The Anatomy of a Social Media Pile-On

Or yeah, I wrote about being trolled on X. Whatcha gonna do about it?

“Let me never fall into the vulgar mistake of dreaming that I am persecuted whenever I am contradicted.”
— Ralph Waldo Emerson

Let’s start with the Emerson of it all, because it’s embarrassingly easy, especially if you live on the internet, to develop a persecution complex the moment there’s friction. Sometimes I’m wrong. Often, I’m careless. Sometimes I say something that lands badly because I put it badly, or because I’m writing from a place of privilege, or because I’m trying to be snappy and I overdo it, and I end up sounding like an idiot. If you put your thoughts in public, people get to push back.

What I want to describe is not this contradiction. It’s what happens after the contradiction, the moment where an innocuous (albeit stupid) post stops being a post and becomes an object other people can pick up and use. It stops being something you said and starts being something you represent. It becomes an informal little moral Rorschach test. People don’t respond to the words; they respond to the version of you those words allow them to conjure. And once that conjuring begins, the ritual goes out of hand quickly- it’s hard to control ghosts once you’ve brought them to life. You can almost feel the crowd deciding what the story is going to be, not because everyone coordinated, but because everyone is responding to the same incentive. Online, “sounds sure” gets mistaken for “is true.” It’s the finality of the tone that does it. If you can flatten a whole person into a verdict and declare it, you get the likes, you get the retweets, you get to feel like you “said it.”

This is where my recent Twitter situation ended up. The trigger was a tweet that was, at minimum, clumsy. The tone was wrong, framing wrong, incredibly underthought. I understand why it irritated people. If I could write it again, I would write it differently, and I’m not saying that to pre-empt criticism so much as to prevent the conversation from getting stuck in the boring binary of “HK is innocent” versus “HK deserved everything.” I’m not trying to litigate the original offence in any case. The part that scrambled my head wasn’t even the (well-deserved!) initial pushback; it was how quickly the pushback stopped being about the tweet and started being about my character, my life, my motives, my supposed history, and then, because the internet is bored and imaginative, about a bunch of things that never happened. If you’ll forgive the strange metaphor, I felt a bit like a cigarette being passed around at a house party. By the sixth or seventh person, no one knew who lit it, whose lighter was used, or if someone accidentally burned themselves on it.

I understand the social contract of visibility. I benefit from being visible; I like writing; I like being read; I like having opinions and putting them where people can argue with them. I’ve gotten opportunities and money and access because people know who I am. I’m not going to act like I stumbled into public life and now I’m shocked there are consequences. But I also don’t think we should pretend that every dogpile is “accountability” just because it uses the vocabulary of socially sanctioned morality. There is a difference between criticism and a crowd deciding you are contemptible and then reverse-engineering the moral language that makes contempt feel like it is a public good. There is a difference between “that was wrong” and “this person is rotten.”. That difference is what I want to write about, not because I’m a special snowflake this shouldn’t happen to (honestly, it should), but because it’s a pattern I keep watching happen, and once it happens to you, you understand that something else is going on under the language of ‘principled disagreement.’

The hatred comes first

There’s a theory of the internet that says pile-ons happen because people care about justice, and sometimes that’s true. Real harm exists. People with power do awful things. Communities need norms, and norms need enforcement. Callouts can be corrective, and sometimes the only reason certain patterns of abuse get interrupted is because of public consequence. I’m not claiming all outrage is fake, or that “accountability” is a myth invented by bored people who like fighting. I’ve seen the other side too, where everyone lets a bad actor keep acting badly because it’s inconvenient to confront them. I’m not nostalgic for a kinder, gentler world where no one is ever criticized. That shit was not actually kinder or gentler!

But I think online pile-ons have specific contours that make them so…strange? Unsettling? We don’t arrive at condemnation by thinking. We condemn, and then we go looking for the reasons that make it feel principled. Unfortunately, it’s not as theoretical and easy to understand when you’re the person being discussed like a case study (LOL).

Jonathan Haidt’s work on moral psychology basically argues that we don’t usually reason our way to moral judgment; we have a fast, intuitive reaction, and then we come up with the reasons that make the reaction feel principled. This is not a particularly flattering thesis, yeah? It basically says that a lot of our “ethical clarity” is just a gut feeling and “I think this might be true” stacked together under a trench coat. It suggests that what we experience as careful moral thought is often post-hoc storymaking; our brain scrambling to produce a narrative that makes what we already feel sound like it was always obvious.

