Citizens’ Deliberation Online as Will-Formation: The Impact of Media Identity on Policy Discourse Outcomes in Russia

The notion of media’s public identity – understood by this paper as a real or perceived image of a particular media resource – is not necessarily included into the process of examining and explaining deliberation practices online. As far as the public politics and will-formation are concerned, the endless proliferation of media resources that enable public conversations demands greater clarity about the impact of their political allegiances on discourse outcomes. In this paper we attempt to investigate whether position-taking by discourse participants is linked to the media type, especially in terms of endorsing or rejecting government policies. It is assumed that the process of political participation requires taking sides and thus forming (joining) certain solidarities among participants. Being a member of a political party or any other organised entity pursuing certain public goals typically means solidarity with the like-minded peers, particularly in the face-to-face setting. However, participation in online discourses, often anonymous, is a different type of social and personal experience providing more flexibility in choosing a preferred debating community. It is assumed that core political preferences, allegiances and motivations do not change quickly. Yet, ideally, the virtual environment of online conversation may help citizens compare differences, fine-tune the established positions and change them altogether, as a result of talking to the peers. It also implies that when a participant articulates a similar opinion on something that has already been uttered by someone else, he or she joins a virtual ‘solidarity’ with that participant at this discourse moment; this is something that Jürgen Habermas defines – within the framework of his discourse ethics theory – ‘inter-subjective solidarities’ built to share the common values expressed in the public sphere.

We consider online discussions among ordinary Russian citizens as distinctive policy discourses in relation to whether or not the Russian authorities are right to destroy the embargoed food imported from the West.^{Footnote 1} Many digital discourses have emerged on the internet debating both the relevance and morality of food destruction policy, with many of them aiming to stop or amend this policy by, for example, launching e-petitions [1]. The latter have become an especially popular e-participation tool in Russia after the start of the official Russian Public Initiative [2, 3] following in the footsteps of the European Citizen Initiative. We view the act of joining any e-petition or contributing to online discussions as a manifestation of political participation – a public activity undertaken by ordinary citizens voluntarily to influence political authorities and decision-making [4: p. 2], [5: p. 120]. Given the growing accessibility of digital networking for ordinary citizens, opportunities for their political activism increase as well. In addition to the importance of studying policy discourses as such through the lens of public deliberation in order to understand their course and outcomes, the expanding diversity and richness of new digital media suggests taking a closer look at the impact of the media itself on discourse outcomes; that is to say, whether certain solidarities formed on certain issues differ across different media? Whether or not discourse participants prefer those media that are closer to their worldviews? Whether or not participants could be encouraged rather than discouraged joining alternative discourses advocating other, opposing views? Knowing answers to these questions may help lower the level of political polarization that inevitably increases when people prefer talking to their like-minded peers and thus further strengthen the beliefs they already have instead of enlarging the space for potential compromise and dialogue.

Judging by high approval ratings of President Putin by the public,^{Footnote 2} Russia does not seem to be a politically polarized society. However, the lack of competitive politics coupled with a short and not always fair history of political representation does not mean that the ordinary citizen is uninterested in public life and policy-related issues. Despite the limitations of official politics, informal citizen initiatives based on social networking are as vibrant in Russia as anywhere in the connected to the internet world.^{Footnote 3} It is assumed that political cultures dwell on prevailing civic cultures and comprise inseparably both individual competencies (such as knowledge, beliefs, feelings, judgements) that reflect the society’s prevailing norms and values and the personalized attitudes towards the official political system (political allegiances and preferences) that citizens articulate in the course of civic activism and political participation [6].

3.1 Deliberation as Civic Culture

Just as civic and political cultures overlap, the conceptualization of public deliberation is linked with the concept of political participation and democratic legitimacy of liberal democracy. However, Jürgen Habermas viewed the public sphere as a common communication space exclusively for citizens’ life-world which is separate from the state’s ‘system’ occupied with its strategic communication mission (i.e. propaganda), as opposed to free and uncoerced communication flow among equal members of the public. Yet the even highly idealized public sphere concept has a distinctive civic character free from the state influence (e.g. a need for independent media) as a condition to realise democratic deliberation. For Habermas, (rational) deliberation constitutes the core of deliberative democracy. Governing regimes can only be morally and ethically justified if citizens are engaged in truth-tracking moral discourses to understand others [7: p. 52]. The role of public discourse is to ‘uncover topics of relevance to all of society, interpret values, contribute to the resolution of problems, generate good reasons, and debunk bad ones’ [8: p. 452]. Such discoveries are realized discursively, i.e. communicatively and collectively in the course of public discussion by making three types of basic claims to: (a) objective (propositional) truths, i.e. when the speaker refers to the ‘totality of objects or existing states of affairs’ in the objective world; (b) shared values (intersubjective normative rightness), i.e. when the speaker refers to the totality of shared interpersonal relationships of social groups; and (c) individual experiences (subjective truthfulness), i.e. when the speaker refers to the totality of his or her personal world [9: pp. 313–314], [10: p. 52]. These are the explicit speech acts carrying certain intentions. Using the real case of The Jerry Springer Show [11: p. 393], Habermas demonstrates how through claim-making a communicative action becomes ‘meaningful when it engages with high moral standards’.

