Alternating between practical examples and theoretical evaluation, the reader is led step-by-step to a detailed understanding of the issues involved in corpus work and, at the same time, tempted to explore for himself some of the major applications where a corpus-driven methodology can reveal unprecedented insights into linguistic patterning.
The applications where the corpus-driven approach is exemplified are language teaching and contrastive linguistics. The book adopts and exemplifies the parameters of the corpus-driven approach and posits a new unit of linguistic description defined systematically in the light of corpus evidence. The two main approaches to corpus work are discussed as the “corpus-based” and the “corpus-driven” approach and the theoretical positions underlying them explored in detail. Thus, starting from the definition of what is a corpus and why reading a corpus calls for a different methodology from reading a text, the underlying assumptions behind corpus work are discussed. The book offers a combined discussion of the main theoretical, methodological and application issues related to corpus work. The problems, primarily those referring to Google Books digital form and metadata as well as to the retrieval and interpretation of data, will be illustrated with material drawn from my study. Because of its format, searches are inevitably full of obstacles which will be presented here in some detail. Importantly, the collection is immense, free and available at the click of a button, but-unlike traditional corpora and lexical databases-it is not a neat and tidy corpus of machine-readable texts. To verify my hypothesis, I use a handful of headwords from the third edition of the Oxford English Dictionary, which have been checked against the evidence of Google Books. I argue that the collection offers to researchers an extremely rich material which can help, inter alia, to antedate OED3 headwords. The aim of my paper is to show that the Google Books collection, a non-specialized textual resource available from, can be applied successfully for research in historical lexicology and lexicography.
0 kommentar(er)
