ONLINE FORUMS AND THE ENLARGEMENT OF PUBLIC SPACE: RESEARCH FINDINGS FROM A EUROPEAN PROJECT liza tsaliki
Abstract
This paper has focused on online political forums as newly emerged public spaces that contribute to the enhancement of public deliberation. Its aim has been to study the deliberative feature of the Internet and assess the extent to which ordinary citizens contribute to the political process through public debate online. The main research question is concerned with the way in which the Internet facilitates participation in politics, enables democratic deliberation, and provides a forum for reasoned argumentation. The extent to which this occurs has been studied through the content analysis of open public forums in Greece, the Netherlands and Britain. Research findings show a high level of interactive communication, high degree of search for information, diversity of opinions and publics and a moderate degree of substantiated argumentation-indicating an enlargement of public space in principle. However, the analysis stresses that unless netizens test their opinions in public systematically, the notion of the Internet as a tool for democratic deliberation is seriously undermined and runs the risk of being replaced by a push-button democracy. In that respect, cyberspace resembles the familiar world of everyday politics as an arena for the ongoing struggle for power and influence, despite the hype surrounding it.
Liza Tsaliki recently held Marie Curie Post-Doc Fellowship, Department of Communication, University of Nijmegen, email:
iza.tsaliki@sunderland.ac.u
Introduction
One effect of globalisation is the growing level of political apathy towards civic affairs: shrinking electoral participation, increasing dissatisfaction with politicians and parliaments, and decreasing levels of attachment to political parties have become clearly identifiable trends across Europe and the United States in the 1990s (Hale, Musso and Weare 1999; IDEA 1997; Gronlund and Setala 2001).
It is perhaps now, more than ever before, that it is imperative that citizens deliberate on public issues in a rational and critical way for any notion of the public sphere to retain its meaning. Precondition for communicative rationality, however, is an informed and active citizenry, hence the importance of the provision of information and of interactive and deliberative communication in modern politics. The spread of Internet use, particularly in Western countries, meant that the voice of the "new" movements reached beyond the borders of the nation state and was heard by a global audience. Concurrently, modern governments increasingly grasp the potential of information and communications technologies, and especially the Internet, for the provision of some of their services and for the enhancement of the decision-making process. The design and use of two electronic tools, e-petitioner and e-consultant, by the International Teledemocracy Centre in Scotland is only one among numerous schemes in this direction (Macintosh et.al. 2001).
This new kind of participatory politics is conducive to what Habermas (1996) has seen as the locus of discursive democracy where the weight of democracy is put on its ability to generate communication. In fact, he continues, "there is a counterweight of emancipatory potential built into communication structures themselves" (Habermas 1987, 390).
It is at this point that I am interested in online political forums as newly emerged public spaces that contribute towards the enhancement of public deliberation and "the enlargement of public space" (Melucci 1985, 815). The reason for concentrating on conversation as an object of study can be traced to the notions of civil society and the public sphere whereby "talk" is placed at the heart of democratic polity (Dewey 1927; Barber 1984; Carey 1995).
The premise of this paper is Alain Touraine's notion that politics in the modern world is a matter of collective action and interest in decision-making. Individuals must feel that they are citizens and must participate in collective life as voters are first and foremost citizens. It follows that democratic culture cannot exist without a public space and political debate (Touraine 1998, 25-27, 150). The paper is an attempt at a more systematic study of the deliberative feature of the Internet and is part of a larger ongoing project on the development of digital citizenship in the European Union. The project wishes to revisit ideas of democracy and citizenship and examine the role of the Internet as a mechanism for social and democratic change by looking at electronic governance, online political debates and environmental Net activism in different European countries (Finland, the Netherlands, the United Kingdom, Spain and Greece).
The objective here is to examine the role the Internet plays in facilitating participation in politics and in providing access to democratic deliberation. My research aim is to explore further the way in which the Internet can provide a forum for democratic reason, as well as the way it can be applied to the political process in order to overcome "the crisis of civic communication" (Blumler and Gurevitch 1995).
In which ways and to what extent do the "abstract publics" of the Internet become mobilised and start participating in politics? The paper presents the empirical findings from my work on online political debates in Greece, the Netherlands and Britain. In part one, it presents the methodology employed, explains the design of the analysis and offers a brief overview of previous research on public debate in cyberspace. It continues with the discussion of the research findings. The paper will then offer its assessment of the extent to which ordinary citizens contribute in the political process through virtual public debates.
Methodology
The focus of this research is on the study of open public forums since such a forum "must exist for democratic online deliberation to be free from state or corporate control. An open forum would conform theoretically to the ideal discursive conditions for a rational-critical public sphere outlined by Jürgen Habermas" (Coleman 1999, 206). The principle of independence is crucial to the credibility and legitimacy of a public forum for direct deliberation, and the autonomy of an open virtual public space cannot be brokered by governing elites without considerable difficulty.
For the purpose of my research, "online forums" were considered as those forums that allowed citizens to make and read contributions made by others online. This means that on-line facilities where citizens were invited to post their opinions via e-mail, without being able to read other participants' contributions were excluded. In addition, chat-like facilities where contributions were expressed in a matter of seconds, rather than being formed over a period of time, and were accessible only during a very short time, were also excluded (Ranerup 1999).
The decision was made to focus on citizen-to-citizen deliberation rather than concentrate on debates where both citizens and political representatives or other figures of authority were involved. Research has shown that in the latter case the dynamics of the discussion are heavily affected by the online presence of the politician invited with the majority of opinion exchanges taking place between the authority figure and each participant (Jankowski and van Selm 2000), hence a clustering of debate activity around the participant most obviously representing political power. This suits the purpose of the larger project whereby I am in search of digital citizenship across the European Union by looking at the way in which ordinary citizens make use of the deliberative features of Internet technology. I decided from an early stage that the analysis would be strictly text-oriented since there was neither time nor resources to interview people involved in the discussions in order to talk about their contributions.
