Dresden 2026 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

SOE: Fachverband Physik sozio-ökonomischer Systeme

SOE 11: Polarization

SOE 11.5: Talk

Wednesday, March 11, 2026, 17:00–17:15, GÖR/0226

The role of antagonization in political discourse on social media — •Armin Pournaki and Eckehard Olbrich — Max Planck Institute for Mathematics in the Sciences, Leipzig, Germany

Antagonization is a central discursive strategy in political narratives. The creation of an enemy allows for political mobilization through powerful affective mechanisms. It is commonly assumed that social media, through a combination of platform affordances and algorithmic amplification, reward and therefore enhance antagonizing content. However, it is an open empirical question whether the prevalence and nature of antagonization as a discursive strategy depends on the political leaning, or whether different political camps antagonize equally.
The present work aims to address those questions by presenting a novel computational approach to systematically extract traces of antagonization from text, including potential targets of antagonization. On a corpus of 1M tweets from the German Twittersphere, we observe a strong prevalence of antagonizing content, confirming common hypotheses by which social media tend to facilitate such discourse. By combining the method with a large-scale estimation of ideological positions of users based on their retweet behaviour, we show that antagonization is significantly more prevalent in the online discourse of right-leaning than of left-leaning actors. Finally, a systematic analysis of the various antagonized targets provides novel insights into how different political factions employ this strategy to mobilize their audience.

Keywords: networks; natural language processing; populism; polarization