The agency's own briefing on Della's career arrived Monday morning, nine pages, clean and confident, and it began its account of her conduct in 1988.
Inez had it spread beside a second stack — authorized investigative summaries pulled from the Department of Public Justice's file transfer — before I had my coat off. "Read the first page," she said. "Then read this."
The briefing's first page described a trusted officer whose service record showed no irregularity until an undefined point in the late 1980s. The investigative summary, two documents down in the second stack, referenced financial movement and unexplained contact reaching back to 1982.
"Six years," I said.
"At least six years. The summary doesn't claim certainty past that point either — it just goes further than the briefing does."
I read both documents twice, the way I had trained every examiner under me to read anything that mattered. The briefing might have used a narrower evidentiary threshold. It might have omitted earlier material for another reason. Nothing on the page settled that question, and the investigative summary's earlier date did not make every detail behind it certain. What I could establish was that two authorized accounts offered different starting lines.
"Nolan wants this ready for the management meeting Thursday," I said, half to myself.
"Which version."
"He didn't ask for a version. He asked for a chronology. He'll assume it's the short one unless I tell him otherwise."
Inez waited while I thought that through, not pushing, which was its own kind of pressure.
"We don't pick one," I said finally. "We build two timelines, labeled plainly — what the agency's briefing supports, and what the investigative record supports beyond it — and we send both. No synthesis, no guess at which is complete. Just the discrepancy, named as a discrepancy."
"That'll read as an accusation to some people."
"It'll read as a question to honest people. I can't help how dishonest people read it."
We spent the better part of two days constructing it properly, cross-checking dates against travel authorizations, leave records, assignment logs, anything with an official timestamp that could anchor a claim instead of leaving it floating. By Wednesday evening we had two clean columns on a single page: 1988 forward, agency-supported; 1982 forward, investigation-supported, with the gap between them marked rather than filled.
I brought it to Nolan myself rather than routing it through his assistant.
He read it standing at his desk, which was his habit when he wanted to finish quickly and decide quickly. "This is going to make Thursday's meeting longer."
"It's going to make Thursday's meeting accurate."
"The briefing came from people who've spent years on this, Calum. You're telling me a six-year gap just sat there unnoticed."
"I'm telling you two documents disagree by six years and I don't know why yet. I'm not telling you anyone lied. I'm telling you I'm not going to decide which number is comfortable and report that one."
He set the page down, and for a moment I thought he was going to hand it back to me with instructions to simplify it. Instead he folded it once and put it in his own folder, not mine.
"Send both timelines to the committee ahead of the meeting," he said. "Let them see the gap themselves before anyone has a chance to explain it away."
It was, I realized later, the most honest instruction he gave me during the entire review, and I did not yet understand how much it would cost him to keep giving instructions like that once the gap started pointing toward people closer to him than an unnamed briefing author.
That night I drove home past the Alderwick annex, dark now except for a security light over the door, and thought about how a six-year gap in someone else's chronology was easy enough to report plainly when I had no reason yet to suspect it might eventually intersect with mine.