Part A: Why you can’t think your way out

1. It hijacks your reasoning

You have a faculty that senses its way toward what is good. You met it in Chapter 1. It knew you were doing wrong before you’d done any calculating, and it couldn’t tell you how it knew. That isn’t a defect. That’s what it is: a sense, not a rulebook. It can’t define the good any more than your eyes can define red.

And it isn’t a detector. It doesn’t sit there evaluating conduct. Chapter 2 named what it actually is: the part of you that cares—about people, about animals, about the world—and what registers as wrong is whatever cuts against that. Hold onto that. It’s the difference between a book about ethics and a book about what has happened to us.

Because it can’t define it, we stop trusting it. The other part—the one that reasons, that can make a case—can always produce something to show. So we run on that instead, and it feels like trading a vague instrument for a precise one.

We haven’t. Reason cut off from that faculty isn’t a poor substitute for it. It isn’t doing the same job at all. Sensing toward the good is something it never once tries. What it does instead is produce stories about how we already are good, and it is superb at that.

That’s what the first stranger was doing. He wasn’t a bad man who had thrown away his conscience. He was an ordinary man running on the arguing side alone, which produced a story in which he was the good one, and everything he saw confirmed it.

We’re all lost in some version of this disorientation, to some degree, most of the time. Me included.

We don’t notice because the faculty that senses the good is also the only thing that can tell you you’ve drifted from it. Demote it and you’ve silenced the alarm. What’s left is a picture of the good that feels like the good—same word, same warmth—with nothing connecting it to the real thing anymore. So it can drift, and there’s no limit on how far. A counterfeit with no original to check against can become anything, and every step of the way it will feel like being right.

Now scale that up.

Every grand plan for doing good—every ideology, every agenda, every “I’ll accumulate wealth and power now and deploy it for the world later”—is built by the arguing side, and the arguing side’s job is to route around what you already know. So when you ask “is this good?” and your mind hands you a proof, you haven’t learned whether it’s good. You’ve learned that the part of you that manufactures justifications has been put to work.

It’s why the aftermath of an institutional disaster looks the way it does. Someone asks how anyone could have known, and then the inquiry finds the memo. There is nearly always a memo, and someone who raised it, and a reason on file for why the concern didn’t hold. The reason wasn’t a lie. It convinced people. That’s the point: the proof existed so that nobody would have to listen to the thing that already knew.

We build rules and moral systems less to make ourselves good than to let ourselves stop paying attention. It’s easier to follow a rule than to stay fully present to what a situation is asking of you. Rules let you say “I did the right thing” without feeling the weight of it. That’s their appeal, and it’s why goodness can’t be formalized: any definition of “good” can be obeyed to the letter while betraying its spirit, and the parasite is extraordinarily good at finding the letter. The belief that metrics can tell us whether we’re on the right track is the same belief, moved outside the skull.

One test for any grand narrative: does it require you to have contempt for what’s right in front of you—this person, this place, this moment—in service of some future good? If so, it is almost certainly not coming from the part of you that cares, and it will not do what its proof says it will.

None of this is an argument against thinking. Thinking is indispensable, in service of the part that knows; it’s how conscience gets anything done. What it cannot do is replace that part, or tell you whether it’s been hijacked. You need something deeper than thought to check that.

I’ll be accused here of defining conscience as whatever happens to agree with me, since I can’t define it or prove which inner voice is which. The honest answer is that I can give you clues but not rules, and that the impossibility of a rule is the point. A rule is what the counterfeit needs. The clues are in section 12.

So let me say now what that commits me to. Nothing in this book is a rule, and nothing in it will be. I’m not going to tell you that self-interest is wrong, or anger, or punishment, or making an exception of yourself. Each of those is sometimes right. The moment I hand you a rule, your conscience gets a substitute to hide behind, and the part of you that argues gets a letter to satisfy. What I can give you are clues. A clue you can be wrong about. A rule you can obey.


2. Then it hijacks your perception

That’s what it does to your thinking. What makes this a trap is what it does to your seeing.

Once the verdict is in—this is what they are—the person in front of you starts to look like the verdict. Every gesture confirms it. It doesn’t feel like an interpretation you’re adding. It feels like what’s there. You can check this against your own memory: think of someone you once despised and later came to see differently. At the time, were you interpreting them as rotten? Or were you just seeing it?

That’s the hijack. Your world becomes indistinguishable from a world in which they really are that. So in what sense aren’t they? Only this: there is more of them there, and it isn’t reaching you. And you can’t tell that it isn’t, because the only thing that could tell you is the thing you turned away from.

