There is someone in your house who should not be found.
Why does not matter tonight. Perhaps they are frightened. Perhaps they have done something, or perhaps they have done nothing and the wrong people believe otherwise. What matters is that they are real, they are someone you love, and an hour ago you told them they would be safe here. They are in the back room now, silent.
There is a knock at the front door.
You open it, because not opening it would say everything. A man is standing there, calm and reasonable, and he asks you plainly, as though it were nothing: is she here?
You know the kind answer. The kind answer — the one that keeps the person you love alive — is no. One word. No, she is not here; I have not seen her. It is not even hard to say. It costs you nothing and it saves everything.
But there is a voice at the back of your mind — a voice you may have been raised with, or one you would defend out loud in daylight — and it says: do not. A lie is a lie. The moment you say "no" you have made yourself a liar; you have decided that you are the kind of person who will say the false thing when the stakes are high enough. And once that is true of you, the voice says, your word means nothing, to anyone, ever. Tell the truth, it says, and let what follows be his doing, not yours.
And standing in the doorway you think: that voice is mad. Of course I lie. Who would not?
But consider what you are doing. You are standing in your own doorway calling a serious idea mad — and the person who built that idea was neither a monster nor a fool, and he saw this exact case coming. He chose it deliberately. By the end of tonight I do not think you will agree with him. But I do not think you will be able to call him mad either.
Last time we slowed down the half-second in which someone you love asks a question and the true answer and the kind answer come apart, and you have to choose. We found that most of us carry a single instrument for that choice — a scale with honesty at one end and kindness at the other — and treat the only question as where to set it.
This is Philosophy for Us — philosophy for everyone, no degree required. This is the second of eight episodes on the most ordinary hard thing there is: telling the truth, or not, to the people you love.
Tonight we set aside my words for the two ends of that scale and bring in the two thinkers who built them in their real versions rather than the slogans. One held that a lie is always wrong: even the kind one, even at that door. The other held that the kind lie is not a sin to be forgiven but the right thing to do, and that those who put the truth above everything are sacrificing a real person to a rule. I will build both until each has a genuine claim on you. Then I will set them against each other, and show you the one thing they never argue about.
The first view was built by Immanuel Kant, in an essay written in 1797 with a long title: On a Supposed Right to Lie from Philanthropy. From philanthropy — from love of others. He wrote it specifically to deny that any such right exists. Not even out of love. And he used almost exactly the case I have just described: a friend has fled to your house, a would-be murderer comes to the door and asks where they are. Kant says you may not lie to him.
The easy response — the one almost everyone has — is to laugh: to file Kant under the famous philosopher so lost in his own rules that he would hand a friend to a killer rather than tell a small untruth. If that were all there were to it, I would file him there too and we would move on. But that reading throws away the part worth having, which is the reason. Kant is not saying that lying is wrong the way a tidy rule for children is wrong. He is saying something more unsettling, and it is worth setting out properly, because it will bear on you later whether you accept it or not.
Begin with a question. What is speech for? When I assert something to you — when I tell you a thing — the whole point of the act, the reason the words carry any meaning at all, is that I am conveying my thought to you as true. That is the understanding beneath every sentence anyone has ever spoken to anyone. To tell you something is to offer it to you as true. So when I lie — when I assert what I believe to be false — I am not merely handing you a false piece of information. I am taking the one instrument whose entire purpose is to carry the truth between us and turning it against that purpose. I am using speech to defeat the thing speech is for.
Here is the step that is distinctively Kant's. The reading I will rely on tonight — from the philosopher Christine Korsgaard, who has thought about this essay as carefully as anyone — is that when you lie, the person you wrong is not only the one you deceive. You wrong, in Kant's phrase, humanity in your own person. You carry within you a rational nature: the capacity to reason, to give and weigh reasons, to deal with other people as minds rather than as obstacles. Speech is how that nature meets another. When you lie, you take your own reason — the thing that makes you more than a clever animal — and bend it into a means of managing someone. You degrade the very thing in yourself that the truth was meant to serve. The lie does not only cheat the other person; it is a small act of self-betrayal, every time.
Set that beside the white lie — the dress, the baby, the sister. What gives it its force is that it refuses to excuse the kindness by calling it small. Kant's claim is that there is no such thing as a lie that is purely between you and the listener, contained and harmless and soon forgotten. Every lie spends something that is not yours alone to spend. And here is the part that should make one uncomfortable, because it is aimed directly at you: every time you tell yourself but this lie is different, this one is kind, you are doing something quite specific. You are making your own case the exception. You are saying that the rule holds for everyone — truth is what we owe each other — except here, except me, except now, because my reason for breaking it is a good one. And Kant's reply is simple. Whose reason is not good, to themselves? The confidence trickster thinks his reason is good. Everyone who has ever lied believed, in the moment, that this particular lie was the justified one. If "but mine is kind" is enough to step outside the rule, then there is no rule — only everyone, all the time, judging their own case to be special.
