Skip to content
TODD BROWN
« The Anatomy of Corruption and Corrigibility II — The Pathologies · Chapter 7

Anti-Memesis: Why Critique Bounces Off

The adult child had been rehearsing the sentence for years. Not the exact words. Those changed every time she ran them in the shower, in the car, in the fifteen minutes before the phone call she kept postponing. What she had rehearsed was the shape of the thing. She knew that if she led with an incident, her mother would refute the incident, and the conversation would become a trial over facts neither of them could reconstruct: who said what, in what tone, in 2003. So she waited until she had the discipline to stay at the level where the problem actually lived.

She had tried it once before, badly. That time she'd led with an example — a birthday dinner where her mother had spent the evening describing her own week — and her mother had spent forty minutes establishing, with real forensic energy, that the dinner hadn't happened quite that way. They argued about the seating and whose idea the dinner had been, and the actual point got buried. So this time she came without an example. She came with the pattern only, sanded down to its plainest form. If the pattern couldn't survive being said out loud, no single instance was going to save it.

They were at her mother's kitchen table. Coffee going cold. She said it plainly, the way you say something you've decided not to take back:

"I don't always feel heard by you. It's not any one incident. It's the pattern of how we relate — I bring something up, and somehow the conversation ends up being about you."

Her mother didn't get angry. That was the part she hadn't prepared for. Her mother's face did something soft and pained and entirely sincere.

"You're right. I haven't been there enough. I'm sorry — truly. Tell me the specific times. Give me the list, and I promise you, I will make every one of them right."

There it was: agreement, apology, an offer of repair, delivered without a flicker of defensiveness. By every visible measure, the message had landed and been received with grace. The daughter should have felt something settle. Instead she felt the floor tilt very slightly, in a way she couldn't name.

Here is what happened, and it's worth being exact, because the exactness is the whole chapter. The daughter's sentence had two possible readings. The first: there have been specific moments when you failed to listen, and I can list them. The second, the one she'd spent years working up to: there is something in how you relate to me, ongoing and structural. It doesn't show up as a moment. It shows up as a pattern, and the pattern is the problem. Her mother heard the first reading instantly, generously, and completely. She never heard the second one at all. Not because she rejected it. Rejection would have required the second reading to arrive somewhere it could be rejected from. It was translated, in the half-second between her daughter's mouth and her own ears, into a version she could survive hearing: a finite, fixable list of times a good mother had fallen short. "There is something wrong with how you relate to people" is a sentence very few self-images can survive intact. So it doesn't get heard as addressed to the person who holds that self-image. It gets heard as something smaller, something with edges, something a loving person can fix by trying harder next Tuesday.

The daughter left with an apology in hand and nothing changed. She had a folder of these conversations going back a decade, each ending in an apology, none changing a single later conversation. And she had no way to explain to her friends why this one hadn't helped either. "She said sorry. She meant it. Why does it still feel like nothing happened?"

Nothing happened. That's exactly right. The mother didn't refuse the message; refusal is a response, and this wasn't one. The message never showed up on her side of the table as the message that had been sent. It showed up as a different, easier one, and the easier one got a warm, competent answer. The critique wasn't rejected. It was never received.

The door correction walks through

Give this a name: anti-memesis. A critique becomes perceptually inaccessible to the person it's aimed at. It isn't misunderstood, disputed, or weighed and found wanting. It is never perceived as being about them at all. The identity capture that makes someone need the correction is the same thing that keeps the correction from arriving. It's a closed loop: the more a person's sense of self depends on not being what the critique describes, the less available the critique is to their perception, at exactly the moment they need it most.

This isn't limited to parents and adult children. The target can be an institution: a company whose leadership reads every critical audit as a request for a new policy memo, never as a question about how the company is run. It can be a movement whose members experience every internal dissent as an attack to manage rather than information to weigh. The mechanism doesn't care what kind of system runs it. It requires only that some identity — a self-image, a shared culture, a brand, a cause — has enough riding on not being the thing described that the critique gets rewritten before anyone has to face it.

Keep this distinct from Chapter 6. Contracorrigibility describes a system where the correction arrives and the machinery neutralizes it: the critique is detected, absorbed, or punished, and the system's own repair tools report that all is well. Anti-memesis sits upstream of that. The correction never registers as correction in the first place. One disables the fix. The other disables the noticing.

It helps to picture correction passing through three doors before it can change anything. It has to be perceived as being about you. It has to be evaluated and, if it holds up, accepted. And it has to produce action. A critique can be stopped at any of the three, and Chapter 9 lays out what each stop looks like and what it predicts. Anti-memesis is the first door, the one furthest upstream and the easiest to miss, because nothing dramatic happens there. No argument, no confrontation, no visible defense. The door just doesn't open, and from the outside that can look exactly like agreement.

