## Chapter 8: The Price of Knowledge

  **Shenyang, Spring 1993**

  His father died on a Tuesday. Eleven-thirty in the morning. His mother had called the previous evening to say the end was close, and Lin Wei had taken the overnight train and arrived at seven AM and sat with his father for four hours.

  His father was mostly not conscious. When he was, he didn't speak much. Near the end, he opened his eyes and looked at the ceiling and said: "The gear that drives the hand. Do you remember?"

  "I remember," Lin Wei said.

  "Good." His father closed his eyes. "That's enough."

  He died thirty minutes later.

  Lin Wei helped his mother with the practical things. The paperwork, the apartment, the objects a life accumulated. He spent two weeks in Shenyang.

  On the last night, he sat at the kitchen table with a cup of tea and the fifteen mechanical engineering textbooks on the shelf above him. Correspondence course. Never finished. Kept.

  He thought about his father, who had wanted to understand how things worked and never got the chance to build anything that mattered.

  He thought about himself — what he was building, and whether it mattered, and what *mattered* meant when you were building things that killed people.

  He did not reach a satisfying conclusion. He went to sleep. He took the morning train back to Beijing.

  He was more careful after that. About which things he pushed and which he let go. About the difference between *this is strategically useful* and *this is right*.

  The difference was smaller than he had previously allowed himself to believe. It was larger than he sometimes pretended.

  ---

  **Beijing, 1994. The Trap.**

  It started as a feeling.

  Not a specific observation — a pattern of small things. A project document accessed at a time that didn't fit the access logs. A technical detail from a working group conversation appearing in a foreign assessment that had no other source for it. A reference in a non-attributed intelligence analysis to a specific test result that had not been in any open document.

  Lin Wei noticed these things the way an engineer notices resonance — not immediately, not with certainty, but with the accumulating unease of a system vibrating slightly wrong.

  He said nothing. He watched.

  After three months of watching, he was certain enough: someone was passing technical information from close to his program. The leak was not from his core team — he had eliminated each member through careful analysis of what information each one had access to. The leak was from the administrative and logistics structure around the program.

  He stayed late one night and sat with a legal pad and wrote out what he knew and what he didn't. Then he wrote out what he could do.

  He could report a suspicion. Suspicion without evidence produced investigations that destroyed institutional trust and rarely found anything.

  Or he could generate evidence.

  He stayed at the institute until eleven on a Thursday, after the last administrative staff had left. He went to the program documentation server — he had legitimate access, he was listed as principal technical contributor — and pulled up the seeker design specification for the anti-ship missile's terminal phase.

  He saved three copies under slightly different version numbers. Then, in each copy, he made one change.

  Version A: the phase detection circuit's threshold voltage. He shifted the value by 4%, which would cause the seeker to lose lock 0.3 seconds earlier in high-clutter conditions. Wrong, but plausibly a rounding error in transcription.

  The story has been taken without consent; if you see it on Amazon, report the incident.

  Version B: the detector quadrant spacing. Off by 1.5 millimeters. Enough to shift the centroid calculation by a consistent bias in one direction. The kind of error that would survive a casual review.

  Version C: the booster trigger range. 4.2 kilometers in the real spec. Version C said 3.8. Close enough to look like an earlier draft. Different enough to be distinctive.

  He distributed them by hand, which required going to three separate filing locations: the program coordinator's physical distribution folder, where Version A went into the Tuesday batch. The technical review committee's document shelf on the third floor, where Version B went into the current-review stack. The logistics support team's shared cabinet in the annex, where Version C went into the support documentation folder.

  He was in the building until midnight. No one saw him at any of the three locations, because he had chosen Thursday evening carefully.

  Then he went home and waited.

  Four months. He was not a patient man by temperament. He was patient when patience was the correct tool.

  He ran the program meetings. He solved the problems that needed solving. He watched the administrative staff the way you watch a machine running slightly wrong — not obviously, not to anyone who didn't know what smooth running looked like.

