Law 17 closes on a caution, and it is easy to read past because it arrives after four pages of recommendation:
"A warning: Unpredictability can work against you sometimes, especially if you are in a subordinate position. There are times when it is better to let people feel comfortable and settled around you than to disturb them. Too much unpredictability will be seen as a sign of indecisiveness, or even of some more serious psychic problem. Patterns are powerful, and you can terrify people by disrupting them. Such power should only be used judiciously."1
Read what that concedes. The identical behaviour — the same inexplicable move, executed with the same skill — produces terror when performed from above and a diagnosis when performed from below.
The mechanism is not in the act. It is in what an observer does with an act they cannot explain.
Faced with an unexplained move by someone with standing, the observer supplies a hidden reason. There must be something I cannot see. Picasso states this exactly in the same chapter: "once you have attained a certain level of recognition, others generally figure that when you do something, it's for an intelligent reason."2
Faced with the same move by someone without standing, the observer supplies no reason at all. There is nothing to explain — the man is simply erratic.
So the operator does not control which reading occurs. They control only the act, and the act is identical in both cases. The variable is a prior held by other people about whether your behaviour has reasons.
Indecisiveness — you cannot make up your mind. A competence judgement.
A more serious psychic problem — you are unwell. Not a competence judgement at all, and the more damaging of the two, because it is not something performance can correct.
The warning is arithmetically worse than it looks, and the chapter does not do the arithmetic.
For the person with standing, the technique's failure mode is that it does not work. The observer is not unsettled, shrugs, and continues. Cost: the move was wasted.
For the subordinate, the failure mode is not that it fails to unsettle — it is that it succeeds in communicating something else. The observer forms a view: unreliable, unwell, not to be given anything important. And that view then organises how everything else you do is read, for a long time.
One is a wasted attempt; the other is an acquired reputation. Greene presents them as two sides of the same caution.
There is also no way to try a small amount first. You cannot try a small amount and see how it lands, because the reading is formed by the observer's existing prior, which is exactly the thing you were trying to change.
Put the warning next to the chapter's own cases and something uncomfortable falls out.
Visconti is a duke. Fischer is the challenger for a world title with nothing to lose. Picasso is the most famous living artist and says the precondition out loud. Jackson is a general commanding an army. Ali is the former heavyweight champion of the world.
Every operator in Law 17 already had standing, and the one who states the mechanism most clearly — Picasso — names the standing as the requirement.
So the law's applicability is the reverse of how it reads. The technique is not a route to power; it is something power already permits, and the chapter's four pages of recommendation are addressed to a reader who, if the warning is correct, mostly cannot use it.
That is the same structure this build has now logged in three consecutive laws: Law 15's annihilation requires the ability to banish; Law 16's withdrawal requires established presence; Law 17's unpredictability requires a prior that your actions have reasons. Three laws in a row whose stated instruction is available only to someone who already occupies the position the book is ostensibly teaching you to reach.
The warning's positive content is one clause: "there are times when it is better to let people feel comfortable and settled around you than to disturb them."
Which is the Reversal — predictability as an asset — arriving as a consolation rather than as a strategy.
And it is a better instrument than the consolation framing suggests, because the Reversal's own account says the pattern is a deposit. The subordinate who behaves consistently is not settling for second best. They are accumulating the only version of this that will ever be available to them: one break, later, from a position where people have a model to be surprised by.
Greene does not say this. But the chapter's two halves, taken together, give the subordinate a sequence rather than a prohibition — and a sequence is a considerably more useful thing than this is not for you.
You have been thinking about being less predictable at work.
Not dramatically. Just less available, harder to read, less immediately forthcoming — because you have noticed that the people who are difficult to pin down seem to be taken more seriously than you are.
Before you do, ask one question, and answer it honestly: when you do something people cannot immediately explain, do they assume you have a reason?
You can check this. Think of the last time you made a call that surprised your colleagues. Did anyone ask you to explain it — or did they assume there was a reason and wait to find out? Asked to explain means the prior is not there. It is not an insult; it is information about where you stand, and it is the exact variable this law runs on.