Previous research has employed the method of content analysis in a variety of ways in order to study people's online presence. From Suzan Herring's (1993; 1996) work on the gender dynamics of online conversation to research of citizen participation in the political process, content analysis occupies a prominent place. Schneider (1996) has used the same method in order to analyse the postings of the talk.abortion Usenet group from the point of view of equality, diversity, reciprocity and quality of participation. The work of Jankowski and van Selm (2000), and Jankowski et al. (1999) is involved with assessing the style of conversing employed by participants. Ranerup (1999) has been working on experiences of on-line dis-
cussion forums at the local level in Sweden as part of a larger project on improving democracy in local government. Content analysis was the method she employed as was the case with Benson's (1996) work on on-line debates that take place on bulletin boards.
Design of the Analysis
Contrary to arguments that call for the dissociation of the notion of democracy from deliberation (Schudson 1997), I am, however, interested in the deliberative feature of the Internet and the extent to which it can be seen as a virtual public sphere. For the purposes of my research I applied and appropriated Anthony Wilhelm's (1999) model of democratic deliberation in cyberspace whereby the extent to which online political forums can act as "sounding boards" of the larger civil society is investigated.Hence, my main quest is the extent to which cyberspace enables and fosters democratic deliberation in the public sphere. The argument is that diversity of opinion and universal access to online forums is a necessary but not sufficient condition for realizing a democratic polity- instead, "discussions need to be deliberative" (Wilhelm 1999, 154). The next step in the empirical investigation of the deliberative feature of virtual public forums is to find ways to operatio/ nalize the content analysis of online debates. The following four questions do exactly that and thus aim to facilitate our understanding and investigation of deliberation on the Internet.
The first question to be raised is: To what extent do participants on online political debates offer or search for information? This question examines how people inquire about political matters. It argues that people need to seek and provide information in order to complement their thoughts and ideas.
The second one asks: In which ways do participants interact and converse with fellow citizens? The argument is that people need to interact and converse with each other if deliberative democracy is to exist. Otherwise, deliberative political process is replaced by "push button" democracy where people merely register their preferences without engaging in exchange of opinions and interaction with fellow citizens.
The third question addresses the extent to which there is in-group homogeneity taking place in online discussions. It addresses in-group favouritism, in the sense that people prefer to affiliate themselves with those who agree with them. Homogeneity is defined as the extent to which postings demonstrate a certain political affiliation, in the form of supporting a person, an argument, an ideology or issue. What we are looking for is political discussions where different viewpoints are expressed and where citizens can revisit their own position in the light of new information. After all, the ultimate goal of online discussions is not that everyone arrives at a consensus. Instead, the online forum (as its real life counterpart) may serve as an arena for a discussion between participants with different standpoints (Robins 1996).
The last questions aims to establish the critical-rational dimension of online political discussions by asking: What is the extent to which contributors participate in the debate in a rational way? Rationality emerges through online practice despite the fragmentation of the self that takes place in cyberspace. This rationality is, hence, discursively constructed, developed in the process of online interaction and not self-centred. This is in harmony with the Habermasian conception of the public
sphere as subjects develop autonomy and reflexivity only though deliberation (Dahlberg 2001). For the purposes of the content analysis, people are expected to give valid reasons for the statements they make instead of resorting to arbitrary comments and rhetoric. The degree of rationality of somebody's statement is assessed according to the reliability of knowledge embedded in it. That means that knowledge can be defended against criticism. In this context, forum participants need to supply reasons in defence of a position they have taken. If not, the arguments are not rational and do not stand public scrutiny.
In order to address the above issues, the following content categories have been used:
For question one, two categories were constructed. The first one, the category PROVIDE/PLANT, describes a message that solely provides information to the forum or plants a seed for the start-up of a discussion thread. A message coded in this category does not make references to other postings and does not make inquiries about further information. The second one, category SEARCH, involves those messages that search for information in the form of open-ended remarks, direct questions etc.
For question two, the categories INCORPORATE and REPLY were developed. A posting falls within the former category when it introduces ideas and information from outside the digital debate. The category REPLY involves a message that responds to other already posted messages. Whereas the first question addresses the extent to which participants engage in online talk to amplify their own views, the second one focuses on the extent to which political discussions are indeed interactive.
In needs to be noted at this point that the first two questions of the content analysis can be answered by "multiple selection." This means that, in question one, for example, a posting can PROVIDE/PLANT, can SEARCH, can do both, or may not do any of the aforementioned. The latter may happen in case a posting merely continues the discussion held, responds directly to the start-up question of the discussion in a monosyllabic way, is totally irrelevant to the discussion, is a swear word or a one-liner trying to appease the tone of the conversation.
The third question, which addresses in-group homogeneity, is further operationalised through the measurement of the scale of political affiliation that messages exhibit. The categories of AFFILIATION, NON-AFFILIATION, and DISAFFILIATION were developed for this purpose. AFFILIATION refers to postings that show solidarity to a person, platform, ideology or argument; NON-AFFILIATION involves postings that do not show any such solidarity (in other words, they are neutral), while
DISAFFILIATION is used to characterise messages that exhibit negativity or opposition towards a person, platform, ideology, or argument. In order to assess the level of political affiliation of each posting, coders were asked to canvass the discussion-tree in order to determine the prevailing trend. Individual messages were then coded in relation to the overall climate of the discussion. Clearly, although this approach is not foolproof, it can produce a rough estimate of in-group homogeneity.
