Democracy in disguise: the use of social media in reviewing the Icelandic Constitution

Method and theory

The reviewing of a Constitution touches upon issues of wide societal relevance. The proposed bill differentiates between basic elements, human rights and nature, Parliament (Althingi), President of Iceland, ministers and the cabinet, judiciary, local governments, foreign affairs, and final provisions. The actual work prior to the bill proposal was, of course, much more extensive as information and discussions were gathered from a variety of sources. This was the case in the upper structures of political public spheres, with a bill submitted by the Prime Minister to Parliament, discussions and presentations by government and Parliament, and the issuing and writing of an extensive report by a Constitutional Committee. Apart from this, numerous interactions were addressed in the cultural public spheres. These included a large National Forum in 2010, all communication enabled by social media, the Constitutional Council’s own website, and discussions in the media.

In order to identify relevant themes that surfaced in these multiple information streams and particularly in different social media, I conducted a qualitative content analysis (Hsieh and Shannon, 2005; Mayring, 2000) on the site used by the Constitutional Council (stjornlagarad.is) as well on the streams that this communication produced on the Council’s profiles on Facebook, Flickr, Twitter, and YouTube. Included in this data material is all communication that took place on the stjornlagarad.is site, all communication on the Council’s Facebook page (from 28 February 2011 to 10 April 2012), the 171 tweets on Twitter (from 4 April 2011 to 4 August 2011), the 51 videos on YouTube (from 20 April 2011to 29 July 2011), and the 522 photos posted on Flickr (from 13 April 2011 to 4 August 2011).

According to Hsieh and Shannon, qualitative content analysis ‘is defined as a research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns’ (2005: 1278). The qualitative content analysis thus aimed to identify relevant themes for the framing of discussions concerning social media, and how members of the Constitutional Council received and treated this communication. This was done by conducting an inductive category development, followed by a deductive category application (Mayring, 2000). This allowed me to identify the following themes relevant to exploring social media’s participatory potential for the rewriting of the Icelandic Constitution: practical information, exclamations, the ‘loop of dissemination’, and the deliberation – statement axis. I address each theme separately in the analysis.

Most of the communication that took place on stjornlagarad.is and the various social media services was conducted in Icelandic and English. English communication is cited directly and Icelandic communication has been translated by the author.

The theoretical framework underscoring the analysis is derived from writings on the multimodal and multichannel characteristics of digital communication and online participatory cultures, as well as from writings on participation, civic cultures, civic engagement, cultural and political public spheres, and networked publics. The aim is thus to analyse the kinds of civic agency expressed through the online networked publics operating on the social media services and consider whether these influenced the Council’s decision-making.

From Assembly to Council

On 29 July 2011, the Constitutional Council presented the Speaker of Parliament with the bill for a new Constitution. The bill was unanimously approved by all 25 members of the Council, who agreed that the general population of Iceland should be given the chance to vote on their proposition prior to Parliament’s final vote. However, as noted above, this was just part of a longer process that was formally initiated on 4 November 2009, when the Prime Minister submitted a bill on an advisory Constitutional Assembly. The foundation for the Constitution was still that of the Kingdom of Iceland from 1874, which does not reflect current Icelandic societal reality. More importantly, there had never been an extensive democratic discourse that included the general public on constitutional issues. It is in light of this that the inclusion of social media becomes particularly interesting.

However, the path from the Prime Minister’s submission to the actual bill was a thorny one. Parliament had already made amendments to the Prime Minister’s bill before it became an Act. The Assembly was, for instance, only permitted to take 2–4 months, instead of the 11 months originally proposed, and a special Constitutional Committee was asked to prepare a report that the Assembly was obliged to take into consideration. Most importantly, the Act maintained that a National Forum of 1000 people be held in order to determine core values for the Assembly to take into account. Furthermore, the Act initially lacked general acceptance by Members of Parliament as it was supported by just 39 votes (one voted against, 11 were absent, and 11 abstained). These results were not uplifting, and they coloured parliamentary practice for the remainder of the process, as old quarrels between MPs and established political parties resurfaced throughout the period – and, indeed, still do.

