ABSTRACT
Critical discourse analysis (CDA) is an interdisciplinary research movement interested in examining how language use in discourse relates to power relations, particularly with a focus on social inequality and social change. This chapter explores how studies which take a CDA perspective can benefit from approaches associated with the digital humanities by outlining corpus-assisted CDA. The chapter begins by briefly describing CDA along with its aims and methods, going on to explore how it can be utilised on large collections of digital texts or corpora. We then describe a case study which involves the analysis of two related corpora taken from the website of the British newspaper the Daily Express. Many studies in corpus-assisted CDA have analysed large collections of newspaper data, and in our study we consider how Express journalists represented Romanians in the decade leading up to the point where British people voted to leave the European Union. CDA not only considers the language within texts though but requires analysis to consider different levels of context. For this reason, we also consider a form of reception surrounding the news articles by collecting and analysing a second corpus consisting of readers’ online comments to the same articles in the corpus. The creation of a corpus of online comments has resonance for studies in digital humanities and we describe the web scraping procedure that was used to acquire this data. We then use a technique within corpus linguistics called keywords in order to identify words which occur statistically more often than expected in each of the two corpora, which acts as the basis for funnelling the analysis around a smaller set of salient words. Then an additional technique employs the tool ProtAnt in order to identify the most linguistically prototypical articles in the news corpus, allowing for a close qualitative analysis of a smaller, but principled, sample of the dataset.
