The message arrived at Station G on a wet Tuesday evening in November, and by the time it reached Brigadier Alastair Finch on the third floor of the BND headquarters in Pullach, it had passed through six hands and meant something entirely different from what the original sender had intended.

This was not unusual. In the autumn of 1962, information in Berlin was a perishable commodity. Every transmission passed through layers of encoding, decoding, translation, interpretation, and the inevitable corruption that came from too many people trying to extract too much meaning from too few facts. The West German intelligence service, the Bundesnachrichtendienst, was less than a decade old and still learning the trade. Its senior officers had been trained by the Abwehr, the Gestapo, or the Americans — three institutions whose approaches to intelligence work were as compatible as oil, water, and fire.

Finch was British, a loan from MI6 who had been seconded to the BND as a "liaison officer," which meant he spent most of his time fixing problems the Germans had not yet learned to recognize as problems. He was forty-three years old, wore tweed jackets that were too heavy for the Munich climate, and drank tea at four o'clock every afternoon with a stubbornness that his German colleagues found both admirable and faintly ridiculous.

The message, in its original form, had been a report from a field agent in East Berlin — a woman known only by the codename Primrose, who had been cultivating a source inside the Stasi's counterintelligence division for the past eighteen months. Primrose's report was three sentences long:

"Source confirms Soviet nuclear warheads have been moved from the Krampnitz depot to an undisclosed location. Movement occurred on November 3 under cover of darkness. Source believes destination is within fifty kilometers of the Berlin sector."

That was what Primrose had transmitted. What reached Finch's desk was something else entirely.

"CONFIRMED: 12 Soviet tactical nuclear warheads positioned along Berlin sector boundary. Launch authority delegated to local Soviet commander. Estimated response window after detection: 7 minutes. Recommend immediate evacuation of non-essential personnel from West Berlin."

Finch read the message twice. Then he walked down the corridor to the office of Oberst Klaus Behrendt, his German counterpart, and placed the paper on Behrendt's desk.

"Who wrote this?"

Behrendt looked at the message. "It came through the normal chain. Primrose to Station G, Station G to Analysis, Analysis to Operations, Operations to my desk, my desk to yours."

"Six people touched it?"

"Seven, if you count the typist."

"And nobody thought to verify it?"

"The verification is in the chain. Each officer adds his assessment."

Finch sank into the chair across from Behrendt's desk. "Klaus, this isn't an assessment. This is a completely different message. Primrose reported that warheads were moved from a depot. Just moved. She didn't say where. She didn't say how many. She didn't say anything about launch authority or response windows or evacuations. Someone in your chain decided the plain facts weren't dramatic enough and added an entire second act."

Behrendt studied the message more carefully. He was a methodical man, a former signals officer who had survived Stalingrad and the subsequent Soviet captivity, and he did not jump to conclusions. He traced the message back through its chain of custody: the raw intercept, the decryption, the translation from Russian to German, the first analysis, the second analysis, the operational assessment, the final summary.

By the time he had finished, the source of the corruption was clear. It had not been one person. It had been everyone.

The translation from Russian to German had introduced an ambiguity — the Russian word for "moved" could also mean "deployed," depending on context. The first analyst had chosen "deployed" because it was more alarming and therefore more likely to be taken seriously. The second analyst had added the number twelve because Primrose's source had mentioned twelve warheads in a previous report, and it seemed reasonable to assume the same twelve were involved. The operations officer had added the launch authority because that was standard procedure for deployed warheads. The typist had added the evacuation recommendation because she had a sister in Berlin and could not bear the thought of leaving her behind.

No one had acted maliciously. No one had acted incompetently, by the standards of their training. Each person had simply added what seemed reasonable, what seemed probable, what seemed necessary to make the message actionable. And the result of six reasonable additions was a message that bore no relationship to the truth.

"This is how wars start," Finch said quietly.

"This is how intelligence works," Behrendt replied. "The signal is always weaker than the noise. The question is whether you can reconstruct the original before it's too late."

Behrendt called Primrose back from East Berlin — a dangerous extraction that cost them the source she had spent eighteen months cultivating. Primrose came home, debriefed, and confirmed what Finch had already deduced: the warheads had been moved from the Krampnitz depot to a storage facility near Potsdam, a routine repositioning that the Soviets had not even bothered to conceal. There had been no deployment. No launch authority. No twelve warheads pointed at Berlin.

Finch wrote the corrected report himself and delivered it to the British ambassador. The evacuation was canceled. The Americans, who had gotten wind of the original message and were preparing their own response, stood down. Berlin remained Berlin — tense, divided, but not at war.

