Participation and deliberation : a case study of readers ’ comments to news stories on the Brazilian presidential campaign

The Internet has created new “dialogical spaces” (Oblak, 2005) where issues of common concern can be discussed, serving to democratize the public sphere. As a potential deliberative section and a civic forum, readers’ comments in newspapers’ websites constitute a locus for public debate and ideas exchange provided by the mainstream media. As a case study, this article intends to assess the quality of audience participation in online news sites, by analysing the readers’ comments in the news about the Brazilian presidential campaign (September-November 2010) in the online versions of two Portuguese newspapers.


The Internet and the public sphere
In recent years, the Internet has made a clear contribution to expanding the forms of public participation, while imposing some challenges on current journalism.Since it possesses in itself the potential to facilitate the relationship between the citizens and political decision makers (Dahlgren, 2005: 148), the Internet also changed the form of production and consumption of information (Scott, 2005: 92), altering, in a sense, the power balance between the sender and the receiver and the unidirectional communication models (Brants & de Haan, 2010: 412).
Several authors have underlined the potential of the Internet for public communication (and the subsequent consolidation of democracy), as well as for journalism itself and for the relationship that it establishes with its recipients.These authors have stressed, among others: the production of news loci for information, debate and interaction (Dahlberg, 2001: 1); the possibility of a more active participation in deliberation processes (Esteves, 2007: 220;Papacharissi, 2002: 11), thus contributing to the revitalisation of political communication (Coleman & Blumler, 2009: 10); or the extension of the social actors that intervene in the construction of the news story, as a result of the expansion of the locus of news production (Fenton, 2010: 10-11).Besides, when compared to the traditional mass media, the Internet, as a communicational environment, provides a wider range of possibilities for debate on issues of collective interest, creating new "dialogical spaces" (Oblak, 2005) and becoming itself a potential forum of political deliberation (Papacharissi, 2002: 11).
The embedding of the Internet in the traditional media therefore highlights the relevance of the deliberative democracy theory, as a conceptualisation of democracy that builds upon an ideal of political autonomy based on the practical rationality of the citizens.The public deliberation of these citizens has the purpose of influencing the political system decision-making process, while at the same time intensifying the vitality of democratic institutions (Dahlgren, 2006: 29).The Internet is then understood by those who propose the deliberative democracy "as the ultimate medium of promotion of the spaces required for rational debate, upon which deliberation is generated" (Ferreira, 2010a: 103).It furthermore has the possibility of establishing itself as a "sounding board" that has the ability to influence the diverse agents of the political sphere (Ferreira, ibid: 105;2010b: 107).
Although several studies have underlined the Internet democratization features, doubts remain as to the quality of the debates that it hosts, as well as to the actual possibility of challenging the power holders (Papacharissi, ibid;Coleman & Blumler, 2009;Dahlgren, 2011a).On the other hand, authors like Dahlgren argue that the use of the Internet for political aims is minor when compared to other activities (Dahlgren, 2005(Dahlgren, : 151, 2011b: 91): 91), such as entertainment, consumption, online chat or non-political networking (Redden & Witschge, 2010: 182).
However, the fact that the Internet provides additional spaces for political debate, stimulating a wider participation, cannot be disregarded.But do the online debates in fact promote a democratic exchange of ideas and opinions?To what extent -and howdoes this new debate environment introduce alterations to the state of the public sphere?
The work of Jürgen Habermas and his theories of the public sphere and rational communication (Habermas, 1994(Habermas, [1962]], 2001) has attracted additional attention with the emergence of new bottom-up communication tools made available by the Internet (De Keyser & Raeymaeckers, 2010: 2).Indeed, building upon the Habermasian concept of public sphere and upon the notion of communicative action, Lincoln Dahlberg proposes specific criteria to assess in detail the democratic quality of daily interactions (Dahlberg, 2004: 3).He also discusses to what extent the online discursive environment expands the critical-rational deliberation (ibid., 2001: 1-2).
