I've revised the section three times.

The first version was too analytical — I wrote it the way I write archive notes, precise and distanced, the register of someone cataloguing an observation rather than someone inside it. The second version overcorrected into something that felt too close, too much of my own response bleeding into what was supposed to be a description of his. The third version found the balance, or something near enough to balance: his arc described from the outside, with enough of my perspective to make it legible but not so much that it became about me.

I read it again this morning at the table by the window, before the settlement was fully awake.

It's close. Not finished — it will need more revision when I understand more, when I've lived alongside what I'm describing for longer. But it's close enough to show.

I fold it and put it in my coat pocket.

Then I go to the archive.

---

The northern boundary records are on the table when I arrive, exactly where I left them. This is one of the things I've learned to rely on about the archive: its consistency. What you leave in the evening will be there in the morning, in the same configuration, because Theo and Aldren and Kael all understand the difference between material that is in active use and material that is complete. No one moves what doesn't need to be moved.

Kael is at his table with a stack of documents I don't recognize — not the monthly patrol reports, not his usual correspondence. Something different. The quality of his attention is specific: the frown he gets when he's reading something that requires legal interpretation, the slight forward lean that means he's working through a complication rather than absorbing straightforward information.

I settle into my own work.

The notation error I was tracing last week turns out to run deeper than I thought — twenty-five years of faithful copying of an incorrect convention, which means I need to go back further in the records before I can establish where it started. I've been building a methodology for this kind of archaeological correction work, a way of moving backward through the records systematically without losing track of where I am. It's slow. It requires the kind of attention that takes over the whole mind, the pleasant disappearance into a problem that has a solution if you look carefully enough.

I'm in the middle of this when Kael speaks.

"The precedent records," he says. "Northern sector, inheritance disputes. Do you know where they'd be?"

I look up.

He's looking at me with the direct quality he has when he wants information rather than conversation — the working look, the same look he'd give Aldren or Theo.

"Pre-reform or post?" I ask.

"Both, ideally. I need a continuous record."

I think about the archive's organization — the upper archive where I've been working, the lower archive where Aldren keeps the older materials. The inheritance records I've encountered in the northern boundary work have been filed by district and date. A continuous record would require pulling from both sections.

"I can find them," I say. "Give me a few minutes."

I go to the relevant section of the upper archive first — the post-reform records, the past fifteen years, organized in the system I know well. I find the inheritance disputes for the northern sector and pull the relevant files. Then I go down to the lower archive for the older materials, navigating Aldren's system with the confidence I've been building since the day I found the working notebook on the table. The pre-reform records are in the third cabinet to the left — I know this now the way I know the east wall of Dara's boarding house, from having found it more than once.

I bring both stacks upstairs and set them on Kael's table.

He looks at them. Then at me.

"Thank you," he says — and reaches for the oldest file.

I go back to my notation corrections. But something has shifted slightly: he asked me to find something in his domain. Not a collaboration initiated by a shared problem, not me sitting down uninvited at his table. He needed something and he asked me, which means he's been watching me navigate the archive system long enough to know I could find it.

I note this without making a thing of it.

I go back to work.

---

An hour passes. Maybe more. I'm deep in a notation error from twenty-seven years ago when I hear the quality of Kael's silence change — the specific shift I've learned to read as: something found that requires more than reading.

I look up.

He has a file open in front of him. He's reading it with the quality of someone who has found what they were looking for and is not entirely pleased with what they found.

He looks at me. "Come look at this."

I cross to his table.

The file is a case record from forty years ago — an inheritance dispute in the northern sector, resolved by the then-lord's office. The resolution is detailed, the legal reasoning laid out in the register of someone who wanted the precedent to be clear and applicable to future cases. And it is clear: the ruling established that in cases of disputed inheritance where the primary heir had been absent from the holding for more than three years, the secondary heir's claim takes precedence regardless of the original grant terms.

I read it.

I read the date.

I read it again.

Forty years ago. Chapter four of *The Dark Covenant* — I wrote chapter four in a coffee shop on a Tuesday afternoon in February, two years ago. I was building the legal texture of the world, the sense that it had a history and a system that had evolved through real cases and real decisions. I wanted the system to feel imperfect, not cruel but limited — a system that had made reasonable decisions in some contexts and problematic ones in others, the way real legal systems work.

I wrote this case as an example of a problematic precedent.

A background detail. A texture element. Something that said: this world has legal history, and some of that history is complicated.

I didn't write it to be applied to an actual family forty years later.

"I wrote this," I say.

My voice comes out even — the even quality of someone who has realized something and is deciding how to handle the realization.

Kael is looking at me. Not with the expression I'd expect — not surprise, not the adjustment of someone who has been reminded that the world they live in was made by someone sitting at a desk. He already knew that. He's known it for a long time.

"The current case," he says. "Family in the northern settlement. Primary heir was away for four years managing a crisis in another territory — not abandonment, necessity. But under the precedent, his absence creates ambiguity in the claim." He looks at the file. "If I follow the precedent strictly, the secondary heir has the stronger case. That's not the right outcome."

