Conversations in Critical Psychiatry: Allen Frances, MD

May 23, 2019

A series aimed at engaging prominent individuals who have made meaningful criticisms of psychiatry and have offered constructive alternative perspectives. First interview: Allen Frances, MD, Chair of the DSM-IV Task Force.

Conversations in Critical Psychiatry is an interview series aimed to engage prominent individuals who have made meaningful criticisms of psychiatry and have offered constructive alternative perspectives to the current status quo. It is Dr Aftab’s hope that these discussions will stimulate a much-needed debate in the psychiatric community.


Dr Aftab is a geriatric psychiatry fellow at University of California San Diego in La Jolla, CA, and has been actively involved in initiatives to educate psychiatrists and trainees on the intersection of philosophy and psychiatry. He is also a member of the Psychiatric Times Advisory Board. He can be reached at

Allen Frances, MD, is Professor Emeritus and former Chair, Department of Psychiatry, Duke University. He was Chair of the DSM-IV Task Force and is the author of Saving Normaland Essentials of Psychiatric Diagnosis.

I have always been intrigued by Dr Allen Frances’ views on psychiatric diagnosis. I was in medical school when DSM 5 was in development, and it was fascinating to see his relentless commentary on issues related to diagnostic inflation and diagnostic validity as an inside critic. His critique also served as a portal for me (and many others) to explore larger philosophical issues in psychiatric diagnosis. This made him an ideal candidate to converse with for the launch of this interview series on critical psychiatry.

Awais Aftab, MD: In many ways, you are one of the architects of modern psychiatry, yet you have also emerged as one of its most prominent critics. How do you see your own relationship with the field?

Allen Frances, MD: I think psychiatry is among the noblest of professions, but I think that it has drifted astray from best practice. It is heartbreaking to me that 600,000 of our most severely ill patients are either in jail or homeless and that we have done so little to advocate for the community mental health centers and affordable housing that would have freed them from confinement and ended the shameless neglect.

I fear that too many psychiatrists are now reduced to pill pushing, with far too little time to really know their patients well and to apply the rounded biopsychosocial model that is absolutely essential to good care. We also have done far too little to educate the primary care doctors who prescribe 80% of psychiatric meds on the principles of cautious prescribing, proper indications, full consideration of risks, and the value of watchful waiting and tincture of time.

 I despair the diagnostic inflation that results from a too loose diagnostic system, aggressive drug company marketing, careless assessment, and insurance company pressure to rush to judgement. Diagnoses should be written in pencil, and under-diagnosis is almost always safer and more accurate than over-diagnosis. And, finally, I object to the National Institute of Mental Health (NIMH) research agenda that is narrowly brain reductionistic; it has achieved great intellectual masterpieces, but so far has not yet helped a single patient. So, in sum, I have loved being a psychiatrist, but wish we were better organized to end psychiatric suffering.

Awais Aftab, MD: When it comes to a philosophical understanding of mental disorders, you have eloquently expressed on several occasions that you neither see them as real diseases nor as fanciful myths, but rather something in between, as useful constructs that represent our current best guess on how to sort psychiatric distress. I am reminded a bit of the argument by Thomas Szasz that there are only two possible realities behind mental health conditions: They are either brain diseases or they are problems in living, and there is no in-between. Do you accept this dichotomy between diseases and problems of living? And, if so, by calling mental disorders as constructs, are we admitting that we don't really know which conditions are diseases and which are problems in living?

Allen Frances, MD: I like to refer questions on the philosophy of diagnosis to the three umpires:

Umpire 1: There are balls and strikes, and I call them as they are.

Umpire 2: There are balls and strikes, and I call them as I see them.

Umpire 3: There are no balls and no strikes until I call them.

Robert Spitzer, MD1 was an Umpire 1 and, until recently, so were most biological psychiatrists. The credibility of this model has been destroyed as we have learned more about the unfathomable complexity of the human brain and the complete failure of genetics and neuroscience to provide useful answers about what causes psychiatric problems.

Szasz could argue the Umpire 3 position only because he strenuously and successfully avoided ever seeing severely ill patients. Schizophrenia and other severe psychiatric disorders clearly have a strong biological contribution and cause a qualitatively, as well as quantitatively, different level of suffering far beyond the everyday problems of living.

I, and I think most people now, humbly aspire only to be Umpire 2, doing our best to define mental disorders in ways that are most currently useful, without any pretention that this is the only way or that current constructs will withstand the test of time.

Sharp disagreements do remain on how inclusive the diagnostic system should be. I am most fearful of the risks of over-diagnosis and have argued strongly for a narrower system with higher diagnostic thresholds. Experts in each area always want to expand their pet diagnoses and worry much more about missed patients than mislabeled patients. We contained the experts in preparing DSM IV; they were given a free rein in DSM 5, leading to a much more inclusive system.