Dahlberg thus suggests a typology of public sphere normative conditions, so as to face the afore-mentioned assumptions against the online practices, as a means to assess the effective realisations of the Internet to deepen public communication (ibid.: 3-15; ibid., 2004; 7-10).This includes: thematisation (arguments directed to those present in the conversation, as well as to all those that can be affected by the debate); reflexivity (to transcend personal choices, to critically analyse values, assumptions and interests); role taking (to be able to change their own beliefs based on the reasons presented by other participants); sincerity (effort to make visible all the information, intentions and interests that are relevant to the debate); formal equality (formal inclusion conditions that give the participants an equal opportunity to express their attitudes, desires and needs); and discursive equality (prevalence of the "best argument" or substantive inclusion conditions referring to the autonomy of the power relations within the discourse, which can be expressed, for example, by means of threats, dogmatic approaches or manipulation attempts), as well as an autonomy in relation to the State and the corporative power.
The potential advantages of the Internet to democracy still are, however, subject to much debate (Redden & Witschge, 2010: 182).In fact, the new technologies cannot unilaterally transform the nature of the political sphere.It can therefore be said that the Internet has contributed to improving citizen participation, but only to a certain extent (Papacharissi, 2002: 20).As a matter o fact, Robert Goodin highlights that posting a comment on the Internet is not, in itself, a deliberative act: "there must also be uptake and engagement -other people must hear or read, internalize and respond for that public sphere activity to count as remotely deliberative" (cited in Crawford, 2009: 454).
Several studies argue, furthermore, that some factors may operate as obstacles to the expansion of the online public sphere (Dahlberg, 2001: 1), suggesting that, "what is technologically possible may not be socially favourable" (Rasmussen, 2008: 75).Diverse authors have argued that political discussions in online discursive environments are known for their fragmentation (Morais, 2011: 225-226), as they are dominated by few participants (Strandberg, 2008: 82, Ferreira, 2010b: 109;Gomes, 2005: 221).They work as an extension of the inequalities that characterise offline public debates, and are too specific to meet the Habermasian criteria of rational consensus (Papacharissi, 2009: 10;Dahlgren, 2005: 152).Despite the inclusive potential of the Internet, which offers additional room for debate, segmentation prevails as a result of the domination of interest groups (Ferreira, 2010b: 110).Other problems are mentioned, such as the encouragement of communication between individuals with similar ideas (instead of promoting diversity) or the arousal of a sense of activity, instead of a genuine involvement, which can have an impact even on the participants' passiveness (Correia, 2011: 44).
Besides, wider participation opportunities do not necessarily correspond to the launch of a debate promoting the democratic ideals.Online communication is often marked by the expression of "rushed" opinions (Papacharissi, 2002: 16).Additionally, aspects like "flaming" (posts of abusive nature), "trolling" (posts that aim to misinform, deceive, poke, destabilize or trivialise the debate), or verbal violence make online discussions particularly vulnerable to disruption and, at the same time, challenge the normative requirements of tolerance and sincerity (Dahlberg, 2001: 7-10).
The anonymity of many debate participants also raises problems of verification, responsibility and accuracy (Fenton, 2010: 10), which can influence the deliberative nature of online discussions (Crawford, 2009: 458).Some authors argue that anonymity and the use of pseudonyms can encourage a more open debate, since the participants feel more comfortable to express their opinion (Wallace cited in Ferreira, 2010a: 110), as this can encourage the expression of those parts of the "self" that remain hidden in offline interactions (Danet cited in Dahlberg, 2001: 10) and suppress the fear of being excluded from the debate (De Keyser and Raeymaeckers, 2010: 9), while obliterating social differences.Other authors, however, claim that this provides an opportunity for a conscious identity fraud, which can undermine trust within online groups and lead to misinformation (Dahlberg,, as well as to an increasing number of verbal attacks (insult, humiliation), among other forms of incivility (Ferreira, ibid).This is a concept that can be defined as "gratuitous asides that suggested a lack of respect and/or frustration with the opposition" (Mutz & Reeves cited in Sobieraj & Berry, 2011: 20).
Other aspects besides anonymity can bring disadvantages to online debates: the posts and comments frequently focus on personal points of view, rather than reflecting a confrontation with other arguments or a response to the other participants (Ferreira,ibid: 111;Wilhem cited in Dahlberg,ibid: 3).The pressure to respond immediately can impose limitations on the participants' capacity to deepen their own positions in face of other participants' arguments (Dahlberg,ibid: 6), and lead to an attention monopoly of particular groups or individuals.
