Isolde came to the archive on a Wednesday and did not bring documentation.

This was unusual enough that Elianor registered it immediately — Isolde was always prepared, always had something specific to contribute, always arrived with the materials for the thing she had come to do. Today she arrived with her coat and her coffee and nothing else.

She sat at the reading room table. She looked at the space for a moment. Then she said: "I need to tell you something."

Elianor put down her pen.

Einar did not look up from his screen, but she felt the quality of his listening change.

Isolde said: "Veyrline's independent review team has contacted the observatory. They want access to the historical aurora dataset."

Elianor looked at her.

"My dataset," Isolde said. "The forty-year record."

"You submitted it to the panel."

"Yes. And they can request it from the panel through discovery. Which means they will have it regardless." She folded her hands on the table. "But they've contacted me directly. Asking me to cooperate voluntarily with their review process."

"Will you?"

Isolde looked at the table. "I've been thinking about that since Monday."

The archive was quiet. Outside the narrow window, the March day continued.

Isolde said: "The data is the data. The same dataset that supports the objection will not suddenly support Veyrline's case just because they're analyzing it. My concern is how they'll frame the analysis." She looked up. "I've seen corporate technical reviews before. The methodology is legitimate, the conclusions are shaped by the question you're asking. If they ask the wrong question—"

"What question would that be?" Einar asked. He was still looking at his screen.

"Whether the aurora interference, while present, was within the operational tolerance of the navigation system." Isolde's voice was precise. "It's the same data. Different framing."

Elianor thought about signal and noise. About the forty-seven pages of methodology documentation. About the difference between a question that was designed to find something and a question that was designed to confirm something was not there.

She said: "Would they be wrong? Technically?"

"No." Isolde looked at her. "That's the problem. They can construct a technically valid analysis using my data that produces a different conclusion. Not a false conclusion — a different emphasis. That's how corporate reviews work."

Einar said: "What does cooperating with them require?"

"Providing the raw data files and being available to explain the methodology." Isolde looked at the table. "I've already submitted the data to the panel. It's accessible through discovery anyway. Refusing to cooperate voluntarily doesn't protect anything — it just makes me look like I'm hiding something."

"Which you're not," Elianor said.

"No. But perception matters in these processes."

Elianor sat with this for a moment. She thought about the case, about the months of careful documentation, about the way each piece had been built to withstand exactly this kind of scrutiny. She thought about what Isolde was being asked to do and what it meant.

She said: "Cooperate."

Isolde looked at her.

"You have nothing to hide in the data," Elianor said. "You documented the methodology. If they ask the wrong question, we document why it's the wrong question in our supplementary filings." She looked at Einar. "We anticipated this."

He looked up from his screen. "Yes. The methodology notes in the original submission specifically address the framing question. That section addresses exactly the alternative analysis you're describing." He met Isolde's eyes. "We were thorough."

Isolde was quiet for a moment. Then: "All right." She unfolded her hands. "I'll cooperate. And I'll document every question they ask and every answer I give."

"Good," Elianor said. "Send me everything in real time."

"I will." Isolde stood. She picked up her coffee. She looked at the archive — at the monitoring systems, at the workstations, at the way the three of them occupied the space. Something in her expression was different from the composure she usually presented. Not vulnerable exactly. But present.

She said: "You have a good case."

"Yes," Elianor said. "We do."

"They know that. That's why they're trying to complicate the process." She looked at the table. "They know the data doesn't support them. They're hoping the process will."

Einar said: "It won't."

Isolde looked at him. A long moment.

Then she picked up her coat and left.

The archive settled back into its quiet. Elianor sat at her workstation. Einar went back to his screen.

She thought about Veyrline and their simultaneous review and their corporate technical team and their forty-page procedural challenges. She thought about six months of Einar at the side table and seven years of work and a forty-seven page technical assessment and thirty years of atmospheric data and Isolde's methodological documentation and twelve families and the panel of seven experts who had already read all of it.

She thought: they don't have what we have. They have resources. We have the record.

She went back to work.

---

The week continued. The panel filed a response to Veyrline's review announcement, confirming that the panel's timeline would not be affected by any parallel processes. Veyrline's legal team filed a challenge to this response. The panel acknowledged the challenge and confirmed again that the timeline would not be affected. This was, Elianor understood, the procedural version of very expensive, very boring attrition — Veyrline generating paperwork in the hope that at some point the sheer volume of it would slow something down.

She and Einar responded to each filing on the day it arrived. They had gotten fast at this — the collaborative rhythm of the past months had refined itself into something that was very efficient. He drafted the technical sections, she drafted the procedural sections, they reviewed each other's work, filed. One hour per filing, often less.

On Thursday evening, after a particularly dense day of challenge-and-response, she said: "I think they're going to run out of challenges before we run out of answers."

He was packing up his laptop. "Yes. The methodology documentation is comprehensive enough that there's a finite number of ways to challenge it." A pause. "We anticipated well."

