The entry, and the number others could stand on

"The word unconfirmed doesn't undo the fact of the claim sitting in an official record. If command pulls that log — and they will, eventually, everything on this ship gets pulled eventually — they see a documented correlation between a storm pattern and a solar feed with your name attached to it, and the fact that you hedged it in one clause doesn't change what the entry is."

"It changes what it says." Bram set the sensor down, careful, the way he was careful with everything that wasn't his own composure. "I didn't write that there's a correlation. I wrote that I noticed something that might be one, and that I haven't confirmed it, and that somebody should look. That's not a claim. That's an honest description of what I actually have, which is: not much, and I said so."

"Then it shouldn't be in the log at all. It should be in your own notes until you do have something, and then it goes in the log as a finding, properly supported, not as a — " William searched for the word, and did not love the one he landed on, but used it anyway, because it was accurate. "Not as a guess with a disclaimer stapled to it."

"It's not a guess. I saw two spikes on the monitor feed and I saw two clusters of eddies and the timing was close. I don't know what that means. I wrote down that I don't know what it means. What I'm not going to do is wait until I'm sure before I write down that I saw it, because if I wait for sure, and it turns out to matter, the honest version of what I actually noticed today is gone. Nobody has it. You'd rather have nothing on the record than something with a question mark on it."

"I'd rather have something correct on the record."

"It is correct. It's correct that I don't know. That's the whole content of the entry." Bram said this without heat, which somehow made it land harder than if he'd raised his voice, the way a flat fact lands harder than an argued one. "You keep treating uncertain like it's the same thing as wrong. It isn't. Wrong is when you say something's true and it turns out it wasn't. Uncertain is just — telling the truth about not knowing yet. I don't understand why that's the version you don't want written down."

William did not have a ready answer to this, which he found, in the moment, more unsettling than the entry itself had been.

Next chapter, the second review:

William's second review came back that afternoon, and it came back the wrong way for William and the right way for everyone else: the eddy timing and the monitor spikes weren't roughly aligned, they were aligned to within the granularity of the sampling itself, six events, six matches, no outliers, a correlation clean enough that even a man built to distrust a hedge couldn't hedge his way out of it.

Two chapters on:

Later that afternoon, reviewing the day's data uploads, William flagged Bram's original correlation entry — the one that had started the argument — and cited it, without further discussion, as the baseline reference for the new lander's expected signal window. He did not apologize. Bram did not expect him to, and would, if pressed, have said honestly that he didn't need one; what he'd wanted, the whole time, wasn't to be told he'd been right. It was simply for the number to be used the way a number was supposed to be used, which was as something other people could stand on without checking first whether it would hold.

It held. Nobody said so out loud. William simply built on top of it, the way you build on top of a foundation you've stopped worrying about, and Bram understood, watching him do it, that this was the actual apology, delivered in the only currency William Wolfgang Talbot had ever reliably dealt in.