Pluralism can serve as an end in itself, with the availability of a plurality of media institutions, outlets and views representing a symbol of a healthy communicative environment. Pluralism can also contribute to the promotion of informed national and global debates where a variety of voices can compete and co-exist.
This distinction is particularly relevant to analysing the paradox that has emerged with the increasing reliance on social networking platforms, apps, and search engines as gateways for accessing information: while information is becoming more diverse and easily available, many individuals seem less likely to access material that challenges their pre-existing views. According to the first understanding of pluralism highlighted above, the proliferation of online sources represents a vector leading to an increase in pluralism. According to the second, algorithms used by social networking platforms and search engines to provide users with a personalized experience based on their individual preferences represent a challenge to pluralism, restricting individuals’ exposure to differing viewpoints and newsfeeds. The images of ‘echo-chambers’ and ‘filter-bubbles’ have become common metaphors to describe this phenomenon. Algorithms are used to offer a service to users, responding to the excess of information occurring in a communicative
32 Wikimedia Foundation 2017. 33 Graham, Straumann, and Hogan 2015. 34 Geiß, Leidecker, and Roessing 2015.
environment where attention has become an increasingly scarce resource, but they also risk distorting the perception of reality. Algorithms present users with more sources of information that users have selected in the past, and with fewer that they have ignored or do not know about.
This process is similar to what has traditionally characterized content selection. In the case of print or broadcast media, for example, users, based on their tastes and leanings, and entrusted their favourite radio or newspaper to collect and analyse information, without necessarily seeking alternative channels that could challenge their worldview. What is distinctive in the case of ‘filter- bubbles’ in the age of the internet is how individual choices are influenced by algorithms whose functioning is hidden from users, combining individual preferences and computational bias. Over time, this combination risks building the artificial perception that a particular representation of reality or viewpoint is not only verifiable or persuasive, but is widely shared. The consequences of this type of media consumption gained attention especially following the 2016 referendum on membership of the European Union in the UK and the US presidential elections, when many confessed their surprise at results that seemed very distant from their expectations. This paradox is deeply rooted in the new reality, and in the contradictions of ‘algorithmic pluralism’. In the past, when users assembled distinctive bouquets of media outlets for their news diet, it was easier for them to assess whether their radio stations or newspapers of choice belonged
to a niche or the mainstream, and whether or not the views embraced by those outlets were gaining or losing traction. As news sources now appear in similar formats on social networking platforms, connecting niche blogs, aggregators, and mainstream media, it becomes increasingly challenging to understand and weigh where information is coming from and where the tide of opinion is shifting. The range of pluralism is obscured by the personalized individualism of the services and the way it diminishes choice. Research on echo chambers has also gained momentum, offering a more nuanced picture of the phenomenon and its variations in different contexts. While the majority of research has focused on Western Europe and North America, studies on the functioning and consequences of echo chambers in Asia-Pacific and Africa have also been conducted.35 The evidence emerging from these studies suggests that use of social media and search engines tends to increase ideological distance among individuals, but there is no consensus on the magnitude of this phenomenon. Comparisons between online and off-line segregation have indicated how segregation tends to be higher in face-to- face interactions with neighbours, co-workers, or family members,36 and reviews of existing research have indicated how available empirical evidence does not support the most pessimistic views about polarization.37 A study conducted by researchers from Facebook and the University of Michigan, for example, has suggested that individuals’ own choices drive algorithmic filtering, limiting exposure to a range of content.38 While algorithms may not be causing polarization,
35 Flaxman, Goel, and Rao 2016; Pariser 2011; Grömping 2014; Gagliardone et al. 2016.
36 Gentzkow and Shapiro 2011. 37 Zuiderveen Borgesius et al. 2016.
38 Bakshy, Messing, and Adamic 2015. 39 Hargittai 2015; Sandvig 2015.
they could amplify it, representing a significant component of the new information landscape.39
‘Fake news’
The phenomenon of so-called ‘fake news’ gained attention in the aftermath of the unexpected outcomes of electoral contests in Western Europe and North America in 2016. ‘Fake news’ is not simply ‘false’ news. Its nature is determined by
fraudulent content in news format as well as by an ability to travel as much as, and in some cases, even more than, authentic news. As a recent analysis of the origin and proliferation of ‘fake news’ has suggested, it is when a deliberate lie ‘is picked up by dozens of other blogs, retransmitted by hundreds of websites, cross-posted over thousands of social media accounts and read by hundreds of thousands’ that it then effectively becomes ‘fake news’.40 (see Figure 2-8)
Figure 2-8: The roots of ‘fake news’
The power of ‘fake news’ to masquerade as verified information and be shared widely, is closely connected to its ability to offer information conforming to and strengthening existing beliefs among a community of like-minded individuals.41 It also rides on more transformations that have affected information online: online business models and the nature of trust.
First, the business model pursued by social networking platforms and search engines encourages the production of information that is ‘click-worthy’, independently of its accuracy or public interest. This dynamic has been exploited by groups of individuals who are producing hoax articles attracting millions of clicks and shares, which can be turned into revenues through services such as Google AdSense.42 Electoral contests that took place in Western Europe and North America in 2016 have highlighted what has been seen as the transnational nature of this dynamic, and its potential dramatic consequences for trust and political debates at a national and international level. Actors residing in Central and Eastern Europe and South East Asia have attracted attention for their reported ability to produce news items widely shared by users in Western Europe and North America, while knowing relatively little of the political systems and candidates about which they were writing.43 The second and related phenomenon is connected to the nature of trust. It depends on assumptions that non-institutional forms of communication are freer from power and more able to report information that mainstream media are perceived as unable or unwilling to reveal. There are numerous cases where blogs and
social media have been able to uncover scandals or put pressure on public authorities. There are also evident shortcomings in mainstream media coverage. Authors of ‘fake news’, however, have exploited a belief in the independence of content shared across social media and turned it on its head, exploiting people’s credulity to gain economic or political advantage. Declines in confidence in much traditional media44 and expert knowledge45 have created fertile grounds for alternative, and often obscure sources of information to appear as authoritative and credible. This ultimately leaves users confused about basic facts.46
Large internet companies, whose credibility has been threatened by this phenomenon, have sought to develop new responses to limit ‘fake news’ and reduce the financial incentives for its proliferation47 [see TRENDS IN MEDIA INDEPENDENCE]. Growing public attention to this issue is likely to transform how users relate to information, especially if combined with more aggressive media and information literacy campaigns.