Rees, Geraint Paul2025-05-152025-05-152022Rees GP. Using corpora to write dictionaries. In: O'Keeffe A, McCarthy MJ, editors. The Routledge Handbook of Corpus Linguistics. Abingdon: Routledge; 2022. p. 387-404.9780367076399http://hdl.handle.net/10230/70401Lexicographers are often characterised as pioneer corpus linguists. This argument is valid to a degree. In the past, lexicographers were supported in their work by ‘corpora’ comprising millions of slips of paper recording examples of words in use. However, taking a more pedantic stance, as lexicographers are apt to do, the extent to which these collections of citations constitute what is currently understood as a corpus in corpus linguistics is questionable. Citation slips are not complete texts but rather extracts of texts recorded by citation readers. The decision on which extracts to record is subject to readers’ intuitions and prejudices. Corpus evidence has shown us that intuitions are an unreliable basis on which to build up an accurate picture of language use, since we tend to notice utterances which are unusual in some way rather than those that are typical or mundane. Prejudice comes into play since we tend to have preconceived, frequently inaccurate, ideas about what constitutes normal language use.application/pdfengThis is an Accepted Manuscript of a book chapter published by Routledge/CRC Press in The Routledge Handbook of Corpus Linguistics on 2022, available online: http://www.routledge.com/9780367076399 or http://www.crcpress.com/9780367076399info:eu-repo/semantics/bookPartCorporaDiccionarisLexicografiaCorpusinfo:eu-repo/semantics/openAccess