A visual cliff is a psychology experiment designed to test whether infants and animals can perceive depth. First published in 1960 by Eleanor J. Gibson and Richard D. Walk in Scientific American, it uses a clever glass-topped table that creates the illusion of a sudden drop-off while keeping the subject perfectly safe. The experiment became one of the most famous studies in developmental psychology and sparked decades of research into how and when depth perception develops.
How the Apparatus Works
The visual cliff is built around a raised platform with a sturdy sheet of clear glass extending out from both sides. On one side, a checkered pattern sits directly beneath the glass, making the surface look solid and shallow. On the other side, the same checkered pattern is placed far below the glass, at the bottom of a visible drop-off. This is the “cliff” side. Even though the glass is equally solid on both sides, the deep side looks like a dangerous fall to anyone relying on their eyes.
A narrow center bridge separates the two sides. In infant studies, a baby is placed on this bridge, and a caregiver stands at the far end of either the shallow or deep side, encouraging the child to crawl across. Researchers then record whether the infant willingly crosses the deep side or refuses to move past the visual edge.
The Original Experiment
Gibson’s interest in depth perception began with an unlikely subject: rats. She and Walk raised rats in complete darkness, then placed them on the visual cliff apparatus. To their surprise, even the dark-reared rats avoided the deep side, suggesting that depth perception didn’t depend entirely on visual experience. The researchers went on to test a wide range of animals, including chicks, lambs, kid goats, pigs, turtles, kittens, dogs, and monkeys. Species that can walk at birth, like chicks and goats, avoided the deep side immediately, demonstrating that for these animals, depth perception is functional from the very start of life.
When Gibson and Walk tested human infants (ranging from about 6 to 14 months old), most refused to crawl across the deep side, even when their mothers called to them from the other end. The conclusion from the 1960 paper was that infants can perceive depth by the time they learn to crawl. This finding landed in virtually every introductory psychology textbook and made the visual cliff one of the best-known experiments in the field.
Is Depth Perception Innate or Learned?
The visual cliff was originally designed to settle a longstanding debate: are we born knowing that a drop-off is dangerous, or do we learn it through experience? Gibson initially concluded that avoidance of the deep side was innate. But her thinking evolved significantly over the following decades.
By the early 2000s, Gibson herself had shifted her position. She came to argue that while basic depth perception may be present very early, the ability to recognize that a drop-off is actually dangerous for your body, given your current abilities, is something that has to be learned. Animals that walk from birth (like goats) appear to come pre-equipped with this judgment. But species that develop locomotion gradually, including humans, need experience to connect what they see with what their bodies can and cannot do.
One important finding supports this: negative feedback from actually falling isn’t what teaches infants to avoid drop-offs. Dark-reared kittens who had never experienced a fall still avoided the deep side once they could see. And human infants learn cliff avoidance without needing to tumble off an edge first. Instead, the learning seems to come from weeks of moving through space and discovering the relationship between what things look like and what the body can handle.
Crawling Age Matters More Than Experience
A study testing 49 infants between 7 and 13 months old revealed a surprising pattern. Researchers expected that more crawling experience would predict whether a baby avoided the deep side. Instead, the age at which the infant first started crawling was the better predictor. Babies who began crawling early were more likely to cross the deep side, while those who started crawling later tended to avoid it. The amount of time spent crawling didn’t reliably separate crossers from avoiders. A follow-up with 40 infants confirmed the same result: crawling-onset age, not the infant’s current age or total crawling experience, predicted behavior on the cliff.
Depth Judgment Is Tied to Posture
Some of the most striking modern findings come from research showing that depth avoidance doesn’t transfer between different ways of moving. In one study, experienced 12-month-old crawlers reliably refused to crawl over large drop-offs. But 12-month-old novice walkers, many of whom had been cautious crawlers just weeks earlier, stepped right over the edge repeatedly. Six of the walkers in the study never refused at all.
The pattern held even at extreme heights. None of the experienced crawlers attempted a 90-centimeter drop-off. But 63% of the novice walkers tried to walk right over it. When 9-month-olds were tested in a sitting posture (which they’d had months of practice with), they accurately judged whether they could lean across a gap. Tested in a crawling posture they were still learning, they made errors even at impossibly wide 90-centimeter gaps.
The takeaway is that infants don’t develop a general “fear of heights.” They learn, through weeks of practice in a specific posture, to judge what’s safe for that posture. Crawling experience teaches them about crawling dangers. Walking experience teaches them about walking dangers. Months of careful crawling do not carry over to the first weeks of walking, which is why new walkers are at higher risk of tumbling off edges that they would have avoided as crawlers.
The Role of Caregiver Emotion
The visual cliff also became a key tool for studying social referencing, the tendency of infants to look at a caregiver’s face when they’re uncertain about a situation. In a well-known 1985 study, mothers stood at the far side of the deep end and posed specific facial expressions. When mothers showed joy or interest, most infants crawled across the cliff. When mothers expressed fear or anger, very few crossed. The infants were using their mothers’ emotional signals to make a decision about an ambiguous situation, essentially asking “Is this safe?” with a glance and adjusting their behavior based on the answer.
Why the Visual Cliff Still Matters
The visual cliff experiment reshaped how psychologists think about perception in infancy. Before Gibson and Walk’s work, the dominant assumption in many circles was that newborns experience the world as a “blooming, buzzing confusion,” with meaningful perception developing only after months or years of learning. The cliff showed that depth perception is present far earlier than many expected, and the decades of research it inspired revealed a more nuanced picture: perception develops through an ongoing conversation between the body’s capabilities and the brain’s interpretation of visual information. It remains a foundational concept in developmental psychology courses and a starting point for understanding how infants learn to navigate the physical world.