A Research On The Contribution And Application Of The Corpus Of Contemporary American English To English Synonym Discrimination

Vocabulary is the basis of language learning, and since English is a language whichhas a great deal of synonyms, clear differentiation and proper usage of synonyms are thekeys to improve the language learning skills. Traditional ways of synonymsdiscrimination mostly rely on the intuition and introspection; it does not really havepositive effect on the differentiating behavior. As the booming of computer technologyand the appearing of lots of large--scale and well-designed corpora and concordancesoftware, corpus-based approach has been more and more widely used in languageresearch.Corpus of Contemporary American English (COCA) developed by Professor MarkDavies who is from Brigham Young University of USA contributes to English synonymdiscrimination: more than425million words of text is equally divided among spoken,fiction, popular magazines, newspapers, and academic texts, which provides rich andauthentic language material. Through observing the frequency/genre distributions andthe collocates of synonyms, the users are capable of distinguishing the synonyms, andavoiding the pragmatic mistakes. The architecture and interface of COCA make itunique to other corpora: an integrated thesaurus with more than60,000synsets allowsfor powerful synonym-based queries; the function of “synonym chain” is uniquelyavailable for the users to see an entire web of related words easily and quickly.In this chapter, the author applies COCA to distinguish sampled groups ofsynonyms: acknowledge, admit, and confess; apparently, obviously, evidently; absurd,bizarre, weird. These words are selected from CET-6glossary and assisted by COCAand WordNet software, according to the frequency rank and mutual information, threegroups of synonyms are settle down. Based on the competitive corpus, the authordifferentiates the sampled synonyms from genre/frequency distribution, collocationalfeature and semantic prosody. By analyzing the data and tables, the author figures outthat there are no exact synonyms and the intention to distinguish them merely throughone method is inappropriate, these methods should be verified complementally.The present thesis is composed of five chapters: Chapter One is the introduction,consisting of the background information, significance of the study and the researchquestion involved; Chapter Two literature reviews mainly focus on the relatedknowledge of synonyms, since it is the research object, and also in this chapter a thorough review on the methods of synonym discrimination in traditional way andcorpus-base method. Chapter Three includes two main parts: one part is to introduce thecorpus linguistics, and the other part is the introduction of COCA, which including thecomposition of the corpus and the competitive advantage in synonyms discriminationthan other corpora. Chapter Four is the particular process of data collecting, processingand analyzing. Chapter five is the conclusion of the thesis, summarizing the results andshowing the pedagogic implication of the paper, as well as the limitation of thisresearch.Corpus-based study is a complement for the traditional way to distinguishsynonyms, which provides more reliable and convincing results. This approach has asignificant implication in pedagogical field.