Szasz Myth Of Mental Illness

The Problem with Calling Everything a Disease

Thomas Szasz published The Myth of Mental Illness in 1961 and it still comes up in every psychiatric ethics board I've sat through. Most people who bring it up have read one chapter and decided they either love it or hate it. I've spent enough time on both sides of hospital ethics committees to say the book has some actual teeth in it, even if Szasz oversimplified half his own argument. The core claim is that mental illness doesn't exist as a medical disease the way pneumonia or diabetes exist. Not because the suffering isn't real. Because psychiatric diagnoses lack the kind of objective biomarkers that legitimate medical diagnoses have. You can measure blood glucose. You can look at a chest X-ray. You cannot run a lab test and confirm schizophrenia. What you get instead is a cluster of behavioral descriptions that get wrapped in medical language and treated like they're equivalent to a broken leg.

How the Szasz Myth Of Mental Illness Actually Plays Out

Here's what happens when you take Szasz seriously in a clinical setting. You start noticing that the language of medicine does a lot of work that pure social or moral language doesn't. When someone gets a diagnosis of depression, insurance covers treatment. When someone is described as "struggling with sadness," they don't. The label is a key to resources. It's also a key to social control. Szasz saw that clearly. But the counterargument that he dismissed too quickly is that brain chemistry absolutely matters. We don't need a single biomarker the way he demanded. We have medication response data. We have neuroimaging showing structural differences in chronic cases. We have genetics that explain partial susceptibility. The picture isn't clean. It's never clean in medicine. Heart disease had its "myth" periods too before we understood atherosclerosis. I worked on a case a few years back that made me reconsider both sides. An admitted patient, early twenties, refused all medication and told the attending physician that his auditory hallucinations were actually a spiritual awakening. He was articulate and self-aware enough that everyone on the team agreed he wasn't in the typical crisis pattern. Under a purely medical model, the doctor could initiate involuntary treatment. Under Szasz's reading, this person was simply holding a different set of beliefs and the state shouldn't coerce him into conforming.

The workaround our team used was to shift the conversation from diagnosis to function. We stopped asking whether he had a disease and started asking whether he could care for himself safely. That's actually closer to what the law requires anyway — competence evaluations are functional, not syndromal. It removed the ideological framing that was paralyzing the discussion and got us to a decision in about forty minutes instead of dragging through a committee meeting that would have gone nowhere for two days. That experience taught me that the Szasz argument is most useful as a check on medical overreach, not as a framework for deciding individual cases. The danger is when clinicians internalize the medical model so completely that they stop listening to patients as anything other than collections of symptoms. I've seen that happen. It's not theoretical. I've watched good doctors prescribe through a chart instead of looking at the person sitting across from them. On the other side, and this is where Szasz gets dangerous if taken too far, is the libertarian conclusion that no one should ever be subject to psychiatric intervention against their will. That position collapses the moment you deal with someone who genuinely lacks insight into their own condition and is causing harm. The "problems in living" framing works fine for existential distress and social nonconformity. It doesn't hold up well against acute psychosis where the person has no connection to reality and the consequences are measurable and severe.

There's also a practical nuance that most casual readers miss. Szasz drew a hard line between "mind problems" and "brain problems." That line has gotten fuzzier since he wrote. We now have treatments that clearly modulate neurochemistry and produce measurable behavioral change. Lithium doesn't work because it changes your moral philosophy. It works because it affects serotonin and glutamate signaling. Whether that makes bipolar disorder a "real" disease is partly a definition game. But dismissing the biological component entirely is no longer defensible in any serious clinical conversation. The DSM has moved in the opposite direction from what Szasz wanted. Rather than expanding medical categories, the latest editions have tried to narrow them by requiring more specific criteria. This has reduced some overdiagnosis but introduced new problems with reliability between different clinicians reading the same manual. Inter-rater reliability for several DSM-5 diagnoses still sits below 0.70, which is mediocre by any medical standard. If you want to engage with this material properly, read the original 1961 text first before anyone's interpretation of it. Szasz revised his views over the decades and some of the quotes people circulate come from interviews where he was being deliberately provocative. The actual book is more measured than the soundbites suggest. His later work on involuntary commitment and pharmaceutical industry influence is where the argument gets more polemical and less rigorous.

The useful takeaway isn't that mental illness is fake. It's that the medical model has blind spots and that language matters. Calling someone depressed rather than saying they are experiencing persistent low mood and anhedonia changes how other people treat them and how they treat themselves. That doesn't make the suffering less real. It makes the label more powerful, and power deserves scrutiny.