<![CDATA[This study investigates how cohesion is realized in English project videos produced by communication studies students within an English for Specific Purposes (ESP) context. While cohesion has traditionally been examined as a structural feature of written discourse, limited attention has been given to its role in spoken and multimodal communication. Adopting a qualitative discourse-analytic approach grounded in Halliday and Hasan’s framework, this study analyzes 13 student-generated self-introduction videos to identify patterns of grammatical and lexical cohesion and their functional roles in meaning-making. The analysis identified 90 cohesive instances across five dominant categories: personal reference, conjunction, lexical repetition, demonstrative reference, and collocation. The findings reveal that personal reference serves as the primary organizing mechanism, supporting identity construction and audience alignment, while lexical repetition functions as a strategic resource for thematic continuity. Conjunctions provide minimalist yet effective logical structuring, and demonstrative references frequently operate across both verbal (spoken language) and visual (on-screen gesture and imagery) channels, indicating the presence of multimodal cohesion. These patterns suggest that students employ cohesion not merely as a structural linguistic system but as an adaptive communicative strategy shaped by real-time production demands and digital discourse contexts. Based on this limited dataset, this study offers a preliminary contribution toward reframing cohesion as a functional, strategic, and multimodal resource rather than a purely structural feature, warranting further investigation with larger corpora. Pedagogically, the findings highlight the importance of integrating discourse-based and multimodal approaches in ESP instruction to support students’ communicative competence in digital environments.]]>
