Picture two armies frozen at the border, neither moving. Then one side starts to look weak. Its emissaries come over with carts — jade, silk, fine horses, even boys and girls — apparently begging for peace. The enemy commander watches the procession and something curdles in him: contempt. They're afraid. They're buying me off. He gets careless. And that carelessness, not the jade, was the entire point. While he savors his own superiority, elite troops are already moving by several roads to fall on his slackened camp.1
That is the first of Hsü Tung's eight methods, and it tells you what kind of document this is. This is not philosophy about spies. This is a recipe book. Around the year 1000, in his Hu-ling Ching ("Tiger Seal Classic"), Hsü Tung wrote out eight named, repeatable techniques for using covert agents — the first systematic operational how-to in the Chinese intelligence tradition.2 Before this, the corpus mostly theorized spies. Hsü Tung tells you, step by step, how to run them.
The plain version: there are eight specific tricks for getting inside an enemy's head and his court, and each one comes with instructions. Sawyer notes most martial writers discussed spies' functions; "only Hsü Tung seems to have pondered their nonmilitary utilization" in detail.3 This page lays the eight out.
The chapter is titled "Employing Spies." It opens, characteristically for the tradition, by reaching back for authority — to the Hsing-jen of the Chou, the official "who traveled about the feudal states and reported back to the king on their plans," whom Hsü Tung flatly identifies as "actually a spy."4 From there it declares the governing rule before any technique: agents must be run in total secrecy, and "apart from the very wisest, who can attain this?"5
Then come the eight. Sawyer organizes them by function: the first two "apply exclusively to military situations," several are "fully developed counterintelligence practices that convert enemy agents to useful assets, whether knowingly or not," and the remainder "are designed to acquire essential knowledge by subverting enemy officials."6 In Hsü Tung's own sequence:7
The chapter closes by restating the master rule: agents are "the critical essence of the military theorists," but without "sagacious wisdom and moral worth" you cannot run them, because the whole Tao "lies in the subtle, secretive, clandestine, and submerged."8
What makes this a genuine taxonomy and not just a list is that the eight methods sort cleanly by which way information and influence are moving — and Hsü Tung clearly knows it.
Three of the eight (methods one, two, three) are feeding the enemy false information — disinformation that flows out from you, dressed up to look stolen. The gift-cart sells false weakness; the fed prisoner carries a fake plan; the acquired fake plan is bait the enemy thinks he took. In all three the enemy's own confidence in what he "discovered" is the weapon. The harder a man works to acquire a secret, the more he trusts it — so you make him work for the lie.
Two (methods four and six and seven and eight, the bribery cluster) are pulling true information in by buying enemy insiders — but they ladder up in sensitivity. Method four buys an enemy agent outright. Method six buys the families of court favorites. Method eight buys trusted local handlers. Method seven doesn't buy information at all — it buys destruction, paying confidants to slander the enemy's best strategists into mutual ruin. That's the cruelest and most strategic of the set: you don't learn the enemy's plans, you make sure he no longer has anyone capable of making good ones.
And method five is the odd, brilliant one: a double-bluff through deliberately obvious agents. You send men so clumsy the enemy spots them as spies, and the spotting itself transmits the message — "they're afraid, they're reduced to sending bunglers to bluster at us." The leak is the medium. The enemy's own counterintelligence becomes the channel for your disinformation.
The unifying mechanism across all eight: Hsü Tung never treats the enemy's intelligence apparatus as an obstacle to defeat. He treats it as infrastructure to hijack. Enemy prisoners, enemy agents, enemy favorites, enemy strategists, enemy locals — every one is a wire already running into the enemy's decision-making, and the method is always to seize the wire and send your signal down it.
This page is the operational floor under the abstract programs. Systematic Covert Programs argues that Chinese statecraft eventually treated covert action as an organized, repeatable enterprise rather than ad hoc trickery. Hsü Tung is the proof: eight named, transmissible techniques you could hand to a commander as a manual. The "system" in systematic covert programs has, here, an actual table of contents.
It gives the agent-typology pages a verb list. Chieh Hsüan's Sixteen Agencies catalogs the kinds of harm a covert program can do; Hsü Tung's eight methods are the techniques by which several of those harms are actually inflicted. Where one page taxonomizes ends, this page taxonomizes means.
