While fallacies have been a major focus of the study of arguments since antiquity, scholars in argumentation theory are still struggling for suitable frameworks to approach them. A fundamental problem is that there seems to be no unique category or kind such as ‘fallacy’, and arguments can be seen as fallacious for many various reasons. This heterogeneity does not invalidate the need to study fallacies, but it poses serious difficulties for general systematic approaches. On the other hand, the numerous repeated attempts to find satisfactory perspectives and tools, together with the critical discussions of these attempts, have increasingly contributed to our understanding of the more local situations where different types of fallacies appear, of how and in what circumstances they are fallacious, and, of which contexts and disciplinary areas are relevant to the study of certain types of fallacies.

This paper [i] aims to illustrate these issues by selecting one fallacy type as its subject, the argumentum ad verecundiam. The main thesis is that argumentation studies can gain a reasonable profit from consulting a field, the social studies of science, where the problem of appeals to authority has lately become a central issue. The first section summarizes and modestly evaluates some recent approaches to ad verecundiam arguments in argumentation studies. The second section overviews the problem of expert dependence as discussed in social epistemology and science studies. The third section presents a rough empirical survey of expert authority appeals in a context suggested by the previous section. The paper concludes by making some evaluative remarks.

1. The problem of ad verecundiam arguments
An argumentum ad verecundiam can loosely be defined as an inappropriate appeal to authority. As there are different types of authority, ranging from formal situations to informal contexts, the function and success of authority appeals can vary broadly. This paper is concerned with one type of authority, namely cognitive or epistemic authority, i.e. those people who have, or who are attributed by others, an outstanding knowledge and understanding of a certain subject or field – in modern terms, with experts. While not all authorities are experts and, arguably, not all experts are epistemic authorities (as we move from ‘know-that’ to ‘know-how’ types of expert knowledge), the paper is restricted to the problem of epistemic authority appeals, or, in short, appeals to experts.

To problematize the definition of ad verecundiam, let us distinguish between two questions: (1) What does it mean for an appeal to authority to be inappropriate? (2) How do we know if an appeal to authority is inappropriate? From the analytical point of view, the first question is primary since one can identify an ad verecundiam argument only if one knows what it is, and, conversely, once we know how an authority appeal can be inappropriate we are, albeit not necessarily immediately, in the position to distinguish a correct appeal from an incorrect one. However, a more epistemological perspective suggests, as will be illustrated below, that one cannot tell what it means for an appeal to be incorrect before one knows how to find it out, and any specific expansion of the above definition is likely to fail when ignoring the more practical dimension opened by the second question.

In order to spell out this problem in a bit more detail, it is worth considering two recent influential approaches: Douglas Walton’s inferential approach and the functional approach by the pragma-dialectical school. Walton suggests that appeals to authority can be reconstructed according to the following argument scheme (Walton 1997, p. 258):
E is an expert in domain D
E asserts that A is known to be true
A is within D
Therefore, A may plausibly be taken to be true

If appeals to authority are implicit inferences, then the first question (What does it mean for an appeal to authority to be inappropriate?) may be answered by analyzing and evaluating the inference: either the inference form is unsound, or some of the premises fail to be true. The soundness of the argument raises serious problems, for it is obviously not deductively valid, nor can it be classified as an inductive inference in any traditional sense (generalizing or statistical, analogical, causal, etc.), but we can certainly attribute to it a degree of ‘plausibility’ the conclusion claims and put aside further investigations into argument evaluation. What Walton seems to suggest is that it is the failure of the premises that renders the conclusion unacceptable. And this means that in order to be able to answer the second question (How do we know if an appeal to authority is inappropriate?), one needs simply to know who is expert in which area, what they assert, and to which area these assertions belong.

The situation becomes more complicated at a closer look. Walton lists a number of questions one has to ask to establish the truth of the premises (ibid., p. 25):
1. Expertise Question: How credible is E as an expert source?
2. Field Question: Is E an expert in the field that A is in?
3. Opinion Question: What did E assert that implies A?
4. Trustworthiness Question: Is E personally reliable as a source?
5. Consistency Question: Is A consistent with what other experts assert?
6. Backup Evidence Question: Is E’s assertion based on evidence?

