Keywords: Social Networks, public sphere, Internet fragmentation

Scholars in political sociology have suggested that the public sphere may be seen as tiered or compartmentalised with regard to the nature of the interaction, degree of formalisation or of justification of viewpoints. The idea is that the diversity may itself play a constructive role in the self-organisation of the public sphere. The challenge of meaning convergence, that is, of making order out of complexity or will formation out of opinion formation, has been analysed on two levels deriving from disciplines that until recently belonged to quite different research traditions. One tradition came out of network analysis and focused on the Internet as a socio-technological structure, whereas the other tradition came out of the Habermasian theory of the political public sphere and set out to do empirical studies of local, national and supranational public spheres in terms of media content. From both positions, a dramatic expansion and pluralisation of communication is observed that subsequently leads to examinations of the possibilities for integration and convergence (Davis 2009).

This chapter, from the point of view of complexity and the generation of a network-like structure on the Internet, particularly addresses Yochai Benkler’s broad analysis of current changes of the public sphere in his Wealth of Networks (2006). I argue that Benkler presents a bold and timely attempt at integrating the notion of the public sphere and the Internet. Network theory and analysis may shed light on much of the dynamic running the growth of the Internet. However, his idealist notion of both leads him toward a mystification of, and an unwarranted optimism about what currently goes on. A less idealistically informed network analysis may account for the concentration of capital, and the related interests in algorithmic surveillance and strict regimes of copyright lie outside his scope.

New media, genres, styles and topics

Since what I have called personal media allowed more people to produce texts and take part in communication, the Internet has offered new forms of access to public authorities, new channels of coordination and influence for social movements, and a multitude of more or less stable settings for chat and discussions (Rasmussen 2013). On all accounts, digital media provide quite different circumstances for communication than the mass media. The most central ones can be listed as follows: (1) social movement activity (Internet, blogs, email, wikis); (2) discussion and chats among citizens (blogs, chatrooms, e-mail lists); (3) citizen access to MPs and public authorities (Internet, e-mail, blogs); (4) online ‘participatory’ journalism (Internet, email, SMS, MMS, blogs); (5) Connections and weak-tie networks (sites such as Facebook and Twitter); (6) search.

In contrast to the public sphere once entirely dominated by public encounters and the mass media, the Internet and personal media propels a more differentiated public sphere, both in terms of topics and styles, as well as with respect to the number and variety of participants. The current public opinion in advanced societies is more niche-oriented because of a more diverse mediascape, but also because of a more ethnically and culturally pluralistic society in general. First, the diversity of topics is broader than in the mass media; it has been argued that the nature of topics in the more recent digital dimension of the public sphere are more particular, private and local than the mass media, in spite of the global reach of the Net (Becker & Wehner 2001, 74). However, it is also the case that global or international issues are constantly discussed, even if in individual and local ways. Second, the span of styles and genres (informality, impulsivity, rhetoric styles, politeness, civility, etc.) far exceeds the mass media. Third, the number and diversity of voices is considerable compared to the mass-mediated public sphere (children, youth, less educated, etc.). Individuals may change between the roles of the generally relatively passive citizen and the more active and specialised communicator.

The differentiation of topics, styles and participants transforms the public sphere as well as how we view it in relation to democracy and culture. With regard to all three differentiation trends, the driving force is the personalisation of media on the Internet, thus enabling the individual to voice opinion directly to public power, to participate in campaigns and social movements, and to exchange opinions on online fora in her own ways and language, drawing upon personal experiences, knowledge, engagements, values and judgements. Because the threshold for speaking out on the Internet decreases, more people do so, and thus increasing participation lowers the threshold even further. And yet, because the threshold is still much higher than watching the television news, more involvement and interest accompany Internet participation. One tends to appear more as a person who is interested in particular themes and interests rather than someone who is simply being a citizen among millions (Becker & Wehner 2001, 74). Whereas the mass media produce homogeneity, the multitude of Internet fora seems to produce a heterogeneity that has difficulties in controlling itself communicatively.

