| Summary: | 碩士 === 國立高雄第一科技大學 === 應用英語系口筆譯碩士班 === 104 === This thesis aims at illustrating the difference between translated and native Chinese and the applicability of Zipf’s law to these two types of Chinese texts with the use of corpus-based translation studies (CTS).
One translated Chinese corpus consisting of texts obtained from the Chinese website of National Geographic Magazine and two native Chinese corpora developed by the Taiwanese Ministry of Education are used in this study in an attempt to reveal how translated Chinese differs from native Chinese in terms of word frequency distribution and Zipf’s law. Not only is the word frequency distribution at the character level analyzed, but this thesis also segments the translated texts into meaningful terms for analysis by applying two divergent segmenting guidelines provided by the two native Chinese corpora.
The findings show that the word frequency distribution of translated Chinese is identical to native Chinese, but does not exhibit Zipf’s law at the character level. At the term level, it is found that the application of the two segmenting guidelines results in divergent curves of word frequency distribution. The guideline that considers certain characters as a ‘term’ can generate curves that are not only similar to each other but comply with Zipf’s law, while the guideline that excludes individual characters from the extraction of terms only partially does. This indicates that the guideline used to segment the Chinese texts may affect the presentation of data.
|
|---|
