Updated Mar 23, 2026 • ~9 min read
Chapter 2: What the application said
DANTE
Her grant application had been thorough enough to flag.
He’d seen a hundred grant applications. Most of them were versions of the same document: regional population studies, habitat corridor assessments, the standard frameworks that produced the standard data that populated the regional databases that he and his predecessors had spent thirty years managing carefully. He read them all. He approved or denied based on criteria that were not in any official guide but that had produced thirty years of clean outcomes: grant the access, manage the terms, deliver the data that fit the framework, protect what needed protecting.
Dr. Camila Reyes’s application had been different.
He’d identified the difference in the second paragraph of the methodology section, where she’d described the anomaly in the population data and the seven months she’d spent verifying that the anomaly was real before deciding the explanation required ground-level access. The language was precise. The statistical reasoning was sound. She hadn’t framed the anomaly as an error to be corrected — she’d framed it as a phenomenon to be explained. That was a different kind of researcher.
He’d sent the application to Rosa, who had read it and said: *she’s going to find the edge of the picture.*
He’d said: *the edge of the picture is manageable.*
Rosa had said: *it always is. Until the person looking at it is too good at looking.*
He’d approved the application.
He’d been telling himself for three months that he’d approved it because population dynamics research was manageable territory — broad category, standard methodology, nothing that required access to the northern sectors where the exposure was real. He’d approved it because the grant window was wet season, which limited the physical range she could cover. He’d approved it because the regional database upload had a gap in the conflict incident data that needed legitimate scientific attention before it became an anomaly someone else flagged from outside.
He sat on the covered veranda in the rain and looked at the notes he’d made during the afternoon session and thought about whether any of those reasons were the actual reason.
She’d been on site for six hours.
In six hours she’d asked about the survey methodology — fine, he’d expected that. She’d asked about the territorial conflict data — also fine, the gap was documented, he’d offered the local dataset, managed transition. Then she’d asked about the prey depletion data and he’d said *adjacent land management* and she’d said *that’s an unusual degree of coordinated land management for a private station’s operational capacity* and the way she’d said it was the way a person said something when they already knew the answer wasn’t sufficient and they were recording that they’d asked.
She was building a record.
He’d seen researchers take notes before. Most of them were recording their observations — the thing itself, what they’d seen or measured. She was recording the shape of the conversation. Not just what she’d learned but what she’d asked and what kind of answer she’d gotten and what the answer implied. That was a different kind of note-taking.
He put down his pen.
The jaguar was alert in a way it hadn’t been in a long time.
This was information he was not prepared to deal with tonight. He noted it, set it aside, and went back to the practical problem, which was: she was going to survey the southern sector tomorrow at dawn with Marco, and Marco was not good at withholding information from people who asked the right questions in the right way.
He’d given Marco a specific briefing before she’d arrived. Stick to the monitoring protocols. Answer methodology questions fully. Redirect anything that moved toward the northern boundary. Marco had said yes, understood, he would handle it.
Marco had also said yes and understood to approximately forty similar briefings over the past six years and had a perfect record of meaning well and delivering something else.
He stood, walked to the edge of the veranda, looked at the river. The wet season had brought the water to the level where the forest was beginning to flood — the lower canopy trees standing in a foot of brown water, the understory submerged, the whole ecosystem in the suspended state of high water when things moved differently and followed different rules. He’d been watching this river for thirty years. He knew every stage of it.
He did not know how to handle a researcher who asked about prey depletion in the first six hours.
He thought about what she’d said before she left: *I’d like to be transparent with you about my research focus.* And then she’d been transparent, with the specific quality of someone who was transparent not because they expected transparency back but because they’d decided it was the approach most likely to produce the outcome they wanted. She was right. He’d responded to it. He’d told her the adjacent territory management existed — he’d phrased it carefully, but he’d confirmed it — because she’d offered him her actual research purpose and the transaction had required something equivalent.
She’d said: *I’m not here to produce a result that fits the model. I’m here to find the actual explanation.*
He thought: *I know.*
He thought: *that’s the problem.*
The practical version of the problem: she was good enough to find the boundary of what the public data showed. In a week, if she worked at the rate she’d established today, she would identify that the territory was being managed at a precision that no private research station had the operational capacity to execute. She would identify that the prey population balance was not the result of natural carrying capacity but of active management across adjacent territories. She would identify that the conflict incident rate was not the result of some unusual territorial stability but of something else entirely, something that didn’t have a name in any ecological framework she’d been trained in.
She would find the edge of the picture.
The question was what she did when she found it.
He’d read her published work before approving the application — three papers, the population modelling methodology chapter, the conference proceedings on territorial range analysis. She wrote carefully. She did not make claims she couldn’t support. When she reached the edge of her data she said so explicitly, in language that distinguished *this is what I found* from *this is what I think it means*, and she was honest about the gap between the two.
That quality was reassuring and alarming in equal measure.
Reassuring: she wouldn’t publish something she couldn’t prove.
Alarming: she would publish everything she could prove, accurately, in careful language, and she would frame the gap between what she found and what it meant as the question her research had opened rather than the question it had failed to answer.
He needed her to find the anomaly — the cover data needed a legitimate scientific account of the population dynamics before the regional database discrepancy attracted attention from somewhere else. He needed her not to find the explanation — the explanation was a species that had no ecological framework, a community of people who maintained a territory that no public institution knew existed, thirty years of land management that operated outside any official channel.
He needed the middle.
The middle was going to require her to reach the edge of the picture and stop.
He had watched this happen once before, in 2014, when a hydrologist had come to study the upstream water chemistry and had found the edge of the picture and had stayed on the right side of it — not because Dante had managed him, but because the researcher had asked *is there something here I shouldn’t document* and Dante had said yes and the man had said *that’s what I thought* and had written his report around the gap with the precision of someone who understood that the gap was intentional and needed to stay.
He did not yet know if Dr. Reyes was that kind of researcher.
He turned back inside.
The rain on the tin roof had settled into the night-rain pattern — steady, heavier than the afternoon rain, the kind that lasted until dawn. He could hear the river from the main building. The forest was doing what it did in high water: contracting into the canopy, concentrating, every animal in the territory adjusted to the new geometry of flood season.
The jaguar was still alert.
He thought: *six hours.*
He thought: *she asked three questions and the third one was the one I didn’t have a clean answer for.*
He thought: *she noticed that I didn’t have a clean answer.*
He wrote in his own record: *Day 1. Three questions. Survey methodology — answered fully. Conflict data — offered local access, she accepted, noted it. Prey depletion — managed explanation insufficient. She filed it. She is building the picture from the data up, not from the framework down.*
He paused.
He wrote: *She said: I don’t know yet what I’ll find. That’s why I’m here.*
He looked at the sentence.
He thought about a researcher who didn’t know what she’d find and was willing to say so. Most researchers knew what they were going to find before they arrived — the grant application was a commitment to a hypothesis, the field work was the evidence collection, the paper was the argument. She’d written a grant application that said: *the data shows something that has no adequate explanation in any framework I know. I’m going to find the explanation.*
That was either very honest or very dangerous.
He put down the pen and listened to the river in the rain and thought that the wet season was going to be longer than he’d planned for.
The jaguar was waiting.
He told it: not now.
It waited anyway.
Reader Reactions