Your mother taught you three words before she let you cross the street alone: stop, look, listen. Haha Lung takes those same three words and aims them at the person lying to your face.1 The drill is identical. You hold still long enough to think. You watch what the body does while the mouth talks. You hear not just the words but the music under them. Done in that order, the three moves close off every exit a liar has — the speed they want, the eye contact they fake, the wording they over-rehearse.
The reason this matters is that lying is not rare. Lung estimates only about 5 percent of people are "natural liars"; everyone else has to work at it, and by adulthood has had decades of practice — the average adult, he says, lies roughly twenty-five times a day, one to six times an hour.2 So the question is never whether the person across the table is capable of lying. It's whether you slowed down enough to catch the seams.
This is the practical, in-the-moment counterpart to knowing what a lie is built from. It is a three-step interrogation posture for use without a polygraph, against a live speaker, in the few seconds you have while they talk.3
Stop means slowing a fast-talking liar down. A liar wants tempo; tempo is how a shaky story gets past you before you can pick at it. You buy time by saying you want to "consider his offer," then you pick apart the line of BS at your own pace.4
Look means watching the eyes and the "shadow-walk" — Lung's name for the body's involuntary betrayals. Body language gives up all but the most accomplished liars. Even a liar who has talked himself into believing his own lie still knows, underneath, that he is lying, and that knowledge leaks out as tells and twitches.5
Listen means hearing tone, not just content. You track which words he leans on, and especially which words and details he tries to slide past. There is one specific tell: a person using contractions — "don't," "didn't" — is, by Lung's reckoning, about 60-40 likely to be telling the truth; the liar tends to over-formalize into "do not" and "did not," and to over-stress the one word the whole lie hangs on.6
A single lie is a fragile thing. It only survives if it stays small and unexamined. The engine inside this method is a simple, cruel insight: every question you ask forces the liar to manufacture new lies to prop up the first one — and each new lie is fresh surface area for error.7
Picture the first lie as a single tent pole jammed into soft ground. It stands as long as nobody touches it. Now you start asking questions, and to keep the first pole upright the liar has to drive in a second pole, then a third, each one improvised, each one needing to agree with all the others. Keep asking and the structure stops being a tent and becomes a house of cards the liar is building in real time while you watch his hands shake. Asking questions does double duty: it generates these new error-prone lies, and it "stifles his flow," breaking the rehearsed rhythm and forcing him to start over.8
The repetition move is the sharpest tool here. You pretend you missed something. You ask him to start the story over, to repeat the questionable points and passages.9 A truthful account is stored as memory and reproduces with rough consistency; a fabricated one is stored as a performance and drifts on each retelling. Forcing the re-run is how you make the drift visible.
This is also why the demographics matter for calibration. Lying is patterned, not random.10 Children under eight are concrete thinkers — they lie about actions ("a big gorilla came in through the window and broke it") and are bad at hiding feelings. Between eight and ten, abstract thinking arrives and the lies get more sophisticated. By puberty, kids are accomplished liars, sometimes lying purely for the thrill and power of fooling adults.11 In adults, Lung reports a gender split: men favor lies of commission — adding information, lying to promote themselves, and leaning in while doing it; women favor lies of omission — withholding, lying to make others feel better, and leaning away.12 The practical upshot is the listening key changes with the speaker: listen to what he says, and listen to what she doesn't.
What this page hands the rest of the vault is the clock on lie-detection — the live, time-pressured procedure that turns static tell-knowledge into something you can run during an actual conversation. Where other pages catalog the anatomy of a lie or the inventory of body-tells, this one says: here is the order of operations, here is the lever (forced repetition), here is the calibration table (who lies how). It converts a field guide into a drill.
It also emits a falsifiable, usable heuristic — the contraction tell at 60-40 — that downstream pages on interrogation, negotiation, and sensory reading can plug straight in, with the honest caveat that 60-40 is a thumb on the scale, not a verdict.
