Picture a senior official leaving his ruler's palace at dawn, climbing into a chariot, and riding for days to a neighbor's capital. He carries a message he must deliver word for word. He also carries something nobody wrote down: a pair of eyes. While he waits in the foreign court, while he is feasted and flattered, he is counting. How many soldiers in the barracks. Whether the roads were kept clear. Which minister hates which other minister. When he gets home, the message he memorized matters less than what he saw. That official had a title in early China: Hsing-jen, "the man who travels." T'ang-dynasty theorists looked back and said this was the first administrative post in China actually responsible for intelligence work.1
The name is plain. Hsing means "to travel" or "to perform some action." Jen means "man" or "human being." Put together it just says "the traveling man." But the post sat near the top of the government. It ranked below the prime minister and stood roughly level with the early specialized ministers for war and works.2 Every state that copied the Chou bureaucratic model kept Hsing-jen on the payroll, sometimes several at once. The job stretched across an enormous range — from common messenger (which carried no title at all) up through diplomatic courier, protocol director, secretary of state, and what we would now call a foreign minister.3 One English word cannot hold all of that, which is exactly the point. The Hsing-jen was wherever the state touched a foreign state, and where states touch is where intelligence is born.
The Hsing-jen was an office, not a person — a slot in the early Chinese bureaucracy that fused diplomacy, protocol, and information collection into one chair. Two sources let us reconstruct it. One is the Chou Li (Rites of Chou), an idealized reconstruction of Chou offices probably composed late in the Warring States period. The other is the historical record preserved in the Ch'un Ch'iu and Tso Chuan, which shows the office actually operating.4
The Chou Li splits the office into two ranks. The Ta Hsing-jen (Great Hsing-jen) runs the rites for important visitors and the formal ceremonies for major guests, all designed "to gain the intimacy of the feudal lords." He stage-manages the king's seasonal assemblies: spring court to plan the affairs of the realm, autumn audiences to compare the achievements of the various states, summer court to display the realm's plans, winter to harmonize everyone's thoughts. And — the line that matters most — "on an annual basis the Ta Hsing-jen makes inquiries among the feudal lords in order to fathom their intentions."5 That phrase, fathom their intentions, is an intelligence mission in ceremonial dress.
The Hsiao Hsing-jen (Minor Hsing-jen) handles the tribute rites of the submissive states and meets the lords' emissaries at the border. But his real assignment is recording. The Chou Li orders him to keep five separate books. One records the causes, profits, and harms touching the people. One records customs, ceremonies, government affairs, instructions, and whether prohibitions were obeyed or violated. One records which feudal lords are "contrary, perverse, brutal, chaotic, conceal hidden plots, or violate edicts." One records deaths, mournings, famines, impoverishment. One records health, happiness, harmony, tranquillity. The stated goal: "to fulfill the king's mandate and thereby thoroughly know the causes and affairs throughout the realm."6 Five filing cabinets of standing intelligence on every neighbor. Sawyer's conclusion is direct — China's first official staff position with intelligence responsibilities was the Hsiao Hsing-jen, operating under the Ta Hsing-jen.7
The drift from messenger to intelligence director happened in the gap between the message and the man. A courier in this period did not carry a sealed document. Often he carried words in his memory and his own observing senses in his head. So even the lowliest version of the job came with eyes. The Tso Chuan is explicit that couriers and envoys "did not simply convey messages with blind eyes and deaf ears." They returned with a reply, yes — but also having "carefully scrutinized the reactions of those receiving the message, noted the situational context, and marked other relevant facts for subsequent analysis."8
You can watch the mechanism in the way Hsing-jen were treated by their hosts. States assumed the visitors had hidden purposes. That is why so many Hsing-jen were seized, detained, even killed — "implying the existence of clandestine aspects to their official roles, whatever the ostensible reasons for their detention."9 When Ch'u seized Cheng's Hsing-jen Liang Hsiao, the surface reason was that Cheng had sent envoys to announce a switch in allegiance from Ch'u to Chin.10 When Chin seized Wei's Hsing-jen Shih Mai, the surface reason was last year's attack on a Chin client state — but Sawyer flags the deeper one: Shih Mai "was an experienced commander and therefore doubtlessly seeking military information as well as fulfilling some innocuous cover mission."11 The cover mission is the tell. A faceless protocol clerk does not need a cover. A general carrying credentials does.
The same logic ran on the battlefield. Messengers went out before clashes to issue noble challenges and demand surrender — and, while there, "undertook a perceptive survey of the enemy camp." After the first day at Yen-ling, Ch'u read a Chin messenger as fearful because his eyes kept moving, and wrongly concluded Chin would withdraw. Protocol said return him unharmed; the fact that he was reading their preparations is exactly why he was there.12 The herald and the scout were the same person.
