaching the material—psychoanalysts themselves—believe that the discipline is untested and untestable. Some psychoanalysts object to research on their work, Shedler says, on the grounds that research is either impossible, unnecessary, or irrelevant. Linda Michaels, cofounder of Psychotherapy Action Network, remembers a discussion on the American Psychoanalytic Association listserv about a year ago. A colleague had posted asking a simple question: “Is there any evidence to back up the work we do?” Coming from an experienced psychoanalyst, the question made Michael’s jaw drop.

Freud himself had an inconsistent relationship to empirical research. By 1926, he was arguing that psychoanalysis should not become the “handmaiden of psychiatry,” that is, a strictly medical specialty backed by research findings. On the one hand, Freud promoted case studies as evidence for the discipline, as recounted by the authors of the 2010 book Bringing Psychotherapy Research to Life: Understanding Change through the Work of Leading Clinical Researchers. On the other hand, he criticized the use of statistical procedures to understand clinical work. At one point, the American psychologist Saul Rosenzweig wrote to Freud, announcing that Freudian ideas about repression had been validated in Rosenzweig’s lab. But Freud replied that empirical testing of repression was superfluous; Freud’s own evidence, McWilliams writes, was deemed sufficient.

Freud bears some responsibility, then, for the field’s dismissive attitude toward research. In his book The Psychoanalyst’s Aversion to Proof, Austin Ratner (editor in chief of this magazine), contends that Freud’s own psychological defenses stopped him from pursuing robust proof of psychoanalysis’s efficacy. By projecting inner criticism onto his audience, for instance, Freud gave power away to his critics. According to Ratner, Freud made psychoanalysis only more vulnerable through an “authoritarian ban on experimental research and public disputation.” By “cloistering” the field, psychoanalysis adopted “closed, unscientific habits.”

Reigning Viennese anti-Semitism led to further emotional interference. Though Freud anticipated this anti-Semitism, he did not purposefully counter it. One of the earliest opponents of Freud’s methods, Alfred Hoche, attacked psychoanalysis using anti-Semitic rhetoric. Together, Freud’s internal defenses and the exterior cultural and political context—namely, his outsider status as a Jew in Viennese society, the “scrutiny of his Jewish background and the perception of psychoanalysis as a Jewish science”—meant that his therapeutic methods remained on the fringes in those early years.

“Saying ‘Freud’ or ‘psychoanalysis’ is like saying ‘Voldemort’ … People have a volatile reaction to it.”

It wasn’t until Karl Popper’s 1934 Logik der Forschung, however, that the criticisms about psychoanalysis’s scientific chops really took wing. In Logik, published in English as The Logic of Scientific Discovery in 1959, Popper presented a philosophy of science which argued that scientific theories must be formulated such that they can be tested empirically, which for Popper meant falsifiability. (See Falsifying Karl Popper below.) These critiques called psychoanalysis’s scientific status into question.

Later, Adolf Grünbaum, another philosopher, critiqued psychoanalysis’s reliance on clinical observation for validation. And then there was Crews, who had once been a disciple of psychoanalysis. That’s partly the issue, McWilliams told me. “One of Freud’s problems is that he promoted idealization. When you do that, you create backlash, and people fall off.” Crews was popular for pointing out inconsistencies in Freud’s work, arguing that Freud’s revisions undermined psychoanalysis’s credibility—rather than seeing those fluctuations as representative of the dynamic, iterative nature of good science.

Popper’s words have had an outsize staying power; the concept of unfalsifiability has plagued psychoanalysis. If they cite anyone, those who write off psychoanalysis in popular media or textbooks cite Popper, engaging in a kind of empty, circular reporting; citations trace back to Popper over and over again. Yet Popper’s conclusions on falsification are not universally accepted ways of thinking about the nature of science. And even in Popper’s terms, researchers have conducted various experiments that could have proved psychoanalysis wrong but did not. Unconscious emotion, for example, once regarded as nonfalsifiable, is now regularly tested and confirmed by methods like mindset priming and fMRI.