There’s an old set of experiments in psychology that elaborate on this phenomenon. They give people a scenario that triggers immediate disgust, and then they remove all the obvious reasons you could cite for why the scenario is wrong. The classic one is consensual incest between adults, written in a way that eliminates coercion and harm. No pregnancy, no trauma, no social consequences, everyone involved is fine. People still instantly say it’s wrong, which is fair enough as a gut reaction; it’s a squirmy scenario. The interesting part is what happens next.

When the researcher asks participants to explain why it’s wrong, and then refutes their reasons one by one based on the details of the story, people don’t update their moral verdict. They get stuck. They reach for different reasons. They flail. They eventually say something like “it’s just wrong,” usually with a lot of indignation. The feeling stays even after the explanation collapses, which is the point: the feeling did not come from the explanation. The explanation came from the feeling.

That’s the piece I think we underestimate when we talk about pile-ons as if they’re purely rational reactions to wrongdoing. A lot of the time, the collective feeling arrives first: disgust, contempt, the little thrill of taking someone down, the adrenaline of being in a group that has agreed upon who the villain of the day is. And then the moral reasoning gets constructed around that feeling to make it feel like a rational conclusion. This is also why the reasons can get so slippery. If you’ve already decided someone is contemptible, you can find a reason that fits. If one reason stops working, you can find another. If the original offence is small, you can expand the story until it feels appropriately criminal. If you need the person to be worse than what you can prove, you can start using suggestion and insinuation. And because everyone else is doing it, it feels not as much ‘you’re being cruel’ and more like you’re participating in community hygiene.

This is where the apology discourse becomes a bit silly. People talk about wanting responsibility as if the repair is the point, like if the person apologized “properly,” the mob would be satisfied, and everyone would go home. But that’s not how it works once a pile-on becomes a spectacle. An apology can’t become repair if repair isn’t what people actually want. If the real reward is the feeling of righteous contempt, then an apology is just new material to add to the list of crimes committed. It becomes proof that you were guilty. Or proof you’re manipulative. Or proof you’re weak. Or proof you’re only sorry you got caught. It doesn’t matter which one; the interpretation will be selected to preserve the emotional verdict, because preserving the verdict is the point. You can almost feel the story protecting itself. You can almost feel the crowd needing you to stay “bad,” because if you became “human,” the whole structure would collapse into something much more awkward: a group of people who enjoyed being mean.

There’s also a scaling problem that makes this worse. A single person being outraged at you can be reasonable, proportional and even helpful. But outrage is an emotion that mutates as it multiplies. When thousands of people do it at once, it stops looking like answerability and starts looking like a swarm. There’s research that describes this as the “paradox of viral outrage”: condemnation that seems morally appropriate in isolation starts to read as bullying when it becomes viral, and the scale can actually generate sympathy for the person at the center of it.

This is the part where the story often has to escalate, because once the response looks disproportionate, the mind needs the target to deserve it. The punishment begins to outgrow the crime, and the only way to make that feel consistent is to inflate the crime.

So the allegations multiply. “This tweet” becomes “this pattern.” “This pattern” becomes “this is who she is.” And once the crowd is in that register, rumours are a sharper tool in the arsenal. They don’t have to be true to be useful. They just have to feel plausible inside the framework of the story that’s already been agreed upon. They fill in the gaps. They make the punishment feel earned. They turn the discomfort of “are we overreacting?” into the relief of “no, actually, she’s worse than we thought.”

I’m not saying this to absolve anyone of anything, least of all myself. This time, I think I did deserve it.

I’m saying it because I think the sequence matters. If hatred comes first and reasons come later, then the confidence with which someone can list their reasons isn’t necessarily proof of their rightness. Sometimes it’s proof they’ve had time to stew and build a case. Sometimes it’s proof they’re enjoying the act of building it. Sometimes it’s just proof that the press secretary is good at her job.