Yet in terms of participation and engagement, the National Forum was a success, boasting 950 participants, and the same is true for the process leading to the elections to the Assembly. Here, 522 candidates expressed an interest, with 9% of Iceland’s population signing letters of commendation for candidates. The results were made known on 30 November 2010: 25 delegates had been elected (15 male, 10 female), and 35.95% of the population had voted. However, in December 2010, three complaints regarding several formal structures of the elections were received by the Icelandic Supreme Court, which invalidated the elections. In response to this unexpected and embarrassing development, an advisory group was appointed, and this recommended that Parliament appoint a Constitutional Council by a parliamentary resolution. The idea was for those delegates that received the greatest number of votes from the invalidated election to the Assembly be offered a seat on the Council. On 24 March 2011, this resolution was approved with only 30 votes: 21 MPs were opposed, 7 abstained, and 5 were absent. It was mainly MPs from the political parties that had held power for a long time prior to the financial crisis who opposed the resolution.

The structure of the process

The Constitutional Council was formed on 6 April 2011 and divided itself into three working groups. Group A worked on basic values, citizenship and national language, the structure of the Constitution, natural resources, environmental issues, human rights, and the state church. Group B worked on the foundation of the Icelandic Constitution; the roles, positions, and responsibilities of the President, Parliament, government, and ministers; the responsibilities of the executive; and the status of municipalities. Finally, Group C worked on public democratic participation, the independence of judicial courts, the supervision of judicial courts regarding other holders of state authority, parliamentary elections, the constituency system and MPs, international contracts, and foreign affairs.

On Mondays and Tuesdays, the working groups worked separately on their topics in closed meetings. On Wednesdays, the results of their work were presented to the delegates from the other working groups. On Thursdays, their recommendations were introduced at open Council meetings. The recommendations were then put on the Council’s website and were entered into a progress document in which the public could participate and provide comments on the delegates’ recommendations online through Facebook’s ‘reply system’. This progress document is important, as it is here that forms of deliberation were detected. This progress document was an ‘open’ document. When the Council meeting accepted propositions, these were placed in the progress document, and the public was again given an opportunity to comment. This meant that, even though recommendations had been placed in the progress document, they were not finalised until the Council accepted them at the very end of the process.

The Constitutional Council’s external communications clearly emphasise the importance of ensuring that the public is involved and kept updated on the Council’s progress. The Council’s website was thus designed to enable people to monitor developments in and comment on the progress document. Furthermore, the Council allowed people to send requests, for instance via email, which then were posted on the website, concerning which people could respond using Facebook’s reply system (as well respond using services supplied by Yahoo!, AOL, and Hotmail). It is important to note that the Council’s own website facilitated the discourse not through a system of its own but instead through commercial social media systems, mainly Facebook, particularly bearing in mind Facebook’s data use policy and statement of rights and responsibilities (Valtysson, 2012), as these affect processes of deliberation.

All things considered, however, it is safe to say that the Council attempted to construct an open and transparent process that greatly encouraged interaction through social media. Short interviews with delegates were put on YouTube and Facebook, and on Thursdays, the Council transmitted live broadcasts of the meetings. The Council was also very precise in providing schedules and minutes for meetings of the Groups, the Board, and the Council, and was clear regarding the work procedures. Additionally, the Council advertised in the media, encouraging people to keep track of the process and participate in it.

These communications do not, however, stand alone since they are always embedded in the multichannel and multimodal communication network of the internet as well as being situated within the convergent ‘dissemination loop’ of the mainstream media. Communication on the Council’s Facebook profile thus cannot be isolated from communication taking place on YouTube or on the Council’s own website inasmuch as these streams of information are interlinked. The different affordances of the social media in question do, however, affect and mould the nature of these discussions. When these are regarded within a wider framework of cultural and political public spheres, opinion formation, and decision-making, the empowering dimensions of the ‘crowdsourcing of the Icelandic Constitution’ are placed in a more critical light.

