TALKING POLITICS ONLINE WITHIN SPACES OF POPULAR CULTURE: THE CASE OF THE BIG BROTHER FORUM Abstract Talking politics online is not bound to spaces dedicated to politics, particularly the everyday political talk crucial to the public sphere. The aim of this article is to move beyond such spaces by examining political talk within a space dedicated to popular culture. The purpose is to see whether a reality TV discussion forum provides both the communicative space, content, and style for politics that both extends the public sphere while moving beyond a conventional notion. The central question is whether it fulfils the requirements of rationality and deliberation. The analysis also moves beyond a formal notion by investigating how expressive speech acts interact and influence the more traditional elements of deliberation. The findings indicate that nearly a quarter of the postings from the Big Brother sample were engaged in political talk, which was often deliberative in nature. It was a communicative space where the use of expressives both facilitated and impeded such talk. TODD GRAHAM Todd Graham is postdoctoral research fellow at the Department of Journalism Studies and Media, University of Groningen; e-mail: toddsgraham@gmail.com. rN ■ LO rN si o o rN £ £ 3 a 0 £ * 1 * (R 0 C ! LO fN Introduction Over the past decade, the potential of the internet in fostering a public sphere where free and open deliberation and the exchange of information among citizens can prosper has been the topic of much debate (see Witschge 2004).1 During this time, we have seen the rise of social media as citizens increasingly engage in e.g. debates in online forums and communities, and storytelling and reporting via blogging and twittering. What this means for the public sphere supposedly is an opening up of the conversations society has with itself, creating new avenues that foster the basic element of the public sphere, i.e. everyday political talk. In its wake, we have seen an increase in research that looks to investigate political talk in online communicative spaces in light of the public sphere. Net-based public sphere researchers have studied these spaces in variety ways. However, the research has focused mostly on political spaces attached to a conventional notion of politics. Such exclusivity is problematic because political talk is not bound to these spaces nor is it to party politics, particularly the everyday political talk crucial to the public sphere. As initial research suggests, other genres of the online communicative landscape such as those tied to reality TV and popular forms of entertainment foster political talk (Graham and Harju forthcoming; Van Zoonen 2007). Political discussions that take place within these spaces also contribute to the web of informal conversations that constitutes the public sphere. Moreover, politics today has become more pervasive. People increasingly organise their political and social meanings around their lifestyle values and the personal narratives that express them as opposed to traditional structures and institutions (Bennett 1998; Giddens 1991). Consequently, any concept of political talk must be capable of capturing issues that may fall outside a traditional notion of politics. The aim of this article is to move beyond politically oriented spaces by examining political talk within a reality TV forum. The purpose is to examine its democratic quality in light of a set of normative conditions of the public sphere. The analysis moved beyond a formal notion of deliberation by also examining the use of expres-y sives. Thus, I present the following two research questions: To what extent does a reality TV forum satisfy the normative conditions of the process of deliberation £ of the public sphere, and what role do expressives play within political talk that emerges in these spaces and in relation to the normative conditions? The answers to these questions look to provide an authentic account of how people talk politics online and provide insight into how such talk occurs outside conventional political communicative spaces. Political Talk and the Public Sphere ^ Net-based public sphere researchers have drawn heavily from deliberative democratic theory. Deliberative democracy involves public deliberation not only as a means of producing public reasoning oriented towards the common good and collective decision-making within formal and semi-formal settings, but also as a process of producing public reasoning and achieving mutual understanding within the more informal communicative spaces of the public sphere (Fearson 1998; Mansbridge 1999; Dryzek 2000). It is through ongoing participation in everyday talk whereby citizens achieve mutual understanding about themselves and each other representing the practical communicative form of what Habermas (1984, 327) c (U II calls communicative action. This web of informal conversations over time prepares citizens and the political system at large for political action. Net-based public sphere researches have been increasingly tapping into politi- ^ cal talk online. Evaluating its democratic value requires normative criteria of the process of deliberation of the public sphere. Researchers have typically drawn from some aspect of Habermas's notion of the public sphere. As Dahlberg (2004) argues, Habermas's work has been both influential and valuable because it provides the most developed critical theory of the public sphere available. Specifically, it is through his pragmatic analysis of everyday conversation where he argues that when participants take up communicative rationality, they refer to several idealising presuppositions. Drawing from these (1984, 1987, 2001), six normative conditions are distinguished, which focus on providing the necessary conditions for achieving understanding during the course of political talk by placing both structural and dispositional requirements on the communicative form, process, and participant.2 First, the process in part must take the form of rational-critical debate. It requires that participants provide reasoned claims, which are critically reflected upon. Such an exchange requires coherence and continuity; participants should stick to the topic of discussion until understanding or some form of agreement is achieved as opposed to withdrawing. The process demands three dispositional requirements, three levels of achieving mutual understanding. Reciprocity, representing the first, requires that participants listen and respond to each other's questions and arguments. However, reciprocity alone does not satisfy the process; reflexivity is required. Reflexivity is the internal process of reflecting another participant's position against one's own. With empathy, one takes a step further and tries to put oneself in the another person's position. It requires an empathic perspective taking in which we not only seek to understand intellectually the position of the other, but we also seek to