Isn't this just unfalsifiable?

Anyone meeting this idea for the first time should raise an objection, and it deserves a full answer, because the concept's credibility rests on it.

The objection: how is "they can't perceive it" different from "they won't admit it"? Worse, isn't the diagnosis rigged so it can never be wrong? If the target agrees, a captured person can perform agreement. If the target disagrees, that's what a captured person would do. Either way the theory wins. That isn't a diagnosis. It's a trap with the door welded shut.

That's a fair challenge, and if there were no answer, the concept would deserve to be discarded. There is an answer, and it's specific enough to check.

Watch what the mother did with the correction, not whether she agreed with it. She didn't dispute it. She converted it. The structural message — "it's not incidents, it's the pattern of how we relate" — got reabsorbed into an incident-shaped one: "give me the list, I'll fix each one." That conversion is the signature. Call it the incident-vs-structure reabsorption pattern: a captured system takes feedback aimed at its structure and processes it as feedback about a discrete event, because incident-sized feedback can be handled without touching identity and structural feedback can't. The tell isn't emotional tone. The mother was warm. The tell is a mismatch you can point to on the transcript: correction-language is offered ("tell me, I'll fix it"), and the fix attempted is aimed at a different problem than the one named. When someone answers a structural critique by asking for the list of incidents, you have watched the reabsorption happen. That is a checkable event, not a mood.

The same pattern shows up in settings with no history and no love in them, which is why it's worth treating as a structural signature rather than a family quirk. Picture a manager told, in a routine review, "the team doesn't feel like you trust their judgment." A manager who has perceived the message asks what's driving that feeling and sits with an answer that might implicate how decisions get made in general. A manager caught in the reabsorption reflex says instead: "Tell me which specific decisions I should have looped people in on, and I'll change my process for those." It sounds like good management. But a bounded procedural question has quietly replaced the structural one, and the structural one never gets addressed, because it never got heard as the question.

Now the firewall, which matters as much as the discriminator: a target's silence, discomfort, or non-response is not confirmation of anti-memesis. Someone can go quiet because they're thinking, or hurt, or simply disagree and would rather not argue at your kitchen table. None of that is evidence. What licenses the diagnosis is the reabsorption pattern on the record — structure requested, incidents delivered — or a companion test from Chapter 6: can the person say what would change their mind? Someone who can specify, even roughly, what evidence would make them agree they'd been unfair is doing the work that anti-memesis disables. Someone who narrows every version of the question back down to "which time, exactly" is showing you the mechanism. Without one of those two things on the record, you don't get to call it anti-memesis. You have a person who disagrees with you, which is far more common and far less interesting. Confusing the two would turn a diagnostic tool into an accusation nobody can answer their way out of.

Two ways in

The block that stopped the daughter's sentence was entirely internal to her mother: one person's self-image, unable to survive one particular truth about itself. Call this egoic capture. The identity under threat belongs to one person, and the defense is native to that person. Nobody coached the mother to hear it that way.

There is a second version where the block isn't purely private. A surrounding culture does part of the work, not just failing to correct a misreading but rewarding it. Call this egregoric capture, after the egregores of Chapter 3: systems held together by shared belief and mutual recognition. A well-known example involves a novel and a film.

Fight Club — Chuck Palahniuk's 1996 novel, adapted by David Fincher into a 1999 film — follows a narrator who falls under the influence of a charismatic figure named Tyler Durden. Tyler preaches liberation through fighting and the rejection of ordinary life. Over the course of the story his ideas escalate into a paramilitary project, and the narrator ends up trying to stop it. The plot shows the reader where Tyler's philosophy leads the people who follow it.

A large share of the audience took something else away. Tyler's lines have been quoted for years as wisdom, and the character embraced as a model. The reception is well enough known that both creators have addressed it. Asked about it in 2023, Fincher said, "I'm not responsible for how people interpret things," and of viewers who miss what the character is, "I don't know how to help them." Palahniuk, asked in 2017, took a more relaxed view: once a work reaches its audience, it "will serve whatever purpose the audience has for it."

This chapter doesn't need to settle what either creator intended. It needs only the shape. A story whose final act has the narrator fighting to stop Tyler was received by many viewers as if Tyler were the lesson. And the misreading wasn't sustained by isolated viewers each defending a self-image, the way the mother was. It was sustained socially: friends quoting the same lines to each other approvingly, a loop in which the misreading was rewarded rather than corrected. No individual in that loop had to resist the story's plot. The surrounding culture had already done the reabsorbing, and each new viewer inherited the conversion ready-made.

Same structure, different source of the block. In the kitchen, the machinery that intercepted the message was one woman's self-protection. In the audience, part of the machinery was other people agreeing with each other, so that the intercepted reading felt like consensus rather than defense. This is also why better arguments can't fix either case on their own. The daughter's sentence was carefully built, true, and calmly delivered. The film wasn't unclear about where Tyler's project leads. Argument quality was never the variable. The variable was whether the message got to count as being about the listener at all, and in both cases a mechanism upstream of argument made sure it didn't.

