How Do You Know You're Not the Woman With the Milk?

Last time I handed you a question and asked you to keep your answer close. The question was: how do you know you're not the woman with the milk? She was certain her neighbour had cursed it — a complete, bodily certainty, the kind you would stake a life on, and she did. And she was wrong. There was no curse. So how do you know that your certainty — that the man in the recording did something wrong, that the wrongness is a real feature of the act — is not the same certainty pointing at the same kind of nothing?

And you had answers. Because I'm right and she was wrong. Because I'm so sure. Because I can see that it's there. I told you to keep them, and not to worry, because I was not going to touch them that night. Tonight I am going to take them up, one at a time.

Before we start, notice one thing you took away last time. It was not that you lost the argument; you didn't. It was that you reached for the answer to how do I know I'm not her — reached confidently, because it seems obvious — and found a reflex you had never had to work out. You had assumed it your whole life. You had never once had to state it. That is where we begin.

This is Philosophy for Us — philosophy for everyone, no degree required. This is the second of eight episodes. Last time: you hold a firm certainty that torturing a child for amusement is wrong; the episode never touched that certainty, and never will — but it showed that you cannot say what kind of fact "wrong" is, and it left you with the woman and the milk. Tonight we follow that question through to the end. I will tell you now how it comes out, because I don't deal in suspense. The certainty survives. It leaves this hour exactly as firm as it arrived. What does not survive is one particular answer you have been giving, without noticing, to the question of how you know.

Let me state the limit of tonight's claim plainly, because the argument comes near a conclusion it does not actually reach, and I don't want you drawing the wrong one. I am not going to argue that the act is not wrong. I am not going to argue that there is no such thing as wrong. Neither of those is tonight's business. Tonight does something narrower: it takes one thing you have been relying on to be sure — the thought that I couldn't feel this strongly about something that wasn't real — and removes it. Not the certainty. Only that one assumption. Keep the distinction in view; it is the whole episode.

Let's do something you have been avoiding for as long as you've known about her. Let's set your reasons for being sure beside hers and compare them honestly, one at a time, giving her no less than her due.

Start with her, and grant her every advantage, because the way we make ourselves feel safe from her is to picture her as a fool, and she was not a fool. She lived in a world where curses were not a wild hypothesis. They were the ordinary explanation, agreed on by everyone she trusted, the priest included, handed down from people who had never once been caught out on it in her lifetime. A child fell sick after a quarrel with a neighbour; the milk turned; the neighbour had been heard muttering. In that world, "she cursed it" was not the raving of a lone crank. It was the calm, obvious reading of the evidence — the reading a reasonable person gave. So when she said she was certain, she meant what you mean. Not "I have a hunch." She meant: I can see it. It is right there. Only someone being deliberately difficult would pretend not to see it.

Now set out what she had, and check each item against what you have.

She had certainty — complete, bodily, the kind that arrives before you have finished the sentence, the kind you would stake your life on. You have that. Yours arrived in a single word, wrong, before you could think. The two match.

She had evidence she could hold in her hands. The milk really had soured; she could smell it. The child really was sick; she could see it. These were not hallucinations. They were facts, available to anyone. You have that too. The act in the recording really happened. The body really was damaged. The fear was real. Hard, public facts, every one. The two match.

She had a track record — a lifetime of the feeling being right. Every other time the world had handed her that flat sense of this is so, it had been so. The bread was stale; it was. The winter would be hard; it was. Why would this be the one occasion the faculty that had never failed her suddenly failed? You have that too. Your sense of wrong has guided you your whole life and directed you correctly a thousand times — kept you decent, warned you off, and never, as far as you noticed, misled you. The two match.

And she had the willingness to act on it — to risk everything. She staked the neighbour's life. You would stake yours: you would put yourself between that child and the man without a second's hesitation, and you would be right to. The two match.

So go down the list. Certainty: she had it, you have it. Evidence in hand: she had it, you have it. A trusted track record: she had it, you have it. Willingness to stake everything: she had it, you have it. Every reason you would name — if I asked you why are you so sure, and you began listing reasons — every one of them she could match, item for item, in her own voice.

And she was pointing at nothing.

Now, the obvious objection is rising, and it is the right objection, so let us not pretend it isn't there. Fine, the two lists look the same from the outside. But mine is the one that is actually true. She matched me point for point and was still wrong, because there was no curse — and that is exactly the point: it is not the matching list that makes me right, it is that the wrongness is really there and the curse really was not. Good. Hold onto that. You have identified the only thing that could save you, and it will take up the next twenty minutes — because notice what you just did. You reached past every item on the list — past the certainty, past the evidence, past the track record — and took hold of the one thing that is not on the list and cannot be: but mine is true. Which is exactly what she would have said. She did not think she was guessing. She thought the curse was really there. The claim "but mine is the real one" is not a further reason you have and she lacked. It is the very thing in question. You cannot settle the argument by asserting your side of it.

