ISSA Proceedings 2006 – A Pragma-Dialectical Analysis Of Online Political Argumentation
1. Argumentation and Online Political Discussions
In recent years, we have witnessed a growing interest for the practice of argumentation using electronic conferencing systems.[i] Research has been conducted to understand how this asynchronous technology could facilitate the learning and practice of argumentation in the classroom (Marttunen, 1994, 1997; Marttunen & Laurinen, 2002 ; Schroeder & Zarinna, 1999), and Campos (2003, p. 300) even argue that “networked (many-to-many) communication has unique cognitive characteristics that are bound to collaborative argumentation”.
However, literature on argumentative practices in online political groups is much more limited. Most frequently, studies of political discussions online are bounded in the larger problematic of the ‘Internet and the Public Sphere’ and refer to the work of Habermas (1989) on argumentation and public deliberations in bourgeois society. Although general conclusions tend to be pessimistic, these studies note a high level of argumentation in online discussions (for a review in French, see Chaput, forthcoming). But on a closer look, one can find that their analysis is restricted to measuring the number of arguments in messages, and thus considering argumentation strictly as a product, which implies in turn to neglect the argumentation as process (cf. Blair & Johnson, 1987).
Our study aims therefore to understand the dynamic dimension of argumentation in online conferencing systems, by adopting what Plantin (2005, chap. 4) refers to as the “dialogical model of argumentation” in which interlocutors confront opposing viewpoints. We thus adopted the pragma-dialectical approach for it proposes “a systematic theory of argumentation” (cf. van Eemeren & Grootendorst, 2003) and, as we thought, can account for the high level of interactivity occurring in online discussion groups. For that purpose, we selected four (4) discussion threads from a lively online group in the Canadian province of Quebec called Politiquébec – a contraction of the words ‘politics’ and ‘Quebec’- whose mission is “To provide a space for constructive discussions about political issues in Quebec”.
Following a brief description of the pragma-dialectical method of analysis we used in this study, we present a synthesis of our results and identify some characteristics of networked communication that can complicate the resolution of critical discussions. Finally, we discuss the specifics of political argumentation and provide some appreciation of the pragma-dialectical method of analysis for online argumentative discourse.
2. Theory, Method and Data
Critical Discussion: Inspired by critical rationalism and speech act theory, Frans H. van Eemeren and Rob Grootendorst (1984, 1992) propose a theory of argumentation as critical discussion aimed at resolving a difference of opinion and going through four stages. During the confrontation stage, a viewpoint expressed by a party is put in doubt or rejected by another party; in the opening stage, the parties implied adopt the roles of protagonist and antagonist and respectively engage to defend or criticize the disputed statement. Common points of departure and rules are accepted at this stage. During argumentation stage, each party presents arguments to criticize or defend the disputed proposition, and finally, in the conclusion stage, we assist at the end of the dispute if the proposition is abandoned by the protagonist or the antagonist abandons its critique of the standpoint. As noted by pragma-dialecticians, critical discussion should be considered primarily as a tool for analysis:
The critical discussion model is a theory of how discourse would be structured if it were purely resolution oriented. It is not a theory of how discourse is structured nor is it a claim about what functions are or are not pursued in actual argumentation. Nevertheless, it plays an important role in the analysis of actual argumentation (van Eemeren, Grootendorst, Jackson & Jacobs, 1993, p. 26; italics added by the authors).
Reconstruction: In order to analyze online political discussions, we referred to the method of reconstruction proposed by pragma-dialectics, which is to consider empirical discourse as part of a critical discussion. The Procedure of reconstruction requires an analytic overview of the corpus studied (van Eemeren and Grootendorst, 1992, p. 93-94), that include the identification of the controversial viewpoints, the positions adopted by parties and the end of the discussion. It also serves to specify explicit as well as implicit premises and conclusions of arguments, the schemes of the arguments and the structure of argumentations. Descriptive analysis is combined with a critical counterpart, where the goal is to identify violation in the rules of a critical discussion.
Data: The corpus of our analysis consisted of four discussion threads retrieved between January and March 2005 in the Politiquébec online community. Basically, a discussion thread is a series of messages linked by the same theme. Published messages appear in chronological order and are organized in a linear way. Discussions are held on a daily basis about many political issues, are subjected to both formal and informal rules (charter and netiquette), and are under the surveillance of moderators. The threads we selected covered a wide range of political issues, from the independence of Quebec to democratic reform and students strike in post-secondary institutions.