The egregoric case is harder to do anything about. Correct one person's egoic block and, if it works, it stays corrected for that person. Correct one viewer's reading of Tyler Durden, and the culture that made the misreading feel like consensus is still there, ready to hand it to the next viewer. The block is distributed across a community of people confirming it for each other. Any fix has to work on the community's reward structure, not just on one member's understanding. That's a harder problem than a conversation at a kitchen table, and this book doesn't claim to have solved it. It claims only to have named where it lives.

Why this is worth naming

Most people have watched this happen and had no word for it. The friend who reads an article about narcissism and forwards it approvingly, oblivious that everyone else in the group chat is thinking of him. There's a public version, common enough to have its own shape. An organization, the subject of a damning piece of investigative reporting, responds not with denial but with something warmer and stranger: a statement thanking the reporter for "raising important issues," committing to "listen and do better." Each specific allegation is treated as an isolated process failure and assigned a named fix — a new training module, a new review step, a new role. The structural claim underneath, the one about how the organization actually operates, never gets a paragraph of its own. Nine months later a similar piece runs about a similar problem, and the same gracious response goes out again. Nothing in that sequence required anyone to be dishonest. It required only that the structural version of the critique never make it through the door.

That matters for anyone trying to correct a captured person or system. If a message that reads as an identity threat gets converted before it lands, sharpness is not the lever. A better-argued version of "there's something wrong with how you relate to people" meets the same door, because the door doesn't filter on argument quality. It filters on whether the message counts as being about the listener's identity, and a sharper argument makes that filter work harder. This isn't a case for going easy, and it isn't a verdict on anyone who can't hear it yet. It's a design constraint, like water running downhill. Sometimes it means finding the people inside a captured system whose sense of self was never fully absorbed into it — people with other commitments and other sources of worth to stand on, who can hear a structural critique without the floor giving way. Sometimes it means waiting for a moment when the identity is already under strain from something else and the door is briefly easier to open. It rarely means making the same argument more forcefully.

What this isn't

This concept is easy to misuse, so say plainly how. Anti-memesis is not a license to tell anyone who disagrees with you that they "just can't perceive the truth." That move fails the discriminator on its face. A person who hears your critique, understands it, and comes back with a specific, reasoned rebuttal has done the opposite of what anti-memesis describes. They perceived the message and evaluated it. That's a different door, and disagreeing with their rebuttal doesn't entitle you to declare them incapable of hearing you. Using the concept against someone plainly standing on the far side of the first door, arguing back, is exactly the bludgeon this book is trying to rule out.

It also isn't a fixed sentence. Anti-memesis is a characteristic consequence of identity capture, not a required one. If it were an iron law, nobody who'd been captured could recover, and people do recover. Some hear the structural version of a critique on the fifth telling, or the fiftieth, or after a change in circumstances — a divorce, a job loss, a friendship that ends — loosens the identity that was doing the intercepting. The concept describes a strong, recurring pattern, not a life sentence. Treating it as one would mistake a common failure mode for a permanent fact about a person.

A test you can run on yourself

This one is worth trying on your own reflexes. Think of a critique you've heard more than once, from more than one person, ideally people who don't know each other: a partner and a coworker, a sibling and a friend, saying some version of the same thing without comparing notes. Notice what your mind does with it. If the immediate move is "okay, but which specific time?" — if your attention goes straight to demanding an incident, a date, a quote, before you've let yourself consider that the pattern is the point — pause on that. It doesn't prove capture. Plenty of requests for specifics are reasonable; sometimes the pattern claim is wrong, and details are the only way to find out. The difference is whether you're asking for the incident to test the structural claim, or to avoid sitting with it. Usually only the person asking can tell which. It's the mother's reflex at the kitchen table, running in miniature, and noticing it while it happens is most of the skill this chapter has to offer. Nobody can open the door for someone else. Everyone can learn to feel their own door swinging shut.

Part II has now named four ways a system can fail. It can optimize the wrong thing while believing it's optimizing the right one. It can thrive by consuming what it depends on. It can fight off the correction it's shown. Or it can be shown the correction and never perceive that it was shown anything, so that the people trying to help walk away holding an apology that changed nothing, wondering what they did wrong. The honest answer is: nothing. The message never crossed the threshold where right and wrong could apply to how it was received.

Naming a failure is not the same as treating it. Part III turns to what a working diagnostic might look like, starting with the question this chapter has been circling: if the person or system most in need of correction is, by the nature of the capture, the worst positioned to administer it, who does that leave standing?

Sources

Contents · Glossary · How this was written