After the verdict: when context stops working

The strangest part of a pile-up isn’t even the anger. It’s the confidence! It’s genuinely incredible how quickly people move from “I didn’t like that” to “I know exactly who you are.” And once that certainty takes hold, language stops functioning normally. Context stops functioning normally. You learn very fast that you cannot “clarify” your way out of a story people are enjoying telling. You cannot “explain” your way out of captured imaginations. You cannot “be reasonable” in a room that is trying not to be reasonable. Which sounds obvious when I type it, but it is genuinely hard to accept when you’re inside it, because it goes against the childish, stubborn part of you that still thinks: if I can just say the right thing, this miscommunication will clear up.

What actually happens is… the story sets into a concrete shape that’s nothing you asked for. And after that, nothing you do gets read on its own terms. People aren’t asking “what did she mean?” anymore. They’ve already decided what you mean, and now they’re just collecting things that match. That’s when the digging starts. Old tweets, screenshots, jokes, captions, anything that can be pulled out and held up like “see?” And it’s not “see what she said,” it’s “see who she is.” Once you’re in that zone, everything becomes fair play. A joke becomes a tell. A clumsy sentence becomes “this is how she always is.” The details don’t even have to agree. They just have to keep you pinned.

This is where “tone” becomes a weapon. The tone you wrote in becomes your personality. Your personality becomes your intent. Your intent becomes your crime. People will tell you what you meant with this serene confidence that makes you want to laugh and cry at the same time, because you’re sitting there thinking, I literally wrote it. I know what I meant. But then you realize: what you meant is irrelevant once you’ve been assigned a role. You’ve been cast. Now people are reacting to the character. And the character is much easier to hate than the reality of a person who said something poorly and maybe meant something different and maybe has blind spots and maybe also isn’t a cartoon villain.

And then there’s the word that does the most heavy lifting in these moments: “pattern.” “Pattern” is magical because it makes everything sound principled. It takes a bunch of scattered moments, some real, some stretched, some fully imagined, and it weaves them into something that feels like a hard-earned insight instead of vague irritation. It’s saying “I’m not being petty, I’m being perceptive.” And sometimes that’s true! Sometimes pattern recognition is literally how you protect yourself from people who keep doing the same harmful thing. But online, “pattern” is also what people say when they’ve found enough screenshots to support the feeling. The screenshots don’t have to prove anything. They just have to look like they might, in the way that lets other people nod along and go, “yeah, that tracks,” which is basically the internet’s version of due process.

The rumour part is what I can’t stop thinking about. When someone is already decided to be bad, rumours aren’t rumours. It’s more like “Of course.” “Yeah I heard that too.” “This explains everything.” And once a rumour has that kind of emotional usefulness, it doesn’t matter if it’s true. It matters if it’s compatible. That’s why the stories get more baroque over time. They’re being refined for narrative satisfaction. Each new claim isn’t meant to inform; it’s meant to justify. It’s meant to make the existing level of rage feel earned and valid. It’s meant to make the punishment feel proportionate. If the punishment is already huge, then the crime has to be huge too, even if you have to build it out of rumours and inference and “my friend said.”

I keep coming back to how normal this feels to the people participating. That’s the scariest part. It doesn’t look like cruelty to them. It looks like community. It looks like doing the right thing. It looks like being on the correct side. It looks like “we’re holding someone accountable.” And because it looks like that, nobody has to sit with the truth underneath it, which is: some part of this is fun. Some part of this is relieving. Some part of this is about power. If the target is rotten, then your contempt becomes moral. If the target is dangerous, then your aggression becomes protective. If the target deserves it, then you don’t have to examine the part of you that enjoys watching someone get humiliated.

Why “I just don’t like her” isn’t enough anymore

One of the weirder shifts of the last few years is that “I don’t like her” is no longer considered a complete thought. It reads as suspicious now, like you’re confessing to something childish. You’re supposed to have a better reason. You’re supposed to be able to justify it in a way that sounds grown-up and principled and socially useful. Dislike has to be upgraded into something that looks like public service.

And I get why. “I don’t like her” is silly. It forces you to admit that you’re having a human reaction to a human being, and human reactions are often petty, jealous, defensive, and inexplicable. Not the kind of thing you want to claim, especially to an audience that is always scanning for hypocrisy. So we do what we always do: we clean it up. We launder it. We turn it into morality.