There is one more point here, quieter, and I think it is the one that weighs most on a decent person. Consider what the kind lie assumes about the other person. When I decide that my sister cannot handle what I really think — when I decide, on her behalf and without asking her, that the truth is too much for her and that she must be managed with a softened version of reality — I have made a judgment about her. I have decided that she is not quite a full adult, not quite someone who can stand in the real world alongside me. I have placed her, gently, among the people who must be handled. And I did it for her own good, which is what makes it so hard to recognise as a kind of contempt. Kant would say: look again. The most respectful thing one can do for another rational being is to treat them as one — as someone who receives the truth and decides for themselves what to do with it. The white lie, however warm, quietly treats them as less than that.
So that is the first view, built — not as a rule for children but as a claim about what one does to the truth, to oneself, and to the person one is sparing, every time one decides that kindness has earned an exception. Now back to the door — the place where this serious idea looks as though it is about to get someone killed.
Here is the problem, and being clever about it does not make it go away. The person you love is in the back room. The man on the step asks: is she here? And Kant, in that 1797 essay, looks straight at this and says you still may not lie. Tell him the truth, or say nothing, but do not assert what is false. And if you tell the truth and he finds her and it ends the way we both know it ends, that, Kant says, is not on you. You did the right thing. The wrong is entirely his.
Everything in one revolts against this. Whatever was compelling about the argument a minute ago, it cannot survive this case. If a moral theory hands a frightened person through the door to be killed and tells you to feel unstained about it, then the theory has broken, not you. One does not need a philosophy degree to see that; one needs only to be a person standing in a doorway.
So which is it? Was the whole thing a sham — a fine-looking argument that collapses the first time it costs anything real?
This is where it becomes interesting, and why I brought Korsgaard in rather than Kant alone. Among the people who have thought hardest about this — the ones who take Kant most seriously — there is a serious line of argument that Kant got his own answer wrong. Not all of them: some accept the hard conclusion and say the duty really is that strict, and they have their reasons. But the reading I am relying on tonight, Korsgaard's, holds that Kant misapplied his own central idea. Her argument runs like this. The wrong-making feature we built in the last segment — using reason against reason, treating a person as a thing to be managed — depends on the other person being inside the arrangement. The understanding of truthfulness is an understanding among people who are dealing honestly with each other. It is a relationship between two rational beings, both bound by it.
But the murderer at the door has stepped outside that relationship. He is not there to reason with you. He is there to use the truth as a weapon — he is asking for your honesty precisely so that he can do violence with it. He has already torn up the agreement that made your truthfulness owed to him in the first place. And to a person who has done that, Korsgaard argues, you owe no truth at all. Lying to him is not making your own case the special exception — the thing Kant rightly feared. It is something different in kind: it is refusing to let your own honesty be turned into his instrument. On Korsgaard's reading you may lie to the murderer at the door and still be acting on Kant's own deepest principle; she argues that Kant failed to see how his best idea applied at the edge. That argument is contested — other careful Kantians push back — but it is a serious reading, not a dodge.
There is a second philosopher worth a sentence, Helga Varden, who approaches it through Kant's legal and political philosophy. Her point is narrower, and worth stating precisely: the crude headline — that Kant says you must tell the killer at the door where the family is hiding — is not Kant's considered position once the essay is read alongside his philosophy of right. That does not make the lie simply fine on her account; lying remains a wrong in the eyes of right. It means only that the crude duty to hand over the family is a caricature. Two serious readers, then, both holding that the headline rigorist is not the real Kant — and neither of them treating lying as simply permitted.
Now, why does this matter for us, at the kitchen table, nowhere near a murderer? Because notice what has happened. We took the case that was supposed to destroy the first view, and the view survived — but it came back changed. It did not survive as "lying is always wrong, no exceptions, hand over your friend." It survived as something narrower and harder to dismiss: the wrong of the lie is real — it concerns what you do to the truth, to yourself, and to the person you deceive — and the exceptions are not yours to grant merely because your reason feels kind to you. They have to be earned by more than "but I meant well."
And that is the version that stays with you afterwards. The murderer at the door comes once in a lifetime, if ever. The dress, the baby, the sister, the friend's dull new partner — that is Tuesday. And Kant's real question, the one left standing even after Korsgaard rescues you at the door, sits at the kitchen table with you: when you grant yourself the white lie, is your case really like the murderer's — someone who has forfeited your honesty — or is it just you, deciding your own reason is good enough? The door does not release you from that question. It only clears away the crude reading so that you have to face the real one.
