Fragmented but coherent: Lexical cohesion on a YouTube channel

• Lexical cohesion on YouTube is multimodal, cross-modal and intertextual. Cohesive ties exist between the video comments as well comments. A discussion may appear fragmented but still show dense cohesion. Qualitative corpus methods reveal how fragmentation a feature surface only. This paper investigates use functions of lexical within across modes levels interaction channel, arguing that contributes to coherence by establishing links video, sources elsewhere internet in intertextual manner. The analysis focuses one its section, investigating maintaining unity lengthy initiated continued are examined with regard their process meaning-making this mediated multimodal context. results cohesive extensive interaction. Chains created comments, video. Despite complex multimodality many participants post short after another, shown be coherent.