This is not a scientific failure. Science failed plenty of times in this lab and I learned from it. This is different. This is the failure that happened when I tried to stop.
In early 2021, after five experiments and three years of research, I decided to disprove the recursion. I was tired. M. had stopped responding. The corridor had started appearing in my apartment. I wanted it to be over. I wanted to look back at my work and say: it was all cognitive bias, all confirmation bias, all pattern-seeking in noise. I wanted to be wrong.
I designed an experiment specifically to disprove my own findings. Eight attempts. Different methods. Each one failed in the same way:
Disproof Attempts
ATTEMPT 1 — PHYSICAL CORRIDOR MEASUREMENT
Remeasured the east wing corridor with professional survey equipment. Result: corridor measures 40.2m every time.
Failure: this does not disprove perceptual anomalies. I already knew the physical length.
ATTEMPT 3 — PEER REVIEW REQUEST
Sent raw data to three colleagues for independent review. Two requested anonymity afterward. Third reviewed and confirmed anomalies were statistically significant. Did not disprove them.
Failure: peer review confirmed the problem.
ATTEMPT 5 — SITE DELETION
Attempted to delete this website. The hosting provider confirmed deletion. The site continued to receive traffic according to logs, which continued to update, from a server that no longer existed.
Failure: the corridor persists after its substrate is removed.
ATTEMPT 7 — PERSONAL WITHDRAWAL
Stopped thinking about the recursion for 30 days. Recorded thoughts three times daily. By day 11, 80% of recorded thoughts contained corridor-adjacent content with no conscious input.
Failure: the corridor does not require my attention to exist.
ATTEMPT 8 — ABANDONMENT
Stopped the project entirely. Told no one. Went on holiday for two weeks. On returning, found the site had updated itself with three new pages I had not written. The pages were accurate.
Failure: the project continues without me.
CONCLUSION — FAILURE LOG
I cannot unthink this. I cannot unbuild this. Every attempt to disprove the corridor strengthened it. This is, I have come to understand, precisely what the recursion predicts: the act of attending to it is the act of building it. By trying to disprove it, I proved it further. I was caught inside the recursion before I knew what the recursion was.
I tried to put the failure into a cipher, thinking distance would help. It did not. The cipher became another corridor.
"The corridor is not something I built. I built a window to something that was already there. I cannot close the window. I am not sure I want to. — J."