"In other words, online discussion forums may increase the volume of participation, but this does not necessarily correspond to a greater participation diversity, as often these participants are ultimately the 'usual suspects'.The virtual public sphere is structured so as to replicate traditional politics.As a consequence, the online debate is ultimately dominated, both quantitatively and qualitatively, by individuals that transfer to the virtual space their real-world domination" (Ferreira, 2012: 69).
We contend, therefore, that cyberspace allows us to explore "new meanings of community and transcendence of identity physical subjugations" (Silveirinha, 2004: 257), even if it does not cease to be "plagued by the inadequacies of our political system" (Papacharissi, 2002: 11).These findings can, indeed, be related to the concept of political poverty created by Bohman, which consists of "the inability of groups of citizens to participate effectively in the democratic process.The consequences of such poverty are two-sided: public exclusion and political inclusion.On the one hand, politically impoverished groups cannot avoid public exclusion (...).On the other hand, such groups cannot avoid political inclusion either (...).(...) [T]heir silence is turned into consent" (Bohman & Rehg, 1997: 333).Power, then, permeates the discursive form, the style and the contents, which can lead to the exclusion or silencing of minority groups (Silveirinha, 2005: 24).Privileging certain types of discourse leads, therefore, to an inequality that translates into an "internalized sense of the right one has to speak or not to speak, and from the devaluation of some people's style of speech and the elevation of others" (Young cited in Ferreira, 2012: 29).

Online discussions: beyond Habermas?
Online environment can, on the other hand, stimulate the flow of discursive modes beyond the critical-rational argumentation, shaping a heterogeneous public space, both in terms of language and expression, and in terms of discussion subjects (Esteves, 2003: 203).It is, therefore, relevant to ask to what extent the online environment should be assessed uniquely based on its correspondence with the traditional public sphere, in a strictly Habermasian sense, and whether the assessment of these discursive spaces should include additional categories.Indeed, recent research has shown that online discussions do not always follow the ideal criteria of deliberative democracy, although these high standards "are useful and necessary to define directions" (Dahlgren, 2005: 156).The communicative character of the political discussion does not always promote the civic ideal, and the discourse is not always as rational as the Habermasian criteria.Much of the political discussion is isolated and unpleasant, which may mean that its contributions to democratic will formation cannot always be assumed (Wilhelm cited in Dahlgren, 2005: 151).
Additionally, what Dahlgren and other authors call "the rationalist bias" tends to discount a wide array of discursive modes that can be of importance for democracy, "including the affective, the poetic, the humorous, the ironic, and so forth" (Dahlgren, ibid), or even the quasi-orality and more unconventional ways of argumentation (Rasmussen, 2008: 77).Dominant theories of democracy place a considerable emphasis on the rationality, but tend to ignore anything that approaches the emotional or affective (Dahlgren, 2009: 83).According to some authors, however, emotion can also be productive, by heightening our attention to particular issues and promoting political participation (Sobieraj & Berry, 2011: 23).
Iris Marion Young is one of the authors who, building upon feminist premises, designs a deliberative democracy model based on a plural communicative concept that aims at acknowledging the differences (Young, 1997: 60 and subsequent pages), which implies taking into account multiple ways and styles of communication.Young, who is more generically concerned with issues of social justice and oppression, criticises "a conception of democracy that foregrounds a culture of rational argumentation, a form of discourse that does not admit difference in speaking or listening.Deliberation, in these terms, heightens the discourse itself, rather than the listening, since the aim of rational deliberation, which paves the ground for political commitment, favours the affirmative and confrontational, formal and logical, impartial and bodiless discourse" (Silveirinha, 2005: 20).
The author therefore identifies three formal elements of communication that can enable that same plurality and transform the difference into a resource for public debate (ibid: 67 and subsequent pages): narrative (storytelling and personal experience as a way of promoting the understanding of difference); rhetoric (speaker's stand in relation to the audience, who can express himself/herself by means of humour, language games and rhetorical resources, as well as images); and greeting (acknowledgement of the other participants' subjectivity as a means of feeding trust).