While these questions are clearly relevant, it is important for us to note that in order to be able to tell whether an authority appeal is correct, one needs to possess a huge amount of knowledge. Elements of this knowledge are of various nature: knowledge of ‘fields’ (like scientific disciplines and sub-specializations), degrees of credibility (like scientific rankings, credentials, institutions and statuses),  logical relations of assertions in a technical field, other experts and their claims, personal details, matters concerning what it means to be evidentiary support, etc. In the pessimistic reading this scenario suggests that laypersons will hardly be able to acquire all this knowledge, appeals to authority will generally be insufficiently supported, and that the interlocutors of a discussion (if they themselves are not experts in the field in question) will rarely be able to tell whether an appeal to authority is appropriate or not. In the optimistic reading it points out themes and areas that are primarily relevant to the first question, through the second question to which the first is intimately connected, and it embeds the problem of ad verecundiams in a specific theoretical context in which they can be analyzed.

While Walton’s approach focuses on what it means for an expert claim to be unreliable (‘incorrect authority’), the pragma-dialecticians place the emphasis on the use of authority appeals (‘incorrect appeal’). According to their functionalization principle, one needs to look at the function of an assertion within the discourse in order to tell whether it contributes to the final dialectical aim of rationally resolving differences of opinion. Fallacies are treated as violations of those rules of rational discussion that facilitate this resolution. In one of their book (Eemeren & Grootendorst 1992, pp. 212-217), they use the ad verecundiam to illustrate that the same type of fallacy (as understood traditionally) can violate different rules at different stages of the dispute, and thus it can serve various purposes. An ad verecundiam argument can thus violate the Argument Scheme Rule at the argumentation stage, i.e. the interlocutor can present an appeal to authority instead of a correctly applied and appropriate argument scheme when defending her standpoint. But ad verecundiams can also be used at the opening stage to violate the Obligation-to-defend Rule: a party refuses to provide adequate argumentative support for her claim when asked, and offers an appeal to authority instead. Moreover, they can violate the Relevance Rule in the argumentation stage again, when authority appeals are used as non-argumentative means of persuasion.

Just as the pragma-dialectical approach offers a radically different answer from Walton’s to the question of what it means for an appeal to authority to be incorrect, the possible answers to the question of how to recognize these incorrect appeals are also strikingly different in the two cases. For pragma-dialecticians, one needs to identify the function of such appeals in the context of the entire dispute as reconstructed according to a fully-fledged theory with its stages and rules and further assumptions. Pragma-dialectics offers an exciting framework in which one can focus on the pragmatic use of elements in argumentation, but it pays less attention to the study of what is used. Surely, an appeal to authority can often be used as to evade the burden of proof, or to intimidate the other party by non-argumentative means, but in many other cases it is simply unavoidable to defer to expert testimonies, even among rational discussants engaged in a critical dispute. As the next section argues, such appeals are actually so widespread and indispensable that the study of abusive appeals seems only secondary in importance.

This paper studies problems that are more similar to Walton’s questions than to the issues raised by the pragma-dialectical approach, although it does not accept the inferentialist framework with its interest in argument schemes (in that the focus will be on elements of knowledge answering Walton’s questions, rather than seeing these elements as connected in an argument scheme). The possibility of ad verecundiam arguments, just as the possibility of correct authority appeals, depends on non-experts’ ability to evaluate the reliability of expert claims. In the followings, recent philosophical and sociological discussions will be summarized in order to investigate such possibilities.

2. Some recent approaches to expertise
It is a common recognition among many fields that, in present cultures, the epistemic division of labor has reached a degree where trust in expert opinions is not only indispensible in many walks of life, but also ubiquitous and constitutive of social existence. Thus the problem of expertise has gained increasing focus in psychology (Ericsson et al. 2006), in philosophy (Selinger and Crease 2006), or in the social studies of science where the initiative paper by Collins and Evans (2002) has become one of the most frequent points of reference in the field. Other forms of an ‘expertise-hype’ can be seen in the theory of management, in risk assessment, in artificial intelligence research, in didactics, and in a number of other fields having to do with the concept of ‘expert’.

For the present purposes, a useful distinction is borrowed from recent literature on the public understanding of science. Two approaches are contrasted to frame the expert-layperson relationship for the case of science: the deficit model and the contextual model (Gross 1994, Gregory and Miller 2001). In the deficit model the layperson is viewed as someone yet ignorant of science but capable of having their head ‘filled’ with knowledge diffusing from science. Such a ‘filling process’ increases, first, laypeople’s scientific literacy (and their ability to solve related technical problems), second, their degree of rationality (following the rules of scientific method), and third, their trust in and respect for science. Recently, this model has been criticized as outdated and suggested to be replaced by the contextual model, according to which members of the public do not need scientific knowledge for solving their problems, nor do they have ‘empty memory slots’ to receive scientific knowledge at all. Instead, the public’s mind is fully stuffed with intellectual strategies to cope with problems they encounter during their lives, and some of these problems are related to science. So the public turn to science actively (instead of passive reception), more precisely to scientific experts, with questions framed in the context of their everyday lives.

