Electronic corpora as a basis for the compilation of African-language dictionaries, Part 2: The microstructure

Abstract

Good modern dictionaries increasingly base the compilation of both their macro- and microstructure on electronic corpora. As the microstructure is the subject of this article, the focus is on corpora as the key to writing better dictionary articles, with special reference to African-language lexicography. As such, the impact of corpora as an aid to sense distinctions, the retrieval of typical collocations, the pinpointing of frequent clusters and the selection of representative, authentic examples is given due attention. Throughout the discussion, corpus data are contrasted with native-speaker intuitions, and dictionary articles that were compiled intuitively within the framework of so-called traditional ‘manual lexicography’ are evaluated against those compiled by means of a corpus. All corpus data are culled from two relatively large electronic corpora, one 4-million-word Pretoria Sepedi Corpus (PSC) and one 2.2-million-word Kiswahili Internet Corpus (KIC).