What Is Meant by a Psychiatric Diagnosis?

Psychiatric diagnoses are not merely descriptive; they reflect genuine illness.

Posted Jun 07, 2020

This article is co-authored with Ronald W. Pies, M.D., clinical professor of psychiatry at Tufts University Medical School and professor emeritus of psychiatry at the State University of New York Upstate Medical University.

It has become fashionable for some in the social sciences to assert that psychiatric diagnoses represent "constructs" and not genuine disorders or diseases. During a recent Twitter exchange, one of us (Mark Ruffalo) was pointed to an article published here on Psychology Today in 2019 by the psychoanalytic psychologist Jonathan Shedler, Ph.D., titled, "A Psychiatric Diagnosis Is Not a Disease."

We wish here to counter the claims made in the Shedler piece, particularly as they pertain to the meaning and implication of a psychiatric diagnosis. One of us (Ronald Pies), a psychiatrist, has spent a large part of his career thinking and writing about the philosophical foundations of psychiatry; and the other (Mark Ruffalo), a psychoanalytic psychotherapist, has developed a keen interest in discussions surrounding the meaning of psychiatric diagnosis.

It is worth noting that answering the question "What is a psychiatric diagnosis?" is not merely a pedantic philosophical exercise; it carries with it great weight and consequence for the treatment of the many people living with psychiatric illness.

Dr. Shedler's Claims

In his 2019 article, Shedler writes, 

Psychiatric diagnoses are categorically different [from medical diagnoses] because they are merely descriptive, not explanatory. They sound like medical diseases, especially with the ominously-appended disorder, but they aren't. If we speak of generalized anxiety disorder and major depressive disorder as if they are equivalent to pneumonia or diabetes, we are committing a category error.

Shedler (2019) also states,

Confusion arises because medical diagnoses point to etiology—underlying biological causes.

He says, of psychiatric diagnoses:

They are agreed-upon labels—a kind of shorthand—for describing symptoms. [For example,] generalized anxiety disorder means a person has been anxious or worried for six months or longer and it's bad enough to cause problems—nothing else [italics added].

Thus, explicitly or implicitly, Shedler is making at least four claims: 1. That medical diagnoses "describe underlying biological causes" and psychiatric diagnoses do not; 2. That psychiatric disorders and medical disorders are categorically different (i.e., belong in different categories); 3. That psychiatric diagnoses provide nothing more than a label or a description of the person's symptoms; 4. That psychiatric disorders and their diagnostic criteria do not reflect or identify the cause or etiology of the patient's problems ("psychiatric diagnoses are…not explanatory.")

In the statements above, Shedler effectively aligns himself with the late psychiatrist Thomas Szasz who often used the term "category error" to describe what he saw as the misattribution of the term "disease" (or disease-like properties) to what he called, "problems in living." Szasz also argued that medical diagnoses point to causes, whereas psychiatric diagnoses are merely labels (see, for example, Szasz, 1987).

In so far as he agrees with Szasz, Shedler is endorsing what psychiatrist Awais Aftab (2020), in his own blog on the topic, recently called an "essentialistic approach" to the disease concept; i.e., the idea that a putative "disease" entity must entail demonstrable biological dysfunction or identify an underlying (biological) cause, to be called a disease.

Claim #1: Medical diagnoses describe "underlying biological causes"—and psychiatric diagnoses do not.

A careful reading of history teaches us that there is no "essential" definition of disease universally accepted by physicians (or by philosophers of science); however, historically, the concept of "disease" has always been more intimately tied to the degree of suffering and incapacity experienced by the individual person than to demonstrable biological dysfunction (see Pies, 1979, 2019). While abnormal biological or laboratory findings can sometimes aid in the diagnosis of a disease—e.g., as confirmatory tests—they are neither necessary nor sufficient for an entity to be considered a disease, nor for the diagnosis of disease.

For instance, a person may have an abnormally shaped ear lobe or an unusually high serum albumin level and not be "diseased" or "ill" in any clinically relevant sense. Similarly, an abnormal electrocardiogram (EKG) does not necessarily point to heart disease, nor does a normal EKG rule out serious heart disease. And history is replete with examples of disease states whose pathophysiological mechanisms were unknown for decades after the disease had first been described clinically. Parkinson's disease is perhaps the best-known example.

On Dr. Shedler's view, no physician in 1817—not even James Parkinson!—could have told a patient with "the shaking palsy" that he or she had bona fide disease—because, for Shedler, the sine qua non of a bona fide disease diagnosis requires a known etiology—or at least, the diagnosis must "point to" etiology, whatever that means. The illogical consequence of this view is that no patient with what we now recognize as Parkinson's disease could have had actual disease until the etiology or pathophysiology was identified, in the 1960s.

Even today, many diseases—readily identified as such—have no known underlying biological cause. Alzheimer's disease, migraine disorders, Kawasaki's disease, fibromyalgia, and amyotrophic lateral sclerosis (Lou Gehrig's disease) are but a few examples of conditions for which the exact cause is either unknown or poorly understood. 

Thus, the claim that medical diagnoses essentially or necessarily "point to" etiologies is false. Historically, disease has been conceptualized as a problem or entity that causes prolonged or severe distress and impairment—or suffering and incapacity—in the individual, living human being, regardless of whether there is a known cause or etiology. As psychiatrist R.E. Kendell (1975) has observed,

Historically, it seems likely that the concept of disease originated as an explanation for the onset of suffering and incapacity in the absence of obvious injury, and that the concept of health was a later development, implying the absence of disease (p. 307).