Such discursive practices are not only defined as communicative – because participants coordinate their actions consensually via validity claims; they are also rational, because communication actors seek to motivate reciprocal discursiveness instead of influencing each other through less honest ‘strategic communication’ actions that are often accompanied by threat or coercion rather than consensual communication intentions [7: p. 58]. Being a rational discourse participant means being a responsible, reciprocal discussant. Rationality is not necessarily a formulaic robot-style exchange of arguments and reasons. It is also emotions, values and motivation. By claiming values and truths as they understand them, citizens demonstrate their communicative rationality ‘in the form of uncoerced and undistorted interaction among competent individuals’ [12: p. 12].

While many agree that citizens as discourse participants should cooperate in pursuing together the morally justified common good, the terms of such cooperation are not always clear [13: p. 9, p. 27]. For example, there is a lack of clarity with regard to the role of consensual practices. Gutmann and Thompson [13, 14] resist the traditional Habermasian emphasis on the discursively accomplished consensus. They believe that there are instances when participants cannot agree, ‘no matter how respectfully they deliberate with their opponents’, and no matter how morally strong the opponents’ positions may be [13: p. 20]. There could be ‘deliberative disagreements’ that despite, for example, moral differences, discussants still can find a mutually acceptable solution by disputing one another’s position in a reciprocal manner; but there could also be ‘nondeliberative disagreements’, when positions are mutually exclusive and could not be reconciled through reciprocal discourse. In many cases, people will not change their opinion, and therefore discussions will not ‘always or even usually’ end in agreement [13: p. 7].

3.2 Rules and Principles of Deliberation

Internet-based discussions this paper investigates are viewed through the concept public deliberation. There are two main notions of deliberation. The ‘procedural’ concept argues that for a discussion to be qualified as deliberative certain strict conditions must be met as far as the course of deliberation is concerned [13, 15]. The other position does not prescribe any specific rules that the discussion should follow in order to be counted as deliberative [16–19]. Joshua Cohen [15, 20] represents a more cautious position by emphasising certain conditioning principles for deliberative politics such as: (a) independence (from authorities) and continuous character; (b) equality of participants, who must agree on the ground rules of deliberation; (c) commitment to mutual cooperation; (d) diversity of participants and pluralism of opinions, but excluding pronounced political or ideological biases; (e) respect for internal deliberative procedures (as the source of legitimacy); (f) ability to deliberate competently; (g) use of reason and better argument; and (h) adherence to decisions by rationally-motivated consensus (voting is accepted if no consensus is possible). Gutmann and Thompson [13] also add the importance of deliberation scale and agenda, since only large-scale public debates discussing publicly important policy issues could count as deliberations.

In contrast, a more loosely defined approach toward deliberation admits the inclusion of smaller-scale discussion events, such as ‘deliberative polls’ that can discuss any issue relevant to a particular community of citizens [17]. Such discussions are usually moderated and short in duration. Diana Mutz [19] supports a more flexible interpretation of deliberation and objects Cohen’s procedural model as too ‘all-encompassing’. She argues that such an idealized and strictly conditioned deliberation would be virtually impossible to realize in practice as, for example, routine political conversations between neighbours, family members or co-workers would be unjustly excluded [18, 19]. For Bohman [16], one of the major weaknesses of procedure-based deliberation lies in the lack of genuine interest in debate, which would demand the demonstration of strong commitment on the part of prospective discussants. The bias toward the procedural side of deliberation may, for example, lead to the rationally produced argumentation that is non-dialogical and lacking the potential benefits of knowledge sharing and mutual learning. Hardly the focus on deliberation at the expense of the reciprocal ‘mutual understanding’ could result is a successful discourse, for deliberation should be viewed as a ‘particular social activity that can be performed only through public discourse’, and hence is ‘imbedded in the social action of dialogue’ [16: p. 32]. In this respect, Gutmann and Thompson [13: p. 3] agree that deliberation can be both procedurally strict and dialogic. The main challenge, however, is striking the right balance between the both sides of deliberation.