The strongly asymmetrical relationship between experts and the public suggested by the deficit model is at the background of a groundbreaking paper by the philosopher John Hardwig (1985), who coined the term ‘epistemic dependence’. His starting point is the recognition that much of what we take to be known is indirect for us in the sense that it is based on our trust in other people’s direct knowledge, and the greater the cultural complexity is, the more it is so. Hardwig takes issue with the dominantly empiricist epistemological tradition, where these elements of belief are not considered rational inasmuch as their acceptance is not based on rational evidence (since the testimony of others does not seem to be a rational evidence).

Hardwig takes a pessimistic position regarding the possibility of laypeople’s assessment of expert opinions: since laypeople are, by definition, those who fall back on the testimony of experts, they have hardly any means of rationally evaluating expert claims. Of course, laypeople can ponder on the reliability of certain experts, or rank the relative reliability of several experts, but it can only be rationally done by asking further experts and relying on their assessments – in which case we only lengthened our chain of epistemic dependence, instead of getting rid of it (p. 341). So, according to Hardwig, we have to fully accept our epistemic inferiority to experts, and either rely uncritically on expert claims or, even when criticizing these claims, we have to rely uncritically on experts’ replies to our critical remarks (p. 342).

However, at one point even Hardwig admits that laypeople’s otherwise necessary inferiority can be suspended in a certain type of situations that he calls ad hominem (p. 342):
The layman can assert that the expert is not a disinterested, neutral witness; that his interest in the outcome of the discussion prejudices his testimony. Or that he is not operating in good faith – that he is lying, for example, or refusing to acknowledge a mistake in his views because to do so would tend to undermine his claim to special competence. Or that he is covering for his peers or knuckling under to social pressure from others in his field, etc., etc.

But Hardwig warns us that these ad hominems “seem and perhaps are much more admissible, important, and damning in a layman’s discussions with experts than they are in dialogues among peers”, since ad hominems are easy to find out in science via testing and evaluating claims (p. 343). And apart from these rare and obvious cases, laypeople have no other choice left than blindly relying on expert testimonies.

Nevertheless, Hardwig’s examples imply that in some cases it is rational and justified for a layperson to question expert testimonies. Recent studies on science have pointed out various reasons for exploiting such possibilities. For instance, there are formal contexts at the interfaces between science and the public, such as legal court trials with scientific experts and non-expert juries, where laypeople’s evaluations of expert claims are indispensible. Such situations are considered by the philosopher of law Scott Brewer (1998), who lists what he identifies as possible routes to ‘warranted epistemic deference’, i.e. means of non-expert evaluation of expert claims.

Substantive second guessing means that the layperson has, at least to some degree, epistemic access to the content of expert argument and she can understand and assess the evidences supporting the expert claim. Of course, as Brewer admits, such situations are rare since scientific arguments are usually highly technical. But even with technical arguments one has the option of using general canons of rational evidentiary support. If an expert argument is incoherent (e.g. self-contradicting) or unable to make or follow basic distinctions (in his example, between causing and not preventing) then, even for the layperson, it becomes evident that such an argument is unreliable. Laypersons can also judge by evaluating the demeanor of the expert: they may try to weigh up how sincere, confident, unbiased, committed etc. the expert is, and this obviously influences to what degree non-experts tend to rely on expert claims. However, all this belongs to the ethos of the speaker and Brewer emphasizes the abusive potential in demeanor often exploited by the American legal system. The most reliable route, according to him, is the evaluation of the expert’s credentials, including scientific reputation. He adopts the credentialist position  even while acknowledging that it is laden with serious theoretical difficulties, such as the regress problem (ranking similar credentials requires asking additional experts), or the underdetermination problem (similar credentials underdetermine our choice between rivaling experts).

Another reason for focusing on the possibility of lay evaluations of expert claims is the recognition that experts do not always agree with one another, and such situations are impossible to cope with in terms of simple epistemic deference. According to the contextual model, the public need answers to questions they find important (regarding health, nutrition, environmental issues, etc.), and these questions typically lack readymade consensual answers in science. Alvin Goldman, a central figure in social epistemology, tries to identify those sources of evidence that laypeople can call upon when choosing from rivaling expert opinions – in situations where epistemic solutions of ‘blind reliance’ break down (Goldman 2001).

