This paper describes the process of writing a bilingual English-French dictionary in the late 1960s, and contrasts it with a similar task in the early 21 century, with all the benefits of a large text corpus and sophisticated query tools. The lexicography focuses on the entries for cook and cooking, designed for use both by an encoding English speaker and a decoding French speaker. Lexicographic evidence for the new entries comes from a lOO-million-word corpus of British English, a small corpus of French for the words cuire, cuisine, cuisiner, cuisinier, and cuisson, and their inflected forms, and a small parallel corpus of English and French texts. The use of KWIC concordancing, the Word Sketch program and the FrameNet database is described in detail, together with the problems of equivalence encountered. An account is given of the way in which these problems are tackled and the dictionary entry is drafted. Dictionaries exist... toprovide a series ofhints and associations connecting the unknown with the known. So runs Bolinger's dictum, quoted by Patrick Hanks in his inspiring address to the last Euralex Congress [Hanks 2000]. However, about thirty-five years ago I began writing a dictionary whose raison d'etre was although I couldn't have said so then to provide a series of hints connecting the known with the unknown. Eleven years and fifty-odd colleagues later, this had become the Collins-Robert English-French Dictionary (CREFD), now in its fifth edition and still going strong. It has latterly benefited from the linguistic resources ofHarperCollins' Bank ofEnglish, and from an editorial eye more knowledgeable than my own, yet still I see in its entries the lingering ghosts of those I wrote in my first stumbling years as a lexicographer. In this paper I shall consider how practical lexicography has changed over the past thirty-five or so years, and set these changes in the context of a handful of closely-related EnglishFrench entries. I shall not concern myself with the great changes which the computer has brought to the consultation of the dictionaries, but solely with the writing of them. I shall look first at how the entries were written in 1967, and then at how they might be written in 2002. Any such comparison must include (as well as consideration of changes in the language itself): (i) the technical skill and linguistic knowledge ofthe lexicographer; (ii) the instructions or guidelines given for the production ofthe dictionary; (iii) the sources ofevidence on which lexicographic decisions are based; (iv) any reference works available for consultation; (v) aids to production oftext; and
Journal Article Lexicographic Relevance: Selecting Information From Corpus Evidence Get access Sue Atkins, Sue Atkins Search for other works by this author on: Oxford Academic Google Scholar Charles J. Fillmore, Charles J. Fillmore Search for other works by this author on: Oxford Academic Google Scholar Christopher R. Johnson Christopher R. Johnson Search for other works by this author on: Oxford Academic Google Scholar International Journal of Lexicography, Volume 16, Issue 3, September 2003, Pages 251–280, https://doi.org/10.1093/ijl/16.3.251 Published: 01 September 2003