A parenting WhatsApp group. Two people going back and forth about screen time, both getting sharper, both getting louder, neither one moving an inch. You've seen this a hundred times — maybe you've been one of the two people. Krisang was watching it happen in his own community and did something most people don't do in that moment: he stepped in and told both sides to stop trying to win.1
Not stop arguing. Stop trying to win. There's a difference, and the difference is the whole page.
Here's the actual instruction Krisang gives, and it's stricter than "listen before you speak." Whenever you start a debate, the first move isn't your point. It's their point — stated so completely that they don't need to explain it to you anymore.2 Not "I hear what you're saying." Not a paraphrase good enough to move past. A restatement good enough that the other person thinks: this guy knows my position better than I do.
That's the test. Not "did I understand them." Did I understand them better than they understand themselves. Most people fail this test constantly and don't notice, because the bar they're actually clearing is much lower — "I let them finish talking before I responded." Letting someone finish is not the same as being able to make their case for them.
This has a name in the tradition Krisang is drawing from: pūrva-pakṣa, the "prior position" — you lay out the opponent's view in full, honestly, before you lay out your own (the uttara-pakṣa, the "later position").3 It's not a debate opener. It's the whole discipline.
Modern debate and classical vād look identical from a distance — two people, opposing claims, an audience. Krisang's point is that they are not the same activity at all, because they're not aiming at the same thing.
Modern debate: prove your point. You walk in with a position, you walk out having defended it, and defending it successfully is what winning means. Losing means your position got knocked down. Nobody in this format is rooting for their own position to fail.
Vād: find the truth. "People never cared if they were right or wrong — what they cared about is what is the truth."4 Losing, in this format, isn't a bad outcome. It's the outcome you were actually hoping for if it turns out you were wrong, because you weren't in the room to protect your position. You were in the room to find out whether your position deserved protecting.
Same room. Same two chairs. Completely different game being played, and you can't tell which game you're in from the outside — only from what happens when someone gets shown they're wrong.
You're in the comments under your own post, and someone disagrees with your take. Your first instinct is to respond to their disagreement. Don't. Write their position back to them first — not the weak version, the version so complete they'd sign off on it without edits. Post that before you post your own view. Notice how much harder this is than it sounds; you'll catch yourself reaching for the strawman, the slightly-weaker phrasing that makes your rebuttal land easier. Rewrite it until the strawman is gone.
You're moderating a disagreement in your own community, the way Krisang was in that parenting group. Don't referee who's right. Interrupt the frame. Say out loud: we're not here to win this, we're here to find out what's true, and if that's not what either of you wants, this isn't the place. Watch what happens to the temperature in the room when the goal changes from "prove it" to "find it."
You're on the losing end of an exchange — someone has just shown you your position doesn't hold. Notice the two available reactions. One is the low-energy reaction: defend, deflect, get smaller. The other is the high-energy reaction: get excited, because you just traded a wrong belief for a true one and that's a good trade.5 Which one you have access to in the moment is the real measure of whether you've internalized vād or you're still playing modern debate with better manners.
Krisang doesn't sell this as easy. Explaining someone's position better than they can requires empathy — fine, expected. But underneath the empathy requirement is an energy requirement most people skip past: "low energy people get destroyed when they realize that what they know is not true. High energy people get excited when they know that what they knew was not true and they found something that is true."6
This reframes what being wrong actually costs you. If your energy is low, being wrong is a small death — a piece of your identity collapsing in public. If your energy is high, being wrong is a trade you were hoping to make. The pūrva-pakṣa discipline doesn't work as a communication technique bolted onto a fragile ego. It only works once you've done something upstream — built enough reserve that the discovery of wrongness reads as gain, not loss.
Krisang is careful to frame this as a skill inside a larger system, not a personality trait some people have and others don't. He names it as one piece — out of sixteen to eighteen elements — inside the classical discipline of nyāya, the logic tradition, and mentions a course being built specifically to teach the "apply logic to find truth" version of debate.7 The framing matters: pūrva-pakṣa isn't "be a nicer arguer." It's a trainable move inside a much larger apparatus for reasoning in public, and Krisang's claim is that learning this one piece alone — restate their side better than they can — already moves your communication "to the next level."7
Apply this to the actual job of making content and the failure mode sharpens. A creator under a comment-section attack is rarely debating to find the truth about their work. They're defending a position they've already published, already monetized, already staked identity on. The incentive structure of content creation — you need people to trust your positions, your positions are attached to your income — pulls hard toward "prove your point" and away from "find the truth," even for creators who'd say, if asked, that they value the truth. The pull is structural, not a character flaw. Pūrva-pakṣa is a deliberate counter-pressure against that pull: a rule you follow specifically because the incentives around you won't produce it on their own.
