Home > Positioning > Persons > Gibson, James J.

James J. Gibson (1904–1979)

Gibson proposed that perception needs no construction. The account descending from Helmholtz held that the senses deliver impoverished input — a flat, ambiguous retinal image — from which the mind must infer a world; perception on that account is unconscious inference, and the interesting work happens inside. Gibson’s ecological approach denies the premise. For an animal moving through its environment, the light arriving from all directions — the ambient optic array — is structured, and its structure is rich: gradients, occlusions, flow patterns, and the invariants that persist beneath transformation specify the layout of surfaces and the animal’s own movement through them. Perception is the pickup of that information, not the decoration of a poor signal with stored knowledge. The proposal relocates the explanatory work — out of the head and into the animal–environment relation — and the tradition that formed around it is ecological psychology.

Life

Born 1904 in McConnelsville, Ohio. Graduate work at Princeton, where Edwin Holt — the behaviourist-realist then visiting — became his mentor; the doctorate itself (1928) was supervised by Herbert Langfeld, on memory for visual form, refuting a Gestalt claim that remembered figures change spontaneously toward better shape. It is the first instance of the pattern that organises the whole career: take a received claim about what the mind adds, and test whether the mind adds it. From 1928 he taught at Smith College, where Kurt Koffka was his senior colleague — Gibson never accepted Gestalt psychology, but he took from Koffka the conviction that perception is psychology’s central problem, and that the world as perceived comes already meaningful. His experimental reputation in the 1930s was built on adaptation and aftereffects, most famously the curvature aftereffect. From 1942 the war took him to the Army Air Forces’ aviation psychology programme, directing research on pilot testing and landing: the problem of what a pilot actually sees when approaching a runway gave him optic flow — the structured streaming of the whole visual field that specifies where one is heading — and convinced him that the laboratory’s frozen, pictorial stimuli had been the wrong starting point. From 1949 he was at Cornell, where the approach was worked out across three books: The Perception of the Visual World (1950), The Senses Considered as Perceptual Systems (1966), and The Ecological Approach to Visual Perception (1979), finished in the last year of his life.

The ecological approach

The information is sufficient. The founding empirical claim is about what is available to be perceived. Classical experiments impoverished the stimulus deliberately — a point of light in a dark room, a frozen picture in a tachistoscope — and concluded from the resulting ambiguity that perception must supplement. Gibson’s countermove was to describe the stimulation of an active animal in a normal environment: an ambient array sampled by a moving point of observation, transforming lawfully as the animal moves. Under transformation some structure persists — the invariants — and the invariants specify their sources. Ambiguity, he argued, is an artefact of the laboratory; in the ecological case the information is sufficient, and nothing needs to be inferred.

Ecological optics, and the environment described at the animal’s scale. To carry the claim, Gibson built a discipline for it: ecological optics, the description of light as it structures an ambient array rather than as rays forming a retinal image. And the 1979 book opens not with perception but with an account of what the environment is for an animal — medium, substances, and surfaces: air to move and see through, solids that resist, and the surfaces between them, where nearly everything an animal does takes place. The environment of perception, he insisted, is this furnished, scaled world — not the physicist’s space.

Perception and action are one system. The array is only rich for an animal that moves, and movement is only controlled by what is perceived: perception and action are not input and output of a central processor but a single continuous coupling. The 1966 book recast the senses accordingly — not passive channels but active, exploratory perceptual systems: the eyes in a turning head on a locomoting body, seeking information rather than receiving stimulation. And the array specifies the perceiver along with the world — optic flow carries where I am heading, occlusion carries where my body ends — so that perception of the environment and perception of oneself in it are one pickup with two poles.

Affordances. The 1979 book’s central concept names what perception is of. What an animal perceives, Gibson held, is first of all what the environment affords it — surfaces that support walking, objects that permit grasping, gaps that permit passage. An affordance is a real property of the environment, but a relational one: the same surface is walk-on-able for one animal and not another, and the fact is fixed by the relation between the surface and the animal’s capabilities. The concept was built to cut across the objective–subjective dichotomy — an affordance “points both ways,” to the environment and to the animal. Its considerable afterlife beyond psychology — the design world’s borrowed “affordance” among the rest — belongs to the tradition’s page.

What followed. A research school formed around the approach and formalised it, and a dispute with the computational theory of perception — Fodor and Pylyshyn’s 1981 critique, Marr’s assessment, and their descendants — has run from his last years to the present. Both are carried on the ecological psychology page.

Where Gibson stops

Gibson named two boundaries of the programme himself, and kept them separate. The first is mechanism: how the nervous system accomplishes pickup he deliberately declined to specify — the perceptual systems “resonate” to information, a metaphor he acknowledged as a placeholder, referring the question to another level of analysis. The second is scale: what an ecological account looks like past perception — the absent, the remembered, the imagined, language — the 1979 book gestures at in its closing chapters but does not develop. He handed both questions on rather than claiming them solved.

Key works

See also: Eleanor Gibson · Ecological psychology · Enactivism · Fodor · Merleau-Ponty