conceptualise empathically both cognitively and affectively how others would be affected by the issues under discussion.3 Expressives and Deliberation Some democratic theorists maintain that rational discourse needs to be broadened, allowing for communicative forms such as greeting, gossip, rhetoric, and storytelling (Young 1996; Dryzek 2000). Young (1996, 129) argues that such forms "supplement argument by providing ways of speaking across differences in the absence of significant shared understanding." Others have argued that emotions and humour are essential to any notion of good deliberation (Basu 1999; Rosenberg 2004). Rosenberg (2004) maintains that productive deliberation requires the formation of emotional bonds between participants. Such connections fuel a participant's effort to understand other positions and arguments. Basu (1999) argues that humour warrants inclusion in any robust conception of deliberation. Humour benefits political talk in three ways: it acts as a social lubricant; it creates a more civil and productive discursive environment; and it can act as social glue (1999, 390-394). In short, deliberative democratic theorists have begun incorporating emotions and alternative communicative forms within deliberation. However, net-based public researchers have tended to neglect expressives by typically operationalising a formal notion.4 This is problematic because when people talk politics, they not only draw from their cognitive and rational capacities, but they also draw on their emotions. Indeed, expressives are inherent to political talk, and as some of the authors above have argued, they may play an important role in enhancing it. Thus, in the analysis that follows, the use of expressives is investigated. By expressives, I am referring to humour, emotional comments, and acknowledgements. Humour represents complex emotional speech acts that excite and amuse for instance jokes and wisecracks. Emotional comments are speech acts that express one's feelings or attitude, while acknowledgements represent speech acts that acknowledge the presence, departure, or conversational action of another person, such as greeting, thanking, and complementing. Methods The forum selected came from bbfans.com, which is a website ran by and dedicated to fans of Big Brother UK. The site maintains thousands of participants, which have contributed hundreds of thousands of postings. The data collected came from the sub-forum Celebrity Big Brother.5 Channel 4's (UK) Celebrity Big Brother series features a number of celebrities living in the Big Brother house, who try to avoid eviction by the public with the aim of winning a cash prize to be donated to their nominated charity. The 2006 series, which the data reflects, consisted of 11 housemates initially, for example: Michael Barrymore the comedian, Traci Bingham the model/actress, Dennis Rodman the basketball star, and Pete Burns the singer/ songwriter. What makes the 2006 series interesting is that one of the housemates was, at the time, the British MP George Galloway. Thus, it was selected because it offered a unique communicative space i.e. a nonpolitically oriented forum influenced by a political personality. The data gathered consisted of the individual postings and the threads in which they were situated. The selection of the data was based on the broadcasting dates of the series, which represented the month of January 2006. The initial sample contained 345 threads consisting of 6803 postings. This sample was first coded for political talk. The goal was to allow also for a more individualised, lifestyle-based approach to politics. All those threads that contained a posting where (i) a partici-£ pant made a connection from a particular experience, interest, issue, or topic in general to society, which (ii) stimulated reflection and a response by at least one other participant, were coded as political threads (Graham 2008, 22-23). The criteria will now be applied to postings from the forum:6 William: Funny you should say that, I have seen him checking him out... but Pete confuses me, he was married for 16 years yet now has a boy friend yet says hes not gay... my head just | explodes..why didnt they teach us this in school i just cant keep up Anne: To quote Barrymore - you should get out more. People are multisexual and not ev-Q eryone fits into a convenient box. This thread begins with a discussion on the lifestyle choices of housemates Dennis Rodman and Pete Burns. In the first posting, William states his confusion over ___________________________________._____r.__________________________________________ Burn's sexuality and ends his post by making a connection to society. A political discussion on multi-sexuality emerges when Anne reflects upon William's posting and replies accordingly. Once identified, political threads were then subjected to three phases of coding (see Figure 1). The coding scheme and instruments adopted for analysis are ^ based on the methodological approach developed in Graham (2008). During the ^ first phase, postings were coded for message type: reasoned claims, non-reasoned claims, and non-claim responses. Those messages that provided reasoning for their claims were coded as reasoned claims (arguments), while those that did not were coded as non-reasoned claims (assertions). Regarding non-claim responses, postings were coded for commissives and expressives. Those messages that assented, conceded (partial assent), or agreed-to-disagree with/to another participant's claim or argument from an opposing position were coded as a commissive. Messages were coded as an expressive response if they conveyed a participant's feeling or attitude towards him-/herself, another participant, or state of affairs, which consisted of the categories humour, emotional comments, and acknowledgements (as defined above). The unit of analysis during this phase was the individual message. Note that these categories were not mutually exclusive. Figure 1: Coding Scheme Overview Message type a) Reasoned claim (Argument) b) Non-reasoned claim (Assertion) c) Non-claim response- a) Commissive b) Expressive i. Humour ii. Emotional comment iii. Acknowledgement Evidence type a) Fact/Source b) Comparison c) Example d) Experience Argument style a) Reflexive argument Empathy a) Communicative Empathy Once all messages were coded, phase two of the scheme began; messages that provided reasoned claims were advanced. During this phase, the coding categories were divided into two groups: evidence type and argument style. Messages were first coded for the type of evidence used (fact/source, comparison, experience, and example), after which, selected messages were coded again for a reflexive argument (defined below). The unit of analysis during this phase was the argument. During the final phase of analysis, all messages were coded for communicative empathy. Messages suggesting that the author had imagined his- or herself in another participant's