Lung gives the liar's side of the same coin in his "Ten Commandments of Lying," and the most revealing one is Commandment VIII — "Stick to your guns." His illustration: when they catch you "with your hand on the knife still sticking in his throat," you swear to God you were only trying to give the man a tracheotomy to save his life.13 Read it from the trapper's chair and the whole Stop-Look-Listen method snaps into focus.
The tracheotomy defense is a lie of commission of the boldest kind — it adds an elaborate, self-flattering counter-story on top of damning physical evidence. By Lung's own rules it is a bad bet, because his Commandment II warns liars to keep lies "within the bounds of reality" and Commandment IX cites Lincoln's line that no man has a memory good enough to be a convincing liar.14 The bold lie violates both. So how do you trap it? You Stop — you don't react to the dramatic confidence, you slow it down. You Look — confidence in the voice, but the eyes and hands are doing the shadow-walk. You Listen — and here the liar betrays himself precisely as Lung predicts, over-emphasizing the load-bearing word ("I was trying to save him") and, under stress, formalizing ("I did not stab him") rather than the truthful-sounding "I didn't touch him."
Then you spring the multiplier. Tell me again how the knife got there. Walk me through where you were standing. Each retelling of the tracheotomy story demands new improvised detail — the angle, the lighting, why he had the knife in hand at all — and because none of it is memory, the second telling will not match the first. The bold liar's own commandment to "always have supporting evidence ready" becomes the trap: invented supporting evidence is exactly the new error surface the questioning method is designed to expose.15
The case study cuts both ways, which is the point. Lung is teaching the lie and the trap in the same chapter, from the same toolbox. The trapper and the liar are studying the same physics; whoever runs the drill more coldly wins.
He's already talking fast when you sit down. The story spills out clean and quick — too clean, too quick — and you feel the pull to nod along just to keep pace with him. Don't. You lean back instead of forward, and you say it out loud: "Hang on. Let me make sure I've got this." The tempo breaks. He waits. Now you own the clock.
You watch his eyes while the next sentence comes. Not staring — just present. There's a hitch, a flick away and back, the hand that touches the collar. The shadow-walk. You file it without showing your face changed.
You listen to the music, not just the lyrics. He says, "I did not go anywhere near that room." Not didn't — did not. And he sits on the word "near" like it's load-bearing. Two flags in one breath. You don't pounce. You say, mild as milk, "Sorry — I think I missed something. Start from where you left the office." He starts over. And the second version has a detail the first one didn't, and the timing is half a beat off from what he said two minutes ago.
So you keep asking. Which door? Who else was there? What time exactly? Each question makes him build a new pole to hold up the old lie, and you can hear him doing carpentry in real time, the flow stuttering, the story thickening with invention. You're not arguing. You're just curious, endlessly, patiently curious, and curiosity is the acid that dissolves a rehearsed lie. By the fourth retelling the seams are showing to anyone who slowed down enough to look. You did. That was the whole trick.
The method fails most often not in the liar but in the trapper, and the failure has a tell of its own: you matched his speed. The moment you let the fast talker set the tempo, Stop is dead, and Look and Listen come apart in your hands because you're too busy keeping up to notice anything.
A second failure is treating the contraction tell as a verdict instead of a lean. Sixty-forty is a coin weighted slightly, not a confession. Build your whole read on one "do not" and you'll hang an honest nervous person and clear a fluent liar who happens to use contractions naturally.16 The tell is one vote in a count, never the count.
The third sign is the trapper who gets the proof and then can't hold it. You catch him, you have it cold, and the itch to humiliate him on the spot takes over. Lung names this as a choice, not a reflex: once you've trapped a liar you can spend it or bank it.17 Spending it — confronting him, humiliating him in front of his girlfriend and peers — is instant gratification and often the weakest play. Banking it — sitting on the proof, "a rabid rabbit you can pull out of your top hat any time you need it" — keeps the leverage live. The strongest variant: let him know you know, and convert the catch into blackmail or into a kept secret that buys you an ally, or at least a co-conspirator.18 The failure is reacting where you could have been positioning.