Finally, question four, which deals with the degree of rationality of postings, is assessed through the content categories of VALID and NON-VALID. VALID
postings are those which offer reasons and arguments for the stance they take. On the contrary, NON-VALID postings do not offer reasons and arguments for the stance they take and resort to arbitrary or non-sensical statements. An example will help illustrate this operationalisation. In the Greek discussion thread on the Communist Party Youth (KNE) one of the participants utters:
In my humble opinion, we should examine what Communist parties and Party Youths have offered and will offer to the labour and youth movements. These differ depending on the time period we refer to. [To give an example] until 1968 there was no other alternative for all those fighting the system. Stalinism was the only left ideology in the West, and despite my disagreement with KNE's policy in Greece, I salute those people who fought for a better world -even if that entailed many wrong attitudes.
This discussant has construed a coherent viewpoint where his personal opinion emerges as a result of rational argumentation. The reply to this (Valid) utterance verges on racism: "The fatter the communist, the better the soap." This contribution, full of anti-semitic connotation, replicates a graffiti slogan that rhymes in Greek. It has no intention of substantiating the (Non-Valid) utterance any further, neither gives any reasons in support of the assertion. In other words, if reasons are not offered in the message itself, then its validity as a statement that may enhance the deliberative process is diminished.
Questions three and four are characterized by "exclusive selection." This means that in these questions postings can only be one thing to the exclusion of others. For example, in question three, a message can be categorized either as affiliated, non-affiliated, or disaffiliated. Having said that, the case may also be that a posting cannot be classified under any of the sub-headings of question three, if for example it concerns a swear word or a one-liner. In question four, however, a message will definitely be either valid or non-valid whether a one-liner or not.
I would like to expand now on the choice of forums. The object of study was online contributions that could be defined as "political." In Wilhelm's work, the name of Usenet newsgroups (alt.politics, talk.politics) was seen as a sufficient criterion for the discussion to be characterised as political. In this work it is the topicality of discussions that makes them the object of study, following Agneta Ranerup, whereby "political" is taken to describe a wide range of issues - anything from economic and environmental themes, to issues about the society at large, traffic and childcare. Discussions pertaining to politics directly (i.e., about political parties, local elections, political candidates) were primarily targeted, while the title of the debate was taken as another indication that they should function as arenas for political discussion. Such discussions were taken to be in contradistinction to conversations about hobbies, leisure activities, sports, clubbing and more casual conversations. The topicality of the forums, that is the agenda discussed, was determined by the participants in the sense that somebody would raise a question and a body of citizens would respond to this and to each other.
Furthermore, following the view that newsgroups are no longer the main vehicles for influencing a mass public - being text-based they represent the old Internet - and that, instead, it is the flashier Web with its graphical interface and hypertext facilities that is more suitable to a mass-audience (Resnick 1998, 49-50), I concentrated on Web forums.
Discussions were also chosen on the basis of their popularity (that is how many postings they had produced). A popular debate would provide sufficient space for the development of arguments. In addition, the relevance of the discussions to the political agenda at the time of analysis was also taken into account. This was seen as an indication of a lively and informed exchange of opinions.
However, the selection of discussions was also determined by the availability of an ample pool of such debates at the time of research in each country. This means that it was not always possible to find appropriate subjects for analysis. More particularly, in the Netherlands, I was limited by what kind of discussions were available online at the time of analysis. Two out of four Dutch discussions did not generate more than 35 messages in total. The same applied in the Greek case, where even the task of finding online discussion forums (any kind of forums, and particularly so when it came to Usenet groups interested in politics) was a difficult one, let alone finding discussions pertaining to political issues. For this reason the Greek case focused on commercial mailing lists. This difficulty is justified by the overall low level of Internet penetration in the country. Once the appropriate discussions were located, it was evident that the majority of them did not exceed 3035 messages in length. This reality informed my decision about how many messages I should analyze from each debate: the first thirty messages were analyzed in all cases.
The British content analysis was conducted on messages posted between March and April 2000, analysis of the Dutch case was done on messages posted between December 2000 and January 2001, and the Greek analysis was on postings between March and May 2001.
Description of Forums
There were eight discussions from Greece, seven of which were found at the "Politics" and "Society and Culture" categories of the site www.pathfinder.gr and one from the forum on www.flash.gr. Four discussions were examined from the Netherlands, two of which originated from the website of the Volkskrant newspaper, a broadsheet at the left of centre (www.volkskrant.nl/forum), the third from the advocacy group Forum voor Democratie (www.forumdemocratie.nl) , and the last one from the discussion forum hosted at the homepage of Roger van Boxtel, the Dutch minister of Integration Policy (www.ministervanboxtel.nl). Finally, another four discussions were analysed from the UK, originating from the Guardian's website (category "Politics"), another liberal broadsheet (http:// talk.newsunlimited.co.uk).
Coders in each country carried out the content analysis using a code sheet to register the data found. They were all provided with a detailed description of instructions as to how to conduct the content analysis. Although no standard procedures regarding inter-coder reliability were followed, informal checks and controls were initiated and followed in each case.
Britain
All discussions originate from the category Politics at the Talk section of the Guardian Online. The first discussion thread "Should Ken stand as an independent?" is about Ken Livingstone's MP decision to stand as a mayoral candidate for
London in May 2000 contrary to the Labour party's decision. The discussion generated 342 messages. The second thread, entitled "Brown - unwaged should take any job" was about a comment made by the Chancellor of the Exchequer Gordon Brown that the unemployed should take up any job instead of signing on for benefits. The discussion produced 295 messages. The third topic, under the title "10 good reasons to join the Euro," reflects the ambivalence over the European Single Market that characterizes Britain. Some 220 messages comprise this discussion. The final thread examined, "Ulster government collapse inevitable?" revolves around the politics of Northern Ireland and produced 350 messages.
