Once a site of promise for democratizing mass communication, the
internet has also become a site of problematic information and polarized affect. Contrary to
claims that polarization is not necessarily encouraged by social media platforms; our
two-year, mixed-methods study of affect and narratives of race and national belonging in
social media discourses of the 2019 Canadian and 2020 U.S. federal elections reveals clearly
polarized collective political storytelling constructing conflicting meta-narratives marked
by a highly affective moralizing tone and clear binaries of us versus them and good versus
evil. Surprisingly, there is very little research that has drawn on either narrative
emotions analysis or melodrama to understand the kinds of polarization that take place
within social media platforms. This talk shares our finding; achieved through our innovative
approach to affective discourse analysis developed through iterative, grounded theoretical
qualitative study; that discourse communities formed according to social as well as
political identities construct these polarized meta-narratives in the genre of melodrama,
readily ensuring the emotional engagement of social media users through “sensationalism and
predictable plot lines of good battling evil, plots and characters that do not encourage
reflection, and refusal of nuance” (Loseke, 2018, p. 517).