Consider a man at the bottom of a flight of concrete steps. He cannot get into the building at the top.
The ordinary way to describe this is to locate the difficulty in him. He uses a wheelchair; his legs will not climb. The word that comes to mind, his included, is can't — and it looks like a plain fact about his body, read off him the way one might read his height.
But consider the steps themselves. Someone designed them. Before the building existed, the steps were a line on a drawing, and the person with the pencil chose stairs rather than a ramp. A ramp would have cost about the same. He chose stairs because every body he pictured approaching that door was a body like his own, one that climbs. He was not cruel; he simply was not thinking about the man in the wheelchair at all. And that is the point. He designed the entrance for people shaped like himself, and so a large part of the man's can't is in the concrete — put there years before the man arrived, by someone who never once pictured him.
So we can ask where the can't actually sits: in the man, or in the steps. The honest answer is that it lies in both — and that we habitually credit the whole of it to the man.
That question is the whole of this episode, and it extends well beyond a single building. Last week we found the people who wrote the test for who counts as a full person, and saw that they had drawn it in their own image. Tonight the subject is the one thing that has seemed solid underneath all of this — the facts the test reads. The incapacities. The can'ts. Throughout the series we have assumed that, whatever else is disputable, these at least are simply true: that when the test records "cannot reason," "cannot follow," "cannot keep up," it is reporting something that is just there in the body, the way the man's can't looks to be in his legs. Tonight we ask how much of it is in the steps.
This is Philosophy for Us — philosophy for everyone, no degree required.
This is the seventh of eight episodes on who counts as a full person, and who decides. Last week we met the people who wrote the test, and watched the line it draws pass through our own future bodies — and still, at the end of the hour, the test had not been refuted. Tonight we do not return to who wrote it. We turn to who built the setting in which it is applied. And then to something stranger underneath even that: the possibility that the very words the test uses do not hold still while we use them — that the word laid on a person reaches back and alters the person it was laid on.
Let me set out the shape of the episode, because there is a tempting error in it. At some point in the next hour it will begin to seem that the whole thing is invented — that if the can't is partly in the steps, and the words do not hold still, then personhood is only a story we tell, with no fact of the matter, and the question can finally be dismissed. That is the most comfortable conclusion in the series. It is also mistaken, and I will explain why it does not follow. Made is not the same as unreal. A thing can be built by us, entirely, and still affect a real person with real force. So: the setting, the words, and the one conclusion that does not follow.
The idea has a name, and it did not come from philosophers. It came from the people the test is run on — which, after six weeks on the question of who never gets asked, is worth noting.
In 1976, a group of disabled people in Britain — the Union of the Physically Impaired Against Segregation — set down a distinction that opens the whole question. They argued that one word was being used for two different things. There is the body, and there is what the world does with the body, and these are not the same.
Call the first impairment. That is the fact about the body — the legs that do not climb, the eyes that do not see, the ears that do not hear, the part of the nervous system that runs on a different clock. It is real, and located in the body; no one disputes it.
Call the second disability — and the distinction turns on this. On their account, disability is not the impairment. Disability is what a world built for some bodies and not others adds on top of the impairment. Their own formulation was blunt: it is society which disables physically impaired people. The man cannot enter the building. The impairment is his legs. The disability — the failure to get in — is the stairs, and the stairs were a decision. Seven years later, in 1983, Mike Oliver, a disabled scholar, gave the idea its name — the social model of disability — and developed it into a way of seeing the whole landscape.
Once stated, it is hard to stop noticing. Speech over sign: a world that runs on talking, so a deaf person is called "impaired" at communicating — when one could as easily say the hearing world is impaired at signing and has built everything around its own impairment. Print over braille. Stairs over ramps. And — this is the one that matters most tonight — a single tempo imposed over all the others. The meeting conducted at one speed. The form written one way. The classroom pitched at one pace of mind, so that a child whose mind runs slower, or sideways, or in a different gear entirely, is recorded as unable to keep up — when no one asked whether the room could have kept a different pace.
The movement's slogan was four words: nothing about us without us. And in those four words the direction of the series reverses: the people the test classifies are speaking back to it. The test we have studied for six weeks issues its verdicts about people, from the outside. These are the people replying: you built the room without us in it, you set the pace without asking us, and then you read our difficulty, in your room and at your pace, as a fact about our bodies.
