University of Minnesota
November 10, 2009
Research on babies helps reveal how human vision develops.
In a baby's reach, clues to the development of human vision
By Deane Morrison
Sitting on her mother's lap, 5 1/2-month-old Fiona looks at a wire figure hanging in front of her. She reaches out and grabs it, unaware that a camera records her every move.
Fiona is one of many young babies helping researcher Al Yonas tackle what he calls "the toughest problem in the Universe"—how we construct a meaningful world from a 2-D pattern of light and color on the back of our eyeballs.
A professor in the University of Minnesota's Institute of Child Development, Yonas studies young babies to find out at what age they develop the ability to interpret the 3-D form of the wire figure, which is much like a line drawing.
That interpretive ability is described as sensitivity to pictorial depth cues. People who have it don't need to move in order to perceive the depth of an unfamiliar scene. If the scene is Yonas' wire figure, 7-month-old babies see, and thus react to, the figure the same as adults do; that is, they see the shape as sticking out toward them.
And younger babies? The jury is still out.
Yonas pursues his research both for the sheer joy of discovery and because it may someday yield dividends, perhaps by helping children blind from birth develop sensitivity to pictorial depth cues after having their sight restored later in life.
To understand the experiment, play the accompanying video and freeze it at the :27 mark. The wire figure will probably look like a solid 3-D object, with three sides that meet at a corner in the middle. The middle part appears to be closest to you.
Now freeze the video at the 1:14 mark. The figure is oriented the same way but, clearly, the center of the form is farthest away.
If you were sitting where our camera was, you'd be able to tell that the center is most distant (assuming you have normal binocular vision).
But if you covered one eye, your binocular vision would disappear and the figure would look like a convex object whose nearest point is—you guessed it—the central Y-shaped vertex.
Older babies with one eye covered tend to reach toward the Y because that seems to be the closest part of the object, as it does for adults using one eye. But when older babies use both eyes they tend to perceive, correctly, that the outer corners are closer and reach for one of them instead.
"With 7-month-olds, the difference is reliable," says Yonas. "We're still collecting data for 5-month-olds."
When children learn to perceive spatial layout, they depend partly on their experiences navigating the 3-D environment, using not only motion but also binocular vision, touch, and other cues, according to Yonas. But what about the ability to perceive depth with only one eye in a motionless environment or a picture? He thinks that the ability to use "static monocular depth cues"—which depends on the recognition of particular patterns—is complex and may develop later than sensitivity to other kinds of depth cues.
"The point of our work is to determine when babies first respond to pictorial depth cues like shading and perspective," he says. In Yonas' wire figure, the pictorial depth cue is the way the lines intersect.
"I think that the ability to use these cues develops last and is most fragile," Yonas says. He cites the case of one Michael May, a business executive who was blinded by a chemical explosion at age 3 and had his sight partially restored at age 46. May lacks pictorial depth perception (remember, it's fragile) and can't recognize anything, even his wife, unless he or the object is moving, says Yonas.