A kid at school has hair that sticks up like a brush. So they call him Brushy.
Everyone finds it hilarious. It's genuinely funny — the observation is accurate, the word is good, and the whole class is laughing. They'll still remember it at a reunion in twenty years.
He loses his entire self-esteem.
Pashupati tells this in about fifteen seconds and it's the sharpest thing in the section: they start calling him Brushy, and he loses his whole self-esteem. But everyone else is having a great laugh, and they remember it after twenty years also.1
That's the failure mode with the mechanism showing. Nothing about the joke was badly made. It worked exactly as jokes work — accurate observation, compressed into a word, memorable enough to survive two decades. The craft was fine. The base was missing.
He gives it a name, and the name is doing real work:
Hāsya without śṛṅgāra is offensive, man. People get offended, big time. Great way to get cancelled — hāsya without śṛṅgāra.2
And then: hāsya without śṛṅgāra is like Asuri Hāsya. It's a mean hāsya. Insulting hāsya — where you have to roast someone. That's not funny. It's just not funny.3
Asuri points back to the origin story in the first part: the asuras attacking the stage. The demonic version of the rasa. Not a lesser humour — a different one, running the same machinery for the opposite purpose.
The positive version he states as a preposition change: laughing with, not laughing at.4 Hāsya with śṛṅgāra is laughing together, not laughing at someone alone.
This is the part that isn't obvious and is the page's actual contribution.
You'd expect the guard against cruel humour to be karuṇā — care, kindness, don't hurt people. That would be the intuitive pairing and it's what most ethical accounts of comedy reach for.
He doesn't say that. He says the base is śṛṅgāra: if you haven't captured your organs with your śṛṅgāra and you try to do hāsya, it'll become offensive.5 And: if your śṛṅgāra is high, then the kind of hāsya you do will also be tasteful. You won't put anyone down. You'll just make it funny.6
Which means the guard isn't moral. It's perceptual.
Follow the mechanism. Humour of any kind requires noticing an incongruity — the thing that doesn't fit, the unexpected turn, the mismatch. If your senses are open, incongruities are everywhere: in language, in objects, in situations, in the shape of a sentence. You have an enormous supply and no reason to reach for any particular one.
If your senses are shut, you have almost no supply. And the one source of incongruity always available, requiring no perception at all, is a person's deviation from a norm. Hair that sticks up. A way of walking. An accent.
So the roast isn't a moral failure that a kinder person would avoid. It's what humour looks like when produced by someone with nothing else to work from. Cruelty is the default output of a closed sensorium trying to be funny — which is a much more useful diagnosis than "be nicer," because it points at a fixable upstream condition rather than at character.
What he offers instead is entirely structural, and his example is deliberately trivial.
Someone says no need to don't worry. A double negative that lands wrong, and he's still laughing about it — should I worry, or should I not worry?7 Another: I cannot be able to.8
Nobody is the target. There's no deviation from a norm being punished. The incongruity is in the language itself, and it's available to anyone paying attention to how sentences actually behave.
It's just stimulating, by the sheer unpredictability of your words. It just makes people laugh.9
That's the with-śṛṅgāra version: incongruity harvested from perception rather than from a person.
He also names the boundary against a near neighbour — being clever is not being funny. A clever play on words doesn't make it funny; hāsya needs playfulness. If you're sitting with a reverse dictionary working out which words to use, you don't have it.10 Cleverness is constructed. Playfulness is perceived.
The Brushy case has a detail that's easy to skip and does most of the moral work: everyone else is having a great laugh.
The laughter is real. Nobody in that room is faking it, and nobody experiences themselves as participating in cruelty — they're experiencing themselves as amused, which is a passive thing that happens to you.
That's the mechanism's cover, and it's what makes asuri hāsya so much more destructive than an insult. An insult is legible: it comes from one person, it can be objected to, and the room can decline to join. A joke arrives as a shared event with no author. The laugh is distributed across everyone present, so no individual is responsible for it, and the target is faced with a room rather than an antagonist.
Which is why the twenty-year memory is bilateral in an ugly way — the class remembers it as a funny thing that happened, and he remembers it as what he became.
For a creator with an audience, the relevant translation is that you are not making a joke, you are convening a room. The people laughing along are having a mild pleasant experience they'll forget. Whoever is on the other end is having the load-bearing one. The asymmetry is total and entirely invisible from where you're standing.
He frames the cost as getting cancelled — great way to get cancelled — and it's worth naming that this is the weakest part of his account.
Cancellation is a consequence to the joker. It's a career risk, and framing it that way makes the whole thing a question of prudence: be careful, that kind of humour will cost you.
But he's just told the Brushy story, which shows he knows what's actually at stake. The cost lands on a child who lost his self-esteem, and the class faced no consequence whatsoever — that's precisely why the pattern is stable.
The two framings recommend different behaviour under different conditions. Prudence says: be careful when there's a risk of being caught, which means the calculation flips the moment you have a sympathetic audience or the target has no platform. Perception says: this is what you produce when you're depleted, and it's bad output regardless of whether anyone objects.
Only the second holds when nobody's watching, and the vault takes the second as the real doctrine — supported by his own example over his own framing.
You're editing the newsletter and the funniest line in it is about a specific person.
Not named. But anyone who follows this corner of the industry will know exactly who, and the line is good — it's the sharpest sentence in the draft and you knew it was good the second you typed it.
So before you decide anything, you run the diagnostic backwards: where did this come from?
And you can trace it. You wrote it Thursday afternoon, at the end of a bad day, after reading something that annoyed you. You were not in a rich state. You were in a thin one, and the one incongruity available to you was a person being ridiculous.