This is why good intentions don’t fix it. You cannot simply decide to look. Deciding is done with the equipment that’s already turned, so the decision comes out turned. This is not a fight between a good part of you and a bad part, which you could win by trying harder. There’s nothing there to fight. There’s a not-looking, and it ends when looking becomes possible again.

Which is why the second stranger worked and the first didn’t. He didn’t win an argument with you. He made looking survivable, by taking what you are off the table, and then you looked. And it’s why contempt can’t do that, however deserved. It puts what you are on trial, and looking becomes the most dangerous thing you could do.

So this is where “evil people” live. In a world where their victims really are what they seem. Not a world they chose. A world that looks, from inside, like the truth. Which raises the obvious objection: then why can’t we just do what the second stranger did, to the worst people, and fix them? I have two answers, and neither is comforting. The deeper someone is in, the more has to be true around them before looking is survivable, and one person’s dignity is usually nowhere near enough. And I don’t know whether there’s a point past which nobody can be reached. I suspect there is. What I do know is which direction contempt moves them, and it’s not toward. When we hold the verdict, we aren’t opening the cage. We’re welding it shut.

This is the drift from section 1, seen from inside. The counterfeit good and the real one share a word—righteous—and at first they share a feeling. You can only choose the counterfeit by forgetting what the real one feels like, and the further you follow it, the less of the real one you have left to compare it to. Eventually there’s nothing to compare it to. The soldier isn’t confused about whether he’s righteous. He’s in a world where he is.

Here is the claim everything in this book depends on: two acts that look identical from outside can have opposite effects, because they come from different places. You verified it with the two strangers. The difference wasn’t in what they said or how hard they said it. It was in whether they were looking. Nothing auditable captures that. Which is why our arguments about policy, which are arguments about the auditable part, so rarely touch what decides the outcome. And it’s why you can’t tell from the outside whether you’re the second stranger or the first. From outside, they look the same. Even to you.

I’m avoiding the words good and harmful here on purpose. Whether an act helps is a further question, and often a hard one; there are people for whom the most helpful thing is prison. What Chapter 1 showed you is narrower and more certain than that. One of those men made it easier for you to turn away from what you knew. The other made it harder. That difference is real, you felt it, and nothing in any record of the encounter would show it.

One clue you can use right now: the next time you’re acting from contempt, notice how sure you are that it’s good. The real thing doesn’t have that quality. It doesn’t need to be sure. It’s just doing what’s in front of it.


3. What contempt is, from the inside

If you take one sentence from this book, take this one.

When you feel contempt, what you’re feeling is a tug to be recruited into the thing you hate.

It isn’t a signal that you’ve spotted evil. It’s the feeling of evil recruiting you, and it uses your accurate perception of it as the hook. You saw that with the stranger: he was right that there was darkness in you. He was recruited by it anyway, into doing the thing that fed it. The accuracy is what makes the tug so hard to refuse.

There’s a name for the form this takes when the harm is real, and I’ll use it: self-righteous contempt, not righteous, because that word has to keep its real meaning in this book. The second stranger was righteous, and nothing about him was contemptuous. The “self” is the tell. It says whose account the righteousness is being credited to, and that’s the difference between the real thing and the counterfeit that dresses in its clothes. Two things to check about it, in order, and the first is easy. Look straight at what you feel toward the worst person you can think of, at the moment the verdict lands, and see whether it’s anything but malevolence. I don’t mean the wish to stop them. I mean the heat, the part that wants them to know what they are. I’ve never found anything else in there. Look and see what you find. The second is harder, and you reach it the same way. If that’s what it’s made of, then it’s made of the very thing you’re aiming it at, and its natural direction is to reproduce itself: it reaches for the same thing in the other person and tries to wake it up. It doesn’t always succeed. Sometimes the person turns toward anyway. But whatever turned them didn’t come from the contemptuous part. The skill isn’t seeing that; you can see it now. The skill is not hiding it from yourself again by the time you’ve finished the sentence.

Look at the thing that’s being pushed aside. Ask a room of hardened soldiers whether it’s wrong to desecrate an enemy’s corpse. They’ll say yes, nearly all of them, instantly. Then ask them what the harm is. The person is dead. Nobody’s watching. No rule is being enforced. They won’t be able to tell you, and they won’t budge either. That’s conscience caught in the act: it knows, it can’t prove, and it doesn’t need to.

Now the cost. I think this is where the damage to whole societies begins. Every act of contempt teaches the same lesson to everyone it touches: that the inner knowing is worthless. If good and evil can’t be told apart from the outside, and the world only rewards the outside, why listen to the part that can’t prove anything? That is how a culture that trusts only proof gets built, one interaction at a time. And a culture that trusts only proof is one in which the single faculty capable of detecting this whole thing has been demoted to the “spiritual,” where it can be safely ignored.


