Branding brexit. A big data textual approach

Political branding responds to the need of parties to effectively differentiate themselves from others and to produce suitable messages in a roster of digital media, by drawing on data-driven insights (Susila et al., 2019; Jungherr et al., 2020). For t h i s reason, understanding how political brands communicate and engage with audiences continues to be a core area of research which calls for more insights into the use of artificial intelligence (AI) in political marketing (Falkowski & Jablonska, 2019). In this chapter, we explore the impact of digital media on political branding by pursuing a big data approach (O’Halloran et al., 2022- in press), involving the analysis of linguistic resources in political parties’ official communication on Twitter. The approach involves the integration of Critical Discursive approaches (e.g. van Leeuwen, 2005; Wodak & Meyer, 2016) with computational tools and AI, by building on the capabilities of Multimodal Analysis Platform (MAP), a cloud-based platform for searching, storing, and analyzing online media texts (e.g. social media, news media and websites) (O’Halloran et al., 2021). Our multidisciplinary approach combining quantitative and qualitative analyses is demonstrated through a case-study that investigates how the Conservative Party and the Labour Party in the UK branded Brexit during the debate over the agreement and the implementation of a ‘deal’ with the European Union. We highlight key discursive patterns through which the two parties strategically differentiated themselves from each other while aligning Brexit with specific ideological visions. The study also highlights possible ways forward for large-scale analysis, combining semiotic and computational approaches.