And it hands the companion disinformation page its first three entries as a foundation. Hsü Tung on Deceiving the Enemy elaborates two further disinformation techniques in a separate chapter; methods one through three here are the same author's earlier, simpler versions of the move — feed the enemy a lie he thinks he stole. The two pages together are a single mind's complete disinformation toolkit, the "Employing Spies" chapter supplying the broad eightfold frame and "Deceiving the Enemy" zooming into two ruthless special cases.
Take apart the seventh method, because it is the one that reaches deepest. The goal is not to learn what the enemy's best strategist is planning. The goal is to get him killed by his own ruler.9
The mechanism runs on three moves. First, bribe the ruler's confidants — the men with his ear — to slander the strategist in court. Second, "appropriately respond to the slanders from outside the state" — meaning you arrange external events or signals so that the slander looks confirmed by reality. The traitor-accusation lands, and then conveniently something happens that seems to prove it. Third, let suspicion do the work: ruler and strategist "grow mutually suspicious," and "instigating them to ruin and harm each other," the partnership collapses from inside.10
What's surgical about this is the target selection. Hsü Tung doesn't aim at the army or the treasury. He aims at the single relationship that converts a state's resources into good decisions: the trust between a ruler and the one man who can out-think you. Break that bond and the enemy keeps every soldier and loses the only mind that knows how to use them. It is the same insight that animates Estrangement Techniques — that a relationship is more vulnerable than a fortress — applied with a named procedure and an external confirming step that most estrangement plays leave implicit.
The "respond from outside" detail is the connoisseur's touch. Anyone can plant a rumor. Hsü Tung plants the rumor and stages the evidence, so the paranoid ruler's own observation appears to corroborate the lie. The disinformation closes a loop: the slander predicts the strategist is treacherous, and then the world (which you have quietly arranged) seems to deliver proof.
You are the commander, and your army has been staring at theirs across the same river for a month. Nothing moves. So you decide to lose, on purpose, in their eyes.
You let your camp look tired. You let rumor drift across — quietly, as if you didn't mean it to — that gifts are coming for their general's favorites, that you're desperate to buy peace. Then you send the carts. Real jade. Real silk. Good horses, the kind a man notices. Children among the gifts, the full humiliating spectacle of a beaten state begging. Your emissaries bow low and ask for reconciliation.11
Now you watch the other commander, because his face is your intelligence report. You are not waiting to hear whether he accepts. You are waiting for the arrogance — the lift of the chin, the insult he can't resist, the order to stand his men down a notch because clearly the threat has passed. The moment you see it, the gifts have done their work. They were never a bribe. They were a lever for his contempt.
That night, or before dawn — early and late, by several roads, so no single column tips your hand — your elite troops are already moving. The enemy is lax exactly where the gift-cart told him he could afford to be. You did not buy peace. You bought his negligence, at the price of some jade and a little of your dignity, and you are about to collect.
The methods rot in recognizable ways.
The gift looks like a gift. Method one collapses the instant the enemy commander reads the carts as a trap rather than a tribute. If your feigned weakness is too theatrical, or your reputation too dangerous to be plausibly afraid, the arrogance never comes — and you've just handed a clever enemy real jade and a clear signal that you're about to attack. The bluff requires a believable bluffer.
The leak is too clean. Methods two and three depend on the enemy working for the false plan, because effort breeds trust. Leak it too easily and a careful enemy smells the bait. The fake must be expensive to acquire or it will be cheap to disbelieve.
The bought man stays bought by both sides. Methods four, six, and eight rest on bribery, and a bribed insider can take your gold and report you to his master, becoming a channel running the wrong way. Hsü Tung's whole closing warning — secrecy, "the subtle, secretive, clandestine, and submerged" — is the antidote, but a handler who can't keep the relationship submerged turns his asset into the enemy's.12
The slander outruns the staging. Method seven fails when the rumor lands but the external "response" never arrives to confirm it, or arrives too obviously. A paranoid ruler who later sees the slander was manufactured doesn't just exonerate his strategist — he hunts the source, and the source is you.
And the master failure, the one Hsü Tung names: a fool runs the program. "Without sagacious wisdom and moral worth, one cannot employ agents."13 These are not push-button tricks. Each demands reading the enemy commander's psychology in real time. A handler who can't tell arrogance from caution, or a believable bluff from a transparent one, will execute every method correctly and lose every time.
Sawyer's framing claim — that "only Hsü Tung seems to have pondered their nonmilitary utilization"14 — is a strong one, and worth holding lightly. It rests on what survives. The Hu-ling Ching may be the earliest extant systematic treatment without being the earliest one that existed; absence of earlier manuals in the record is not proof of their absence in fact. Sawyer's continuity narrative across the corpus is textual, and silence in a textual tradition is a weak negative.
