Everyone wants to know what happens next. It is the oldest want there is, and there has never been a shortage of people offering to supply it.
Law 14 opens by naming that want as the whole game: "In the realm of power, your goal is a degree of control over future events."1 The obstacle is stated just as plainly — people will not tell you their thoughts, their plans, or their ulterior motives, so "you cannot predict their moves, and are constantly in the dark."2
What follows in this chapter is an answer to that, and the answer is startlingly narrow. There is exactly one place the future comes from, and three separate voices in this chapter say so without ever being introduced to each other.
The law's Authority is unusual, because it argues by elimination rather than by assertion:
"the reason a brilliant sovereign and a wise general conquer the enemy whenever they move, and their achievements surpass those of ordinary men, is their foreknowledge of the enemy situation. This 'foreknowledge' cannot be elicited from spirits, nor from gods, nor by analogy with past events, nor by astrologic calculations. It must be obtained from men who know the enemy situation—from spirits."3
Four named methods ruled out. Spirits. Gods. Analogy with past events. Astrological calculation.
Three of those are what we would file under superstition. The third one is not. Analogy with past events is history, precedent, pattern-matching — the ordinary way any competent person forecasts anything. Sun-tzu puts it in the same bin as consulting a spirit, and neither he nor Greene remarks on how aggressive that is.
Greene independently makes the same elimination in the body text, in a register closer to a shrug: "Let others consult the horoscope, or read tarot cards: You have more concrete means of seeing into the future."4
Same structure. Name the discredited method, dismiss it, claim the replacement. And the replacement is not analysis, judgement, or experience. It is a friendly front that lets you secretly gather information — people, again.5
He does not connect his sentence to the Sun-tzu passage he prints thirty-seven lines later, though they are the same argument.
The chapter's third statement is the strangest and it sits in the margin:
"Rulers see through spies, as cows through smell, Brahmins through scriptures and the rest of the people through their normal eyes."6
This is not advice. It is anatomy. Kautilya is not saying rulers should use spies; he is saying that spies are the organ by which a ruler perceives, the way smell is how a cow perceives — with the implication that a ruler without spies is not badly informed but blind in the relevant sense, the way a cow with no nose is.
And it distributes the organs by station. Brahmins perceive through scripture. Ordinary people perceive through their eyes, which in this schema is the impoverished option. What you are able to see is a function of what you are.
Three claims, one chapter: the future is only available from people (Sun-tzu), the alternatives are for other people (Greene), and the channel you perceive through is set by your position (Kautilya). Nobody assembles them.
Take Sun-tzu's four exclusions seriously and see what survives.
Spirits and gods — ruled out as sources of information about the enemy. Fine, and unsurprising in a fourth-century military text.
Astrological calculation — ruled out. Also a claim about a method that was, at the time, an expensive and institutionally respectable practice with staff and instruments.
Analogy with past events — and here is where the argument gets interesting rather than merely sensible. Ruling out precedent means ruling out the enemy's own track record. What he did last time. What generals of his type usually do. The whole apparatus of inference from history.
What survives is only this: a person who is currently in a position to observe the specific situation, reporting what they currently see.
That is an extraordinarily restrictive epistemology, and its justification is never given. The implied reason has to be that adversaries are not stationary — a track record describes a person who has since learned, changed, or is deliberately playing against his own pattern. Which would make sense of it. But nobody in the chapter says so, and Greene, who prints the passage, does not notice that his own book is four hundred pages of reasoning by analogy with past events.
The argument has a hole in it and the chapter's own material shows where.
Reports from men who know the situation are only as good as the men. They can be wrong, bought, replaced, or turned — and the same chapter closes with the V-1 campaign, where an entire targeting programme ran for months on reports from "men who know the enemy situation" who were working for the other side.7
So the eliminated methods and the endorsed one share the defect. A horoscope tells you what the horoscope says. A turned agent tells you what the enemy wants you to hear. Neither has an independent check inside it.
The real distinction is not source-type at all. It is whether the channel can be cross-checked against a second, unrelated one — and on that criterion human intelligence is not categorically better, it is just usually multiple. Sun-tzu's argument names the wrong variable and lands on roughly the right practice.
You have a decision to make and you do not know what the other party will do.
Notice what you reach for. You reach for the record — what they did last time, what people in that position usually do, what happened the last two times someone tried this. It feels like the responsible move. It is free, it is fast, and it produces a confident answer.
Sun-tzu's line says that answer is worth about as much as a horoscope, and you should sit with the discomfort of that for a second rather than dismissing it, because the reason is not mystical. A pattern describes who they were. It cannot see the conversation they had last week.
So you do the expensive thing instead. You find one person who is in the room this month — not senior, not authoritative, just present — and you spend an hour with them. Not asking what will happen. Asking what is happening.
Then the part that costs the most and gets skipped: you find a second one, unconnected to the first. Not because the first is dishonest. Because a single channel and a horoscope have the same structural property, which is that nothing inside them can tell you they are wrong. And when the two accounts disagree, that disagreement is the most valuable thing you will acquire all month — it is the only moment you get to see the shape of what neither of them can see.
