199: The Final Year

I told Wanqing about the February 14 meeting on Thursday.

The next day. She was at the bench already when I arrived — the mid-February campus, the bare maple, the cold that had not yet shifted toward anything you could call pre-spring. She had the spring seminar preparation materials arranged on the bench beside her and the thermos between us as soon as I sat down. She'd already poured.

"She asked a different question than December," she said. Not a question. She'd been thinking about it since I'd mentioned in January that the February meeting was coming.

"The December question became the February question."

"How do you know what's worth building."

"Yes."

She turned a page of the seminar preparation notes. She was working through the first unit's structure — the problem sets were spread across her lap in the order she'd be using them. She didn't look up.

"And you answered it."

"Yes."

"What did you say."

I told her. The words I'd used — that you build what holds the people you're responsible for. That you know something is worth building when the people inside it become more themselves than they were before they were inside it. That the fund held Father, the guild held the people in the three AM sessions, the workshop held the remaining years instead of the years that had been in danger.

She was quiet for a moment after I finished.

The mid-February campus. A student crossed the path in the distance, moving fast in the cold. The bare maple held its specific February quality — not dormant-bare, waiting-bare. The first buds were five or six weeks away.

"That's accurate," she said.

"Yes."

She turned another page. "And she has something to tell you."

"When she's ready. Spring."

"She said spring."

"She said she'd contact me when she was ready. Not in-game."

Wanqing turned a third page, then stopped. She set the pen down on top of the problem set. She looked at the bare maple for a moment — the way she looked at things when she was deciding whether to say something, which was different from the way she looked at things when she was processing data.

"You should know," she said, "that whatever she says — I'll still be here. At this bench."

It was the most direct thing she'd said in two and a half years. The clearest version of something that had been in the space between us for longer than that — the bench, the thermos, the fourth December, the third February. Something she'd been present for without naming and had now named without elaborating.

I looked at the bare maple.

The specific pre-spring bare quality that was different from winter-bare — not the aftermath of a completed fall, not the full dormancy of January, but the still period before the work begins again. The tree at the beginning of its next cycle. The third year I'd watched it at this stage at this bench.

"I know," I said.

She picked up the pen and turned back to the problem set.

We sat for another hour. The cold afternoon. The campus gradually emptying as the afternoon shifted toward evening. The thermos went through two pours and I finished the second cup and she finished hers and neither of us said anything else that needed to be said.

Nothing else did need saying.

***

The spring seminar started March 5. Wanqing's pattern resumed — Tuesdays and Thursdays, the bench, the seminar materials on one side and the thermos on the other. The pattern that had been running since February 2016 with the seasonal breaks. The early-bud maple arrived in the second week of March. The fourth time I'd watched it come in at this bench, at this tree, in this same sequence.

The thesis outline was submitted February 28. Professor Liang reviewed it and returned three comments: two about citation format — the standard adjustment for a thesis that drew on game-domain literature alongside academic sources — and one substantive note about the formation optimization model's assumptions in the third case study section. He'd asked for a clearer articulation of what the model was assuming about coordination latency under Phase transition stress.

I revised the section and resubmitted March 1. The revision was three paragraphs — the kind of elaboration that clarified without changing the underlying argument.

He approved on March 3.

The thesis writing proper had begun.

The guild sessions continued at the standard schedule. Wenqing was accumulating Volume 2 — the post-championship sessions, the January and February clear times, the resonance integration deepening without a named ceiling. The January Floor 20 average of 3h 22m had moved to 3h 14m by March's first session.

Wenqing's note on March 7: *The integration is continuing. I don't have a model for when it plateaus. The transitional class had an accumulation ceiling — 100% was the documented limit. The full designation doesn't have a documented limit. I'm treating it as an ongoing process until the data indicates otherwise.*

*What does that mean for the projection,* I sent.

*It means I can't project a ceiling yet. The data would need to stop improving for me to estimate where it stops.* He paused. *It's still improving.*

The class was still developing. A year from the full designation, it was still deepening the formation. Wenqing was building a model for something that hadn't finished arriving yet.

The same problem Wanqing had had with the referral network: the standard model assumed decay, and the actual behavior was something that didn't decay in the expected way. Different domain, same type of problem — building a model for something that was performing better than the standard assumptions predicted.

***

Xiaoyu's message came on March 10.

She'd been in Osaka since October. The messages were regular — shorter in the week, longer on weekends, calibrated to the schedule she'd settled into. She'd adapted to the host family, the school system, the city's specific texture. She reported finding things: a good morning market three blocks from the school, a secondhand bookshop with a Mandarin section, a route from the school to the canal district that took twelve minutes on bicycle.

She'd been building knowledge of the city the same way she built everything — methodically, full attention, looking for what was actually there.

The March 10 message: *I found the street.*

She'd attached a photograph.

I opened it.

