Corpora are a natural source of data for cognitive linguists, since corpora, more than any other source of data, reflect "usage" - a notion which is often claimed to be of critical importance to the field of cognitive linguistics. Corpora are relevant to all the main topics of interest in cognitive linguistics: metaphor, polysemy, synonymy, prototypes, and constructional analysis. I consider each of these topics in turn and offer suggestions about which methods of analysis can be profitably used with available corpora to explore these topics further. In addition, I consider how the design and content of currently used corpora need to be rethought if corpora are to provide all the types of usage data that cognitive linguists require.
corpora; cognitive linguistics; metaphor; polysemy; synonymy; prototypes; constructional analysis; statistics; R statistical program
