1 Introduction

This article reviews the process of translating Shakespeare's sonnets into Chinese, from the first translation by Shen Mo in 1929 to Zhu Xiang's selection of translations in the 1930s, and then to the addition of translators such as Liang Zongdai and Fang Ping, demonstrating different translators' understanding and reproduction of the original work. The evaluation of poetry translation usually adopts qualitative or qualitative research methods, such as Li Zhengshuan and Wang Xin comparing the translations of Tu An and Gu Zhengkun from a cultural perspective, Lu Wei analyzing multiple translations from the perspectives of rhyme, rhythm, and ideological communication, and Huang Bikang emphasizing the poetry and philosophical thinking of translations by Liang Zongdai, Tu An, and others. After the reform and opening up, translators such as Gu Zhengkun and Cao Minglun pursued formal similarity while remaining faithful to the original poetic conception. With the development of corpus linguistics, quantitative research methods have been introduced into translation studies, such as Liu Zequan's style comparison of the English translation of "Dream of the Red Chamber", Yuan Yanling and Ge Lingling's analysis of the humor mechanism of the English translation of "Two Horses", Qin Ying and Wen Qiufang's automatic evaluation system for English Chinese translation, and Wang Jinquan's classification of translation quality evaluation methods. Pan Xiaxing and others compared the original and translated versions of Shakespeare's sonnets from the perspectives of vocabulary, word frequency, and part of speech distribution. Luo Yimin and Liu Jian compared the translation styles of Liang Zongdai and Gu Zhengkun's translations based on corpus methods. This study attempts to explore a quantitative evaluation method for poetry translation based on word frequency distribution indicators and its effectiveness.

2 Research Methods

Detailed research methods were elaborated, including text processing, calculation and analysis of word frequency distribution indicators. Firstly, the researchers included Shakespeare's 154 sonnets and their five Chinese translations for analysis, and used SEGTag software to automatically segment the Chinese translations. By counting the number of words, it was found that the average number of words in the translated text is generally higher than that in the original text. Next, frequency distribution indicators such as h-point, a-point, R1, k-point, b-point, R2, m-point, and R3 were introduced, which are used to measure the usage of function words and content words, the ratio of commonly used words to non commonly used words, and the distribution of personality words and common words in text. The researchers also proposed a modified formula to eliminate the influence of text length on the indicators. Finally, the research question was clarified, which is to compare the similarities and differences in word frequency distribution indicators between the original and translated versions, and to conduct data analysis using methods such as analysis of variance, multidimensional scaling analysis, and cluster analysis.

3 Research Results

The research results show that there are significant differences in word frequency distribution indicators among Chinese translations of Shakespeare's sonnets by different translators. Specifically, the analysis of variance for indicators h, a, R1, k, b, R2, m, and R3 all showed significant differences among different translators. For example, the translations of Cao Minglun and Gu Zhengkun are significantly lower than other translations in indicators h, a, R2, and m, while the Tu An translation is significantly lower than the original text in indicators R1 and R3. Multidimensional scaling analysis shows that there are certain differences in word frequency distribution between the original text and other translations. The translations by Liang Zongdai, Liang Shiqiu, and Tu An have higher similarity to the original text, while the translations by Gu Zhengkun and Cao Minglun have lower similarity to the original text. Cluster analysis further confirms that the translations of Cao Minglun and Gu Zhengkun have significant differences from the original text, while the translations of Liang Shiqiu, Tu An, and Liang Zongdai have high similarity to the original text. These results provide a quantitative basis for understanding the translation styles and strategies of different translators.

4 Research conclusions

There is a significant difference in word frequency distribution between the original and translated versions of Shakespeare's sonnets, and analysis of variance shows that at least one text is significantly different from the others. The specific differences vary depending on the indicators, for example, indicator h shows no significant difference between the original text and the translation, while indicators R1 and b show significant differences between the original text and all translations. Multidimensional scaling analysis reveals that the original text is independent on a two-dimensional plane, and the translations of Liang Zongdai, Tu An, and Liang Shiqiu have a high degree of similarity with the original text.

5 Conclusion

Through analysis of variance, it was found that there are differences in word frequency distribution indicators between different translations and the original text; Multidimensional scaling and clustering analysis revealed that some translations have a high degree of similarity with the original text; The quantitative analysis method is consistent with the results of traditional qualitative research.