Now apply this not to the building but to the test. The test reads a "can't" off a person and records incapacity: present or absent, full person or not. But if the social model is even half right, some unknown share of every "can't" the test has ever found was never in the body at all. It was in the steps. In the tempo. In a world built by people for people like themselves — the same competent, independent, reasoning adults we met last week, who wrote the test and poured the concrete. The line the test draws — this side counts, that side does not — is drawn by reading a "can't" off a body. And part of that line, perhaps a large part, was drawn long before the test arrived, by a designer who was only ever picturing himself.
Here I must be as careful as anywhere tonight, because the social model is clean and liberating, and it is easy to overstate it.
In its strongest form it can begin to sound like this: there is no real limit, only a badly built world. Every "can't" is a staircase. Build enough ramps, rebuild every room, correct the world throughout, and impairment costs no one anything — it becomes simply one more way of having a body, no harder than being left-handed among right-handed scissors.
And that goes one step too far — a point first made from inside the movement, by the disabled scholar Tom Shakespeare, not by a critic outside it. Some of it no ramp can reach. Recall the accounts we heard two weeks ago, when we let the people the test sorts speak for themselves. The pain that comes with some bodies is not the fault of the staircase, and no design can remove it. The illness that takes more of a person each month is not waiting for a better building. And the profound cases — the child who, for reasons that have nothing to do with stairs or tempo or anyone's drawing, genuinely cannot hold another person in mind — are not honoured by pretending the difference is entirely the world's doing and would vanish once the room was rebuilt. That is not liberation. It is a second way of declining to look at him. The honest form of the social model says as much: there is a residue — something the body carries that the world did not put there and cannot remove.
And once the residue is admitted, we reach the point where the criterion survives after all — and I will state that plainly rather than pass over it.
The consistent man from week three has been listening throughout, and he is untroubled. Give him the entire social model; let him grant every word of it. He does so without hesitation. Yes, he says. Build every ramp. Run the meeting at three different speeds. Rename every category twice over. I grant all of it — I grant that a great share of what my test read as "can't" was the staircase and not the man, and that I have credited the body with the architecture for two hundred years. Take it. And when you have finished — when the world is rebuilt and nothing remains but the residue — there will still be a child who cannot hold another person in mind, for reasons no design produced, and a mind in late dementia that no ramp reaches. And that, not the stairs, is what I was reading all along. You have not shown me wrong. You have shown me how much of my reading was the building. I accept the correction. The rest remains, and it is still what I count.
Notice what has happened, because it is the discipline of the whole series. The social model did not defeat the test. It removed the test's innocence — the cover of pretending that every "can't" it found was simply nature, simply the facts, nothing to do with us. Half of it was us. Half of it was the steps. The test concedes this, as the consistent man just did, and is not thereby defeated — because where the body really does fall short, after every ramp is built, it is still reading something genuinely present. The social model was not, after all, the thing that finally releases the woman in front of us from the test. It is the thing that tells us how much of the test was the world's doing — and then hands back the part that was not.
So the world supplies half the can't. That is the setting. Now the words — and this is the stranger point, and harder to hold, so I will build it slowly.
Ian Hacking, a philosopher of science, spent years on a question that sounds dry and is not: what is the difference between the things we classify in nature and the people we classify? Take gold. We have the category "gold," and we sort the world into it and out of it. And gold is indifferent to this. It sits in the ground being exactly what it is whether or not anyone has ever named it. Call it gold, call it nothing, revise the definition — the metal does not move. The classification reaches out to the gold; the gold returns nothing. Hacking called such categories indifferent: the name and the thing do not touch.
Now try the same with a person. Lay a name on a person — autistic, incapable, slow, gone — and something occurs that never occurs with gold. The person learns of it. The person hears the name, lives under it, is treated through it, grows up inside it — and is changed by it. That alters what the name picks out, which sends us back to revise the name, which alters the people again. The name and the person move each other, back and forth. Hacking called these interactive kinds, named the whole circling process the looping effect, and described, more bluntly, what we are doing when we sort people this way. We are, he said, making up people — not inventing a fiction, but assembling a real thing that was not there in that form before, the way one makes up a bed or makes up a team.
Return to the child in the report, the one carried since the first week — the eight-year-old with the phrase beside his name, "lacks theory of mind." Once that phrase is in the file, it does not only describe him; it begins to shape the setting he grows up in. Teachers read it and expect a little less. Specialists route him one way rather than another. He learns, in a thousand small ways, that he is the child who cannot read other people — and he forms a self around that, as anyone forms a self around what he is told he is. By adulthood the phrase has helped make the very person it claimed only to report. It did not hold still and point; it reached in and shaped. And then he and his generation push back against it, the category has to move, and the next cohort grows up inside a different name — and so the loop continues.
