Bounded by Hate: How X’s ‘Communities’ Feature Curates Paranoid Misogyny in Incel Spaces

Digital platforms' algorithmic curation and recommendation systems shape how illiberal communities form and organize. This paper examines how X's (formerly Twitter) Communities feature, revamped in May 2024 to algorithmically suggest topic-based groups, facilitates community-building among users promoting illiberal ideologies. Through a case study of the IncelTalk Community, an extension of the incels.is forum operating on X, we analyze how platform affordances enable the collective construction of extremist discourse in mainstream social media spaces. We collected 17,030 posts from IncelTalk between June 15 and September 15, 2025, including 6,032 native posts and 10,998 replies. Using BERT topic modeling, social network analysis, and critical discourse analysis, we identify core thematic patterns and trace how influential users curate engagement through reply threads. Our findings reveal that users collaboratively develop and reinforce conspiratorial narratives about gender while strategically moderating their language to minimize perceived extremity. Key themes include biological determinism regarding attractiveness, collective victimhood narratives, and the normalization of misogynistic worldviews through repeated rhetorical patterns. This research contributes to understanding how algorithmic community recommendation features on major platforms facilitate the organization and mainstreaming of illiberal movements. It highlights a critically overlooked aspect of content organization on one of the world's largest social media platforms with implications for platform governance, content moderation policy, and the study of online radicalization.