A year after the diagnosis, his wife said something that stayed with him. She said: you used to lose your keys. Now you have ADHD.
She did not mean it unkindly. She meant it as a description, and he could not answer it, because she was right and he could not say what was wrong with what she had said.
Consider how things stood before. Before the word, he was a man who lost his keys, who started four things and finished one, who stopped following a sentence his wife was still speaking. He had always been that man, and he had always carried the usual verdict about it — careless, a bit lazy, not trying hard enough — the same verdict carried by the woman with the late-night questionnaire we met at the start of this series. Then, at thirty-four, sitting across from someone with a clipboard, he was given the name. The name was a relief, a genuine one, and nothing in this episode takes that back.
But consider what the name did over the year that followed, because this is the part that usually goes undescribed. He did not simply learn a fact about himself. He reorganised around it. He went back through his whole life and read it again in the light of the word — the job he lost at twenty-six, the degree he never finished, the friendships that went quiet — and each one now read as ADHD, where it had not before. He found the online forums and learned the vocabulary — time blindness, body doubling, rejection sensitivity, executive function — words he had never had, which now fitted so well that he could not remember thinking without them. He began saying that's the ADHD about ordinary things: the keys, the drifting, a flash of temper. And here is what his wife had noticed. It was true that he had always lost his keys. But the man who loses his keys and the man who has ADHD are not living the same life, even when they are the same man doing the same thing. One of them is making a mistake. The other is being a kind of person. And over that year, without a single new fact arriving about him, he had moved quietly from the first description to the second.
So was it always there, or did the name make it? He cannot say. Not because he is confused, but because the question may have no clean answer. The lost keys were real before anyone named them; he is not inventing that, and neither is the name. But the shape of it — the weight of it, the way his whole past now gathers around a single name — that arrived with the word. The word did not find a finished thing in him and label it. The word changed how the thing was held. And he cannot get back to before the word to check.
He is not lying. He is not cured. He is simply no longer sure, when he says that's the ADHD, whether he is reporting something or becoming something. And he cannot tell the difference from the inside — and neither, if you have ever taken a name on yourself, can you.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the fourth of eight episodes on the most personal question this show asks: is something wrong with me? Last time we built the hardest-headed answer there is — that a disorder is the objective failure of a real function, as real as a failing kidney in any century you care to name — and then found its one deep problem: which departures count as failure looked like a judgment, not a plain reading. Tonight we turn to the people who come at the question from the opposite direction. They say that with a human mind, naming a thing can have a hand in making it; that the categories are visibly moving and expanding; and that the quiet question who gets to say? is a serious one. And, as we did with the realist, we will build their case at full strength — and then turn and test it.
If you walked out of last week thinking it's objective after all, settled, that is exactly the response these thinkers are waiting for. So keep last week's case in hand. We are about to see what argues against it.
The man who reorganised around his diagnosis is not an oddity. He is the clearest case of an idea that belongs to a Canadian philosopher, Ian Hacking, who spent his career on a question the realist never has to ask: what happens when the thing you are classifying can read the classification?
Begin with a contrast, because the whole idea rests on it. Take a kidney — last week's example, used once more. You can study a kidney for a century, name everything it does, sort it into types, and the kidney is unaffected. It does not find out it has been classified. It does not read the textbook and adjust. It does not filter differently because a doctor wrote "kidney" on a chart. The same holds for an electron, a rock, a strain of bacteria. Hacking called these indifferent kinds: the classification is one thing, the object another, and naming it changes nothing about it. The name describes the thing; the thing is unaffected by being described.
Now classify a person. Call someone depressed, or autistic, or an addict, or traumatised — and something happens that never happens to the kidney. The person finds out. They learn the word, and the word carries a description, a set of expectations, other people who bear it, and an account of what it means and where it leads. And then — this is Hacking's central point — the person lives differently in the light of the word. They act in new ways, or they read their old behaviour in new ways, or they meet others who bear the name and learn how it is meant to go. Because they live differently, the thing the word pointed at changes. The category alters the people it names, and the altered people alter what the category picks out. Hacking called this the looping effect, and he called these interactive kinds: kinds that interact with the very people they classify, so that the name and the named each keep moving the other.
The kidney does not loop. The depressed person does. That is the whole difference, and once you have noticed it, it is hard to set aside.
