Abstract
Stylistic fidelity remains a persistent challenge in translation, particularly when navigating between languages with differing rhetorical and cultural conventions. While much research has addressed semantic equivalence, less attention has been given to how stylistic features such as tone, register, and syntactic variation are preserved or altered during translation. This study investigates recurring stylistic discrepancies in translated texts across literary and non-literary genres. A qualitative-comparative method was applied, analyzing source texts and their translations in English and a target language, focusing on sentence structure, lexical choices, idiomatic expressions, and voice. The analysis revealed frequent shifts in modality, overuse of nominalization, and simplification of figurative language, particularly in literary translations. In non-literary texts, translators often adopted more formal or neutral styles, occasionally leading to loss of original tone or emphasis. These findings suggest that stylistic variation is not always systematically addressed in translation practice or training, resulting in a degree of homogenization in target texts. The study highlights the need for a more nuanced approach to stylistic sensitivity in translator education and quality assessment frameworks. Future work could expand into multilingual corpora to identify broader patterns and improve stylistic calibration in machine translation tools.