Revealing Human and Machine Translation Differences Through Annotation

This study aims to provide an annotated translation of figures of speech in the short story “You Perfect, Broken Thing” by C.L. Clark in Uncanny Magazine Issue Thirty-Two and to examine the translation strategies applied in human and machine translation. The data was analyzed using a descriptive qualitative and quantitative method. The main theory applied is Chesterman’s (2016) translation strategies, focusing on syntactic and semantic strategies. A total of 61 data are categorized by type of figure of speech: 30 metaphors, 14 idioms, and 17 paradoxical sentences were identified along with their translation processes. The result shows that human translation applies semantic strategies in all of metaphors, idioms, and paradoxical sentences, and syntactic strategies in metaphors (33,3%), idioms (64,2%), and paradoxical sentences (64,7%). Meanwhile, machine translation applies semantic strategies in metaphors (66,7%), idioms (66,3%), and paradoxical sentences (35,3%), and syntactic strategies in metaphors (66,7%), idioms (35,7%), and paradoxical sentences (100%). This shows that human translation is more dominant in applying semantic strategies (100%) that prioritize the translation in contextual meaning. In contrast, machine translation is more dominant in applying syntactic strategies (68.85%) that tend to keep the source text’s structure.