Last time, we ended with a task. The question belonged to the realist, and the instruction was to run it on a wrong of your own choosing. Name everything that happened — every plain fact of the case — and then answer the one thing the episode turned on: once you have named all of it, is the wrong one more fact in the pile, or is it the thing standing on top of the pile? If you ran it honestly, you got stuck, and stuck in exactly the place where the philosophers who work on this are stuck. There were two objections, and neither could be answered. Suppose the wrong is one of the facts. Then why, once you had named every fact, was the question but is it wrong? still open? That is the naturalist's objection — the open question — and it would not close. Suppose instead the wrong is an extra fact, sitting on top of the natural ones. Then how do we ever reach it — how does a human mind make contact with a fact that is nowhere and made of nothing? That is the non-naturalist's objection — access — and it could not be answered either. Two positions, both insisting there is a fact, both left standing, and each with an objection it could not meet. So we come back where we started: certain the act is wrong, and no nearer to saying what kind of thing the wrongness is.
Before we take a step, here is one thought to carry in — and then to leave alone until the episode has earned it. For three weeks the question has been: which fact is the wrong? Every answer has cost something. But notice what every one of those answers had in common. Each took the question itself for a good question. They disagreed about which fact the wrong is. None of them stopped to ask whether "wrong" is in the business of naming a fact at all. Perhaps that is the hidden assumption — not which fact, but the quieter premise underneath the whole search: that when we say "wrong," we are trying to state a fact in the first place. If we are not — if that was never the word's job — then the objections we could not answer were objections to a claim we were never making. You cannot fail to find a fact you were never describing.
I am not going to tell you that is right. Not tonight, and perhaps not ever — it is one of three positions, and the other two hold that it is mistaken. But there is a third person we have not yet met, and the first thing they do is refuse the question. Ask them which fact the wrong is, and they answer: none of them. There is no fact there. You were never detecting anything. And before you decide this simply means it is your feeling — we settled that in the first episode — stop. Because it is not that. And the whole episode is about why it is not that.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the fourth of eight episodes. Last time we opened the first position, the one that says yes, there is a fact — and it turned out to hold two views, not one. The naturalist, who posits no extra entities and says the wrong simply is the suffering, and who owes us the open question. The non-naturalist, who keeps the wrong as a real, additional fact and owes us access — an account of how an evolved human mind ever makes contact with it. Both left standing. Neither objection met. Tonight we turn to the second position, and behind it is a person who says there is no fact out there at all — not the natural kind, not the non-natural kind — and who then hands back almost everything we cared about. That a moral claim is true. That we can be mistaken. That the act would be wrong even if we had been raised to cheer it. Everything the realists were defending — without either of the objections they could not answer. The natural reply is that this must be a trick, that you cannot have the truth without the fact. That is the question we open tonight.
Go back to the response you reached for in the very first episode, because the third person will be mistaken for it, and the whole episode depends on seeing why the two are not the same. Back then, when you could not find the wrong anywhere in the act, you reached for feeling. There is no wrongness out there; it is in me. "Wrong" just means I am disgusted, it turns my stomach, I disapprove. And you had already closed that response yourself, in your own words, because you do not mean I disapprove. You mean the act would be wrong even if you had been raised to cheer it — and a version of you raised to cheer, reporting approval, would be making no mistake, which you cannot accept. So the appeal to feeling failed. Call it flat subjectivism: the view that "wrong" reports a feeling — that "torturing is wrong" describes you, and says something true or false about what is going on inside you. It failed, and you are the one who showed that it fails.
Now here is the third person, and notice how close they come to that failed view without holding it. They agree with the first half. There is no wrongness in the act — you looked, it is not there, no fact, no property, nothing for the word to point at. So far, this is the appeal to feeling. But then they add something the appeal to feeling never said. When you call the act wrong, they say, you are not reporting your feeling either. You are not describing yourself. You are doing something else with the word — something the word "report" does not cover.
Here is the difference, and the whole episode depends on it, so take it slowly. There are two different things a person can do with a feeling. You can report it — "I am in pain" — which is a plain description, true if you are in pain and false if you are pretending. Or you can give it voice — "ow." "Ow" does not describe your pain. It is not true or false. It comes out of the pain; it is the pain made audible. Now carry that distinction to morality. The appeal to feeling said that "wrong" reports an attitude — describes your disapproval, in the way "I am in pain" describes your pain. The third person says no: "wrong" gives voice to an attitude. It does not describe your hostility to the act; it expresses it. And it does so in a particular register — not merely "I am against this," but "this is to be opposed, by anyone, by you as well" — expressing the attitude as a demand that everyone share it.
