Antagonism: conveying and responding to negative evaluations in YouTube comment thread interactions

This thesis explores the problem of antagonism between ideologically opposed groups in online comment threads on YouTube from an interactional sociolinguistic perspective. The case study discourse under examination either supports or criticises evangelical Christian faith in the context of personal video narratives of religious exit and “deconversion”. The research aims to better understand how negative evaluations of interlocutors, social groups and ideas expressed in online interactions propel discussions towards antagonism and further polarisation of stances. This type of discourse is currently under-researched, particularly in terms of the functions performed by pragmalinguistic devices in negative evaluative language and escalation of antagonism. YouTube is a massive online platform with billions of users, many of whom actively engage in comments by posting evaluations of videos and aligning or disaligning with other users’ stances (Garcés-Conejos Blitvich 2010b). Impolite discourse is a socioculturally ‘situated phenomenon that is interactionally constructed’ and antagonism is common in YouTube comment sections (Garcés-Conejos Blitvich 2009, p.274; Pihlaja 2014a), making it a rich source of data for sociolinguistic analysis. Commenters employ linguistic, pragmatic and textual resources to construct evaluations which convey stance. In response, the current analysis casts a wide net in terms of pragmalinguistic and graphic elements of interpersonal communication online, as well as taking sociocultural influences into consideration. Pragmalinguistic aspects of online comments discourse are used to convey varying degrees of negative evaluation, by which commenters assess ideas, individuals, and social groups. Disaligned interactions are ‘mediated by the technological and situational characteristics’ (Haugh 2010, p.7) in the comment threads, as in other digitally mediated modes. Multimodality is a feature of the YouTube platform, so graphic elements such as punctuation, text type, and emoji also contribute to evaluations (Zappavigna 2018a). This study applies the APPRAISAL framework (Martin & White 2005) to classify nine categories of discoursal element with functions in negative evaluations: (1) insults; (2) directives; (3) adverbials and adjectives which upgrade or downgrade evaluations; (4) acronyms and initialisms; (5) capitalisation of words; (6) emoji and emoticons; (7) repetition across comments; (8) quoting of interlocutors; and (9) externally sourced information and arguments. The functions of these resources in conveying and provoking antagonistic stance are then analysed using a range of analytical approaches, as appropriate for each resource, to identify antagonism triggers, and then map patterns of escalating antagonism across comment threads. Finally, the findings are contextualised in relation to real-world social issues related to antagonism and polarisation online.