Claim #2: Psychiatric disorders and medical disorders are categorically different (i.e., are not "equivalent").

This claim is a red herring, because psychiatrists do not assert—nor does the DSM-IV or DSM-5—that generalized anxiety disorder or major depressive disorder are "equivalent" to pneumonia or diabetes, any more than an internist would assert that pneumonia is "equivalent" to diabetes. Psychiatrists (and many philosophers of psychiatry) recognize that different disease entities have, among them, very substantial differences; e.g., pneumonia involves histological alterations of lung tissue, whereas diabetes primarily involves dysregulation of glucose metabolism, and temporal lobe epilepsy involves subtle alterations of consciousness. But all three conditions fall under the category of "disease," owing to their tendency to produce certain kinds of suffering/distress and incapacity/dysfunction. This is their "family resemblance," in Ludwig Wittgenstein's terms.

Similarly, serious mental illnesses (SMI) like schizophrenia, bipolar disorder, and major depression are rightly grouped in the broad family of disease entities not because they are "equivalent" to diabetes or pneumonia, but because—as Wittgenstein explained in his later work (e.g., Philosophical Investigations)—there are "family resemblances" between SMI and certain conventionally "medical" disease states; i.e., like diabetes or pneumonia, schizophrenia (or major depression, bipolar disorder, autism, and others) produce characteristic types of suffering and incapacity. (It turns out that many serious mental disorders also have numerous biological risk factors and neurobiological correlates, but this feature is neither necessary nor sufficient to impute the term "disease" to these conditions).

Claim #3: Psychiatric diagnoses provide nothing more than a label or a description of the person's problems.

This claim is based on the fallacy that psychiatric diagnoses are made solely on the basis of the patient's symptoms (subjective complaints). In reality, they also include signs (observable features) such as psychomotor agitation, weight loss, abnormal sleep patterns, cognitive impairment (as demonstrated with psychometric testing) and other objectively observable phenomena. Thus, insofar as this claim alleges that psychiatric diagnoses describe only symptoms, it is false.

Furthermore, psychiatric diagnoses provide more than a shorthand or label. Many also have predictive validity, genetic risk factors, neurobiological correlates, and psychometric characteristics that may be objectively demonstrated. For example, DSM-5 states that one-third of the risk of developing generalized anxiety disorder is genetic, and that these genetic factors overlap with the risk of "neuroticism." Thus, Shedler's claim that a diagnosis of generalized anxiety disorder means "nothing else" than a description of the patient's prolonged worry or anxiety is false.

Claim #4: Psychiatric disorders and their diagnostic criteria cannot be considered the "cause" of the patient's problems.

Whereas Shedler (2019) claims that psychiatric diagnoses do not "cause" symptoms, we contend that it is completely appropriate and correct to state, for instance, that a patient's extreme mood swings are caused by their underlying bipolar disorder; or that a patient's auditory hallucinations, paranoid delusions, and thought process disorder are caused by their having schizophrenia.

Shedler appears to take the stance that (1) psychiatric disorders and their diagnostic criteria do not speak to the cause or etiology of the particular condition (which is generally true); and therefore, (2) (a diagnosis of) condition X cannot be considered a cause of the patient's problem.

This is a non sequitur, because we are really talking about two distinct kinds of causality: let's call them "cause 1" and "cause 2." That we don't know the "cause 1" of, say, schizophrenia does not mean that schizophrenia (by current criteria) is not the "cause 2" of the patient's hallucinations, delusions, etc. To put it another way, we may not know the cause of schizophrenia, but we can still say, quite accurately, that schizophrenia is the cause of the patient's suffering and incapacity. In his 2019 piece, Shedler confuses and conflates these two very different types of causality.


A historical and philosophical investigation of the disease concept reveals that psychiatric disorders—and their referent diagnoses—meet historically-based and well-accepted criteria for classification as disease. It is Shedler, in fact, who makes a "category error" by supposing that diseases—or disease categories—are necessarily causal (cause-based) entities. He also errs in supposing that traditional "medical" diagnoses necessarily or invariably "describe underlying biological causes." While psychiatric diagnoses often describe the patient's experience, they are made on the basis of symptoms and signs; and frequently demonstrate predictive validity, genetic risk factors, neurological correlates, and other psychometric characteristics. Finally, a psychiatric diagnosis can reasonably be said to be the cause of a patient's emotional, behavioral, and cognitive problems, even though the underlying cause of the condition itself may not be known.


Aftab, A. (2020, May 30). Mental disorders: From definition to nature. A Myth in Creation.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Author.

Kendell, R. E. (1975). The concept of disease and its implications for psychiatry. The British Journal of Psychiatry, 127, 305–315.

Pies, R. W. (1979). On myths and countermyths: More on Szaszian fallacies. Archives of General Psychiatry, 36(2), 139-44.

Pies R. W. (2019). Thomas Szasz and the language of mental illness. In: Thomas Szasz: An appraisal of his legacy. Edited by C. V. Haldipur, J. L. Knoll IV, E. v .d. Luft. Oxford University Press, 2019, pp. 155-66.

Shedler, J. (2019, July 27). A psychiatric diagnosis is not a disease. Psychology Today.

Szasz, T. S. (1987). Insanity: The idea and its consequences. Wiley.