Taking into account the online discursive environment and our previous discussion -both in relation to the public sphere expansion/transformation possibilities, as much as to the multiplicity of ways of expression existing in its diverse spaces -, this article aims to provide, via a case study, a distinct analysis of the online public sphere that preserves the basic criteria of deliberative democracy, while acknowledging the existence of diverse ways of public discussion and a multiplicity of participation modes.

Scope and methodology of this study
Traditional mass media -such as radio, television and newspapers -experienced deep alterations on their operation as a result of the emergence of the Internet.These alterations affected, not only the journalists' professional routine, but also their relationship with the receivers of the news message.Presently, most of the mainstream media, both national and international, have their own websites where the readers are offered different participation possibilities, including, for example, the ability to comment on news pieces produced by journalists.
The journalists' email addresses, the existence of discussion forums, the establishment of virtual communities or the blog phenomenon reveal different degrees of audience participation, resulting from the online presence of the traditional media (Domingo, 2008: 687-94;Rasmussen, 2008: 76).In fact, media organizations have made considerable efforts to create visible and instantaneous feedback tools, so as to allow for additional information, alternative angles and even error correction by the public (Keyser and Raeymaeckers, 2010: 4).This may "provide journalism with an agility that increases its sensitivity to peripheral phenomena, sometimes scarcely apprehensible by the institutionalised practices and sources" (Correia, 2010: 93).Since they are a potentially deliberative space, newsreaders' comments on the traditional media websites allow them, therefore, to express their opinion on a certain topic, by reacting to a news piece.News reader's comments can, therefore, be viewed as civic forums -which are frequently understood as the paradigmatic model of the online public sphere (Dahlgren, 2009: 168) -, where the participants exchange ideas, beliefs, arguments and opinions.
The main aim of this paper is to present a case study on quality assessment of online public deliberation.An analysis is conducted of the reader's comments to the six news stories with most readers' participation, on the Brazilian presidential campaign (October -November 2010), in the online version of two Portuguese newspapers (Público and Expresso).We analysed 303 readers' comments to online news stories (139 in Público and 164 in Expresso), using the quantitative and qualitative analysis software package Nvivo.
Since this case study is part of a wider research project on the possibility of a "feminine public sphere"1 , our aim was to assess the quality of the debates encouraged by the online loci, using variables such as the argumentation rationality or reciprocity.In this respect, we make reference to the study conducted by Tsaliki (2002) on online political discussion forums in Greece, the Netherlands and the United Kingdom, as well as to categories related to the discursive elements suggested by Young (1997), as shall be explained.Each paragraph corresponds to one unit of analysis, and was coded according to one or more categories/subcategories.
It is also important to mention the different strategies used by the traditional media to handle the readers' contribution.Overall, two main systems can be identified: the first, being more "interventionist", is based on pre-moderation, prior to publication; the second one, being "looser", is based on post-moderation, after publication (Noci et al., Participation and deliberation: a case study of readers' comments to news stories on the Brazilian presidential campaign .Marisa Torres da Silva 2010: 3).In the latter case, the comment is published automatically, but should its contents be considered abusive by the other users, the case can be reported and the comment subsequently removed.When this case study was conducted, both Público and Expresso used a post-moderation system, which required prior user registration (name, password and email) 2 .

Discursive modes: participation and deliberation
In order to analyse the discursive modes of the comments to the news stories on the Brazilian presidential campaign in the newspapers indicated above, a grid was designed which includes the following variables and categories: -degree of argumentation legitimacy: rational.Categories: justification (reasons/arguments provided to claim/sustain a certain premise/idea/opinion); complexity (subcategories: existence of different ideas; ways of expressing the same idea); civility (subcategories: respect for democracy; respect for other social groups; respect for other points of view); and politeness (subcategories: cooperating attitude; avoidance of disagreement); -degree of argumentation legitimacy: non-rational.Categories: non-civility (subcategories: threats to democracy; threats to the rights of other individuals and/or social groups) and impoliteness (subcategories: insults/personal attacks; libel; non-cooperation; other types of behaviour, such as sarcasm or the use of all-caps); -critical-rational judgement.Categories: self-criticism and criticism of others; -degree of rational legitimacy.Categories: valid (offers reasons and arguments for the stand taken) and non-valid (subcategories: does not offer reasons or arguments for the stand taken; arbitrary comments; fallacies); -elocution positions.Categories: raise/launch topic, search for information (for example, using direct prompts), incorporate (to complement other participants' arguments with ideas and information from outside the debate) and response (to reply to other comment or participant); -elements of plurality maintenance.Categories: narrative, rhetoric and greeting.
