Abstract This paper investigates coherence and cohesion in multimodal translation, with a particular focus on audio description (AD) as a representative case. While coherence and cohesion have been extensively studied in interlingual translation, less attention has been paid to their role in multimodal contexts, where meaning emerges from the interplay of linguistic, visual, and auditory modes. Drawing on theories of textual cohesion and multimodal discourse analysis, the study situates AD as a form of multimodal translation that must balance internal textual cohesion with cross-modal coherence. In audio description, cohesion primarily serves to create textual continuity through devices like reference and conjunction. Coherence, in contrast, is built upon a framework of narrative logic, clear temporal-spatial sequencing, and the strategic integration of dialogue and sound effects. The analysis highlights that cohesion secures the internal flow of the AD text, whereas coherence emerges from multimodal alignment and audience cognitive processing. These findings suggest that coherence in multimodal translation is distributed across semiotic resources and must be addressed holistically. The study concludes by emphasizing the theoretical and pedagogical implications of treating coherence as a central concern in multimodal translation.
Jie Lü (2025) studied this question.