She thought about the months of preparation. About the care with which the documentation had been constructed, layer by layer, each section designed to hold up under exactly this kind of pressure.

She said: "You did. The methodology notes specifically."

He looked at her. "We both did."

She looked at the workstation. She thought about seven years of work that had been hers alone and the months of work that had been theirs together, and the difference between those two things.

She said: "I'm not used to that."

"I know," he said. He said it simply, without elaboration. She thought: he does know. He spent seven years working alone on data he couldn't open, maintaining something he had no one to give it to. He knows exactly what it is to work without a partner.

She said: "I'm getting used to it."

He picked up his bag. He said: "Good."

He left. She sat in the archive for a while longer, processing the day's filings, running a routine monitoring pass on the evening's captures. The harbor bell rang. The drives stored.

She thought: *we both did.*

She thought: that is the right framing. That is the record, accurately stated.

She liked it.

She went home.

---

The following Saturday she played him a second file from the tower.

She had been thinking about it all week — which file, when, how much. She had decided: the file called *Einar-thinking.* The sound of a fifteen-year-old recording his friend at a keyboard, in a stone tower on a ridge in Aurevik, on a winter afternoon.

She went to the tower on her own that morning. He had a phone call with the institute — something administrative about the consulting position's formal documentation — and she had said she would go ahead and start the equipment checks. He would come up when he was done.

She had a specific plan and she was following it.

She did the equipment checks. She sat at the secondary workstation. She navigated to the folder she knew the location of. She found the file. She read the label again: *Einar-thinking (keyboard + ambient) — 4m22s — NV.*

She sat with it for a moment.

She pressed play.

The recording began with the room sound — the ventilation pipe, the equipment hum, the specific quiet of a stone building. And then: the keyboard. Not dramatic, not loud. Just a consistent, measured sound: someone working at a computer, key by key, with the particular rhythm of someone whose thinking moved through their hands.

It was mundane. It was entirely mundane. And it was also the most intimate thing she had heard in one of Niven's recordings, because she knew who was at that keyboard and she knew he was not aware of being recorded and the recording had been made because a fifteen-year-old boy had decided this was worth keeping.

She listened to four minutes and twenty-two seconds of a man she loved working at a keyboard in a room he had not been back to in seven years, preserved by a recording made by her dead brother, who had understood that sounds were worth keeping.

She stopped the file when she heard footsteps on the stairs.

Einar came in. He looked at the workstation. He looked at her.

He said: "You played it."

"Yes."

He was quiet for a moment. Then he came to stand beside the secondary workstation. "And?"

She looked at the screen. "It sounds like exactly what it is." She looked at him. "He was right. You think in your hands. I can hear it."

He looked at the screen. He did not say anything for a long time.

She said: "He made a good recording."

"Yes," Einar said. His voice was even. "He was getting better."

She looked at the screen. She thought about all the files she had not yet played — the eight others with his voice, the fourteen ambient recordings, the storm recordings. They were here, waiting. She would get to them.

She said: "I'd like to start hearing them. Gradually. One at a time."

He looked at her. "Yes."

"Not all at once. Just when it feels right."

"Whatever pace you need."

She looked at the monitoring room — the equipment running, the printouts on the walls, the mug on the shelf. The room that had been Niven's second favorite place in Aurevik.

She thought: I'm learning it. I'm learning it through the recordings, one at a time. That's also how you learn a place — not all at once, but gradually, until it becomes part of you.

She stood up from the secondary workstation.

She said: "Equipment checks are done. Let's go."

They went down the stairs and out into the March morning. The ridge was warming — the last of the snow was mostly gone from the path, the mud firming up in the strengthening spring light.

She looked at the city below as she always did at the bottom of the path.

She thought: harbor at the center. Everything else arranged around it.

She thought: yes. That's the pattern.

She walked toward the archive.

---

The days accumulated. The panel filings continued. Veyrline's review progressed without producing anything that threatened the objection. Isolde documented every question and sent the transcripts; Einar reviewed them and confirmed that the framing remained manageable. The twelve families received weekly updates. Rowan's ice safety team continued their seasonal assessments.

And in the archive, the ordinary work continued alongside all of it: the harbor monitoring captures, the restoration queue, the ongoing accumulation of the atmospheric record. She processed two restoration files that had nothing to do with Frostline, filed three equipment maintenance reports, updated the monitoring protocols for the spring season.

She noticed that she was doing the ordinary work with a different quality of attention than she had brought to it in years. Not more effort — the same careful precision she had always applied. Something more like presence. The work landing differently when it was not being done in the shadow of the other work.

She thought: this is what it means to have done the thing. The ordinary work is just the ordinary work again. It is not secondary to something more important. It is just the work, valuable in itself, worth doing for its own reasons.

She thought: Niven would have understood this. The work was always the point for him too — not what the recording was for, but the act of listening carefully enough to make it.

She archived a harbor radio capture from Thursday's monitoring. She logged it. She moved to the next file.

