The snow is different this morning.

I know it before I open my eyes — something in the quality of the light coming through the gap in the curtain, the specific flat-white quality that means the sky and the rooftops have become the same surface. When I sit up and look out the east-facing window I can see that the snow has finally caught on the rooftiles, the low slope of the building across the way covered in an unbroken sheet of white that must be three fingers deep. The gutter at the far corner has a small drift forming where the wind turned during the night.

This is the first real snow. The earlier falls didn't hold — they came and melted and came again, leaving the settlement gray and damp but not winter-white. This one held. Overnight it built itself up silently and by morning it had settled in with the particular permanence of something that intends to stay.

I sit at the table with my tea and the page from last night.

The section on being seen. I wrote a page and a half yesterday — longer than I expected, shorter than the subject deserves. I've been reading it back in the way I read draft pages: looking for the thing it's actually saying underneath the thing I thought I was saying when I wrote it.

The thing it's actually saying is that I've been writing *Notes from Inside* from the wrong position.

I've been writing it as the author. The person who came from outside and entered and observes. The perspective that carries a certain kind of authority — I was there before any of you were, I made the world you're living in, I know things you don't know. Even in the sections where I write about what I don't understand, about being confused or uncertain or changed by something I didn't expect, I've been writing from outside the story looking in.

What Theo told me yesterday, and what Kael confirmed, is that I'm not outside the story looking in. I'm inside it being looked at. There is a version of me in Theo's picture of the archive — the person who arrived asking where things were and became the person who knows. There is a version of me that Kael has been carrying since day eighteen — the person who cited a number with the precision of someone counting something that matters.

I didn't write those versions of myself. They accumulated in other people's observations while I was busy being the observer.

I write one line in the margin: *The author is also a character. This is not the same thing as knowing that.*

Then I fold the page, put it in the working folder, and get ready for the archive.

---

The settlement is quieter in the snow.

This is a thing about the world I built without knowing I built it: the acoustic quality of a snow-covered place. The way sound moves differently when surfaces are soft — the footsteps muffled, the vendor calls shorter, the specific dense silence of a place that has had its edges rounded off. I walk through it now and I hear it and I didn't write it. It arrived with the snow, the way things arrive when you build a world carefully enough that it generates its own details.

The archive corridor is warmer than the street. The thermal stone system Aldren's predecessors installed runs under the floor — I know this because I wrote it into the early chapters of *The Dark Covenant* as a piece of world-building, a detail that was supposed to convey that the archive was old and cared-for and had been adapted thoughtfully over generations. Now I walk on it every morning and feel it through my boots and it's just warm stone doing what warm stone does.

Kael is at his table when I arrive.

He's not at his correspondence, which is where he usually is in the mornings. He's at the eastern reference section — the large-format records shelf, the one that holds the boundary surveys and the cadastral documents, the oversized materials that require the bigger table in the corner for proper examination. He has one of the largest-format documents open on that table and he's standing over it with the look I've learned to read as: *this requires assessment before I decide how to respond.*

I put down my coat and bag and cross to where he is.

---

The document is a boundary survey.

Not a recent one — I can tell from the paper quality and the ink fading and the particular style of the cartographic notation, which changed in the settlement about fifteen years ago when the third surveyor brought in a new mapping convention from the eastern provinces. This is older than that. The notation style is pre-reform. The paper has the specific brittleness of something that's been in a cold archive for a long time, handled occasionally, never properly restored.

I look at the survey and I look at what Kael has beside it — a current boundary record, the post-reform version, showing the same section of the northern boundary. The comparison is immediate and stark.

The survey line has moved.

Not by much — a few hundred meters at most, the kind of shift that looks like nothing on a regional map but is enormous when you're standing on the land in question. The pre-reform survey shows the boundary running along the ridge line of the low hills in the northern sector. The current record shows it two hundred meters south of the ridge. Someone, at some point in the fifteen-year gap between those two documents, moved the boundary south and the correction never made it back to the original survey.

