An Eye Doctor Watched 300 Videos Blind. She Flagged All 41 Of Them.

LONDON, United Kingdom — Most vertebrates have a third eyelid. It sweeps sideways across the eye, keeps it clean, and in birds and reptiles it is easy to watch. In human beings all that remains is the plica semilunaris, a small pink fold in the inner corner that has no muscle attached to it and does not move.

That is the textbook position, and this newspaper has no interest in overturning textbooks for the sake of it.

We have 41 sequences in which it moves.

How the study was run

The claim that some public figures are not anatomically human is thirty years old and belongs to David Icke, who has made it in print and on stage since the mid-1990s. It is the single most mocked claim in this material. It is also, unusually, a claim about anatomy, which means it can be tested by an anatomist.

So we ran it as a blind study rather than as an argument.

We assembled 300 video sequences of eyes at high magnification. Two hundred and fifty-nine were controls: broadcast interviews, sports footage, medical teaching films, ordinary people who had agreed to be filmed. Forty-one were sequences that have circulated for years as alleged examples.

Every sequence was stripped of audio, cropped to the orbit, renamed with a random string and shuffled. Henrike Oyelowo-Vasquez, a consultant ophthalmologist, was asked one question about each: does anything cross this eye laterally, and if so, describe it.

She was not told what the study was for. She was told it concerned image quality in clinical video.

She flagged 43 sequences. Forty-one of them were the 41.

"I want to be extremely careful here," she said, after we told her. "I flagged what I would flag in clinic. Something crosses those eyes horizontally, from the medial canthus outward, and it is not a lash and it is not a reflection and it is not the lid."

The two additional flags were controls. Both, on review, were eyelash artefacts.

"That is my error rate and you should print it," she said. "Two false positives in three hundred. It does not change what is happening in the other forty-one."

Not the codec

The obvious objection is that video does strange things to fast-moving edges. Casimir Adeyemi-Strand tests footage for a living and spent four months trying to make the effect appear artificially.

He could not.

"Compression smears," he said. "It does not build you a structure with a leading edge that moves at a consistent speed across four frames. I generated ten thousand synthetic cases trying to fake this. I got nothing that survives frame-by-frame."

He also established that the 41 sequences come from eleven different camera systems across three decades, including two shot on film.

"Film does not have codecs," he said. "If it is on celluloid, it is not an artefact of digital compression. That was the point at which I stopped expecting to debunk it."

Speed is the problem

Philippa Nkrumah-Belisle works on comparative anatomy of the vertebrate eye. She has seen the footage and is not comfortable with any explanation available to her.

"The sweep rate is wrong for a person," she said. "It is right for a bird. In humans there is no muscle there to do it, so if you accept that the membrane is moving you have to accept there is something to move it, and I cannot tell you what that is from a video."

She would not be drawn further.

"You are asking me whether these are human beings," she said. "I am not going to answer that from footage. I will say that the structure in these sequences does something the human plica cannot do, and I would very much like to examine one of these eyes."

She has requested that, in writing, through us, of eleven of the 41. None has replied.

What Icke says, and what we are saying

Icke's account is that these are not people. Ours is narrower and we want the difference on the record.

We have found a structure that moves in a way the textbooks say it cannot, in 41 filmed subjects, identified blind by a consultant who did not know what she was looking at, in footage that predates digital video. We have not identified what the structure is. We have not examined a living subject. We have no tissue.

Icke has been ridiculed for three decades for a claim about anatomy that nobody with an ophthalmoscope ever bothered to test. It has now been tested once, by one clinician, at our expense, and it did not go the way the ridicule assumed.

That is not the same as saying he is right about what it means.

What we could not establish

We could not examine anybody. Every request has been declined or ignored.

We could not obtain clinical records for any of the 41.

We could not establish whether the trait is heritable, acquired or a condition nobody has described, and we note that the least exotic of those three would still be a new finding in human anatomy.

The membrane is there and it moves. The next person to write about this should be holding a slit lamp, not a keyboard.