He found Miron in the Archive on the twelfth day.

This was not where he had expected to find Miron, who had told him, over the course of their conversations, that he spent his professional time in the council records office and his personal time avoiding anything that reminded him of the council records office. The Archive, with its walls of organized documentation and its atmosphere of institutional memory, seemed like exactly the kind of place that would make such a reminder.

But there he was, in the section that held the oldest linguistic materials, sitting cross-legged on the floor between two shelves with a large volume open across his knees and the expression of a man who has found something that has made all the professional inconvenience of his life worth it.

Renji said: I did not know you came here.

Miron looked up with the startled blink of someone brought back from a significant distance. He said: Oh. I come here sometimes. When I need something. He looked back at the volume and then up at Renji with a different quality of attention. You have been here for thirteen days now.

Renji said: Thirteen, yes.

Miron said: You do not look like someone who has been in an unfamiliar world for thirteen days.

Renji said: I look like someone who has been in a difficult archive for twelve days.

Miron said: Yes. That is what you look like. He considered. It suits you.

Renji came around to see what Miron was reading. The volume was in a script he did not recognize, dense, with annotations in three different inks. He said: What is this?

Miron said: Etymological records. Language evolution maps. This is the source material for a project I have been working on for seven years. He paused. Eleven years if you count the previous iterations that did not survive my methodology changes.

Renji said: What is the project?

Miron said: The relationship between certain key terms in the Verdant Prophecy and their precursor forms in the oldest available linguistic records. He looked at Renji with the expression of a man who has been waiting a long time for someone to ask him about this project. Would you like to know what I have found?

Renji sat down on the floor across from him. He said: Yes.

Miron opened to a section marked with a piece of cord. He said: The word Drakonir, in the oldest available records, is a compound of two root terms. Drako, meaning stone or deep earth, and nir, which in the oldest script systems means people or community. The compound means, most literally, the people of the deep stone. Or the stone-earth community. Or, and this is the rendering I find most accurate to the original semantic field, the ones who belong to the deep ground.

Renji said: Not dragon.

Miron said: The word for dragon in Elyndor's language evolved separately, from a completely different root. The word in the oldest texts that was subsequently translated as dragon in references to the Drakonir is this compound, Drakonir itself. Someone, at some point, rendered it as dragon. Whether through genuine linguistic confusion or deliberate mistranslation, he paused, is what I have been trying to establish for eleven years.

Renji said: When did the rendering change?

Miron said: Approximately three hundred years ago. Before that, the oldest records that mention the Drakonir describe them consistently as a people, a community, a political entity. The diplomatic records describe negotiations, agreements, correspondence. After the change, references to Drakonir in new documents describe a creature. A threat. An it rather than a they.

Renji said: The translation changed when the Prophecy was modified.

Miron said, with the expression of a man who has spent eleven years waiting for this confirmation: Yes. Exactly then. I have been trying to establish the causal relationship versus the coincidental one for years, but I lacked access to the Archive materials that would confirm it.

Across two shelving units, behind a stack of old administrative volumes, something small and moss-green moved. Renji caught it in his peripheral vision: Fen, apparently having been present for this conversation longer than either of them had known, repositioning on a higher shelf with the serene air of a creature attending a presentation it had arranged.

Miron had not seen the fox. He was still reading. He said: I have also found, in the oldest linguistic records, references to a text called the Veylwood Covenant. The term appears in six places in these etymological records, always as a reference to a document that the text assumes its reader is familiar with. Never as a citation, always as an assumption. The way you reference a legal framework that everyone knows.

Renji said: Not the Verdant Prophecy.

Miron said: A different name. Entirely. A document that predates the name Verdant Prophecy by, based on the dating of the linguistic records that reference it, at least two centuries. A document that the language of its time treats as a foundational covenant rather than a prophetic vision.

He looked at Renji over the top of his volume.

He said: I have been trying to tell someone this for eleven years. There was a covenant. Before there was a prophecy, there was a covenant. And then the covenant became a prophecy, and the Drakonir became a dragon, and everything that had been an agreement became a destiny that only one side was required to fulfill.

Renji wrote: Veylwood Covenant. Original name. Covenant not prophecy. Agreement not vision. The Drakonir, a people, not the Black Dragon, a creature. The language was changed when the document was changed. Three hundred years ago. Same period. Same mechanism. The modification was total: the document itself, the surrounding documentation, the language used to describe the parties involved.

He said: Miron. I need to ask you something.

Miron said: Please.

He said: Have you told anyone else this?