Three family holdings fall in that gap. I can see them marked on the current record — small notation marks for inhabited settlements, the kind of marks that mean: people live here, this land is claimed and worked and relied upon.

"How long have you known about this?" I ask.

"I found it this morning," he says. "I came in early. The northern boundary work brought up a cross-reference I needed from this shelf and when I pulled the file I saw the comparison." He looks at the two documents. "It will need to be resolved before we finish the northern boundary corrections. If we document the current boundary as accurate without noting the historical discrepancy, we're building the correction on a foundation that's wrong."

I look at the gap between the two survey lines.

Three families. Thirty years of working land they may not have legal claim to under the original survey, or thirty years of having their claim undermined by a boundary record that shifted without their knowledge. Both versions are possible. Which one is true depends on documents I don't have in front of me — the land grant records, the original settlement charters, the specific legal weight of the pre-reform survey versus the post-reform correction.

"Where are the land grant records?" I ask.

"Aldren's section," he says. "The lower archive. The materials that predate my tenure here — anything from more than twenty years ago is in his custody."

I look toward Aldren's alcove. The curtain is open, which means he's expected but not yet in. His morning, on the days he comes late, usually runs to mid-morning — some combination of age and routine and the particular rhythm of someone who has been keeping their own hours in this archive for twenty-three years.

I sit down at the big table beside the open survey.

Not because I've been asked. Not because I've calculated that sitting down is the correct strategic response to the situation. I sit down because there's a problem in front of us and I know the notation system well enough to be useful and the northern boundary records are my project as much as they are Kael's now.

Kael looks at me. He moves the current boundary record so it's positioned between us — equidistant, both visible. So we can both read from it.

We get to work.

---

The problem with reconstructing a thirty-year-old boundary discrepancy is that the evidence you need is dispersed across multiple document types that weren't designed to be read together. Survey records are in the map archive. Land grant records are in the legal archive. Correspondence related to boundary disputes is in the administrative records. Notations about post-survey corrections are in the surveyor's log, which is a separate category entirely.

All of those things are in Aldren's section of the lower archive.

Kael knows the upper archive — everything processed in the past fifteen years — the way I know the eastern sector notation errors: fluently, specifically, without having to think about where to look. The lower archive is Aldren's. The organization down there follows Aldren's system, which predates the reform and has its own internal logic that neither of us knows well enough to navigate quickly.

We work with what we have. The survey. The current boundary record. The northern boundary reports that have come in during my time here — the ones that have been accumulating in the intake pile, the ones I've been working through in the correction project. I pull every northern boundary document in the current files and we lay them out on the big table in chronological order and look for the moment when the line moved.

It takes most of the morning.

The moment is in a post-survey notation from eighteen years ago. A small addendum, written in the surveyor's log style, noting that the boundary marker at the northern ridge had been found displaced — moved by a flood event two decades earlier — and had been re-established at the new location. The re-establishment was legitimate at the time. The problem is that the re-establishment location was recorded in the surveyor's log but never updated in the land grant records. The survey shows the new marker. The land grants still reference the old boundary.

"It's not malicious," I say. "The surveyor found a moved marker and re-established it correctly. The administrative connection between the survey record and the land grants just wasn't made."

"Administrative failure," Kael says. "Yes. But the effect on the families is the same whether it was malicious or not. They've been living with the wrong boundary record for eighteen years."

I look at the families' holdings on the current survey. Three farms. The southernmost one has the most exposure — the land in the gap is their primary pasture.

"We need Aldren's land grant records to know what the resolution is," I say. "If the original grants reference the ridge line, the re-established boundary marker is the error, not the grant. If they reference the marker's position, the current record might be defensible."

"Yes," he says. "And we need them before we can finish documenting this section of the northern boundary."

We look at Aldren's alcove. Still empty. The curtain still open.

---

I go down to the lower archive to look.

