A Corpus-based Stylistic Analysis Of Ernest Hemingway's Works

Hemingway is one of the most celebrated American writers, a remarkable stylist and the winner of the Nobel Prize for Literature in 1954. His outstanding style has won him numerous readers, and a great many scholars all over the world have spent much academic energy on his works from different perspectives. However, most of the studies are oriented to literary rather than linguistic aspects. This thesis aims at analyzing the stylistic features of Hemingway's works from linguistic point of view, and tries to compensate for the shortage of qualitative analysis in the previous studies by using quantitative analysis. A corpus of Ernest Hemingway's works (EHW) is built, and another corpus Brown (K-R) is also employed to better generalize the stylistic features of EHW, since the distinction of a style cannot be perceived without contrast.Based on the theories of modern stylistics and with the facility of computer software and programs (SPSS and FoxPro), a series of computation and tests are carried out on the level of graphology, lexis, syntax, and semantics. It is concluded that there exists significant difference on the four levels in the two corpora. Hemingway's preference for shorter paragraphs and sentences, his choice of common Anglo-Saxon words, his economy in the use of adjectives and adverbs, his frequent use of repetition, etc., establish his unique stylistic features. In this way, Hemingway succeeds in carrying out his principle, "Less is more", and urging readers to dig out the implications below the surface.