I look at the precedent I wrote.

I wrote it to be imperfect. I wrote it to illustrate the limitations of a legal system that had evolved through case-by-case decisions without consistent principle. I wrote it at two in the afternoon on a Tuesday and didn't think about it again.

"I wrote it as an example of the system getting it wrong," I say. "Not as something that should be applied indefinitely."

Kael looks at me steadily. "Then we don't apply it indefinitely."

"You can't just ignore a precedent."

"No," he says. "But I can distinguish the current case from the precedent on material grounds, and establish a new ruling that better reflects the intent of the original grant system." He looks at the case file. "The precedent assumed absence meant abandonment. This case has documented evidence of necessity. Those are materially different facts." He looks at me. "Help me write it."

---

It takes two hours.

Not because the legal reasoning is complicated — once Kael has identified the distinction, the reasoning follows clearly enough. Because this is careful work. The kind that needs to say exactly what it means and no more, that can't afford ambiguity in the language or gaps in the logic. We work through it the way we've worked through the northern boundary corrections: he writes, I cross-reference, he revises, I verify. The parallel rhythm we've built over a hundred and ten days applied to a different kind of document.

I find three supporting precedents from the pre-reform records that help — cases where absence due to necessity had been treated differently from absence due to abandonment, cases that Kael can cite as showing the original intent of the grant system. They're not definitive, but they're enough. They give his reasoning a foundation in the archive's own history rather than in his personal judgment alone.

At some point Aldren comes out of his alcove and reads over my shoulder.

He doesn't say anything. He looks at the precedents I've found, looks at the draft resolution Kael is working on, looks at me. He nods once — the specific nod that means: this is right, proceed — and goes back to his alcove.

The archive's memory approves, apparently.

We finish in the early afternoon.

Kael reads the completed resolution once more, then sets it down and looks at it with the quality of attention he gives finished things: checking not for errors but for completeness, for the sense that it says what it needs to say.

"The family's claim is protected," he says. "The precedent is distinguished on grounds that will hold. The new ruling establishes a clearer standard for future cases." He looks at me. "This is better than what was there before."

"What was there before was my mistake," I say.

He looks at me.

"You wrote a world," he says. "Not every detail of a world can be right. The mistakes are part of it — they're how the world becomes real rather than designed. Real things have flaws." A pause. "What matters is what you do when you find one."

I look at the completed resolution.

I wrote a bad precedent into a legal system on a Tuesday afternoon, not thinking about it. Today I sat in the archive it became part of and helped write the correction.

"Thank you," Kael says.

Not the administrative thank-you — not the acknowledgment of a task completed. The register of something genuine, given to someone who has done something that mattered.

I receive it without deflecting.

"Thank you for asking me to help," I say.

He picks up his pen to sign the resolution. Official signature, official seal, the specific formality of a document that will become part of the archive's record. The resolution will go into the files I've been working in, will be cross-referenced with the original precedent, will become the new foundation for any similar case that comes in the northern sector.

I wrote the mistake. I helped write the correction. Both of them are in the record now.

---

The afternoon is quieter.

I go back to the northern boundary notation work. Kael drafts the letter that will accompany the resolution to the northern settlement — the specific administrative prose of someone informing a family that their claim has been upheld. Aldren processes a stack of materials that have been accumulating in his alcove all week.

The archive settles into its working rhythm.

I'm finishing the last notation correction of the day when I remember the section in my coat pocket. The three-times-revised pages about Kael's arc, about the difference between a wrong belief and a truer one, about what satisfied looks like when it replaces relieved.

I've been carrying them all day.

The right moment, I think, is not a moment I'll recognize in advance. I said that to myself last week when I was waiting to ask Theo, and the rightness was in the beginning, not before it. The moment is something you make by starting.

I take the pages out of my pocket.

I don't walk them over to his table — that would make it a presentation, a thing being offered for response. Instead, when I get up to put a corrected file back on the shelf, I pass his table and set the pages down beside his current work. Not on top of it, not with ceremony. Just: here, if you want to read this.

He looks at the pages. Then at me.

I've already turned back toward the shelves.

"What is this?" he says.

"A section of *Notes from Inside,*" I say, without turning around. "The one about you."

I put the file back on the shelf. I go back to my table. I pick up my pencil and I look at the next record in the stack.

I don't watch him read.

He's quiet for a while — long enough that I've worked through two more records by the time I'm aware, in my peripheral vision, that he's set the pages down. Not pushed them away. Set them down with the deliberate care of someone returning something they want to keep.

He doesn't say anything immediately.

I don't prompt him.

The archive does its winter afternoon thing — the light shifting, the thermal stone steady underfoot, the smell of old paper and the specific quality of a room where two people are sitting with something that has been given and received.

Then he says: "The part about satisfaction replacing relief."

I look up.

He's looking at the pages. Then at me.

"That's right," he says. "That's what it feels like." A pause. "I didn't have language for it. You found it."

I look at him across the archive.

"You gave me the language last week," I say. "Knowing it versus knowing it. I just applied it to what I've been watching."

He looks at me for a moment longer.

