Article Preview
Top1. Introduction
Citizens use social media platforms to be informed, share their viewpoints, and engage with others. Simultaneously, these platforms have come under growing scrutiny and pressure from the public to better regulate the use of social media through design. Three concerns, in particular, are hate speech and bullying on social media, false and misleading information, and the question of whether users should be encouraged to debate with those who hold opposing viewpoints (see, for example, Guiora & Park, 2017, on hate speech; Farkas & Schou, 2019, on post-truth and fake news; and Pariser, 2011, on filter bubbles). In response to these concerns, social media companies have implemented various new design features. For example, users are discouraged from bullying with questions such as ‘Are you sure you want to post this?’;Meta works with impartial fact checkers who review and rate content on Facebook, Instagram, and Whatsapp;1 users are referred to alternative sources; and Twitter has Community Notes2 – a community-based approach to addressing misleading information. These design solutions, however, are not uncontroversial. When social media platforms blocked the account of former President Donald Trump in response to the Capitol raid on January 6, 2021, newspaper headlines expressed concern that silencing a President in such a way might be undemocratic.3 This episode illustrates how complex the notion of democracy is, conveying, for example, both the values of safety and free speech and the range of questions to be answered when designing social media for democracy. Should we design for free speech, or is censoring speech – perhaps even users – at times justified and democratic? Should platforms ensure that all users have an equal voice online through design mechanisms? What functionality should we design recommendation algorithms for? In short, what design choices should engineers make if they are to build a (more) democratic social media?
The public concerns about social media platforms and related design choices are discussed at length in the academic literature.4 However, to our knowledge, there is little literature that discusses social media design principles as a whole, viewed in light of different theories of democracy. Some notable exceptions are Dahlberg (2011) and Bozdag and van den Hoven (2015). Dahlberg’s paper sketches four different democratic theories (liberal-individualist, deliberative, counter-publics, and autonomist Marxist) and their relation to digital technologies, but he does not go so far as to investigate specific design choices for social media. It is exactly this task that Bozdag and van den Hoven (2015) call for as they reflect on the design solutions to the so-called filter bubble effect. They argue that in order to strengthen and diversify designs solutions, engineers should be exposed to various traditions of democracy that embody different democratic norms and thus bear alternative implications for the (re)design of social media. There are a variety of different philosophical approaches to democracy, ranging from branches of classical liberalism (which emphasise the rights and freedoms of individuals) to libertarian accounts (which stress the importance of freedom and opt for minimal governmental intervention) to deliberative approaches (which emphasise the importance of public deliberation over the mere aggregation of votes) to more recent calls for ‘radical democracy’ (which emphasise the agnostic character of public debate and are sceptical of enlightenment ideals of rationality). Given the scope of this paper, we cannot explore all these different accounts. Rather we choose to focus on two prominent theories. By exploring the traditions of liberal democracy and deliberative democracy, we underline the call by Bozdag and van den Hoven (2015) and illustrate how varied the design directions can be when we design for a democratic social media.