This paper supports a more flexible idea of deliberation as a civic and political activity, since it fits better the nature and character of online discussions among ordinary citizens. Through discursive interaction, they are capable of assuming the role of rational discussants who are willing to debate difficult and sensitive public issues in a dialogical manner. The typical participant is also often emotional and sometimes uncivil. Expressiveness is not necessarily a drawback in peer-to-peer discussions; it is rather an objective discourse trait helping to maintain communication and overcome disputes in a casual conversational manner. Citizens are well aware of what and how they are communicating. They also appreciate the importance of public debate and take participation seriously as a special civic activity and at the same time as a political participation act.

4.1 Research Assumptions and Objectives

As mentioned above, the study’s main research question is to assess the role of the media in will-formation by hosting online discussions; that is, whether a particular media identity – its type, public status, ownership, political allegiances – has an impact on the quality and outputs of online discources. Indirectly, knowing the answer to this question could provide some insights into (a) the degree of awareness of discourse participants about the media’s role and hence their decision to choose a certain media outlet (e.g. in order to further strengthen the already existing personal political preferences among the like-minded people); (b) knowing more about whether such an awareness might prevent them from participating in discourses that advocate opposite points of view on the publicly important issues that divide and polarize citizens? The topic of food destruction was found a highly dividing issue in August 2015 among the public in Russia and abroad.

The concept of public deliberation refers primarily to the pre-digital, offline context when participants debate issues in the face-to-face mode which is highly different from the anonymous environment of internet discussions. Face-to-face public discussions were championed by James Fishkin [17, 22] in the form of ‘deliberative polls’ in the United States. But Diana Mutz [19] asks in this regard whether routine, real-world political activism can be truly deliberative, since people usually prefer to be ‘surrounded by those who agree with them (and) thus reinforce… their own political views’ since such an agreeable environment does not necessarily encourage political discussion [19: p. 3]. Bohman [16: p. 25] believes that genuine deliberation requires a more diverse and heterogeneous setting to become meaningful; in addition, it should be impersonal to be truly public (and thus more democratic). Individuals tend to exchange their opinions more eagerly if they are unaware of other participants’ social identities. It is then assumed that in online debates citizens may prefer ‘to debate with diverse others’ [21: pp. 2–3, p. 31], contrary to face-to-face discussions. If so, the very deliberation context might be crucial for determining its outcomes. Settings where participants are anonymous might diminish possible intimidation or confusion created by status, appearance or group pressure. Another important question that arises in this context is whether discussion can help change established opinions through a more flexible mode of side-taking.

Drawing on these considerations, the research hypothesised that in an online environment, discourse participants would tend to choose that media resource which has a distinguishable identity in terms of whether it supports or opposes their own views regarding government policies. The logic behind such a hypothesis was grounded in the assumption that people would like to see their beliefs and opinions – for example, a position supporting the government’s policy of food destruction – reinforced by other like-minded peers rather than diverse others; that discussions on pro-government media would approve of destroying the embargoed food, whilst the discources hosted by independent media would instead say “No” to such policies.

4.2 Research Sample

According to Denis McQuail, there are no commonly agreed media typology [23: p. 29]. As a general entry point, this study considers three main types of media by applying the following criteria of audience and thematic focus: (a) partisan, political media, (2) elitist media, and (3) the mass media. Pippa Norris [24] also mentions such characteristics, as coverage (national, regional, local), publishing frequency (daily, evening/Sunday), content quality. Media ownership plays a key role in determining thematic orientation and political allegiance. A clear focus on promoting specific content and involving particular contributors determine to a large degree the main target audience and dominant topics [38: p. 238]. Loyalty to the current political order and governing elite – either openly when the media is owned by the government or implicitly when it controls the owners – has a direct impact on the media’s editorial policy and target audience [39: p. 405]. That is especially important for a democratically immature society where the insufficient representation of certain groups in government and hence their inadequate participation in decision-making may demand additional and often informal communication channels to express their views. Digital media, with its many-to-many communication capabilities, provides such channels and somewhat offsets the democratic deficiency of the official public communication system [25]. Even for a more democratic western society, there is a challenge of overcoming the excessive reliance on mediated politics that diminishes the participatory potential of public communication instruments [26–29]. Based on the above considerations, as well as taking into account the research focus on examining the will-formation process as citizens’ reaction to government’s policy of food destruction, the presented study was designed to examine the following seven internet discussions that formed the research sample, namely:

1. 1.