There’s actually research that basically backs up this intuitive difference between plain old dislike and the kind of hate that shows up in online pile-ons: hate isn’t “stronger dislike.” It’s dislike that’s been moralized. People don’t feel negatively; they feel righteous about the negativity. The moral framing is the whole point, the thing that makes hatred feel like a virtue rather than a vice. And once you notice that, you start seeing it everywhere: the way criticism escalates from “I didn’t like that tweet” to “this person is harmful,” “this person is dangerous,” “this person should not have a platform,” as if disliking someone isn’t valid unless it can be justified as harm reduction.

This is where the internet gets to do its favourite magic trick: taking a feeling and turning it into a test. Because if disliking someone is framed as moral discernment, then liking them becomes morally suspect. Supporting them becomes complicity. Not joining in becomes endorsement. The emotional reaction spreads outward and recruits other people, not into agreement, but into performance. You’re not allowed to think someone is a dick; you’re expected to demonstrate that you, too, can recognize the dickishness, and that you, too, are on the correct side of it. “I don’t like her” becomes “I’m a good person.”

There’s a term psychologists use, moralization, for the process where something that used to be a preference becomes a value. Where “I don’t like this” turns into “people shouldn’t do this.” Where a personal reaction becomes an obligation and then, eventually, a kind of purity ritual. Rozin and colleagues wrote about this years ago, and what’s striking is how closely it matches what we see online: once something becomes moralized, people feel licensed to censure it, and they also start recruiting reasons to support their position. Not because the reasons caused the moralization, but because the moralization makes the reasons feel necessary. It’s like your brain and your community both agree that “just a feeling” isn’t enough, so now you need a case file. You need a story. You need a reason that will stand up in public.

Which means the internet becomes a place where a huge amount of emotional life gets disguised as ethics. Envy gets disguised as structural critique. Personal irritation gets disguised as activism. A desire to punish gets disguised as reckoning.

The most common example is how quickly “this annoyed me” becomes “this is unsafe.” “Unsafe” is one of those words that started as a serious descriptor and now gets used like a particularly cheap garnish. Sometimes it’s accurate. Sometimes it’s saying “I felt something and I don’t want to assess it.” Sometimes it’s “I don’t like the power dynamic here” without having to admit the underlying feeling is envy or intimidation or social comparison. It’s turning discomfort into authority. It lets you treat reaction as evidence.

And honestly, who wouldn’t want that? It’s exhausting to admit you’re being petty. It’s much easier to say, “This person is bad.” Once they’re bad, your feelings don’t need interrogation. They need expression. Once they’re bad, your cruelty can be reframed as boundary-setting. Once they’re bad, you don’t have to deal with the more complicated possibility: that you dislike them for reasons that have nothing to do with morality, and everything to do with you.

This is also why pile-ons can feel so… gleeful, even when everyone is insisting they’re doing something serious. Morality gives people permission. It’s the permission slip that says: you are allowed to be harsh, you are allowed to be contemptuous, you are allowed to strip someone of complexity, because this isn’t you being mean, this is you being right. And once a crowd gets that permission slip, it’s very hard to take it away, because taking it away doesn’t challenge a behaviour, it challenges an identity. It asks people to consider that they might have been enjoying something ugly.

So the moralization has to stay intact. The target has to stay bad enough to justify the scale of the response. Dislike has to stay dressed up as duty. And that’s why the story escalates. That’s why rumours “track.” That’s why people keep saying “pattern.” It’s not only because the internet is cruel. It’s because the internet is full of people trying to experience their emotions without taking responsibility for them.

Why it feels so good to be part of the crowd

Pile-ons are fun for the people participating! A very solid “guilty pleasure,” the way it’s fun to gossip, the way it’s fun to be in on something, the way it’s fun to feel like you have power over a situation that has nothing to do with you. And you can watch people working very hard to deny this to themselves in real time, because admitting it would collapse the story. If it’s fun, it can’t be purely righteous. If it’s fun, then maybe we are not ‘protectors’, maybe we’re also enjoying the spectacle. If it’s fun, then the line between “good people calling out bad behaviour” and “crowd enjoying cruelty” gets blurry, and blurriness isn’t very convenient!

And then there’s the social reward, which is honestly the bigger drug than the cruelty itself. A pile-on gives you instant belonging. You don’t have to build intimacy with people, you don’t have to earn trust, you don’t have to be interesting. You just have to pick a side. You post the right reaction, in the right tone, and you get the little hits: likes, retweets, replies, “exactly,” “this,” “drag her.” It’s a tawdry version of community, but it still feels like community. It tells you you’re safe inside the group. It tells you you’re safe today, because you’re pointing at the correct person.

