Stylistic analysis using machine translation as a tool

This pilot study investigates the potential of automatic translation engines to be used as a tool for literary analysis. Two sample texts, one prose and one poetry, were chosen as representative of their genre and translated from English into one of 62 languages. Re-translated back into English, the resulting texts were compared with a stylistic analysis of the original. Many stylistic features were retained although more subtle features of the texts were lost in the translation process. The similarity index varied widely between languages – between 90% and 32.9% -with corresponding stylistic feature retention observed. The similarity index gives a good indication of the literary quality of the text, suggesting that this process may be useful as a preliminary “filtering” technique in deeper studies.