Goldman distinguishes between two types of argumentative justification. ‘Direct’ justification means that the non-expert understands the expert’s argument and is able to evaluate it, similarly to what Brewer means by substantive second guessing. But when arguments are formulated in an unavoidably esoteric language, non-experts still have the possibility to give ‘indirect’ justification by evaluating what Goldman calls argumentative performance: certain features of the arguer’s behavior in controversies (quickness of replies, handling counter-arguments, etc.) indicate the degree of competence, without requiring from the non-expert to share the competences of the expert. Additional experts can be used in two ways in Goldman’s classification: either by asking which of the rivaling opinions is agreed upon by a greater number of experts, or by asking meta-experts (i.e. experts evaluating other experts, including credentials) for judgment on the expert making the claims. Similarly to Hardwig’s ad hominem cases, Goldman also considers the possibility of identifying interests and biases in the arguer’s position. But what he sees as the most reliable source of evidence is track-record. He argues that even highly esoteric domains can produce exoteric results or performances (e.g. predictions) on the basis of which the non-expert becomes able to evaluate the cognitive success of the expert.

Despite their different answers to the question of most reliable decision criteria, Brewer and Goldman agree that sounder evaluation needs special attention, either by studying the institutional structure of science (to weigh up credentials) or by examining specialists’ track-records. But why should the public take the effort of improving their knowledge about science? If we turn from philosophical epistemology to the social studies of science and technology, we find an answer at the core of the discipline: because laypeople’s lives are embedded in a world in which both science and experts play a crucial role, but where not all experts represent science and even those who do, represent various, often incompatible, claims from which laypeople have to choose what to believe.

The program called ‘studies of expertise and experience’ (SEE) evolved in a framework shaped by these presuppositions, initiated by science studies guru Harry Collins and Robert Evans (2002, later expanded to 2007). Their initial problem is that “the speed of politics exceeds the speed of scientific consensus formation” (Collins and Evans 2007: 8), meaning that decision making processes outside science (politics, economy, the public sphere, etc.) are usually faster than similar processes in science. This gives rise to what they call ‘the problem of legitimacy’ (Collins and Evans 2002: 237): how is technological decision making possible given the growing social uncertainty? They claim that solutions are already achieved, or pointed to, in the field of ‘public participation in science’. However, a related but yet unsolved problem is ‘the problem of extension’, i.e. to what degree should the public be engaged in technical decision making? The program of SEE is meant to provide normative answers to this question.

In this framework the term ‘expert’ has a wide range of applications, since experts are defined as those “who know what they are talking about” (Collins and Evans 2007: 2), which is based on immersion in communicative life forms. Forms of expertise range from ubiquitous skills (such as native language usage) to the highest degree of scientific specialization, as summarized in ‘the periodic table of expertises’ (p. 14). This table includes, in addition to types of specialist expertise, those forms of ‘meta-expertise’ that can be used to judge and evaluate specialist expertise.

According to the SEE, the public live in a society where they are conditioned to acquire skills and ‘social intelligence’ needed to cope in an expert culture. Non-experts are able to come to decisions regarding technical questions on non-technical grounds, based on their general social intelligence and discrimination. As Collins and Evans claim (p. 45), the “judgment turns on whether the author of a scientific claim appears to have the appropriate scientific demeanor and/or the appropriate location within the social networks of scientists and/or not too much in the way of a political and financial interest in the claim”. So people (or at least sufficiently informed people) in Western societies have enough social skills to form correct judgments (in their examples, about astrology, or manned moon landings, or cold fusion) without possessing field-specific technical knowledge. Also in their ‘periodic table’ one can find ‘meta-criteria’ for evaluating experts, such as credentials, past experience and track record, but all these criteria need special focus on the layperson’s side to asses, apart from their basic general social skills.

To sum up the main points of this section: It seems clear that despite all the possible theoretical difficulties, laypeople can and do make evaluations of expert claims, and since laypeople are not experts in terms of their cognitive domains, these evaluations are based on criteria external to the specialist domain. Also, such external evaluations are not only frequent but generally unavoidable in a world of rivaling experts and consensus-lacking controversial issues. But while these philosophical analyses give rise to different while partly overlapping normative solutions, it remains unclear whether these solutions are really functional in real life situations. The next section attempts to examine this question.

