Akurasi Penerjemahan Lirik Lagu "Qolbi Fil Madinah" Oleh Maher Zain & Haris J Melalui Penerjemahan Chat Gpt

Abstract

This study aims to analyse the accuracy of the translation of the lyrics of the song ‘Qolbi Fil Madinah’ by Maher Zain and Haris J using ChatGPT. The type of research is descriptive qualitative with a translation accuracy approach model from the perspective of Nababan et al. Data collection techniques were carried out by recording and translating song lyrics into the ChatGPT application. Meanwhile, the data analysis technique used Miles and Huberman's interactive model, namely data condensation, data presentation, and conclusion drawing. The results showed that 9 lyrics (64.3%) were accurate in translating doxological sentences, but 5 lyrics (35.7%) showed distortion of meaning, especially in Sufi metaphors. Three dominant error patterns were identified: (1) pragmatic errors (57.1%) involving a reduction in spiritual meaning, (2) lexical errors (28.6%), and (3) grammatical errors (14.3%) in poetic structure. A critical finding is ChatGPT's failure to maintain the semiotic hierarchy of Sufi texts, particularly in translating cultural symbols. This study concludes that ChatGPT is still unreliable for translating the nuances of contemporary Sufi lyrics, thus requiring post-translation editing.