So set aside all the reasons that turned out to match, and go straight to the one you reached for — the real one, underneath all the others. Strip away the evidence and the track record and what remains is this: I could not be this sure if it weren't real. The sheer strength of it — the way it is not a preference but something I cannot think my way past — that strength is itself the proof. You don't get certainty like this about nothing. That is the reason you actually trust. Everything else was secondary; this is the one that does the work. So let us examine it more closely than you ever have, because it is the most natural thought in the world, and it does not survive examination.

Here is the question that undoes it, and it is a plain one. Where does the strength of a conviction come from? Not where you would like it to come from — where does it actually come from? It comes from you: from the kind of animal you are, the brain you have, the way certainty is produced in a creature like you. A feeling of firm this-is-so is an event inside a person. It is a fact about the believer. And here is the part that matters most: that machinery — the part of you that produces the feeling of complete certainty — has no connection to the world that would let it check. It produces the feeling. It does not, in producing it, also confirm that anything out there answers to it. It cannot. It works from the inside, and it runs the same whether you are right or wrong.

You already know this, because you have watched it from the outside many times. The woman with the milk had the feeling at full strength and there was nothing beneath it. The man convinced past all argument that his rival had poisoned the well — full strength, nothing beneath it. Everyone who has ever been completely certain of something that turned out to be false — and the past is full of them, some far cleverer than you — had, from the inside, exactly what you have now. The complete certainty. The this-is-so. It felt to them precisely the way yours feels to you. There is no special quality that the true certainties carry and the false ones lack. If there were, the woman would have noticed that hers was one of the false ones. She did not, because she could not, because no such quality exists.

So consider what this does to your last reason. I could not be this sure if it weren't real. But the woman was exactly that sure, and it was not real. So plainly you can be exactly this sure of something that is not real — it happens constantly; it is one of the commonest things a person does. Which means the strength of your certainty, on its own, cannot tell you which of the two you are holding. It feels like evidence. It is the most convincing thing you have. And it is the one thing the woman had in full, equal measure, while pointing at nothing at all. The reason you were going to give to prove you are not her — I am simply too sure for it to be nothing — is the very reason she had. It was always hers as much as yours. You can insist on it as firmly as you like; it supports her position and yours equally, which means it supports neither.

That is the conclusion, and I want to be exact about what it removes, because the temptation now is to think it removes more than it does; and if you let it, you will spend the rest of the series resisting a claim that was never made.

What it removes is one inference. One. The step from I am certain to therefore it is real — the step that treats the strength of the feeling as though it revealed the fact directly. That step is gone. It was never valid; you never had it pointed out, because no one makes you. You cannot get from how sure I am to what is out there, because the first is a fact about you and the second is a fact about the world, and the woman is the standing proof that there was never any step from one to the other.

What it does not remove — and listen closely, because this is where the mistake is made — is your conviction that the act is wrong. I have not touched it. I have not given you one reason to think torturing that child was acceptable. I have said nothing about the act at all. All I have done is take away a bad argument you had for your conclusion. The conclusion is still there, exactly where it was. You came in certain; you are still certain; you should be. Losing a bad reason for a true belief does not make the belief false. It only means you have to find out whether there is a good reason, instead of resting on the one that turned out to be the milk-woman's. What you were sure of may well be so. You have only learned that "I can feel how firm it is" was never how you knew it.

Now I have to stop you going too far with this, because there is a wrong way to take the conclusion, and it is the way most people take it, and it is a cheat — a cheat in the opposite direction from the one we started with.

The wrong way is to hear all that and think: so it is just like the curse, then — there is nothing there, it is all in my head, the wrongness is a curse I am laying on the act. No. Slow down. I did not say that; the woman does not show that; and if you go that way you have made the milk-woman's mistake a second time — taken a feeling of sudden clarity and treated it as a finding. So let us separate carefully the two different things that were true of her, because you are about to need both.

The first: she was right about the milk. The milk really had soured. The child really was sick. Those were facts; her senses reported them correctly; and nothing about the absent curse takes them back. When the curse turned out to be nothing, the sour milk did not become fresh again. Hold that.

The second: she was wrong about the curse — the extra item, the one she added on top of the souring, the thing that turned "the milk went off" into "she did this to me, and it is an outrage." That, and only that, was the part that pointed at nothing.

Now make the same division in your own case, because it falls in exactly the same place. There is what happened — the body, the damage, the terror, the man's enjoyment of it. That is real. That is your soured milk: hard, in the world, reported correctly by your senses, and nothing in this episode touches it. The child was harmed. That stays true however the rest comes out. And then there is the extra thing you add on top — that it is wrong, forbidden, an outrage against the way things ought to be — the thing over and above the bare damage, the thing you were so sure you could see in the act itself. That, and only that, is the part the woman's certainty matched. That is the part your last reason was meant to prove. And that reason has just failed.

