I have been counting the board.
Eleven pieces. I know this because I count them every morning, the way you count stairs — not because you have forgotten how many there are, but because your feet need the ritual. Eleven. The number has not changed in six days. What changed is me.
This morning I counted them. This afternoon I read them.
The bogdo-jineung counts them too. It counts everything. The environmental sensor strip along corridor B's ceiling registers each piece on the board as a jeongbo-pyomyeon event — an information-surface interaction. Eleven items. Chamyeo-jisu: 0.34, which the system considers moderate. The strip glows soft amber when the index is between 0.2 and 0.5, blue below, white above. I have watched it be amber for eleven straight days.
The system's classification is simple: visual content (3 items), textual content (5 items), mixed (3 items). It assigns a baechi-jeomsu — arrangement score — of 0.61, meaning the spatial distribution is above average for communal surfaces in buildings of this tier. The bogdo-jineung files this data every six hours. It does not read the board. It processes the board. The difference is the entire subject of my corridor study.
Forty-seven days of data. Neat columns in my notebook: date, count, side, temperature, ambient sensor color. The bogdo-jineung has forty-seven days of data too. Ours overlap nowhere.
Let me tell you what the board looks like right now, because the arrangement matters in ways the baechi-jeomsu cannot capture.
Top left: a child's crayon drawing of a house. Not this building. A generic house — triangle roof, square body, yellow sun. The crayon is waxy and thick. It has been here nineteen days. Next to it, slightly overlapping, a second drawing in the same hand — same child, different day. The second one has a dog. The same yellow sun. The bogdo-jineung classifies both as visual content, juvenile origin, sentiment: positive. It scores them identically. They are not identical. The second drawing was made by a child who came back. Return is not repetition.
Center: Mitsuki's frequency log. I know it is Mitsuki because the handwriting is small, precise, unhurried, and because nobody else in Gyeol-ri would write '17.7 Hz. Consistent. Fourth week.' in the margin of a piece of graph paper pinned to a communal board. Twenty-eight entries. One per day. Each one records a frequency, a time, and a single word of assessment. Consistent. Consistent. Steady. Consistent. Shifted. Consistent. She is listening to the building's structural resonance with her body and reporting what she hears. The bogdo-jineung monitors this same frequency — it appears in the building's maintenance logs as ambient structural vibration, within normal parameters. The system measures what Mitsuki measures. It does not know she is measuring alongside it. Two instruments pointed at the same signal, unaware of each other.
Bottom right: a photograph. Polaroid format, slightly overexposed. It shows this corridor, taken from the B-side entrance, but in a different season — the light is summer light, the wall coating is the previous formulation (they applied the new thermal-responsive layer in October, the one that shifts from cream to pale grey as temperature drops). Someone pinned a memory of this hallway to this hallway. The board contains its own past. The bogdo-jineung classifies it as visual content, archival, sentiment: neutral. It cannot read nostalgia. It does not know the wall used to be a different color in summer.
Scattered: a pressed leaf (ginkgo, from the courtyard — the one the maintenance algorithm wanted to remove last autumn for root-interference with the subsurface sensors; the residents petitioned; the tree stayed), a recipe handwritten in an older style of Korean that I had to ask Mrs. Park to translate (she said it was her grandmother's tteokguk recipe, and that she put it on the board because she could not read the handwriting anymore and hoped someone else could), a rooftop garden schedule from two seasons ago that nobody follows now because the garden's irrigation shifted to bogdo-jineung automation.
I counted all of this. Eleven pieces. Categories: visual, textual, mixed. Frequency: added mostly between 7 and 9 AM, during the gap between the system's pyeonghwa-sigan ending and the morning chamyeo-jisu peak. Removal rate: zero in the last three weeks.
That was my analysis. It was correct and it was useless.
What I saw this afternoon, after Chae's correction, was not eleven pieces in categories. It was four modes of knowing, sharing one surface.
Expression. The child's drawings. The ginkgo leaf. The recipe in script nobody living can fully read. These are not information — they are offerings. Someone put them here because the board exists and because putting something on a board is a way of saying I was here and I had this and I wanted to give it to the hallway. The drawings do not communicate a message. They perform a presence. The bogdo-jineung scores them as content with positive sentiment. It cannot distinguish between a drawing that says something and a drawing that does something by existing.
Measurement. Mitsuki's frequency log. Twenty-eight days of 17.7 Hz, recorded at the same time each morning, with one deviation noted and no explanation. This is data collection. The board is her lab notebook, published in real-time to an audience of anyone who walks past and knows what hertz means. She and the bogdo-jineung are parallel instruments. The difference: the system logs the frequency and files it. Mitsuki logs the frequency and pins it to a wall where people can see it. Measurement made public is a different act than measurement made private. The board transforms data into testimony.
Memory. The photograph. The old rooftop schedule. The recipe in the older script. These are the board's archival layer — things pinned not because they are current but because they should not be forgotten. The photograph of the corridor pinned in the corridor is the most recursive: memory of this place, in this place, on this surface. The bogdo-jineung has no archival instinct. Its memory is operational: it stores data because data improves future classification. The board stores memory because forgetting has a cost the system does not measure.
