What the clinic actually means by a delusion of grandeur—and why the people who turned out to be right kept failing the test from the wrong end.
“Delusions of grandeur” is one of those phrases that escaped the clinic, got out into the world, and went feral. Out in the wild it means, roughly, someone with more ambition than I am comfortable with. Inside the clinic it means something far narrower, and the narrowness is the whole point, because by the strict definition almost nobody who gets called grandiose actually qualifies. Some of the people doing the calling, on the other hand, qualify rather well.
What a delusion actually is
Open the DSM-5 and you find a definition that has quietly shifted under everyone’s feet. Delusions, it says, are “fixed beliefs that are not amenable to change in light of conflicting evidence” (American Psychiatric Association, 2013). Read it twice, because the older versions most people still carry around defined a delusion as a false belief, and the manual dropped “false” on purpose. The committee was admitting something every working clinician already knew. You usually cannot establish whether a belief is true at the moment a frightened human being is sitting across from you. What you can assess is whether the belief moves when the evidence moves.
So the modern hallmark of a delusion is not its size. It is its rigidity. Not this belief is wrong but this belief will not budge, whatever the world does. Hold on to that, because it is about to do some damage.
The grandiose part
Grandiose is simply one of the themes a delusion can wear, alongside persecutory, referential, somatic and religious (American Psychiatric Association, 2013). A grandiose delusion is a fixed, unbudgeable conviction of one’s own special power, importance, mission or identity. The operative word is still unbudgeable. A man who believes he is going to overturn physics, and adjusts the instant the equations decline to cooperate, is not having a grandiose delusion. He is doing physics. A man who believes it and cannot be moved by any equation, any colleague, any result, is in different country, and it is country I have sat in a room with. The two can sound identical for the first ten minutes. The difference only shows up when you introduce evidence and watch what the belief does with it.
Who in the room had the fixed belief?
This is where the history gets awkward for the people usually cast as the grown-ups.
In 1935 a twenty-four-year-old named Subrahmanyan Chandrasekhar presented the Royal Astronomical Society with the maths for what we now call a black hole. Arthur Eddington, the most powerful astrophysicist alive, rose and ridiculed him: not with a counter-calculation, which he did not possess, but with contempt, which he had in quantity. Chandrasekhar collected the Nobel for it in 1983, by which time Eddington had been dead thirty-nine years and the universe had carried on manufacturing black holes throughout, never once having been informed it was being absurd (Wali, 1991).
Now apply the definition. Whose belief, in that room, was fixed and not amenable to change in light of conflicting evidence? The young man revised, defended, published, and was vindicated by the data. The eminent man held his position against the arithmetic until the day he died. I am not handing Eddington a diagnosis. That would be precisely the lazy move this essay exists to complain about, and you cannot diagnose the dead by séance. But notice that the epistemic signature the manual uses to define a delusion, the refusal to update in the face of disconfirming evidence, sat on the side of the institution, not the boy it was busy calling absurd.
The pattern repeats with a tedious reliability. Ignaz Semmelweis demonstrated in 1847 that washing hands in chlorinated lime collapsed the death rate on his maternity ward; his profession treated him as an irritant and he finished in an asylum, dead at forty-seven of an infection of the precise kind he had spent his life trying to prevent (Science History Institute, n.d.). Katalin Karikó spent two decades being demoted and defunded by her own university for believing messenger RNA could be turned into medicine; the medicine then went into several billion arms, and she shared the Nobel in 2023 (The Nobel Assembly at Karolinska Institutet, 2023). In each case the evidence was sitting there, countable and ignored, and the immovable belief belonged to the body with the letterhead.
Why institutions do this on purpose
You might file all this under the occasional Great Man having a bad century. It is more structural than that, and we can measure it.
Azoulay, Fons-Rosen and Graff Zivin (2019) tracked what happens to a scientific subfield when one of its superstars dies suddenly and prematurely. The morbid finding: after the death, contributions from non-collaborators, people who had never worked in the field, rise by close to nine per cent, and those incoming papers are disproportionately likely to become highly cited. The outsiders had the ideas the whole time. They simply would not come in while the great man was alive to block them, entering only once a field became less hostile to imported thinking. The authors’ dry conclusion is that dominant scientists “hold on to their exalted position a bit too long.”
That is gatekeeping with a standard error attached. Not villainy, not conspiracy, just the ordinary physics of status: the people best placed to recognise the next idea are the people most invested in the last one.
My own file
I have a stake in this, and I will declare it. I am a counselling psychologist, which means I have sat on the issuing side of the label and know exactly what it costs to get it wrong. I also spent most of my adult life on the receiving side of one that was wrong. The profession decided I had bipolar II, and went on deciding it for decades, with the serene confidence of people who had a form to complete and a drawer to file me in. The actual answer—that I am autistic and ADHD, a nervous system built to a different specification and quietly paying the surcharge the entire time—did not turn up until I was sixty-six.
For years I took that personally, as a private failure of the particular clinicians involved. The research suggests it was a pattern wearing my name. Kentrou and colleagues (2024) surveyed autistic adults and found that roughly one in four, and around one in three of the women, reported a prior psychiatric diagnosis they regarded as a misdiagnosis, with mood disorders well represented among them. My file was not an anomaly. It was a statistic nobody had got round to collecting yet. I wrote a book about it and called it Misdiagnosed, on the grounds that accuracy in titling is the least a misdiagnosed man can offer.
The test, applied honestly
Here is the part I cannot make tidy, and I am not going to pretend otherwise. The visionary who is right and early and the person in genuine grandiose crisis can look identical from the front. Both are certain. Both are alone. Conviction, I have come to think, runs at one temperature regardless of who is holding it, which is exactly why it is useless as evidence. The only thing that has ever separated the two is the boring, available, unglamorous act the institutions kept declining to perform: checking whether the belief moves when the facts do.
That is the entire clinical distinction, and it cuts both ways. It is the test we are trained to apply to the frightened person in the consulting room. It is also the test the consulting room, the faculty, the journal and the newspaper so reliably fail themselves, clipboard in hand. A delusion is a belief that will not change in light of conflicting evidence. Read that one more time, then ask who, in the histories, it actually describes.
My own correction came at sixty-six. Late, but on the right side of the grass. I would rather we did the looking while the person is still in the room, and still able to hear that we were wrong.
References
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). https://doi.org/10.1176/appi.books.9780890425596
Azoulay, P., Fons-Rosen, C., & Graff Zivin, J. S. (2019). Does science advance one funeral at a time? American Economic Review, 109(8), 2889–2920. https://doi.org/10.1257/aer.20161574
Kentrou, V., Livingston, L. A., Grove, R., Hoekstra, R. A., & Begeer, S. (2024). Perceived misdiagnosis of psychiatric conditions in autistic adults. eClinicalMedicine, 71, 102586. https://doi.org/10.1016/j.eclinm.2024.102586
The Nobel Assembly at Karolinska Institutet. (2023, October 2). The Nobel Prize in Physiology or Medicine 2023[Press release].
Science History Institute. (n.d.). Ignaz Semmelweis. https://www.sciencehistory.org/education/scientific-biographies/ignaz-semmelweis/
Wali, K. C. (1991). Chandra: A biography of S. Chandrasekhar. University of Chicago Press.
Leave a Reply