Miron said: I have submitted three papers to the council linguistic archive in eleven years. All three were filed and not distributed. He paused. I was told the subject matter was insufficiently relevant to current administrative practice.

Renji said: And you stayed anyway.

Miron said, with the expression of a man who has been making peace with a complicated decision for years: The wine is very good. He paused. And I could not leave without knowing whether I was right. He paused again. I am right, am I not.

Renji said: Yes.

Miron was quiet for a moment. Then he said, quietly, with the specific weight of someone who has been carrying a thing alone for a very long time: Good. That is. Good.

From above, Fen made a small sound.

He wrote that evening: three of us now. Miron with the linguistic evidence. Lysera with the archival evidence. Me with the document analysis. Each working on a different angle of the same problem. Miron for eleven years. Lysera for sixty. Me for thirteen days. We should have been working together from the beginning. We are working together now. That is the relevant thing.

He thought about the eleven years and the sixty years and the thirteen days, and he thought about the specific shape of a problem that had been correctly identified by multiple people in isolation and had remained unsolvable in isolation, not because the problem was unsolvable but because solving it required a kind of collaboration that the institution containing the problem was designed to prevent. The Archive kept its keepers separate from the linguists. The linguists were not given access to the sealed documents. The sealed documents were not accessible to anyone without explicit authorization that was controlled by the person with the most to hide.

The isolation was the mechanism. He was going to break the isolation.

In the days after finding Miron in the Archive, the three of them began to work in a more coordinated way than they had before, though the coordination was largely unspoken: each of them had a domain, and the domains overlapped at the edges in ways that produced the clearest picture.

Miron's domain was the linguistic history. He had eleven years of notes that he began, gradually, to share with Renji and Lysera: the evolution of key terms in the Prophecy, the relationship between those terms and their precursor forms, the places where the etymology diverged from the expected path of a natural linguistic evolution and suggested instead a deliberate intervention in how a word was understood.

The word Drakonir was the most important. But it was not the only one. The term for the sacred forest itself had shifted, over the modification period, from a word meaning the jointly held green, with the jointly held being the significant qualifier, to a word meaning the protected forest, with no qualifier about who was doing the protecting or who had rights in the protecting. The shift was subtle. The difference in meaning was not.

Miron had found seven such shifts, each clustering around the modification period, each moving in the same direction: from language that described a bilateral relationship to language that described a unilateral one.

He said, when he presented these to Renji and Lysera in the small reading room: the linguistic modification was systematic. It was not a single decision about a single word. It was a program of revision across the vocabulary used to describe the relationship between Elyndor and the Drakonir.

Lysera said: A program of revision implies a plan.

Miron said: Yes. Exactly. This was not opportunistic. This was designed.

Renji said: The administrative modification was also systematic. The document modifications, the reclassification program, the creation of the three-tier restricted system to protect the materials that would expose the modifications. These are all elements of the same program.

Miron said: The linguistic program and the administrative program were coordinated.

Renji said: Yes.

They sat with this for a moment.

Miron said: Someone had the authority to direct both the administrative archive and the linguistic use of official language at the same time. Those are different domains. In any institution, coordinating them requires significant authority.

Renji said: The High Oracle.

Lysera said, and it was the first time she had said it directly, in that specific phrasing: Caerion.

She said it with the specific quality of someone who has been circling a thing for a long time and has decided to stop circling. Not dramatically. Just: decided. Caerion.

The name sat in the room for a moment.

Renji said: Yes.

Miron said, after a pause: He was High Oracle three hundred years ago.

Lysera said: He has been High Oracle for two hundred and forty years. He became High Oracle sixty years before the modification period. He has held the position continuously since.

Renji said: He has held the position for long enough to have designed the program, implemented it, and spent the following three centuries managing the concealment.

Miron said: That is a very long time to maintain a deception.

Renji said: He is not maintaining a deception. He is maintaining an interpretation. That is different. A deception requires you to know the truth and conceal it. An interpretation, once it becomes institutional, becomes the truth within the institution. He may not think of himself as concealing the original Covenant. He may think of himself as protecting the Prophecy, which is what he believes the Covenant is.

Lysera said: Or he may be well aware of what he is doing and has had three hundred years to become comfortable with it.

Renji said: Both are possible. The distinction matters for understanding him. It does not matter for the documentation.

Miron said: We cannot prove which it is from the evidence we have.

Renji said: We do not need to. We need to prove that the Prophecy is a modification of the Covenant and that the modification changed the terms in specific, documentable ways. The question of Caerion's internal experience of what he is doing is for someone else to determine after the fact.