Consequently, the Internet accelerates the differentiation of the public sphere in terms of topics debated, styles applied and persons involved. The diversity of communication on the Internet is in part caused by anonymity and quasi-orality (and therefore more extreme viewpoints and considerably uncivil characteristics, as well as unconventional ways of argumentation), diversity of communication forms (from chat fora to blogs and homepages with comment functions), and diversity of intertextual connections between fora (hyperlinks, RSS feeds, network sites). In its reciprocity, heterogeneity and resistance to censorship, it stands in a complementary position to the mass media. Particularly the national and international mass media enable broad attention around some prioritised public topics ‘of national interest’, and so serves as a resonance for national and international politics. The Internet and personal gatherings underline the individualisation and segmentation of modern societies, in that attention and engagements are spread among a wider range of topics, which make a political focus difficult to trace.

As a modern response to a dynamic democracy, the digital dimension of the public sphere offers less guidance for politics but more possibilities for expression. Compared to the journalism of mass media, online journalism tends to be more compartmentalised and based upon self-selection and personalisation. The criteria of selection are to be composed by the individual. Rather than offering carefully edited information, it offers a differentiated space for interaction and for presenting user-composed information, which tends to be rather specialised, and is also closer to personal opinion, rumour and unconfirmed information. Whereas the mass media works toward conformity and common denominators, the Internet is more oriented towards particular interests. It is located ‘between’ the mass media public and face-to-face interaction such as public meetings, rallies, etc. (Becker & Wehner 2001, 75).

Studies that I will refer to below indicate negative side effects of the new fora: polarisation of debates, isolation of issue-based groups, unequal participation, lack of responsiveness and respect in debates and incivility. For these reasons, the value of digital fora in a public sphere context is questioned. A main problem addresses their numerous, local, segmented character, which makes it difficult to see how their normative communication may integrate into larger sentiments of public opinion. What seems to be lacking in many fora on the Internet is a culture for civil, public communication, or simply a public culture. Due to the lack of personal experience in an open space and the absence of editing functions, communication often has a private style, in spite of its open and widely accessible nature. In spite of being public, it draws on genres for private communication. This has two negative consequences: (1) the discussion has problems with the complexity of the issues discussed, and (2) the discussion has problems reaching a self-referential, self-critical level where the normative aims of the discussion are subject for discussion. In other words, the responsibility of public communication (publicity) is not taken into consideration in the nature of the interaction.

Still, the vast majority of these fora fulfil some basic requirements of a public sphere: They are (as are the mass media and local meetings), committed to improve social conditions in one way or another, and also to free speech and open dialogue. They are also committed to make themselves understood and to understand others in an open space of an assumed indefinite audience, if for no other reason than to make rhetorical shortcuts or reach compromises. Some sort of cooperative action may seem to be at work here.

At first glance, there are few and weak functional equivalents to editing and regulatory agencies, like editors, journalists and judges (Bohman 2004, 143). However, there are in fact plenty of intermediaries on the Internet, as in online journalism and moderators, filters and other software systems, the norms of social movements and organisations, which all serve to normalise communication in one way or another. In spite of its ‘anarchic’ nature, much of the communication on the Net is embedded in larger normative frameworks that tend to discipline interaction. Second and more importantly, we should not assume that the Internet is isolated from the mass media and face-to-face meetings as a platform for a public sphere. The intertextuality of meaning and communication in and out of fora in the public sphere are innumerable. The lack of intermediaries on the Internet is less of a problem than it may seem, precisely because it is so integrated with face-to-face and mass mediated interaction.

Still the term ‘community’, though often used to characterise value-based collectives of consensus-building and loyalty, is inadequate in accounting for online interaction, assuming too much fellowship and coercion while clarifying little about the empirical nature of public interaction and communication. In its place, an extended interpretation of the public sphere is needed (1) that recognises the role of the media without discriminating crudely between online and offline interaction, (2) that accounts for the construction and transformation of multiple and overlapping networks and the building of power that stem from such constellations, and (3) that accounts for purposive/instrumental, affective and expressive practices. If the ritualistic and cultural aspects of the term ‘community’ are toned down, it may capture the individual orientation towards stable social and political networks, while the term ‘networked community’ more aptly covers the decision-oriented (rather than tradition-oriented) nature of current political expression.