What if I told you that there is plenty of research on whether, and how exactly, psychoanalysis works?

Psychoanalysis has embraced research in fits and starts over the years. In the 1920s, several psychoanalytic institutes began collecting data on analyses and patient outcomes, trying to answer the question of whether treatment worked. Early clinicians were often doing the research themselves. But as the demands of academia increased throughout the 20th century, academic psychologists had to work much harder. The number of researchers with clinical practices diminished. Researchers, in turn, “stopped asking whether their work was generalizable to the real world,” Michaels laments. By the 1940s, research was effectively discontinued by the institutes. But none of the experts I spoke to fully understand why that was.

Proponents of psychoanalysis eventually began sounding the alarm: According to Ratner, Edward Glover warned his colleagues that they were neglecting their research responsibilities in 1952. This was the heyday of psychoanalytic practice—the decades of dominance that psychoanalysis enjoyed in the medical establishment and in popular culture. For a few decades, at least, to head a psychiatry department, one often had to be a psychoanalyst.

Being at the top of that kind of hierarchy, McWilliams noted to me, means that you are expected to participate in a set of social rituals. The psychoanalytic establishment shied away from having their beliefs examined from the outside. As McWilliams notes in her 2004 book, this resistance may have emanated from “the complacency that comes along with being an elite.”

There were costs to this position. In 1940, as the war broke out, the American Psychoanalytic Association enacted a resolution it had proposed to the International Psychoanalytic Association in 1938. The “1938 Rule,” as it was known, afforded APsA exclusive rights for training in the US—“a right which was restricted practically to physicians,” said Otto Kernberg in 1996—excluding involvement of nonphysician analysts from practice and training in the States. Redefining psychoanalysis as a highly technical procedure that only physicians could carry out meant that psychoanalysis was limited to medical schools and institutes and excluded lay researchers. The atmosphere that took root in training centers, argues McWilliams, was “somewhat cult-like.”

Those decades did real damage to the research landscape. The ban on non-MDs served to intensify the culture of intellectual isolation that Freud had birthed. Certainly, psychoanalysis continued to be a dynamic, exciting field—but it developed outside academia, in freestanding, insular, often-inaccessible institutes, wrote Shedler in 2006. Though physicians are immersed in science during their training, and might ostensibly value research, psychoanalytic leaders—including A. A. Brill, who considered himself the father of American psychoanalysis—were staunchly opposed to the practice of psychoanalysis by non-MDs. Brill worked to promote psychoanalysis as a subspecialty of psychiatry. Brill and others compensated for psychoanalysis’s aversion to research, insisting that the field was scientific—after all, American psychoanalysts were all required to be MDs. This gave the profession a kind of scientific veneer.

Academic psychology, meanwhile, did little to correct misconceptions about psychoanalysis. The introduction of the PsyD degree in the 1970s potentially worsened the issue, creating two separate degrees granted by different institutions and shrinking spaces of intellectual overlap and connection: PsyDs who were trained to practice psychotherapy and PhDs who were often trained to do psychological research without clinical training.

Until 1988, analytic institutes continued to be composed primarily of psychiatrists. In 1986, Bryant Welch and three coplaintiffs sued the American Psychoanalytic Association for denying psychologists the right to practice psychoanalysis. Arguing that the rule constituted restraint of free trade, the plaintiffs won a settlement worth hundreds of thousands of dollars, which effectively mandated that psychoanalytic institutes welcome nonmedical practitioners.

Research did pick up. Even in the late 1950s and early 1960s, changes to the profession—more conferences, the foundation of multiple new organizations, new journals, and book publications—aided in a small renaissance. The National Institute of Mental Health also began shaping research. As Christopher Muran, dean of the school of psychology at Adelphi University, details in his 2010 “A Brief Introduction to Psychotherapy Research,” a critical intellectual turn meant that by the 1970s, researchers began conducting analyses of psychoanalytically oriented variables like the working alliance, rupture and repair, empathy, positive regard, and countertransference.