position, either cognitively or emotionally, were coded as an empathetic exchange. The unit of analysis here was the individual message. In all three phases, the context unit of analysis was the discussion thread; the relationship between messages within a single thread were analysed. I refer the reader to Graham (2008, 23-32) for a more comprehensive and detailed account of the coding categories, the coding scheme, and an operationalisation of the six conditions. Regarding expressives, the aim was not only to identify them, but also to see how they were used and whether they tended to facilitate or impede deliberation. Consequently, the above analysis represented only the first step. Additionally, several separate in-depth readings on the use of expressives for each were carried out with specific attention being paid to indentifying the type, analysing their social structure, and examining their use in relation to the normative conditions. In each case, the selected material was read, re-read, and worked through. Additional literature aided in the analysis; Shibles (1997) taxonomy of humour and Shaver's et al. (2001) categorisation of primary and secondary emotions were consulted as a means of categorisation. For a systematic account and breakdown of these analyses, see Graham (2009, 61-63). Identifying Political Talk Political talk was no stranger to the Big Brother forum. Thirty-eight threads containing 1479 postings, which represented 22 percent of the initial sample, were coded as political threads. What were the political topics of these discussions? This question was addressed by categorising the political discussions, which consisted of 1176 postings, into broad topics based on the issues discussed within the various coherent lines of discussion.7 There were 13 topics identified by the analysis including George Galloway's politics; bullying and codes of conduct; animal rights and conservation; the judicial system; health and the body; gender, sexuality, and discrimination; immigration, multiculturalism, and racism; the media; parliamentary politics; reality TV and society; the Iraq War and foreign policy; political philosophy; and education. The dominant topic of discussion was George Galloway's politics, consisting of 436 postings, which represented more than a third of the political discussions. It seems that Galloway's presence in the Big Brother house got participants talking politics. Much of the debate here dealt with his motives for appearing on the show and on whether a sitting MP should be allowed to participate in a reality TV series. How-y ever, the political discussions on Galloway were not always confined to these issues. Occasionally, the discussions branched off into debates on MPs and parliament in general. Moreover, participants here frequently discussed Galloway's politics, e.g. his political arguments, his position on the Iraq War, and his character, behaviour, and performance as an MP. Galloway was not the only political topic of discussion. Participants often engaged in discussions on a variety of issues. Moreover, these topics were not always driven by conventional political issues. From bullying to sexuality, 42 percent of the discussions centered on issues that were more individualised and lifestyle ^ oriented. Results: The Normative Conditions Q Rational-critical debate requires that political talk be guided by rationality and critical reflection.8 In terms of rationality, arguments are preferred over assertions. There were 825 claims made. Out of these claims, 591 were reasoned, which represented 72 percent of all claims, indicating that providing reasoning with a claim was the norm. In terms of postings, nearly 40 percent provided arguments, whereas fv^ only 16 percent contained assertions. Together, the exchange of claims, which rep- c (U resented 54 percent (796 postings) of the postings, was the guiding communicative form. In terms of critical reflection, all those arguments that directly challenged or contradicted another claim or argument were considered to have achieved critical reflection. Forty-two percent of all arguments contained critical reflection, which represented 17 percent of the postings. Coherence requires that participants stick to the topic of discussion. Thus, postings within each thread were first analysed and then categorised into lines of discussion based on the issues discussed. By determining the number of topic changes and more importantly, the relevance of those changes, the level of coherence was ascertained. Within the 38 discussion threads, 98 lines of discussion were identified. Participants did not diverge at all from the original topic in only nine of these threads. That said, within the remaining 29 threads, there were 40 lines of discussion, which consisted of only 193 postings, coded as complete departures.9 In other words, 87 percent of the postings were coherent. Continuity requires that a discussion carry on until some form of agreement is achieved as opposed to abandoning it. Continuity was examined by determining the level of extended debate and convergence. The level of extended debate was measured via the presence of strong-strings, i.e. the depth of the exchange arguments. A strong-string refers to a minimum of a three-argument interaction, ideally in the form of critical reflection. There were 53 strong-strings. The average number was nearly nine with the largest totalling 42 claims. Fifty-five percent of all claims (455 claims) were involved in strong-string exchanges, which represented 30 percent of the postings. Furthermore, 88 percent of strong-string claims were reasoned with arguments containing critical reflection representing slightly more than half, indicating the rational and critical nature of these exchanges. Convergence was the second indicator of continuity, which gauged the level of agreement achieved during the course of a discussion by identifying commissive speech acts. There were 30 commissives identified, which represented only two percent of the postings. In order to determine the level of convergence, the number of commissives was compared with the number of lines of discussion. The sample consisted of 38 threads, which contained 47 political coherent lines of discussion.10 The average number of commissives per line of discussion was 0.64. Furthermore, 29 percent (or 14 lines) contained at least one act of convergence. Reciprocity requires that participants read and reply to each other's posts. It was assessed by determining and combining the level of replies with a degree of centralisation measurement. First, as Figure 2 shows, the level of replies was moderately high. Twelve out of the 38 threads had a reply percentage indicator of > 75 percent. While nearly half of the threads (18 threads) contained a percentage of replies of > 50 percent but < 75 percent. The percentage of replies for the whole sample was 65 percent. Regarding the degree of centralisation, the measurement is set