The networked public sphere

Yochai Benkler's study from 2006 clearly stands out as the most ambitious attempt to synthesise the idea of a rational and enlightened public sphere with the theory of social networks. While Manuel Castells’ voluminous work on the global ‘network society’ caught interest in the beginning of the century, he did not address public opinion as such. Jan van Dijk’s book (2012) The Network Society addresses the implications of digital media for democracy without discussing mediated public debate. In contrast to the reserved assessments of Habermas, Barber, Schudson and others, Benkler (2006) argued in this work that the Internet represents a significant change towards a more democratic and responsive public sphere. He enthusiastically argued that the transition from a mass media-structured public sphere to a distributed discursive architecture with multidirectional links among all nodes in the information environment has eliminated barriers to communication and fundamentally changed the possibilities for participation in the public sphere. In democratic countries, the lowered threshold for mediated conversation leads to a radical expansion of the public sphere. For totalitarian nations, this new architecture represents bad news – a new threat that they find very difficult to handle, due to the decentred, end-to-end structure of the Internet. If at all, in what ways does the Internet democratise the public sphere?

A network-like description of the public sphere suits the pre-Internet era as much as it does the current era. However, with Internet access on all computers, the development has accelerated in different directions. The Internet has intensified the circulation in time and space. It has radically lowered the barrier for interaction of all forms by combining the written and audio-visual speed and scale of the mass media with the interpersonal dialogue of face-to-face interaction. Since the emergence of e-mail in 1972, the Internet has come up with software that has presented a wide range of communication structures. The distinction in mass media between producer and audience is suspended, as is the case with the distinction in face-to-face interaction between physical place and social space. This creates the possibility for a wide range of recombined communication situations. What seems to emerge is a more network-like, distributive dimension of the public sphere (Benkler 2003). This carries with it, Benkler argues, a greater democratic potential in terms of participation, but with some democratic deficit in terms of focus and reflexivity. This however, is about to change, or ‘normalise’ itself. The Web may structure the organisations that apply it, but it also becomes heavily influenced by them. The Web in our context is not technology, but a possibility for communication that quite naturally takes shape from its institutional contexts.

What is of importance in Benkler’s understanding of a ‘networked public sphere’ is that smaller Web sites are linked thematically together, constituting clusters of public communication. However, there is also a concentration of attention on a limited number of Web sites. As Benkler (2006, 235) argues, ‘A tiny number of sites are read by the vast majority of readers, while many sites are never visited by anyone. In this regard, the Internet is replicating the mass media model, perhaps adding a few channels, but not genuinely changing anything structural.’ Exactly how concentrated the Web structure is in terms of linking and attention has been measured with network analysis. The results span from power law structure, to a distributed network structure. Still, search engines like Google point the reader towards relevant information. Against Cass Sunstein’s (2009) argument that the Web is increasingly fragmentary, Benkler argues for a networked public sphere of interlinked sites and arenas of communication. The polyphony of debate, argumentation, agitation and mobilisation constitutes, in an abstract sense, a complex sphere of public communication dedicated to matters relevant for all.

Social media networks generate social capital through a number of structural features, often conceptualised as the ‘architecture’ or ‘topography’ of networks. From the insight of network theory, I will mention four such structural features drawn from network analysis that also underlie Benkler’s argument (Rasmussen 2008, 2014):

Clusters: By mediating clusters of strong ties, and with high density of connections along with unifying norms and trust, social media mediate small community-like groups of sociality and loyalty. The effect of such bonding groups is a high degree of homogeneity. Such groups provide direct links to others, who each have links out of the group. However, such groups are relatively introverted, with relatively few connections out of the group compared to the connections inside the group. In other words, the connections are quite far from random. The redundancy is high inside the group, but low out of the group. The kind of social capital from such groups (best described by Coleman) is also called ‘thick’ social capital. Such clusters inhibit small world connections for the individual in the group. But when such clusters appear as nodes in larger networks, they serve the small world effect from the local level.

Short-cuts/redundancy: In networks of relatively random connections between the nodes, enabled by a large share of weak ties, bridges will also develop between networks. In large, random-like networks, there are fewer clusters and more random connections, and therefore a shorter path distance than in a world of clusters. A large share of my connections also know someone I don’t know, and some of them serve as bridges to other networks. The network is more individualistic, redundant and heterogeneous. In combination with clusters, it more easily connects dispersed groups.

Supernodes: Online and offline networks tend to generate a certain bias, favouring the nodes that enter the network early or possess other advantages. Merton called this the Matthew effect; others call it the Pareto effect. The long tail thesis demonstrates the same thing. The web-topology develops relatively few supernodes (extraordinarily popular sites), and a very large number of smaller sites. Web use creates ‘mountain-peaks’ in the horizon, which everyone sees regardless of where one is positioned, and which dramatically reduce the number of steps between randomly selected nodes. Power-law distributions emerge, due to knowledge, conformity and time. Herbert Simon pointed towards time and attraction as the two central mechanisms that tend to create bias. A self-enforcing differentiation takes place, which is strategically developed, as in Google’s search method, and in rating methods (Slashdot, Digg). Popularity leads to more popularity. This point, I argue, has proven itself highly valid the last decade, and could not be sufficiently addressed by Benkler.

Cascades: Increasing exposure to practices of others with lower thresholds than oneself leads to lowering thresholds for joining campaigns, petitions, spreading ‘memes’ (jokes, images), etc. and by spreading scandalous information, snowball effects more easily emerge and end. Virus-like phenomena appear due to the speed and reach of information, which for some reason seems important, scandalous or provocative. As accumulated side-effects, certain forms of information are imitated and duplicated locally, and escalate. New information of some considered importance (again popularity) reaches the individual.

Successful social media allow for all these features in a productive combination:

They enhance a small world effect by aiding the users to navigate locally and long-distance, to identify changes and pick up news and groups, and to establish new connections. They potentially magnify the effects of individual actions, and so instigate social change. They all remain dependent on individual action. Most of all, and absent from Benkler’s analysis, is the accumulation of capital and power that this network logic enhances.

What Benkler describes is not only relevant for public communication on the Internet, but for the public sphere in general. The most viable way of understanding the current public sphere in Europe is precisely as ‘polyphony’ – as a network of networks of sites of communication. As noted, Habermas (1996) too has used ‘network’ as a metaphor for the current public sphere without elaborating on this further. Empirical research on the quality of the public sphere needs to consider the Internet as a functional complement to the mass media and face-to-face interactions. A vibrant and democratic public sphere depends on its internal composition and dynamics, particularly whether the direct interaction, the mass media and personal media based on the Internet are integrated with one another through overlapping networks of individuals, themes, opinions and knowledge crossing in various ways and shapes between its ‘compartments’ and realms.

Blogs are a case in point. Benkler observes that bloggers work from a different logic than the mass media and so may pursue issues independently of the agendas of media and politics and may also pursue stories over a longer time than the mass media with their short news cycle. Nor are bloggers dependent on norms of objectivity and neutrality, which inhibit mass media from participation in campaigns against negative forces in society. From the perspective of conventional journalistic norms, engagement is considered unprofessional in the mass media. Nevertheless, stories that are brought up by blogs or smaller websites need to be picked up by the mass media in order to affect the political public sphere. The inner circles of the public sphere close to the centre of political decision-making are still dominated by the national mass media. But the new sources of insight and debate more frequently come from the networked information environment. The bloggers are feeding the watchdogs.

A research strategy to clarify the networked character of the public sphere would be to apply network analysis software to detect hypertext links between local, national and international websites. Another approach would be to focus on certain distinct issues, to see how argumentation in various media refers to each other or is interlinked in other ways. Regardless of the specific results, such analysis would conclude on degrees of connectedness within a population.


With the growing complexity of digital and analogue communication, how can argumentation and deliberation survive in this communicative cacophony? How can the threads of communication somehow be untangled and filtered towards a more unified voice, and into an underlying sense of community or even a true public opinion? The key is the network structure of the Internet and its media. The internal dynamics of filtering and synthesis by way of links (references, feeds, retweeting, etc.) enable those who are interested in the topic and connected to any of the participatory sites to get familiar with opinions and facts on the issue. By way of the structural features of the Net, and in spite of its dramatic and dynamic growth, thousands of views can in fact coalesce around certain central arguments, events or sites, presenting petitions, boycotts, campaigns, etc.

This relates foremost to structural aspects of the Internet and web predominantly addressed by network analysis. Insights from mathematics-informed sociological network analysis have shown how the Net operates in synthetic ways that create some sense of order on the Net despite its enormous activity and incredible diversity. This form of analysis has some direct implications for studies of the Internet as an engine for the political public sphere, because they can counter allegations about the Internet’s non-democratic effects, such as information overload, centralisation and digital divides. And most importantly, this analysis addresses the claim that the widespread use of the Net may undermine the critical and investigative function of the mass media and possibly the whole existence of the paper-based press.

Consequently, in this network perspective the political public sphere is considered as an open and complex network of networks of ideas and arguments about public themes, a place where agents and powers seek to make judgements and legitimise statements and actions. This understanding accomplishes the following: (1) it stresses the function of the public sphere as a space of communicative legitimacy; (2) it excludes the motivations of individual and collective actors; (3) it excludes the actual media and arenas of communication; (4) it includes the imperative of both diversity and focus of communication, and (5) it includes both the chat-like and the deliberative character of public communication. However, in this perspective, there are no counter-public spheres detached from the main public sphere and no plurality of public spheres, only interconnected ‘nodes’ of debates and counter-debates, publics and counter-publics. The connections between discursive nodes expand the range of arguments, problems and solutions, and widen the range of quality of argumentation.

An additional point is that this approach is not only normatively reconstructive in a Habermasian sense, it even enthusiastically welcomes a dramatic democratic change, due to the emphasis on network-enhancing digital technology. Citizens are no longer absorbed in observing and listening audiences but experience themselves as more responsible for local and national affairs. They are no longer simply private observers but subjects involved in public and political communication (Benkler 2006, 213). Even more dramatic is the structural changes of the public sphere itself. Benkler argues that the production of the public sphere benefits significantly from the networked information economy that has emerged since the ’90s, with the World Wide Web and web-based media like blogs, wikis and other social media. They offer insights and commentary of a rather different character than do mass media that are often dominated by conventional views considered to be accepted by the public, and by the views of their owners. Internet-based media are set up by NGOs or individuals that have different approaches to issues and that present their case in different, often unconventional and non-journalistic ways. The facts presented are different and so are the views and strategies that contextualise them. Typically, politically important revelations published in the mass media are often picked up from low-budget blogs and sites like Wikileaks. Many of the sites and blogs depend neither on advertising nor do they express mainstream tastes and opinions, since the motivation for publishing is anchored in personal engagement.

Fragmentation and concentration

Among the most frequent arguments against the Internet as a contributor to the production of the public sphere is that the enormous activity on the Net creates a chaotic, fragmented discourse, which in turn may lead to ‘balkanisation’, or parallel communities: isolated groups cultivating introvert, sometimes extreme views. The critique is that this Internet-based public sphere will not be able to serve its democratic function vis-à-vis the political decision-making domain. Alternative ideas and constructive views will tend to get lost in the jungle of fragmented communication. Cass Sunstein (2009) argued that the omnipresence of information combined with a weakened press to condense and synthesise would undermine the common base for political discourse. Related is the point that fragmentation leads to polarisation and reinforcing of views and beliefs. When contrary or conflicting positions rarely get the possibility to meet and challenge one another, they tend to develop more extreme views and develop further distance from one another. This is precisely what is not supposed to happen in the public sphere, according to the Habermasian approach. Group escapism, what Todd Gitlin (1998) called ‘sphericules’, goes dramatically against the value of sharing ideas, encountering new viewpoints and confronting arguments with arguments. On the contrary, it may in turn lead to islands of extremism and ignorance. Research has, as we will see, analysed alleged tendencies of ‘homophily’ and found that if people with some shared interests and opinions form clusters, this makes it less likely that people of different opinions find one another in debate. But they may find each other indirectly through news feeds and linking. Clusters of similar interest may trigger some collective engagement, but they may also form isolated groups that conform to biased images of society.

Illustrative of this is the debate in the United States about whether the Internet is to blame for an increasingly polarised political atmosphere between adherents of the two main political parties (Fiorina and Abrams 2008; McCarty et al. 2006). Analysis of the American blogosphere showed that political bloggers tend to belong either to a right wing or a liberal cluster, and that they link almost only to blogs on their own side of the political gap (Adamic and Glance 2005; Hargittai et al. 2008). A similar tendency has been found on Twitter (Conover et al. 2011). Readers of political blogs also tend to be more politically engaged than others and to follow blogs that support their own view (Lawrence et al. 2010). Farrell suggests from various findings that a large group of blog readers tends to be exposed to a wide variety of online sources and views, whereas another smaller group of generally more politically active people preferentially follows political blogs of their own view and to a far lesser extent reads blogs of different views (Farrell 2012, 42). Research indicates that the more connections to other political leanings people have, the less likely they will engage politically (Mutz 2006). Cross-cutting relationships among friends and work is considered to be a good thing but tends to weaken political engagement, while increasing political engagement stimulates group isolation and polarisation. There thus seems to be a trade-off, or a dilemma, between political involvement and cross-political interaction. Causes and effects in such cases are, however, very difficult to unpack, and research conducted by Gentzkow and Shapiro (2010), based on a large sample of blogs, did not support the polarisation thesis.

Another type of critique takes the view that the Internet has developed concentration and marginalisation. While the vast majority of speakers on the Net are never heard, a tiny minority of sites or individuals is highly linked and receives the vast majority of visits and attention. This critique is based on studies of Internet linking and addressing, which clearly shows a skewed distribution of linking and visits on the Internet. There is a low probability that a web site will be linked to from many other sites, and a very high probability that for a given web site, one or no other site will link to it. The blogosphere is another similar case: While the proliferation of blogs was seen as an important democratic innovation, as it allowed anyone to become a publisher at little or no cost, others commented that the blogosphere preserved the biased profile of the mass media society, because a group of elite bloggers attracted most of the attention; whereas the majority of bloggers receive few visits (Farrell and Drezner 2008; Shirky 2009). American research also indicated that successful bloggers tended to be narrowly recruited socio-economically (Hindman 2008).

The distribution of linking forms what is called a right-skewed distribution and a power-law distribution (Benkler 2006, 244). If we are to associate democracy with a statistically normal distribution, the Internet is far from democratic. For reasons of ‘preferential attachment’, the early established and thus highly visible and frequently visited nodes on the Net tend to grow at the expense of the late-coming majority of less visible ones, because we tend to do what others do and have done: Linking tends to be conforming behaviour: It generates conservative patterns. New nodes tend, on average, to be less frequently visited than the ‘first riders’. Of course, there are important exceptions to this tendency, such as Google. Over time, however, the dynamic Internet topology has established a limited number of supernodes and a multitude of rarely visited sites. The concentration of linking among certain giant nodes, Benkler argues, has a strikingly democratic effect. The new infrastructure that lies under a wide range of digital media seems to work in ways that favour freedom of expression, diversity and conversation. Filtering and synthesis mechanisms seem to yield clustering and redundancy, with the happy effect of convergence without control. Benkler argues that the networked public sphere allows hundreds of millions of people to publish without consequences of disintegration and concentration.

We know that the network at all its various layers follows a degree of order, where some sites are vastly more visible than most. This order is loose enough, however, and exhibits a sufficient number of redundant paths from an enormous number of sites to another enormous number, that the effect is fundamentally different from the small number of commercially professional editors of the mass media (Benkler 2006, 253).

Benkler argues from this that the topology of the Net is ordered through some highly visible peaks supplemented with sufficient redundancy of links so that no node controls the flow of information. The power-law distribution neutralises the fragmentation argument, and the redundancy argument neutralises the concentration and polarisation arguments. Following Benkler, the Internet cancels out all main obstacles to ordered diversity, which makes it perfect as a platform for the democratic public sphere. It is both more resistant to control, and less susceptible to trivialisation than are the mass media.

Benkler shares Habermas’ view on the public sphere, along with the deliberative democracy approach, although he locates the Internet in a different factual position. Both emphasise the importance of generating involvement in the civil sphere around controversial issues. Still, Benkler goes further and argues that the Internet more than the mass media enables users to take part in dialogues where they experience themselves to be real participants in debates on society. Benkler's argument about the networked information economy as a platform for the public sphere is based on normative arguments about human reason and normative political theory, supplemented with an account of technological change: On the distributed, end-to-end nature of the Internet as a network, Benkler’s position is highly normative in that his objective is to show what we lose if the structure of the Internet becomes corrupted. The Internet can only function in a liberal and liberating way if it functions along the lines it has done so far. If any of the global Internet corporations such as Google or Facebook become so powerful that they can lock our communication into their proprietary services, the Internet as a democratic platform may get lost. This applies too, we may add, if national intelligence agencies ally with the multinational Internet industry in conducting widespread eavesdropping and surveillance of both ordinary citizens and politicians. Nevertheless, Benkler clearly stands out as a liberal idealist.

Benkler’s analysis ignores the problems of the concentration of proprietary control and ownership, leading to what I have called supernodes. (Rasmussen 2014; Fuchs 2014). All emphasis is on open structure of the Internet and the lower threshold for access to the public sphere. This kind of idealism is widespread in the writing about the Internet. Another more realist-oriented view would take into consideration the structure and content of websites and mega-corporations like Apple, Google and Facebook. In spite of the distributed networked structure of its original and initial structure, the Internet is becoming centralised in terms of services like social media, search, streaming and consumer retailing. I have addressed the social and conflict-oriented history of the Internet in another work (Rasmussen 2007). Here, I can only indicate the most central problem areas: (1) the privatisation of messaging services in corporate social media, and of the portals to the web; (2) the surveillance, big data analysis and reselling of personal data for targeted advertising; (3) the cooperation between large media companies and public intelligence agencies like NSA on exchange of private data revealed by Edward Snowden, and (4) The unwillingness of large Internet companies to operate in accord with public regulations regarding data protection and market regulation. User data are thus raw products for private advertising industry and for international state pervasive surveillance.


These aspects cannot be ignored in the assessment of the Internet-mediated public sphere since it affects the way social interaction operates freely and actually serves the intended purpose of creating a rational foundation for public decision-making. Public communication without privacy is hard to imagine in questions of controversy. We have already seen how dangerous whistle-blowing is in questions of national and diplomatic secrecy. It is also a fact that encryption has become the recommended solution for groups fearing state monitoring. To the extent that this image of rational public sphere is undermined (and has been for a long time), another conception of public debate needs to be developed that more accurately describes the reality. This reality can be described as public participation under the condition of social control, similar to what Habermas once characterised as feudalisation and colonisation. We are not here talking about a liberal public sphere suffering some temporary deficiencies, but rather public opinion as a sphere for conflict, profit and self-legitimation of power. To explore this process of meaning-formation and control, a more accurate image of politics and public opinion is to be found, not in Kantian and contractual approaches, but in the more realist tradition from Machiavelli and Hobbes via Weber to Bernard Williams.