If the prior is not there, unpredictability does not buy you gravity. It buys you a reputation for being hard to work with, and that reputation will be applied retroactively to things you did well.
What is actually available is the deposit. Be legible, repeatedly, in one specific thing, until people can forecast you — and understand that you are not being safe, you are accumulating. The single break is a real instrument and it requires a pattern to break.
And if you are on the other side of this — if someone junior has become hard to read and you have started to wonder about them — check whether you are running the diagnosis Greene names. The same conduct from someone senior would have been read as having reasons. That asymmetry is in you, not in them, and it is worth knowing you have it before it becomes a view.
The warning's strength is that it is Greene's own, unprompted, and it concedes something structural rather than hedging on degree — the same act reads differently by position, not by intensity.
Tension: it inverts the chapter and arrives last. Four pages of recommendation, then a paragraph establishing that the technique is unavailable to most readers. No case is revisited in light of it.
Tension: every operator in the chapter already had standing, and Picasso states the precondition explicitly a page earlier. The chapter therefore contains its own scope condition twice and applies it neither time.
Tension: the downside is not symmetric with the upside, and the warning's phrasing (can work against you) suggests it is. A failed attempt from above is a wasted move; from below it is an acquired reputation.
🚩 SINGLE SOURCE — this is Greene's own assertion with no case, no evidence and no source. The claim that unpredictability reads as "some more serious psychic problem" from a subordinate is presented as an observation about how people are perceived, with nothing behind it. [POPULAR SOURCE] · [LOW CONFIDENCE]
Open questions. Is the prior about you specifically or about your role — could a person be granted the assumption-of-reasons in one relationship and denied it in another simultaneously, and if so is this a per-room variable like Law 16's threshold? And: how does anyone acquire the prior in the first place, given that this chapter's instrument requires it and the corpus never says where it comes from?
The convergence with Law 16's Reversal is exact, and neither chapter references the other. Law 16: leave too early and you do not increase your respect, you are simply forgotten. Law 17: unpredictability from below reads as indecisiveness or illness.
Both are the same claim about a threshold, on two different instruments, and both arrive in the closing paragraph after the technique has been recommended at length.
Set that beside Law 15's requirement that you be able to banish an enemy, and the pattern is three laws deep. This build should carry it forward as a corpus-level finding rather than three separate observations: the book's instructions are consistently conditional on a standing the book does not teach you to acquire, and the condition is stated, when it is stated at all, at the end of the chapter in a paragraph that does not revisit any case.
Within the chapter, this warning and Visconti are the same phenomenon from the two ends of a court. The duke's oscillation destroys his courtiers' capacity for independent judgement and is filed as mastery. A courtier behaving identically would be filed as unwell — and Greene supplies both halves without noticing that his own warning describes what Visconti's method looks like performed by anyone below the throne.
Psychology — Stereotype and A Priori Judgment
That page's territory is the categorisation that arrives before the evidence and then organises how the evidence is read.
This warning is that mechanism applied to status, and the pair makes the asymmetry legible.
The observer is not evaluating the act. They have already categorised the actor — has-reasons or does-not — and the category determines what the ambiguous behaviour means. Identical conduct, two categories, two readings.
What the pair produces: the person being read has no access to the variable that decides their case. Everything they control is the act, and the act is not where the meaning is assigned. This is why the usual advice to junior people — be more confident, act like you belong — misfires so reliably: it operates on conduct, and conduct is the input the prior is applied to, not the prior itself.
It also explains something the chapter treats as a quirk. The two bad readings Greene names — indecisive, or unwell — are both explanations of last resort, the kind an observer reaches for when a behaviour resists explanation. A person granted the has-reasons prior never generates them, because the observer stops at there is something I cannot see. The difference is not in how charitably you are judged. It is in whether the judging process runs at all.
Sharpest implication. The same act is terror from above and a diagnosis from below, and the operator controls only the act. Which means Law 17 is not a technique for acquiring standing — it is a privilege that standing confers, and the chapter says so in its final paragraph after recommending it for four pages.
Generative questions.