Session Chair: Monika Fratczak
Presentation 1
Digitization and Polarization in Local Context: Contemporary Rural Talk Radio Stations in the US
Rebekah Larsen
Harvard University, United States of America
Talk radio has been an enduring player in the American media landscape for decades. US talk radio tends to be partisan, overwhelmingly conservative, and sometimes painted as alternative ‘infotainment’ on a dying medium. But on national and local levels, the talk radio industry continues to play an outsize role—especially in conservative politics. Rural talk radio stations are a significant and trusted source of news and information in the US, serving to contextualize national events and topics, as well as playing a role in local news, promoting local businesses, and organizing local politics. Drawing from interviews and content analysis from three rural stations in Utah, this paper sheds insight on how these stations are increasingly operating at the intersection of digital and broadcast developments. It situates their choices of programming and audience curation within the wider political economy of conservative news environments—from expanding syndication choices bolstered by podcasting to the rise of digital ‘newspapers’ owned by these local stations. This paper also explores implications around changing media regulation, misinformation spread, and increasing polarization—while pushing back against monolithic conceptions of conservative media.
Presentation 2
GUILTY BY ASSOCIATION? INTRODUCING GUIDED LABEL PROPAGATION FOR IDENTIFYING AFFINITIES IN LARGE INFORMATION SHARING NETWORKS ON SOCIAL MEDIA
Jakob Bæk Kristensen
Roskilde University, Denmark
Network analysis has become a popular approach for analyzing information dissemination and behavior on social media. This paper proposes the method guided label propagation which serves the purpose of detecting clusters of accounts with similar sharing behavior, comparable to general community detection algorithms, but focusing on clusters that pertain to categories, identities or affiliations of interest for a particular study. The general method proposed can theoretically be used for all types of information sharing behavior (retweeting, co-sharing hashtags, liking same content), however it is tested using networks based on social media accounts and their mutual sharing of URLs. The method is tested using a very small sample (27) of left-wing, right-wing and non-partisan alternative news outlets to explore the political orientation of additional accounts that have reshared them at least once (7995 accounts). Sharing a partisan news article only once is not sufficient evidence for determining the partisanship of a social media account. However, guided label propagation iteratively explores the associations between those that share many partisan news articles and other content they share in order to propagate partisan affiliations throughout the wider network. The high precision of the method is validated using manual examination of labelled accounts. The paper demonstrates a viable approach for network analysis of complex information sharing behavior in cases where a more focused way of determining the role of clusters, compared to the completely unsupervised approach of traditional community detection algorithms, is needed.
Presentation 3
Following Lenin and Stalin Through Instagram: Varieties of dissimulative play in left-revolutionary memes
Aleksi Knuutila(1), Jonne Arjoranta(2)
1: University of Helsinki, Finland; 2: University of Jyväskylä, Finland
Politigram (also Theorygram) is a loose subculture of pseudonymous meme accounts on Instagram, focused on debating political ideologies and theories using memes. Politigram’s memes often feature political positions that are either idiosyncratic (e.g. “Feminist-Monarcho-Primitivism”) or outside the so-called Overton window of publicly acceptable political ideologies (such as Leninism or Stalinism). While a growing amount of research addresses how online spaces such as 4chan produce and popularise memes featuring far-right ideology, memes that depict left-revolutionary themes are relatively understudied. Since some authors have begun to view anonymous or pseudonymous online spaces as inherently problematic, understanding the variety of practices and content in online spaces that focus on fringe politics is important. Our analysis follows memes that feature Lenin and Stalin through a large number of Instagram accounts posting memes. We concentrate on Lenin and Stalin because they represent fringe ideologies that remain indirectly relevant in contemporary political discourse. With a visual dataset from 45,000 accounts from 2021, we study the thematic variations, boundary work and political positioning performed by Lenin and Stalin memes. We describe how requirements of theoretical literacy, militancy as a communicative value and parody of conceptions of communism drive the discussion of fringe political positions on Politigram. We conclude by contrasting memetic logics in this context with established accounts of “dissimulative identity play” in anonymous online spaces.
Presentation 4
Unveiling Tiktok's Shadow: A Typology of White Nationalist Narratives as Eudaimonic Entertainment
Julia Niemann-Lenz, Katharina Kleinen-von Königslöw, Leia Kantenwein, Leonie Oelschlägel, Hannah Scherf
University of Hamburg, Germany
In today's social media-dominated landscape, identifying the manifestation of white nationalism on platforms such as TikTok is essential to understanding the spread of extremist content. This research examines the interplay between content and form in TikTok videos, revealing how creators exploit platform-specific features to propagate white nationalist narratives. Rooted in the Identity Approach, which encompasses Social Identity Theory and Self-Categorization Theory, the study elucidates how individuals, particularly on social media, define themselves within interchangeable group memberships, giving rise to in-group/out-group distinctions. Applying this framework to digital platforms highlights the role of content in signaling group membership and shaping collective identity.
By analyzing white nationalist content on TikTok through qualitative content analysis (n = 75), the study identifies seven types, including nationalist militarism, traditionalism, AfD edit, stylized political speech/news video, infographic/documentation, demo footage and patriotic self-presentation. Each type uses different strategies, from glorifying militarism to repackaging political figures for TikTok. The typology sheds light on extremists' strategies for narrative dissemination on TikTok, and offers insights for platform moderation and interventions to mitigate the spread of extremist content. In addition, the study examines the emotional appeal and recruitment strategies embedded in extremist narratives and explores the influence of eudaimonic entertainment. Methodological challenges, ethical considerations, and barriers to research, are discussed, providing valuable insights for future research on online extremism.