3. The Dynamics of Online Argumentations
The pragma-dialectical methodology provides a very detailed analysis of actual discourse that could not be reproduced here. Instead, we summarize some of our main results, insisting on similarities and differences along with the normative ideal of the critical discussion. In order to illustrate our point, we include some excerpts taken from the thread on the students strike.[ii]
Confrontation Stage: We can first observe that the initial message launching a discussion thread is strongly argumentation-oriented. For instance, one interlocutor may advance a proposition by taking position on some issue. In other cases, as in the following example, one does not advance a viewpoint but invite others to debate on some actual or future event:
I’m not very familiar with this issue but I think that my ‘cegep’ and many others are holding a vote for an unlimited strike. I’ve red a little on the subject, and it’s about the savage cutbacks in the student’s grants program. I don’t know yet whether I will be in favor or against it so I’d like to discuss it with you.[iii]
Even without adopting a specific viewpoint and by acknowledging for his or her indecision, this interlocutor nonetheless settles the argumentative question and frames the potential positions of the protagonist and antagonist: those who stand in favor or against the strike. But at this point, the critical discussion is still virtual, because other participants must answer this request for a discussion to be held. Otherwise, the thread could simply ends here.
Opening Stage: As we just noted, the initial intervention is not a sufficient condition for the unfolding of a dispute, since it could stay without any responses. In the present case, it is only with the implication of other participants that a dispute definitely will be launched:
If I still were in ‘cegep’, I would be “against” [the strike].
But if I were in ‘cegep’, I would have voted in favor [of the strike].
Now at the opening stage, we can attribute the roles of protagonist and antagonist to those two interlocutors who entered the discussion. However, in all analyzed threads, interlocutors never express themselves on the attribution of roles or on common definitions, starting points and rules. This absence could confirm the explanation of van Eemeren and Grootendorst (1992) according to whom the opening stage in practical situations is essentially implicit, but it could also mean that participants never come to such an agreement, which would indicate then that the goal of the discussion is not to pursue the resolution of a conflict of opinions. Both these explanations seem equally plausible in absence of further information. In addition, we could take in concern that online discussion groups are already subjected to different kind of rules concerning the respect of opinions and individuals, as well as the commonly-shared knowledge about the practice of argumentation in those sites.
Argumentation Stage: According to the model of the critical discussion, argumentation is advanced by parties once they agreed to resolve their difference of opinion. But the asynchronous mode of communication found in computer conferencing systems enable participants to present arguments even before any disagreement has been expressed. In our example about the students strike, the undecided interlocutor who initiated the thread offers some reasons to support his concern:
I certainly agree to protest against the stupid policies of the government on education, but on the other side, I found the general strike to be a very radical mean, and I fear that it won’t change anything. You know, I really don’t want to miss my classes…
And so, while judging that a strike constitutes a justified mean to express opposition to the government’s decision, on the level of practical consequences for himself, he seems to doubt the success of the enterprise and fear for his own future as a student, thus explaining his indecision. Those arguments can then be reaffirmed or criticized by other interlocutors, or they might be ignored and replaced by stronger arguments. In the next example, the interlocutor formerly identified as antagonist chooses to present a slightly different argument to support his or her viewpoint:
Fist, planners of student manifestations are too disorganized, and then, the government knows very well that it is in fact only a minority of students who are concerned by these cutbacks.
In this case of multiple argumentations where two reasons are formulated to support the standpoint, practical consequences are still evoked but this time on a collective and not on the individual level, that is, concerns for students protest groups and the sum of all students.
And we can notice that both arguments contain some implicit premises and conclusions: first, if students manifestations groups are too disorganized, then they will not be able to succeed and thus strike is not a good option; and secondly, if only a minority of students are concerned, then the majority will not support the strike and so it is not a valuable action to undertake. A second antagonist develops a very similar argument using a deductive reasoning:
We talk about a gain of approximately $ 10,000 for less than a third of students. We talk of a possible loss of half a school year for all students. The choice is obvious, against the strike!
Some arguments can be repeated many times in the course of one dispute, and we can certainly question the effects of this repetition on the outcome of a discussion, for example if it contributes to reinforce an argument, to make it more commonly acceptable and convincing, up to the point where this argument is repeated so many times that it becomes naturalized and taken for an obvious fact. This question, however, goes beyond the objectives of the present paper.
