Battles for Britain : Exploring Drivers of Political Tribalism in the Wake of Brexit

: (Alternative Format Thesis)

Abstract

Tribalism is an innate human trait, but one that can be especially harmful in political settings, because it can lead to polarised societies unable to compromise on issues affecting the whole. In this thesis, I explored how individuals exhibit tribal behaviour online in discussions related to Brexit, specifically focusing on negative language and interactions between Pro-Leave and Pro-Remain sides. Across the three individual studies in this thesis, I drew on several classic theories, namely social identity theory, reactance theory and media framing, and used them to guide a computational analysis of big data sets extracted from social media (Twitter). In the first study, I confirmed and characterised the presence of Brexit-related online tribalism across 32 months of data from Twitter. I then performed anomaly detection to identify abnormal increases and decreases in the levels of tribal behaviour on Twitter over time. Based on this data, I then mapped these anomaly points to specific real life topics using hashtag analysis. In this way, I was able to demonstrate links between real life events in the Brexit timeline and subsequent rises in tribal behaviour online. In the second study, I used thematic analysis and text analysis to compare similarities and differences in use patterns of tribal partisan language around Brexit, in both UK mainstream media and Twitter. Finally, the COVID-19 pandemic provided a unique opportunity to understand how a polarised society responds to a major threat. For the final study, I conducted two surveys across a sample of the UK population, aiming to understand how an individual’s propensity to psychological reactance affects both the response to Brexit and to COVID-19. Overall, this thesis demonstrates that online political tribalism is a real and present problem in the UK context, for example in terms of compromising the population’s likelihood of following urgent public health measures.

Date of Award	24 Jul 2024
Original language	English
Awarding Institution	University of Bath
Supervisor	Lukasz Piwek (Supervisor) & Adam Joinson (Supervisor)