Then he picks up his pen and goes back to the letter he's been drafting.

I go back to my notation corrections.

The section is his now — he's read it, he's confirmed it, he's given it back. It'll go into *Notes from Inside* when I'm ready, and when it does it will carry both of us in it: his arc, my observation, the specific quality of a hundred and ten days of being in the same room and watching a person become something truer.

I fold the pages and put them in my working folder.

Outside, the winter holds.

---

Same time tomorrow.

He says it at the door, coat on, the end-of-day ritual.

"Same time tomorrow," I say.

He goes.

I stay a few minutes longer than usual. Not for any particular reason — the correction log is done, the files are shelved, there's nothing left to do today. I stay because the archive has a quality in the late afternoon that I've been learning to sit with: the specific warmth of a place that holds its heat, the specific quality of light at the angle that means the day is nearly done, the accumulated presence of a room that has been lived in for sixty years and carries that living in its walls.

I've been in this room for a hundred and ten days.

In that time I've corrected the eastern sector's notation errors, established the foundation for the northern boundary corrections, helped resolve a thirty-year boundary discrepancy, found the land grants that protected three families' claims, and today helped write a new precedent that undoes the damage of a bad one I wrote on a Tuesday afternoon two years ago.

None of that was in the manuscript.

Chapter four of *The Dark Covenant* had the precedent in it as a background detail, a texture element, something to make the world feel like it had legal history. The story I was writing didn't have a scene where the protagonist helps correct it. The protagonist in that story — Kael, the dark knight lord moving toward his appointed end at chapter thirty — didn't sit in an archive for a hundred days with his author, working through the records she'd built for him to live in.

The story I was writing didn't have this.

The story I'm in does.

I look at the archive shelves. At the northern boundary records stacked on my table. At the empty chair at Kael's table, the pen capped, the letter to the northern settlement folded and ready to be sent.

*Notes from Inside* is going to need a section about this — about the specific experience of finding your own mistakes made real, about what it means to correct them from the inside. Not self-flagellation, not the performance of guilt. Something more specific: the particular quality of an author discovering that the flaws she built into a world for texture are being lived by actual people, and the specific responsibility that creates.

I wrote a bad precedent. Today I helped write a better one. Both of those things happened. Both of them are in the record.

Real things have flaws, Kael said. What matters is what you do when you find one.

I pick up my coat.

I look at the working notes on my table — the margin fragments from the week, the notation corrections, the scratch paper with Dara's words on it. I gather them into a folder to carry home. The folder has gotten thick over the weeks; I'm going to need to organize it properly, decide what goes into *Notes from Inside* and what can be let go.

There's a lot that goes in.

The being seen section. The being known section. The choice section, still rough, still needs work. The section on what it looks like when a wrong belief becomes a truer one. And now this: the section on finding your own mistakes in the world you built and writing the corrections from the inside.

I've been writing *Notes from Inside* for weeks and I still don't know how it ends. That's the nature of a document you're writing in the middle of living — you can't know the ending because you're not at the ending yet. When I started I thought I'd write about the archive work and the chain and the specific mechanics of how a written world functions when you're inside it. And I have written about those things. But the document has become something larger than that, something I didn't plan.

It's become a record of a person learning to be somewhere.

Not just the mechanics — the actual texture of it. What it feels like to arrive as an author and become, over a hundred days, something that doesn't have a clean name but is real. What it feels like to be seen by a place before you knew you were there, to be counted by someone from day eighteen, to watch a person you wrote become someone truer than what you wrote.

What it feels like to help correct the things you got wrong.

I put on my coat.

I go out into the winter dark.

Two streets to Dara's. The route I walk without thinking about it, the way you walk routes that have become part of you. The settlement is doing its winter evening things — lamps in windows, people moving with purpose toward warmth, the specific quiet of a place that pulls inward when the cold comes.

At Dara's I take off my coat and hang it by the door and go up to my room without stopping in the kitchen. I light the lamp. I sit at the table.

I open *Notes from Inside* to a blank page.

I write: *Today I found a mistake I made two years ago, sitting in a coffee shop on a Tuesday afternoon. It had become real. A family was affected by it. We wrote the correction.*

I look at that sentence.

Then I write: *Real things have flaws. What matters is what you do when you find one.*

I write: *I am in the world I made. I am responsible for what's in it. Not the way an author is responsible — that's distance, that's craft, that's the power to fix things by changing what's on the page. The way a person is responsible. From the inside. With the tools available here, which are not the tools of authorship but the tools of presence: time, attention, collaboration, the specific knowledge that comes from a hundred and ten days of being in a room with the records.*

I write for a while.

When I stop the lamp has burned lower and outside the settlement is quiet and the section isn't finished but it's started, which is what matters, which is always what matters.

I close the folder.

I look out the east-facing window at the dark rooftops and the winter sky beyond them.

Tomorrow I'll go back to the archive. I'll work on the northern boundary records. Kael will be at his table. Aldren will be in his alcove. Theo will come through with the intake. The archive will hold its records and its warmth and its sixty years of memory.

I'll be there.

I'm here.