-loose categories.Off topic and use of irony.
2 Público adopted a pre-moderation system in March 2011, as a result of several readers' complaints, and the subsequent intervention of the newspaper's watchdog, about the growing posting of abusive comments.However, on 22 November 2012, when the website was refreshed, the newspaper implemented a moderation method based on a reputation system.This method consists of allocating the registered newspaper users to different levels, each with more or less power within the community, depending on the quality of their participation, while retaining, however, the pre-moderation of the comments.
The system used by Expresso is relatively similar to the one currently used by Público.It is based on gaining karma points for each comment (similarly to what happens on Youtube or Slashdot).The more prestige points are gained by a participant, the more possibilities he/she has to moderate the other readers' comments, and to rate them as funny, interesting or well written (individual users are not, however, required to have prestige points to report abusive comments).
Participation and deliberation: a case study of readers' comments to news stories on the Brazilian presidential campaign .Marisa Torres da Silva we also observed that many comments offered reasons or arguments for the stand taken (129 instances) 3 .Despite the significant rational legitimacy of the comments, we observed, on the other hand, a substantial volume of impoliteness, including: defamation (49 instances), personal attacks to other users (36 instances) and other types of behaviour such as sarcasm or using all-caps (58 instances).The use of fallacies (108 instances) and irony (72 instances) was also very frequent.
Although the comments do not contain swear words, some of them include words that are far from the politeness expected in these spaces of participation.The concept of politeness implies a commitment to a conversational contract that aims to minimise conflict (Papacharissi, 2004: 260).Therefore, although the traditional media websites include newsreaders' comments moderation rules, the existing considerable amount of insults and instances of defamation may be related to the inexistence of pre-moderation, which could contribute to avoiding offensive/abusive language and behaviour.
Unsurprisingly, criticism of others (e.g. of another user or a politician) is the fourth most commonly used category (189 instances), which suggests a very "aggressive" exchange of ideas on the Brazilian presidential campaign.However, the discussion often diverges from the electoral dispute itself, to focus mainly on the candidates' features and on the users' party differences, as well as on the comparison between Brazil and Portugal.
Notwithstanding the afore-mentioned assertiveness on the comments' debate, we observed the relative presence of discursive modes whereby the participants attempt to establish trust and empathy with the other users -such as greeting (41 instances) 4 .We also observed, even if less frequently, a reference to personal experience (narrative -24 instances), sometimes as a strategy of individual expression/revelation.This corresponds to the characteristics of an individualistic liberal democratic model that gives priority to personal interest (with a cathartic function), to the detriment of dialogue (Freelon, 2010: 7-10) 5 .Sometimes, however, the reference to personal experience is used as a strategy 3 Example: "Dilma won.And therefore, liberalism won.In fact, Lula's major virtue was the deeper liberalisation of the Brazilian economy.In Brazil, it is easy to establish oneself as an entrepreneur; it is easy for employers to dismiss their employees, as the Brazilian labour market is more flexible than the European labour market.It is also easy getting credit to open private companies.Since Lula's government benefited from the price increase of oil and raw materials in general, as well as from the massive agricultural and livestock exports, an unprecedented improvement of the life quality level of millions of Brazilians was made possible.It is true that barrel oil prices increased from 20 to over 80 dollars in ten years, and the production has nearly doubled; the price of raw materials in general increased about four times; and Brazil became the major world exporter of agricultural products and meat.The billions of reais gained with oil, mining and agricultural productions allowed Brazil to significantly improve the living conditions of the poorer.Dilma should, therefore, proceed with Lula's liberal agenda, so as to try and continue benefitting the poorer." 4 Examples: "Congratulations on your comment"; "You are absolutely right, and what you've just said, that knowledge and practice are a major asset, is very interesting"; "Many thanks for your aside"; "My best regards".
