Chapter 8 - What the General Knew
Colonel Adam Ward arrived for his second interview carrying no secret recording, no miracle confession, and nothing anyone could call proof until it was checked against the documents.
He had an appointment calendar, a contemporaneous note, and an account of a meeting that had happened behind an ordinary office door.
"I was told to make the summary defensible," he said. "The general believed the test didn't represent field conditions. He wanted a clear answer for the secretary, not a page of caveats."
"Did you tell him the field incident was linked to the same replay condition?" the investigator asked.
Ward swallowed. "The depot event hadn't happened when we first discussed the wording."
"What about when you circulated the final document?"
"We had an initial incident notice by then. I attached it to the package for review."
"Did General Sterling see it?"
"I put it on his desk. I don't know when he read it."
That uncertainty was important. A document sitting on a desk was not the same as a person knowingly concealing its contents. But Ward's calendar placed a review meeting after the notice arrived, and two staff members remembered discussing the depot discrepancy there. Both provided interviews independently. Their accounts matched on the fact of the discussion and differed on exactly how Sterling had responded.
One remembered him saying, "Don't confuse an incident investigation with a certification." The other remembered, "The exception goes in the next review cycle." Neither claimed he ordered records destroyed. The program's problem was less theatrical and more recognizable: people had taken an unresolved safety issue and moved it past the decision it was meant to inform.
Dr. Brooks's team issued an interim engineering finding. The replay weakness was real under the field-approved fallback configuration. The version of the certification that went to decision-makers did not reflect the known limitation. The Hickory incident data was consistent with that weakness occurring, though the investigation could not attribute Sarah Kim's injury to a single system failure alone.
When I read the finding, I waited for the fierce satisfaction I had imagined at three in the morning, alone with my notes. It never came. The system hadn't been fixed yet. Mara was still out of her office. Sarah was still learning stairs with a brace. An honest technical finding was a beginning, not a victory parade.
My administrative reviewer delivered her determination the following week. I had violated internal source-handling guidance through negligent inclusion of an identifying detail in a distributed summary. She recommended a written reprimand and additional training, with the opportunity to respond in the established process. She also found I had reported the exposure promptly and hadn't instructed Ben to delete Isaac's original statement.
"You may disagree with my findings," she said, sliding the memorandum across the desk. "You may submit a response before final disposition."
I read each line. "I won't dispute what I did. I want to add the steps we took to preserve the source after disclosure."
"That's appropriate."
Outside her office, Ben was waiting with two paper cups of tea.
"You look like somebody handed you a bill," he said.
"They did. It's mine."
He handed me the cup. "I have one too."
His counseling memorandum identified a separate failure to preserve a witness reference and a failure to ask for clarification. He accepted it with less drama than I had expected. Maybe we'd both finally learned that admitting a mistake wasn't the same as surrendering our worth.
In the program office, meanwhile, an urgent operational question arrived. Two field sites could not meet the proposed secondary-confirmation requirement with their current backup radios. The interim mitigation that looked sensible in a Washington briefing wasn't immediately workable everywhere. Dr. Brooks asked site personnel to identify a safe procedure rather than issue a command they couldn't follow.
One remote installation needed a human relay point staffed continuously during alerts. Another had to use a redundant line that failed during heavy rain. Sterling circulated those problems as evidence that the review had gone too far.
Hayes disagreed in a formal policy note. "An instruction that cannot be followed is not a safety measure," she wrote. She directed program managers to correct the implementation plan, prioritize equipment where the alternate communication path was weakest, and document any residual risks that commanders would have to manage.
It reminded me of my first sergeant's warning: a briefing isn't the people who must live with it.
Then Park called with the last interview update.
"Sterling acknowledges attending the meeting and hearing the discussion," she said. "He disputes understanding it as a field-proven defect. The investigators will evaluate that against the technical records."
"So he knew there was a warning."
"He knew there was an unresolved question. Don't simplify it beyond that."
I accepted the correction.
"One more thing," Park said. "Sarah Kim and Mara Ellis have asked to attend the program's safety review. They want to put their own questions on the record."
The meeting would be open only to authorized participants, with technical details handled separately. I had been removed as the briefing lead, and neither woman owed me a conversation.
Still, I wanted to hear them.
May you like
"Do they want me there?" I asked.
"That," Park said, "is for them to decide."
