Lexico-Semantic Analysis of Khayyam's Quatrains: A Comparative Study of Machine Translation and Human Translation

This paper presents a comparative and descriptive analysis aimed at evaluating lexico-semantic translatability errors in Khayyam's Quatrains (1048-1131) when translated using the Persian online translation tool Targoman.Fitzgerald's (1859) human translation of the Rubaiyat serves as a reference point for assessing the quality of translation.The corpus for this study includes 30 quatrains from two translations (outputs from the Targoman Machine Translation & Edward Fitzgerald's translation, which is a human translation) of selected Persian classical poetry in the quatrain genre, authored by Omar Khayyam (1053Khayyam ( -1123)).These quatrains were chosen through a random sampling method from Khayyam's Rubaiyat.The study employs both comparative and descriptive analytical approaches to identify and evaluate the occurrence of lexico-semantic errors across different translation types.Using Liao's (2010) model, the most frequent errors in Targoman translations were Rendition Errors, especially issues with collocations, occurring 48 times.Errors followed this in translating phrasal verbs, which appeared 35 times.The least frequent errors, according to this model, involved the translation of proverbs (8 occurrences) and linguistic errors (1 occurrence).The purpose of this study is to assess the accuracy and effectiveness of Targoman's machine-generated translations of Omar Khayyam's quatrains by comparing them with Fitzgerald's human translations.The research aims to examine how well machine translation preserves the linguistic, semantic, and poetic nuances of the original text and to identify common errors that could affect the accuracy of meaning in translation.Additionally, the study concludes by emphasizing the need for further research into machine translation quality, highlighting its current limitations, and advocating for the use of more effective methods and strategies to address these issues.