You have had this argument many times, and you know both sides of it by heart.
Someone says a thing is not really what you are calling it. It was not really violence — it was a shove, things got heated, nobody planned it. It is not really a disease — it is just where she falls on a normal range of moods. That was not really racism, it was a clumsy joke. She is not really a woman, whatever the document says. And someone else says it is all of those things, and you both reach for examples, and the examples settle nothing. For every case one of you brings, the other has a reason it does not count. So you add facts. More facts. The facts never close it.
Then one evening you notice something about the argument itself. You had not been disagreeing about the thing at all. You had been disagreeing about the word — about where it stops applying, and about what it is for.
The first time you see that, it comes as a relief. Of course. We are not arguing about whether she counts as a woman; we are arguing about what the word "woman" is meant to pick out, and why. We are not arguing about whether his sadness is a disorder; we are arguing about what we want the word "disorder" to do, and whom it is meant to help. The argument gets clearer the moment you stop treating it as a dispute about facts. You have found the question that was actually driving it.
And then it turns on you. If the argument is about what the word should mean, then is the person who redefines it simply cheating — redrawing the boundary so their own side falls inside it, redefining the word in the middle of the argument so they come out on top? You have watched people do exactly that, in plain bad faith, and call it justice. And you have done it yourself, meaning every word of it, and called it clarity. And from the inside, you cannot reliably tell which of the two you were doing.
Underneath that worry is a further one, which you have never quite said out loud. You do not know whether deciding what a word means is a power you have at all. Perhaps it is. Or perhaps you only ever assumed it, while the language went on meaning what it meant regardless of you. You do not know. You have never checked. You simply took it that deciding what a word should mean was the same act as changing what it means.
This is Philosophy for Us — philosophy for everyone, no degree required.
This is the first of eight episodes, and the whole series is about that one move: the move you make when you stop arguing about the facts and start arguing about the word. It is one of the most powerful things you know how to do, and one of the least examined. By the end of these eight episodes it should be neither — not a reflex you trust without looking, and not a trick you are quietly ashamed of.
Over these eight episodes I am not going to tell you which words you should fix, or how. I am going to do something slower. I am going to take that move — the one that felt like insight and felt like cheating in the same moment — and take it apart, until you can see exactly what you were doing each time you made it, and the one question inside it that you have never thought to ask. I do not stand outside any of this. I make the move too; I have called my own redefinitions clarity and hoped nobody would check.
So tonight does one thing. You think this move is something you stumbled onto late, in the middle of a single argument. It is not. You have made it on this show four times, deliberately, and never once watched yourself do it. Let me show you.
The move you have just described — stop arguing about the facts, start arguing about the word — you think you found by accident, in some argument that would not end. Let me name it plainly, because it has a definite form, and once you see the form you will recognise it at once.
Here is the form. You stop asking one kind of question and start asking another. You stop asking what is X — what is the fact of the matter, what does this word really pick out — and you start asking what should this word mean, given what we need it for. That is the whole move. A swap, from is to should. I am going to call it the swap for the rest of the hour, so that we both know what we are pointing at: you trade the question about the fact for a question about the word.
And you have run that swap deliberately four times on this show. Let me take you back through them, because they did not feel like the same thing at the time, and they were.
The first was money. We were asking whether a price could be fair — whether the number on the thing was the right number. Partway through, you stopped asking that and started asking what kind of value a thing can have that a price can never capture. That is not a question about the facts of the market; no measurement answers it. It is a question about what the word "value" should cover — whether we let the market's number be the whole meaning of the word, or hold a second meaning open beside it. You were running the swap on "value," and it did not have a name yet.
The second was machines. We were asking what a language machine actually does — what goes on inside it when it talks. We stopped, and asked instead what "understanding" should mean; and then we watched every test you tried to build, every test meant to deny understanding to the machine, quietly deny it to you as well. That argument was about the edge of a word: where "understanding" stops. The swap, run on "understanding."