In 1999, Lew Aron, then director of NYU’s postdoctoral program in psychotherapy and psychoanalysis, convened a panel on the role and importance of empirical data for psychoanalytic work. Muran told me how he was approached by Aron, who wanted to “bust down some of these walls and bring the empirical into play.” They created a task force to make recommendations for NYU’s program, met regularly for a year, wrote a proposal, and brought it to the program’s governing body. But the proposal was shot down. None of the recommendations were picked up, and the panel’s findings were not followed up on again for several years.

A quarter of a century later, the state of contemporary psychoanalytic research is still mixed. In many European countries, research benefits from state funding and support. Even in the US, the appetite for psychoanalytic research is growing. We may never see another heyday of psychoanalysis as in the 50s and 60s—but that might not be such a bad thing. Muran, for one, says he is seeing more of an appetite for pluralism that is “inclusive of the empirical perspective.” The somewhat arbitrary lines between the psychoanalytic and the cognitive-behavioral worlds have become more porous, he notes, especially as behavioral psychology begins running into its own limits.

The psychoanalytic research we have tells a compelling story. Results consistently show that the benefits of psychodynamic therapy are dose-dependent, meaning they increase with time—including after termination of the treatment. In “In Evidence: Measurement-based care can guide clinical practice in psychoanalysis,” published in this magazine last year, researchers from the Austen Riggs Center conclude that psychodynamic and psychoanalytic treatment is “effective and clinically useful for treating many disorders, including complex psychiatric problems and severe mental illnesses.” Since the 1960s, the researchers write, more than 300 randomized control trials (RCTs) have found that psychodynamic treatment is equal or superior to other “evidence-based” treatments.

But this body of research has not been effectively translated or advertised to the public. In contrast, cognitive-behavioral therapies have done a beautiful job marketing their product. Organizations like the Association for Cognitive and Behavioral Therapies have staff dedicated to getting messages about CBT into the media, and they set strategic goals around funding for research, since research is the juggernaut that fuels academic careers. “We’re facing the consequences of a strong marketing effort by CBT,” Linda Michaels says—but a “weak effort by psychoanalysis—even an aversion to such things.”

Psychoanalysts may be loath to admit it, but there may be something they can learn from CBT.

Developed by psychologists like Aaron T. Beck and Albert Ellis, who were trained psychoanalytically, CBT and its derivative therapies employ various psychoanalytic concepts, though they label and describe these concepts differently. In CBT, some scholars see the translation of psychoanalysis from continental Europe to the United States: an appeal to the American consumer who was discontented with the analyst’s authority and orthodoxy and craved quicker solutions to their woes.

Studies have found that among clinical psychology doctoral students, only five to ten percent identify as “psychodynamic,” yet most are using multiple psychoanalytic techniques, simply under different names. The psychoanalytic field gets hung up on this “narcissistic injury,” Lewis says. What’s more important is that when concepts get coopted, not all of their therapeutic meaning gets transferred.

CBT has been remarkably well-suited to American academia. Unlike psychoanalysis, CBT could be studied quickly. The shortened time frame translated easily into RCTs, inherited from the pharmaceutical industry. The time frame also paved the way for researchers to get tenure: conduct a study, publish, rinse-and-repeat. Those tenured researchers were then the ones to educate graduate students and therapists of the next generation. They became the leaders of the APA and provided information to lawmakers in DC that informed mental health policy.