Outside the narrow window, April was beginning to arrive — the light changing again, the days growing perceptibly longer, the harbor active with the early fishing season. Aurevik in spring had a different quality from Aurevik in winter: less compressed, less internal, the city beginning to turn outward toward the water and the mountains.

She had always loved this transition. She had forgotten, for a few years, that she loved it.

She remembered now.

At noon she took her lunch to the bench by the harbor — the one where she and Niven had eaten ice cream in summer, the one she had passed every day for seven years without stopping. She sat on it and ate her sandwich and watched the harbor.

At the archive three blocks away, Einar was at the side table working on the aurora data. At the tower on the ridge, the equipment was running. At her mother's house up the hill, Maera was probably gardening — she always started gardening too early in spring and always complained that she had started too early. The city was doing its midday things.

She sat on the bench and ate her lunch and felt the spring sun on her face and thought: I am here.

I have always been here.

I am so glad to be here.

She finished her sandwich. She went back to the archive.

She passed the bench she had just been sitting on every day for the rest of that spring, and sometimes she stopped, and sometimes she didn't, and both were fine.

Both were fine.

---

The second time Einar came to dinner, he brought bread.

Not wine this time — bread from the bakery on the corner of the harbor road, the one she had pointed out to him when they walked there the week after the tower visit. Dark rye, the kind her mother made with root vegetables, the kind Niven had liked.

She noticed he had remembered.

She said: "That's the right bread."

"I thought so." He set it on the counter. He did not say more about it.

She had made soup — the lamb stew from her mother's recipe, the one she made in winter and early spring, the one that took all afternoon and was worth it. The apartment smelled good. The kitchen table was set. The lantern was burning low in the other room.

He sat at the table. She sat across from him.

They ate and talked about the week's filings and the panel's timeline and the observatory's response to Veyrline's review request, and then they talked about other things: the monitoring data from the latest moderate aurora event, the acoustic properties of the harbor in spring versus winter, what Niven had recorded on his school trip to the mountains three years before the accident.

She had the field notebook at the table — she had been reading it again, slower this time, taking notes on the observation methodology. She told him about the mountain trip entry. He had not known about that trip. Niven had mentioned going, but not in detail.

She read him the relevant section. He listened.

He said: "He was trying to apply the same framework to a different environment. To see if the methodology transferred."

"Yes. He concluded that mountains had better fundamental frequencies but worse harmonic resolution than harbors." She looked at the notebook. "He was planning a comparison study."

"He would have done it."

"He would have." She looked at the table. "He had a whole system. He was going to do the cathedral next. He thought the acoustic properties would be the most interesting of anything he'd measured."

"He was probably right."

"He was probably right about most things." She looked at the notebook. "I've been reading it slowly because I don't want to run out of it."

He looked at her. "There are more recordings."

"I know." She looked at the window. "I'll get to them. But the notebook is him at fourteen, and the recordings are him at fifteen, and I want to spend enough time with each before I go to the next."

He said: "That's the right approach."

She looked at him. "You've been waiting. Seven years."

"Yes."

"For me to be ready to receive them."

"Not waiting for you specifically." He looked at the table. "Keeping them until the right time. I didn't know the time would involve you. But it makes sense that it does."

She looked at the bread he had brought. The dark rye, the right kind.

She thought: he remembered which bread was the right bread because I told him once on a walk. He pays attention to details the way she pays attention to details. He builds the record carefully, piece by piece.

She thought: that is also why Niven trusted him.

She said: "He made a list of sounds he wanted to record in the tower. The list you found on the shelf."

"Yes."

"One of them was labeled *the sound of Einar typing* — his exact words, in his filing system." She looked at him. "He was already planning to record you specifically. Before he ever mentioned you to me."

Einar was very still for a moment. "I didn't know that. I knew about the recording. I didn't know it was on the list."

"It was on the list," she said. "First draft plans. Then he did it."

He looked at the table. He was quiet for a long time.

She said: "He thought highly of you. Not just as someone who knew things about the aurora. As a person."

"I know," Einar said. His voice was even, the specific evenness of something that was being held carefully. "I thought highly of him too."

She looked at the notebook. She looked at him.

She said: "I know."

They finished dinner. They cleared the table. He put on his coat in the front hall.

At the door he said: "The bread was the right one."

She said: "Yes. It was."

He left.

She went back to the kitchen. She looked at the table. The two plates, the empty soup bowls, the bread with a piece broken off. The notebook at the side where she had placed it.

She thought: this is what it looks like now. Dinner and bread and a field notebook at the table, and the person at the window side is learning to be someone who comes back.

She thought: I think he is going to be someone who comes back.

She thought: yes. I think that's right.

She washed the dishes. She put the notebook carefully back in its archival folder. She went to check the lantern.

It was burning steady and low.

She looked at it for a moment.

She said, very quietly: "Niven. The bread was the right bread."

She went to bed.

She slept well.