Twenty-five participants produced the first 30 messages of the thread on Ken Livingstone. One of them generated 17% of the postings under examination, while there was a more or less equal participation by the rest of the discussants. Seven people took part in the exchange of the first 30 messages on Gordon Brown's statement, one of which contributed 50% of the messages and two more were responsible for another 30%. The part of the discussion on the Euro that was analysed consisted of 11 participants with two of them being responsible for half of the contributions. Finally, 12 people were involved in the Ulster thread with two participants being responsible for 50% of the contributions examined.
The Netherlands
The first two Dutch discussion threads came from www.volkskrant.nl/forum, the online service of the Volkskrant newspaper. The first, under the title of "PvdA," is about the Dutch social-democratic party chaired by the Dutch Prime Minister, Wim Kok. In autumn 2000 there was some internal disagreement regarding the need to modernize the party and Kok's role was openly discussed. The discussion took place between November and December 2000 and attracted a total of 13 contributors who generated 35 postings. To the extent that it was possible to ascertain the rate of involvement of the participants,1 two people were responsible for 27% of messages exchanged, while the rest had an equal number of contributions, more or less.
The second discussion thread, "ontwikkelingshulp" (Development Aid), revolved around the taboo issue of criticising the way the business of development aid is conducted. Nineteen people contributed to this debate accounting for 35 messages overall. The debate lasted from November to December 2000. It was not possible to assess the rate of involvement in this case, as the discussion webmaster did not facilitate the tracing of such information.
The third discussion thread, entitled "Democratie & bestuurskracht in Europa" (Democracy and the power to reign in Europe), originated from the www.forumdemocratie.nl. The theme of the debate is whether or not the Dutch feel increasingly alienated from the European Commons. Twenty people participated in this part of the discussion (sub-total of 30 messages) which took place from November 2000 until January 2001 and generated about 80 messages overall. Two of them contributed 33% of the total number of messages and another two 13%. The rest of the participants made single contributions.
The last discussion thread from the Netherlands originates from the webpage of Roger van Boxtel, www.ministervanboxtel.nl, and bears the title "Elektronische overheid" (Electronic government). The topic of this discussion is whether or not
electronic government in the Netherlands is working for the better. It took place in December 2000 and was part of Roger van Boxtel's routine practice of a monthly debate. Seventeen participants contributed in the first part of the discussion (subtotal of 30 messages) which accounted for 120 messages overall. One person was responsible for 20% of the contributions under examination and two more for another 20%. The rest of the discussion flow was not dominated by anyone in particular, with the majority of discussants making single contributions.
Greece
The discussion threads examined involved the following eight issues:
1.	The arrest of Slobodan Milosevic, whereby discussants were asked to comment on whether or not they considered this act as fair or as a sign of international hypocrisy. This was one of the lengthier discussions in Greece generating 95 messages. Twelve people participated in the debate and five of them dominated the exchange of opinions.
2.	The visit to Greece of Bush Senior where the main question was about whether or not this trip had a more sinister and obscure purpose (drawing from the consensual view in Greece that the US influences heavily Greek internal politics and foreign policy). This question attracted 10 people who produced 35 messages in total. Three members of this group contributed the bulk of the messages.
3.	The Greek Communist Party Youth and the role it has played in the political scene in Greece since 1974. The discussion attracted 13 participants, generating 32 messages overall. Again, three discussants generated the majority of contributions.
4.	The introduction of the draft bill for social security, a very controversial social security policy the Socialist government tried to pass which was later withdrawn after huge public unrest. This is another lengthy theme, consisting of 140 messages. The first 30 of these were produced by three people, while in the overall discussion 30 people were involved. All of the participants were men.
5.	The fate of the Greek Right and the way it should try to overcome internal fragmentation. The discussion consists of 28 messages, 23 of which were produced on the same day. Eleven people were involved, with 3 of them being responsible for the majority of the traffic of messages.
6.	"Avramopoulos-Vozenbergh" referring to the Centre Right-wing mayor of Athens, Dimitris Avramopoulos, his newly-launched party which split the Conservatives in Greece, and the appointment of Ms Vozenbergh as the party's spokesperson. This theme comprised 29 messages, and 12 participants.
7.	Another discussion about Dimitris Avramopoulos and the way in which his new party will affect the Greek Right. There were 30 messages in total and 13 participants. Two of them were the most visible, and another five had a moderate presence.
8.	A discussion about press freedom, its limits and the way it impinges on personal freedom. This discussion comprised 30 messages and eight contributors, 3 of which had a dominant presence.
Results of Content Analysis
The results of the content analysis per country are presented in Tables 1-3. The total number of messages for each country, as already mentioned, is 30. The tables present the number of postings found belonging to each category out of the aggregate of messages that are examined in each discussion. For example, in Greece, the first discussion to be examined is concerned about the arrest of Slobodan Milocevitch. It shows that 14 out of 30 messages provided or planted information, three out of 30 sought for information, while there were 13 postings which did neither. At this point, it needs to be stressed that because of the "multiple selection" and "exclusive selection" features of questions regarding the content analysis, described earlier, values in Tables 1-4 do not always sum to 30. To be more specific, messages in categories Provide/Plant and Search (corresponding to question one) and Incorporate and Reply (referring to question two) will not sum up to 30, as one particular posting can fall into both categories in both cases. It may also be that a posting cannot be put in any category as already explained. The latter equally applies for the Affiliated/Disaffiliated/Nonaffiliated categories (corresponding to question three) in spite of the "exclusive selection" characteristic. The reason, again, is that certain messages, such as one-liners, are difficult to decipher and categorize accordingly. On the other hand, categories Valid/Non-valid (referring to question four) always add up to thirty since a posting can only be valid or non-valid, even in the most extreme cases.
Table 4 translates the results of the content analysis from absolute figures into averages (i.e., in the Greek case, the analysis has shown that 56% of Greeks, or 16.8 on average, provide/plant information when talking online, while 21% search for information, or 6.4 on average). This has been done in order to overcome the problem of having a different number of discussions analysed from Greece (8) in comparison to the Netherlands and Britain (4 in each country).