The strongest evidence is the triple statement: three sources from three traditions and three millennia — Sun-tzu (fourth century B.C.), Kautilya (third century B.C.), and Greene — making the same elimination in one chapter. That convergence is real and partially answers the corpus-wide 🚩 SINGLE SOURCE flag on this book, since the two classical sources are independent of each other and of Greene.
Tension: Greene's own book is disqualified by the Authority he prints. The 48 Laws of Power is, structurally, an argument from analogy with past events — forty-eight patterns extracted from historical anecdote and offered as predictive. Sun-tzu rules that method out by name. The contradiction sits inside one chapter and is never noticed.
Tension: Kautilya's claim and Sun-tzu's are not the same claim. Sun-tzu's is methodological — this technique works, those do not. Kautilya's is positional — what you can perceive is determined by your station, and the ordinary person's eyes are simply a worse organ. One is advice anybody could take; the other says the channel is allocated by rank. Greene prints both as endorsements of spying.
🚩 SECONDARY WITHOUT PRIMARY — the Sun-tzu passage is attributed to The Art of War, fourth century B.C., with no translator or edition; the Kautilya line carries only a name and a century. [POPULAR SOURCE]
Open questions. Is the exclusion of precedent a real doctrine or a rhetorical flourish — did Sun-tzu mean never reason from the past, or only never rely on it against a live adversary? And does the argument survive contact with domains where the adversary is not adaptive: does foreknowledge about weather, markets, or machines have the same single source?
Within Law 14 itself, this page is the premise the three case pages spend their evidence on — and they do not all confirm it.
Duveen confirms it exactly: years of purchased human reporting, and a coincidence engineered from it. Talleyrand confirms it with a wrinkle — his best method extracts from people who do not know they are reporting, which is still men, but not men "who know the enemy situation" in Sun-tzu's sense.
The V-1 campaign is the counter-case, and it is in the same chapter. There, human reporting is the vulnerability. The Germans did exactly what Sun-tzu prescribed — obtained foreknowledge from men in position to observe — and were destroyed by it, precisely because they had only that. Had they possessed a single astrological calculation, it would at least have been independent of the enemy.
That is the finding the chapter contains and cannot state: Sun-tzu's rule optimises for accuracy and ignores capture. A method that is more accurate and fully controllable by your adversary can be worse than a method that is less accurate and beyond his reach. The chapter recommends the first and then documents the second bringing down a bombing campaign, forty lines apart.
Eastern spirituality — The Astrologer as Objective Witness, Not Fortune-Teller
Sun-tzu rules out astrological calculation and Greene tells you to leave the horoscope to other people. That page opens by demolishing exactly the picture both of them are dismissing — the robed figure predicting the future — and replaces it with three roles: objective witness, guide, and provider of a map for harmonising the purposes of a life. Explicitly not prophet, not fate-pronouncer.
So the two traditions are not disagreeing about astrology. They are describing different practices that share a name.
The insight the pair produces: the argument is really about whether the question is what will happen or what am I dealing with. Sun-tzu wants the first, and he is right that a chart cannot supply it. The jyotiṣa page disclaims the first and claims the second — bringing hidden aspects of a life to light for someone who then does their own work. And on the second question Sun-tzu's own method is no better positioned, because a spy in the enemy camp reports the enemy's movements and can say nothing whatever about the client's own material. Which means the elimination is narrower than it sounds: it disqualifies divination as military intelligence, and leaves entirely untouched the use both traditions actually make of their instruments, which is characterising a situation rather than forecasting an event. Greene's dismissal is aimed at a practice that neither he nor his source has described.
Business — Specific Knowledge — Action First, Research Second
Nobody learned to ride a bike by reading about bikes. That page's rule is that the valuable knowledge in any craft is only acquirable by doing the thing, in that order — take off first, read the book to fix whatever is rattling. The runway is the trap.
Sun-tzu's elimination is the same shape aimed at a different target. Both rule out mediated knowledge in favour of direct contact with the situation: one rejects precedent and calculation in favour of men presently in the enemy camp, the other rejects the course and the cohort in favour of being on the court.
What the pair produces jointly, and what neither says: the two rules point in opposite directions about who has to be present. The business page requires that you be the one in contact — the prize is only handed out on the court, and specific knowledge is by definition non-transferable. Sun-tzu's rule outsources presence entirely; his sovereign never goes anywhere, he buys men who have been. Put together, they identify a category the business page treats as impossible and Law 14 treats as routine: knowledge that is both situational and transferable. Duveen's valets supplied it; a valet's report on Mellon's coat is direct contact, purchased. So the real variable is not direct-versus-mediated at all. It is whether the knowledge survives being told to someone — and comings and goings survive, while how to set up a story does not. Neither page has the distinction, and both of their rules are special cases of it.
Sharpest implication. Every source has the property that nothing inside it can tell you it is wrong. Sun-tzu's elimination sorts sources by dignity — men over spirits, observation over calculation — and that sorting does no work, because the failure it should be guarding against is capture, not credulity. The only property that has ever protected anyone is having a second channel that the first one cannot reach.
Generative questions.