A canal-side street in morning light. The architecture older than most of the surrounding city but in good condition — the kind of maintenance that happened when a district had cultural or commercial significance worth preserving. The bridges had the ironwork specific to the prewar Japanese canal districts, the kind of metalwork that had a particular weight to it, a density that modern versions didn't replicate correctly. A woman with a bicycle crossed one of the bridges in the background, partially blurred by the motion. The canal caught the morning light in the flat way canals caught morning light — direct, without depth.

*Grandmother's letters described this street,* Xiaoyu wrote. *She said she walked it every morning from the boarding house to the textile district where she worked. I looked for the address she mentioned and found it — the boarding house is a café now. They've kept the original woodwork inside. The owner's grandmother used to work the front desk of the boarding house in the 1960s. We talked for an hour.*

She'd walked the city until she found the block that matched the description. Grandmother's handwriting from forty years ago had described the canal-side street and the ironwork bridges and the route to the textile district. Xiaoyu had arrived at the same street from the description and been standing in it.

*How does it feel,* I sent.

*Like two places are the same place,* she replied. *The one in the letters and the one I'm standing in. They're not the same exactly — the boarding house is a café and the textile district is apartments now and there's a bicycle lane where the street used to be wider. But the street is the street. The canal is the canal. The bridges are the bridges.*

*The letters were right.*

*Yes.* A pause. *Send my love to Father. Tell him the bridge is still there. He'll know which bridge.*

I called that evening.

Father answered on the third ring. The Suzhou flat — I could hear the canal sounds in the background, the specific ambient quality of the canal district at night.

I told him what Xiaoyu had said about the street and the canal and the bridge.

He was quiet for longer than usual. Not uncomfortable-quiet — the processing quiet of someone working through a weight of feeling that wasn't negative.

"Which bridge," he said finally.

"She said you'd know."

Another pause. "The iron arch one," he said. "At the end of the morning route." His voice had the quality it got when he was remembering something specific — not emotional in a loose sense, but focused. Arriving at the exact thing. "My mother used to say the arch looked like a song stopped partway through. She meant it as a comfort — things stopped partway through weren't finished, they were waiting. But she said it with a smile, so I knew she'd looked at it that way first before deciding it was a comfort."

He was quiet again. The canal sound.

"Tell her," he said. "Tell her I remember."

***

The March bench. Early-bud maple. The fourth year of this bench at this stage of this tree's cycle.

Wanqing had the spring seminar's second unit materials. She'd moved through the first unit faster than projected — the seminar groups this year were working at a higher pace than last year's cohorts, which meant she'd been adjusting the unit timing as she went.

I told her about Xiaoyu's message. The street, the photograph, the café that had been the boarding house, the ironwork bridge.

She listened without looking up from the problem set she was reviewing. Then she set the pen down.

She looked at the early-bud maple. The buds at this stage were the size of small fingers, the green just beginning to differentiate from the bark color. In two more weeks they'd be fully open. In six weeks they'd be the full young-leaf green of April.

"She found what she was looking for," Wanqing said.

"Yes."

"And it was both things at once."

"The street in the letters and the street she was standing in. Yes."

She looked at the maple for another moment. "That's how it is," she said, "with places that matter to someone who isn't there anymore. You build a picture of them from other sources — letters, descriptions, the things the people who were there carry with them and pass on. Over years the picture becomes very detailed. And then you arrive and the place is itself, which is different from the picture in specific ways. But the picture is also still true. Because the picture was real when it was made. Both versions are present at once."

She was describing Osaka and also describing something else. The bench under both descriptions at the same time. Something she'd been sitting with that she'd found a way to say at a distance.

"Yes," I said.

She picked up the pen. "The thesis."

"Writing. First chapter done, second in progress."

"Professor Liang's comments were substantive?"

"One was. The coordination latency assumption in the third case study. I revised and he accepted."

"The defense is June 22."

"June 22. The committee confirmed the slot."

She made a note in the margin of the problem set. Then she set the pen down again, which was unusual — she kept working through conversations and only stopped when the conversation required it.

"After the defense," she said. "What does Hangzhou look like."

I looked at the early-bud maple. The specific stage it was at. The fourth year.

"I don't know yet," I said.

"That's honest."

"Yes."

She looked at me. The April light was still three weeks away, but the March light was already different from February's — longer, a degree brighter, the quality of it slightly warmer even when the temperature was still cold.

"Tell me when you know," she said.

"I will."

She opened the seminar materials and went back to the problem set. I poured the last of the thermos into my cup. The early-bud maple. The fourth March at this bench.

Somewhere in Osaka, Xiaoyu was standing on the street from Grandmother's letters with both things true at the same time. The street in the writing and the street she was standing in. The arch that looked like a song stopped partway through.

Father had said: *it was stopped partway through, not finished. Waiting.*

That was also accurate.