He wrote, that evening: the three of us see different aspects of the same thing. Miron sees the language. Lysera sees the administrative structure. I see the documentation patterns. Together we have a picture that none of us would have had alone. This is the thing that the institutional isolation was designed to prevent. The Archive keeper was not supposed to know the linguist. The linguist was not supposed to have access to the administrative records. The outside auditor was not supposed to meet either of them in an informal setting.

We met anyway.

I think we were supposed to.

He thought, after the meeting in which the three of them had first worked as a coordinated unit rather than as three isolated people who happened to be investigating the same thing, about the specific quality of what that coordination produced.

Individually, each of them had a portion of the picture. Miron had the linguistic evidence, which said what the Drakonir were and when the word for them had changed. Lysera had the archival evidence, which said what documents had been moved and when and in what pattern. He had the documentary evidence, which said what had been changed in the text itself and how the changes had been concealed.

Separately, each portion was suggestive but not conclusive. It was the kind of evidence that an experienced institutional defender could argue around, pointing to alternative explanations for any individual piece while acknowledging the overall pattern was interesting.

Together, the three portions were different. Together they triangulated. Each piece of evidence that Miron had found was consistent with what Lysera had found, which was consistent with what he had found, and the consistency was not coincidental because the things they had each found independently were different parts of the same event described from different angles. Three different disciplines, three different archives, three different starting points, one conclusion.

A conclusion that three independent lines of evidence converged on was of a different quality than a conclusion that one line of evidence pointed toward. It was the quality of a thing that was very probably true rather than the quality of a thing that might be true. And very probably true, in his professional life, was the category of conclusion that produced results.

He thought: we are close. We have been close for a while without knowing it. We needed each other to see how close.

He thought: I am glad Fen bit the corner of that document. He thought: I would like to understand what Fen is and why it does the things it does. He thought: that is probably not the most urgent question. He wrote it down anyway, in the margin of his notes, as a reminder that some questions were worth asking even if they were not urgent.

He thought, late that night, about the nature of the thing they were doing.

Three people, in a kingdom that had been organized around a lie for three hundred years, working to expose that lie. None of them with formal authority to do so. None of them with institutional backing for the work. Each of them with a different kind of standing: he had the mandate of the verification consultant role, narrow and temporary; Lysera had the institutional standing of the Archive Keeper, which was real but which Caerion could revoke; Miron had no standing at all in the formal sense, only eleven years of research that no one had been willing to distribute.

And yet together they had something that formal authority rarely produced: independent lines of evidence converging on the same conclusion. That convergence was its own kind of standing. Evidence did not require permission to exist. It was either there or it was not.

He had learned this in his first year of audit work, when he was twenty-six and had assumed that finding the wrong numbers and documenting them correctly would be sufficient to produce a response. It had not been sufficient, because the wrong numbers were in someone's interest and the documentation was in no one's interest and interest, not evidence, was what most institutional decisions were made on.

He had learned, in the years since, to look for the conditions under which evidence became the decisive factor rather than an inconvenient complication. Those conditions existed. They were specific and they were not easily manufactured, but they were real. They required the evidence to be complete and undeniable, the exposure to be public enough that managing it was worse than acknowledging it, and the cost of acknowledgment to be lower than the cost of continued concealment.

He was working on all three conditions. He had eight days.

He thought about the specific shape of the collaboration that had developed between the three of them.

Miron was the most expressive of them, the most likely to narrate his own thinking, to explain his process as well as his conclusions. He had eleven years of work that he had been unable to share and had developed, in the absence of an audience, the habit of explaining things very completely, as if making up for the years of having no one to explain them to. This was occasionally more explanation than was needed but was consistently useful, because the completeness meant that nothing was assumed and everything was available to be cross-referenced.

Lysera was the most precise of them, the most careful about the distinction between what she knew and what she inferred, the most likely to correct imprecise language before it could be mistaken for established fact. She had sixty years of working in an institution where the cost of imprecision was high, and she had internalized the standard of precision to a degree that was now indistinguishable from character.

He was, he thought, somewhere between them: more willing than Miron to hold back the explanation and more willing than Lysera to allow the inference to be stated before it was fully established, as a working hypothesis rather than a finding. This had gotten him transferred four times. It had also gotten him to conclusions faster than a more conservative approach would have, because hypotheses, stated clearly and tested rigorously, were a more efficient route to conclusions than waiting until every inference was fully established before stating it.

Together they made a good team. He thought this was not coincidence. He thought it might be the reason they had each been drawn to the same problem from different angles: problems of this kind required exactly the combination of approaches they embodied.