3. A rough case study
The recent worldwide public interest in the H1N1 influenza pandemic threat , and in the corresponding issues concerning vaccination, provides a highly suitable test study for the above theoretical approaches. First, the case clearly represents a technical topic about which various and often contradicting testimonies were, and still are, available. Second, despite the lack of scientific consensus, decisions had to be made under uncertain circumstances, both at the level of medical policy and at the level of individual citizens who wanted to decide eagerly whether vaccination (and which vaccination) is desirable. Huge numbers of non-experts were thus forced to assess expert claims, and come to decisions concerning technical matters lacking the sufficient testimonial support.

Luckily, the internet documented an overwhelming amount of lay opinions, mostly available in the form of blog comments. In order to see how laypeople do assess expert claims, I looked at four Hungarian blog discussions (as different as possible) on the issue, examined 600 comments (from October-November 2009) trying to identify explicitly stated criteria of evaluative decisions that I found in 110 cases.[ii] The work is rather rudimentary and methodologically rough at the moment, but it may suffice to yield some general results to be tested and elaborated by future work. I approached the material with a ready-made typology of warrants abstracted from the theoretical literature, and I counted the number of instances of the abstract types. I disregarded those comments which did not contain any clear opinion, or where arguments (reasons, warrants) were not given in favor of (or against) the standpoint, or which were redundant with respect to earlier comments by the same user. Some comments contained more than one type of argument or warrant, where all different instances were considered. The tested categories distilled from the literature cited in the previous section are the following.

(1) The first group is argument evaluation by the content, i.e. Brewer’s ‘substantive second guessing’ or Goldman’s ‘direct argument justification’, when laypersons interiorize technical arguments as their own and act as if they had sufficient cognitive access to the domain of expertise. Example: “I won’t take the vaccine, even if it’s for free in the first round. The reason is simple: the vaccine needs some weeks before it takes effect, and the virus has a two week latency. And the epidemic has already begun…” (cotcot 2009, at 10.06.13:06).

(2) The second group contains those contextual discursive factors that are indirectly tied up with the epistemic virtue of arguments. (2a) Such is the consistency (and also coherence) of arguments, clarity of argument structure, supporting relations between premises and conclusions, etc. Example: “Many of those who go for this David Icke type humbug are afraid of the crusade against overpopulation, so they’re against inoculation, which is a contradiction again” (cotcot 2009, at 10.05.22.:43). (2b) A similar matter is the degree of reliability of argument scheme used by the expert. Arguments can be weakened, albeit at the same time increased in persuasive potential, by different appeals to emotions and sentiments, or by abusive applications of ad hominems, or by irrelevant or misleading appeals to authority, etc. Also, dialectical attitude (instead of dialectical performance) can be highly informative, i.e. moves and strategies in controversies, including conscious or unnoticed fallacies such as straw man, red herring, question begging, shifting the burden of proof, and more generally, breaking implicit rules of rational discussion. I found that these kind of assessments are very rare, still an arguable example is: “It is a bad argument that something is a good business. Safety belt is also a good business for someone, and I still use it.” (vastagbor 2009, at 11.04.14:52)

(3) Hardwig, Goldman and the SEE all emphasize the role of detecting interests and biases. Considering these factors belongs to the field of ‘social intelligence’, and precisely because these are ubiquitous they do not need focused effort and training to improve (as opposed to the argumentative factors mentioned above). Example: “I’d be stupid to take the vaccine. All this mess is but a huge medicine business.” (vastagbor 2009, at 11.04.12:26)

(4) Social intelligence covers the ability to evaluate the reliability of experts, instead of judging the arguments. (4a) The simplest case is unreflected deference or blind trust. Example: “My aunt is a virologist and microbiologist. She never wants to persuade me to take any vaccination against seasonal flu, but this time it is different…” (reakcio 2009, at 11.14.15:21) (4b) As the credentialist solution suggests, laypeople can estimate the formal authority of different experts by judging their ranks or positions. Example: “So, when according to the Minister of Healthcare, and also to Czeizel [often referred to as “the doctor of the nation”], and also to Mikola [ex-Minister of Healthcare], Hungarian vaccine is good, then whom the hell would I believe when he says that it isn’t?” (szanalmas 2009, at 11.04.12:22) (4c) Also, quite similarly, one may discredit testimonies by claiming that the expert is a wrong or illegitimate authority. Example: “Why should I want to believe the doctor who tried to convince my wife not to take the vaccine a few days ago, and then tried to rope her in Forever Living Products? Or the doctor who does not even know that this vaccine contains dead virus, not live? […] So these are the experts? These are the doctors to protect our health? ” (szanalmas 2009, at 11.04.12:22)

(5) Finally, there are various forms of commonsensical social judgments not explicitly dealing with interests or authorities, as expected by the SEE programme. Three examples: “Let us not forget that first there wasn’t even a date of expiry on the vaccine” (vastagbor 2009, at 11.04.12:00). “This huge panic and hype surrounding it makes things very suspicious” (vastagbor 2009, at 11.04.12:02). “The vaccine comes from an unknown producer, and the formula is classified for 20 years…” (szanalmas 2009, at 11.03.16:01).