The contraction heuristic is the soft spot. Lung gives "60-40 in favor of his being truthful" with no study citation, and the broader deception-detection literature it sits near is exactly the kind of pre-2015 behavioral-science claim that has fared badly under replication — humans detect lies at barely-above-chance rates, and most single tells are weak.19 The figure is best treated as a practitioner's rule of thumb, not a measured base rate.
There is also a tension inside Lung's own framing. He insists the body betrays "all but the most accomplished of liars" — but his Ten Commandments are precisely a manual for becoming that accomplished liar (match your body language to the lie, put your heart into it, believe it yourself).20 So the method's reliability is implicitly bounded by the skill of the target: it works well on the 95 percent who have to work at it, and degrades against the 5 percent natural liars and the trained. Open question: against a trained liar, does the forced-repetition multiplier still bite, or does rehearsal immunize the retelling?
This page draws on a single source, so the live tension is internal to Haha Lung rather than between authors. Lung writes the trap and the lie in the same breath — Stop-Look-Listen sits a page away from the Ten Commandments of Lying, and the two are mirror images. That is not sloppiness; it is his whole posture. The trapper and the liar run the same physics from opposite chairs, and Lung's wager is that the colder operator wins. The honest strain in the material is that his confidence in body-tells ("betrays all but the most accomplished") quietly concedes that his own coaching aims to manufacture the exception. The method is strong against amateurs by the author's own logic, and he never fully reconciles how strong it stays as the target gets better — a tension worth preserving rather than smoothing.
Plainly put: this page tells you how to run a lie-catch in real time, and it borrows its raw materials from the vault's pages on what lies are made of and how bodies leak — but it adds the one thing those pages don't, which is a clock and a lever.
The first handshake is with The Anatomy of a Lie and Its Entourage. That page is the dissection table — it lays out the lie and the supporting cast (the emphasis, the fabricated detail, the emotional fuel) as static structure. This page is the dissection in motion. The structural parallel is exact: where the anatomy page says a lie travels with an entourage of supporting fabrications, this page weaponizes that fact by forcing the entourage to grow through questioning until it collapses under its own coordination load. The insight you only get by holding both: a lie's entourage is simultaneously its armor and its fault line — the very supporting details that make it convincing on first telling are the surface that fails on the third. Knowing the anatomy tells you the lie has parts; knowing the trap tells you which part to push to make the whole thing buckle.
The second handshake is with Shadow-Walkers: Reading Posture and Tells. That page is the field guide to the body's involuntary betrayals — the catalog of what the shadow-walk looks like. The "Look" step here is the field guide put on a stopwatch. The tension between them is instructive: the tells page implies reading is passive observation (watch and the body tells you), while this page insists the best tells are provoked — the gender-coded lean (men in, women away), the stress-twitch, the over-formal phrasing all intensify under the pressure of forced repetition. The combined insight neither gives alone: a tell at rest is faint and deniable, but a tell under interrogation pressure is amplified, because Lung's whole stress claim — "the tighter you squeeze a rat, the more he squeals" — means the trapper isn't just reading the shadow-walk, the trapper is manufacturing it by raising the stakes.21
A third, lighter handshake runs to Watchers, Listeners, Touchers: Sensory-Mode Classification. The Look/Listen split in this method maps onto the vault's sensory-mode taxonomy: a trapper strong in the visual mode will catch the eye-flick and the lean; one strong in the auditory mode will catch the "do not" and the over-stressed word. The insight is a warning — your own dominant sensory mode determines which half of the trap you run well and which half you're blind in, so the under-developed channel is exactly where a liar matched to your weakness will walk free.
The Sharpest Implication. The deepest move in this method isn't detection — it's what you do after you've caught him. Lung treats the catch as a fork: spend it, bank it, or convert it to leverage. That reframes lie-detection away from a courtroom model (find the truth, render judgment) toward a power model (acquire a fact about someone that they desperately don't want known, then decide its market value). The trap doesn't end an interaction; it opens a position. Most people stop at "gotcha." The operator starts there.
Generative Questions.