   E-Petitions site Change.org, part 1 e-Petition ‘#Don’tCrushFood’; 100 first posts coded; 7 August 2015; media identity type – international, independent, language; Russian.

2. 2.

   E-Petition site Change.org, part 2 ‘e-Petition Don’tCrashFood Update: We will achieve our goal! Preparing a conference (Mы дoбьeмcя peзyльтaтa! Гoтoвим кoнфepeнцию); 77 last posts coded (out of 77); 30 August 2015; media identity type – international, independent, language: Russian.

3. 3.

   Web site Business newspaper Vzglayd (Дeлoвaя гaзeтa “Bзгляд”). Lead article title: Polish Minister writes to Putin calling food destruction a sin (Пoльcкий миниcтp в oбpaщeнии к Пyтинy нaзвaл yничтoжeниe пpoдyктoв «гpexoм»); 7 aвгycтa 2015; 100 posts coded (out of 240); media identity type – Russia, openly pro-government; language: Russian.

4. 4.

   Russia Today TV channel (Poccия Ceгoдня, in Russian). Lead article: Dmitry Peskov (President Putin’s press representative) commented on the reaction following the destruction of embargoed food (Дмитpий Пecкoв пpoкoммeнтиpoвaл peaкцию нa yничтoжeниe caнкциoнныx пpoдyктoв); 5 August 2015; 97 first posts coded (out of 344); media identity type – Russia, government-owned; language: Russian.

5. 5.

   Novaya Gazeta (Hoвaя гaзeтa). Lead article: Authorities staged a risky experiment by publicly destroying food in a country whose people are still following a food cult (Ocытeнeли. Bлacти пocтaвили pиcкoвaнный экcпepимeнт c пyбличным yничтoжeниeм eды в cтpaнe, нaceлeниe кoтopoй вce eщe пoклoняeтcя кyльтy жpaчки); 6 August 2016; 100 first posts coded (out of 544); media identity type – Russia, independent, language: Russian.

6. 6.

   Playground – gamers’ forums (Playground, Бoлтaлкa/Oбщecтвo). Lead article: Destruction of embargoed food. Your opinion (Уничтoжeниe caнкциoнныx пpoдyктoв. Baшe мнeниe); 8 August 2015; 73 first posts coded (out of 74); media identity type – Russia, independent, language: Russian.

7. 7.

   The Guardian. Lead article: Russians despair at food destruction as Moscow says it is having desired effect; 7 August 2015; 100 first posts coded (out of over 2,000); media identity type – UK, independent, language: English.

The sample was randomly selected by searching the internet. The main selection criteria were as follows: (a) the dominant discussion topic must be food destruction; (b) there must be at least 70 messages posted by users after the reports; (c) at least one discussion must take place in a western country (in English); (d) presence of independent, pro-government and government-owned media outlets hosting the food destruction discussions; (e) at least one media outlet must not be connected with public politics in any obvious way.

In total, 647 messages posted by 417 unique participants on seven discussion threads were coded and analyzed to reveal the implicitly or explicitly expressed positions either supporting or rejecting government policy of destroying the embargoed food imported from the West.

4.3 Content Coding Methodology

The content coding method aimed at discovering the dominant positions expressed by discourse participants in their assessment of the government policy of food destruction. Specifically, the content analysis looked at the presence of the mutually excluding statements ‘For food destruction’ and ‘Against food destruction’. The important aspect of the validity-based method is that it requires to disclose the intended meaning contained in the uttered message, i.e. to reveal its illocutionary force in terms of the Austian and Searlean notions of intentional ‘speech acts’. The notion of the illocutionary speech act is borrowed from linguistics pragmatics as part of the philosophy of language and communication [30, 31]. For Searle [31: p. 16; 32: p. 45], just as for Habermas, ‘speaking a language is performing speech acts, acts such as making statements, giving commands, asking questions, making promises and so on’; this is done according to certain rules as an attempt to communicate an intention to obtain the hearer’s recognition of a certain assertion or to convince in the truth of a particular proposition. Illocutionary acts are based on the fundamentally dialogical relationship of mutual recognition between the speaker and the hearer.