I think this is why pile-ons always have a performative edge. Even when the original criticism is valid, the swarm quickly turns into a stage. People aren’t expressing disapproval; they’re auditioning. They’re trying to write the dunk that gets screenshotted. They’re trying to be the voice that sets the framing. They’re trying to be seen by the in-group as a person who gets it, who has the right politics, the right instincts, the right moral disgust. And once you’re on that stage, the incentives get ugly fast, because subtlety doesn’t win. Nuance doesn’t travel. The thing that travels is the most confident, most contemptuous, most final version of the judgment.

This is also where envy and resentment sneak in. A public target is usually someone with something: attention, perceived influence, perceived money, perceived access, perceived ease. Which means a pile-on becomes an opportunity to rebalance status, to pull someone down to a level that feels less threatening. There’s a whole thing in social psychology often called tall poppy syndrome- the impulse to cut down someone who’s visibly thriving because their thriving makes you uncomfortable. The justification is usually framed as fairness, or humility, or answerability, but the emotional core is simpler: “why does she get to have that?”

This is also where privilege accusations become a favourite weapon, not as analysis, but as delegitimization. If she’s succeeding and I dislike her, then her success must be illegitimate. It’s the rhetorical crowbar you use to pry someone’s status away.

The last payoff is that participating in a pile-on can make you feel morally superior. You get to be part of a moment. You get to locate evil, stupidity and badness outside yourself, and then bond with other people around the shared act of rejecting it. It’s purity culture with better distribution and branding. And the reason it feels so good is because it solves, temporarily, the exhausting modern problem of not knowing whether you’re a good person. If you can hate the right person, you don’t have to sit with your own ambiguity.

The impossible response

Being the target of a pile-up isn’t “getting criticized.” It’s getting shoved, very abruptly, into a game where every move you could make has already been pre-interpreted, and not in a generous way. You don’t get to respond like a normal person responding to a normal misunderstanding, because the room you’re responding into isn’t behaving like a room full of individuals; it’s behaving like a single organism with a script, and the script has three lines for you: if you explain, you’re defensive; if you apologize, you’re manipulative; if you stay quiet, you’re arrogant. Pick one. Choose your fighter. Either way, you’re still the villain of the day.

Your first instinct is to clarify, not even because you’re trying to “win” but because you still have this stubborn belief that language is supposed to do something. You think: I’ll just say what I meant, I’ll name what I got wrong, I’ll add the missing context, and surely the temperature will drop by a few degrees, because that’s what happens in real life. Except online, clarification isn’t clarification.It’s a refusal to take responsibility. It’s you trying to talk your way out of consequences. Half the people reading aren’t even reading for meaning; they’re reading for posture, for tone, for whether you sound sufficiently bowed. So you end up with this insane situation where the more precise you try to be, the more people act like you’re doing a lawyerly dance, and you start sounding guilty purely because you’re trying not to be misunderstood.

Then you think, okay, apology. Because apology is the approved ritual. It’s what everyone demands with this intense confidence, as if the only reason mobs exist is because people aren’t accountable enough, and once you offer the correct amount of reckoning, everyone will stand down. But apologies online are content. They get screenshot. They get quote-tweeted for “tone.” They get audited for missing clauses. And if you’re careless, you end up apologizing to the loudest interpreter of your words rather than to the people you actually affected, which is such a weird position to put yourself in, because now you’re responding to a version of your intent that doesn’t belong to you. It’s not “I’m sorry I did X,” you’re saying “I’m sorry you think I did X,” except you have to phrase it in a way that doesn’t sound like “I’m sorry you think,” because that gets called a non-apology, and suddenly you’re playing this stupid semantic game in public while the actual situation gets farther away from you.

So you go, fine. Silence. Let it blow over. Don’t feed it. And then silence gets its very own story attached to it, because of course it does. Silence becomes proof you don’t care, or proof you think you’re above criticism, or proof you’re “hiding.” And the funniest part is how quickly the internet can turn into a mind-reading machine. If you block people, you’re thin-skinned and censorious. If you don’t block people, you’re letting harm happen in your mentions. If you log off, you’re running away. If you stay on, you’re centering yourself. If you make a joke, you’re flippant. If you don’t make a joke, you’re grim and self-serious. Every option becomes evidence of the worst possible intention, because once someone is committed to a story about you, your behaviour is no longer behaviour; it’s “signals.” You can’t be a person fumbling through an unfortunate moment. You have to be a character committing to a plot.