The dialectical mode of argumentation explains that a question must be properly responded, that each argument advanced to support a viewpoint must be subject to criticism by the other party. But if some arguments are actually criticized by an opponent, many responses merely takes the aspect of a reframing move. Here, the protagonist also evokes the consequences of this decision, but on the scale of the long term period:
If you don’t protest now, you [the students in ‘cegep’], as future students of university, are going to pay a lot more than actual university students. It mainly depends on your long term vision. The objective of manifesting in ‘cegep’ is more for your future in university than for your present in ‘cegep’…
The issue is no more related to missing your classes today, but about paying a lot more tomorrow for higher education. The argumentation stage is thus built trough many messages that justify to vote for or against a students strike in the ‘cegeps’.
Conclusion Stage: Following some exchanges between one protagonist and many antagonists, the latter messages published in that thread are all by antagonists, and the discussion is put to an end when no further contributions are presented. A similar dynamics was observed in the other threads that we analyzed. It thus appears that the main contrast between the model of the critical discussion and the analyzed online discussions refers to the lack of resolution of disputes in online political groups. Many factors could explain why parties practically never agree at the end of a discussion, and we leave this subject for the next section.
4. Instabilities in Online Discussions
Various elements that characterize interactions in electronic conferencing may render difficult the good development of a critical discussion and limit the possibilities of a resolution of a difference of opinion using argumentation. Those characteristics concern the modes of participation, ideological antagonisms as well as the level of fallacies.
Modes of participation: For a critical discussion to be held, “participants must agree that there is some hope of resolving the disagreement through discussion and must enter into a cooperative search for resolution within a set of shared expectations about the way the search will be conducted” (van Eemeren & al., 1993, p. 27). However, the literature on online political discussions and our own observations suggest that such an engagement toward a common goal may not be the primary goal of participants and can be therefore difficult to maintain. First, many-to-many communications in computer conferencing resemble less to ‘dia-logue’ than to what we might call ‘multi-logue’ or ‘poly-logue’, in the sense that these exchanges are held simultaneously by numerous participants. This could in turn be explained by the device which make every intervention ‘public’ and offer the opportunity to break in or out an ongoing discussion. Furthermore, many participants participate simultaneously in many interactions (cf. Bentivegna, 1998), thus contributing to decrease collective attention and increase the fragmentation and multiplication of discussions (for a presentation of some features on commuter-mediated communication, see Marcoccia, 1998, pp. 17-18).
Multiple participations also create inequalities in the distribution of viewpoints among participants and that can complicate the resolution of disputes. In the threads we analyzed, we observe that the protagonist often faces a greater number of antagonists, and while not directly determining the issue of disputes, we can nonetheless suppose that a higher number of opponents will be more difficult to convince than one adversary. For example, the ratio between the number of protagonists and antagonists was from one to six in the second and third threads, and from two against seven in the fourth one (about the students strike).
Ideological Antagonisms: A second source of instability come from the opinions on political questions that often implicate value systems or ideologies, and as Walton mentions (1992, p.16), “political differences between right and left ideologies, it could be said, are precisely the sort of conflicts that do not lend themselves to resolution through simple discussions”. Windish, Amey and Grétillat (1995) illustrate this point in the nuclear debate in Switzerland, showing that conflicting parties developed entirely different worldviews that are totally incompatible. In similar cases where the contested viewpoints are irreconcilable, we begin to see more clearly why disputes are not resolved.
But this does not mean that viewpoints are deeply frozen and discussions impossible, because alongside with ideological antagonisms, Benoit-Barné (2002, p.163) asserts that public debate in electronic discussion groups can have a positive influence of the viewpoints shared by citizens: “Through this process, citizens assert, evaluate, and potentially reshape their taken-for-granted-assumptions about the principles that govern their lives”. However, the changes of viewpoints are more susceptible to occur in the long term than at the end of a sole discussion. In addition to that, when taking into consideration that those argumentations occur in the context of an online community where participants develop share a mutual knowledge or endoxa (cf. Tardini, 2005) where personal reputation may be involved, we better understand why a participant in a discussion might prefer to leave rather than having to abandon a standpoint, as in the following example where the protagonist could not convince the antagonists of his proposition:
It’s obvious that it won’t change anytime soon, and I am going to be preaching in the desert for a very long time. But I don’t care; I don’t want to be part of the gang…
And thus, in a public electronic conferencing system, we should not neglect this aspect of ‘performance’ on the part of the interlocutors, and implications for the preservation of one’s image or reputation in the community can have for consequences to limit the possibilities of truly critical discussions.
