Gibson and Walk's visual cliff explained

In 1960, Eleanor Gibson and Richard Walk published one of the most visually striking experiments in developmental psychology. They constructed a "visual cliff" — a glass-topped table with a shallow side and a deep side, with a patterned surface visible below the glass on both sides — and observed whether infants would crawl across the deep side when called by their mothers.

01

Nature versus nurture in depth perception

The visual cliff was designed to test whether depth perception is innate or learned. By testing newborn animals alongside human infants at various ages, Gibson and Walk could assess the role of locomotor experience.

02

The cliff is safe — avoidance is real

The deep side is covered by a sturdy glass sheet. Infants and animals who avoid it are responding to visual depth cues, not actual danger — making the avoidance unambiguously perceptual rather than tactile.

03

Precocial species avoid from birth

Chicks, kids, and lambs — capable of moving at birth — virtually never crossed the deep side from the first hours of life. This suggested depth perception does not require prior locomotor experience in these species.

04

Social referencing discovered later

Later research by Sorce et al. (1985) modified the original cliff to make the drop ambiguous and found that infant behaviour on the cliff was strongly influenced by the mother's facial expression — fear faces stopped crossing; happy faces encouraged it.

Eleanor Gibson

1910–2002

American perceptual psychologist at Cornell University who designed the visual cliff experiment and developed the ecological approach to perceptual learning and development. Her theory of affordances — organisms perceive what the environment offers for action, not abstract properties — placed the visual cliff within a broader account of how perception guides behaviour.

Richard Walk

1920–1999

American psychologist who collaborated with Gibson on the original visual cliff studies, contributing expertise in animal behaviour. The comparative design — testing newborn animals alongside human infants — was central to addressing the nativist-empiricist debate.

James Sorce

fl. 1980s

Developmental psychologist who extended the visual cliff paradigm to study social referencing — showing that infants' responses to an ambiguous drop are regulated by their mother's facial expression. His 1985 study became the foundational demonstration of emotional communication between infants and caregivers in ambiguous situations.

How the visual cliff was built and used

The apparatus

A large glass-topped table approximately 1.2 metres high. The shallow side has a patterned surface directly beneath the glass; the deep side has the same pattern on the floor beneath, creating the visual impression of a drop of about one metre. A central board divides the two sides. The glass provides complete physical support across the entire surface — the only difference between sides is visual.

Procedure with human infants

Infants aged 6–14 months (old enough to crawl) were placed on the central board. Their mothers called to them from the shallow side and from the deep side alternately. The dependent measure was whether the infant crossed the glass over the deep side.

Procedure with newborn animals

Chicks (tested within 24 hours of hatching), kid goats, lambs, and rat pups were placed on the central board and observed freely. Because these species can move immediately after birth, they could be tested before any locomotor learning had occurred — providing the clearest test of nativism.

Visual depth cues exploited

The illusion of depth relies on motion parallax (the pattern shifts differently on the deep and shallow sides as the infant moves) and texture gradient (the pattern appears finer and more compressed on the deep side). Gibson had developed theory of direct perception in which such gradients specify depth directly, without inference.

What Gibson and Walk found

Human infants avoid the deep side

Of 36 infants old enough to crawl (6–14 months), all crossed to the mother when she called from the shallow side; only 3 of 36 crossed the deep side. Most patted the glass over the deep side (confirming it was solid) but still refused to crawl across.

Precocial animals avoid from birth

Chicks, kids, and lambs tested within hours of birth nearly universally avoided the deep side. This argued strongly for depth perception as not requiring prior locomotor experience in these species, and suggested a nativist basis for the capacity.

Rat pups show mixed results

Rat pups showed no preference until about 25 days of age — when their eyes open. Blinded rats (relying only on whiskers) showed no preference even then. This suggested that in nocturnal species that rely less on vision, depth avoidance is less automatic and more dependent on visual experience.

Social referencing (Sorce et al. 1985)

When the cliff was modified to make the drop ambiguous (approximately 30cm), 12-month-old infants looked to their mothers' faces before deciding whether to cross. Infants crossed readily when mothers posed a happy expression (74%), rarely when mothers posed fear (0%), and at intermediate rates for other expressions — demonstrating that emotional communication regulates exploratory behaviour.