Citizen engagement in the informational hurricane

There has been a degree of conventional optimism regarding the participatory potential of the internet (Castells, 2001; Grossman, 1995; Negroponte, 1996; Poster, 2001), particularly in terms of economic transformation, global understanding, and democracy (Curran, 2012). From the perspective of participatory democracy, this is to a certain extent true. As the case of the rewriting of the Icelandic Constitution shows, digital communication and digital media provide multiple channels of communication that enable both ‘senders’ and ‘receivers’ to communicate in different ways. These kinds of communications are multichannel and multimodal as they refer both to different technologies of communication and to the organisational arrangements of the sources generating these kinds of communications (Castells, 2009). Indeed, the communication in question undergoes constant transformation depending on the inherent affordances of given digital media services, such as Facebook, YouTube, and Twitter, as well as on the people sending, receiving, and transforming these communications.

These ‘people’ or ‘users’ go under different names, such as produsers (Bruns, 2008), creative audience, users-turned-producers (Castells, 2009), prosumers (Toffler, 1980), and interactive audience (Jenkins, 2006), implying fundamental changes in the relationship between senders/producers and receivers/consumers/users. This sort of ‘communicative landscape’ is characterised as a convergent media landscape, referring to mediatising convergence, industrial convergence, technological convergence, regulatory convergence, and Web 2.0 and network convergence (Dwyer, 2010). Castells follows a similar line of thought in linking the mass self-communication of the creative audience to the concept of convergence as the communicative subjects ‘integrate various modes and channels of communication in their practice and in their interaction with each other’ (2009: 135).

This was certainly the case with the rewriting of the Icelandic Constitution as the inherent affordances of the different social media constituting venues for communication were strategically used to encourage discussions. Furthermore, these communications were integrated into the different modes and channels of communications characterising the contemporary convergent online and offline media milieu. Generally speaking, we see this every day in the communication hurricane of broadcasting, narrowcasting, and the interlinking of ‘mass’ media, social media, and other forms of communication enabled by the network society’s informational infrastructure. This does not, however, say much about whether such communication provides viable, meaningful, and empowering communication channels that encourage increased participation and deliberation in democratic societies.

As Dahlberg (2001) points out, if the public sphere is to be extended through the internet, it must develop spaces suitable for deliberation as well as attract citizen participation. This is in line with Dahlgren (2009), who argues that, if this kind of citizen participation is to ‘take off’, it must reach down to the ‘cultural realm’ of the lifeworld, with its cooperative necessities of communicative action and the transmission and renewal of cultural knowledge, cultural reproduction, social integration, formation of solidarity, and personal identities (Habermas, 1984: 137–138). These kinds of engaging, participatory communications taking place in the cultural public spheres of the lifeworld are affective and ‘help people to think reflexively about their own lifeworld situations and how to negotiate their way in and through systems that may seem beyond anyone’s control on the terrain of everyday life’ (McGuigan, 2005: 435). This is not to say that these kinds of communication do not take place within political public spheres but rather that, as in Habermas’s (1989) original blueprint, these processes in the cultural public spheres serve as pre-conditions for processes in the political public spheres.