The Habermasian claims to validity have a lot in common with the Searlean illocutionary force. The latter is increasingly applied in the argumentation and artificial intelligence (AI) research. The focus on argumentation underpins much of the existing scholarship on the topic of deliberation (both offline and online). In effect, argumentation is considered as deliberation itself according to Frank Fischer and Gerber Gottweis for whom deliberation is a ‘procedurally governed form of collective argumentation’ [33: p. 9]. They further argue that argumentation drives social relations and practices when ‘concepts and ideas…shape argumentation’ and turn discourse participants into knowledge agents ‘to give meaning to physical and social relations’ through which ‘people experience the world’ [33: pp. 10–11]. The AI research domain ‘argument and computation’ aims at establishing ‘inferences between propositions in the domain of discourse’ that should be revealed and recorded in the spoken utterances (speech acts) along with the understanding of the ‘means of persuasion used by discourse participants’ [34]. Argumentation is viewed – and analyzed – here as a special type of speech act which usually contain implicit inferences by which discourse participants influence and persuade one another in a dialogical manner. As a consequence, online deliberative processes are viewed as a series of argumentative dialogues that can be examined to map arguments by ‘argument structures’ created with the help of argument mapping software tools [35].^{Footnote 4}

Argumentation aspects are important for understanding the value of policy discourses for decision-making. In the similar vein, the claim validation method also includes revealing and documenting factual evidence and other forms of argumentation presented by discourse participants to make their claims more convincing. Such evidence has been collected and recorded in the process of content coding to ensure that the expressed statements and positions were indeed validated by discourse participants. However, the argumentation aspect of discourse analysis presented in this paper has not been its main objective. Moreover, this paper does not tend to support the view that exchanging arguments in a rational manner would be the main motivational factor of online discourses among ordinary citizens (that might be the case, though, in procedural debates among experts). Being a knowledge agent would entail more than presenting arguments. Citizens’ discourses are moral conversations in the first place. Studying the claims to normative rightness helps maintain the emphasis on morality and social values underpinned by the Habermasian ‘intersubjective solidarities’. It is believed that citizens judge government policies on ethical grounds as well, in addition to looking at the rational and argumentative sides of policy decisions.

5.1 Coding Steps

Each discourse was analyzed by coding the content of its messages as they were posted on discussion threads in a chronological order. As explained earlier, the coding aimed at identifying the Habermasian claims for validity of the second type, i.e. claiming the normative rightness of certain worldviews and positions. The coding process comprised the following steps. Firstly, each post was assigned a unique ID so as to connect it to the post’s author.^{Footnote 5} The second step included identification of the content type which would point at the presence of one of three basic claims to validity, i.e.: content about fact-based objective lifeworld known for all, something that is hard to dispute – validity claim type 1; content revealing value/morale-based statements about social worlds for some, not for all, something that is disputable – validity claim type 2; content about personal sincerity, something that is deeply individuals shared with others, a proof of openness – validity claim type 3. Types 1 and 3 were needed to be identified in order to discern more clearly type 2 – the main aim of content coding. The next, third step, focused on the validation of the claims made, i.e. to separate those that were actually responded via agreement or disagreement. As argued above, the presence of arguments was essential as part of the validation act. The latter did not mean direct responses alone, since the objective was to validate – by supporting or rejecting – the intended and often hidden meaning, its propositional (illocutionary force) the posts could contain. Posts with uncertain content were removed from the analysis, just as were personal messages that had little to do with the main discourse theme of food destruction. At the fourth stage, the validated claims were summarized to look like the moral statements about food destruction that were either disputed or agreed with by others. The list of such statements could be rather long depending how focused the discussion was. Behind each statement was a certain number of supporters exemplifying a ‘social weight’ attached to such statements as a measure of the statement’s public significance shared by other discourse participants. The final fifth step included further aggregation of the agreed-disagreed statements into a shorter list of the broader “For” and “Against” positions.

5.2 Dominant “For” and “Against” Positions

Tables 1 and 2 below summarize the aggregated position statements represented in The Guardian newspaper discourse (as an example).

In the case of the Guardian discourse, there were a total of 54 distinctive statements identified, of which 25 were articulated via agreement and 29 – via disagreement. The most prominent was a position against food destruction which supported by 11 validated claims. The most popular “For” statement was exemplified by eight validated claims supporting the government policy of food destruction. Each “For” and “Against” side supplied different sets of argumentation to substantiate their claims. Discourse participants touched upon a wider range of issues that were not necessarily related to the theme of food destruction alone; in fact, many of them did not discuss the main topic of the lead article that opened the discussion. While it is not the best example of online deliberation quality, nonetheless it is a telling case for this particular media resource.

