Chapter 2: What the Documents Know
The brief landed on my desk at 7:04 AM. Gregory Sloane dropped it without ceremony — a manila folder thick enough to require two rubber bands and a sticky note with his handwriting: DUE DILIGENCE. MERGER. 2 DAYS.
"Dressler-Hawkins acquisition," he said. "Mid-size tech merger, nothing complicated. Two boxes of documents in Conference Room C. Client wants a clean memo by Wednesday noon."
"Understood."
Gregory was already moving toward his next task. "And Calder — don't make me look bad."
"I won't."
The door closed. I pulled the rubber bands off the folder and started reading.
Conference Room C smelled like cardboard and recycled air. The two boxes sat on the table like a dare — four inches of documents each, organized with the kind of haphazard precision that suggested a paralegal had done the initial sorting and then given up halfway through.
I opened the first box. Pulled the first document. The Ledger turned.
Synthesis. Ninety seconds.
The Dressler-Hawkins merger assembled itself in my mind like a jigsaw puzzle solving itself: acquisition terms, asset schedules, liability disclosures, the standard representations and warranties that protected both parties from post-closing surprises. The pattern was clean. Too clean. I pushed deeper, and there it was — a thread connecting three documents across different exhibit tabs, a liability disclosure in Exhibit D that contradicted a warranty in Exhibit F that contradicted a footnote in the financial summary.
Gregory had missed it. His review notes were in the margin of Exhibit D, and he'd initialed the page without flagging the contradiction.
[CASE FILE OMNISCIENCE: Dressler-Hawkins Acquisition — LIABILITY THREAD IDENTIFIED]
[WARNING: Two facts in synthesis may reflect user interpretation bias. External verification recommended.]
The corruption warning again. Two facts wrong. I'd accepted two corrupted facts in the Whitfield synthesis, and now I was accepting two more here. The Ledger's accuracy wasn't the problem — the Ledger showed me patterns that existed. The problem was that my brain filled gaps with what I wanted to see, and the Ledger couldn't tell me which gaps were filled with truth and which with wishful thinking.
I pulled out my notepad. The liability thread was solid — I could trace it through the originals, document by document, citation by citation. That one was real. The secondary implications I'd drawn from it, though? The conclusions about opposing counsel's strategy? The assumptions about the seller's motivations?
Two of those were wrong. I just didn't know which two.
I accepted them anyway. I was on a deadline.
Day four. The memo was three-quarters finished when Louis Litt walked past the associate bullpen.
I didn't notice him at first — I was deep in a cross-reference, tracking a citation through four separate documents — but the shift in air pressure was impossible to ignore. Other associates straightened in their chairs. Keyboards went quiet. The ambient hum of focused work became something sharper.
Louis stopped at my desk.
I looked up. He was shorter than the show suggested, but the intensity was accurate — eyes that moved too fast, cataloguing everything, filing details for later use.
He didn't speak. He looked at my document organization: color-coded tabs, three-column margin notes, a system I'd developed over three days of Ledger-assisted synthesis. It wasn't something I'd learned in law school. It was something the synthesis had taught me, a method for keeping track of which facts were verified and which were still pending external confirmation.
Louis looked at it for six seconds. Then he kept walking.
I watched him go. He didn't look back.
"Day four," I noted. "Louis noticed."
That wasn't a predicted behavior. Canon Louis didn't pay attention to new associates this early unless they were doing something wrong. I wasn't doing anything wrong. I was doing something different.
Different was, apparently, interesting enough to earn Louis Litt's attention on day four.
Gregory found me at noon.
"Calder." He dropped into the chair across from my desk, holding a printed copy of my memo. "The liability thread."
I set down my pen. "The contradiction between Exhibit D and Exhibit F. It's also referenced in the footnote on—"
"I saw." Gregory flipped to the second page. His expression was unreadable. "I reviewed that section twice. I didn't catch it."
"The connection is easier to see if you read the documents in a non-linear order. The footnote cross-references a schedule that cross-references the warranty, and the warranty's language is ambiguous enough that the contradiction doesn't surface until you map the definitions."
Gregory stared at me. "You mapped the definitions."
"I work fast."
Silence. Gregory looked at the memo again. Something shifted in his posture — not quite tension, not quite relaxation. Recognition, maybe. Or calculation.
"Good catch," he said finally.
He didn't look up when he said it. He was already reading the next paragraph, already moving on to the next evaluation.
It was the first genuine professional acknowledgment I'd received in this world. I noticed it. I noted that it cost me nothing to feel something about it.
I went back to work.
The memo went to the partner at 11:47 AM Wednesday. Gregory routed it with his own name on the cover sheet, which was standard practice for associate work. The liability thread was attributed to "associate review" — no name, no credit, just the acknowledgment that someone had found it.
I didn't mind. Credit wasn't the point yet. The point was demonstrating value without demonstrating visibility. The point was making Gregory look good so Gregory would keep routing work my way.
The Ledger pressed against my sternum as I walked back to my desk. Two synthesis sessions in four days. Two sets of corrupted facts accepted. The errors were compounding somewhere in my working memory, building a foundation on assumptions I couldn't verify.
"External verification," the system kept warning me. "Recommended."
I'd verify later. I always told myself I'd verify later.
The merger memo was accurate. The liability thread was real. The two corrupted facts were probably irrelevant anyway — background assumptions, secondary implications, nothing that would surface in the final document.
Probably.
The Ledger turned another page in my chest, and I felt the weight shift: new debt recorded, new liability logged. The corruption always sounded true. That was the problem. The facts I wanted to believe were the ones most likely to be wrong, and I wanted to believe the merger was clean, and the synthesis had shown me exactly what I wanted to see.
I sat at my desk. Opened the next file. Kept working.
Two boxes down. Seventeen notes in my pad. Somewhere in those notes, at least four facts were wrong.
I just didn't know which four.
Get Early Access to New Chapters
Thank you for reading. For those who want to skip the wait, my Patreon is currently 21 chapters ahead of the public sites.
Schedule: 7 new chapters released every 10 days.
Benefit: Gain a significant lead of 7 to 21 chapters depending on your tier.
Support the project and start reading the next arc now: Patreon.com/IsekaiStories