Consider what that does to the realist's central claim. Last week the realist insisted on one thing above all: the word does not make the thing. The failure is in you whether or not anyone names it; the naming only reads what is already there. For a kidney, that is exactly right. Hacking's claim is that for a human kind it can be exactly wrong — or at least far less clean than the realist needs. Because with a person, naming does not simply describe something already finished. Naming is one of the things that shapes it. You cannot stand outside and read off the disorder the way you read off the kidney, because the moment you apply the word, the word becomes part of how the person acts. Describing them changes what is described.
And the evidence is not a thought experiment. It is in the historical record, where you can watch a kind change in real time. Hacking's own leading case was a diagnosis that barely existed in nineteen seventy-two and then, over the next two decades, expanded enormously — thousands of people, many of whom found the diagnosis first and the symptoms afterwards; the typical presentation shifting as books and films and talk shows taught everyone what it was supposed to look like; and then contracting again when the experts lost confidence in it. Here was a way of being a suffering person that appeared, spread along the very channels that taught people the name, took the shape the culture expected, and then receded. Not because anyone was faking — Hacking was firm on this — but because a human kind is the sort of thing that can be partly made by being named, described, and circulated. The category was real in the way that mattered: people genuinely lived inside it and genuinely suffered inside it. And it was also, plainly, partly a product of its own naming.
Now I have to stop you before you draw the easy conclusion, because it is right here, and it is the thing that wrecks this position when people get hold of it. You are about to hear "partly made by the naming" and turn it, in your head, into "so it's not real." So the man's ADHD is not real. So the depression is invented. So everyone simply talked themselves into it and could talk themselves out. That is not what looping says, and Hacking spent much of his career saying so. A kind being interactive does not make it unreal. It makes it interactive. The man's trouble with his keys, his unfinished degree, the friendships that went quiet — those are real. The depressed person under the name is genuinely suffering. Looping does not say the pain is invented; it says the category the pain is sorted into is not a fixed, found-in-nature thing — it is something that moves when we name it, that we partly shape in applying it. You can hold both claims at once: this suffering is entirely real, and the category it lives under was partly made by our naming it. In fact you have to hold both, because both are true, and any version of this position that drops the first half is not rigour. It is the old cruelty — it's all in your head, you made it up — in a new form. We rejected that last week, and we are not taking it up again tonight.
So here is what the man's wife had seen, named properly. You used to lose your keys; now you have ADHD. She was not saying he was faking. She was saying she had watched the looping happen — watched the name become the centre his life reorganised around, until losing his keys was no longer something he did and became part of what he was. The trait did not change. The kind he was living under did. And that is the doubt the woman had at one in the morning, in the first hour of this series — was I always like this, or did the name make me this? — which I said then was one of the most important questions a person can ask. Now you can see why. It is not ingratitude, and it is not weakness. It is Hacking's question, it is genuine philosophy, and the realist's account has no place for it.
Hacking shows you a single kind changing under its own name. Now look wider, because a second thinker measured what happens when this occurs across a whole vocabulary, over a whole generation — and what he found is what the woman noticed at one in the morning when she saw that everyone has a word now.
His name is Nick Haslam, a psychologist in Australia, and he gave the pattern a plain name: concept creep. The claim is simple and, once seen, hard to stop seeing. The concepts we use for harm and for sickness — trauma, abuse, addiction, bullying, disorder — have been steadily expanding. They began by naming a narrow band of severe cases, and over a few decades they widened to cover milder and milder ones, until the word that once meant a soldier destroyed by a war now also covers a hard week, and the word that once meant a beating now also covers a thoughtless remark. The category does not hold still at its old border. It expands.
And notice — this is Haslam's care, and worth keeping — the expansion runs in two directions at once. It runs outward, reaching kinds of experience the word never used to touch: trauma now covers situations it would not have been used for in nineteen eighty. And it runs downward, reaching gentler and gentler versions of what it already covered — until things that used to be ordinary bad, a rude boss, a sleepless stretch, a child's normal misery, are re-sorted as mild instances of a clinical kind. New territory in one direction, milder cases in the other. The vocabulary of "something is wrong here" keeps covering more of what used simply to be called living.
Put Hacking and Haslam together and you have the constructionist's whole case, and it is formidable. Hacking gives you the mechanism — a human kind moves because the people in it live under the name. Haslam gives you the evidence — you can watch the kinds moving, outward and downward, across the culture, within your own lifetime. The man re-reading his past as ADHD is the mechanism. The fact that far more people carry that word now than a generation ago is the evidence. And the woman's late-night reasoning was exactly right: either every single person she knows genuinely became ill within the same twenty years — or the words for "ill" widened to cover a great deal of ordinary life, and most people, reasonably, crossed the new border without noticing it had moved.
