Prototype Theory
Is a penguin less of a bird than a sparrow?
Prototype theory holds that many categories are organised around their best examples, so members differ in how typical they are and the edges of a category can be fuzzy.
Ask a class to name a bird and you will probably hear sparrow or pigeon long before penguin or chicken, yet all four are birds. In the 1970s Eleanor Rosch turned this everyday fact into a research programme.
Prototype theory holds that many categories are organised around central, best examples rather than a checklist of necessary and sufficient features. When Rosch (1975) asked students to rate items as examples of furniture, chair and sofa came top and telephone came last. Rosch and Mervis (1975) linked such typicality to family resemblance: the most typical members share the most features with other members. William Labov (1973) showed fuzzy edges directly: as drawn containers grew wider, people shifted from calling them cup to bowl, and imagining them filled with mashed potatoes rather than coffee moved the boundary.
Typicality effects are robust; what they reveal is contested. Armstrong, Gleitman and Gleitman (1983) obtained typicality ratings even for odd number, a category with perfectly sharp membership, so graded judgements do not prove that a category lacks clear boundaries. Rosch herself (1978) later warned that prototype effects do not, on their own, tell us how categories are stored in the mind.
Cognitive linguists carried the idea into word meaning. Lakoff (1987), building on Brugman, and Tyler and Evans (2003) treat a preposition such as over as a family of related senses spreading out from a spatial core. On such accounts, senses like over the limit and the film is over are extensions of that core, not unrelated homonyms. For learners, a core-and-extension map is easier to hold than a list, and it shows where their interlanguage may be stretching an L1 category onto English.
In actionA typicality warm-up
Rank these from best to worst example of furniture: chair, sofa, wardrobe, lamp, rug, piano, telephone. Then argue about the bottom three.
An invented task modelled on Rosch’s rating studies. The arguments at the bottom of the list, not the agreement at the top, show learners where category edges, and their L1 categories, differ.
In the classroom
- 1
Teach the central sense of a polysemous word first (over as above and across), then present extended senses as a chain from it, asking learners to explain each link, sometimes through metaphor.
- 2
Introduce a new word or pattern with typical examples, as with give for the ditransitive construction, then deliberately add peripheral ones so learners do not under- or overextend it.
- 3
Ask learners to sketch or rank what counts as breakfast, a house or a friend in their culture, then compare with English texts to see whose prototype the coursebook assumes.
Myth and reality
Prototype theory shows that categories have no real boundaries, so meaning is just a matter of opinion.
Typicality is graded, but membership is often clear: a penguin is a less typical bird, not half a bird. Some categories, like cup and bowl, do have fuzzy edges; others, like odd number, show typicality effects despite sharp ones.
Reflect
When you explain a new word, do you give learners its most typical example first, and whose idea of typical is that?
Worth a minute before your next lesson, or a conversation in the staffroom.
SourcesChecked against Crossref and library records
Rosch, who first published as Eleanor Heider, developed the idea from studies of colour and shape categories, including work with the Dani of New Guinea, and extended it to semantic categories in 1975. Wittgenstein’s notion of family resemblance was an acknowledged precursor.
Rosch, E. H. (1973). Natural categories. Cognitive Psychology, 4(3), 328–350.
Rosch, E. (1975). Cognitive representations of semantic categories. Journal of Experimental Psychology: General, 104(3), 192–233.
The classic typicality ratings for everyday categories like birds, fruit and furniture, and evidence that typicality shapes how people process them.
Rosch, E., & Mervis, C. B. (1975). Family resemblances: Studies in the internal structure of categories. Cognitive Psychology, 7(4), 573–605.
Explains why some members feel more typical: they share more attributes with the rest of the category.
Armstrong, S. L., Gleitman, L. R., & Gleitman, H. (1983). What some concepts might not be. Cognition, 13(3), 263–308.
The sharpest critique: typicality effects appear even for well-defined categories, so they cannot by themselves reveal category structure.
Lindstromberg, S. (2010). English prepositions explained (Rev. ed.). John Benjamins.
A teacher-friendly account that builds the many meanings of English prepositions out from their spatial cores.
ConnectionsSee it in the graph
- Conceptual MetaphorMetaphor is one main route by which word senses extend from a prototype.
- Image Schemasnot yet writtenThe spatial cores of prepositions are often described as image schemas.
Know a teacher who’d like this?
Send them the card. It’s made for Facebook, Zalo, LinkedIn and Instagram, and links back here.