Dahlgren maintains that engagement and participation are key factors in generating these kinds of communication and that they are closely linked to civic agency: ‘For engagement to become embodied in participation and thereby give rise to civic agency there must be some connection to practical, do-able activities, where citizens can feel empowered’ (2009: 80–1). When citizens engage and participate in the process of rewriting the Icelandic Constitution, they form collectives with other citizens and connect with them through civic bonds, which Dahlgren (2009) regards as laying the foundations for vibrant civic cultures. These civic cultures represent six dimensions (knowledge, values, trust, spaces, practices, and identities), all of which are important in the multichannel, multimodal media environment that was created to facilitate the rewriting of the Constitution. Access and interaction are significant in this context since the communication channels applied by the Constitutional Council create space and invite the public to use that space to promote their preferred democratic values, exchange knowledge, and put these into action through concrete practices. However, as Carpentier remarks, ‘access and interaction remain important conditions of possibility of participation, but they cannot be equated with participation’ (2011: 354).

This is an important point since – even though the Constitutional Council created this specific media environment, drawing on the diverse affordances of stjornlagarad.is, Facebook, YouTube, Flickr, and Twitter – numerous hurdles have to be overcome on the route from ‘speaking’ to being ‘listened to’ and seeing one’s ideas materialise in the ‘upper’ political public spheres. Although the internet provides manifold streams for multichannel and multimodal communication, there is no guarantee that a message will move successfully from the cultural public spheres to the political ones. Indeed, critical voices encourage toning down the ‘often-celebratory enthusiasm displayed by many authors regarding the participatory potentials of the Internet’ (Cammaerts, 2008: 372) and instead conceptualising user agency in its full complexity. Here, van Dijck advocates taking into account the ‘cultural role as a facilitator of civic engagement and participation’, the ‘economic meaning as a producer, consumer and data provider’, and ‘the user’s volatile position in the labour market’ (2009: 55).

This is extremely important since the media milieu created by the Constitutional Council is grounded in a convergent space in which commercial interests – in the form of Facebook, Twitter, Flickr, and YouTube – intertwine with ideals of deliberative democracy in Habermas’s tradition, with participants reaching consensus through communicative action. As will be evident in my analysis of the communication conducted on stjornlagarad.is and the affiliated social media, much emphasis was placed on such deliberation. The interesting question, however, is whether these communicative acts were taken into consideration by the Constitutional Council and in the subsequent processes in the political public sphere or whether these were always constituted as weak publics and remained so, even after having been taken into account by the Council.

Networked publics in online participatory cultures

In later works, Habermas applies the metaphor of the network to describe the formation and communication carried out by publics within different public spheres. He maintains that in ‘complex societies the public sphere consists of an intermediary structure between the political system, on the one hand, and the private sectors of the lifeworld and functional systems, on the other’ (1996: 373). These structures form complex networks of overlapping areas – be they international, national, local, or subcultural, segmenting the public sphere into the episodic, the occasional, and the abstract. According to Habermas, this differentiation depends on density of communication, organisational complexity, and range.

This sort of differentiation of public spheres and the publics constituted within such spheres is useful and conforms with what Fraser terms post-bourgeois public spheres, indicating the adaptable characteristics of public spheres in constituting communication within and between different publics. Corresponding with Habermas’s (1989) initial distinction between cultural and political public spheres, Fraser (1992) refers to strong and weak publics. According to Fraser, strong publics generate discourses that include both opinion formation and decision-making whereas weak publics generate practices of opinion formation. This is an important distinction that can be transferred to the communication that takes place on the online forums in the process of rewriting the Icelandic Constitution. As will become clear in the analysis, the comments, discussions, and attempts at deliberation generated by citizens are indeed constituted as weak publics since power remains in the hands of the 25 delegates on the Constitutional Council, which is responsible for inserting these ideas into the Constitutional bill. The situation is further complicated by the Constitutional Council’s own status as an intermediate public between the weak networked publics, constituted by the produsers/users-turned-producers/interactive audience, and the strong publics, constituted by the decision-makers in Parliament. These ‘users’, constituted as weak networked publics, galvanise their civic agency to engage and participate in the rewriting process, seeking to influence the Constitutional Council’s work and have their views represented. However, the same can be said for the Constitutional Council itself, which can do no more than present the actual decision-makers in Parliament with a bill for a new Constitution.