This is Simon Blackburn, and the name for the first half of what he is doing is projection. You have an attitude — a settled, deep hostility to the cruelty. You take that attitude and direct it outward, onto the act; you speak as though the wrongness were a feature sitting out there in the world. You project your own attitude onto the act, Blackburn says, and then read it back off the act as though you had found it there. And this is the part that matters: that is exactly why "wrong" feels as though it points outward rather than inward. Recall your own complaint from the first episode — it does not feel as though I am talking about myself; it feels as though I am talking about the act. Just so. Expressing your hostility as a demand that others share it is not the same as describing yourself. It points at the act. It is about the act. It keeps your attention on the cruelty and not on your own state. The appeal to feeling could never explain why "wrong" points outward — it had you covertly describing yourself the whole time, and you knew you were not. Blackburn accounts for the outward feel instead of explaining it away. Yes, he says, it points outward — that is what expressing an attitude in this register does — and there is still no fact out there. The outward feel was never evidence of a fact. It was evidence of what you are doing with the word.
Set that beside the realists from last week, because the contrast is the point. They looked at the same outward feel — the way "wrong" seems to be about the act — and said it feels that way because there is a fact out there that the word tracks. They posited a fact to explain the feel. Blackburn explains the same feel and posits no fact: the wrongness is in your speech because you put it there, not because you found it. The same appearance, with no fact underneath. That is what makes this a serious position, and why the philosophers who study it take it as seriously as they take realism. It is not the dismissal, and it is not mere venting. It accounts for your experience as fully as realism does — it just reads the outward feel a different way.
And here you have to be careful, because you can feel the old objection rising. If it is all just me giving voice to my attitudes, you think, then it is still just attitudes, and we are back at "it is all made up, grow up" — back at the dismissal I have been reaching for since the first episode. Hold onto that thought, because it is the right one, and it is exactly what Blackburn has to answer. The crude version of this view — the old "boo and hurrah" idea, that "wrong" is just a boo and "good" just a cheer — had precisely that problem: it could not tell a moral claim from a noise. If "wrong" is only "boo," you cannot be mistaken, you cannot reason, you cannot be wrong even if you had been raised to cheer — a boo is just a boo. And if that were all this position had, you would be right to walk away from it. But Blackburn's whole project — the reason it took him three books and a name of its own, quasi-realism — is to begin from you are voicing an attitude, there is no fact and then build the entire realist surface back on top of it, one feature at a time, until you have everything you thought you had lost. Truth. Mistakes. Reasoning. All of it. That reconstruction is where we go next. And it is why this is the position the listener was never shown.
So the third person begins where the dismissal begins — no fact, you are voicing an attitude — and then does what the dismissal never attempted. They earn the whole surface back. Watch how, because if it works you get everything back with no fact underneath any of it — and whether that is an achievement or a trick is a question worth keeping open the whole way through.
Start with the largest word. True. We want to be able to say it is true that torturing a child for fun is wrong — flatly true, not true-for-me. The realist says we can, because there is a fact and the sentence matches it. The third person has no fact, so how do they earn the word? Like this. When you say "it is true that torturing is wrong," look at what you are actually doing. You are not laying your sentence against a fact and checking the fit. You are endorsing the attitude — voicing it again, and adding, and I will stand behind that. A whole community of people voicing attitudes, arguing about them, correcting one another, discarding the ones that do not survive scrutiny, produces exactly this way of talking. "True" is what they say about an attitude that has been tested and held up; "false," about one that has not. And Blackburn's claim is the striking one: a truth earned that way is not a cheap imitation of the realist's truth. It is the only kind of truth there ever was. The realist thought "true" needed a fact beneath it. The quasi-realist says: watch me use the word correctly, doing everything you wanted it to do, with no fact beneath it anywhere. The word was never resting on a fact. It was resting on the practice all along.
Now mistaken. We want to be able to be wrong about a moral question — really wrong, not merely out of step with those around us. The third person earns this too. Your attitudes can be in poor condition. They can rest on false beliefs about what happened. They can be inconsistent — you condemn the cruelty in a stranger and excuse it in a friend. They can be the kind you would give up at once if you looked at them harder. To hold a wrong moral view is to hold a tangled, badly built, half-informed set of attitudes — which is a real way to be wrong, with real work to do to put it right. No fact required. Only attitudes that do not hold up.