So that is the first view: built, tested, and still standing — wounded and qualified, but not mad. Now I owe you the other one, and I will build it just as carefully, because if I do not, you will think this was a setup.
So change sides — and I mean genuinely change sides, not consider the other view politely before returning to the one we prefer. By the end of this segment the case for it should stand as strong as the one you came in holding.
Return to the sister in the kitchen. Three weeks out of a ten-year marriage. The lease is signed, the boxes unpacked; there is no version of the night in which she goes back. And she asks: was I right to leave him? The first view, even in the careful, wounded version we have just built, wants me to notice certain things — that there is a real answer, that softening it makes my own case special, that telling her the warm thing quietly treats her as someone who cannot handle the world.
And the second view says: that is a fine argument, and you should be ashamed of yourself for bringing it anywhere near that woman's kitchen.
Here is the second view at its strongest. Morality is not about the condition of your own conscience — not about protecting your own moral record, or keeping your record of true statements unbroken. It is about what actually happens to actual people. So look at what actually happens. If I tell my sister the blunt truth — that I never understood why she stayed or why she left, that from the outside it looked as though she fought the same fight twice — what is the outcome? She is in more pain. That is the whole of it. She does not un-leave him. She does not gain some valuable self-knowledge she will thank me for in five years. She simply hurts, tonight, when she is already at the lowest point. I have bought nothing with her pain except the private satisfaction of having told the truth.
And if I tell her the kind thing — yes, you did the right thing — the outcome is that a person I love, on one of the worst nights of her life, feels less alone. Where in that is the wrong? Whom, exactly, have I wronged? Name the victim. The view that says I did something bad there — that I corrupted myself, spent some abstract trust, lowered her dignity — is so busy admiring its own principles that it has stopped looking at the one real person in the room.
This is the part the first view never quite answers, and you should feel its weight: some truths are simply gratuitous. They do not inform a decision — the decision is made. They do not right a wrong. They do not even make the other person freer. They only hurt, and the person delivering them gets to feel honest. And there is a suspicion that hangs over that — that "I was only being honest" is sometimes a respectable cover for "I wanted to say the cutting thing." Note it lightly. We will return to it, hard, later in this series; there is much in it that I am not opening tonight. For now, register only this: the truth-teller is not automatically in the right. Sometimes the blunt truth costs the other person everything and costs the teller nothing, and we still call it brave.
And the second view has one further argument, and it is the door. Recall Kant, before Korsgaard's rescue, saying: tell the murderer the truth, and the death is on him. The second view says that this is what your rule finally amounts to. This is the reductio — the point at which valuing truth-telling for its own sake shows itself for what it is, a willingness to sacrifice a real person to an abstraction. If your principle, followed honestly, hands your friend through the door, then your principle is the problem. Outcomes are not one consideration among many; they are the whole of it. A morality that can look at a comforted, grieving woman and a dead friend and say "but my statements stayed true" has lost sight of what morality was ever for.
That is the second view. And notice how much one wants it. It does not merely permit the kind lie; it commends it. It tells you that the warm thing you said to your sister was not a small, forgivable failure of nerve but the right thing, plainly, and that the version of you who would have told her the truth is the one who should feel bad. That is an enormous relief. Keep an eye on how good the relief feels — not because it is wrong; I have not shown you it is wrong, and I will not tonight — but because anything that feels this much like permission is worth watching as it arrives.
So now both are built. One holds that the lie wrongs the truth, the liar, and the person being managed, and that "but it is kind" is merely you making your own exception. The other holds that the lie spares a real person real pain and harms no one, and that it is the truth-teller who is left with the damage. Both at full strength. Both, at this moment, standing. Let us set them in the same room.
Set them in the same room and they do not disagree politely. Each makes a serious charge against the other, and each charge is true.
The first view looks at the second and says: you have built a method that can justify anything. Every liar who ever lived said "no one was really harmed" and "the truth would have hurt more." That is not the exception to your view; it is your view, handed to everyone — every coward and every manipulator — as a licence. You have made the wrongness of a lie depend entirely on a calculation of harm, and who performs the calculation? The person who wants to tell the lie. You have put the defendant in charge of the verdict, and you call it morality.
The second view looks straight back and says: and you have built a method that licenses cruelty and keeps the conscience clean. You would rather be right than kind. You have turned truth from something that serves people into something people are sacrificed to. When it mattered most — at the door, with a life in the balance — your view said "tell him; it is not your fault," and you had to be rescued from your own theory by an ingenious rereading. A theory you have to be rescued from is not a theory to steer by.