4. How much of your life is in there

So far we’ve looked at one kind of moment: being caught, or being the one doing the catching. Now much more widely, and this part isn’t comfortable.

The claim is that limbo isn’t a rare exception. It’s the ordinary condition of most of what we do.

Consider the feeling of an action that serves two purposes. “Yes, this benefits me, but it also helps others.” You can even feel the balance; it feels like a reasonable compromise. Now check whether a deeper part of you knows that the “helps others” half is the fig leaf, and the other half is the point. Not always. But far more often than we’d like.

Consider the small thrill of getting away with something. Underreporting a little on your taxes. Keeping the extra change. I’m just sticking it to the man. And notice that the thrill is itself the evidence: some part of you registered exactly what you did.

Consider how easy it is to see all this in your enemies and how hard in your allies, and that the difference has nothing to do with who’s doing it more.

It can be much subtler than that. Being unwilling to really hear someone’s experience. “That person is stupid for feeling that way.” Looking at the people around you, on some days, as though they’re extras in your movie. Each of these is a turning away from the life in someone, and each carries the same signature, however faint.

And one more, which anyone who argues for a living will recognize. The uncharitable reading. You take what your opponent said and read it so that it’s stupid, or cruel, or self-serving, when a version that isn’t was right there and you knew it. You’d never accept that reading of your own words. That’s the tell. It’s the same exception, applied to how you read people, and it needs the same not-looking, because the fair reading was right there and had to be pushed out of view. I’ve done it in this book, to thinkers I admire, and had to go back. Watch for it in yourself. It’s the most respectable form the parasite takes.

And there’s a kind of limbo that doesn’t look like contempt at all, and nobody calls it that, and it may do more damage than the obvious kind. Watch someone pushing an agenda that happens to serve them or their group. A policy, a hiring decision, a family argument about who gets what. Nothing hostile is said. But there’s a look, a slight hardening, a readiness, and everyone in the room reads it. The person knows they’re wearing it and doesn’t. That look is limbo. It’s the stranger’s tell without the stranger’s heat. And it teaches everyone watching the same lesson contempt teaches: that the game is what’s real, and the part that knows better is for suckers. Most of what we call politics is this.

And here’s a move I made for years: I can see self-deception so clearly in other people. If I were doing it, I’d see it in myself. I don’t, so I’m not. The clarity about them became the proof about me. That’s the mechanism using its own detector as cover.

And notice what we call it. When it’s mine, it’s “unconscious.” When it’s yours, it’s “conscious,” and you knew exactly what you were doing. It’s the same state with two names, and which name it gets depends on whose it is. Both names leave out the thing that makes it what it is, which is that the knowledge was there and wasn’t admitted, and each of them serves the person doing the naming. “Unconscious” gets me off the hook. “Conscious” licenses what I do to you.

There are places where the softer word looks like the kind one. Tell a room of decent people that their bias is intentional and they stop listening. Tell them it’s unconscious and they relax. But look at what the soft word does. It picks a side. It tells them the knowledge wasn’t available, and it was, and saying it wasn’t is the demotion the first stranger taught you. That isn’t kindness. It’s niceness, which is the other way of collapsing it.

The second stranger didn’t do that. He held both halves and it cost him nothing, because both are true. What you did is unacceptable, and you knew, and you are better than it. That’s why he was harsh, and it’s why it landed as kindness. The honest frame is his frame. You knew. You didn’t let yourself know. And that is not a verdict on what you are.

Psychology has names for pieces of this, and I should say how they relate to what I’ve been describing, because a reader who knows them will be waiting. Motivated reasoning: my conclusion counts for more than the truth. Status defense: I count for more. Shame avoidance: my self-image counts for more than accuracy. The reward of dominance: I’m above. These aren’t rivals to the account. They’re what the parasite is built from, and every one of them is the same assertion at a different level: an exception, made for oneself, that has to be kept out of view. What makes that more than relabeling is that it predicts something. If they’re one thing, they should share one blind spot of one shape, and it should fail the same way for all of them: put the thing looked away from back in view, and the mechanism stops. For motivated reasoning that’s already been found. People stop searching the moment they reach the conclusion they wanted, and keep searching, past it, when the conclusion is one they didn’t want. That’s a not-looking with a direction. It’s the stranger’s tell, measured.

The point of this section isn’t to make you feel rotten. It’s to establish scope. Nothing about what happened between you and the stranger was a special event. It’s the water we swim in.



Back to top

A draft, changing often. Tell me where it fails.