Online conspiracy theories: Politecnico di Milano research can predict which users will ...
eurekalert.org
A 500-million-comment Reddit study reveals that conspiracy believers betray themselves through language alone — and a new model can spot them before they ever post a conspiracy theory.
Radicalization TheoryLinguistic RelativityFilter Bubble TheorySocial Contagion Theory
Theory Briefing
- Politecnico di Milano analyzed over 500 million Reddit comments to identify linguistic fingerprints of conspiracy-community members.
- Users in conspiracy groups use distinct language patterns even in unrelated subreddits, making their beliefs detectable before they spread misinformation.
- The predictive model challenges filter-bubble theory by showing radicalization leaves traceable cognitive traces across all online behavior, not just echo chambers.