The results are summarized by the table below:

Table 1. Number of argument type instances in blog comments

Judgment by content (type 1) is quite frequent, contrary to the recommendation of normative approaches emphasizing that the demarcation between experts and laypeople correlates with the distinction between those who have the ability to understand technical arguments and those who do not. There are several possible reasons for this. One is that laypeople do not like to regard themselves as epistemically inferior, and try to weigh up expert arguments by content even if they lack the relevant competences. Another is that the publicly relevant technical aspects of the H1N1 vaccine issue are far less esoteric than for many other scientific issues, and there is a lot to understand here even for non-virologists and non-epidemiologists. Another is that while people form their opinions on testimonial grounds, they often refrain from referring explicitly to their expert sources (especially in blog comments resembling everyday conversations), and state their opinion as if they themselves were the genuine source.

In contrast, assessment informed by argument structure and form (types 2a and 2b) is pretty rare, even when it seems plausible to assume that, in some respect, judgments on general argumentative merits require different competences from the specialist judgments based on content. But just as most people are not virologists, they are very rarely argumentation theorists, so they are usually not aware of the formal structure or type of arguments they face, or the relevant fallacies.

The identification of interests and biases (type 3) is a really popular attitude in the examined material. While part of the reason for this might be that the studied case is untypical in that very clear interests were at play (the vaccine producer company seemed to have some connections with certain politicians), this popularity is nevertheless in line with the expectation shared by most of the cited authors about the relative importance of such considerations.

Also, simple deference (type 4a) is a relatively widespread attitude, despite the fact that contradicting expert testimonies were obviously available in this specific case. While Brewer and Goldman suggest ranking and comparing expert authorities, it seems that such ranking is pretty rare in actual arguments. Neither considering formal or institutional indicators of authority (type 4b) nor questioning the legitimacy of putative experts (type 4c) seem frequent. Perhaps this is partly because people tend to base their trust on personal acquaintances (the SEE calls this ‘local discrimination’). Another likely reason is the public’s relative ignorance in the field of scientific culture and social dimension of the workings of science: unlike other important cultural spheres like that of politics, economy, or sports, about which laypeople are more likely to make reliable social evaluations, science as a social system is hardly known by the public.

What I found to feature most often in laypeople’s decisions is ‘commonsensical’ forms of social judgments, practically those that consider factors other than direct interests or expert authorities. Obviously, social structures and mechanisms are easier to understand (based on our fundamental experience with them) than technical arguments, even if peculiar features of the social world of science are much less widely known than the social reality in general.

In sum, public assessment of expert claims is based on skills and competences acquired through everyday social interaction, and the applicability of these skills in restricted cognitive domains is generally presupposed without further reflection. While the deficit model suggests either blind reliance or the acquisition of the same domain-specific cognitive skills shared by experts, the contextual model points to the possibility of a kind of contextual knowledge that would enable the public to assess expert claims more reliably than merely adopting the most general social discriminations, without having to become experts themselves in all the fields in which they need to consult experts. However, it seems that the evaluative criteria suggested by normative accounts are rarely used in actual decisions.

4. Conclusion
If we set aside the question of how expert authority appeals are used inappropriately and, instead, focus on what it requires to tell whether an expert argument is reliable at all – which is essential when critical discussions are aimed at rational decisions – then it turns out that the depth and range of knowledge required from the public seems to escape the confines of the study of argumentation in general. Surely, evaluations of expert claims supported by arguments can be significantly improved by awareness of some basic concepts in argumentation studies, regarding e.g. the consistency (and also coherence) of arguments, clarity of argument structure, relations between premises and conclusions, argument schemes and their contexts, fallacious argument types, etc. However, it is important to realize that an even more efficient support to such evaluations can be gained by some familiarity with the social dimension of science (as opposed to technical knowledge in science, restricted to experts): credentials, hierarchies of statuses and institutions, types and functions of qualifications and ranks, patterns of communication in science, the role of different publications and citations, mechanisms of consensus formation, disciplinary structures, the nature of interdisciplinary epistemic dependence and resulting forms of cooperation, etc.