The Hsing-jen is the institutional anchor for the rest of the vault's early-China intelligence material. Where Sun-tzu's Five Spies gives the doctrine of organized espionage, the Hsing-jen gives the organization chart that predates the doctrine — proof that a recognized office, not just freelance cunning, was collecting on neighbors. It supplies Systematic Covert Programs with its earliest legitimate home: a budgeted, ranked, record-keeping post rather than an ad hoc agent network.
To Field Observation Methods, the Hsing-jen offers the platform the methods ran on — a body authorized to stand inside a foreign court and watch. To Agent Qualification and Selection it offers the first selection criteria on record: Cheng deliberately chose men like Kung-sun Hui precisely because they could "know about the activities of the four surrounding states, distinguish the clan and family names of the high officials, their positions, rank, and whether they were capable or not."13 That is a job description for an intelligence analyst, written in the sixth century BC.
The single richest illustration is the state of Cheng, a small power that flourished though ringed by giants — and Sawyer locates its survival partly in two men. A minister from Wei, appraising Cheng's strength, singled out its two Hsing-jen, Tzu Yü and Kung-sun Hui, for their comprehensive knowledge of other states.14 The Tso Chuan then describes how Cheng's chief minister Tzu Ch'an ran government as a kind of intelligence cycle. He selected the capable and assigned each to his strength. Kung-sun Hui knew the affairs of the four surrounding states down to the names, clans, ranks, and competence of their officials, and could draft orders. Feng Chien-tzu decided major affairs. P'i Ch'en had planning ability that worked in the countryside but failed in the city. Tzu Ta-shu, handsome and refined, delivered the final word to honored guests.15
Read it as a workflow and it is startlingly modern. When Cheng had a pending matter with the feudal lords, Tzu Ch'an first inquired about the actions of the surrounding states from Tzu Yü — collection. He then had Tzu Yü draft several versions of the relevant edicts — options. He rode out into the countryside with P'i Ch'en to decide whether the plans would work — assessment, done in the field where P'i Ch'en's judgment was sharp. He informed Feng Chien-tzu to render the decision — decision. He entrusted the finished response to Tzu Ta-shu to transmit to the guests — execution. "For this reason they rarely encountered failure."16 Collection, analysis, options, decision, delivery — separated by role, routed through specialists. The Hsing-jen sat at the head of it as the collector. Sawyer's verdict: Cheng's Hsing-jen, "while nominally protocol directors, were clearly masters of foreign information, essentially state intelligence directors."17
Tzu Yü himself is the proof of professionalization. In the Analects he is praised for polishing the drafts of communications with other states; the commentators tie him to the Chou Li duty "to oversee the rites connected with communicating with other states."18 His fame, Sawyer writes, "derived from his superlative performance in executing foreign missions, handling visiting dignitaries, gathering and evaluating information, and planning state affairs, as befits a Ta Hsing-jen."19 A Confucian disciple whose claim to greatness was running a foreign-intelligence desk well.
You are the Hsing-jen of Cheng, sent to a neighbor with a routine message about tribute dates. You arrive at the border and the protocol officer meets you — or does not, and you note which. Inside the capital you are feasted. You watch who pours your wine and who does not bother. You let the host's steward talk; servants tell their counterparts everything. When the foreign ruler's envoy fails to react to the Odes performed at the banquet as a man of breeding should, you say nothing in the hall — but later your own ruler will have his protocol chief privately question that envoy "on his mystifying actions," sparing everyone embarrassment and opening a quiet channel for the real exchange.20 On the ride home you are not composing your message. You are composing your report: roads clear or choked, granaries full or thin, ministers loyal or quarreling. You hand it up, it gets written into one of the five books, and the next time Cheng must decide whether to fight, flatter, or wait, your morning in that court is on the shelf.
The Hsing-jen breaks when the man forgets he is two things at once. The first failure mode is the courier who edits. In Chin, the prime minister passed over one Hsing-jen for a mission specifically because the man "often changed the words of communiqués to interject his own opinions," choosing Tzu Yün instead.21 An envoy who substitutes his judgment for his ruler's words corrupts both the message and the intelligence — you can no longer tell what was reported from what was invented.
The second failure is the envoy who serves himself before the state. Sung's Hsing-jen Yüeh Hao-li was entertained grandly by Chao Chien and, against his own steward's protest, gave him sixty willow shields as a present. At the capital he was denounced for "giving precedence to personal affairs, neglecting official state business, and delaying the execution of his mission," and imprisoned — Sawyer suspecting he was really suspected of arming one of Chin's factions.22 When an intelligence officer starts trading on his access for private advantage, the host stops believing the cover and the home state stops trusting the take.