The first question this content analysis examines was how people inquire about political matters and addresses the extent to which participants on online debates offer or search for information in order to complement their arguments. From a
Table 1: Analysis of Greek Discussion Threads (N = 30 messages)
	Milo-	Bush	E N K	Social	GR Right	Avramo-	Avramo-	Press
	cevitch	visit		security		poulos 1	poulos 2	freedom
Provide/Plant:	14	9	12	22	22	21	17	17
Search	3	3	6	11	5	9	5	9
None of the above	13	18	16	2	4	4	12	12
Incorporate	5	6	10	7	15	21	2	4
Reply	21	16	19	23	21	11	24	22
None of the above	8	13	7	5	2	2	5	6
Affiliation	4	4	12	13	5	9	19	6
Non-affiliation	6	5	8	2	16	14	3	6
Disaffiliation	6	0	5	10	6	6	3	2
Cannot be classified	14	21	5	5	3	1	6	16
Valid	16	10	20	17	12	11	16	20
Non-valid	14	20	10	13	18	19	14	10
Ln O
AVERAGE	Greece	(%)	Nether.	(%)	Britain	(%)
Provide/Plant:	16.8	(56)	25.3	(84)	20.8	(69)
Search	6.4	(21)	11.3	(38)	8.5	(28)
None of the above	10.1	(34)	2.5	(08)	4.8	(16)
Incorporate	8.8	(29)	15.3	(51)	16.3	(54)
Reply	19.6	(65)	17.0	(57)	21.3	(71)
None of the above	6	(20)	3	(10)	3.5	(12)
Affiliation	9.0	(30)	5.0	(17)	7.0	(23)
Non-affiliation	7.5	(25)	16.8	(55)	7.5	(25)
Disaffiliation	4.8	(15)	8.3	(28)	14.8	(49)
Cannot be classified	8.9	(30)	0	(0)	0.7	(03)
Valid	15.3	(51)	22.5	(75)	17.0	(57)
Non-valid	14.8	(49)	7.5	(25)	13.0	(43)
Table 2: Analysis of Dutch Discussion Threads (N = 30 messages)
	PvdA	Devel. Aid	Democ. Man.	e- govern.
Provide/Plant:	27	25	24	25
Search	11	12	10	12
None of the above	2	4	2	2
Incorporate	14	16	14	17
Reply	19	14	17	18
None of the above	1	3	4	4
Affiliation	3	6	5	6
Non-affiliation	17	16	16	18
Disaffiliation	10	8	9	6
Cannot be classified	-	-	-	-
Valid	23	24	21	22
Non-valid	7	6	9	8
Table 3: Analysis of British Discussion Threads (N			= 30 messages)	
	K.Livingstone	G. Brown	Euro	Ulster
Provide/Plant:	16	25	17	25
Search	14	7	6	7
None of the above	-	4	12	3
Incorporate	15	14	14	22
Reply	12	21	28	24
None of the above	3	4	2	5
Affiliation	9	0	12	7
Non-affiliation	12	11	5	2
Disaffiliation	7	19	12	21
Cannot be classified	2	-	1	-
Valid	11	12	21	24
Non-valid	19	18	9	6
Table 4: National Averages of Discussion Threads (N = 30 messages)
cross-national point of view, the evidence shows that in all three countries, participants are willing to provide information. In this way, they express their opinions and have their say. They inquire about issues to a lesser extent, though. Messages posing a query provide a point of departure for further conversation. However, if no one responds to them, they do not lead anywhere and die out. Table 4 shows that the Dutch, with 84%, have the best profile in terms of provision of information while online in relation to the British (69%) and the Greeks (56%). They are also ahead when searching for information (with 38%, followed by 28% by the British and 21% by the Greeks).
The second question is about the way in which participants interact and converse with fellow citizens and argued the need for dialogical communication. Do people incorporate opinions and information of others (outside the debate) in their conversations? Is this knowledge appropriated in any discursive way? When it comes to introducing new knowledge from outside the digital domain, the Dutch and British participants appear to be better informed or at least keener to bring into the discussion opinions from the off-line world (51% and 54% respectively as opposed to 29% from Greece). The reply factor seems to be higher in British online discussions (71%), followed by 65% in the Greek discussions and 57% in the Dutch ones.
However, further qualification is needed here. It is not enough for people to respond to each other's postings. Responses need to advance the discussion overall rather than merely aim to give an answer to a personal comment. This analysis has found that a large number of debates involve dialogical communication of the second, more personal, style. This is another thing to take into account when arguing about the deliberative features of the Internet and the way they can enhance democracy.
The third question of the content analysis addresses the degree of in-group homogeneity. Findings have shown that overall there is, indeed, a diversity of opinions and viewpoints expressed on online political debates, as the categories AFFILIATION, NON-AFFILIATION and DISAFFILIATION indicate. There are people who agree, disagree or take a neutral stance about political issues in all three countries, an indication of a healthy democracy where asymmetries of opinion are inherent. After all, a high level of in-group homogeneity would mean that a society is not very pluralistic and amenable to new identities. Disagreement, as we have previously seen, though, is important because it furnishes the occasion for new information search and triggers partners to rationalise and justify their opinions (Stromer-Galley 2002). In addition, these results show the similarities between online and real-life conversations where the point is not to resolve an issue and come to a definite answer to a problem - rather to exchange views about it. Table 4 shows that there is a higher rate of neutrality in the discussions that take place in the three countries, followed by a rather high level of dissensus, and a much lower level of consensual viewpoints. The Greeks are those who take a positive stance to a larger extent than the British and the Dutch (30%, 23%, 17% respectively). The British come across as those airing the more dissenting voice (49% of the British discussants disagree with the dominant view expressed in the debates, followed by 28% of the Dutch and 16% of the Greeks). The Dutch seem to adopt a neutral stance much more easily than the rest (56%, followed by 25% of Greeks and British).
The fourth question in this content analysis searches for the degree of rationality demonstrated by contributors by looking into the validity of their answers. Overall, the VALID answers outnumber the NON-VALID ones, which indicates that people engage in articulate and critical conversation when online. Having said that, the number of non-valid, unsubstantiated contributions is quite high, but this is hardly surprising either. As this research has found, there are quite a few participants who cannot support their stance and are carried away by personal feelings. The Dutch provide a more concrete substantiation of their arguments when online, with 75% of all discussions in the category "Valid," followed by 57% in the British case and 51% in the Greek one.
Discussion
In some cases, participants did not seem to approach the forum in order to engage in conversation with fellow citizens and exchange opinions. Instead, it looked as if people simply wished to air their own views without being particularly interested in listening to what others had to say. This is no different from dialogical communication in real life where similar practices can occur. However, we need to keep this in mind when advocating the ability of the Internet to generate dialogue and thus enhance democracy. As if it was not enough that the outcome of discussion forums is rarely followed up by appropriate policy action (Jankowski and van Selm 2000), now it looks that the people themselves, the producers and consumers of online talk, are more involved in sound-bites - their own - rather than in creative dialogical communication.2
Research on the degree of involvement of individual participants (e.g., Jankowski and van Selm 2000; Schneider 1996; Ranerup 1999; Wilhelm 1999) has shown that in most cases a very small number of discussants dominates the debate and is responsible for the majority of contributions. Participant involvement, that is, ranges between low and moderate levels, which indicates that the interactive element of the Internet is used under capacity. Additionally, it appears that there are some recurrent users who are in pursuit of participation - in the same way that members of the public become addicted to participating in quiz and game shows. After a while, closer relationships are formed online and fellow discussants engage in a repeat bilateral exchange of opinions.
This research has yielded similar results in the sense that once again only a few participants - some of them "running" in a participation circuit - are responsible for the majority of contributions made. This is certainly true in the Greek case, while it is less pronounced in the British and Dutch discussions. In saying that, I take into account the fact that some of the debates analysed in the latter instances were quite long, and hence the part of the discussions that was examined, comprising the first thirty postings, did not necessarily represent the true dynamics of the group of participants. Furthermore, in the Netherlands it was not always technically possible to ascertain the identity of each contributor and monitor their response rate.
In terms of gender, women participants in Greece appear to have a lesser presence in the debates.3 This is not only a result of the small number of women taking part in the debates, but mainly of the fact that Internet use in Greece is dominated by men. The Internet usage rate in November 2001 was 56% amongst men and
40% amongst women in the European Union. The equivalent rate for Greece - a country usually ranked lowest in terms of penetration and spread of Internet technology and use within the European Union - was 17%, compared to 57% for the United Kingdom and 69% for the Netherlands (Eurobarometer Report 112, 2001). Research has shown that Greek Internet users are following the main social patterns and demographic profiles of Internet use4 and thus reproduce existing gender inequalities on cyberspace as well (Boudourides and Drakou 2000; http:// hyperion.math.upatras.gr/survey/grinet.html). For those (few) women who do get online, the medium is a new experience and they can be easily silenced by their male counterparts (for more on this issue, see Herring 1993; 1996). Unfortunately, it was not possible to make similar remarks for the Dutch debates in any consistent way for technical reasons. Equally, it was impossible to discuss the gender ratio in the British online debates since participants did not volunteer any such information. Although a strong online female presence would be expected in Britain and the Netherlands due to the high level of technological development in both countries, it would not reverse the overall male-dominated trend in the digital domain.
Another interesting feature is the fact that online discussions sometimes stray away from the original question and develop a life of their own. For example, in the Greek case, the majority of discussions begin with serious and articulate argumentation only to subsequently resort into a series of personal attacks. The conversation degenerates very quickly and the site becomes an arena to settle personal scores rather than a forum for democratic deliberation. This is no more an agreement to disagree over politics - it is about "getting your own back."
The tone of discussion, particularly in those examined from Greece and Britain, is very informal and relaxed, though it tends to revert into bad language at times. Such language is not necessarily meant to be demeaning to other participants, as contributors sometimes tend to know each other from other discussions - hence the level of intimacy between "pals." Although such an observation is not surprising, it should again make us think about how many people actually use the Internet for democratic dialogue. These discussions seem, on occasion, to be a "close circle of friends" who take part in several online interactions at the same time.
In terms of the limitations of this research, the question might be asked: "are discussions comparable?"5 A number of steps were taken towards comparability of discussions. Hence, all of them were considered part of the world of politics, that is, they originated from forums where discussions categorized as "political" were being conducted. The notion of "politics" was taken in its broadest sense to encompass a wide range of issues. It is true that the discussions were set within their particular national contexts, hence cultural differences about how to conduct a debate would come into play. Although this was not seen as an obstacle for the cross-national content analysis at this stage, a more thorough insight into cultural differences following an ethnographic analysis would have been useful.
The choice of postings analyzed in the larger discussions in the British case is also worth mentioning. One of the limitations here may be that as the content analysis focuses on the first thirty messages of each thread, the discussions examined were not fully appreciated insofar as the number of participants and the way they contributed to the conversation is concerned. On hindsight, it may have been preferable to select shorter discussions, so that they matched the discussions in the
Netherlands and Greece (which generated around 30 messages altogether). This would have also resolved the possibility of a related problem, that of homogeneity of answers as discussions tend to be more homogenous in the beginning and heterogeneous later.
It needs to be stressed, though, is that this analysis is not meant as the ultimate answer to the way people converse about political matters on the Internet, nor does it pretend to draw any definite conclusions. It should be seen, instead, as a step towards a better understanding of the deliberative features of Internet technology. Future research should also look into more discussions from the Netherlands and Britain, so that there is an equal number of debates from all participant countries, and a larger overall sample. In fact, this paper should be read with the caveat of using a small number of case studies.
More systematic and in-depth work is also needed in the language and style of conversation employed across different national settings, as well as in the gender dynamics and the logic of argumentation within different cultures. Not knowing enough about the identity of participants, which would have required interviews and some ethnographic work, is also a drawback.
Another contentious issue has to do with the fact that while searching for the role of the Internet in enhancing participatory democracy I have focused on online debates, and hence I have targeted a niche of people who take part in such discussions anyway. All these are issues that should be addressed more systematically and thoroughly in the future.
Conclusion
This paper has focused on online political forums as newly emerged public spaces that contribute to the enhancement of public deliberation. Its aim has been to study the deliberative feature of the Internet and assess the extent to which ordinary citizens contribute in the political process through public debate online. The main research question has been the way in which the Internet facilitates participation in politics, enables democratic deliberation, and provides a forum for reasoned argumentation. The extent to which this occurs has been studied through the content analysis of open public forums in three countries, Greece, the Netherlands and Britain. Four questions were posed that consider how people inquire about political matters, how they interact and converse when talking politics online, whether there is in-group homogeneity, and finally whether conversations held amongst participants are rational.
At first glance, the results of the content analysis are encouraging: high level of interactive communication, high degree of search for information, diversity of opinions and publics, and a moderate degree of substantiated argumentation takes place - all indicating an enlargement of public space. The latter, however, is not seen in the sense of one, broadened, public sphere, but, rather, in terms of multiple micropublic spheres. In this sense, this analysis has come to a different conclusion than others regarding engagement with online deliberation about political issues. For example, in her work on alternative media on the Internet (2001), Suzan O'Donnell has argued that the relationship between alternative media online and the mainstream public sphere is weak and susceptible to various political and social constraints. Moreover, it is debatable the extent to which alternative media online can
be extended to participants unknown to each other, particularly when it concerns discussion of politics.
What are the consequences for deliberative democracy, however? The deliberative potential seems to be there but the question is who uses it, for what purpose and to what extent? It is evident that there are cases where only a few people are actually involved in meaningful conversation online. This may have to do with the degree of access to the Internet and its technology in various national contexts (and Greece is the laggard in relation to the Netherlands and Britain as mentioned above). It may also have to do with the standards of public debate and the notion of "conducting an argument" within specific national cultures (hence in Greece things are more chaotic and temperamental while in the Northern cultures politics is conducted in a more phlegmatic way). It certainly is evident that online citizen forums are not a panacea that will automatically result in a meaningful debate. On the other hand, perhaps the purpose is to stay online and develop a "sociable model of conversation" (Schudson 1997, 307) instead of a problem-solving one. The analysis, however, draws attention to the fact that unless netizens test their opinions in public in a systematic way, the notion of the Internet as a tool for democratic deliberation is in jeopardy, paving the way for more push-button democracy. One way to circumvent this obstacle is through policy that will introduce the normalisation of a right to reply amongst participants.
The research findings of this study resemble those from similar investigations (Downey 2001; Downey and Fenton, 2001; Resnick 1998; Wilhelm 1999), which view the relationship between the Internet and civic engagement with growing scepticism. One of the greatest disillusionments has been the realisation that the use of Internet technology for enhancing democratic deliberation is very much limited in the sense that it is about "preaching to the converted," and in that respect the Internet accentuates the digital divide between the haves and have-nots (Downey 2001).
It is true that the Web has transformed cyberspace, creating new opportunities for the politically skilled and sophisticated. However, despite the hype surrounding it, cyberspace very much resembles our old familiar world of everyday politics. Rather than romanticizing about the ability of the Net to radically transform politics within advanced capitalism on its own and to make politics like a conversation among equals, it would be more realistic to accept that cyberspace is merely another arena for the ongoing struggle for power, wealth and political influence (Resnick 1998). Cyberspace, no longer a wide-open frontier but normalised by now, has become a settled territory. While the Internet at large will continue to harbour a diversity of opinion and controversy, the Web is more likely to remain a vehicle for those already interested in politics.
Regarding policy recommendations, perhaps the solution lies in the view supported by Blumler and Gurevitch (2001, 2) who acknowledge the potential of the Internet to enhance public communication, but assert this is a "vulnerable" one. However, with appropriate policies and institutional support some of that potential may be realised. The challenge is to fashion and apply a public service remit for the Net so that a "civic commons" could be forged in cyberspace. Policy recommendations involve the establishment of an authority with responsibilities for arranging, publicising, moderating and reporting on a wide range of initiatives in digital democracy. The numerous initiatives which have been taking place in the
past few years could be brought under this authority which would subsequently back them up with production resources and expertise. It could be institutionalised as part of a public broadcaster such as the BBC, or as an new entity set up as an impartial public body (Blumler and Gurevitch 2001, 9-10). Such a policy step could forge an enlarged political culture where democracy becomes a continuing conversation rather than the occasional spin opportunity and technology can duly facilitate and increase public involvement and connection with politics.
Notes:
1.	The webmaster of this discussion did not provide for a constant monitoring of who contributes what.
2.	Although there is current work arguing that when people are online they engage in a free exchange of opinions and they do so because they enjoy the diversity of opinions available (Stromer-Galley, 2002).
3.	In most cases, it was possible to check the gender identity of Greek participants following the biographical note supplied by them. This means that I hypothesised that discussants were telling the truth about their identity, although there is no way to verify this since it was impossible to include any interviews in this research.