5.3 Position Differentiation Across Media Type

Other discourses were analyzed in a similar way to reveal the dominant position statements on which discourse participants agree and disagree when discussing the government policy of food destruction and to determine the scale of such support and rejection. Table 3 contains the coding results across the entire discourse sample. These data reveal, first of all, that discourses on the pro-government media outlets Vzglyad and the Russia Today TV channel clearly supported food destruction (55 % and 72 % of all validated positions accordingly), whereas the discourses hosted by independent media rejected food destruction (with the rate of 80 % and above). This is an interesting case of opinion polarization suggesting that the media identity itself can be a dividing factor and can significantly, although not decisively, influence discourse outcomes. People indeed tend to participate in those discussions that generate positions similar to their own preferences, as it would happen in face-to-face discussions (according to [19]). On the other hand, these analytics also reveal that, in general, apart from the Russia Today TV channel (which was expected), the support to food destruction policies was rather low; even on the pro-government resources such support just slightly exceeds 50 %, while on the Playground Gamers’ forum is as low as 3 %. Otherwise speaking, the impact of media identity has its limits and should not be over-emphasized. The Playground Gamers’ discourse is especially interesting, as it is an apolitical media resource by design. Still its participants were unanimous in being against food destruction. One of the explanation of that could be the stronger presence of presumably young gamers than on other resources (it is also presumed that young people might be more critical towards authorities). This case also points at the earlier discussed close connection between civic and political activism.

These discourse outcomes point at a number of other essential for online deliberation factors, such as the focus on the discourse main topic which is measured by (a) deliberation reciprocity (“dialogicality”) – a percentage of all validated positions on all topics to all posts, and (b) deliberation consistency – a percentage of all validated positions related to the topic of food destruction in the total number of all positions on all topics. According to these measures, all discourses (with some exception of the discussions run on the Guardian and Gamers’ Forum) were dialogical and reciprocal, while the e-petitions Change.org media, the Russia Today TV channel and the Gamers’ Forum hosted the most focused on the main theme discussions displaying little “deliberation noise”.

On surface, media identity, both objectively and as perceived by participants, seems to determine the discourse outcome. However, in reality, it might be the case that discourse participants simply prefer a particular media outlet that they are comfortable with, that generates statements and positions close to their own worldviews. It is not clear, therefore, what would be the discourse outcome on other media were the participant audiences as heterogeneous as possible. That doubt leaves the study’s main research question not fully answered. More research would be needed to deeper examine the role of media in online policy discourses attracting a more diverse range of participants.

As argued above, the media type and its identity does have a role to play in determining deliberation outcomes. That is, the pro-government and openly government media outlets host discussions that are supportive to government policies and actions. In the same manner, independent from government control media is more critical towards authorities. However, the impact of media identity has its limits. The study results support the conclusions of those researchers who argue that, when it comes to a political conversation, citizens prefer talking to their like-minded peers than diverse others. Hence, those who support the government on some principled and moral basis – say for patriotism reasons – would turn to the media which identity is known to be, for example, pro-government. Participation in the discourses hosted by such media outlets would reinforce the already existing beliefs in favour of the authorities’ actions, even if they might seem unusual, such as food dumping into the soil by tractors. On the other hand, those who are critical towards the government on the differently understood moral grounds would more likely contribute to the discourses that reflect best upon their own worldviews. However, these are only the general observations. The discussion on the politically neutral Gamers’ Forum illustrates suggests that at least the spectrum of those criticizing the government is rather broad. We still don’t know what would be the discourse outcome on the media resources that are equally attractive for the bearers of opposing positions and whether, as a result, their positions would be swayed.

Additional studies are needed on other topics and more diverse media samples, including via focus groups and surveys, to examine the motivational side of participation in online policy discourses versus offline deliberation. Understanding better the relationships between civic and political activism in the digital realm remains an essential research agenda too.

This study has not been able to answer a great deal of other important questions, such as: how entrenched and unchangeable the articulated position statements are at the participant level; and to which extent participants themselves see their positions changeable and under which conditions they are ready to change them. Inability to answer these and other questions was a major limitation of the study. Yet we believe that the presented study contributes to the scholarship concerning online deliberation theories and practices by arguing that citizens’ casual conversations online represent legitimate policy discourses that can radically expand the level and scope of public knowledge (and specific expertise) needed for better policy-making [37].