A Corpus Stylistic Analysis Of Henry James’s Fiction

Corpus stylistics is an emerging research field which combines corpus linguistics and traditional stylistics, aiming to make a quantitative description and qualitative analysis of stylistic features in literary texts by adopting corpus techniques and statistical methods.Corpus approaches can complement approaches in traditional stylistics and literary criticism,and provide the possibilities to reveal stylistic features or linguistic facts that have largely remained hidden from human observation.Henry James (1843-1916) is a great novelist, critic and literary theorist in the transitional period of English and American fiction from traditional realism in the 19th century to the 20th century modernism. He occupies an important position in both American and British literature. A great many literary studies on James’s fiction have been conducted,but very few studies have dealt with his fiction from the corpus stylistic perspective. The present paper adopts three corpus linguistic theoretical principles as the theoretical foundations for corpus stylistics,namely, language as a social phenomenon, the equivalence between frequency and significance, and the unity of form and meaning. It makes an analysis of James’s fiction on the basis of two corpora: the James corpus containing his 20 novels, and the 20th century corpus including 23 novels by 16 American writers in his time, focusing on the keywords and five-word key clusters extracted by using the KeyWords function in WordSmith Tools 4.0. It aims to answer the two questions: (1)what stylistic features in James’s fiction could be disclosed by the five-word key clusters data? and (2) what stylistic features in his fiction could be revealed by the keywords data?Based on the data analysis and comparison, the results show that the mean word length and sentence length in the James corpus are longer than those in the 20th century corpus. The findings also show that, compared with his contemporaries, (1) for words,James uses significantly more speech act verbs, mental process verbs, and intensifiers to describe the characters, speech behaviors or highlight their feelings and attitudes; (2) for key clusters, he tends to use significantly more As If clusters, Body Part clusters, Speech clusters, Negation clusters, and It/There clusters to depict the characters or events in his fiction.