She had not asked for it; Hendricks had simply included it in the authorization, a line of foresight she registered and filed away. The program's own incident binder for the previous spring listed four training injuries serious enough to record. The aid-station log for the same months listed nine visits tagged to combatives training hours — nine names, nine dates, nine dispositions in a medic's cramped handwriting. Four of the nine matched the binder. Five did not appear in the program's records at all.
She checked the dates twice, then a third time, because the first rule of the work was that the discrepancy was usually your own error. It was not her error. The five unmatched visits were not hidden, exactly; nobody had falsified anything. They had simply been classified at the point of origin — walked off, iced, returned to training, no report filed — by whoever had been running the mat that day, exercising a judgment that the system left entirely to them.
One entry she read three times. A private first class, twenty-three years old, sent down in the late afternoon with a shoulder that had, in the medic's phrase, "come out and gone back." Disposition: sling, light duty, follow-up. In the binder: nothing. On that week's schedule, in Voss's slanting hand, the session was marked with a small star — his symbol for the days the cadre called, among themselves, full-send days.
Claire copied the entry into her pad without comment, in the same even hand as everything else. She did not yet know what it meant. A single unmatched record meant a busy afternoon and a judgment call. Five meant a threshold — a line, drawn somewhere inside the program's culture, below which an injury was not considered an injury. And a threshold like that was never drawn by accident, and almost never drawn by the person whose name would end up next to it in a briefing.
She closed the binder, aligned the stack by year, and sat a moment in the ticking quiet of the office. Through the wall came the faint, rhythmic thump of the hall's heavy bag, though the building should have been empty; someone was working late, alone, hitting something that did not hit back. She did not need to guess who it was, or what he was rehearsing, or how it would end — the same kick, thrown at the same bag, landing exactly where he aimed it, proving nothing at all.
"You can't fix a record by re-running it," she said quietly, to the empty room, in the words of a woman four years gone from her life. Then she opened the next binder.
The briefing room had a projector that worked now, because Claire had fixed it during her weeks of invisibility, and she thought about that while the cadre filed in.
Hendricks sat at the side, deliberately not at the head. Voss sat at the head, because it was his program's room, and his cadre arranged themselves around him the way iron filings arrange around a magnet: Felix Grant to his right, two staff sergeants, the master fitness trainer, a medic with the aid-station log. Captain Baird had come without being required to, and sat at the back with a notebook she did not open.
"Lieutenant Donovan," Hendricks said. "The floor is yours. Facts only, as agreed."
Claire put up the first slide. It had no unit crest, no motto, no photograph. It had a number.
"Over the last three training years," she said, "this program recorded one hundred and twelve injuries in combatives training. That is the program's own count. The aid-station log records one hundred and seventy-four visits in the same period coded to combatives hours. The gap between those two numbers is my first finding, and I want to be precise about what it is. It is not fraud. Nobody falsified a record — I checked, and I want that on the record too. The gap is a reporting threshold. Injuries below a certain severity are being classified, at the mat, as not injuries. Walked off. Iced. Returned. The judgment is made by whoever is running the session, and the system, as written, allows it."
She let that sit. Voss's face was a wall being built in real time.
"Second finding." The slide changed: a rate, per thousand training hours, three lines across three years. "Fort Calder's rate of training injuries in combatives is roughly double that of the two comparable programs I was given data for. The trend is not flat. It rises in each of the three years, and it rises in step with total training hours — which means it is not explained by one bad month, or one cohort, or one instructor's rotation."