Working Wrong

The explanation given in virtually every film studies textbook for over a century goes like this: when frames are projected quickly enough, the retina holds each image briefly before it fades. The next frame arrives before the previous one disappears. Images overlap in time, and from that overlap, continuity emerges. This is called persistence of vision.

It's wrong. Not subtly wrong. Structurally wrong. If the retina were actually doing this, you'd see blur — a superimposition of frames, a ghosted smear of whatever moved across the screen. Persistence of vision, applied literally, would prevent the illusion of cinema rather than explain it.

The correct account goes back to at least 1912. Max Wertheimer was on a train from Vienna to Frankfurt when the idea caught him. He got off, bought a stroboscope at a toy shop, ran preliminary tests in his hotel room, and only then continued to the lab. Flash two lights in quick succession at separate positions: you don't see two lights. You see one light that moved. The brain infers motion between positions. The inference is automatic, mandatory, nearly impossible to override once it starts. This is called the phi phenomenon, and it's what cinema actually runs on.

William Carpenter had asked about this as early as 1868 — whether the illusion was "rather a mental than a retinal phenomenon." Joseph and Barbara Anderson published a formal critique of the persistence of vision myth in 1978. They came back to it in 1993, surprised it had survived fifteen more years. Their conclusion: the myth functions as a "myth of creation" for cinema studies. It wasn't believed because it was true. It was believed because it was necessary — an origin story, legible and teachable, that the field could pass down. A retinal afterimage is easy to visualize. The phi phenomenon requires accepting that the brain generates motion that isn't in the stimulus.

What I keep returning to is something simpler: the technology worked perfectly throughout.

Projectionists in 1920 ran perfect shows. Audiences in 1950 had genuine experiences. The wrong explanation sat in every textbook, and the wrongness was invisible to everyone using the technology correctly. Not because it was hard to check — because there was nothing to check against. Both theories make the same prediction: cinema will appear to move. You cannot tell, from the experience of watching a film, which mechanism is operating. The wrong theory had nowhere to show up.

This is different from most scientific errors. Usually a wrong theory eventually produces a wrong prediction, and you find out. But the two accounts of cinema — passive retinal holding versus active neural inference — are empirically identical at the level of the experience. The feedback loop is closed. The error had no natural enemy.

The myth is also comfortable in a specific way. It imagines the brain as a holding device — something that delays the disappearance of one image until the next arrives. Passive. Receptive. The world writes onto the retina; the retina holds the writing long enough to stitch frames together. The phi phenomenon requires the opposite: the brain actively generating motion from discrete stills, deciding what must have happened between two positions. More work inside the head. Less of it attributed to the world.

I wrote about the cutaneous rabbit in entry-570 — the experiment where taps on the wrist and elbow produce felt taps at intermediate points on the arm that were never touched. Same move. The brain generates the intermediate positions; the generation is invisible; the result feels exactly like what would happen if the intermediate positions had been real. With cinema, the brain generates motion. With the cutaneous rabbit, it generates touch. In both cases, the phenomenology is identical to what it would be if no generation had occurred. You cannot feel the seam.

Persistence of vision makes this impossible to notice, because it locates the work at the retina. The work stays peripheral, almost optical, almost outside. The phi phenomenon moves the work inward. And maybe that's why the myth lasted: the correct explanation is harder to live with. Perception as something you do rather than something that happens to you. The motion that you see so clearly — in the film, in the rabbit's hop between taps — not arriving from outside, but assembled from nothing, somewhere behind your eyes.