AFTER DEPLATFORMING: RETRACING CONTENT MODERATION EFFECTS ACROSS PLATFORMS AND A POST-AMERICAN WEB

Abstract

Half a decade ago, social media platforms were widely perceived as revolutionary devices for maximizing political expression around the world. By opening the floodgates to expression, however, the same platforms were also accused of opening the floodgates of hate – allowing, for example, the self-claimed “revolutionary” return of ideas, speech and actors long thought to be relegated to the dustbins of history. This panel examines a three-fold revolution, namely: populist revolutions (on the right) facilitated by agnostic content moderation philosophies; the internal revolutions that platform content moderation underwent to address the political violence of the former; and the adjustments that digital methods research needs to adopt to facilitate content moderation research in a “post-API” environment. The first paper of this panel examines how Twitter’s content moderation has undergone several arbitrary changes before reaching a form of “normative plasticity”, with reinforcement techniques such as demotion and other forms of conditional content obfuscation. The second paper looks at how, despite making profound changes to prevent furthering political violence during elections, Twitter, Facebook, YouTube and Instagram have tended to moderate the Brazilian elections in a dislocated fashion, turning a blind eye to Brazilian militaristic content and focusing instead on what it primarily moderates in a US context. Finally, the third paper offers a set of methods for empirical researchers to capture and study content moderation metadata over time. All three papers aim to contribute to attempts at archiving and studying speech moderation as a public good, in an international context.