CBT research accepted the paradigms of pharmaceutical companies: its prioritization of quantitative benefits, of efficacy coefficients, of observable symptoms. The structure of RCTs dictated that research participants have “pure,” not comorbid, psychological disorders—a rarity in reality. In her presentations on behalf of the Psychotherapy Action Network, Michaels argues that what really counts for patients and their clinicians is too often excluded from results: challenging personality issues, insecure attachment, difficult social and emotional styles. These are rigged rules, rigged paradigms, writes Shedler. Psychoanalytic research is being done with our hands tied behind our backs, he protests.

The focus on rapid symptom reduction is particularly short-sighted. Research that focuses exclusively on symptom reduction tells us little about how people are functioning in their daily lives, says Meehan. Do patients have relationships that provide intimacy and connection? Do they have work they find gratifying?

It matters what kinds of therapy get studied. We already know that psychoanalysis works, but further psychoanalytic research can guide and improve care and continue to build trust in psychoanalytic ideas and practices. As we aim to narrow the chasm between clinicians and researchers, researchers must deliver data that is practitioner-friendly. Practitioners rely on the research to tell us what direction we need to move in. A lack of understanding and effective advertising of the research that has been done has put a generation of therapists in a more vulnerable spot, positioning us to have less of a voice in decisions about mental health research and funding.

For all of CBT’s entrenchment in research and public opinion, the evidence about it is deeply troubling. A shocking 40 to 70 percent of CBT participants relapse to major depression within 12 months, and the majority of patients are left to seek treatment again. Studies show that brief, manualized treatments like CBT are ineffective for most depressed patients most of the time. In recent years, researchers have downgraded the efficacy of CBT. But that change has not been widely publicized. It is, in part, psychoanalysts’ responsibility to let the public know that CBT is not effective as a standalone treatment. Effective therapy clearly utilizes a relationship, not just a technique. Cognitive-behavioral researchers will likely not be shouting this from the rooftops.

---

Lily Meyersohn is a freelance writer and a student of social work at CUNY-Hunter College.

---

Falsifying Karl Popper

By TAP editors

How do you know you’re doing science? If your theory has consequences that could potentially prove it wrong. So goes Karl Popper’s theory of falsificationism, beloved of textbook editors who need a cursory introduction to scientific method. It’s an intuitive idea that captures something important about science. The problem for psychoanalysis is that Popper famously—and wrongly—chose Freud’s theory as one of his pet examples of unfalsifiable pseudoscience.

The idea that falsifiability is the demarcation criterion for science—separating science from other forms of human endeavor such as myth-making, poetry, and senseless blabber—might seem like a deep feature of modern thinking stemming from the Enlightenment. In reality, Popper introduced it in 1934 in his book Logik der Forschung. The going view at the time was that verifiability is what makes a theory scientific, but Popper didn’t think verifiability was a strict enough criterion to rule out all kinds of nonscientific ideas.

Popper held that psychoanalysis is unfalsifiable because a given piece of human behavior can always be interpreted to accommodate the theory. He made similar complaints about Marx’s theory of history and Darwin’s theory of natural selection: These theories consist of untestable “just so” stories about why something—a behavior or historical event or organic trait—had come about. They explain things too easily. While Popper eventually retracted this complaint about Darwin, admitting that aspects of natural selection are amenable to testing, he never made a similar retraction for psychoanalysis.

Popper was wrong. Psychoanalytic theory does generate falsifiable hypotheses, as Freud recognized whenever he altered his theory based on clinical observations. The theory that childhood sexual abuse is necessary for hysteria cannot withstand the observation of hysteria without childhood sexual abuse, for instance.

The question is not whether some analysts provide overconfident, dogmatic interpretations—for instance, claiming that whenever the patient disagrees with an interpretation, this must be a reaction-formation that in fact proves the interpretation. Bad analysts don’t make psychoanalysis unscientific any more than bad radiologists make radiology unscientific. (Moreover, reaction-formation may sometimes be correctly attributed to the patient, with evidence.) The question is whether key psychoanalytic hypotheses are falsifiable in principle and whether they can be used by adept clinicians and researchers to pursue therapeutic goals such as insight, growth, and the alleviation of suffering.