There is also a genre tension. Hsü Tung opens by anchoring the eight methods in the Chou Hsing-jen, presenting them as the elaboration of an ancient office.15 But the Hsing-jen, on the evidence, was a reporting official — an overt envoy who observed and informed. Hsü Tung's methods are aggressive covert operations: bribery, slander, staged defectors. Calling the Hsing-jen "actually a spy" backdates a sophisticated operational doctrine onto a much simpler diplomatic role. The continuity is rhetorical authority-borrowing as much as genuine lineage — the same move the tradition makes everywhere, dressing innovation as restoration.
Finally, the chapter never resolves the obvious internal friction: methods six, seven, and eight all assume you can reliably bribe insiders into an enemy court without those insiders betraying you to their ruler, while the disinformation methods assume the enemy's intelligence is sharp enough to be hijacked. An enemy clever enough to have his agents stolen from him (method four) is presumably clever enough to detect a bribed favorite (method six). Hsü Tung treats the enemy as alternately shrewd and gullible depending on which method needs which — a tension the manual leaves entirely open.
Sawyer the synthesizer and Hsü Tung the practitioner pull in different directions on what this chapter is. Sawyer slots it into a long, continuous Tao of spycraft, reading the Hsing-jen opening as evidence the doctrine descends unbroken from the Chou. Hsü Tung, writing a working manual a millennium later, is doing something more like systematization — taking scattered tricks and giving them names, numbers, and a governing rule. The convergence is that both see a coherent tradition. The tension is that Sawyer's continuity flattens what is genuinely new in Hsü Tung: the taxonomic move itself. Earlier texts have the tricks. Hsü Tung is, on Sawyer's own account, the first to lay them out as an organized, transmissible set — which is less "continuation of the Chou" than "invention of the operational manual." Sawyer's framing both reveals this (he flags Hsü Tung as singular) and partly obscures it (by embedding the singularity in a continuity narrative).
To Systematic Covert Programs. The systematic-programs page makes the historical argument that Chinese statecraft moved from ad hoc trickery toward organized covert enterprise. This page is the textual hinge of that argument: eight named, numbered, transmissible techniques are exactly what "systematic" means in practice. The insight neither page produces alone is about what makes covert action a program rather than a stunt — it is enumeration and transmissibility. A clever ruse executed once by a genius is not a system; eight methods written down so a competent commander can run any of them is. Hsü Tung supplies the missing artifact the systematic-programs thesis needs to be more than an inference: a surviving manual with a table of contents. But the handshake also surfaces a limit — a manual proves the doctrine was systematized, not that the practice was; the gap between a written eightfold method and an actual standing intelligence service is exactly the gap the systematic-programs page must keep honest.
To Estrangement Techniques. Method seven — slander the enemy's strategist into the ruler's distrust — is a fully specified estrangement operation, and comparing the two pages sharpens both. The general estrangement doctrine says: break the relationships that hold an enemy together. Hsü Tung adds the procedural detail that the doctrine usually leaves abstract — the three-step sequence of bribing confidants, staging external confirmation, and letting paranoia compound. The insight that emerges from holding them together: estrangement is not merely planting a lie, it is closing a loop — the rumor must be married to apparently independent confirmation, so the target's own judgment becomes the executioner. The most effective estrangement doesn't ask the ruler to believe a slander; it arranges the world so the ruler believes he discovered the treachery himself. That "respond from outside" step is the difference between gossip and an operation.
To Chieh Hsüan's Sixteen Agencies. Chieh Hsüan catalogs the sixteen agencies or categories of covert harm a program can pursue; Hsü Tung catalogs eight methods by which several of those harms are produced. Read together they form a means-ends matrix: the sixteen agencies are the objectives, the eight methods are the verbs. The insight is structural — a mature covert tradition develops two taxonomies, one of effects sought and one of techniques used, and the existence of both (in different authors) is itself the strongest evidence that Chinese intelligence thought had matured past single-trick cleverness into something a working professional could be trained in.
Sharpest implication. Hsü Tung's deepest move is to treat the enemy's own intelligence apparatus not as a wall to breach but as a wire to commandeer. Every one of the eight methods seizes a channel the enemy already trusts — his prisoners, his agents, his favorites, his strategists, his locals — and sends a counterfeit signal down it. The enemy's capacity to "know" becomes the surface you write on.
Generative questions.