Fallacies: The last cause of instability during online political discussions on the Internet is related to the presence of fallacies in the course of verbal interactions. As specified by van Eemeren and Grootendorst (1987, p. 284), “In our pragma-dialectical conception, the term ‘fallacy’ is reserved for speech acts which hinder in any way the resolution of a dispute in a critical discussion”. Throughout all the messages we analyzed, we can fin many times when one of the ten normative discussion rules has been broken, and we could class those fallacies in three categories. First, there are those fallacies linked to invalid forms of arguments; in these cases, others participants are often able to refute the arguments or to reject them as unacceptable. More problematic are those fallacies that deal with ambiguity or unclear statements. Those are frequently due to a lack of good formulations on the part of the authors, but are also caused by the specific style of communication in computer conferencing that combines oral and written forms and can provoke misunderstandings. Problems related to interpretation of others contributions can sometimes degenerate into clashes: “Because of the very nature of the language and the multiplicity of meanings that words have, clashes can arise due to the different interpretations that texts can provide to readers” (Campos, 2002, online).
This brings us to the third and most important kind of fallacies, the personal attacks or argumentum ad hominem. The enflamed discourse of one interlocutor against another is not automatically harmful for human relations, as is argued by Papacharissi (2004) who insists that impoliteness does note necessarily means a lack of civism. But on the other side, an attack against an opponent risks to provoke a counter-attack, and thus transforming a critical discussion into an open quarrel. This frequent phenomenon in debates on the Internet is also known as ‘flame wars’ (for one illustration, see Herzog, Dinoff & Page, 1997, p. 411-413). Among the threads we analyzed, one transformed into a quarrel of this type where 11 out of 16 messages contained irritating elements and argumentation seemed pointless in those conditions. In another case, the publication of a hostile message had for consequence to put a premature end to the ongoing discussion.
In summary, it seems that a lack of engagement towards other participants in discussion, a relative absence of respect for adversaries, refusal to accept criticisms and a relatively high level of irritating messages can make it difficult to resolve difference of opinions in the course of online political discussions. According to these results, political discussions in networked communities fail to qualify as critical discussions. To conclude this paper, then, we question the specificities of political argumentations and consider the implications of pragma-dialectics as a tool for the analysis of verbal interactions in electronic conferencing.
On Political Argumentation: In our study, we considered online political argumentations strictly as a form of critical discussion. However, according to Walton (1992, p.130), political argumentations like partisan debates qualify as a kind of ‘mixed dialogue’ that combines elements of the quarrel and the critical discussion, democratic requirements and effective persuasions. This dual aspect of political argumentation has not been inquired in our study nor is considered by the model of the critical discussion, but recent efforts by pragma-dialecticians to integrate both rhetorical and dialectical aspects of argumentation (cf. van Eemeren & Houtlosser, 1999, 2000) could lead to interesting results in this field.
Another questioning is tied to the ‘political’ character of everyday argumentations in online discussions. While much has been written about the democratic potential of online political discussion groups, we could wonder if argumentation gets political by referring to political issues or by its enunciation in political contexts, which is not exactly the same. For instance, a café discussion about world politics can have little or no practical consequences, where a parliamentary debate will conclude argumentations with a decision, by voting a law, etc. According to Latour (2003), who is usually associated with the field of sociology of science and technology, a speech becomes political not for its content but as it is a container aimed at constituting and assembling groups. Therefore, we could conclude that online political discussions are not fully political argumentations, and following authors like Marcoccia (2003) or Wilhelm (1999), qualify these interactions as “laymen discussions” or “sociable conversations” oriented more towards the pleasure of exchanging opinions with others.
On Pragma-dialectics: We understand that the main advantage of the critical discussion model is that it offers a more elaborate method for the study of online political discussions than those found in the literature. As we noted earlier, most analysis of online political discussions limit argumentation to the presence or absence of arguments in a posted message, and fail to consider the situated and interactive aspect of online discussions. Secondly, the reconstruction procedure enables to reassemble and organize the numerous messages of a discussion thread that first appears to be ‘anarchic’. Finally, it allows a more detailed understanding of discussion dynamics even when exchanges appear, at first sight, to be the opposite of the ideal model of critical discussion. We also agree with the evaluation of pragma-dialectics by Bonevac (2003), who praises this theory for being dynamic, context-sensitive, and multi-agent, also offering a theory of fallacy and argumentative structure. All these characteristics make pragma-dialectics a strong toolkit for the analysis of practical argumentations in everyday situations.
