She also began to study, in the evenings, after her shifts, with a secondhand technical Korean textbook she bought at a used bookstore near the ferry terminal, and a notebook in which she wrote down terms she did not know and the sentences in which she had found them. She did not study because she believed studying would make her a different person. She studied because she had learned, in a private dining room, that the distance between knowing a language and being able to use it responsibly was a distance you crossed by work, not by confidence.
And then, one morning, an envelope arrived at the restaurant, addressed to her, with Northgate's return address on it, and inside it was a letter signed by Aaron Kline. It was not a long letter, and it did not praise her in the way letters of praise usually do, by exaggerating. It described, exactly, what she had done: she had translated a sentence chosen by Chairman Han, she had translated a technical paragraph accurately, she had reviewed a Korean disclosure packet and identified a difference between the summary and the record concerning patent title and laboratory certification, and she had done all of it while stating clearly what she could and could not certify. It did not call her an engineer. It did not call her a certified translator. It called her disciplined, and it said that her discipline in distinguishing what evidence had established from what had only been assumed or hoped for was the quality that had made her work useful.
Selma read the letter three times, and then she put it in the drawer with her notebook and the fund agreement, and she went to work, because work was what she did when she did not know what else to do with a feeling too large to name.
By the fourth week, the technical team had finished the additional testing, in a wider range of operating conditions than the original schedule had allowed, and the results were neither a triumph nor a disaster. They were, like most honest results, mixed. The module met some of the expected parameters, and it met them well, under normal conditions and under several kinds of stress. But one test, run at a temperature and load combination the original twenty-one-day window had not covered, found an issue that needed more work, more time, and more money. It was not the kind of issue that meant the module was worthless. It was the kind of issue that meant the original schedule had been unrealistic, and that anyone who had promised a closing date based on that schedule had promised something the technology could not yet deliver.
The technical report was sent to both companies, and this time every conclusion came with the evidence that supported it, and every gap came with a note saying it was a gap. Lee Ji-yeon read it with her pencil, and she made fewer marks than usual, because there was less to argue with. Kim Dong-hyun read it and nodded, the way an engineer nods when a report finally describes the thing he built instead of the thing someone wished he had built.
The deal team revised the financial model to account for the additional work and the revised schedule, and the number at the top of the model, three hundred and eighty million dollars, stopped being treated as a prize that could not be questioned and became what it had always been: a starting point for a more precise discussion. The modelers, who were quiet people accustomed to being ignored until a number was needed, found themselves listened to for the first time in months, and they produced a model that was less impressive and more true, which is the best thing a model can be.
The internal review, when it finished, cleared Farah Nasser completely. It said she had followed the correct process, that her warning had been specific and timely, and that she bore no responsibility for the decision to change the summary, a decision that had not been hers. It said Dennis Vogel had been removed from verification responsibility, and that he would be stripped of his leadership role and required to complete additional compliance training before he could hold a position of authority over a deal again. It said the deal team would also be reviewed, not because every member had done something wrong, but because a process that allowed a summary to go out unchecked was a process that needed to be examined, not just a person.