P1.P.5Gregory's constructivist theory of perception and the role of inference

Where Gibson saw perception as direct, Richard Gregory (1970, 1980) saw it as indirect — an active construction by the brain that uses sensory data plus prior knowledge to make a best guess about the world. His theory is the strongest top-down account.

Perception as hypothesis

Gregory described perception as the brain forming hypotheses about the world from limited sensory input. The retinal image is impoverished and ambiguous — the brain fills in gaps using schemas, expectations and past experience.

Inference and constancies

Size constancy is a direct example. A friend walking 50 metres away projects a much smaller retinal image than a friend 5 metres away, yet you do not perceive the distant friend as physically tiny. The brain infers their real size by combining retinal image + depth cues.

Similarly shape constancy (a door looks rectangular even when it appears as a trapezoid in the retinal image as it opens) and colour constancy (a white shirt looks white in yellow lamplight) all rely on inference.

Evidence

• Visual illusions: Müller–Lyer and Ponzo are explained by size constancy misfiring on inappropriate depth cues. The brain is making the wrong hypothesis.
• Hollow face illusion: a hollow mask is perceived as a normal protruding face, even when you know it is hollow — the brain's hypothesis (faces protrude) overrides the depth data.
• Ambiguous figures (Necker cube): the brain alternates between hypotheses when no single interpretation wins.
• Bruner & Minturn (1955): an ambiguous figure between letters was read as a "B" but between numbers as "13" — context (top-down expectation) shaped perception.

Strengths

• Explains illusions and ambiguous figures cleanly.
• Accounts for the role of culture, expectation and emotion in perception.
• Compatible with what we know about cortical feedback connections — there are more "downward" connections in the visual system than "upward" ones.

Weaknesses

• Most evidence comes from artificial 2-D stimuli (illusion drawings); real-world dynamic perception is mostly accurate.
• Cannot easily explain how perception is fast and automatic in clear conditions — too slow to fit Gregory's deliberate-inference picture.
• The role of prior knowledge is not always clear; some illusions persist even when you know about them.

Synthesis with Gibson

Most modern psychologists treat perception as interactive — direct in clear, dynamic, ecologically rich conditions (Gibson) and constructive in degraded, static or ambiguous conditions (Gregory). For a top-band GCSE answer, name both, give a brief contrast, and conclude that they are complementary, not contradictory.