A week later, Finch sat in his office with Primrose, who was recovering from the extraction and trying to decide whether to return to the field. She was twenty-nine years old, sharp as a blade, and angry about the loss of her source in a way she was trying very hard not to show. The anger was professional — a year and a half of cultivation, gone — but it was also personal. She had known her source, a Stasi clerk named Dieter, not just as an asset but as a human being. He had a wife and two daughters and a dachshund named Fritz, and he had been passing information to the West not out of ideology but out of exhaustion — a bone-deep weariness with the lies and the surveillance and the constant fear. And now Dieter was gone, disappeared into the machinery of East German counterintelligence, and Primrose did not know whether he was alive or dead, and she would probably never know.

"I understand why you're angry," Finch said. "I would be too."

"I'm not angry at you."

"I know. You're angry at the system. At the entropy. At the fact that six reasonable people, each doing their jobs, each making defensible assumptions, turned a routine report into a war scare." He paused. "The question is what you do with that anger."

"What do you suggest?"

"I suggest you go back to Berlin. Not to run Dieter again — that network is burned. But to build new networks, with shorter chains, with fewer hands. To train your analysts to doubt their own assumptions. To make sure that the next time a report like that crosses your desk, you catch it before it reaches mine."

Primrose was silent for a long moment. Then, to Finch's surprise, she smiled. It was a thin smile, a winter smile, but it was genuine. "You're offering me the chance to fix the system from the inside," she said.

"I'm offering you the chance to try. Whether you succeed is another question entirely."

She went back to Berlin the following spring, not as a field agent but as a trainer — teaching a new generation of analysts to recognize the fingerprints of entropy on their own reports, to distinguish signal from noise, to be suspicious of certainty and comfortable with ambiguity. It was less exciting than the field work, less romantic, less likely to earn her a medal or a mention in the history books. But it was also, Finch believed, more important. Wars began in briefing rooms, not on battlefields. And wars were prevented in briefing rooms too, by people who were willing to read the original message before six other people had rewritten it.

"They destroyed a year and a half of my work," she said. "Not a mole. Not a leak. Just... entropy. Six people making reasonable assumptions."

"That's the frightening thing," Finch said. "There was no enemy. No sabotage. No one trying to start a war. Just the natural tendency of information to degrade as it moves through a system. Every hand that touches it adds noise. Every mind that interprets it adds bias. And by the time it reaches the person who has to act on it, it's been transformed into something that would have been unrecognizable to the person who sent it."

He walked to the window and looked out at the Bavarian countryside. The trees were bare, the fields brown, the sky the color of a bruise. Somewhere beyond that sky, on the other side of an arbitrary line drawn by men who had never met each other, there were Soviet analysts doing exactly the same thing he was doing — taking fragments of information and assembling them into a picture that might or might not correspond to reality. And somewhere in the middle, in the gap between what was true and what was believed, nuclear war was waiting like a patient predator.

"The Cuban missile crisis was eight weeks ago," Finch said quietly. "The world came closer to annihilation than anyone outside a handful of briefing rooms will ever know. And do you know what I learned from that?"

"That the Americans and the Soviets are both insane?"

"That every piece of intelligence that crossed my desk during those thirteen days was wrong. Not maliciously wrong. Not incompetently wrong. Just wrong — degraded by the same process that degraded your report. The signal had passed through so many hands that by the time it reached the people making decisions, it was mostly noise. And the decisions were still made. And the world is still here. But not because anyone knew what they were doing. Because we were lucky."

Primrose was silent for a long moment. Then she said, "You're saying our entire profession is built on noise."

"I'm saying our profession is built on the hope that the signal is stronger than the noise. And that hope is not always justified."

"So what do we do?"

"We trust fewer hands. We write shorter messages. We doubt every piece of intelligence that has passed through more than three people." He paused. "And we remember that the most dangerous thing in this business is not the enemy's lies. It's our own reasonable assumptions."

Primrose looked at him for a long moment. Then she poured herself a cup of his tea — the first time any German colleague had done so — and said, "You're not as stupid as your jacket makes you look."

"High praise," Finch said. "Coming from you."

Outside the window, the Bavarian autumn was turning to winter. The trees were bare now, the fields brown, the sky the color of old pewter. Somewhere in Berlin, a file clerk who had never known the chain of events she had set in motion was stamping routine paperwork and thinking about what to have for lunch. And somewhere in a bunker beneath the divided city, a report that had never been written was being filed under a heading that no one would ever read: The War That Did Not Happen, Because A British Officer Preferred To Verify His Sources. And somewhere in East Berlin, a Stasi clerk named Dieter — alive, it would turn out, transferred to a desk job in Leipzig — was teaching his daughters to ride bicycles in a park that overlooked a wall that would, in twenty-seven years, come down.