Now run the same question on yourself; you cannot feel this one at a distance. Your own word — Anxiety. Depression. ADHD. Trauma. Whatever the name is that you have, or suspect, or have been circling. Hold it, and ask the looping question about your own taking it on. Not "is the suffering real" — leave that alone; it may be entirely real. Ask the other thing. When you took the word on — or when you began to — did you only describe yourself with it? Or did you, like the man with the keys, begin to live under it: re-read your past through it, learn its vocabulary, find the others who carry it, start saying that's the anxiety, that's the trauma, until the word stopped being something you had and became part of how you live? And here is the hard part, the part you will not be able to settle, and I am not going to settle it for you: can you tell the difference? Can you get back to the version of yourself from before the word and check whether it found something already finished in you, or shaped you around it? Try it honestly. I think you will find what the man found. You cannot get back to before the word. There is no clean reading of yourself with the word removed, because by the time you are asking, the word is already part of how you ask.
That is the exercise. But do not let it tip into the easy conclusion, because the same mistake is available here too. The word having a hand in shaping you does not mean nothing was there. You did not invent the dread before work or the mornings you could not get out of bed; the creep is real and the suffering at the mild end of the creep can be real, both at once. What looping and creep take from you is not your pain. It is the comfortable belief that the name for your pain was simply waiting in nature to be found — neutral, fixed, the same word it will be in fifty years. That belief is gone now. The name has a hand in the thing. You have felt it have a hand in you.
You may have met this idea once before, in a lighter form — a while back, in another of these series, watching a different category, person, change the very beings it was used to sort, where you held "this category is partly made" without sliding into "so it's arbitrary, so it doesn't matter." That was a lighter version. This is the same thing in full force, because this time the category is not "person" in the abstract. It is the one you have applied to yourself.
So far the case has been about the kind — that the category is partly made and always moving. But there are two positions close to Hacking's that are harder than his, in different directions, and you need both, because the listener who came in with a real diagnosis is probably already holding one of them without yet having the word for it.
The first rejects Hacking's approach from the opposite side. Hacking says the kind is partly made in the naming. This position says: no — the trait is entirely real, and there is nothing invented about it; what is invented is calling it a disorder at all. This is the neurodiversity view, together with its companion, the social model of disability. Be careful here, because it sounds like construction and it is the opposite of construction.
Take the man with ADHD again, and hear what this view says to him. It does not say your trait is a moving target the name invented. It says the reverse: your trait is real, it is stable, it has probably always been roughly what it is — and it is not a defect. It is a variation. Human brains come in more than one design. Some people attend narrowly and steadily; others attend broadly and in bursts. Some bodies sit still for eight hours; others do not. None of these is the broken version of the others. They are the range. And then the second half, which is the sharp one: what disables you is not the variation. It is a world built for one of the designs only. The eight-hour shift, the open-plan office, the lecture hall, the form to be filled in silence — these were built around one cognitive style, and they turn every other style into a deficit. Change the setting, and much of what gets diagnosed as a disorder turns back into a difference in the person and a mismatch with the setting. So the honest question, this view says, is not what is wrong with me? It is what is wrong with a world built with no place for me in it?
This is the social model of disability, developed forcefully by writers such as the British scholar Mike Oliver; and the term for its brain version, neurodiversity, entered academic use in the late nineteen-nineties — the credit for the word is genuinely contested, claimed in different forms by people within the autistic community, which is itself part of the point: this is a position with a movement behind it, not only a theory. And you must not file it next to Hacking, because they make opposite claims. Hacking is anti-realist about the kind — the category moves; it is partly our making. The neurodiversity view is realist about the trait — the trait is solid, real, yours, not going anywhere — and it relocates the whole problem from inside your body out into the built world. One says the thing is made by the naming. The other says the thing is entirely real, and calling it a disorder is the only mistake. For the listener the cold open described — self-diagnosed, reading at midnight — this is the live option Hacking is not: not maybe I made it up, but maybe there was never anything to fix in me, only a world with no place built for me.