The third was a word closer to home. "Disorder." Is a disorder a breakdown of function — a mind or a body failing to do the job it is built for? Or is it a failure to flourish — a life going badly by some picture of a good life? You felt, rightly, that the whole argument turned on which of those we decide the word should mean. Nobody was going to settle it by looking harder at the symptoms. The swap, run on the word you reach for when you ask whether something is wrong with you.
The fourth was race. We stopped asking what race really is — biologically, under the skin, in the blood, where no answer held up — and asked instead what the word should pick out, given the work we need it to do. And you met two answers that both ran the swap. One said the word is broken beyond repair: drop it. The other said keep it, and aim it deliberately at justice. Drop it or repurpose it — either way the question was what should this word mean, not what is it. The swap, on "race."
Four times. Set them side by side and the point is plain: this is not four clever tricks but one move, made four times. And every time, it felt like reaching the real question at last — like you had stopped repeating the same argument and reached the one that actually mattered. That feeling was not wrong. But it kept you from noticing that you were doing something with a form, a name, a method, and — we will come to this — a serious catch. Tonight you need only see that it is one thing, and that it is yours. You did not take it from me. You ran it four times before either of us called it anything.
Notice one more thing the swap leaves behind, because we will come back and examine it properly later. The old question — what is X, really — carries an assumption so ordinary you have never inspected it: that there is a fact about what the word truly means, a real definition waiting to be discovered, and your job is to find it. That feels like plain common sense. It is not. It is a particular idea, with a history and nearly a date of birth, and in a few weeks we will trace where it came from. For tonight, just notice that the moment you asked what the word should mean, you stopped hunting for the one true meaning and started arguing about which meaning we ought to build. You changed the task, from finding to making, and you did not announce it, even to yourself.
So before we go on, do this. Take a word you actually argue about — not one of mine, one of yours. "Merit." "Trauma." "Abuse." "Faith." Whichever you have genuinely argued over, the one that raises your temper. Ask yourself which question you were really in. Were you asking what it is, or what it should be? Be honest. Most of the time, when the argument would not die however many facts were thrown at it, you were already in the second question. You had already made the swap. You just never said so out loud, perhaps not even to yourself.
Now the part the episode is really for — the one thing worth carrying out of here.
When you make the swap — from what is this word to what should it mean — you are not asking one question. You are asking two, at once, in the same breath, and you have been treating them as one. They are not one. Let me separate them and set them down side by side, because once you can see both you will not be able to fuse them again.
Here is the first. What should this word mean? Given what we need it for, given whom it is meant to serve, given the work we want it to do — what is the better meaning? Call it the design question. It is a real question, with better and worse answers. You can do it carefully or carelessly, argue it well or badly, get it right or get it wrong. And — hold on to this, because it matters more at the end than it does now — the design question is yours. Nobody gets to tell you that you may not ask what "merit" should mean, or what "violence" should cover, or what "disease" is for. That question is open to you by right. You have every business asking it.
Here is the second, and this is the one. Once you have decided what the word should mean — once your case is made and your better meaning is finished and defensible — can you make it mean that? By what mechanism? Who does it? When you redefine "merit" at the dinner table, or in an essay nobody who disagrees with you will read, or at a meeting that ends and everyone goes home, does the word's meaning actually change — out there, in the mouths of the people you were arguing against, who never heard a word of your case? Call it the implementation question: can we make the word mean what we decided it should.
And here is what you have been doing — what almost everyone does, and what many of the loudest public arguments do without noticing. You have been treating those two questions as a single act, as if deciding the better meaning and installing it were the same thing — as if winning the argument about what "woman" should mean simply is changing what "woman" means. You decide, and in the same motion, with no join you can point to, it is done. The design becomes the fact. Should becomes is, on its own, because you settled the should.
But say it slowly and the difference is plain, without any help from me. Deciding what a word should mean is something you can do this afternoon, alone, with a pen and an hour. Changing what a word means — across a whole living language, among the millions who use it every day and never asked your opinion and are not going to — is what, exactly? The same act? A second, separate act? Something that happens later, by other means, through other hands? You do not know. And the reason you do not know is not that the answer is hard. It is that you have never held the two questions apart long enough to ask the second one.
