Fostering Civil Discourse Online: Linguistic Behavior in Comments of #MeToo Articles across Political Perspectives

Linguistic style and affect shape how users perceive and assess political content on social media. Using linguistic methods to compare political discourse on far-left, mainstream and alt-right news articles covering the #MeToo movement, we reveal rhetorical similarities and differences in commenting behavior across the political spectrum. We employed natural language processing techniques and qualitative methods on a corpus of approximately 30,000 Facebook comments from three politically distinct news publishers. Our findings show that commenting behavior reflects how social movements are framed and understood within a particular political orientation. Surprisingly, these data reveal that the structural patterns of discourse among commenters from the two alternative news sites are similar in terms of their relationship to those from the mainstream - exhibiting polarization, generalization, and othering of perspectives in political conversation. These data have implications for understanding the possibility for civil discourse in online venues and the role of commenting behavior in polarizing media sources in undermining such discourse.