The first method in Law 33's plan opens with a quotation:
*As Sigmund Freud remarked, "No mortal can keep a secret. If his lips are silent, he chatters with his fingertips; betrayal oozes out of him at every pore."*1
And the principle Greene draws from it:
Remember: Since we all try to hide our weaknesses, there is little to be learned from our conscious behavior. What oozes out in the little things outside our conscious control is what you want to know.2
That second sentence is the actual doctrine and it is stronger than the method it introduces. Prepared material is not merely less informative — it is worth nothing, because effort has been spent making it uninformative. Attention is a budget and this says spend none of it on anything anybody has thought about in advance.
His guidance on collection is more specific than the method's title suggests:
Everyday conversation supplies the richest mine of weaknesses, so train yourself to listen. Start by always seeming interested—the appearance of a sympathetic ear will spur anyone to talk.1
Then the details:
Train your eye for details—how someone tips a waiter, what delights a person, the hidden messages in clothes. Find people's idols, the things they worship and will do anything to get.3
Note that only one item on that list is a gesture. Tipping is a transaction; clothes are a decision made in the morning; what delights someone is a reaction. The method is called gestures and unconscious signals and most of its collection points are neither — they are unguarded choices, which is Schopenhauer's territory rather than Freud's. See Character Shows in Trifles.
The one technique in the method that is an action rather than an observation:
A clever trick, often used by the nineteenth-century French statesman Talleyrand, is to appear to open up to the other person, to share a secret with them. It can be completely made up, or it can be real but of no great importance to you—the important thing is that it should seem to come from the heart. This will usually elicit a response that is not only as frank as yours but more genuine—a response that reveals a weakness.1
The mechanism is reciprocity applied to disclosure, and the asymmetry is the whole point: your confidence is manufactured and theirs is not.
More genuine is doing real work in that sentence. The exchange does not merely balance — the other party overshoots. Which is what you would expect if the pressure being created is social rather than informational: a person who has just received what feels like trust does not calibrate their reply to the content of what they got, they calibrate it to the gesture, and matching a gesture of intimacy requires supplying something that costs them.
Greene's version costs nothing and buys something expensive, which is the definition of an exploit rather than an exchange.
A short passage in the middle of the method deserves attention because it is methodologically the weakest thing in the chapter:
If you suspect that someone has a particular soft spot, probe for it indirectly. If, for instance, you sense that a man has a need to be loved, openly flatter him. If he laps up your compliments, no matter how obvious, you are on the right track.3
Run that as an experiment.
Hypothesis: this person needs to be loved. Test: flatter him obviously. Confirming result: he responds well. Disconfirming result: — none is specified.
And there is no plausible one, because a person who responds coolly to obvious flattery might be secure, or might be suspicious of you, or might need validation from a different source, or might be embarrassed by it — all of which are readily narrated as consistent with the original hypothesis.
This is not a probe, it is a confirmation. And the flattery is not neutral: it supplies validation, which makes a person warmer toward you regardless of whether they needed it, so the test partly creates its own positive result.
That is a real methodological defect and it runs through the whole search plan. The chapter has no case of a wrong reading, and this passage shows why — the instrument does not have a needle that can point the other way.
The largest omission is invisible because Greene's cases all quietly satisfy it.
You cannot read a deviation without knowing the person's normal. Nervousness with a concierge means nothing unless you know how this man usually talks to staff. A tip is only legible against what he tips ordinarily.
Every worked case in the chapter had that baseline and none of them mentions acquiring it. Richelieu "studied the workings of the court as if it were the machinery of a watch."4 Lustig watched Loller for days before speaking to him. Duveen spent years.
The reading is presented as the skill and the surveillance is presented as nothing, which makes the method look fast. It is not fast. The observation is cheap and the baseline is the entire cost, and it is paid in weeks.
You want to understand somebody and you have been listening to what they say about themselves.
Stop. That material is prepared, and by this method's own doctrine it is worth nothing.
Redirect the same attention to four things that cost nothing extra to notice:
How they treat people who cannot affect them. The waiter, the assistant, the person on the phone. Nobody performs for an audience that cannot help them, which is exactly why it is legible.
What they return to unprompted. Not the topic you raised — the one that keeps coming back regardless of what you raised. Nobody guards repetition because nobody hears themselves repeat.
Where the reaction is the wrong size. Disproportion needs no interpretation, only arithmetic.