Try it on a clean case. Take "merit" again. Suppose you have done the design question well. You have worked out exactly what "merit" should mean — what it ought to reward, what it ought to stop quietly building in, whom it has been failing — and your case is sound, as good as anyone has done on the word. Now: what happens next? Set out the sequence of events by which "merit," out in the world — in the hiring committees, the admissions offices, the arguments at other people's dinner tables — comes to mean your better thing. Who has to do what, and in what order? The question becomes puzzling, does it not. It is not that you have an answer and it is merely hard to carry out. It is that there is no answer where you were sure one would be. You had assumed the design did the work — that a good enough answer to should simply becomes the new is, by itself, the moment the argument is won.
That second question — whether the word's meaning can actually be moved, and by whom, and by what means — is what this whole series is really about. Not whether you are allowed to redefine a word. Whether you can. I am not going to settle it tonight; it takes the rest of the episodes, and the answer is stranger and less settled than either of us would like. Tonight the two questions need only sit apart where you can see both at once. Because once you have seen the second one — once you have looked straight at it and found nothing where you expected an answer — you cannot return to not having seen it. You will not again quite believe that deciding what a word should mean is the same thing as changing what it means.
So go back to the thing that was troubling you at the start: the worry that when you redefine a word to win an argument, you are cheating — redefining it midway so your own side ends up on the right side of the line. Look at that worry again now that the two questions are set apart, because it turns out the worry was leaning on one of them the whole time.
The charge of cheating assumes the move works. It assumes that when you redefine the word, its meaning really does change — that you really are redefining the shared word, and the only open question is whether you were allowed to. But that is the second question, the implementation one, the one we have just found you never ask. You cannot cheat by changing something you were never able to change. The whole accusation — you redefined the word to win — stands only if redefining the word is a thing you can do.
And we have just found that you do not know whether it is. So look honestly at where that leaves the worry, because it does not leave it settled. Perhaps the worry is exactly right: perhaps you really have been bending words to suit yourself and calling it clarity, and you should be uneasy about it. Or perhaps the worry is mistaken at the root: perhaps you were never redefining the shared word at all, because that was never something one person in one argument can do, and the whole time you were only proposing — only making a case to a language that was always going to decide for itself, in its own time, by means that have nothing to do with whether your argument was any good. I am not going to tell you which. Not to be coy, but because I do not think you can answer it yet, and neither can I, until we have done the work the next seven episodes are for.
But look at what you are holding now that you were not holding an hour ago. You came in with one move that felt like insight and felt like cheating in the same breath, and no way to tell them apart. You leave knowing the move is real — that you have made it four times on purpose, and that it has a name and a method, which serious people have spent careers on. And you leave knowing the thing worth more than any of that: the move was never one question. It was two, treated as one. What should the word mean — which is yours, fully, by right. And can you make it mean that — which you have somehow never once asked. That distinction, between the two, is what everything after this turns on.
This is the first of eight. Next time, the first easy answer — the one most people reach for the moment the worry about cheating arises. It says: you cannot just redefine words; words mean what they mean; changing one to win is always a cheat, full stop. It sounds like hard, clear-eyed common sense, and it is very nearly the most popular answer there is. So that is where we start. We take that rule and hold it against its own examples — the times a word's meaning plainly did change — and ask whether it survives them. I have a view about how that goes, but I will show you the test and let you reach your own verdict before I give you mine.
Before then, do one thing this week. Take the word you most want to fix — the real one, the one that has been quietly deciding things you care about. The next time you catch yourself reaching for it, arguing about what it really means, stop for a second and ask which question you are actually in. Are you deciding what it should mean? Or are you trying to change what it means? And — the hard one — can you even tell, from the inside, which of the two you are doing? You have been making this move your whole life. This week, for the first time, watch yourself make it.
Thanks for listening. I'll see you next time.