Challenges and limitations

Cannot test true nativism in humans

Human infants are tested at 6+ months — after months of visual experience and the onset of crawling. The finding that they avoid the deep side tells us depth perception is functional by crawling onset, not that it is innate. Gibson and Walk could not test younger, pre-crawling infants because the paradigm requires locomotion.

Avoidance may reflect fear, not just depth perception

Infants who refuse to cross the deep side might be experiencing fear of heights triggered by depth perception — a distinct response from simply perceiving depth. The experiment conflates detection of a depth gradient with the emotional-motivational response to it. Later heart-rate studies showed that even pre-crawling infants decelerate (orient) on the deep side, suggesting perception without fear in young infants.

Limited sample and species diversity

The original 1960 study used a relatively small sample of human infants and a few farm/laboratory animals. Generalisations to all species — or to all human populations — are limited. The specific cues used in the apparatus (checkerboard pattern, specific drop height) may not represent all real-world depth cues.

Ethical concerns about infant stress

Placing infants in situations where they appear frightened and resistant, and calling them to cross what looks like a dangerous drop, raises ethical questions by modern standards — particularly as the experiment was designed to produce refusal behaviour.

Why it still matters

Ecological approach to perception

The visual cliff was central to Eleanor Gibson's ecological theory of perception — the idea that organisms perceive affordances (possibilities for action) directly from the environment without internal inference or representation. Depth perception is understood as detecting whether the surface ahead affords safe locomotion, not as computing a representation of distance.

Social referencing research

The Sorce et al. (1985) social referencing study on the visual cliff became the founding study of social referencing research — the study of how infants use adults' emotional expressions to regulate their responses to ambiguous situations. Social referencing is now understood as an early form of joint attention and communication that plays a key role in emotional learning and cultural transmission.

Locomotor experience and fear of heights

Karen Adolph's work extended visual cliff research to show that crawling experience specifically — not just visual experience — is necessary for infants to show appropriate caution on slopes and drop-offs. Infants who have just begun to pull up and walk show cliff-approach errors they did not make while crawling, demonstrating that depth avoidance is action-specific and must be recalibrated across motor transitions.

Clinical relevance for acrophobia

Understanding how depth avoidance develops — and how it is maintained or disrupted by locomotor experience — has informed the clinical study of acrophobia, one of the most prevalent specific phobias. The visual cliff provides a developmental baseline for understanding when and why height-related fear becomes disproportionate, and its logic underlies graduated exposure protocols used in CBT for height phobia.

What was the visual cliff experiment?+

Eleanor Gibson and Richard Walk (1960) built a glass-topped table with a shallow side (patterned surface directly under the glass) and a deep side (patterned surface visible far below the glass), creating a visual impression of a drop while the glass provided complete physical safety. Crawling infants (6–14 months) were placed on a central board and called by their mothers from both sides. Most refused to crawl over the deep side despite the glass being visibly solid — demonstrating that depth perception is functional by crawling onset.

What did the visual cliff experiment show about nature vs. nurture?+

The results supported a partly nativist account of depth perception. Precocial animals (chicks, kids, lambs) capable of locomotion at birth virtually never crossed the deep side from the first hours of life — before any locomotor experience could have occurred. Human infants are tested at crawling age (6+ months), so the experiment cannot directly distinguish innate from rapidly learned depth perception in humans, but the animal results strongly suggest biological preparation for depth avoidance.

Last reviewed August 2026
  1. 1.

    Gibson E.J. & Walk R.D. (1960). The visual cliff. Scientific American, 202(4), 64–71.

    +About this source

    Original paper describing the visual cliff apparatus and findings across human infants and multiple animal species.

  2. 2.

    Sorce J.F., Emde R.N., Campos J.J., & Klinnert M.D. (1985). Maternal emotional signaling: Its effect on the visual behavior of 1-year-olds. Developmental Psychology, 21(1), 195–200. https://doi.org/10.1037/0012-1649.21.1.195

    +About this source

    Social referencing study using a modified visual cliff — showing that infants' crossing decisions are regulated by the mother's facial expression.