말뭉치 바탕 구어 연구

Abstract

This paper provides an overview of currently available language corpora in linguistic scholarship, and describes the practical and theoretical import of utilizing spoken language corpora in studies of language. In so doing it is intended to promote the interest of the linguistic academia in developing and using spoken language corpora, by showing that spoken language, contra written language, is inherently dynamic and has its own distinctive grammatical structures and idiosyncrasies. In traditional research paradigms it has often been the case that written data or made-up data are used to support a purported argument or to disprove certain arguments, Despite their obvious theoretical merits by themselves, these paradigms fell short of providing dynamic, true-to-life linguistic descriptions and explanations thereof. This fact that spoken language data show more concrete representations of language in use not only presents a contrast with the written language and, for that reason, linguistic analyses drawing upon it, but also has a theoretical implication that analyses based on spoken corpora have higher level of explanations and predictabilities of linguistic phenomena. This is explicated in this paper by presenting quantitative research that reveals pragmatic tendencies and tactics, and distinctive features of spoken language. This paper also argues for the necessity of developing systematic corpora based on naturally-occurring spoken language in diverse real-life circumstances, a task which in turn requires as its prerequisite corpora-tagging specialists, arrangements for copyrights, and standardization of transcription and annotation.