Machine-readable text corpora and the linguistic description of languages

Abstract

"To understand the role of machine-readable text corpora in linguistics it is necessary to consider the four possible sources of data for the linguist, viz. (1) the analyst's own introspection/ intuition, (2) more or less systematically conducted elicitation experiments with groups of native speakers of the language studied, (3) collections of authentic spoken or written citations gathered unsystematically, and (4) evidence extracted systematically from a well-defined corpus of texts. After a discussion of the advantages and disadvantages of the various sources of data, I will briefly exemplify recent advances made in the corpus-based description of languages that have become possible as a result of the application of computer technology to linguistics and then go on to present the major databases currently available for the study of English and German." (author's abstract)