Algorithmic rhetoric

With the Brexit political drama and the Trump presidency, analysts and journalists alike have accounted for a transformation of the common sense and the dealignment of the political landscape by focusing on the use of language, questioning the limits of acceptability and the effectiveness of rhetoric. In parallel to this shifting political reality, rhetorical analysis has been flourishing in political studies: for instance, studies have focused on the ritualistic character of speech moments in the British political calendar, on the use of anecdotes or on the use of melodrama in conflictual rhetoric. We now have useful tools to study changes in rhetorical culture, however what is often missing in this debate is the role of technology in transforming speeches. What does social media do to rhetoric?

In my recent article in Politics, I examine in detail the role played by social media platforms and their algorithms in the production and delivery of political speeches. Using the recent literature on critical algorithms studies, I develop a new approach in rhetorical political analysis. I demonstrate that by ignoring the means of communication, we continue to adopt an outdated conception of technology as a set of simple tools used by orators to get their ideas across. On the contrary, social media platforms and their algorithms condition our experience of political speeches as well as the design of speech situations. More specifically in my article, I break down this influence by turning to four conditionalities of algorithms on rhetoric:

  1. Programmed speech content: social media platforms favour specific content and influencers have trained themselves to get maximum exposure (by ‘gaming’ the algorithms) by adapting both the form and the content of their posts. Online celebrities are now using politics and the so-called ‘culture wars’ to make a living, becoming ‘rhetorical entrepreneurs’ (through donations on websites like Patreon).
  2. The verticalisation of political communication: I show the processes by which certain political speeches get prioritised and stay on top of news feeds. Rhetoric is not an even playing field and mediators categorise and filter the content presented to users. I have called this process a verticalisation of speech to help us make sense of the multitude of operations that help certain speeches get seen and heard.
  3. Digital biases: when algorithms select content to be prioritised, they do so according to social norms, values and assumptions. Rhetoric is not immune from the digital biases that compose our technology-enhanced reality. I therefore conclude that this is one of the reasons why rhetoric is so white. A solution is not simply to fix the algorithms or ‘racist bots’ but to tackle the social reality of racism itself.
  4. The rhetorical machine learning: speechwriters and politicians are hiring Artificial Intelligence (AI) companies to help with campaigning. Computer scientists created a basic AI machine ‘Political Speech Generation’ to demonstrate the possible uses of machine learning in rhetoric. I show how this example can help us imagine the future effects of AI on the creativity of rhetoric.

In sum, I provide a new framework in rhetoric that integrates into it a technological analysis in the hope that it alerts us to the role of algorithms in speech interventions while contributing more widely to the growing debate on rhetoric and the media.

Benoit Dillet

Benoit Dillet

Benoit Dillet is a Lecturer in French Politics and Society at the University of Bath. His research focuses on the political theory of technology. His first monograph The Political Space of Art: The Dardenne Brothers, Arundhati Roy, Ai Weiwei and Burial (co-authored with Tara Puri) was published in 2016, and he is the author of numerous articles in Political Studies Review, Deleuze Studies, Parallax, and boundary 2.

Leave a Comment

Your email address will not be published. Required fields are marked *