And now the hard case, because it is the exact sentence that closed off the appeal to feeling back in the first episode, and the third person walks straight up to it and meets it. It would be wrong even if I had been raised to cheer. Recall why that sentence killed flat subjectivism: if "wrong" simply reports your disapproval, then a version of you raised to approve reports approval, makes no mistake, and the act comes out fine — which you could not accept. Flat subjectivism could not survive that sentence. Now watch the quasi-realist take it on. What are you doing when you say "it would be wrong even if I had been raised to cheer"? You are not reporting a feeling. You are taking a position about your own attitudes. You are saying: I am against this — and I am also against being the kind of person who could be raised out of being against it. It is an attitude directed at attitudes, your own possible ones included, and it holds whatever your upbringing might have done to you, because it is precisely a refusal of what your upbringing might have done to you. The response you reached for in the first episode and had to give up — the one that said perhaps it is all just my attitudes — is available again, and this time the sentence that broke it before passes straight through. That is what the listener was never shown: not the appeal to feeling, but the appeal to feeling rebuilt until it can say the one thing that was supposed to be impossible for it.
There is a second builder, standing on slightly different ground — Allan Gibbard. Where Blackburn speaks of attitudes, Gibbard speaks of norms and of plans. To judge that the act is wrong, Gibbard says, is to accept a system of norms — rules for living, for feeling, for what to be angry at — that rules the act out. When you say "wrong," you are not detecting a fact and you are not merely venting; you are committing yourself, signing on to a plan for how to live and what to forbid. And on that base Gibbard sets out to rebuild truth and logic — to show that the whole structure of moral reasoning, the way claims support and conflict with one another, is the structure of plans that hang together or come apart. Keep his machinery in mind — the norms and the plans — because in a few minutes it returns, as the most fully worked-out attempt anyone has made to answer the one objection this position faces.
But first, take the relief properly, because you earned it and it is real. Look at what you can now say. You can say it is true. You can be mistaken. You can say it would be wrong however you had been raised. You can reason about it, argue, correct, improve. That is the entire surface the realist was defending — every part of it. Now notice what you are not answering. Not the naturalist's objection, the open question, the but is it wrong? that would not go quiet — because you are not claiming the wrong is one of the natural facts, so no gap opens between the facts and the wrong. And not the non-naturalist's objection, access — because you posited no extra fact, no strange property located nowhere and made of nothing, so there is no faculty you owe an account of, no evolved mind reaching out to touch a thing it cannot reach. The realists could not stop answering that objection. The expressivist sets it down unanswered and keeps everything it was protecting.
And here is the clearest measure of the relief. Recall the cold thought that started all of this — how do I know I am not the woman certain the curse soured her milk, sure of something she could never reach? That worry weighed on the realist, who does claim to reach a fact and so must say how. The expressivist is the one position the worry does not reach. They never claimed to reach a fact at all. There is no out-there thing whose existence they might be wrong about, the way the woman was wrong about the curse — because they were never pointing at an out-there thing. The whole difficulty of the milk was being certain of something in the world that was not in the world, and the expressivist was never describing the world. So: almost the whole surface kept, and the objection the others could not meet simply does not arise. The relief is real, and you should have it cleanly, because the objection is about to arrive. Relief like this always comes with a cost. And this position's objection is stranger than either of the realist ones, because it is not about facts at all. It is about a single word, sitting inside an if.
Here is the objection, and it is an unusual one, because it is not a demand for a fact. The realists owed you facts they could not produce. The expressivist owes something smaller and harder: an account of a single word doing its work in a sentence where no one is asserting anything. It has a name — the Frege–Geach problem — and it is precise, so let me build it slowly.
The whole picture rests on this: "wrong" voices an attitude. When you say "torturing a child is wrong," you are voicing your hostility, as a demand that everyone share it. Good. That works when you come straight out and say it. But you do not only use the word that way. Consider this perfectly ordinary sentence: "If torturing a child is wrong, then getting your little brother to do it is wrong too." Now notice what you did not do. You did not say that torturing a child is wrong. You supposed it. You put it inside an "if" — you could say that whole sentence in a flat voice, as a piece of plain logic, and assert nothing about torture at all. So put the expressivist's own question back to them: what does "wrong" mean in there? It cannot mean "I am against this," because you are against nothing; you are inside an "if." The voicing is switched off. And yet the word is plainly doing something. The sentence plainly makes sense.