Krisang's claims here are self-reported practice, not documented outcomes — he describes the parenting-group intervention as something he did, not something he measured the effect of. The energy-and-being-wrong claim ("low energy people get destroyed... high energy people get excited") is asserted without a mechanism for how "energy" gets built in the first place; it's plausible as phenomenology, untested as psychology. [PARAPHRASED]
Open question: is "restate their position so well they don't need to explain it" achievable for genuinely opposed worldviews, or does it work best for disagreements that are actually adjacent — close enough that steelmanning is hard but not impossible? Krisang's own example (two parents on screen time) is a low-stakes, close-adjacency case. Whether the discipline holds under higher-stakes ideological distance is untested in the source.
Krisang never mentions Kauṭilya by name, but he's standing in the same shastric room. Pull the Arthashastra's own pūrva-pakṣa/uttara-pakṣa method — the one where Kauṭilya greets a rival tradition's teacher before his own, states the predecessor's case, then states his own — next to Krisang's parenting-group intervention, and the identical vocabulary is doing two different jobs.8 Kauṭilya's version is a reading discipline: which book do you pick up first, whose position do you study before forming your own. Krisang's version is a speaking discipline: what do you say first when two live people disagree in real time. One operates on a library. The other operates on a comment thread. Same shastric grammar, applied at two different speeds — the slow speed of a scholar building a position over years, the fast speed of a creator responding in the moment a disagreement lands in their notifications. Neither Krisang nor the Kauṭilya-reading tradition treats the other's speed as illegitimate; they're the same discipline running at different clock rates.
History — Kautilya's Shastric Method: Purvapaksha-Uttarapaksha and the Sukra-First Inversion. This is the direct lineage connection, and it's worth being precise about what each page actually contributes, because they don't just repeat each other. Kauṭilya's method (via Pillai) is about sequencing what you read: greet the demon-guru first, study the suppressed tradition before the dominant one, because the loser had to think harder under more pressure. It's a discipline for building your own position honestly before you've spoken a word to anyone. Krisang's method is about sequencing what you say: when a live disagreement is already happening, state their case first, so completely they feel over-served rather than under-heard. Put the two together and a claim emerges that neither page states on its own: pūrva-pakṣa is not one technique but a family of techniques unified by a single move — deliberately handicap your own position's head start by fully arming the opposing one first — and the family covers both solitary reading and live conversation. The Kauṭilya page's four-piece reasoning (open-mindedness, good-in-bad, hear-both-sides, the-wicked-think-harder) explains why you'd bother doing this when reading alone, with no audience watching. Krisang's version explains why the same discomfort is worth tolerating in front of an audience that is watching, in real time, whether you flinch. Read together, the discomfort itself becomes the tell: if handicapping the other side doesn't cost you something — in study time, in social standing, in the risk of losing — you probably haven't actually done it.
Psychology — Theory of Mind: The Ability to Know What Others Don't Know. Krisang's test — they should feel like you know their position better than they do — is, underneath the Sanskrit, a demand for unusually high-resolution theory of mind: modeling not just that the other person has a position, but modeling it with enough fidelity to out-articulate them. The psychology page's most unsettling finding is that theory of mind and empathy are separable — a sociopath can have "perfect Theory of Mind... and use that understanding to manipulate, coerce, and harm." High-resolution modeling of someone else's mind is not, by itself, a virtue. It's a capability, and capabilities serve whatever goal points them. This is exactly the fork Krisang is drawing when he separates vād (find the truth) from modern debate (prove your point): both games can be played with the same trained capacity for modeling an opponent's position with precision. The difference between vād and manipulation is not skill level — a skilled manipulator can restate your position as well as any vād practitioner. The difference is which goal the modeling serves once it's built. Neither page names this fork explicitly on its own — the psychology page flags the ToM-without-empathy trap in the abstract, Krisang asserts the truth-not-winning goal without explaining what happens when the same skill is aimed elsewhere. Together they specify the actual safeguard: pūrva-pakṣa is theory-of-mind training with the goal pre-committed before the modeling begins ("I am not here to prove my point, I am here to find the truth," stated out loud, at the start4), which is precisely the declaration a manipulator has no reason to make.
Sharpest implication. If the test for real pūrva-pakṣa is "they feel like you know their position better than they do," then most public disagreement — including most disagreement performed by people who'd call themselves fair-minded — never clears the bar, because clearing it requires work most disagreements don't get: rewriting the other side's case until it's stronger than what they actually said. The bar isn't listening. The bar is doing their homework for them, better than they did it, before you do your own.
Generative questions.