The third failure is structural: when the office is the hapless victim, killed for tidings that were not its fault. Ch'u executed Ch'en's envoy though "the guilt did not lie with the Hsing-jen."23 A service that loses its officers to other men's quarrels bleeds its own institutional memory — the books stay open, but the men who fill them stop coming home.
The deepest tension is the one Sawyer himself plants: the Chou Li "elaborates a coherent vision of Chou government that was realized only in part."24 The five-book recording system may be a Warring States idealization projected backward, not a sixth-century reality. So the clean intelligence-cycle reading of the Hsiao Hsing-jen rests partly on a document describing how things should have worked. Against that, the Ch'un Ch'iu and Tso Chuan give the harder ground — actual seizures, actual cover missions, actual Cheng workflow — which "essentially substantiate the claim."25 The historical record corroborates the idealized chart without proving its filing system existed.
A second tension lives inside the title itself. The Kung Yang commentary insists one is a Hsing-jen "only when acting on official business, not private concerns,"26 yet the office's whole intelligence value came from the unofficial watching done under official cover. The cleaner the diplomatic mandate, the better the espionage platform — a contradiction the sources never resolve and that recurs through the history of diplomacy.
The notes also widen the picture: the Hsing-jen was not alone. The Chou Li names several lesser officials who "probably performed intelligence-gathering duties because of their extensive contacts with foreign officials" — the Ssu-yi, who handled foreign contacts; the Hsing-fu, a junior aide to the Hsing-jen "stuck with all the miserable jobs"; the Hsing-hsü, who managed envoys from the barbarian states and oversaw their communications with the Chou king; and the Chang-k'o, nominally "handling guests" but apparently supervising the meal service.27 Ch'in seals later preserve a renamed version, the Ch'in-hsing, alongside posts like the tien-song managing submissive barbarian peoples.28 An open question: was there a coordinated system of these offices, or just a scatter of court roles that each happened to touch foreigners? The evidence does not yet say.
Sawyer is reading two kinds of source against each other and mostly making them agree. The Chou Li is prescriptive, possibly fictional in its tidiness; the Tso Chuan is narrative, anchored to dated events. His move is to let the Tso Chuan's real seizures and the Cheng workflow validate the Chou Li's job description — to say the idealized chart was "realized only in part" but pointed at something true. This is persuasive but worth flagging: Sawyer's broader habit of arguing cross-dynasty continuity (Chou office → Han usage → enduring practice) leans on documents compiled centuries after the events they describe, and the Tso Chuan itself was edited in the Warring States period when the relevant vocabulary had already become commonplace. The convergence is real but the dating chain needs independent verification, exactly the caution that applies wherever a translator-synthesizer threads one institution through many dynasties.
To The Semblances Problem. The Hsing-jen's entire value rested on a thing being read for what it is not: a herald who is really a scout, a protocol clerk who is really a foreign minister, a tribute mission that is really a collection run. The semblances problem — that surfaces and depths come apart, that the appearance is engineered to mislead — is usually framed as a danger to the observer. The Hsing-jen inverts it. Here the institution itself is a deliberate semblance: the state builds an office whose ceremonial face hides its intelligence function, and counts on the host being unable to separate the two. The insight neither domain reaches alone: the semblances problem is not only an epistemic hazard to be defended against but a resource to be manufactured. A society sophisticated enough to fear false appearances is sophisticated enough to build false appearances as standing infrastructure — and the same court that worries an envoy may be a spy is the court that sends its own envoys as spies. The fear and the tool are the same competence, pointed in two directions.
To Confirmation Bias as an Ancient Problem. The five-book system of the Hsiao Hsing-jen is, read structurally, an early answer to confirmation bias built into bureaucracy. By forcing observations into separate fixed categories — one book for the contrary and perverse, one for famines, one for harmony — the system refuses to let a single dominant impression of a state swallow all the evidence. A ruler who "knows" Ch'en is doomed cannot file everything under doom; the harvest data goes in one book, the diplomatic insults in another, and they must be reconciled, not blurred. The insight the pairing produces: the cure for biased intelligence in early China was not better individual judgment but enforced categorical separation of the take — a procedural firewall against the mind's tendency to confirm what it already expects. Modern intelligence calls this structured analytic technique; the Hsiao Hsing-jen's five books are a two-and-a-half-thousand-year-old version of the same defense, achieving by filing what no single observer's discipline could.
Sharpest implication: Intelligence in early China was institutionalized before it was theorized. The Hsing-jen office had ranks, recording duties, selection criteria, and a working analytic cycle (the Cheng shop) generations before Sun-tzu wrote the doctrine of spies. The chart preceded the manual — which means the famous Warring States theory was codifying a practice that already had a bureaucracy.
Generative questions: