21 september 2018
During its early days, social media was met with great optimism. It was expected to foster greater cooperation by bringing together diverse groups of people to freely discuss and solve major societal challenges. Instead, as recent events have shown, the effects of social media have been characterised by two rather problematic tendencies. First, it seems rather to have increased separation and polarisation between groups by allowing users to seek out individuals who hold similar views and opinions, resulting in ‘echo chambers’ that reinforce existing beliefs and fuel biases. Second, this development has gone in tandem with a decline in trust of credible, balanced and factually correct news sources and a corresponding rise of disinformation. This parallel development is leading some observers to suspect a link between the two phenomena, but the exact nature of such a link has proven hard to establish, partly because online information spreads like viruses on complex network structures.
For his study, Törnberg, assistant professor of Sociology at the UvA, used a computer model to simulate misinformation spreading virally on social media as a result of users deciding whether or not to share it with their friends. After first running his simulation on millions of generated social networks, he then used Twitter retweet data of interaction between European politicians to generate empirically grounded networks in order to validate that these dynamics in fact also play out in ‘real-world’ social networks structures.
‘What my simulations revealed was that the presence of a group of like-minded users in a social media cluster (i.e. echo chamber) will tend to increase the diffusion of viral news and rumours that align with the views of the cluster’, says Törnberg. ‘This seems counter-intuitive: one would think that if a group of users are less connected to the rest of the network, they will have a harder time spreading their views. The reason for this is perhaps best understood through the lens of a metaphor: if we think of the viral spread of misinformation in a social network as akin to a wildfire, an echo chamber has the same effect as a dry pile of tinder in the forest – it provides the fuel for an initial small flame, that can spread to larger sticks, branches, trees, to finally engulf the forest.’
Törnberg’s findings open the way for further research on the link between echo chambers and the viral spread of fake news, as well as possible measures to counteract this. ‘The World Economic Forum recently listed the online spread of misinformation as one of the primary threats to human society, and rightly so. In my opinion, this phenomenon points to how these so-called social technologies are in fact making it harder for us to connect with others and to form a joint understanding of our shared world. The truth is we do not even fully understand the effects that these technologies are having on our society. We need studies like this to understand how to shape our digital meeting places to help us build the foundation for a common world, and help weave, rather than fray, our social fabric. And I do think it is possible to design social technology that brings us together rather than tears us apart.’
Törnberg’s study forms part of the larger EU project ‘Opinion Dynamics and Cultural Conflict in European Spaces’ (ODYCCEUS). As part of this project, Törnberg will continue to investigate how the dynamics of new social media are impacting societal polarisation and cultural conflict. He is currently finishing a study on the impact of the online platform Airbnb on gentrification and racial conflicts in urban areas, which will provide novel insights into how social media platforms are shaping social life.
Petter Törnberg, ‘Echo chambers and viral misinformation: modeling fake news as complex contagion’ in PLOS ONE (21 September 2018). Doi.org/10.1371/journal.pone.0203958