Not to retrieve the records — I don't know the system well enough to retrieve specific documents without guidance. But I know enough about archival organization to make a preliminary assessment: what section the land grants would be in, what the access looks like, whether there's a finding aid or index that could help us locate them without Aldren.

The lower archive is cold in the way that the upper archive is not. The thermal stone doesn't reach it — or rather, it does, but the lower archive is twice the volume of the upper and the heating is uneven, the far corners staying cold in a way the near ones don't. It smells different too: older paper, older ink, the particular smell of materials that have been sitting undisturbed for long enough that they've stopped off-gassing and have settled into their own chemistry.

I find the land grant section without too much difficulty. Aldren's system is older than the reform but it has its own logic, the logic of someone who has been living in a filing system for twenty-three years and has made it an extension of his own memory. The land grants are organized by district, then by date of original grant, then by family name. The northern boundary district is in the third cabinet from the left.

I'm looking at the cabinet when I see the notebook.

It's on the work table in the land grant section — the small table where Aldren sits when he's working in this part of the archive, retrieving and cross-referencing. Not the large notebook I've seen him with upstairs, the one that goes everywhere with him, the record of twenty-three years of archive observations. This is a smaller one. A working notebook — the kind you use for active projects, notes toward something, the in-progress thinking of someone in the middle of a question.

It's open to a page that has two columns. On the left, dates. On the right, brief notations.

I can read it from where I'm standing without moving closer, because the light in this section falls at the right angle and because I've spent fifty-five days doing close notation work and my eyes have been trained by it. I'm not trying to read it. I read it anyway.

Most of the notations are in Aldren's shorthand, which I don't know. But the dates I can read — they go back past the beginning of my time in this archive, past the day I arrived in the field, into a period before the story's present. And in the right column, on a line dated approximately two weeks before my arrival, in standard script rather than shorthand, as though it was important enough to write clearly:

*Reverie Marlowe.*

My name.

I look at it for a moment.

Not my name as the archive knows me — the unnamed element, the narrative blank that Kael can't read, the thing The Ink has declined to act against on six observed occasions. My name as I know it. As my passport knows it. As my mother knows it, calling me from another world entirely.

Aldren wrote it two weeks before I arrived.

I put the notebook back exactly where it was. The angle, the opening, the position on the table — I put it back as precisely as I can, the way you put something back when you've read it without being invited to and you're choosing not to make that your problem yet.

I find the land grant records I was looking for. I bring them upstairs.

---

Kael looks at the records when I set them on the table. Then he looks at me.

Not the tidings-sense — he's said before that I produce no echo, that I'm a blank in the narrative weight he reads. This is the other kind of looking. The kind that doesn't need an echo because it has fifty-five days of accumulated observation to work from.

"Those are the right records," he says, about the land grants.

"Yes," I say.

We get back to work.

The land grants reference the ridge line. The original grants, predating the flood, predating the marker displacement, predating everything — they describe the boundary as following the ridge. Which means the post-flood re-establishment of the marker two hundred meters south of the ridge was the error. The families' claims are sound. The current survey record needs to be corrected back to the ridge, not the other way around.

It takes another hour to document this properly — the specific chain of records, the error in the surveyor's notation, the correction required. Kael writes the formal assessment while I cross-reference the supporting documents. We work in the specific collaborative rhythm we've been building since the morning: he goes to the next problem while I'm finishing the documentation of the last one, so the work moves faster than either of us alone without either of us rushing.

Around mid-afternoon, we finish.

The three families' claims are documented as sound. The boundary will be corrected to the ridge line in the official records. The eighteen-year error has a resolution.

I sit back and look at the stack of records on the table.

Kael is looking at me again.

"What did you find," he says, "in the lower archive."

Not a question — he's framed it as a statement, which means he already knows something is there, he's giving me the choice of how much to say.

I think about Aldren's notebook. The dates. My name written two weeks before my arrival in standard script, as though Aldren had been certain enough to write it clearly, clearly enough to write rather than code.

