Persuasion in the digital age: a theoretical model of persuasion in terse text

This thesis explores how the increasingly prevalent terse text format of Social Media communication has affected the way we seek to persuade one another and whether it has impacted the applicability of existing models of persuasion, influence and attitude change. Over the past few decades, communication behaviour has evolved dramatically. As a society we increasingly consume information in the format of short messages, rather than lengthy text and verbose speech. Meanwhile our understanding of persuasion has hardly moved on from the 1980’s and continues to be spread across a variety of academic disciplines, such as Behavioural Science/Psychology, Philosophy/Rhetoric, and various sub-fields of linguistics. Existing models of persuasion are to date lacking interdisciplinarity and applicability to the terse text format found in Social Media. The data used in this research is in the format of Twitter microblogs gathered throughout a number of recent political campaigns, such as the 2016 UK Brexit referendum and the 2016 US General Election. The research purpose is fundamental, rather than applied, meaning that it seeks to expand knowledge by increasing the understanding of fundamental principles, rather than answering specific questions and offering a precise solution to a practical problem. The research philosophy that has been adopted for this project is interpretivism. The research approach is idiographic, and the methodology is predominantly qualitative, with occasional use of descriptive statistics. The research was conducted in several distinct phases, starting with the construction of the theoretical model, followed by two validation exercises and further experimental exploration by means of a recall test and computational linguistic analysis, culminating in a revised model of terse text persuasion. This research draws upon and collates existing knowledge from behavioural science, rhetoric, linguistics, and cognitive science and develops a comprehensive understanding of how we seek to persuade through terse text media, based on data collected around a number of recent political campaigns and topics of debate. The research demonstrates that existing models of persuasion, such as the Elaboration Likelihood Model (Petty and Cacioppo, 1986) and the Heuristic Systematic Model (Chaiken et al., 1989)cannot be applied to the terse text context without significant modification. A new theoretical model of persuasion in terse text is proposed and evaluated. The findings also show that there is a distinct preference for heuristic over systematic cues in terse text messages with persuasive intent, and – in terms of Aristotelian rhetorical appeals – a preference for appeals to credibility (ethos) and emotion (pathos) over appeals to reason (logos). Additionally, the research explores, by means of a recall test, the most memorable subcategories of terse text microblogs, as well as the examining message structure and features through computational linguistic tools. Although this research focusses on political persuasion in terse text Social Media, the findings have implications that reach far beyond the political sphere into activism, marketing, social engineering, strategic communication and (human centred) information warfare