Perception may be less a recording than a wager. A growing view holds that the brain does not passively receive the world but actively predicts it, correcting itself only where prediction fails. Seeing, on this account, is controlled hallucination disciplined by error.
Prediction before sensation
The proposal reverses the intuitive order. Expectation comes first, and sensation serves mainly to correct it. The mind meets the world already anticipating it.
The economy of error
Efficiency motivates the scheme. Transmitting only what is unexpected conserves the brain’s scarce resources. Surprise, not stimulus, is the currency.
Hierarchy of expectation
Prediction is layered. Higher levels forecast the activity of lower ones, and mismatches propagate upward as correction. The architecture is a cascade of guesses.
Illusion as evidence
Perceptual illusions become clues rather than curiosities. They reveal the priors the brain imposes when evidence is thin. Error exposes assumption.
Action as inference
The framework dissolves the line between perceiving and acting. To act is to fulfill a prediction by changing the world to match it. Behavior becomes inference in motion.
The self as model
The boldest extension turns the lens inward. The sense of being a self may itself be a prediction the brain maintains. Identity becomes a durable inference.
The larger point
The predictive view casts perception as anticipation corrected by error rather than passive reception. It explains illusion, unifies perceiving and acting, and reaches even the sense of self. Seeing is a disciplined guess.