Awais Aftab, MD: A lot of people feel that something went awry between the creation of Spitzer's Research Diagnostic Criteria (RDC) and the creation of DSM III. The two dozen psychiatric diagnoses in the RDC were well supported by a body of research, but DSM III created diagnostic criteria out of thin air (based on consensus alone) for so many more disorders.

What if psychiatry had continued to expand the research diagnostic criteria patiently and iteratively through research studies instead of relying on DSM in this fashion? Might psychiatry be in a better shape now?

Allen Frances, MD: The narrow research focus of the RDC made it a completely inadequate guide to clinical practice. Going overboard in the other direction, the diagnostic exuberance of DSM 5 confuses mental disorder with the everyday sadness, anxiety, grief, disappointments, and stress responses that are an inescapable part of the human condition. DSM 5 ambitiously mislabels normal diversity and childhood immaturity as disorder, creating stigma and promoting the excess use of medications. We need a Goldilocks-just right balance between the risks of missing patients and of mislabeling them. At the moment, mislabeling rules.

Awais Aftab, MD: You have expressed that the DSM approach has been far too influential, leading to a reification of diagnostic constructs and a checklist approach to diagnosis that the creators of DSM had never intended. Why do we still need the DSM? Most other fields of medicine don’t rely on an official manual in this fashion.

Allen Frances, MD: DSM is a system of limited, but essential, value. I don't trust clinicians who know only DSM, but I equally don't trust clinicians who don't know it at all or use it carelessly. The DSM evaluation should be just a small part of an initial interview and not done as a rote checklist. Eliciting symptoms in a natural way expresses empathy and understanding for what the patient is experiencing and can be a giant step toward a strong therapeutic relationship. DSM also plays a central role in differential diagnosis (particularly in ruling out medical, medication, and substance causes for symptoms), in differential selection of treatments, and in predicting course and prognosis. Reductionistic systems are useless or harmful in clinical practice;  see for example NIMH’s Research Domain Criteria2 and “Understanding Psychosis and Schizophrenia” by the British Psychological Society.3

Awais Aftab, MD: You recognize the DSM as an imperfect document, yet you also advocate for diagnostic conservatism-that current diagnostic constructs should be maintained unless there is convincing meta-analytic evidence to support a change. When groups of researchers come up with proposals for diagnostic criteria for psychiatric conditions and their supporting evidence falls short of meta-analytic evidence, those proposals are not given much legitimacy,  even though the amount of evidence would likely still be far more evidence than Robert Spitzer ever had for DSM III. Does that not seem like an unscientific state of affairs?

Allen Frances, MD: With all its flaws and lack of empirical base, I think the field would have been better off sticking with DSM III. Changes since that edition have been consistently exploited to increase diagnostic inflation. If anything in DSM can be misused, it will be misused. We had excellent empirical evidence for including both bipolar II and Asperger’s in DSM IV, but both did more harm than good. Data drawn from research studies on highly selected patients in the hothouse environment of a university research clinic generalize very poorly to the hustle and bustle of primary care.

Awais Aftab, MD: What are your thoughts on the role and relevance of Freud and his ideas in contemporary psychiatry?

Allen Frances, MD: Freud is punished now for being unduly worshipped during his heyday 100 years ago; he gets far too little credit for presciently anticipating much of modern cognitive theory, neuroscience, and psychotherapy technique. Freud's emphasis on the power of instinct and unconscious mental functioning successfully applied Darwinian psychology to a wide variety of clinical problems. But psychoanalysis was too important to be left in the hands of the antiquated psychoanalytic institutes that adhere religiously to old ideas that Freud himself would have surely abandoned as modern science made them obsolete and quaintly silly. It is tragic that many residents now get so little psychotherapy training; it explains why some psychiatrists become mindless pill-pushers.

Awais Aftab, MD: Is it appropriate in your view that one single organization, the American Psychiatric Association, has total control over the DSM? Some would say that the APA uses the book to solidify its authority and for the generation of financial revenue. It seems odd for a manual of such importance to be tied to organizational professional politics.

Allen Frances, MD: I have worked with a group studying how guidelines are developed throughout medicine. We recommend that specialty groups never be permitted sole power to determine the diagnostic guidelines for that specialty. There is an inherent financial, intellectual, and emotional conflict of interest that leads every specialty to recommend over-diagnosis. Diagnostic guideline development needs specialists as consultants, but they should never be allowed to call the shots. Contributions from primary care, public health, health economics, and consumers are also important, and methodologists without specialty affiliation should do the review and evaluation of evidence. Moreover, APA has a special conflict of interest because the DSMs are such a valuable publishing property-essential for meeting its budget. This makes frequent revision too tempting and results in an unseemly hyping of the product.