4.	The social and demographic profile of household and office access to the Internet across the European Union involves mostly young men, educated, living in a metropolitan area and belonging to the most fortunate social classes (executives, liberal professions and self-employed) (Eurobarometer Report 112, 2001)
5.	Particular thanks to Nick Jankowski, Slavko Splichal, and Matthew Hibberd at the Euricom Colloquium in Piran, Slovenia (15-20 September 2001) for their constructive comments.
References:
Barber, Benjamin. 1984. Strong Democracy: Participatory Politics for a New Age. Berkley:
University of California Press. Benson, T. W. 1996. Rhetoric, Civility and Community. Political Debate on Computer Bulletin
Boards. Communication Quarterly 44, 3, 359-378. Blumler, Jay and Michael Gurevitch. 1995. The Crisis of Public Communication. London: Routledge.
Blumler, Jay and Michael Gurevitch. 2001. The New Media and our Political Communication
Discontents: Democratizing Cyberspace. Information, Communication & Society 4, 1, 1-14. Boudourides, Moses and Evangelia, Drakou. 2000. Gender @Cyberspace. Paper contributed to the 4th European Feminist Research Conference, Bologna, Italy, available at http:// www.women.it/cyberarchive/files/boudourides_drakou.htm. Carey, James. 1995. The Press, Public Opinion, and Public Discourse. In T. Glasser and C. Salmon
(eds.), PublicOpinion and the Communication of Consent, 114-137. New York: Guilford. Coleman, Stephen. 1999. Cutting Out the Middle Man: From Virtual Representation to Direct Deliberation. In B. Hague and B. Loader (eds.), Digital Democracy, 195-211. London: Routledge.
Dahlberg, Lincoln. 2001. The Habermasian Public Sphere Encounters Cyber-Reality. Javnost/The
Public 8, 3, 83-96. Dewey, John. 1927. The Public and its Problems. New York: Henry Holt. Downey, John and Nancy Fenton. 2001. Constructing a Counter Public Sphere: The N30 Demonstrations and the World Wide Web. MeCCSA conference paper, Stanford Hall, Loughborough, UK.
Downey, John. 2001. In Search of the Net Election. Paper presented at the Euricom Colloquium
E-Networks: Arenas for Democratic Engagement, Piran, Slovenia, 15-20 September. Eurobarometer Report 112, Internet and the General Public, November 2001, available at http://
europa.eu.int/ISPO/basics/measuNng/i_mesunn.htmL
Gronlund, Kimmo and Maija Setala. 2001. ICTs and Transformations of Democracy. Research proposal, Dep. of Political Science, Abo Akademi University, and University of Turku, Finland.
Habermas, Jürgen. 1987. The Theory of Communicative Action. Voi. 2: Lifeworid and System: A Critique of Functionaiist Reason. Boston: Beacon Press.
Habermas, Jürgen. 1996 Between Facts and Norms: Contributions to a Discourse Theory of Law and Democracy. Cambridge: Polity.
Hale, Matthew, Julite Musso, and Christopher Weare. 1999. Developing Digital Democracy: Evidence from Californian Municipal Web Pages. In B. Hague and B. Loader (eds.) Digitai Democracy. London: Routledge.
Held, David. 1996. Modeis of Democracy (2nd edition). Cambridge: Polity Press.
Herring, Suzan. 1993. Gender and Democracy in Computer-mediated Communication. Eiectronic Journai of Communication, 3, 2.
Herring, Suzan. 1996. Posting in a Different Voice: Gender and Ethics in CMC. In C. Ess (ed.), Phiiosophicai Perspectives on Computer-Mediated-Communication, 115-145. Albany NY: State University of New York Press.
IDEA. 1997. Voter Turnout from 1945 to 1997: A Global Report on Political Participation. Stockholm: International Institute for Democracy and Electoral Assistance.
Internet Use in Greece. Online survey (in Greek) available at http://hyperion.math.upatras.gr/ survey/grinet.html.
Jankowski, Nickolas and Martine van Selm. 2000. The Promise and Practice of Public Debate in Cyberspace. In K. L. Hacker and J. v. Dijk (eds.), Digitai Democracy: Issues of Theory and Practice, 149-166. London: Sage.
Macintosh, Ann, Anna Malina, and Angus Whyte. 2001. Designing E-democracy for Scotland. Paper presented at the Euricom Colloquium E-Networks: Arenas for Democratic Engagement, Piran, Slovenia, 15-20 September.
Melucci, Alberto. 1985. The Symbolic Challenge of Contemporary Movements. Sociai Research 52, 789-816.
O'Donnell, Susan. 2001. Analysing the Internet and the Public Sphere: The Case of Womensiink. Javnost/The Pubiic 8, 1, 39-58.
Ranerup, Agneta. 1999. Contradictions When the Internet is Used in Local Government. In R. Heeks (ed.), Reinventing Government in the Information Age. London: Routledge.
Resnick, David. 1998. Politics on the Internet: The Normalization of Cyberspace. In C. Toulouse and T. Luke (eds.), The Poiitics of Cyberspace: a New Poiiticai Science Reader, 48-69. London: Routledge.
Robins, Kevin. 1996. Cyberspace and the World We Live In. In j. Dovey (ed.), Fractai Dreams: New Media in Sociai Context, 1-31. London: Lawrence & Wishart.
Schneider, Steven. 1996. A Case Study of Abortion Conversation on the Internet. Sociai Science Computer Review, 14, 4, 373-393.
Schudson, Michael. 1997. Why Conversation is Not the Soul of Democracy. Criticai Studies in Mass Communication 14, 297-309.
Stromer-Galley, Jennifer. 2002. New Voices in the Political Sphere: A Comparative Analysis of Interpersonal and Online Political Talk. Javnost/The Pubiic 9, 2, VSTAVI STRANI.
Touraine, Alain. 1998. What is Democracy? Transl. D. Macey. Oxford: Westview Press.