Consequently, in this case, there are three different publics: weak networked publics acting mainly within the online communicative spheres of stjornlagarad.is, Facebook, YouTube, Flickr, and Twitter; the intermediate public consisting of the 25 elected delegates in the Constitutional Council, who, despite their legitimacy, are still considerably weakened after the invalidation of the elections for the Constitutional Assembly; and, finally, the strong public of the elected MPs who had been given a clear decision-making mandate in the previous parliamentary elections.

All of these publics operate within different though sometimes overlapping public spheres that differ in density of communication, organisational complexity, and range. At this point, it is important to recall that public spheres themselves do not act or communicate but, rather, that communication is carried out by publics operating within these specific communicative spaces (Splichal, 2010). These communicative spaces do, however, tailor the communication inasmuch as they possess different affordances, which can be said to produce different discourses, comments, and in some cases deliberation. When scrutiny is granted to stjornlagarad.is, Facebook, Twitter, Flickr, and YouTube, as well as to the multimodal, multichannel communication that flows across these services, it becomes quite evident that they all produce specific forms of communication and interaction between citizens and between citizens and members of the Constitutional Council. In all cases, this interaction and communication constitutes networked publics since it simultaneously institutes space and the collection of people (boyd, 2011) – hence my referring to these publics as weak networked publics. According to boyd, these share certain characteristics in terms of persistence, replicability, scalability, and searchability. Furthermore, their central dynamics are based on the invisibility of audiences, on collapsed contexts (lack of spatial, social, and temporal boundaries), and on the blurring of the public and the private (boyd, 2011). While these characteristics of networked publics can be beneficial to the dissemination of information, to the multimodal and multichannel transformation of information, and to this information’s accessibility, these kinds of information streams can also be detrimental to democratic deliberation due to the lack of legitimacy, invisibility of audience, and collapsed context such publics constitute.

When stjornlagarad.is, Facebook, Twitter, YouTube, and Flickr institute networked publics at the intersection of people, technology, and practice, they create specific media environments with specific affordances that generate specific forms for communication. As will be illustrated in the next section, the Constitutional Council repeatedly emphasised the importance of rational deliberation and sought consensus through communicative action, in the Habermasian spirit. The Council succeeded up to a point, at least in its communication as an intermediate public. Furthermore, the networked publics that participated in rewriting the Constitution online respected the ‘communication pact’ made by the Council, and although much of the communication can be thematised as taking the form of statements, rather than of deliberation, the tone is courteous and generally positive.

However, it is quite clear that the inherent affordances of, for instance, Facebook (the most used social media service in the rewriting process) are quite limited when it comes to realising the Habermasian discourse model, which already in Habermas’s early writings takes as a given the disregarding of status, the object of common concern, and full accessibility for all (Habermas, 1989). Furthermore, it is not evident how the affordances of the social media in question facilitate the common rationality of communicative action in which ‘participants in communication encounter one another in a horizon of unrestricted possibilities of mutual understanding’ (Habermas, 1987: 149). This is certainly not the case when we consider the accessibility, user terms and conditions, data use policy, and user manoeuvrability of a service like Facebook (Valtysson, 2012) since the networked publics constituted within Facebook are significantly limited by the service’s affordances. This is also the case for YouTube, Twitter, Flickr, and stjornlagarad.is.

Nevertheless, the Constitutional Council clearly related to the tradition of deliberative democracy as it repeatedly reiterated the importance of rational consensus based on conditions of ideal discourse. Mouffe describes the contours of such discourse as follows: ‘[T]he more equal and impartial, the more open that process is and the less participants are coerced and ready to be guided by the force of the better argument, the more likely truly generalizable interests will be accepted by all persons relevantly affected’ (1999: 748). The question remains whether these social media sites offer spheres or surroundings in which the constituted networked publics can reach deliberation as ‘an inter-subjective performative process that involves the transformation of privately-oriented selves into publicly-oriented “citizens”, and pre-deliberative positions into critical-reflexive public opinions’ (Dahlberg, 2007: 50). Alternatively, do the uneven power relations between the weak networked publics constituted online, the intermediate publics of the Constitutional Council, and the strong parliamentarian publics in the political public sphere generate inter-discursive contestation that more clearly address issues related to normalisation, exclusion, and cultural power struggles (Dahlberg, 2007)? The latter view is agonistic as it places political struggle and conflict at the crux of democracy, a view that Mouffe (2000) advocates with her emphasis on conflictual consensus.