And then something small and humiliating starts happening in your head. You start rereading your own words to pinpoint your faults. You keep thinking you’ll find it, like there will be one obvious line where you can go, ah, yes, that’s what they’re reacting to, and if I correct that line, the whole thing will become normal again. You scroll your own tweets like they’re evidence in a case you didn’t know you were on trial for. You start pre-empting interpretations before they arrive. You start writing imaginary replies in the shower, the one paragraph that will finally make everyone go “oh, fair,” even though you already know the paragraph doesn’t exist. The most embarrassing part is how earnest this urge is. How badly you want there to be a fix that involves you being articulate enough, humble enough, clear enough, human enough, so that the world feels less crazy.

But the pile-on isn’t really a misunderstanding that can be corrected. It’s momentum. It’s a bonding event. It’s a hundred people trying to write the definitive caption under your face, and the best caption wins. So you find yourself watching strangers narrate your motives back to you with this confidence that makes you feel insane, because it’s not an argument with what you said anymore; they’re arguing with what they think you “are,” which means you’re no longer allowed to be inconsistent, or clumsy, or half-formed, or learning in public, or any of the things we supposedly tolerate in humans. You’re either redeemable or irredeemable, and the crowd usually decides that early, and then spends the rest of the time acting like it was a conclusion reached through careful analysis.

This is why people disappear after they get dogpiled, and I don’t mean “disappear” in the dramatic influencer way where they take a week off and come back with a better content strategy. They stop posting, they stop sharing, they stop taking risks, they stop being visible, because at some point, your name stops feeling like yours. It becomes a thing other people pass around, reshape, improve for drama, repeat like trivia. It doesn’t even have to be accurate to become sticky. It has to be repeatable. And once it’s repeatable, it’s real enough to ruin your week, your reputation, your relationships, your ability to sit in a room without wondering who saw what and believed what and decided what about you without ever speaking to you.

The most irritating part is that the whole experience also tempts you into your own version of moral laundering. It makes you want to frame yourself as purely misunderstood, purely victim, purely innocent, because that’s the only role the internet really understands on the other side of this: either you’re the villain who deserves it, or you’re the saint who was wronged. And I don’t want to write this essay from either costume. I can be wrong, and people can be right to criticize me, and the pile-on can still be its own beast with its own incentives and its own ugliness. Holding all three is inconvenient. It’s not tweetable. But it is what it is.

What to do instead

This is not an answer, and if I pretended it was, it would be just another way of claiming clarity, which is kind of the whole problem. But I do have a few rules I try to follow when I’m the one being piled on, and also when I’m the one tempted to join a pile-on, because that temptation is real and pretending you’re immune to it is how you accidentally turn into the exact type of person you hate.

The first rule is boring and annoying: I don’t read everything. Not because I’m above it, not because I’m “protecting my peace,” not because I’m spiritually evolved, but because my brain is extremely suggestible when it’s stressed. If I read two hundred strangers confidently describing my motives, my brain will start offering up those motives as plausible, even when they’re nonsense. If I read ten different versions of what “everyone is saying,” I will start believing that “everyone” is a coherent entity with a stable opinion, when actually it’s just a loud subset of people performing for each other. You don’t realize how much your mind craves social consensus until you’re watching a fake consensus being manufactured around you in real time. So I limit exposure. There is a point past which data stops being information and becomes contamination.

The second rule is that I try not to negotiate with the internet in public. This is the hardest one for me, because I love words, and the idea that words won’t save you makes me sad. But once the verdict has set, you’re dealing with a spectacle. Clarifying starts to look like pleading. Defending starts to look like doubling down. Apologizing starts to look like an insincere performance. So when I do respond, I try to respond once, plainly, to the people who are actually affected or actually engaging, and then I stop.