And here is what I have to be straight with you about, because it is the real state of the matter and I would be misleading you to tie it off neatly: both charges are true. The consequentialist view really does risk handing every self-server a licence — the calculation really is run by the interested party, the one who wants the lie. And the strict view really does, left to itself, lead toward monstrous results and need rescuing at the edge. They are not each half-right, with the truth sitting comfortably in the middle. Each charge is genuinely damaging, and neither view is refuted.
So let me do what this programme does, which is decline to choose for you and instead put the choice in your hands, on a case that is actually yours. Take a real one. Not the murderer — that is once in a lifetime. Take an ordinary one of your own: the friend's manuscript; the "honestly, how do I look?"; the "did Mum seem all right to you?" when Mum did not seem all right. Choose one where you still remember the half-second.
Now run the first view through it, all the way, without flinching. Say the true thing, in your head, now. Track what it does — to them; and track, too, what it does to you: the small note of self-respect, I did not lie, I respected them enough to be straight. Then run the second. Say the kind thing. Track that — the conversation staying warm, the friend still pleased, and the small private knowledge that you smoothed something over, that there is now a version of you they trust that is slightly off from the real one.
Here is what to notice, and it is not "which was right." It is that you cannot get the two costs onto the same scale. The first view is pricing one thing — what you did to the truth, and to yourself as a truthful person. The second is pricing something else entirely — what happened to the people involved. They are not measuring the same quantity and disagreeing about the total. They are measuring different quantities, and there is no master scale that converts one into the other. That is why it does not resolve. That is why you have been having the same argument with yourself your whole life and never settling it. You have been trying to settle on a single scale a question that has two needles, measuring two different things.
Both views, built to full strength. Both still standing. And you, still stuck — but stuck more precisely than you were an hour ago, which is the only kind of progress philosophy ever really offers. Now, before you take this away, there is one thing the two of them never fought about. They spent the whole episode attacking each other — and on one point they silently, completely agreed. And it is the point this whole series is going to challenge.
Go back over the whole dispute. Kant against the kind lie. The wrong of the lie against the harm of the truth. Humanity in one's own person against the comforted woman at the kitchen table. They disagreed about everything that could be disagreed about — what makes a lie wrong, whether outcomes are the whole of it, whether you owe the murderer the truth. Everything.
Except this. Both of them, throughout, took for granted that the thing under discussion was the false assertion — the spoken sentence that is not true. Kant's wrong is the wrong of asserting what you believe to be false. The kind lie is a lie — the warm sentence that is not accurate. Both views are, underneath, theories about the same single act: saying the untrue thing. They fight bitterly over when you are permitted to do it. Neither ever asks whether that is the whole of what honesty is about.
And it is not — not obviously. Let me raise one more thing, and then I will let you go. You can mislead someone without ever saying a false word. You can answer "did you finish the report?" with "I have been at my desk all day" — every syllable true — and walk away having planted exactly the impression you intended, having never technically lied. Was that honest? Both of tonight's views have surprisingly little to say about it, because both were built around the false assertion, and you did not make one. You told the truth. The misleading truth.
So the question you take away with you is this. Both of these large, serious theories — the ones you would want to commit to, the ones that have run your inner argument your whole life — quietly agree that honesty is a matter of whether the sentence is true. And it is at least possible that that is not what honesty actually consists in: that you can keep every sentence true and still be running a deception, and that you can utter a false sentence and not be. If that is right, then the scale you walked in with last week — the one I called into question at the end of the first episode — is not merely hard to set. It may be measuring the wrong thing altogether.
I am not going to settle Kant against the kind lie tonight, or ever, really — I told you that at the start and I meant it. You leave with both of them standing, both with a genuine claim on you, and no master scale on which to weigh them. That is not a failure of the episode; it is the true shape of the problem, and you can now feel it precisely: two needles, two different quantities, one scale that was never built to read both.
But you also leave with something the two of them handed you by accident, by agreeing without noticing. A suspicion: that the whole dispute you have just watched — the dispute you have been having with yourself — may be taking place at too shallow a level. That before you ask when you may say the false thing, there is a prior question no one in the room tonight thought to ask: what is honesty even about — the sentence, or something beneath it?
That prior question is where we go next — to the people who say the wrong lives not in the sentence but in the trust between two people, and to the literal truth that deceives without a single false word.
Tonight, take away the deadlock and take away the question. The deadlock: two great theories of the false sentence, both standing, neither winning. And the question the whole dispute was standing on: perhaps honesty was never only about the sentence at all.
Thanks for listening. I'll see you next time.