  The foreign intelligence assessment arrived in the fifth month. Routine distribution from Colonel Liu's office — documents relevant to his technical areas, flagged for his comment. He opened it at his desk on a Wednesday afternoon.

  Page four. A technical characterization of the anti-ship missile's seeker behavior in high-clutter environments.

  The threshold voltage figure cited was his Version A value.

  Lin Wei set the document down on his desk.

  He sat very still.

  He had expected this moment. He had planned for it. He had told himself, in the four months of waiting, that confirmation would feel like resolution — the engineering problem solved, the uncertainty collapsed to a data point.

  It did not feel like that.

  It felt like something had been cut. Not dramatically. Just — removed. A component that had been present in the system and was now absent, and the system was still running, and the absence was quiet, and quiet was worse than loud.

  He knew which version. He knew which channel. He had a list, twelve names.

  He did not look at it that afternoon. He finished the workday. He went home. He ate. He came back the next morning and looked at it.

  He went to Colonel Liu the next morning with a folder and laid out the evidence carefully: the three document versions, the distribution analysis, the foreign assessment with the planted characteristic highlighted, and the list of twelve names.

  Liu read through it in silence.

  "You conducted a counterintelligence operation," he said.

  "I identified a security problem and generated evidence," Lin Wei said. "What happens with the evidence is your department's decision."

  Liu looked at him. "You could have come to me earlier with a suspicion."

  "Suspicion without evidence isn't useful. Evidence is."

  Liu was quiet. He turned to the last page — the list of twelve names. He looked at one name for a moment, briefly, then set the folder down.

  "I see," he said quietly. Not the *I see* of the unexpected. The *I see* of confirmation.

  "You already knew," Lin Wei said.

  "I had a suspicion," Liu said. "Without evidence." A pause. "Now I have evidence."

  He closed the folder and stood.

  "The program security protocols will be restructured," he said. "You will not be informed of the specifics of the investigation." He looked at Lin Wei. "And you will not conduct further independent security operations. If you notice something, you come to me. Immediately."

  "Understood."

  "One more thing." Liu looked at him steadily. "The person who was passing information. They had reasons. I will not tell you the reasons and you will not ask. But I want you to know that the situation was not simple."

  Lin Wei thought about this. About the twelve names and the one that Liu had paused at.

  "I know it wasn't simple," he said. "It never is."

  ---

  That night he did not sleep well.

  He lay in the dark and thought about the twelve names. He thought about Liu pausing at one of them — a half-second, barely perceptible, the tell of a man who had already done the work of confirmation and was looking at a name he recognized in a new context.

  He tried to remember the face that went with it. He could. A logistics coordinator, mid-thirties, someone he had passed in the corridor a dozen times. A man with a daughter — Lin Wei had seen a photo on the desk once, a girl in a school uniform, seven or eight years old.

  He did not know what had happened to the man. He would not be told. Liu had been explicit.

  What he did know — what he lay with in the dark — was that the technical support contract that man had passed information through had been running for three years. Three years. Which meant Lin Wei's programs had been observed that long. Which meant the design modifications he had made, the careful degraded specifications on the export versions, the seeker discrimination changes — all of it had been watched.

  The question was what had been done with the watching.

  He did not know that either.

  At 3 AM he got up and made tea and sat at his desk and did not work. He just sat. The city outside was quiet. Somewhere two floors down, a building guard made his rounds; Lin Wei heard the footsteps, regular and unhurried, and then silence again.

  He thought about the girl in the school photo.

  He thought about the crews in Country A, and the crew identifier in the after-action report, and the ways that decisions made in Beijing offices had consequences that arrived, eventually, as specific names.

  He did not reach a conclusion. He went back to bed at four and slept three hours and got up and went back to work, because that was what there was to do.

  The file with the twelve names he kept in his desk drawer. He did not destroy it.