Mapping the Young Far-Right: Network Analysis of Discourse Diffusion on Discord

• In Event: Collective Behavior and Social Movements Section Roundtables
  In Refereed Roundtable (60 minute): Table 7: The Role of Digital Platforms in the Practice and Study of Movements: The Case of the U.S.

Abstract

This study investigates how Discord’s server-to-server network structure shapes the diffusion of varying types of far-right discourse. The generational shift in far-right politics suggests that young extremists increasingly rely on semi-bounded, lightly moderated social platforms for radicalization, such as Discord. Despite its popularity, Discord remains understudied compared to mainstream social media platforms like Reddit. Using an analytic corpus of 12.7 million messages across 72 interconnected servers, we perform network analysis, topic modeling, and fixed-effects regression to examine how structural position (i.e., brokerage) and the type of discourse (minority-related or politics topics vs. others) moderate information flow through the network structure. Our findings provide empirical support for the “complex contagion” hypothesis. Brokers—servers that bridge distant, otherwise isolated communities—amplify general subcultural discourse like gaming and anime but gatekeep more sensitive political topics. Conversely, tightly clustered servers are the primary drivers for the diffusion of politically charged discourse, due to the cohesive social affirmation these echo chambers can provide. Finally, we also find that broker servers initiate minority-related topics proactively, presumably to appeal to the white supremacist sentiment overarching the network. This study suggests that brokers aim to elicit general engagement to maintain their brokerage status, while closed clusters tend to intensify sensitive political ideologies.

Authors

• Youngjin YJ Chae, Rutgers University-New Brunswick

• Alessandro Giuseppe Drago, University of Notre Dame