While this contextual (rather than substantial) knowledge about science may be essential in societies that depend in manifold ways on the sciences, it is not obvious how and why the public attention could turn to these matters. If spontaneous focus on scientific expertise might be unrealistic to expect from the public, there are organized ways to improve cognitive attitudes toward science. One relevant area is school education where, in most countries at present, science teaching consists almost exclusively of scientific knowledge at the expense of knowledge about science (and awareness of argumentation is also rather rare in school curricula). Another area is science communication, including popular science and science news, where contextual information about matters mentioned above is typically missing but would be vital for enhancing understanding. Also, improving forms of public participation in, or engagement with, science is an obvious way to increase public interest and knowledge.

All in all, as our cultural dependence on cognitive experts has been recognized as a fundamental feature of our world, the problem of appeals to expert authorities seems both more complex and more crucial than when viewed simply as an item on the list of fallacy types in argumentation studies. The paper tried to show that the study of argumentation can shed light on some important aspects of authority appeals. However, this does not mean that the problem of expertise is, or should be, a substantive field of argumentation studies, or that argumentation theorists should substantially evaluate claims made by experts. But argumentation studies (as a field of expertise itself) can obviously offer important contributions to the study of expertise, especially when theoretical approaches are supplemented with an empirical study of argumentative practice. Such a perspective may put the emphasis on aspects that are, as seen in pragma-dialectics, rather different from the traditional question of ‘How do we know that the discursive partner appealed to the wrong expert claim?’ The latter problem is also vital, and in order to tell how to answer it one needs to find out a good deal about science and its relation to the public. The best way to do so seems to be to consult, or better cooperate with, those disciplines that take related problems as their proper subject.

NOTES
[i] The work was supported by the Bolyai Research Scholarship, and is part of the HIPST project. For section 2, the paper is partly based on an earlier work to be published in Teorie Vĕdi (‘Contextual knowledge in and around science’), while the empirical work presented in section 3 was done for Kutrovátz (2010).
[ii] The four blogs are: cotcot (2009) – an online fashion and health magazine (mostly for and by women); szanalmas (2009) – an elitist community blog site, often highly esteemed for intellectual autonomy; vastagbor (2009) – a political blog with marked right-wing preferences; reakcio (2009) – a cultural/political blog with right-wing tendencies.

REFERENCES
Brewer, S. (1998). Scientific Expert Testimony and Intellectual Due Process. The Yale Law Journal, 107, 1535-1681.
Collins, H., & Evans, R. (2002). The third wave of science studies: Studies of expertise and experience. Social Studies of Science, 32, 235-296.
Collins, H., & Evans, R. (2007). Rethinking Expertise. Chicago: The University of Chicago Press.
cotcot (2009). http://cotcot.hu/test/cikk/4819. Cited 14 July 2010.
Eemeren, F. H. van, & Grootendorst, R. (1992). Argumentation, Communication, and Fallacies. New Jersey & London: Lawrence Erlbaum Associates.
Ericsson, K. A., Charness, N., Feltovich, P. J., & Hoffman R. R. (Eds.) (2006). The Cambridge handbook of expertise and expert performance. Cambridge: Cambridge University Press.
Goldman, A. I. (2001). Experts: which ones should you trust? Philosophy and Phenomenological Research, 63, 85-109.
Gregory, J., & Miller, S. (2001). Caught in the Crossfire? The Public’s Role in the Science Wars. In J. A. Labinger & H. Collins (Eds.), The One Culture? A Conversation about Science (pp 61-72). Chicago and London: University of Chicago Press.
Gross, A. G. (1994). The roles of rhetoric in the public understanding of science. Public Understanding of Science, 3, 3-23.
Hardwig, J. (1985). Epistemic dependence. The Journal of Philosophy, 82, 335-349.
Kutrovátz, G. (2010). Trust in Experts: Contextual Patterns of Warranted Epistemic Dependence. Balkan Journal of Philosophy, 2, 57-68.
reakcio (2009). http://reakcio.blog.hu/2009/11/04/a_felelem_bere_6_milliard?fullcommentlist=1. Cited 14 July 2010.
Selinger, E., & Crease, R. P. (Eds.) (2006). The philosophy of expertise. New York: Columbia University Press.
szanalmas (2009). http://szanalmas.hu/show_post.php?id=64722. Cited 14 July 2010.
vastagbor (2009). http://vastagbor.blog.hu/2009/11/04/kave_vagy_vakcina_tokmindegy_ki_gyartja. Cited 14 July 2010.
Walton, D. (1997). Appeal to Expert Opinion. University Park, Pa.: Penn State University Press.

