Online intergroup conflict: How the dynamics of online communication drive extremism and violence between groups

Thesis

In the early days of the Internet, the social web was suggested to hold the power to reduce conflict, prejudice, and discrimination, by promoting connection between those with diverse backgrounds and experiences. However, this optimistic view has not materialised. Instead, evidence shows that Western societies are increasingly polarised and divided, and social media has been particularly implicated with the promotion of violent extremism and the spread of hate speech. The exact role social media plays in driving these processes remains poorly understood. This thesis investigates the mechanisms through which social media may foster and exacerbate intergroup conflict both online and offline, and explores different intergroup and intragroup dynamics which may drive this trend. This work firstly investigates the prevalence and impact of online communication between opposing groups on offline intergroup relations, demonstrating how intergroup communication does frequently occur online, but can be confrontational, hostile, and fails to improve intergroup relations, instead predicting future offline violence. Following this, we then develop a novel approach to detect online hate speech by leveraging datasets from multiple social media platforms in order to train natural language machine learning classifiers. We use these classifiers to explore the influence of extreme outgroup denigration in ingroup discussions on fringe social media platforms popular with the far-right. We demonstrate how exposure to online hate plays a role in shaping individual radicalisation trajectories, increasing the likelihood of offline hate crimes, and driving mutual radicalisation with opposing groups. Finally, we explore the effect of hostile manipulation from state actors on the dynamics of online conversations, presenting novel techniques to reveal how manipulation of these discussions increases the polarisation of online conversations from genuine users. Together these findings shed light on the mechanisms by which the Internet promotes intergroup conflict, extremism, and violence, providing new insight into initial steps that could be used to counter these effects.

Authors: Gallacher, JD

• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Social Media; Social Psychology; Internet; Extremism; Polarisation; Intergroup Conflict
• Subjects: Radicalism; Terrorism; Psychology; Conflict (Psychology); Internet