A Study of Comparison in Corpora of New Zealand English

Abstract

Comparison is pervasive in all areas of language use. For both practical and psychological reasons human beings constantly make comparisons of various kinds. The orthodox linguistic description of comparison has centred on the comparative and superlative forms of adjectives and adverbs and on a relatively narrow range of meanings. The research reported in this thesis sets out to demonstrate that both the meanings expressed through comparisons, and the linguistic means which are typically used to do so, are much more diverse than usually acknowledged. More specifically the thesis will report on the corpus-based derivation of a taxonomy of linguistic devices used to make comparisons and on distributional analyses of selected types which are used to express various categories of comparison within this taxonomy. The derivation of the taxonomy was based on the manual analysis of a 100,000-word sample corpus comprising extracts from the Wellington Corpora of Written and Spoken English and the distributional analyses were carried out using a combination of concordancing and manual analysis on the ICE-New Zealand corpus. The analyses described already were supplemented by the compilation and analysis of a pilot corpus and reference to English grammars and thesauri. The results of the research confirm that the previous understanding of comparison as reflected in the linguistic literature has been artificially narrow. Thus statements comparing a state of affairs with what is desirable, normal, or necessary, or with what used to be or may become the case, turn out to be subcategories of comparison in their own right alongside the scalar expressions of rank with which comparison is most often identified. Similarly, the range of linguistic means used to express comparisons proves to be both wider and deeper than normally assumed, encompassing marking by affixes, single words from virtually every part of speech and phrasal items, including a wide array of different sentence frames. The distributional analysis highlights differences between the spoken and written sub-corpora. The claims made are illustrated by examples taken from the corpora used in the study. The thesis attempts to account in general terms for some of the factors influencing the choice of the particular linguistic means employed to express comparisons. It also examines the pedagogical implications of the results regarding the range of meanings which comparisons can be used to express and the range of linguistic means which can be used to make them.