Democracy in disguise

The networked publics constituted in the interaction of people, technology, and practice in the process of rewriting the Icelandic Constitution are significantly influenced by the affordances of social media. However, as already mentioned, the converging characteristics of multichannel and multimodal communication provide spheres in which form and content travel and change from one service to the next. This is quite clear in the case of the videos screened on YouTube, which in many instances also appear on the Council’s Facebook profile, and the same applies for the photos on Flickr, which generate different responses depending on the contexts and affordances of the different social media services.

When these social media are considered individually, the 51 videos uploaded on YouTube cannot be said to have generated much communication, discussion, or deliberation. Indeed, the videos, of which the vast majority are short interviews with Council delegates, only receive around 100 views each, and there is hardly any discussion in the comments section. The same is true for the Council’s Flickr profile, to which 522 pictures were uploaded, consisting mainly of snapshots from the daily work of the Council members and from the formal ceremony at which the delegates presented the bill to the President of the Parliament. Even though Flickr’s affordances have often previously generated vibrant communities, this was not at all the case for the photo stream provided by the Council. Indeed, it did not generate a single comment and therefore did not facilitate any kind of conversation, let alone deliberation or antagonism. When these pictures appeared in the context of the Council’s Facebook page (for instance, when the bill was handed over), they still generated neither much ‘liking’ nor many comments. Here, the photo that generated the most comments only had 38 ‘likes’ and 9 comments from the public, all of which congratulated the Council for its fine work (Council’s Facebook profile, 29 July 2011).

Twitter has on many occasions served as a convenient service for disseminating information, facilitating multichannel and multimodal communication, and organising offline activities, particularly when democratic ruptures have taken place. The activity on the Council’s Twitter profile does not, however, confirm such usage, with only 171 tweets posted during the rewriting process. As of 18 May 2012, the profile had only 552 followers. Twitter’s affordances do not facilitate lengthy deliberations, as each tweet is limited to 140 characters. Therefore, much of the communication can be characterised as practical meta-communication in which the Constitutional Council provides information on the process, often in a quite detailed manner as the following example indicates: ‘16th council meeting will start in 5 minutes’ (Council’s Twitter profile, 12 July 2011) and the next day, ‘16th meeting continues’ (13 July 2011).

Twitter’s affordances are quite useful when it comes to ‘spreading the word’, particularly within the multimodal and multichannel convergence of social media. The Council made use of these features by advocating for Council members’ appearances in other media, providing links to external sites such as CNN and to its own site, stjornlagarad.is. As mentioned above, these two themes – the ‘dissemination loop’ and ‘practical information’ – emerged through the inductive category development conducted on the data material on the social media sites in question, as did ‘exclamations’.

The fourth theme that surfaced is what I refer to as the ‘deliberation–statement axis’, facilitated primarily through the Council’s Facebook profile and the design of the Council’s own website, stjornlagarad.is. Although it is debatable whether deliberation in Habermas’s sense can take place through social media such as Facebook, which capitalises on the communication of its ‘prosumers’ and targets them strategically as consumers, Facebook at least provides space for more detailed discussions than do Twitter, YouTube, and Flickr. Furthermore, Facebook is the most popular social media service in Iceland, meaning that large portions of the Icelandic public are familiar with its interface and general functions. Facebook links directly to stjornlagarad.is as the commentary system on the Council’s website is that of Facebook. Indeed, most discussions take place on stjornlagarad.is (through Facebook’s commentary system), particularly during stages when the Council’s progress document had taken form and people could comment directly on specific articles in the Constitutional draft.