What they never mention. Absences in the account of a life are as informative as anything in it, and they are not guarded because omission does not feel like disclosure.
Then two disciplines Greene does not supply.
Get the baseline before you read the deviation. Three or four ordinary occasions, no agenda, before you conclude anything. If you have not seen someone's normal you are not observing them, you are pattern-matching against yourself.
Write down what would prove you wrong. If nothing would, you have not formed a reading. Greene's flattery probe is the model of what to avoid: a test whose only available outcome is confirmation.
And on the Talleyrand move: the honest version is to actually tell someone something true that costs you a little. It produces the same reciprocity and it does not require you to be running an operation, because you have in fact traded.
Strongest evidence. A named practitioner with a specific reusable technique (Talleyrand's manufactured confidence), a stated doctrine that is coherent and unusually strong, and a collection list that is concrete enough to act on.
Tension — the flattery probe has no disconfirming result and partly manufactures its own confirmation. Documented above; the clearest instance of the search plan's evidential problem.
Tension — the method's own collection points are mostly not gestures. Tipping, clothes and delight are unguarded choices; the Freud quotation is doing rhetorical rather than analytical work.
Tension — no baseline requirement anywhere, though every case in the chapter had one and none says how it was acquired.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — the Freud quotation is given with no work, no date and no edition; it circulates widely in this exact form and its provenance is not established here. Treat as "attributed to Freud" rather than as a citation. 🚩 The Talleyrand claim ("often used") has no instance attached.
Open questions. Does the reciprocity asymmetry hold — do people genuinely overshoot a manufactured confidence, or does the effect depend on the disclosure being real and Greene's "seem to come from the heart" concealing a much harder requirement? And: how long is the baseline in practice, because the answer determines whether this method is available to anyone who is not already inside the target's life.
Talleyrand is on his seventh page in this vault and the corpus shows one instrument used every time.
Talleyrand as Conversational Spy is this method's parent case — the social register as an intake channel. Talleyrand and Manufactured Clairvoyance at Law 30 is the same channel run for output. This page adds the third mode: manufactured intimacy as a pump.
Across seven appearances he never once uses force, position or a threat. The whole career is conducted in conversation, and the book distributes it across seven law numbers as seven techniques.
Within Law 33, this method is the search plan's opening and its weakest member. Look for Contrasts gives a reading rule that can fail; Schopenhauer's trifles gives a collection point with a stated reason. This one gives a quotation, a list, and a probe that cannot return a negative.
Applied observation — the same signals, with the discipline attached. Reading People: Nonverbal Cues covers systematic reading of involuntary behaviour and carries the standing constraints of the practice: establish a baseline, read clusters rather than single cues, and treat any isolated signal as noise.
Greene's method has the collection points and none of the constraints.
The insight neither produces alone: the constraints are not refinements, they are what converts observation into evidence — and their absence explains the chapter's missing failure mode directly. A practitioner working from baselines and clusters gets contradictory readings regularly, and the contradictions are the feedback. An operator reading single cues against no baseline never gets a contradiction, because any single behaviour supports several stories and the operator picks one. The trained version fails often and visibly; the untrained version never fails and never improves — which is exactly what the corpus shows, and it is a stronger objection to this method than any about its ethics.
Sales practice — the manufactured disclosure, and where it breaks. Sales as Transference of Belief Over a Bridge of Trust treats the disclosure exchange as load-bearing and insists the seller's side be real, on the argument that the buyer is calibrating not the content but the cost to the discloser.
That is the precise objection to Talleyrand's version, and it is practical rather than moral.
What the pairing produces: if what the other party is reading is the cost of your disclosure rather than its content, then a secret that is "real but of no great importance to you" is cheap and therefore transmits a small signal — and the reciprocity you get back should be proportionally small. Greene's claim that the reply comes back "more genuine" than the prompt requires the target to misprice what you gave them, which they will do exactly once, and only if the manufactured intimacy is well made. So the technique degrades on repetition with the same person, and it works best on strangers — which is a scope condition that fits Lustig and Duveen and does not fit anyone trying to use it on a colleague.
Sharpest implication. The doctrine is stronger than the method: prepared material is worth nothing, so attention spent on anyone's stated account of themselves is attention wasted. But the method has no baseline requirement and its flattery probe is a test with no disconfirming outcome that partly manufactures its own result — which is why the chapter contains no misreading. The trained version of this skill fails often and visibly; this version never fails and therefore never improves.
Generative questions.