on a scale of zero to one with zero representing the ideal decentralised thread and one the ideal centralised thread.11 First, Figure 2 indicates that only three threads were moderately to highly centralised (threads > .500). These threads resembled more a one-to-many or many-to-one type of discussion rather than a web of interaction. Second, 17 of the 38 threads were moderately decentralised (threads between .250 and .500). In these threads, even though there were still several central participants, the connections were more dispersed. Finally, nearly half of the threads (18 of 38 threads) were highly decentralised (threads < .250). The connections here between participants were distributed more equally. Finally, concerning the combined analysis, those threads within the top left quadrant, strong decentralised web quadrant, were considered to have a moderate to high level of reciprocity. As is shown, 28 of the 38 threads fell within this quadrant. In order to make a sharper distinction between these threads, a second set of criteria was added (represented by the dotted lines) as a way of distinguishing between those threads possessing moderate levels with those containing high levels of reciprocity. As is shown, there were four threads that contained an ideal level of reciprocity (threads > 75 percent and < .250) while six threads maintained a strong, moderately decentralised web of interaction, in other words, a moderately high level of reciprocity (threads > 75 percent and between .250 and .500). Given the modest level of replies, a majority of the threads within this quadrant (18 threads) fell below the dotted line with eight representing highly decentralised threads and 10 moderately decentralised threads. Figure 2: Level of Replies and Degree of Centralisation in the Political Threads of the Channel 4's Celebrity Big Brother Forum (1479 postings in 38 threads; January 2006) il CN CO 1009590858075706560- cu 5555 F c cu U s^ CU J^ JO, 3 50 45403530252015 10 5 0 Bl n n ] * □ 3 □ □ □ d? H □ □ D u 0 a □ n □ □ □ □ 0,000 0,100 0,200 0,300 0,400 0,500 0,600 0,700 0,800 0,900 1,000 Degree of Centralisation Reflexivity requires that participants reflect other participants' arguments against their own. The first step in determining the level of reflexivity is to establish the type and level of evidence use. The use of evidence suggests that a participant has taken the time to reflect upon the opposing position because in order to relate evidence to one's own or opposing argument they must know and to some extent understand the opposing position (Kuhn 1991). Overall, in terms of evidence use, 41 percent of all arguments contained supporting evidence. There were four CO types of evidence identified, which were examples, comparisons, facts/sources, ^ and experiences. Examples were most frequently used, accounting for 45 percent. Examples typically were of the housemates' behaviours and statements (usually in the house) such as their bullying behaviour, their smoking habits, and Galloway's political statements. Comparisons and facts/sources represented 23 and 24 percent respectively, while experiences were the least common at only 8 percent. Regarding the use of facts/sources, participants typically dropped links to news media reports and government/non-government public information sites as the below posting illustrates: Harold: Oh, and if anyone thinks it's despicable that Galloway isn't representing his constituents, go to the Hansard site and have a look at who's said what (or not) and who has attended (or likely not). The second step in ascertaining the level of reflexivity is to identify reflexive arguments. When a posting or series of postings (1) provided a reasoned claim; (2) used evidence to support that claim; (3) was responsive to challenges by providing rebuttals and refutes; (4) and provided evidence in support of that defence or challenge, they were coded as part of a reflexive argument. When these criteria were applied, they identified 20 reflexive arguments, consisting of 85 messages, which represented 6 percent of the postings and 13 percent of all arguments. The average number was slightly more than four messages per argument with the largest totalling eleven. Empathy was gauged by determining the level of communicative empathy. It requires that participants convey their empathetic considerations to others. There was one trend identified, which was the communication of third-person empathy. On occasions, when participants were discussing the behaviour and statements of Big Brother housemates, they would empathise with them and communicate this to fellow forum participants, as Matilda's posting below illustrates: Matilda: That was really uncomfortable viewing. I actually feel like crying myself I'm amazed how how well Traci coped so well with the way she was being treated. WHY did no one step in?? ok so shes a bit all American cheerleady type but there was absolutley no need for Pete to treat her in that way.Ç^j I hope she wins now. I think it touched into when I was bullied at school I really want to cry about it.(H) In this thread, Matilda empathises with Traci Bingham who was being bullied by Pete Burns; she brings her third-person empathy to the discussion. Matilda's posting reveals her bullied past, which eventually ignited a political discussion on bullying and British youth. Moreover, during this discussion, it sparked internal empathetic exchange between participants on their bullied experiences. However, such exchanges were infrequent in comparison to the total number of postings. In particular, there were 22 messages coded as communicative empathy, which represented less than two percent of the postings. Results: Expressives Expressives were a common ingredient of political talk, appearing in 41 percent of the postings. The most common expressive was humour. It accounted for 45 g S 5 0 £ * i * (R G C (U ro percent of expressives, which represented 20 percent of the postings. Overall, the analysis revealed three aspects on the use of humour. First, within the context of everyday conversation, humour may be used for a variety of reasons from expressing frustration and anger towards authority to criticising another (Koller 1988). Three general trends emerged regarding the use of humour. Participants tended to use humour (1) to entertain; (2) as a form of social bonding; and/or (3) to criticise, assess, or provoke thought. The most common use of humour was to entertain. Humour here usually came in the form of wisecracks, caricature, sarcasm, anecdotes, jokes, and banter. There were two focuses. First, humour was used to make fun of the Big Brother housemates. For example, the two postings below come from a thread on Pete Burns's coat, which was confiscated and examined by police due to allegations that it was made of gorilla's fur: James: I'd like to see Galloway wear a pete burns coat. Yeahm you read that right. Lets 'process' pete...and get some good use out of