Now the second position, and this one is harder than either, because it stops asking what the condition is and asks a question about power. Its oldest and most uncompromising voice was a psychiatrist, Thomas Szasz, joined by R. D. Laing, and the question they kept asking is this: never mind whether your suffering is "really" a disorder — look at who gets to say, and look at what they can do to you once they have said it. A diagnosis is not only a description. It authorises powers. When the right person writes the right word on the right form, it permits things that could not be done to you a moment before: being held against your will, being medicated without your consent, having your own account of your own life overruled on the ground that the account is itself a symptom. The word is not only a reading of you. It can become part of an authority exercised over you, by someone else.
And the history here is not subtle, because power has always had a use for the word sick. In the American South before the Civil War, a physician proposed a diagnosis — he called it drapetomania — for enslaved people who tried to escape. The wish to flee bondage was classified as a mental illness, a disorder of the mind, something wrong with the person who wanted to be free. Now, nobody defends that today. But do not let yourself off with "that was the past, and we are better now." Look at the shape of it. A power arrangement that wanted certain people to stay put reached for the medical word, because the medical word does something no other word does — it shifts the trouble from the arrangement to the person. The enslaved man who runs is not responding rightly to an intolerable thing done to him. He is ill. Something is wrong with him. And once it is wrong with him, the thing done to him disappears from view. That is what Szasz wanted you to be able to see, and it is genuinely uncomfortable, because the question it leaves is not safely historical. It is this: when the word disorder lands on someone — on the inconvenient child, on the grieving widow, on the worker who cannot manage the eight-hour shift, on you — is it always only a reading of what is wrong inside them? Or is it sometimes, quietly, doing the drapetomania thing — taking trouble that lives between a person and their world and relocating all of it into the person, where it can be treated instead of answered?
I am going to leave that question standing tonight and not follow where it leads — that comes later, and it leads somewhere we are not ready to go. But let it stand, because it is the most disquieting thing in this whole argument: that who gets to say something is wrong with you may not be a neutral medical fact but an arrangement of power, and that the people the word is used on are rarely the people who get to decide.
I have built all three of those as strongly as I can, and now I am going to do to them what I did to the realist last week — show you the limit of each. Not that any of them fails; none does. But each gives you less than it promised, and the price is worth seeing clearly before you reach for the comfortable version.
Start with the deepest problem, the one that hits Hacking and Haslam together, because it is the one you should not pass over. Everything they showed you is true and not enough, and the space between those two is where people are harmed. Looping and creep undo one belief: that the categories are fixed, found in nature with hard edges. That is real work, and worth doing. But undoing that belief does not tell you, of any particular diagnosis, whether it tracks a real harm or whether it is creep. Hacking can show you that some psychiatric kind once spread along the channels that taught its name — he cannot tell you whether your depression is one of those, or one of the real biological breakdowns that would wreck a person in any century. Haslam can show you that trauma widened outward — he cannot tell you whether the thing that happened to you is at the real end of the word or the widened end. The construction case is very good at showing you that a line was drawn by human hands. It is silent on where the line should be drawn. And an approach that can undo every boundary and locate none is dangerous in exactly one situation: when there is a real boundary and you need to find it.
Here is who bears the cost. "It's all just construction, the categories are made up, people used to cope" — say that with enough confidence, and somewhere a person who has a real, mechanical, treatable breakdown hears it, decides their suffering is a story they have told themselves, and does not go and get the thing that would have helped. The one boundary that mattered most is gone. Remember the man whose brother stopped — nine days under the blanket, the toast still on the plate, everything in him shut down. You cannot say that is partly made by the naming and have said anything true. The naming did not cause it. Some things break in any world, under any vocabulary; they were breaking long before we had the words, and they will break in whatever words come next. Construction has no answer to the brother under the blanket, and the honest constructionist does not pretend to have one.
Now the social model — the neurodiversity view — and its limit is its own best insight pushed too far. It is exactly right that a world built for one kind of mind turns other kinds into deficits, that a great deal of what we call disorder is really mismatch, and that changing the setting changes the diagnosis. Hold all of that. But push it to everything and it goes wrong, because it then has to say that no condition is ever really disabling in itself — that there is no possible world, no redesigned setting, in which the profound, the degenerative, the relentless pain is anything but a difference society failed to accommodate. That is not true, and the people living with those conditions are usually the first to say so. There is a line — a real and hard one — between the disability a better-built world would remove and the disability no building could ever take away. The social model is the best tool we have for finding the first kind. It cannot, by itself, tell you when you have crossed into the second. Relocate all the trouble into the world, and you abandon the person whose trouble genuinely lives in the body.