Awais Aftab, MD: You have acknowledged that conditions have become mental disorders by “accretion and practical necessity” and that mental disorder is “what clinicians treat and researchers research and educators teach and insurance companies pay for.” Does that not imply an unsettling relativism in the shape psychiatry has taken? I am reminded of something Robert Kendell, FRCPsych, wrote in 1975: “The fact is that any definition of disease which boils down to ‘what people complain of’, or ‘what doctors treat’, or some combination of the two, is almost worse than no definition at all. It is free to expand or contract with changes in social attitudes and therapeutic optimism and is at the mercy of idiosyncratic decisions by doctors or patients.”2

Allen Frances, MD: Bob Kendell was the clearest voice in the history of psychiatric diagnosis, and his caution about diagnostic relativism is even more cogent and widely applicable today than when he wrote 45 years ago. Diagnostic inflation exists not just in psychiatry, but in every medical and surgical specialty. The definitions for just about every disease-hypertension, osteoporosis, diabetes, glaucoma, knee, shoulder and back disease, prostate, breast, and thyroid cancers-are far too loose and ever getting looser. Half of adults in the US now have hypertension. Modern medicine is making such rapid advances, soon none of us will be well. If I were in control of psychiatric diagnosis now, I would recommend a reduction in diagnostic inflation through a risk/benefit analysis. We need an evaluation of which diagnoses and what diagnostic thresholds do more harm than good.

Awais Aftab, MD: You have talked about epidemiological studies exaggerating rates of mental disorders because they also include psychiatric symptoms lacking clinical significance in their prevalence estimates. DSM relies heavily on “clinical significance” as a necessary criterion, but the concept is never formally defined in the manual. How do you understand “clinical significance”?

Allen Frances, MD: Never believe the extremely high rates of mental disorders routinely reported by epidemiological studies in psychiatry-usually labelling about 25% of the general population as mentally ill in the past year, about 50% lifetime. This entire literature has a systematic, but unacknowledged, methodological bias that inherently results in over-reporting. Because epidemiology requires such huge samples-in the tens of thousands-it is prohibitively expensive to conduct clinical interviews. Instead phone surveys are done by non-clinicians following a highly structured format that allows no clinical judgment whether the symptoms reported cause sufficient clinically significant distress and impairment to qualify as a mental disorder. Since there is no sharp boundary between normal distress and mental disorder, not assessing for clinical significance includes among those labelled mentally ill many who are merely distressed. The rates reported in studies are really only upper limits, not accurate approximations of true rates. They should be, but never are, reported as such.

"Clinical significance" is an indefinable but essential construct in applying DSM criteria-so important that we repeated its necessity in each and every criteria set despite the fact that we could not operationalize its application. Many of the most essential terms that we use so glibly in everyday life and practice are equally undefinable, eg, mental disorder, disease, illness, impairment, and dysfunction. Words that are categorical lose precision in defining phenomena that are distributed dimensionally. DSM should be seen only as a tool helpful in guiding clinical judgment, not a replacement for it.

Awais Aftab, MD: Are there books that have profoundly influenced how you understand psychiatry that you would like to recommend to other psychiatrists and trainees?

Allen Frances, MD:

  • Ulyssesby James Joyce for the fullest literary description of how people think.

  • The Brothers Karamazov by Fyodor Dostoevsky for the deepest understanding of human motivation and best takedown of expert witness testimony in forensic psychiatry.

  • The Peloponnesian War by Thucydides for its profound insight that interacting complex contingencies make predicting the future impossible no matter how well we know the past.

  • On the Nature of Things by Lucretius for its brilliant scientific intuitions and calm acceptance of the small place we humans occupy in this universe.

Also watch lots of movies, travel, and read widely. See lots of patients, and imagine yourself in their life, feelings, and thoughts. The broader and deeper you are as a person, the better you will be as a psychiatrist. The easiest, and most mindless, part of psychiatry is prescribing meds; be good at it, but not limited by it.

“There's no more terrible pain a man can endure than to see clearly and be able to do nothing.” -Herodotus, The Histories

Awais Aftab, MD: Thank you!

Editor's note: The opinions expressed in the interviews are those of the participants and do not necessarily reflect the opinions of Psychiatric Times.


Dr Aftab and Dr Frances report no relevant financial disclosures.


1. Robert Spitzer, MD was the Chair of the DSM-III Task Force.
2. NIMH. Research Domain Criteria (RDoC). Accessed May 15, 2019.
3. British Psychological Society. Understanding psychosis and schizophrenia. Accessed May 15, 2019.
4. Kendell RE. The concept of disease and its implications for psychiatry. Br J Psychiatry. 1975;127:305-315.

download issueDownload Issue : Psychiatric Times Vol 36, Issue 10