him. On Galloway...Pete might look good. George: Pass the Morsel In this thread, a discussion on animal rights and the fur trade emerged. During the course of the discussion, several participants engaged in a humour fest. As the above postings illustrate, such humour was often accompanied by malicious delight. Humour here tended to be less constructive in relation to the issue under discussion and orientated more towards "having a laugh." Moreover, the use of pictures, like above, to tell jokes or to present caricature was employed, suggesting a culture and commitment to entertaining fellow participants. Second, a substantial portion of humour under "to entertain" focused on good-natured teasing and the exchange of witty remarks between and about participants LO in the form of banter. Banter was the most frequent type of humour used. Banter ^ appeared to serve two functions. In addition to entertaining, banter acted as social glue; it functioned as a means of social bonding. These types of exchanges tended to be playful and flirtatious in nature. They seemed to unite forum participants creating a sense of shared experiences (participants would refer to these types of exchanges even days after they occurred) and fostering a friendly and sociable atmosphere. This sort of good-natured banter was common; 147 of 289 humorous comments (51 percent) were involved in this type of exchange. However, banter led discussions off the topic; 72 percent were off the topic of discussion. The final pattern was to criticise, assess, or provoke thought. Humour has a critical function e.g. questioning, criticising, and assessing politicians, government, or society in general. Humour here usually came in the form of satire via sarcasm, exaggeration, comparison, and anecdotes, as the below postings illustrate: Elizabeth: A Member of the UK parliament is under no obligation to do anything whatsoever during their term. except... to swear allegience once...to HM the Queen/King. Thats all. Ol' Georgie is more than allowed to be there. Edward: Well. The evictions are a little less boringly predictable than General Elections. That's how we should get the vote UP for political elections. If we were voting to EVICT MPs from Parliament, we'd have close to a 100% turnout, iljj The two postings come from a discussion on whether a sitting MP should be allowed to participate on a reality TV series. In both cases, participants use humour to express their cynicism towards the current state of parliament. In the first example, Elisabeth uses sarcasm to criticize MPs' job performance or lack thereof, while in the second example Edward offers a comical remedy to improve voter turnout. Unlike above, humour here was supportive and constructive to the political issues under discussion. The second aspect of humour was its social structure. Humour invited more humour in the form of humour fests. For example, when a participant posted a wisecrack, it often ignited an exchange of humorous comments. Out of the 289 postings containing humour, 56 percent were involved in humour fests. There were 29 fests. The average number was six with the largest totalling 36 postings. The final aspect of humour was its relationship, or lack thereof, with various variables of deliberation. First, humour on a few occasions was used as a weapon of degrading or resulted in flaming; 10 postings were tied to humour in this way. On these occasions, humour was used to make fun of another participant or was interpreted as such. Regarding coherence, humour acted as a distraction to political talk; 41 percent of all humorous comments were off the topic of discussion. Emotional comments accounted for 31 percent of all expressives and appeared in 14 percent of the postings. Overall, the analysis revealed three aspects on their use. First, when participants expressed emotions, they commonly expressed negative emotions. Anger was the most frequently used emotion; 66 percent of emotional comments expressed some form of anger, which was usually directed towards Big Brother housemates. Anger was expressed mostly through statements of dislike, disgust, and annoyance. Though the level of negative emotions was high, participants also posted expressions of appreciation, admiration, approval, and longing. The second aspect of emotional comments was their social structure. Similar to humour, emotional comments fuelled more comments that were emotional in the form of rant sessions. These were exchanges where participants vented their disgust, annoyance, and dislike towards Big Brother housemates, as the postings below illustrate: Victoria: I don't think i have ever seen anyone so self absorbed, disgusting, vile self opinionated, and every horrible word under the sun in my life. What a revolting man. Mary: I don't think I can express how disgusting I think this man is?! It really worries me that he is in a position of power in this country. Well, hopefully was. Surely there is no way he can continue to represent anyone in this country from now? If I lived in Bethnal Green or Bow, I would move. ASAP. Stephen: I just want to wipe that smug smile off his face. (jSj) Mary: How can anyone who he is supposed to represent can ever believe a word that comes out of his mouth now I don't know. He should be kicked out of the show and kicked out of parliament. How can anyone want that vile, nasty, sneaky man as their MP I don't know. He is a bully, a snake, a smug b****d and he makes my blood boil!! Charles: he was a total D*CK on last night's show. In this thread, a discussion on Galloway's attempts to discuss politics in the Big Brother house turns into a rant session on Galloway's behaviour. Participants were more interested in expressing their anger and disgust for Galloway than talking about whether politics and reality TV mix. These types of exchanges were often raw and vulgar. Moreover, they tended to be polarised; they ranted together under a common feeling and not at each other. Out of the 204 postings coded as emotional comments, 43 percent were involved in rant sessions. There were nine sessions. The average number was nine with the largest totalling 19 postings. The final aspect of emotional comments was their relationship with certain variables of deliberation. Emotional comments, when used, were fairly often used £ during the exchange of claims; 42 percent of emotional comments were expressed via arguments. Given the level of intense anger expressed, there was a tendency for these types of arguments to be abrasive, vulgar, and crude, as Jane's posting below illustrates: Jane: George Galloway is a disgusting, corrupt quasi-fascist dictator-loving *******. He is notorious for licking Saddam's arse, but now that Saddam has been toppled, he has taken to licking the butt cheeks of that other murderous tyrant, Syria's President Assad. He was expelled from the Labour party for urging Iraqis to kill British troops. He is an apologist for suicide bombers. He described the fall