The Seam cultural feed linked the corridor piece to the Commission takedown notice on a Sunday.
Gyeol-ri sent Mitsuki a screenshot. Mitsuki looked at it for two minutes before she understood what she was reading. Not a takedown of the corridor installation -- the notice was directed at a different project, something that had cited the Vernacular Index without permission, or used the Commission report as source material in a way the Commission's communications office had decided to contest. The corridor was linked because the algorithm had placed it adjacent. Guilt by recommendation. This is how things spread, and this is how they get contaminated: not by being the target but by being nearby when the target is named.
Gyeol-ri wanted to know what to do.
Mitsuki thought about it for longer than the question seemed to require. The question was: should we respond to the feed post. The answer was: no. But she wanted to know why the answer was no, and the why took longer than the answer.
The corridor piece was not a document that could be taken down.
This was not a legal claim. It was an observation about what the study had become. The Vernacular Index was a clinical document, yes, but it was also filed in Yun's office and cited in the Commission report and referenced in Gyeol-ri's installation and described in Chae-Gyeol's USB document in the drawer below the annotation wall. It existed in too many registers simultaneously to be retractable. Someone could take down a link. They could not take down Geum-hee saying it was just a corridor, or Il-bong saying nobody checked on his daughter there, or Mrs. Park saying the corridor was the right length. Those conversations had happened. They were in a clinical notebook in Mitsuki's office, handwritten, not uploaded to anything. The Commission's communications office did not have access to the handwriting.
She had been thinking for months about the corridor's vulnerability -- the way each layer of documentation added intention to a space that was valuable because it lacked intention. The Ordinary Gradient contaminated itself. The Stratigraphy recorded its own contamination. She had worried that the corridor was fragile precisely because it had been studied.
She was wrong about the direction of the fragility. The corridor was not fragile because it had been studied. It was durable because it had been studied in too many ways by too many people with too many different methods to be reduced to any single document that could be taken down.
The laundry conversations were in a notebook. The session data was in the clinic system. The witness log entries were in a waterproof notebook bolted to relay housing. The photographs were in Bok's studio on a wall she had not been invited to see. The USB was in a drawer. The card on her window ledge said only: What I have not sent anywhere.
None of these were the same record. None of them could stand in for the others. Together they were the study. Apart, each one was just a note someone had made about something that happened in a corridor.
She told Gyeol-ri: do not respond to the feed post. Let the algorithm link what it links.
Then she wrote it in the Stratigraphy, in the margin next to Layer 8: the study cannot be retracted because it was never in one place. This is not a design feature. It is what happens when a space is documented honestly -- the documentation ends up as distributed as the space itself, spread across registers, held in different materials, accessible to different audiences, belonging to no single institution.
The corridor does not need defending.
You cannot remove something that was never in one place.