코퍼스 기반 번역 연구에서의 장르별 분포 편차 분석을 통한 연어 관계 검증

Abstract

This paper aims to confirm that the corpus analysis of the collocational relationships among words, when solely based on their frequency of use, may result in less reliable data and thereby to find a way to overcome this limitation so that non-native translators can make better use of this useful tool. To this end, top 30 collocation nouns of two adjectival synonyms fast and rapid were listed in terms of usage frequency. Among them were proper nouns, ones used predominantly in specific contexts, and others as part of compound words. It was then found that such nouns differ from others in terms of how frequently they are used in COCA’s eight respective genres. Based on this, an index called the genre-specific frequency deviation was devised, which represents how different their frequency in each genre of COCA is from the average frequency pattern of the adjective in question.