of the Soviet Union as the worst day of his life and has virtually admitted to being a Stalinist. His party rests on a coalition with extremist Islamists that means they have eschewed gay rights and women's rights in order to woo Muslim votes. ^ He is utter scum, and I despise him. In a discussion on Galloway's position on the Iraq War, Jane vents her disgust for the politician. As shown, her anger is intense and her statements are both vul-fv^ gar and crude at times contributing little constructively to the debate in question. (R G C Finally, acknowledgements accounted for 25 percent of expressives and appeared in 11 percent of the postings. There were five types identified: complimenting (60 percent), apologising (20 percent), greeting (11 percent), thanking (8 percent), ^ and congratulating (1 percent). Complementing was most common, representing 60 percent of acknowledgements, and appearing in seven percent of the postings. Participants typically complemented another participant's humour or argument with the latter accounting for nearly half. When participants did compliment another participant's argument, it was often directed at an opposing argument as opposed to being polarised. Participants also had a tendency to apologise in advance for posting an opposing position. Statements such as "apologies if anyone is offended" were used when an argument might seem too offensive or too critical. The Normative Analysis To what extent did the Big Brother forum satisfy the normative conditions of the process of deliberation of the public sphere? Overall, Big Brother faired relatively well in light of the normative conditions and past studies on online deliberation. The level of rationality, coherence, and reciprocity were high, while the level of critical reflection and extended debate were moderate. However, when it came to achieving deeper levels of understanding and agreement, Big Brother did not fair well. Rational-critical debate has been one of the most common conditions used among net-based public sphere researchers. Much of the research suggests that within a variety of political forum types, structures, and contexts participants are talking politics online rationally (Wilhelm 1999; Dahlberg 2001; Jensen 2003; Coleman 2004; Jankowski and Van Os 2004; Winkler 2005; Wright and Street 2007). For example, Wilhelm (1999, 173) concluded that participants within asynchronised forums are afforded both the time and anonymity needed to construct political messages, which reflect considered judgment. The results from Big Brother are consistent with these findings. In particular, the exchange of claims was guiding communicative form, which was typically rational in nature. The findings also indicated that a substantial portion of reasoned claims engaged in critical reflection. Regarding coherence, the analysis indicated that when participants talked politics, they rarely strayed off the topic; 87 percent of the postings were coherent. These findings are consistent with past studies (Dahlberg 2001; Jensen 2003; Wright and Street 2007). Moreover, they reveal that coherent discussions are not exclusively reserved for professionally (pre-) moderated forums, as some of the above studies suggest. Indeed, the self- and post- moderation practiced in Big Brother can also be effective in maintaining coherent (political) talk. Continuity was assessed by determining the level of extended debate and convergence. The analysis indicated that a substantial portion of political talk came in the form extended critical debate. This finding is not consistent with past studies (Wilhelm 1999; Brants 2002), which suggest that extended debate on a single issue was uncommon. One possible explanation is that these studies relied mostly on observations as opposed to a systematic operationalisation of extended debate. The finding does seem to fall in line with Beierle's (2004) survey research. Though his research was conducted with participants from a governmentally sponsored forum, it suggests that during the course of online debate participants developed a sense of commitment to that debate. Regarding convergence, it seems that extended critical debate on a particular issue rarely led to convergence of opinions, falling well short of the condition. This finding is consistent with previous research (Jensen 2003; Jankowski and Van Os 2004; Strandberg 2008). Reciprocity is another popular condition employed by past researchers. A couple of studies found low levels of reciprocity in online forums (Wilhelm 1999; Strandberg 2008). For example, Strandberg's (2008, 83) analysis of Finnish political message boards and Usenet news groups showed low levels of reciprocity thus concluding that the condition of reciprocity was hardly met. However, much of the literature does suggest that within a variety of forum types, structures, and contexts online political talk tends to be reciprocal (Dahlberg 2001; Brants 2002; Jensen 2003; Beierle 2004; Winkler 2005; Wright and Street 2007). The findings here are consistent with these latter studies; the level of replies was moderately high. However, as argued elsewhere (Graham 2008), the percentage reply indicator, which was employed by most studies, on its own is inadequate; it neglects a thread's social structure. Consequently, a degree of centralisation measurement was added. The combined analysis revealed that the political discussions maintained a high level of decentralised social interaction, indicating that a web of reciprocity was the norm. Few studies have measured reflexivity within online political talk directly. The studies that do examine it found substantial levels (Dahlberg 2001; Jensen 2003; Winkler 2005). However, unlike these findings, the analysis above revealed a low level with only 13 percent of arguments coded as reflexive. Regarding communicative empathy, to my knowledge, there have been no studies, which have employed this condition of deliberation. Given the lack of research, assessing the level is difficult. That said, the findings suggest that communicative empathy was infrequent, representing less than two percent of the postings, indicating that achieving deeper levels of understanding (or communicating it as such) were rare. y Expressives What role did expressives play within political talk? Expressives appeared in more than a third of the postings. Overall, they played a mixed role in relation to political talk by both facilitating and impeding it at times. Humour was the most common expressive, and it seemed to foster a friendly communicative environment. It seems Basu (1999) was right when suggesting that humour can benefit political talk by acting as a social lubricant and glue. The use of banter in particular seemed to foster social bonds. In some ways, humour appeared to help create a communicative atmosphere where a diversity of opinions on a variety of political issues was allowed to emerge. Humour too on occasions was used in support of