3er Seminario Internacional sobre el Discurso en Twitter

Matteo Fuoli

Associate Professor of Corpus-based Discourse Analysis
Department of English Language and Linguistics

University of Birmingham, Great Britain

Mixing corpus and experimental methods to study polarized discourses on social media

Polarization, defined as the divergence of attitudes towards ideological extremes, is a major contemporary challenge as it engenders societal fragmentation, misinformation and radicalization, making it hard to reach consensus on solutions to pressing global issues such as climate change and the Israel-Palestine conflict. Polarization is exacerbated by social media, which simultaneously segregate people into ‘echo chambers’ and expose them to extreme opposite views. While there is now a substantial body of work examining polarization on social media, linguists have paid relatively little attention to this issue. In this talk, I will share two case studies that show how we can use a mix of corpus-assisted discourse analysis and experiments to investigate polarized discourses on social media. The first case study examines how Twitter users reacted to Elon Musk’s takeover of the social media platform in 2022. We use corpus techniques to identify the main divisive issues and linguistic markers of polarized attitudes. The second case study combines discourse analysis and experimental methods to study the features and effects of evaluative language in online political protest messages. Drawing on SFL’s Appraisal framework, we develop and test a model of how the linguistic framing of political protest messages influences the spread of blame on social media. Together, these two studies demonstrate the potential of a linguistic approach to understanding polarization. This approach provides fresh insights into the communication and interaction factors that drive polarization and can inform potential solutions to address this issue.