That's the tell. Not the content of the joke — its provenance.
You don't delete it yet. You try the harder thing first: can you find a funnier line in the same paragraph that isn't about a person? Fifteen minutes. And you can't, because you're still thin — which is itself the answer.
So you leave it and come back Saturday, after a morning outside, having read something good and cooked something properly.
And on Saturday the sentence reads differently. It's still accurate. It's also, you can now see, slightly pleased with itself in a way you couldn't detect on Thursday. And within ten minutes you've got a better line about the situation rather than the person — funnier, because the situation is stranger than the person is.
The rule that comes out of it: don't judge the joke, judge the state you were in when it arrived. Cruel lines don't announce themselves as cruel. They announce themselves as the best you had.
The most useful case is one where the humour does have a target and works anyway, because it shows the line isn't simply "never be about a person."
Pratish had sent him a video — Pratish being angry about things not being aesthetic. Pashupati watched it at midnight after intense Devi sādhana, during an eclipse, and was, in his phrase, cracking up.11
His reading: somebody with high energy releasing their frustrations is funny. And the counter-example he reaches for is Hanumān being scolded by Rāma and enjoying it — what is this Sanskrit he's calling me? Oh my god, what did he call me? — the unexpectedness tickling the person being scolded.12
So the rule extends: be like Hanumān. Even if it's calling someone out, they should get tickled. They should see your frustration and get tickled.13
The discriminator is what the target experiences. Brushy loses his self-esteem. Rāvaṇa — in the version he's telling — is delighted. Same structural act, opposite reception, and the difference isn't the content of the insult. It's whether the person is being diminished or being included in something that's overflowing.
High energy releasing frustration reads as vitality. Low energy releasing frustration reads as an attack — because that's what it is, and everyone can feel the difference without being able to name it.
Evidence. The perceptual account explains something the moral account can't: why cruel humour clusters in people who are tired, embattled, or working under pressure, rather than in people who are simply unkind.
Tension — the discriminator is internal again. Did they get tickled or diminished is checkable in principle and unavailable in practice at scale. A newsletter's target isn't in the room.
Tension — cancellation as the stated stake. He frames the cost as getting cancelled, which is a consequence to the joker rather than to the target. The Brushy example shows he knows the real cost. The framing doesn't.
Open question. Is there humour that requires a target and is fully legitimate? Satire against power has a strong claim, and nothing in his account distinguishes punching up from the roast — both are hāsya aimed at a person.
Set this beside The Thirteen Vīthī Types — Comedy Architecture and you get the two halves of a comedy education that rarely appear together.
The vīthī material is a taxonomy — thirteen named comic devices, catalogued in the Nāṭyaśāstra, usable as a beat structure. It tells you what the moves are.
This page tells you what state produces good moves, and asserts that the same taxonomy in a depleted person yields the asuri version. The devices are neutral; the operator isn't.
That's a genuine addition rather than a rival. A comedy writer with the taxonomy and no state has thirteen ways to construct something mean. And it explains a pattern anyone who has worked in comedy rooms recognises — the technical skill stays constant while the output curdles over a long, tired season.
Against Krisang's material there's a gap worth naming: he has no comedy pages at all. Hāsya is one of the nine in his Rasa Maṇḍala, assigned to Gyāna-Śakti alongside Śṛṅgāra and Adbhuta — which is the one triad where teacher and student agree completely — and he develops the other two while leaving this one nearly untouched. Whether that's oversight or discomfort isn't recoverable.
To Manipulation and Influence Hub — behavioral-mechanics
The influence corpus documents ridicule as one of the oldest compliance instruments available. Public mockery enforces norms at almost no cost to the enforcer, it's socially deniable — it was a joke — and it produces durable behavioural change in the target and in everyone watching. The cult-formation material documents it as a control mechanism; the propaganda corpus documents it as a tool against out-groups.
Brushy is that mechanism operating in a classroom with nobody in charge of it.
The insight the pairing produces is about deniability as a structural feature rather than an excuse. Ridicule's power comes precisely from its ambiguity: the target cannot object without appearing humourless, which means the enforcement carries a built-in gag on the person being enforced against. That's not incidental — it's what makes it more efficient than direct sanction. Which reframes the śṛṅgāra requirement in a way Pashupati doesn't reach: the reason the with/at distinction matters so much is that "at" humour silences its target by construction, and a creator with an audience running "at" humour is operating a compliance mechanism whether or not they intended one.
To Somatic Trauma Theory Hub — psychology
Shame's clinical signature is global self-evaluation producing concealment and withdrawal, and the developmental literature identifies peer humiliation in childhood as among its more reliable sources — precisely because it arrives with an audience, which converts a private failure into a public identity.
Brushy is a textbook instance and the twenty-year memory is the tell.
What holding these together produces is a bridge back into this corpus's own Kriyā material. That cluster establishes that bībhatsa without karuṇā destroys self-esteem, and names the education system as the case. This page identifies the delivery vehicle: the mechanism by which schools actually do it is very often not the teacher's judgement but peer humour, running unsupervised, at a target selected for perceptible deviation. Which means the two failure modes in this framework — shaming without care, and humour without perception — converge on the same event, and the vault's account of the education system is incomplete without both.
Sharpest implication. Cruel humour is not a moral failure that a kinder person would avoid — it is the default output of a closed sensorium attempting to be funny, because a person's deviation from a norm is the one incongruity available to someone perceiving nothing else. Which relocates the intervention entirely: you don't fix it at the joke, and you don't fix it by resolving to be nicer. You fix it upstream, by having something else to notice.
Generative questions.