rational-critical debate. However, humour did not always contribute constructively to political talk. Humorous comments frequently ignited humour fests, which tended to lead to incoherent political discussions. Emotional comments on the other hand seemed to impede political talk. Though they were used during the exchange of arguments, due to the intense anger that prevailed, these types of arguments tended to be abrasive, vulgar, and crude. As such, they contributed little constructively to the political discussions in question. Moreover, these types of arguments tended to ignited rant sessions. Here partici- c > pants engaged less in reciprocal-critical exchange and more in relieving their frustrations and anger in general by joining in on a rant with fellow participants. Thus, these C^ types of rants usually added little, in terms of understanding, to political talk. ^ Finally, acknowledgements appeared to facilitate political talk. The most common acknowledgement was compliments. Complimenting here was not polarised, that is, participants complimented across argumentative lines. Thus, it tended to encourage a civil and friendly atmosphere between participants on opposing sides of a position. Complimenting along with the use of preemptive apologising seemed to enable participants to express opposing positions and negotiate those positions without falling out. In sum, acknowledgements tended to create an atmosphere conducive for deliberation. Conclusion Talking politics online is not bound to political communicative spaces. The analysis above illustrates that this fundamental element crucial to the public sphere is taking place online in spaces dedicated to popular forms of entertainment. However, net-based public sphere researchers have tended to neglect such spaces. This is problematic because, as recent survey research suggests, those who participate in online discussions are more likely to talk politics in nonpolitically oriented spaces (Wojcieszak and Mutz 2009). Moreover, those participants who talk politics in political spaces probably differ from those who e.g. participate in reality TV forums. Therefore, in order to provide a more comprehensive and accurate account of online political talk and the public sphere, we need to start widening our scope of investigation. Future research should begin identifying political talk not only within spaces dedicated to fans of popular culture but also within sites attached to for example: lifestyles and hobbies, sports, friendship, support and self-help groups, occupations and trades, and consumerism - spaces where everyday political talk is likely to emerge. Research here should not only examine the discursive structure and normative characteristics of political talk in light of the public sphere, but should also investigate the mixing of everyday life, popular culture, and political culture that takes place within these spaces. For example, such spaces offer us an opportunity to explore the relationship between the personal and the political, moments when citizens make connections from their everyday lives to society, offering us insight into their concerns. Moreover, they provide us an opportunity to investigate political talk from citizens who are probably not actively engaged in the formal political process. What makes these spaces interesting too is that the participants who engage in political talk are not there to talk politics and may not believe they are doing so allowing them to avoid to some degree the negative connotations that are typically associated with talking conventional politics today, possibly leading to more deliberative talk. However, this raises the question of whether participants within these spaces regard the more lifestyle-based forms of political talk that I describe above as political. One of the limitations of this study is that it focuses solely on the text thereby neglecting the perceptions of participants. Studies should employ questionnaires, interviews, and/or focus groups in order to explore participants' perceptions, experiences, and motives for engaging in such talk within these spaces. Another question that emerges from this study is whether and to what extent such spaces empower citizens, leading to public engagement and participation in formal politics. Do such spaces foster "proto-political" engagement as Dahlgren (2009) describes? That is, to what extent are these types of performative practices supportive in a movement towards participation in the formal political process? Questions like these call for not only more longitudinal research on participation within these spaces, but also ethnography studies that focus on how (and whether) this connects and transfers into participation in formal politics, something currently lacking in net-based public sphere research.12 Finally, the analysis above reveals that expressives can make a distinct contribution to political talk, to deliberation. Though the philosophical and theoretical debate here is thick, net-based public sphere researchers specifically and political communication scholars in general have tended to neglect the use of expressives. Given the lack of empirical research, there remains a fundamental need for more descriptive studies, studies that focus on how the use of expressive interact and influence the more traditional conditions of deliberation. More work similar to Polletta and Lee's (2006) research on the use of storytelling for example would also add to our understanding. Moreover, as initial research suggests, the context (e.g. political or nonpolitical) and issues of deliberation may make a difference with regard to the role expressives play in political talk (Graham forthcoming). More studies that compare the use of expressive within various contexts on different issues would provide us more insight. Such research for example would help practitioners and researchers develop more effective facilitating and moderating functions for online deliberative initiatives such as e-consultations. Notes: 1. This paper is based on my dissertation (Graham 2009), which is available at the University of Amsterdam's public repository. 2. There are 11 conditions. However, due to the scope of this article, five have been omitted. See Graham (2009) for a comprehensive account. 3. Habermas focuses on the cognitive process of what he calls "ideal role taking" (1996, 228-230), while paying little attention to its affective side. i 4. See Graham (2010) for an analysis on the use of expressives in online political talk. 5. The data was taken from all those threads originating in January 2006. £ 6. When participants posted comments on government, policy, law, etc. criteria one was assumed. Note also that all call signs have been replaced with invented ones. ^ 7. There were 303 postings coded as nonpolitical and/or incoherent, which were not included. ^ 8. It went beyond the scope of this paper to assess the validity of argumentation used. Rather, the G................ o focus was placed on whether opinions stated were supported by argumentation. Note that a single post may have contained multiple claims. 