So here is exactly where you stand, and I want it clear, because the rest of the series is built on getting this one thing right and going no further.

Is the wrongness there, the way the souring was there? Perhaps. Is it not there, the way the curse was not? Perhaps. What you have lost — the only thing you have lost — is the right to settle that question by pointing at how sure you feel. You came in thinking it was not even a question: the wrongness was plainly there, as available as the smell of the milk, and your certainty was the proof. Now you know your certainty proves nothing either way, because hers did not, and yours is produced the same way. So the question that was not a question has become a real one, and it is genuinely open — open in both directions. The woman does not prove there is no wrong, any more than your certainty proved there was one. She proves something narrower and more unsettling: that from where you stand, with what you brought, you cannot yet tell. That is not "it is all invented." It is "you don't get to skip the work."

And that is the difference between tonight and last time. Last time you had complete conviction and did not know there was a question. Tonight you have the same complete conviction — every bit of it — and you can see the question, standing where your certainty used to stand, and you can see that you have not answered it. That is not a smaller place than where you began. It is a more honest one. You have stopped mistaking how firmly you believe for knowledge of what you believe.

I have been placing you next to a woman from a few hundred years ago, and there is something a little too comfortable in that. She is far away. She wears different clothes. It is easy to grant the point about her and quietly assume the lesson stops with her. So let us bring it closer. Let us find the woman with the milk in your own memory, because she is there, and once you have seen her you cannot pretend the point is only about people in old paintings.

Find a moral conviction you once held — held hard, with full certainty behind it — and have since completely reversed. Most people have one; take a moment and actually find yours. Something you were once sure was wrong — shameful, obviously so, not even a question — that you now think is plainly fine. Or the reverse: something you waved off as harmless, and now know, with your whole being, was wrong, and are a little ashamed it took you so long. Don't reach for a small one. Reach for one you would have argued about at the time. One you would have been a little offended to have questioned.

Now hold the two versions of yourself against each other — the one who was certain then and the one who is certain now — and ask the only question that matters here: from the inside, did the old certainty feel any weaker than this one does? It did not. That is the whole point. Back then it was complete too. You would have insisted, then, that this one was different, this one you were actually sure of. And one of those two selves — the earlier or the present — is simply mistaken; they contradict each other; they cannot both be reading the world correctly. The feeling did not mark which. It came at full strength both times, and at least one of those times it was full strength pointing at nothing. You do not need the woman with the milk. You have been her. You have held a complete moral certainty you now believe was false, and from the inside, on the day, it was indistinguishable from the truest thing you know.

And now you will do the thing — you will say: all right, but this one. Torture. This one is genuinely different. That other case does not compare. And perhaps it doesn't. I am not saying it does; I keep telling you the conviction stands, and I mean it. But listen to yourself: this one really is the real one is precisely what you once said about the conviction you later discarded. The feeling of "no, but this one is genuinely different" is not a separate, more reliable feeling. It is the same feeling. It is how complete certainty always feels from the inside — including the certainty that turns out to be empty. Which does not show your conviction is wrong. It shows, once more, and from the inside this time, that the feeling is not the thing that could tell you.

So that is where I leave you tonight, and let me close as I opened, by stating plainly what this is and what it is not, because you have earned that.

I have not told you torturing a child is acceptable. It is not, and I would sooner give up the whole enquiry than spend it arguing otherwise. I have not even told you there is no such thing as wrong — I do not yet know whether there is, and neither, if you are honest, do you, which is the new part. All I have done is remove one thing: the assumption that said I am sure enough that I don't have to look. That assumption was the woman's, and it failed her, and it is the same assumption, so it fails you. You do not get to be sure that way any more. You can still be sure. You just cannot use the sureness as the proof.

Which leaves the question standing where your certainty used to stand, and it is a better question than the one you came in with. Not how do I know I'm not her — you now know the honest answer to that is I can't tell, not by the feeling. The real question is the one underneath: is the wrongness actually there — out in the act, the way the souring was really in the milk — or not? And if it is there, what kind of thing is it, that physics cannot find and certainty cannot reach? That question has serious answers. Three of them, as it happens — three different accounts of what you are doing when you say wrong, each held today by people who have thought about this harder than either of us, and we are going to take all three seriously and hold each one to its cost. We begin next time with the first, and it is the one you will be most relieved to hear, because it answers plainly: yes. There really is a fact there. You were not imagining it. And then it tells you what you owe for the privilege of being right — and the cost is higher than you would expect.

So carry this out with you. The conviction, whole — you lost nothing there. And one fewer way of being sure, which feels like a loss and is in fact the first honest basis you have had for it. You have met the woman with the milk twice now: once a few hundred years ago, and once in your own memory. From here on the thing to do is to stop asking how you differ from her, and start asking the question she never got to ask: all right — so how would I actually tell?

Thanks for listening. I'll see you next time.