ABSTRACT
As economic crises erupted throughout Europe post-2008, the discursive emphasis on economic responsibility and policies of austerity took centre stage in discussions about the future of the European Union and countries within it. In this piece, we analyse the AuBriN corpus using a novel approach to word embedding models with paragraph concordances as a hermeneutically grounded methodology. We demonstrate the utility of this approach through a comparison of conceptions of crisis presented in The Guardian and The Daily Telegraph and further examine how these conceptions are attributed to individual EU member countries in each news source differently. We find that The Guardian tends to conceptualise crisis as a disease or contagion as it draws distinctions between bailout states and countries like Germany that fund the bailouts, while also emphasising the inevitability of the potential fallout to the UK. In contrast, The Daily Telegraph depicts crisis as an extreme disaster, especially for the UK, and connects reported violent events to the violent effects of the downturn through metaphor. We show through our analysis how movements between close and distant readings of texts can be used to mitigate the weaknesses of each and further suggest that this approach offers new possibilities for the ways we incorporate machine learning algorithms into discourse analysis.