9. Eleven of the 58 coherent lines (110 postings) were nonpolitical lines of discussion. 10. Only the commissives posted in the political coherent lines of discussion were included. 11. It is based on De Nooy et alt. (2005, 126) degree of centralisation measurement. 12. See Wright's (Forthcoming) discussion here on a new agenda for online deliberation research. References: ■i—I Basu, Sammy. 1999. Dialogic Ethics and the Virtue of Humor. The Journal of Political Philosophy 7, 4, 378-403. Beierle, Thomas C. 2004. Engaging the Public through Online Policy Dialogues. In P. Shane (ed.), Democracy Online: The Prospects for Democratic Renewal through the Internet, 155-166. New York: Taylor & Francis. Bennett, W. Lance. 1998. The Uncivic Culture: Communication, Identity, and the Rise of Lifestyle Politics. P.S: Political Science and Politics 31, 4, 41-61. Brants, Kees. 2002. Politics is E-verywhere. Communications: The European Journal of Communication Research 27, 2, 171-188. Coleman, Stephen. 2004. Connecting Parliament to the Public via the Internet. Information, Communication & Society 7, 1, 1-22. Dahlberg, Lincoln. 2004. Net-Public Sphere Research: Beyond the "First Phase." Javnost — The Public 11, 1, 5-22. Dahlberg, Lincoln. 2001. Extending the Public Sphere through Cyberspace: The Case of Minnesota E-Democracy. First Monday: Peer-Reviewed Journal on the Internet 6, 3 < http://131.193.153.231/ www/issues/issue6_3/dahlberg/index.html> Dahlgren, Peter. 2009. Media and Political Engagement. Cambridge: Cambridge University Press. De Nooy, Wouter, Andrej Mrvar and Vladimir Batagelj. 2005. Exploratory Social Network Analysis with Pajek. Cambridge: Cambridge University Press. Dryzek, John S. 2000. Deliberative Democracy and Beyond: Liberals, Critics, Contestations. Oxford: Oxford University Press. Fearson, James D. 1998. Deliberation as Discussion. In J. Elster (ed.), Deliberative Democracy, 44-68. Cambridge: Cambridge University Press. Giddens, Anthony. 1991. Modernity and Self Identity: Self and Society in the Late Modern Age. Stanford: Stanford University Press. Graham, Todd. 2008. Needles in a Haystack: A New Approach for Identifying and Assessing Political Talk in Nonpolitical Discussion Forums. Javnost - The Public 15, 2, 17-36. Graham, Todd. 2009. What's Wife Swap Got to Do with It? Talking Politics in the Net-Based Public Sphere. PhD Dissertation. Amsterdam School of Communications Research: University of Amsterdam. Graham, Todd. 2010. The Use of Expressives in Online Political Talk: Impeding or Facilitating the Normative Goals of Deliberation. In E. Tambouris, A. Macintosh, and O. Glassey (eds.), Electronic Participation, LNCS 6229, 26-41. Berlin: Springer. Graham, Todd. Forthcoming. What's Reality Television Got to Do with It? Talking Politics in the Net-Based Public Sphere. In K. Brants and K. Voltmer (eds.), Political Communication in Postmodern Democracy: Challenging the Primacy of Politics, 248-264. Hampshire, UK: Palgrave Macmillan. Graham, Todd and Auli Harju. Forthcoming. Reality TV as a Trigger of Everyday Political Talk in the Net-based Public Sphere. European Journal of Communication 26, 1. Habermas, Jürgen. 1984. The Theory of Communicative Action. Vol. One: Reason and the Rationalization of Society. Boston: Beacon Press. Habermas, Jürgen. 1987. The Theory of Communicative Action. Vol. Two: Lifeworld and System: A Critique of Functionalist Reason. Boston: Beacon Press. Habermas, Jürgen. 1996. Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Cambridge, MA: MIT Press. Habermas, Jürgen. 2001. From Kant's "Ideas" of Pure Reason to the "Idealizing" Presuppositions of Communicative Action: Reflections on the Detranscendentalized "Use of Reason". In W. Rheg and J. Bohman (eds.), Pluralism and the Pragmatic Turn: The Transformation of Critical Theory, 1139. Minnesota: MIT Press. Jankowski, Nicholas W. and Renée Van Os. 2004. Internet-Based Political Discourse: A Case Study of Electronic Democracy in Hoogeveen. In P. Shane (ed.), Democracy Online: The Prospects for Democratic Renewal through the Internet, 181-194. London: Routledge. Jensen, Jakob L. 2003. Public Spheres on the Internet: Anarchic or Government-Sponsored - A Comparison. Scandinavian Political Studies 26, 4, 349-374. u S 5 0 £ * i * (R G C fN Koller, Marvin R. 1988. Humor and Society: Explorations in the Sociology of Humor. Houston: Cap and Gown Press. Kuhn, Deanna. 1991. The Skills of Argument. Cambridge: Cambridge University Press. Mansbridge, Jane. 1999. Everyday Talk in the Deliberative System. In S. Macedo (ed.), Deliberative Politics: Essays on Democracy and Disagreement, 211-239. Oxford: Oxford University Press. Polletta, Francesca and John Lee. 2006. Is Telling Stories Good for Democracy? Rhetoric in Public Deliberation after 9/11. American Sociological Review 71, 5, 699-723. Rosenberg, Shawn W. 2004. Reconstructing the Concept of Democratic Deliberation. Shaver, Phillip, Judith Schwartz, Donald Kirson and Cary O'Connor. 2001. Emotion Knowledge: Further Exploration of a Prototype Approach. In G. W. Parrott (ed.), Emotions in Social Psychology: Essential Readings, 26-56. Philadelphia: Psychology Press. Shibles, Warren A. 1997. Humor Reference Guide: A Comprehensive Classification and Analysis. Carbondale, IL: Southern Illinois University Press. Strandberg, Kim. 2008. Public Deliberation Goes On-line? An Analysis of Citizens' Political Discussions on the Internet Prior to the Finnish Parliamentary Elections in 2007. Javnost- The Public 15, 1, 71-90. Van Zoonen, Liesbet. 2007. Audience Reactions to Hollywood Politics. Media, Culture & Society 29, 4, 531-547. Wilhelm, Anthony G. 1999. Virtual Sounding Boards: How Deliberative is Online Political Discussion? In B. N. Hague and B. D. Loader (eds.), Digital Democracy, 154-178. London: Routledge. Winkler, Roman. 2005. Europeans Have a Say: Online Debates and Consultations in the EU. Vienna: The Austrian Federal Ministry for Education. Witschge, Tamara. 2004. Online Deliberation: Possibilities of the Internet for Deliberative Democracy. In P. Shane (ed.), Democracy Online: The Prospects for Democratic Renewal through the Internet, 109-122. London: Routledge. Wojcieszak, Magdalena E. and Diana C. Mutz. 2009. Online Groups and Political Discourse: Do Online Discussion Spaces Facilitate Exposure to Political Disagreement. Journal of Communication 59, 1, 40-56. Wright, Scott. Forthcoming. Politics as Usual? Revolution, Normalisation and a New Agenda for Online Deliberation. New Media & Society. Wright, Scott and John Street. 2007. Democracy, Deliberation and Design: The Case of Online Discussion Forums. New Media & Society 9, 5, 849-869. Young, Iris M. 1996. Communication and the Other: Beyond Deliberative Democracy. In S. Benhabib (ed.), Democracy and Difference, 120-137. Princeton, NJ: Princeton University Press.