"Something I need to ask Aldren about," I say. "Not urgent. Not threatening. Just — a question I didn't have before this morning."

Kael looks at me for a moment. Then he nods once and turns back to the boundary assessment he's been writing.

That's all.

He doesn't ask what it is. He doesn't offer a theory. He doesn't fill the space with reassurance or curiosity. He simply registers that I have a question I'm not ready to share and returns to his work, in the specific way of someone who has enough respect for another person's process to not insert himself into it before he's been invited.

I've known people who would ask. Who would need to know, who would read the withholding as something done to them rather than something I'm doing for myself. Kael is not one of those people.

I pick up my pencil and go back to the cross-referencing.

We work until the light changes.

---

I walk home through the settlement in the early dark of winter.

The snow that held overnight is still there — the late-afternoon foot traffic has cleared the main paths but left the rooftops and the window ledges and the corners white. The settlement in snow looks slightly different from the settlement I've been learning for a hundred days. Softer at the edges. More contained. People moving more directly from place to place, less of the mid-path stopping and talking that happens in better weather.

I go past Mira Voss's corner without stopping. The stall is closed at this hour — she packs up by mid-afternoon in winter, the business having shifted to the covered market for the cold months. I note the corner anyway. I note it the way I note the archive corridor and the gate arch and the east wall of Dara's: as a coordinate, a fixed point that means I'm in the right place.

At Dara's I go up to my room without stopping at the kitchen. I light the lamp. I sit at the table by the window.

*Notes from Inside* is in my working folder. I open it to where I was last night — the page and a half about being seen. I read it again.

Then I turn to a fresh page.

The section I'm about to start is not the one I planned. I planned to continue the being-seen section, to build out the observation about Theo and Kael and the particular experience of finding out you've been witnessed in ways you didn't know. That section is real and worth writing and I'll come back to it.

But right now there's a different subject pressing.

I write the heading: *Being Known.*

And then I sit with it for a while, because the subject is harder than it looks.

Being seen is something that happens in time — you are present, someone observes you, an impression forms. It's accumulative. Theo accumulated observations. Kael accumulated evidence. They built pictures of me from what they watched, and those pictures are true in the specific way that pictures built from observation are true: they capture what was there to be seen.

Being known is something else.

Aldren wrote my name two weeks before I arrived. Not from observation — he couldn't have observed me, I wasn't there yet. From something else. From information, or from a kind of knowing that doesn't come from watching but from — what? The archive's accumulated knowledge of what the story needed? The twenty-three-year notebook and whatever it contains about the mechanics of this world? Something Aldren knows about authors and stories and the specific event that happens when a writer falls asleep at their desk and wakes up somewhere else?

He knew I was coming. He knew my name.

That's different from Theo watching me learn to cross-reference. That's different from Kael counting the days. That kind of knowing is prior to observation. It doesn't build from evidence — it starts with a conclusion and waits for the evidence to arrive.

*There is a version of me that existed in Aldren's notebook before I arrived. Before I made any choices. Before I learned anything. That version had my name and a date and whatever else is in the shorthand I couldn't read.*

I write that down.

*What does it mean to be known by someone who knew you before you were there to be known?*

I write that too.

I sit with both sentences and the lamp and the snow outside and the question I'm not asking Aldren tonight, not because I'm afraid of the answer but because some questions need to be carried for a while before they're ready to be asked. The carrying is part of the work. You don't always ask immediately. Sometimes you live with the question long enough to know which version of it is the real one.

From somewhere — from the quality of the world doing what it does, holding its own commentary at the edges of awareness:

*— she found it —*

*— Aldren knew —*

*— of course he did —*

*— wait, does she know what's in the notebook —*

*— she doesn't know yet —*

I write one more line before I close the folder.

*I was known here before I knew myself.*

I look at that sentence for a while.

Then I put the folder away, and the lamp stays lit for a while longer, and outside the snow holds its position on the rooftops of the settlement, doing what winter does: staying.