So here is the tempting conclusion, exactly where I said it would appear. If the world builds half the can't, and the names help build the people they claim to find, then surely it follows that the whole thing is invented — there is no fact of the matter about who is a person, it is labels all the way down, the labels can be removed, and the question can finally be dismissed. That is the conclusion. I said at the start that I would bring us to it. Here it is. And it does not follow, for two reasons.
First: constructed is not the same as harmless. Look again at the child. Everything that happened to him is constructed — the phrase, the file, the expectations, the setting. All of it man-made, none of it fixed in nature. And it landed on an actual child and bent the actual course of his actual life. A thing need not be fixed in the nature of the universe to affect a real person with real force. "It's only a label" is precisely the sentence that lets one stop watching what the label does. That something is constructed is no proof that no one is harmed; the construction is the means by which they are.
Second, and this settles the matter: even granting all of it — that the line is made, drawn by us, moved by us — none of that tells us where to draw it. Hacking shows that the names are ours and that they loop. He does not, because he cannot, tell us which name is the right one to lay on the woman in the bed. That a line is man-made is no argument that any particular line is wrong, or that no line should be drawn, or that all the lines are equal. "It is constructed" removes the test's last claim to innocence — the pretence that it was only ever reading nature. It identifies no answer. One can know, with complete certainty, that we made the category — and be exactly as uncertain as before about whether one's grandmother is still present. The construction account takes from the test the one cover it had left — that it was only reading nature, never us. It puts nothing in its place.
So let us take stock, because seven weeks in, we are carrying more than it may seem.
There are six answers to the one question, and none has won. The criterion that a person is a reasoning, self-governing mind — week three. The answer that every human counts regardless, a dignity that cannot be audited — week four. The answer that personhood comes in degrees, with no sharp line — week four. The correction that the independent self was always a fiction, and that each of us is carried by others — last week. And tonight's two: that the world supplies half of every can't, and that the names we lay on people do not hold still, but reach back and remake them. Six answers, each with something real to say, and not one of them able to settle the question. That is not a failure of the programme; it is the honest shape of the question, and anyone offering fewer than six is selling one of them.
Here, then, is the work, and it is yours; I will not do it for you. Do not run it on the man at the steps — that case is easy now. Run it on a can't you actually use. Find one you say of yourself — I can't do numbers; I'm not a people person; I've never been able to — or, harder, one you say of someone you decide for: the relative everyone calls difficult, the parent who "can't manage any more," the child who "can't sit still." Take it and ask the two questions the hour has earned. First: how much of this can't is the body, and how much is the steps — a world, a room, a tempo built for someone else, that you have been crediting to the person? And second, the stranger question: is the name doing something? Has "can't" been laid on this person so long, and so often, that it has helped build the very thing it claims merely to report — that they have grown up, or grown old, inside it? You will not reach a clean answer. That is not the point. The point is that you can no longer read a can't straight off a person and call it a plain fact of nature. You will always have to ask how much of it was built.
And none of this means the whole question has collapsed into arbitrariness. It has not. The criterion survives — we have granted it the world's share and the names' interference, and it has conceded both and gone on reading the residue underneath. The appeal to dignity still cannot be audited; the degrees answer still cannot find the line; and the new, comfortable conclusion — it is all constructed, so relax — does not follow either, because made is not unreal, and the construction is the instrument of the harm, not an excuse to ignore it. We have lost a great deal tonight and been handed nothing to put in its place. That is exactly where, in week one, I said the series would leave us.
That is the seventh of eight. Tonight we found that half of what the test reads was built — into the world, and into the names — and that knowing it was built yields no answer at all. One episode remains, and it is where this stops being a seminar. Next week we stop adding answers; six is enough. Instead we look, directly, at what this line has actually cost in the world — what happens when people stop arguing about who counts and begin acting on it. And then, at the end, I do the thing the series has promised since the first night: I hand the whole question back, and decline, flatly, to draw the line for you. Not out of coyness, but because there is no honest answer to hand down, and to pretend otherwise would be the one thing this programme was built to refuse.
This is the seventh of eight. One week remains.
Thank you for listening. I'll see you next time.