Scientific Methods for Stylistics

• 作者: Willie van Peer
• 来源期刊: 语言与符号学研究
• 学科: 文学
• 年，卷（期）: 2015,（2）
• 页码范围: 151-160
• 页数: 10页
• 分类号: H05

摘要

The present article argues that stylistics as a discipline has made little progress over the past decades,and that one major reason for this is a severe lack of methodological consciousness(and knowledge).This demands the formulation of quite explicit hypotheses,opening the possibility of independent data analysis,often also opening new perspectives on unforeseen aspects of a problem.In doing so a point is reached where statistics becomes unavoidable for the analysis.Two kinds of statistics are needed:descriptive statistics(to detect patterns in the data,patterns that our daily intuitions are unable to uncover),and more importantly,inference statistics(to estimate in how far one can generalize one’s findings).Examples are given to show results that are reliable,interesting and unexpected.They show us which of our intuitions are right and which ones are not.In other words:they inform us about what we are and how we function as human beings.