However, Bonevac pursues his evaluation of pragma-dialectics by considering that it lacks to take in consideration the discussions implicating multiple participants, where there can be more than ‘one protagonist’ and ‘one antagonist’. Even though we accept the comments of van Rees (2003) who correct the point that ‘protagonist’ and ‘antagonist’ refers to roles in the discussion and not to actual persons, the fact that multiple interlocutors participate in online discussions certainly poses a additional difficulty for the analysis of online interactions or group discussions. For example, we noted that some participants join and leave ongoing discussions, and it is not obvious what role to attribute them in the dispute. Furthermore, participants don’t always have an established viewpoint on an issue, like our undecided speaker about the students strike. And the number of interlocutors who adopt one role or the other could have, as we mentioned earlier, an incidence in the outcome of a discussion. Under these conditions, could it be possible to talk of ‘many protagonists’ and ‘many antagonists’? Answer to this question would require more elaborated empirical analysis than we could provide here, and perhaps more consideration for the phenomenon of multi-partied discussions
In this paper, we presented an analysis of online political discussions in which the pragma-dialectical theory of argumentation was applied, and results suggest that argumentations do not lead to a resolution of the differences of opinions. Modes of participation in an online discussions, ideological antagonisms and fallacies could justify these deceiving results. Acknowledging these limits for the accomplishment of critical discussions online, it is then possible to propose different ways to enable better argumentative discourse; those could be linked to the technology itself or to the presence of a facilitator to enhance the unfolding of cooperative argumentations (cf. Campos, 2005), or to the pursue of common objectives that justify collaborative argumentations, just like in the case of shared knowledge in sites of the like of Wikipedia.
[i] This paper is based on the first author master thesis in communication sciences, written under the supervision of the second author (cf. Chaput, 2005).
[ii] In winter and spring of 2005, thousands of students in higher education institutions (universities and cegeps) protested against the decision by the government of Quebec to cutback funds in financial aid for students in difficult situations.
[iii] All original excerpts are written in French, but they have been translated for the purpose of this paper. Original messages can be found at this address: http://www.politiquebec.com/forum/ftopic9249.php
Benoit-Barné, C. (2002). “The Internet as a Space for Civic Discourse: The Case of the Unity Debate in Canada”. In S.D. Ferguson, & L.R. Shade, (Eds.), Civic Discourse in Canada: A Cacophony of Voices. (pp. 155-168). Westport, CT: Greenwood.
Bentivegna, S. (1998). Talking Politics on the Net (Research Paper R-20). Cambridge, MA: John F. Kennedy School of Government, Harvard University/The Joan Shorenstein Center on the Press, Politics, and Public Policy,
Blair, J.A. & R.H. Johnson (1987). “Argumentation as dialectical”. Argumentation, 1(1), 41-56.
Bonevac, D. (2003). “Pragma-dialectics and Beyond”. Argumentation, 17(4), 451-459.
Campos, M.N. (2002). “Competition, Lies and Dissimulation: Lessons from an Online Learning Clash”. Interpersonal Computing and Technology Journal, 8(1), <http://www.aect.org/Intranet/Publications/ipct-j/2002/Campos.asp>.
Campos, M.N. (2003). “The Progressive Construction of Communication: Towards a Model of Cognitive Networked Communication and Knowledge Communities”. Canadian Journal of Communication, 28(3), 291-322.
Campos, M. N. (2005, April). Nurses in action: A participatory design model for the building of a networked community. Paper presented at Practices of Assessment and Intervention in Action-Oriented Science and Technology Studies, Amsterdam.
Chaput, M. (2005). L’argumentation dans la communication : une analyse des interactions au sein d’une communauté politique en réseau [Argumentation in Communication : Analysis of Interactions in a Political Networked Community]. Unpublished Master Thesis, Department of Communication, Université de Montréal, Montréal.
Chaput, M. (forthcoming). « La dynamique argumentative des discussions politiques sur Internet » [The Argumentative Dynamics of Political Discussions on the Internet] COMMposite. <http://commposite.org>.
