Using Corpus Data to Examine Collocational Patterns, Lexicographic Representation, and the Nature of Near-synonymy

Main Article Content

Sara Sowers-Wills


Working from corpus data collected for near-synonymous adjectives, this paper argues that the subtle differences found in the near-synonymous relationships between the senses of lexical units are relevant for the structure of the lexicon and also have implications for lexicography. Whether there is definable nuance in meaning that distinguishes the use of one near-synonymous adjective over another is an important question. If usage trends found in corpus data reveal subtle but definable semantic nuance between near-synonyms, then this fine-grained distinction should be reflected in dictionaries—particularly learner’s dictionaries since language learners depend on their content to achieve native-sounding language. The present study examines the nature of near-synonymy in English by analyzing semantic relationships in two pairs of near-synonymous adjectives (i.e., big/large and enormous/huge) and the nouns with which they frequently collocate. Data were collected from The Corpus of Contemporary American English, and WordNet Search 3.1 online was used to sort the nouns modified by each of the near-synonymous adjectives into semantic categories. Edmonds and Hirst’s (2002) cluster model of lexical knowledge provides suitable representation of the fine-grained aspects of meaning and behavior that distinguish near-synonyms. The model facilitates interpretation of the data presented here, which may benefit the representation of near-synonyms in English learner’s dictionaries in the future.

Article Details