Gravity-defying ramps take illusion prize
Vision scientists award 'Oscar of perception' to Japanese mathematician.
In a packed concert hall, Kokichi Sugihara wields a pickaxe and mimes a blow to the stage. "I am a miner, and I have a secret," Sugihara says, adjusting his hard hat and headlamp. "I have discovered a new super-magnet."
A screen behind Sugihara begins playing a video. A cardboard structure appears, consisting of four ramps ascending to a raised platform. A hand places a wooden ball at the base of a ramp, and it rolls uphill, before stopping on the 'super-magnetized' platform. As the same trick is repeated for the other three ramps, the crowd lets out an "ooh".
Only there is no super-magnet, and Sugihara is no miner. He's a mathematician at Meiji Institute in Kawasaki, Japan, and his display is — officially — the world's best visual illusion.
At least, according to the hundreds of vision scientists in the audience who have come to judge the sixth competition to find the Best Illusion of the Year, a satellite event of the Vision Sciences Society's annual meeting in Naples, Florida. Sugihara's gravity-defying marbles beat nine other finalists, chosen from 84 entries.
Fooling the eye
Vision scientists are not easily fooled — so how did Sugihara do it? All is revealed as the camera angle changes to show that what looked like four ordinary ramps leading up to a platform was really an intricate set-up of sheared support columns, skewed slide angles and ramps of different lengths, all of which were above the platform all along.
From every viewpoint but one the contraption looks like what it really is: a mess of ramps and columns at different angles and eccentricities. But from one spot, the three-dimensional solid becomes ambiguous, and our brains favour the solution that puts the columns at neat right angles, making the ramps seem to be below the platform.
"Humans prefer rectangular solids over stranger shapes," Sugihara says, because our visual systems are used to carpentered environments. Children of kindergarten age, whose perceptual systems have not been so shaped by experience, are seldom fooled by the illusion. "All of my daughter's friends look at me and say, 'What's so funny?'," says Sugihara.
Sugihara's illusion was inspired by a computer program he wrote that turns two-dimensional line drawings into renderings of three-dimensional solids. The program worked beautifully for everyday solids, but Sugihara wondered what would happen if he fed his program some drawings of 'impossible figures', of the kind that might appear in an artwork by M. C. Escher.
"Most of the time, the software rejected the input," Sugihara says. "But once in a while it found a solution." The result was a number of figures that appeared to move in unnatural ways, including the gravity-defying balls that took the prize.
The competition, hosted by the Neural Correlate Society, is the "Oscars of perception," says society president Susana Martinez-Conde. The trophies are themselves visual illusions — wooden sculptures that look vastly different depending on the observer's point of view.
Illusions are close to cognitive scientists' hearts, because they reveal how perceptual systems are organized by making them fail spectacularly.
"Illusions are cool and fun, but they're important because they are mistakes of the visual system that give us clues about how the underlying processes work when they're functioning properly," says vision researcher Peter Tse of Dartmouth College in Hanover, New Hampshire, a finalist in this year's illusion contest.
Tse's entry showed that shifts in attention can change our perceptual experiences dramatically. He took a horizontal German flag and a vertical Greek flag, but in their complementary hues, so that the German tricolour of black, red and yellow became white, turquoise and blue, respectively, and the blue and white Greek flag switched to yellow and black.
Looking at a colour for long enough produces an afterimage of its complementary colour, so staring at the oddly coloured flags produces afterimages of the flags' true colors.
Tse's twist was to ask viewers to attend to either a rectangle with its long edge vertical, like the Greek flag they had just seen, or horizontal, like the German flag. This determined which afterimage they saw. "You can control your own conscious visual experiences using your attention," Tse told the crowd.
Spot the pirate
Another finalist, Daniel Simons of the University of Illinois, Urbana-Champaign, updated his famous "gorillas in our midst" illusion, in which a gorilla struts across a screen undetected by observers instructed to pay attention to white-shirted individuals1.
Clad in a gorilla suit himself, Simons played a clip much like the original gorilla video. Sure enough, a gorilla strutted across the screen, but this time a curtain in the background changed colour and one of the people in the clip ran off-screen. The audience members, who had been expecting a gorilla, didn't see the other changes — nor did they notice a man in a pirate suit creeping onstage.
When the clip ended with the phrase: "Did you see the gorilla?", Simons turned to the audience: "Did you see the pirate?"
- Simons, D. J. & Chabris, C. F. Perception 28, 1059-1074 (1999).