5 Example: "That happened over eight years ago, when I was dating a Brazilian who is now my wife.One day she decided to pay me a visit in Portugal.I called the Immigration Office to learn about the requirements to enter smoothly in Portugal.The official said that she would have to bring an X amount of money and prove such and such, and eventually suggested that, rather than her moving to Portugal, I should consider moving to Brazil instead -even though I had just returned from there one week earlier.She said that we are an immigration country (...).All this is just to say that the Portuguese still have to swallow many bitter pills when it comes to badmouthing Brazil".
to stimulate the understanding of certain points of view6 .A considerable amount of the rhetorical category is also to be noted in several comments (80 instances).This includes language games, humorous propositions or even figures of style, to seek the adherence of the audience7 .

Debate and monopolisation
In his study of the online forum talk.abortion,Schneider (1997: 85) observed that some participants managed to make themselves heard better than the others, and concluded that over 80% of the posts were published by fewer than 5% of the participants.In another analysis of the De Standaard Online newspaper discussion forums over a period of six months, Beyers found that 10,201 messages had been sent by 957 users, which amounts to an average of 10.7 comments per user (Beyers, 2004: 13-14).Dahlberg therefore highlights that, although in theory all the subscribers to a group have the same opportunities to post, in many groups only a small number of participants is responsible for most of the messages, which leads to online discourse attention monopoly (Dahlberg, 2001: 13), and inclusively raises issues of discursive equality and inclusion.
In the present case study, we also observed that the 303 comments analysed were posted by 104 users, i.e. an average of 2.9 comments per user.The domination of certain users was even more marked in Expresso (with an average of 4.7 comments per user).Some users demonstrate participating in the debate on a regular basis.In the case of Público, for example, one user alone posted 31 comments, and in Expresso several users posted more than 10 comments each.
The normative requirement of sincerity can also be challenged by resorting to anonymity and pseudonyms.Although it cannot be stated that omitting personal data automatically leads into a more disrespectful debate, the fact is that several users chose not to identify themselves by name (or, not the least, using a credible name); rather, they preferred to use pseudonyms or write anonymously.Of the 104 users that wrote comments, only 31 identify themselves by providing their name or surname and location.Only 12 of them provide their name, surname and location -even if it is practically impossible to verify the authenticity of these data.

Final remarks
This case study of the readers' comments to Brazilian presidential campaign news stories ascertained that news comments might represent a discursive space with features Participation and deliberation: a case study of readers' comments to news stories on the Brazilian presidential campaign .Marisa Torres da Silva that are distinct from the criteria established by Habermas and other authors for criticalrational deliberation.In fact, analysis categories like justification, complexity (namely, the existence of disparate or contrasting ideas), civility (which indicates a respectful or tolerant attitude in relation to the democratic system, other social groups and other points of view) and politeness (cooperation, expression of agreement, avoidance of disagreement) are among those that were used the least in our corpus.A substantial use of impoliteness was also observed in some comments, which raises issues related to tolerance, as a demonstration of respect for the reasons and perspectives of the others, which is a precondition for deliberation (Bohman, 2003: 93;Dahlberg, 2001).
However, the interaction among the comments' authors is one of the most relevant elements of the comments scrutinised, as are the exchange of information and ideas and the relative existence of discursive styles that aim at identifying, understanding and acknowledging the differences between the participants.This highlights the need to assess the discussion spaces made available by the online environment, by means of a framing that is wider than the one allowed by the strict Habermasian conception.
It can be stated, as we resume the questions asked in the beginning of this paper, that the space for news comments has in itself the potential to promote the dialogue and reciprocity among the traditional media public, within the scope of the public sphere enlargement sustained by the incorporation of the new technologies in the traditional media -notwithstanding the effective realisations that indicate the existence of incivility and impoliteness in the respective debates, as well as issues related to the discursive equality.
We, thus, underline the need for a more comprehensive analysis of the online discursive spaces -which includes news comments -that takes into account the fact that cyberspace is also a ground subject to tension, within a wider social and political order "that often imposes severe limitations on the Internet (and on digital technologies in general), especially in respect of its democratic potential and capacities" (Esteves, 2011: 36).
Translated by Rui Sousa-Silva