余华和莫言都是当代文学的代表人物。他们出生年代相近：童年处于艰辛的五六十年代，青壮年处于剧变的六七十年代，因此大部分作品背景相同、主题相近同时又不失个人特色。本文希望利用计量风格学的方法，对余华和莫言的共10部长篇小说的语言风格进行了更加客观的对比和分析。本文选取10部小说中的段落、句子、分句的长度及词汇频率作为特征量，进行均值、方差、分布列、假设检验、文本聚类等多方面的统计分析，并结合具体文本定性分析数据背后所代表的作品风格，力求得到真实全面、客观可靠的结论。运用n元文法和文本聚类对这些特征进行整体的考查时，发现在1-4的n元字符、n元标点、n元分句长和断句标点、n元词性、1-2的n元词汇以及颜色词上，余华和莫言的10部长篇小说按作者聚成两类，说明这些都是能够区分两个作者风格的特征。具体的观察这些特征，在长度方面，余华的平均分句长和句长比莫言大，但平均段长比莫言小，而莫言却偏爱超长的分句、句子和段落，他更长的平均段长是段落中更多的句子导致的。更多的工作集中在词汇方面，首先通过假设检验的方式对余华和莫言的词表进行分类，再通过寻找近义词对的方式比较两人的语言风格倾向。在标点上，莫言比余华使用了更多的括号、顿号、感叹号、破折号，表现出更丰富的信息量与修辞；在虚词和代词、时间词上，莫言表现出更加文言化、书面化和乡土化的语言倾向；在名词、动词和形容词上，莫言具有更丰富的用词及这种丰富中表现出的更复杂的世界、更激烈的情感；在量词和拟声词上，莫言远大于余华的类符表现出前者更大的词汇选择空间。另外，还选择颜色这一语义场作为对象，研究了两位作者不同构词和搭配的情况，发现了莫言在颜色词上更加多样的构词、更多的超常搭配和连用情况，及以此构筑的魔幻文本世界和带给读者的强烈感官刺激。相比于莫言个人特色的浓烈，余华的文本风格则相对更加口语、现代、冷静一些。一方面，正是这种冷静使两人同样具有的戏谑风格凸显出讽刺的感觉；另一方面，这也是文本风格学研究的重要特征：风格只有在对比中才能体现。
Yu Hua and Mo Yan are both representatives of contemporary literature. These two authors were born in a similar historical period: they had their childhood in the hard 50s and 60s and were both young adults in the booming 70s and 80s. Thus, their novels share the same background and similar themes, while preserving unique personal characteristics. The present thesis sets to investigate and compare Yu Hua and Mo Yan’s literary style, analyzing 10 of their novels using quantitative methods.I have selected length of paragraphs, sentences, clauses and the frequency of words as features to carry out statistical analyses such as mean, variance, distribution, hypothesis testing and text clustering. Related examples are always provided in order to explain the different styles hidden behind these data, in an attempt to make the conclusion objective, comprehensive, reliable and evidence-based. When using n-gram and text clustering to conduct overall examinations of these features, Yu Hua and Mo Yan's novels have been first grouped into two categories according to their authorship, and then further analyzed on the basis of n-gram characters, n-gram words, n-gram punctuation, clause-level punctuation, n-gram parts of speech and color words. This means that these features are treated as distinctive characteristics of these two authors’ literary style.When specifically observing these features, with respect to length, Yu Hua's average clause and sentence length is greater than Mo Yan’s, while his average paragraph length is smaller. However, Mo Yan tends to prefer long clauses, sentences and paragraphs. His longer paragraphs are due to the greater number of sentences found in each paragraph. A large part of the present thesis focuses on vocabulary. Firstly, hypothesis test is used to classify Yu Hua and Mo Yan’s word lists; secondly, synonyms are used to compare their text styles. As for punctuation, Mo uses more brackets, enumeration commas, exclamatory marks and dashes than Yu, showing richer content information and a wider use of rhetorical devices. As for function words, pronouns and time words, Mo’s language is more classical and closer to written style, and makes as well use of local terms and regionalisms. As for nouns, verbs and adjectives, Mo’s lexical choices are richer and more varied, and thus able to depict a more complex world and stronger emotions. As for measure words and onomatopoeic word, Mo’s vocabulary range of choices is also extremely wider than Yu’s. As for color words, Mo is more experimental, he plays with diverse forms of word formation and tries new, innovative collocations: he is able to build an magic-realism literary world and to stir up strong feelings and sensory perceptions in his readers.Compared to Mo's strong personal characteristics, Yu's language is more colloquial, modern and detached, almost cold. On the one hand, Yu’s ‘detached’ style conveys a sense of irony thus makes his novels distinguishable from Mo’s absurd; on the other hand, this shows an important characteristic of quantitative stylistics: style can only be fully reflected through comparison and contrast.
