One of the central questions about the role of the Internet in a digitally augmented public sphere concerns the tension between plurality and polarisation. For several years scholars have sought to establish whether the internet extends the political space, particularly through greater civic participation, or whether it furthers the fragmentation of political discourse into ‘echo chambers’ consisting of like-minded actors, who are apt to perpetuate conflict rather than resolving it and threaten to erode the very structures of political discourse (Sunstein, 2009).
Against these far-reaching implications, research exploring from a hyperlink perspective how actors structure the online space have for the most part examined the linking patterns on single platforms such as Twitter (e.g. Williams, McMurray, Kurz, & Lambert, 2015), the activities of civil society actor types such as bloggers (e.g. Elgesem, Steskal, & Diakopoulos, 2015), and generally focused on instances of heightened controversy between opposing camps in an issue.
The present study builds on this work by extending it in three ways that address existing shortcomings: First, although it takes the question of civic participation as its starting point, it does not restrict the analysis to civil society actors, but investigates how they structure hyperlink issue networks in interaction with other actor types – political actors, the media and the economy. Second, it assesses how controversy and polarisation are related to each other by moving from the predominant case study design to a comparative approach. Taking climate change as a contrasting case, the study compares a country where it generates only moderate amounts of controversy, Switzerland, with one where the conflict between the opposing camps runs deep, the US. Finally, the study presents an inductive approach to identify close-knit environments in hyperlink issue networks and develops a set of indicators that allow us to distinguish whether connections among like-minded actors are within expectable boundaries of a normal ‘economy of attention’ (Benkler, 2006), or whether they represent actual forms of polarisation that are detrimental to public discourse. This allows us to specify thresholds of polarisation that have so far been absent from current research.
As the results show, polarisation is a multi-layered process that can traced on the level of the online issue space as a whole, the communities which make it up, and the tie formation dynamics that lie in turn at their heart, thus distinguishing different levels of polarisation intensity. The analysis suggests that complex forms of fragmentation and polarisation interact and are at least in part responsible factors in structuring both of the online issue spaces examined, though the underlying dynamics vary greatly between the countries – and the single communities. They also show that pockets of diversity exist, but they are pushed to the margins of the actual debate. The paper discusses the ambivalent normative implications of polarisation in the digital age, which requires us not only to develop novel methodological approaches, but also theoretical perspectives that distinguish between ‘benevolent’ forms of polarisation characterising many civil society movements, and the detrimental instances that undermine public debate.