And it gets harder, because that sentence is part of a piece of reasoning that has to hold. Consider: "If torturing a child is wrong, then getting your brother to do it is wrong. Torturing a child is wrong. So getting your brother to do it is wrong." That is a valid argument — the most basic step in all of logic, the one everyone makes without thinking. But look at what validity requires. For that argument to hold together, "wrong" must mean the very same thing inside the "if" — where you assert nothing and voice nothing — as it does in "torturing a child is wrong," where you voice the full attitude. If the word means one thing voiced and another thing supposed, the argument trades on two senses of a single word and falls apart, the way any argument built on a pun falls apart. But the expressivist told us what "wrong" means: the voicing of the attitude. And inside the "if" there is no voicing. So either the word means something else in there, and the reasoning everyone relies on is broken — or it means the same thing in there, in which case its meaning is not the voicing after all, and the whole picture is in trouble. That is the objection. The expressivist owes us a logic for the word when no one is asserting it. It is real, unanswered on its face, and it is not a footnote — it sits at the foundations, because if you cannot reason with moral words, you have lost the very thing the reconstruction was for.
Now — does anyone have an answer? This is where Gibbard's machinery returns, the norms and the plans, and it is the most serious attempt on the table to meet this exact objection. The idea, roughly, is to stop treating the meaning as a voiced boo and start treating it as a plan. To accept "torturing is wrong" is to rule it out in your plan for living. And then the "if" sentence is about how plans fit together — any consistent plan that rules out torturing also rules out getting your brother to do it — and the reasoning works because an inconsistent plan is a real failure, in the way that contradicting yourself is a real failure. That is the shape of the answer, and it is serious work. But notice what I am not saying. I am not telling you it settles the objection. Whether it does — whether the plan machinery really captures everything logic needs, or only pushes the same problem back a step — is what the people who work on this are still arguing about, in print, right now, unsettled. The objection has a serious attempt at an answer resting on it. It does not have "answered" written across it. Keep those two apart.
And beneath that objection there is a deeper one, and this is the one I am going to hand you and not take back, because no one has the answer — not me, not the field. Suppose the reconstruction works. Suppose Blackburn and Gibbard earn back every last part — true, mistaken, mind-independent, reasoning, improvement — the complete surface, nothing missing. Then ask the quiet question that has been growing all hour: in what, now, does this person actually disagree with the realist? They say "it is true that torturing is wrong." So does the realist. They say "you can be mistaken about it." So does the realist. They say "it would be wrong whoever raised you." So does the realist. So where is the difference between them? The expressivist began by saying there is no moral fact out there — but they have now earned the right to say it is true, independently of any mind, that torturing is wrong, and if saying that, fully and sincerely, is not believing in a moral fact, then what would believing in one look like? The harder the expressivist works to keep the word "true," the more the realist can say: you have rebuilt my position, part by part, and now you are standing inside it telling me it is yours. Has the expressivist won the argument? Or has winning it dissolved the argument — left two people saying word for word the same things and calling it a disagreement? I do not know. The field does not know. It is a genuine puzzle, and it is yours to stand in, not mine to settle.
So now run the whole thing yourself, on something small and your own, the way you ran the realist's question last week. Here is a fresh case. A man you will never meet, in a bad week, drags a key down the side of a stranger's parked car — a car that took its owner two years of saved wages to buy — and walks off feeling lighter for having done it. You know what that is. It is wrong. Now run the expressivist reading on it, in your own words. I will ask; you answer; and take my word for none of it. Look for the fact first: is there a wrongness sitting in the scratch itself, in the metal, the way the souring was supposed to be in the milk? Look before you read on. The expressivist predicts you will come up empty-handed in the metal, and find, where the fact was supposed to be, your own hostility to the spite, put into words. See whether that is right. Then watch which way the words point when you say them — at the man, at the act, or back at your own state — because the claim stands or falls there: the position predicts they point outward, at the man, which the appeal to feeling could never explain. Now try to earn the surface back, one part at a time, and grant each part only if you honestly can. Can you say it is true it was wrong — say it and mean it? Could you be mistaken — is there a version of the facts in which your attitude should not survive, in which the stranger had it coming? And the hard one: would it still be wrong if you had been raised to think that keying strangers' cars is good fun? Stay with each question until you have your own answer, not mine. If they came back yes, yes, and yes, then look at what just happened: you kept every word, and you paid nothing for a fact you never posited. Now the harder part. You have just said it is true, independently of any mind, that the keying was wrong, that you could be mistaken about it, and that it would be wrong however you had been raised — and you said all of it with no fact out there in the world at all. So what is left between you and the person who insists there is a moral fact? Both of you say the same words about that car. Standing inside your own mouth, can you even tell which of the two you are? That is the puzzle, on a scratched car, tonight.
