| Linguistic style of literature is the personal characteristics in the activities of using language, which reflects the author's casual habits in writing. To some extent, linguistic style can be characterized by quantitative characters. That is: different people tend to have different habits in talking especially in writing. They have their own styles in words, sentences, language structurse, sentence length, etc.Computational stylistics is a branch of computational linguistics, and it plays an important role in the author ownership identification, text style analysis and plagiarism-detection. Besides it also provides a new research method and research perspective for linguistic style study.Feiyu Bi and Tong Su are both the domestic famous native writer of Jiangsu Province.They have lots of similarities on the age, regional culture, or on the writing subject. At present, the primary means of the studies about their works are introspective methods,besides a few simple quantitative analysis for their works.In this study,we introduced the Multi-Dimensional analysis to study the differences between Feiyu Bi and Tong Su's works, and we made some improvements in technology to make it suitable for comparative analysis of the similar style writers'works. MD analysis emphasizes the continuity and schematism in stylistic feature analyses, and the experimental results show that it can reflect differences between the two writers in detail. In this paper, the research and results can be briefly summarized as the following several parts:First, the establishment of corpus. We built three work corpus of the group of writers, Feiyu Bi and Tong Su, which have a total of more than 440 words. Then we use softwares to process and check the datas repeatedly, which lasts about 4 months to ensure the correctness of the corpus.Second, and the selection of linguistic features. According to the text characteristics and relevant researches,we chose 92 linguistic items,and use softwares to count the frequency of these linguistic items. In order to extract linguistic items that are of significance, we screened 50 linguistic features through the correlation analysis.These linguistic features are the basis of our quantitative analysis study.Third, the analysis linguistic style of Feiyu Bi and Tong Su under the framework of MD analysis. We improve the MD analysis, and analyze the linguistic style of the two writers with groups with the corpus of the group of writers as reference corpus. The results are as follows :(1) on the meaning expression and lexical richness ,Tong Su is higher than Feiyu Bi and the group of writers,and the group of writers is a little higher than Feiyu Bi; (2) on the intensity of mood and narrativity ,Tong Su is much higher than Feiyu Bi and a little higher than the group of writers; (3) on the stability of narration, Feiyu Bi is higher than the group of writers and Tong Su ,and the group of writers is a little higher than Tong Su.Forth, the analysis linguistic style of Feiyu Bi and Tong Su under the Computational Stylistics analysis. We analyze the linguistic items that are of great difference in the two writers through the Computational Stylistics analysis. Our results verify the introspective research results: (1) on the narrative rhythm of texts, Feiyu Bi's works are relaxative,while Tong Su's works are urgent; (2) on the content,the two writers are good at account of the female characters; (3) there are a large amount of free direct speech in Tong Su 's works, which is the "defamiliarization" of punctuation.In conclusion, this paper used computational stylistics and MD analysis methods,and make studies on the work style of the two writers. We can find that the two writers have some differences in linguistic style. And the results demonstrates that the quantitative analysis methods have a good effect on analyzing author style. And this paper introduced the MD analysis in the study of linguistic style ,which provides the theoretical support for the quantitative research in writer writing style researches. |
PDF Full Text Request