#Voteleave or #Strongerin: resonance and rhetoric in the EU referendum

Thesis

This thesis explores the persuasive function of tweets in the context of the 2016 European Union membership referendum. What makes a message resonate with its audience on social media? How can the message be expressed and delivered most effectively? What strategies do individuals and organisations use when they send a tweet and why?

To explain tweet success, or resonance, an interdisciplinary framework centred around rhetoric was established. It is comprised of nine different features: discoverability, organisational authority, message framing and structure, emotions, topical issues, resolution, questions, and playfulness.

A multi-method design is adopted to address the research questions from micro- and macroscopic perspectives. Approximately 9.8 million referendum-related tweets were collected from Twitter during the referendum period (15 April-22 June 2016). A random sample of 2,976 was hand-coded according to the framework. Semi-supervised algorithms were used to label the features in the full dataset, and multivariate regressions were computed to assess the influence of each feature. This quantitative analysis is complemented by interviews with 22 campaign leaders, journalists, politicians, digital marketing experts, and private individuals who tweeted actively during the referendum. Interviewees included the CEOs of Vote Leave and Stronger In, BuzzFeed’s Europe Editor, a Twitter board member, and a former MP and Minister of State for Europe. Finally, exceptional tweets and users were identified from the regression models and interview data and rigorously scrutinised.

The overarching conclusion is that in the fast-paced and high-volume landscape of social media, discoverability is more important for resonance than content or style. Increasing the likelihood that a message is seen matters more than what is said or how it is said, because the vast majority of tweets are not seen. As such, this thesis introduces a revised schema for the resonance of digital cultural objects premised upon algorithmic power, visual media, and organisational authority.

Authors: Lu, Y

• DOI: 10.5287/ora-e9vkvo5jp
• Type of award: DPhil
• Level of award: Doctoral
• Awarding institution: University of Oxford
• Language: English
• Keywords: Brexit; Digital rhetoric; EU referendum; Internet linguistics; Natural language processing; New media; Discourse analysis; Political campaigning; Twitter; Social media
• Subjects: Digital communication; Political communication; Media studies; Sociology of culture; Digital marketing; Information technology--Social aspects