And the power critique — Szasz's, the hardest of the three. It is permanently right that the word disorder can be an authorisation, that who decides matters, that the medical word has been used to shift trouble off an arrangement and into a person, and that drapetomania was not a single lapse but a shape that recurs. Keep every bit of that. But in its strongest form — the form that says mental illness is therefore a myth, that there is no such thing, only labels and power — it fails in the same way as the rest. Tell the man whose brother could not lift his head that his brother is not ill, only labelled, only the victim of a power arrangement, and you have said something obscene. The power was real and the illness was real; the label can be an authorisation and still be tracking something that genuinely went wrong. The strong power critique gains its clarity by denying the brother, and that cost is too high. The serious version keeps the question about power and drops the claim that there is nothing but power.
So look at where you are standing now, because it is not where you were an hour ago, and not where you were last week either. Last week you held the realist — something can be genuinely, objectively wrong — and could not put it down. Tonight you have held the other side at full strength: the category really is partly made and always moving; the fault really can lie in the built world; the power to decide really is held by someone with power over you — and you cannot put any of that down either. Holding both does not give you a position. It leaves you with no place to stand. You cannot rest on "it's objective, settled," because you have watched the categories move under their own names and widen across a generation, and you have felt a name do this to you. And you cannot rest on "it's all just construction, or difference, or power," because some things break in any world, and that undoes the one boundary a suffering person most needs to find. Both claims are real, and they cannot both be rested on — and the natural thing to do, the thing you are probably already doing, is to reach for the thought that surely the truth is somewhere in the middle.
There is such a position. I am not going to give it to you tonight, and I am withholding it on purpose, because the relief on the other side of it is real — and it is also a trap, and the trap cannot be seen from here. You have to stay where you are, wanting the middle, to see why the middle will not save you. So we stay here one more week.
So what do you take from tonight. The second case — and the difficulty of holding it together with the first.
Last week left you holding the realist's case, and you probably walked in tonight thinking it had settled the matter: it's objective after all, the question has a fact of the matter and you can stop worrying at it. Tonight took that away from you, and it should have. You have now stood inside the most serious case there is that the categories are not fixed, found things in nature — that a psychiatric kind is interactive, that it moves when we name it, that a name can become the thing a person reorganises their life around, that the words for harm and sickness have widened across a single generation until ordinary life keeps being pulled across the line. You watched a man become a kind of person where there had been only a man who lost his keys. You ran the same question on your own word and could not get back to before it. That is real, and you cannot put it down again. Anyone who tells you the diagnosis simply reads off what was always there — neutral, fixed, the same word in fifty years — has to answer this first, the same way you did tonight.
And the second thing, which does not reconcile with the first — and is the whole reason tonight ends with no verdict. Everything you just held undoes one belief; it cannot tell you where the real line falls. It undoes the borders; it does not locate them. And some things break in any world, under any name, and "it's all just construction" is the sentence that leaves a real, treatable breakdown untreated, because someone decided their suffering was only a story. So you cannot take the easy dismissal — the categories move, but your pain is not a rumour, and the brother under the blanket is not a social construct. And you cannot take the clean diagnosis either — because you have watched the categories move, and felt one move you. Hold both — and notice how much harder that is this week than last. Last week the two truths only had to sit side by side. This week they pull in opposite directions, and you are held between them.
This is the fourth of eight, and you are not near the end of this. You are in the hardest part of it — two claims, both real, pulling in opposite directions, with no settled position between them. And here is where the next episode goes, because it goes straight at what you are already reaching for. Everyone in a position like this reaches for the middle: surely it's some of each — some of it real, some of it overreach. That is the most reasonable thought a person can have right now, and next week we will give it to you at full strength — the position that says the whole "all real versus all made-up" fight was the wrong fight, a false choice you never had to make. It is the answer most careful people land on, and it will feel like the settled answer you have been missing. And then I am going to show you why even it does not close the question — why the most careful answer in the room does not settle this either. The relief is coming. So is the reason the relief is a trap.
But that is next time. Tonight, the two cases held together, and no verdict between them. The next time you reach for the question — is something wrong with me? — you are no longer allowed either of the easy versions. Not "it's a plain fact, the test will find it," because you have watched the test's own categories move and widen and reach into the people they name. And not "it's all just labels, I can step out of it," because some things really do break, and you met one last week, and no amount of talk about construction got him out of bed. You are asking a real question, about real suffering, through a name that is partly your own culture's making and is still moving as you ask. Carry both. Neither one will let you put the other down, and that is the honest weight of where you now stand.
Thanks for listening. I'll see you next time.