A discourse analysis approach to understanding the characteristics of trolling in social media

Abstract

Computer-mediated communications (CMC) has introduced new ways in which people communicate with each other (Hardaker, 2010). The ease of communication, in a more anonymised social context brings a different set of communication behaviours and patterns, some of which can be negative, such as trolling. YouTube attracts high volumes of trolling (Moor, Heuvelman, & Verleur, 2010). The aims of this study into trolling on YouTube were to examine if and / or how users responded to trolling with further trolling, and if users that interjected in defence of the target found themselves to be the new target of trolling. A subset of trolling data generated by a video of a feminist was analysed using discourse analysis. Trolling did often lead to more trolling, but the trolls did not attack users who intervened with a counter argument. The most overt sexual comments generated more trolling by different users, thus providing further evidence between trolling and sexually aggressive language. The influence of anonymity on trolling presented opportunities for further study. The study provides insight into the patterns of trolling behaviour on YouTube. Whilst it is too limited to be generalised, the study offers a greater understanding into the characteristics of trolling on social media.