Prototype theory
We represent a category by its average, idealized 'best example.'
What it means
Prototype theory proposes that a concept is mentally represented as a prototype — a summary or central tendency of its members' features — and that an item is categorized by its similarity to that ideal rather than by satisfying a strict definition. It elegantly explains typicality effects: a robin is judged a better, faster-verified bird than a penguin because it is closer to the avian prototype, and category boundaries are graded and fuzzy. Prototypes are abstracted across instances, so the theory predicts that people may 'recognize' a never-seen prototype more confidently than the actual items they studied. Its weakness is that pure prototypes discard information about variability and correlated features that people demonstrably retain. It is usually contrasted, and sometimes combined, with exemplar theory.
Examples
Asked to name a bird, most people say 'robin' or 'sparrow' rather than 'ostrich,' because the former sit near the prototype.
Asked to picture a chair, almost everyone sees a four-legged dining chair rather than a beanbag or a stool — and confirms the dining chair is furniture faster.
Shown a run of similar faces, people later say they recognise the blended average of them all with more confidence than the faces they actually saw.
First described in Eleanor Rosch (1973, 1975).