---

Further Reading

Ackerman, S., and K. Lewis. “In Evidence: Measurement-Based Care Can Guide Clinical Practice in Psychoanalysis.” The American Psychoanalyst 57, no. 2 (2023).

Brandell, J. R. “Psychoanalysis in the Halls of Social Work Academe: Can This Patient Be Saved?” Journal of Social Work Practice 27 no. 3 (2013), 235–248.

Danto, E. Freud’s Free Clinics: Psychoanalysis and Social Justice, 1918–1938. Columbia University Press, 2005.

Driessen, et al. “The Efficacy of Cognitive-Behavioral Therapy and Psychodynamic Therapy in the Outpatient Treatment of Major Depression: A Randomized Clinical Trial.” American Journal of Psychiatry 170, no. 9 (2013): 1041–1050.

Herbert, G. L., V. McCormack, and J. L. Callahan. “An Investigation of the Object Relations Theory of Depression.” Psychoanalytic Psychology 27, no. 2 (2010), 219–234.

Kernberg, O. Interview held by Sergio Benvenuto and Raffaele Siniscalco, July 1996, for the Multimedia Encyclopaedia of Philosophical Sciences by RAI (Italian Radio Television).

Levendosky, A. A., J. E. Turchan, X. Luo, and E. Good. “A Re-Introduction of the Psychodynamic Approach to the Standard Clinical Psychology Curriculum.” Journal of Clinical Psychology 79, no. 10 (2023): 2439–2451.

Levy, K. N., and T. Anderson. “Is Clinical Psychology Doctoral Training Becoming Less Intellectually Diverse? And If So, What Can Be Done?” Clinical Psychology 20, no. 2 (2013), 211–220.

Muran, J. C., L. G. Castonguay, and B. Strauss. “A Brief Introduction to Psychotherapy Research.” In Bringing Psychotherapy Research to Life: Understanding Change through the Work of Leading Clinical Researchers, edited by L. G. Castonguay et al., 3–13. American Psychological Association, 2010.

McWilliams, N. Psychoanalytic Psychotherapy: A Practitioner's Guide. The Guilford Press, 2004.

Richards, A. D., and P. W. Mosher. “Abraham Arden Brill, 1874–1948. “American Journal of Psychiatry, published online 2006. DOI: https://doi.org/10.1176/appi.ajp.163.3.386.

Shedler, J. (2006). “That Was Then, This Is Now: Psychoanalytic Psychotherapy for the Rest of Us.” Retrieved from http://jonathanshedler.com/writings/.

Shedler, J. “A New Language for Psychoanalytic Diagnosis.” Journal of the American Psychoanalytic Association 50, no. 2 (2002): 429–56.

Shedler, J. “The Efficacy of Psychodynamic Psychotherapy.” The American Psychologist 65, no. 2 (Feb-Mar 2010): 98–109.

Shedler, J. “Where Is the Evidence for ‘Evidence-Based’ Therapy?” Psychiatric Clinics of North America 41, no. 2 (2018): 319–329.

Tummala-Narra, P. Psychoanalytic Theory and Cultural Competence in Psychotherapy. American Psychological Association, 2016.

Westen, D., C. M. Novotny, and H. Thompson-Brenner. “The Empirical Status of Empirically Supported Psychotherapies: Assumptions, Findings, and Reporting in Controlled Clinical Trials.” Psychological Bulletin 130, no. 4 (2004): 631–663.

Wood, J., M. T. Nezworski, and H. N. Garb. “What’s Right with the Rorschach?” The Scientific Review of Mental Health Practice 2, no. 2 (Fall/Winter 2003): 142–146.

Yoder, S. “Lay Versus Medical Analysis: A Conflict Centered on the Disciplinary and Professional Identity of Psychoanalysis.” Honors Thesis, Portland State University, 2012. DOI: https://doi.org/10.15760/honors.11.