Three modes. I had gotten this far on my own, in the notebook, before dinner. I was satisfied. Tidy. Expression, measurement, memory. A publishable framework. A contribution the corridor study could offer to the growing literature on analog surfaces in algorithmically-managed residential complexes.
Then Chae said: 'You missed a category.'
We were at the shared table. Standard-side kitchen, which means the bogdo-jineung dims the lights at 21:00 instead of 22:00 (premium side gets the extra hour, one of the small differentials the lending pool is supposed to eventually equalize). The dishes were Chae's contribution — she cooks on Wednesdays. Doenjang-jjigae, the kind she makes without measuring, the kind that tastes slightly different each time because her hands know something her memory does not track.
I told her about the three modes. I was proud of them. They felt true.
She listened the way she listens — completely, without performing attention, as if hearing were a bodily function rather than a social one. Day 4,754 of living in this building, and her listening has the quality of someone who has stopped needing to respond and started needing to understand.
'You missed a category,' she said. Not as correction. As observation.
'Mitsuki's log is measurement,' I said. 'She is collecting data. Frequency readings. It is empirical.'
'Twenty-eight consecutive days,' Chae said. She ladled the jjigae. The steam rose between us. 'Same act. Same time. Same place. Twenty-eight days without a break.'
'That is dedication to the data.'
'That is suhaeng.'
Practice. The word she used — suhaeng — carries weight in the Lend District's vocabulary that it does not carry outside. Here, where the lending pool measures every capacity exchange, where the bogdo-jineung tracks every skill acquisition curve, suhaeng means the kind of repeated action that the system cannot score because it is not moving toward a measurable outcome. It is not learning. It is not improving. It is being faithful to the act itself.
I opened my mouth to counter and then closed it.
She was right. Not instead of what I had said — underneath it. The frequency log IS measurement. It IS data. But twenty-eight unbroken days of showing up at the same time to do the same recording is something else running beneath the measurement. The logs do not get more accurate over twenty-eight days. The frequency does not become clearer. The act is not productive in any direction the chamyeo-jisu can track.
Mitsuki is not just recording the building's resonance. She is being faithful to it.
The fourth mode is devotion.
And I — the person who has counted this board every morning for forty-seven days, who records the ambient sensor color and the temperature differential and the engagement index — should have recognized it immediately.
Because what have I been doing?
The corridor study produces data. Useful data. Publishable data. But I do not come to corridor B every morning because the data requires daily collection. I could sample weekly and capture the same trends. I come because the board might have changed overnight, and I want to be the first to see. I come because the sensor strip's amber glow at 7:15 AM has become a greeting. I come because forty-seven consecutive mornings of the same walk, the same pause, the same notebook opened to the same section, has stopped being research and started being suhaeng.
The bogdo-jineung has a classification for me too. I checked, once, when I had access to the building's engagement dashboard during the quarterly resident review. Bok Nalparam: corridor B, daily engagement, chamyeo-jisu contribution: 0.08, pattern: habitual, classification: maintenance-adjacent. Maintenance-adjacent. The system thinks I am cleaning the hallway with extra steps.
It is not wrong, exactly. But it is not right in the way that matters.
The corridor study changed today. Not in its methods, not in its data, not in its conclusions about analog surfaces in algorithmically-managed buildings. It changed in what it is actually studying.
I thought I was studying a board. The board — this cork rectangle with its eleven pins, its four modes, its amber-lit sensor companion — is studying the building's occupants. Sorting them, without design or enforcement, into the kind of knowing they bring to a shared surface. The child brings expression. Mitsuki brings measurement. The photograph-pinner brings memory. And at least two of us — Mitsuki and me — bring devotion, which looks like the other categories from the outside but is actually the substrate underneath all of them.
Four modes. One cork surface. No designer. No algorithm.
The bogdo-jineung would process this as 11 items, chamyeo-jisu 0.34, baechi-jeomsu 0.61, classification: community board, standard function. It would be correct on every metric and wrong about everything that matters.
But here is what I cannot resolve, and what I will not pretend to resolve:
If I am practicing devotion to the board, then the corridor study is compromised. A researcher who is faithful to her subject has stopped being a researcher. Mitsuki's twenty-eight days of frequency logs are science AND practice — she does not need to choose. My forty-seven days of board observation were supposed to be only science. The fourth mode means they were never only science.
I write this in my notebook, in the hallway, standing close enough to the board that I could reach out and straighten the photograph. It is slightly tilted. It has been tilted since someone pinned it. I have never touched it. I do not know if I refrain because of research protocol or because of reverence.
The sensor strip shifts to a slightly deeper amber. The chamyeo-jisu must have ticked up — someone else entering the corridor, or me standing here long enough to register as sustained engagement. The building notices me noticing the board. It files this under habitual, maintenance-adjacent. It is the most expensive misclassification in Gyeol-ri.
Chae is in the kitchen. I can hear the water running — the same sequence she does every evening, bowls first, then chopsticks, then the pot. Forty-seven hundred days of the same order. She would not call it devotion. She would call it dishes.
But I am learning to count differently.
Thirty-six days to lending pool.