Eemeren, F.H. van & R. Grootendorst (1984). Speech Acts in Argumentative Discussions: A Theoretical Model for the Analysis of Discussions Directed towards Solving Conflicts of Opinion. Foris/Mouton de Gruyter, Dordrecht/Berlin.
Eemeren, F.H. van & R. Grootendorst (1987). “Fallacies in Pragma-Dialectical Perspective”. Argumentation, 1(1), 283-301.
Eemeren, F.H. van & R. Grootendorst (1992). Argumentation, Communication, and Fallacies: A Pragma-Dialectical Perspective. Hillsdale, NJ: Lawrence Erlbaum.
Eemeren, F.H. van & R. Grootendorst (2003). A Systematic Theory of Argumentation: The Pragma-Dialectical Approach. Cambridge: Cambridge University Press.
Eemeren, F.H. van & P. Houtlosser (1999). “Strategic manoeuvring in argumentative discourse”. Discourse Studies, 1(4), 479-497.
Eemeren, F.H. van & P. Houtlosser (2000). “Rhetorical Analysis Within a Pragma-Dialectical Framework: The Case of R. J. Reynolds”. Argumentation, 14(3), 293–305.
Eemeren, F.H. van, R. Grootendorst, S. Jackson, & S. Jacobs (1993). Reconstructing Argumentative Discourse. Tuscaloosa: University of Alabama Press.
Habermas, J. (1989). The Structural Transformation and the Public Sphere: An Inquiry into a Category of Bourgeois Society. Cambridge, MA: MIT Press.
Herzog Jr., H.A., B. Dinoff & J.R. Page (1997). “Animal Rights Talk: Moral Debate over the Internet”, Qualitative Sociology, 20(3), 399-418.
Latour, B. (2003). “What if We Were Talking Politics a Little?”. Contemporary Political Theory, 2(2), 143-164.
Marcoccia, M. (1998). « La normalisation des comportements sur Internet: Étude sociopragmatique de la netiquette » [The Normalization of Behaviors on the Internet: A Socio-pragmatic Study of Netiquette], in N. Guéguen & L. Tobin (Eds.), Communication, société et internet : actes du colloque GRESICO de Vannes, Université de Bretagne-Sud, 10 et 11 septembre 1998 (pp. 15-32). Paris; Montréal: L’Harmattan.
Marcoccia, M. (2003). « Parler politique dans un forum de discussion » [Talking Politics in Computer Conferencing]. Langage et société, 104, 9-55.
Marttunen, M. (1994, June). Practicing Argumentation through Computer Conferencing. Paper Presented at the Third International Conference on Argumentation, Amsterdam.
Marttunen, M. (1997). “Electronic Mail as a Pedagogical Delivery System: An Analysis of the Learning of Argumentation”. Research in Higher Education, 38(3), 345-363.
Marttunen, M. & L. Laurinen (2002). “Quality of Students’ Argumentation by E-mail”. Learning Environments Research, 5(1), 99-123.
Papacharissi, Z. (2004). “Democracy Online: Civility, Politeness, and the Democratic Potential of Online Political Discussion Groups”. New Media & Society, 6(2), 259-283.
Plantin, C. (2005). L’argumentation : Histoire, théories et perspectives [Argumentation: History, Theories and Perspectives]. Paris : Presses Universitaires de France.
Rees, M.A. van (2003). “Within Pragma-dialectics: Comments on Bonevac”. Argumentation, 17(4), 461-464.
Schroeder, E.E. & E.A. Zarinnia (1999, February-March). Argumentation Online: The Use of Computer Conferencing. Paper presented at the SITE 99: Society for Information Technology & Teacher Education International Conference, San Antonio, TX.
Tardini, S. (2005). “Endoxa and Communities: Grounding Enthymematic Arguments”. Studies in Communication Sciences: Argumentation in Dialogic Interaction, 279-294.
Walton, D.N. (1992). Plausible Argument in Everyday Conversation. Albany: State University of New York Press.
Wilhelm, A.G. (1999). “Virtual Sounding Boards: How Deliberative is Online Political Discussion?” In B.N. Hague & B.D. Loader (Eds.), Digital Democracy: Discourse and Decision Making in the Information Age (pp. 154-178). London; New York: Routledge.
Windisch, U., P. Amey & F. Grétillat (1995). « Communication et argumentation politiques quotidiennes en démocratie directe » [Everyday Political Communication and Argumentation in Direct Democracy]. Hermès 16: Argumentation et rhétorique 2, 57-72.