The same close as every week: plainly, what moved tonight and what did not.
What did not move, for the fourth week running, is the conviction. I keep putting it on the record because the further into this we go, the easier it is to mistake the questioning for the conviction itself giving way — and it is not giving way. Torturing a child for fun is wrong. Keying a stranger's car is wrong. Nothing tonight touched either claim. If anything, you spent the hour with a person who agrees with you outright — who says it is true that the cruelty is wrong, that you can be mistaken about right and wrong, that it would be wrong however you had been raised — and who says all of it without one moral fact anywhere in the universe. The expressivist is not the enemy of your certainty. They are trying to hand you a way to keep every word of it, and to keep it cheaply. The certainty is exactly where it was.
What shifted is the picture underneath it, and it shifted further than it did last week. You came in three weeks ago thinking the whole dispute was real against made-up. Last week you learned that "real" held two views. Tonight you learned that "made-up" — the dismissal you have been half-reaching for since the first hour, fine, it is all just feelings — held a view inside it you never knew was there, and it is nothing like the dismissal. It is the position most working anti-realists hold. It gives you back the entire realist surface — true, mistaken, mind-independent, reasoned-about — and trades both realist objections for one of its own. That is a serious position, and you had never once been shown it. And the objection it faces is not smaller, only stranger. It owes you a logic for the word "wrong" when no one is asserting it — Frege–Geach — with a serious attempt at an answer on the table and no "answered" written across it. And it opens a question the field cannot close: if it earns back everything the realist had, what is left to tell the two of them apart?
So here is what you carry out tonight, and it is heavier than last week's, not lighter. The second of three positions, held open: no fact out there, and almost nothing lost. Keep the relief — and you know by now not to bank it. This is the second of three. The realists from last week are still standing; they hold that this person is mistaken, and they have not been answered. The expressivist may be wrong. And there is a third person still to come.
Because notice the one thing tonight's position would not give up. It kept "true." Everything it rebuilt, it rebuilt to save that word — no fact, but the truth survives; that is the whole point of it. Next week you meet someone who also says there is no fact — none, not the natural kind, not the non-natural kind, exactly as tonight — and who looks at the expressivist's strenuous rescue of "true," shakes his head, and draws the opposite conclusion. Where tonight's position says no fact, but the truth survives, his says: no fact, so there is no truth to survive. He holds that the rescue fails, and that the honest word for a moral claim is not "true" but false — strictly, plainly false, in the way "the witch cursed the milk" is false. This is the view you have been bracing to hear since the first night, the one your whole body recoils from: that every moral claim you have ever made is, straightforwardly, false. Why he says it — what he thinks we are doing with the word, and why he thinks it misfires — is his hour, not mine to give tonight. The man has a name, and we will hand it to him next time. Tonight just hold the shape of where he stands: the same no fact you met tonight, and the opposite ending. One says no fact, keep the truth. The other says no fact, so there is no truth — only a mistake we all make every time we open our mouths.
Before next week, do one thing. Take your own wrong again — the scratched car, or any one you are sure of — and the reading you just ran on it, where you kept "true," and "mistaken," and "wrong however I was raised," all without a fact. Then ask yourself the one question this hour leaves in your hands: if you can say all of that, and mean it, with no fact anywhere — is there anything left that the realist has and you do not? And if your honest answer is no, nothing — then the question is which way that cuts. Perhaps the anti-realist has quietly won everything. Or perhaps he has merely walked home, rebuilt the realist's position, and put his own name to it. And you will not be able to tell which, from inside your own certainty. That is what I am sending you home with — not an answer. A question you can no longer un-ask.
Thanks for listening. I'll see you next time.