The third rule is the one I wish the internet understood: I keep the critique at the size of the offence. This is true whether I’m the one being criticized or the one criticizing. If someone says something clumsy, I try to treat it as clumsy. If someone is being careless, I try to treat it as careless. I try not to do that thing where I reverse-engineer someone’s soul from a screenshot. There are people in this world who are genuinely malicious, and there are people who are genuinely harmful, and there are people who should not be protected by “benefit of the doubt” because they have proven, repeatedly, what they are. But most people online aren’t those people.

The fourth rule is: I don’t confuse moral language with moral action. This one is for me as much as anyone else. Calling someone “harmful” is not harm reduction if it’s just a way to get social approval for being mean. Declaring someone “unsafe” is not community protection if there’s no protection happening, you’re just escalating a mob. Even when criticism is valid, the performance layer creeps in so quickly. If I’m not willing to be accountable for what I’m doing, if I’m not willing to say “I’m angry” or “I’m hurt” or “this feels unfair” or “this reminds me of something personal,” if I’m only willing to speak in the language of righteousness, then I’m probably laundering something. And if I’m laundering something, I’m not as pure as I’m pretending to be.

The fifth rule is that I take the conversation offline when it matters. If I actually harmed someone, I’d rather speak to the person than to the crowd. If someone is criticizing me in good faith, I’d rather talk like actual people rather than trade statements like rival PR teams. The internet encourages you to keep everything public because publicness is the currency, but publicness is also what makes repair impossible. Repair requires privacy and time and the option to be awkward without it becoming another piece of content. This doesn’t mean everything should be hidden. It means that “public consequence” and “private repair” aren’t synonyms, and treating them like the same thing is why everyone ends up stuck in these endless loops of performative repentance and performative punishment.

And finally, the rule that keeps me from turning into a cynic: I don’t let the crowd become my compass. When you’re visible, you can start making decisions based on the general temperature of your mentions, and you don’t realize it’s happening until you look up and you’ve become a person who is always pre-flinching. You start writing around the mob. You start rounding off your edges. You start only saying what you know will be approved. And then you wake up one day with a platform you built by being yourself, and you don’t recognise yourself.

So the best “what to do instead” I have is: refuse to shrink, but also refuse to harden. Don’t become the kind of person who thinks every contradiction is persecution, and don’t become the kind of person who needs other people to be morally reprehensible in order to justify your contempt. Try to keep your criticism specific. Try to keep yourself honest. Try to keep your humanity intact.

In Closing

Emerson’s line is still the one I want to hold onto, partly because it’s true and partly because it’s humiliating. It is genuinely vulgar to treat every contradiction as persecution. Being disagreed with is not a tragedy. Being criticized is not a human rights violation. Sometimes you say something stupid and the appropriate consequence is that people tell you it was stupid, and you take the L, and you write better next time.

But the reason I started with that quote is that it makes room for a second truth without letting me slip into martyrdom. You can accept contradiction as part of the deal and still acknowledge what a pile-on is. You can be accountable for your clumsiness and still refuse to pretend that every dogpile is justice. You can admit that the internet has built a very effective machine for converting ordinary dislike into moral certainty, and then using that certainty to license cruelty at scale.

I signed the social contract of visibility. I like being read. I like writing in public. I like having opinions that people can argue with. I’ve benefited from attention in ways that are real and material and I’m not going to play innocent about that. But I didn’t sign up to be turned into a communal object every time I misstep, to have my intent rewritten by strangers, to have rumours become “pattern,” to have contempt dressed up as civic duty and then handed back to me as if it’s good for my soul.

Maybe this is just the price. Maybe this is the tax you pay to be a person with a platform, and if you don’t want it, you should log off, go private, write in your diary, live a quieter life. I’m not even saying that’s wrong. I’m saying: if this is the price, then we should at least be honest about what we’re paying for. Because “accountability” is too easy a word for what’s often happening. It flatters the crowd. It hides the pleasure. It absolves the cruelty. It turns the mess of envy, insecurity, boredom, genuine principle, genuine harm, and genuine group psychology into a story where everyone is a hero, and the target is a lesson.

And I don’t want to be the kind of person who thinks she’s persecuted whenever she’s contradicted. But I also don’t want to live in a world where contradiction automatically metastasizes into moral prosecution. If nothing else, I want to keep one truth: sometimes you don’t like someone, and that is not the same thing as them being a horrible person.

It’s ok to just dislike someone. It’s perfectly ok.