However, even though it is quite clear that the affordances of the various social media generate different kinds of communication, the inductive category development and the consequent deductive category application (Mayring, 2000) could not be limited to the affordances of the services alone. The reason for this is twofold: first, in a convergent media landscape, social media do not stand alone but constantly provide multichannel and multimodal links to other digital media; and, second, the identified themes cut across the various social media. Some of these are more prominent on stjornlagarad.is while others are more prominent on Facebook or Twitter. This is why I construct the analysis in the manner described above.

The constitution of networked publics

In the case of the rewriting of the Icelandic Constitution, networked publics were constituted at the intersection of people, technology, and practice, and the affordances of the social media in question shaped how people engaged with and participated in the process. These affordances worked well in the context of providing practical information, disseminating information in the media, and responding to positive exclamations. However, when addressing matters of more substantial Constitutional relevance, the affordances provided by social media came up short, despite obvious attempts from Council members to respond to and conduct discussions. Discussion, however, is not the same as deliberation. Indeed, in the Habermasian tradition, deliberation results in rational consensus. In most cases though, the threads and comments originating in the social media come up short as they are never fully deliberated. Indeed, they usually end with statements such as ‘We discussed this in the Council and decided not to…’ or ‘Thank you for your comments, we will look into it’ or are not responded to at all. Again, the tone is always polite and, surprisingly enough, there is little of the expected antagonism that Mouffe regards as necessary for democratic debate. However, the social media do not host the kind of deliberation that Habermas envisages. Instead, there are open processes, a welcomed dialogue that, at its best, inspires Council members to take proposed issues under consideration – but at its worst, is countered by utter silence.

In either case, these networked publics are constituted as weak publics since the decision-making is never in their hands. The Constitutional Council itself does take on this form of deliberation, however, and emphasises it clearly in all of its external communications, for instance in the videos on YouTube. It is maintained that consensus would be attained through rational deliberation. Indeed, provided with the information obtained online through the various social media services, and given the structure and working methods that the Council set for itself, this was much akin to the open rationalised process in which people reach consensus ‘guided by the force of the superior argument’. However, this was not the case in that, despite engaging and participating (activating civic agency through the six dimensions of civic cultures discussed by Dahlgren), the general public could never assume empowerment since there was no guarantee that its voice would be heard, let alone implemented in the bill.

After the Council had formally handed the bill over to the President of the Parliament, it was decided to call for an advisory national referendum on the Council’s propositions. On 20 October 2012, voters were asked whether they supported the Council’s propositions as foundations for a parliamentary bill as well as five other more detailed questions concerning the Council’s propositions. Of the eligible population, 49% voted, with 64.2% wishing to use these propositions as the foundations for a bill on a new Constitution. Since the vote, MPs and members of the government have interpreted these numbers differently, and the aftermath is currently being debated in Parliament. This again highlights that, compared with the strong decision-making public in Parliament, the Council itself can be constituted as a disempowered intermediary public.

It is thus inaccurate to say that this is, or will be, the Constitution of networked publics. Nevertheless, networked publics were most certainly constituted in the processes of rewriting the Icelandic Constitution. These publics were, however, weak in Fraser’s sense, as they encompassed opinion formation but not decision-making. Indeed, this could be an ideal manifestation of ‘democracy 3.0’, as people engage to participate, with no one really knowing whether this kind of participation has any effect, both because of the inherent affordances of the social media in question and because of the institutional mechanisms that ultimately decide the fate of the reviewed Constitution. As Carpentier puts it: ‘In this sense the expectation that participation in the media is a privileged channel to allow for participation in society can only be considered a naïve fantasy that ignores the complexity of the polis’ (2011: 355).