The Oxford Children's Corpus: Using a Children's Corpus in Lexicography

This paper presents the Oxford Children's Corpus, a 30-million-token corpus of texts written for children aged 5-14. A comparison of keywords and word sketches in the Oxford Children's Corpus and the Oxford English Corpus shows that writing for children differs from writing for adults in sometimes unexpected ways, thus highlighting the need for a separate children's corpus. It is shown that the corpus can be used by lexicographers in various ways: to refine headword lists for children's dictionaries, to identify key collocates and phrases that children are exposed to, to help choose appropriate child-centred examples, and to keep abreast of changes in the language. Future plans for the corpus are also set out, including plans to enhance the newly-included subcorpus of writing by children. Adapted from the source document