Computational cross-media research: tracing divergences between normative Dutch television and social media discourses on the ‘refugee crisis’ (2013-2018)

Abstract

This article examines how the ‘refugee crisis’, sparked by the arrival of refugees from the Syrian civil war and other conflicts around the world, was articulated across Dutch television news programs and social media between 2013 and 2018. This crisis has been described as a key catalyst of the radicalization of European political discourse. Crucially, it took shape during a period of profound transformation of the media landscape, in which mass media lost significant ground to social media as authoritative sources of truth and norms. The research focuses on the crucial but underexplored link between television and social media discourse, which is at the heart of contemporary European public debate. Using a combination of digital methods and NLP techniques, the article compares automatic speech recognition (ASR) transcripts of Dutch televised news on the refugee crisis with responses from publics on Facebook and Twitter. This computational cross-media approach enables a longitudinal analysis of how social media users differ in their interpretation of key events characterizing the crisis, as well as what language is acceptable to debate issues around integration, tolerance and identity. A rejection of mainstream news media editorial guidelines by social media users eventually resulted in their consumption of populist right-wing (‘alternative’) news media and active transgression of anti-discriminatory speech norms.