Sometimes it is hard to know where politics ends and metaphysics begins: when, that is, the stakes of a political dispute concern not simply a clash of competing ideas and values but a clash about what is real and what is not, what can be said to exist on its own and what owes its existence to another.
–       J.M. Bernstein, ”The Very Angry Tea Party” (The New York Times, June 13, 2010)

All modern philosophy hinges round the difficulty of describing the world in terms of subject and predicate, substance and quality, particular and universal. The result always does violence to that immediate experience which we express in our actions, our hopes, our sympathies, our purposes, and which we enjoy in spite of our lack of phrases for its verbal analysis. We find ourselves in a buzzing world, amid a democracy of fellow creatures; whereas. . .orthodox philosophy can only introduce us to solitary substances, each enjoying an illusory experience. . .
–       A.N. Whitehead, Process and Reality, p. 49[i]

We understand argumentation as a political practice, and propose that argumentation theory has neglected to attend to that “clash about what is real and what is not, what can be said to exist on its own and what owes its existence of another” that informs the diverse points of view – the “clash of competing ideas and values” – that is displayed in argumentative engagements. That neglect is due to a powerful presumption that has its roots in the primacy that Aristotle gave to substance, rather than relation, as well as the preeminence that Plato accorded to stable concepts (eternal Ideas) in contrast to changing things (the materiality of our “immediate experience”).[ii]

Questioning and even overturning this powerful presumption of “solitary substances,” which persists in rationalistic, constructive idealist, and empiricist traditions, is not an easy endeavor. The exigency for doing so is strengthened by arguments for the value of argumentation theory and informal logic, rather than formal deductive logic, for analyzing, understanding, and arguing about that “buzzing world” of our “immediate experience.” In this essay we propose that Alfred North Whitehead’s process-relational metaphysics offers an alternative to the “violence” that (as he proposes in our second epigraph) “modern” or “orthodox” philosophy does to “that immediate experience which we express in our actions, our hopes, our sympathies, [and] our purposes.” Thus, we would modify Jay Bernstein’s suggestion: we cannot know “where politics [or, argumentation] ends and metaphysics begins” because – in our “immediate experience” – there is no severance between those activities. However, an implicit traditional metaphysics that gives primacy to particular substances (subjects) and universal predicates (qualities) remains as the ground that nurtures the explicit “clash of competing ideas and values” that are the content of argumentation.

1. Two Proposals
Our first proposal, then, is that our epistemological endeavors would benefit by accepting the need to critically investigate our metaphysical presumptions. That’s because affective, cognitive,  cultural, and social assumptions about what is to be known exert an influence – perhaps, even determine – how we go about epistemic endeavors. Metaphysical inquiry, we would emphasize, is not an optional additional level in, aspect of, or tier within argumentation – because all theory and practice, including argumentation, presupposes some metaphysics. Nor is it a concern with how premises are generated from ideas or beliefs. Rather, this first proposal calls for reflection upon the elements and relations that are presumed as present in arguments (as products), by way of examining the presuppositions that are embedded in the process and procedure of argumentation.

In making this proposal, we focus on the first and second of Joseph Wenzel’s “three different ways of thinking about argumentation” (1990, p. 9), rather than the third way: We regard argumentation as rhetorical and dialectical, rather than as logical; which is to say that we focus upon argumentation as a process of communicative interaction and procedure for organizing what’s articulated in that interaction, rather than as a product that enables evaluation of what’s articulated in order to assess its strength or validity. We recognize that all three perspectives are valuable, yet propose that the first two are more appropriate for analysis of argumentation understood as a political practice concerning the “immediate experience which we express in our actions.” Not coincidentally, the rhetorical and dialectical perspectives emphasize the fluidity, relationality, and contingency of that “buzzing world, amid a democracy of fellow creatures” that characterize our reality, rather than focusing upon abstracted conceptual content that, within the third (logical) perspective, is articulated as its form.

We believe that making these assumptions about the nature of reality explicit and proposing alternative presuppositions enables a re-specified understanding of argumentation that focuses upon what actually happens in our “immediate experience.”. That understanding, in turn, enables us to envision rationales for making decisions that choose among the plethora of affective, cognitive, cultural, and social possibilities for action that compose that “buzzing world.” As in Bernstein’s analysis of the anger that motivates the Tea Party, we can move beyond obsession with the “clash of competing ideas and values” insofar as we acknowledge that we are divided about “what is real and what is not; what can be said to exist on its own and what owes its existence to another.”  This focus on “what actually happens” and on the nature of reality motivates our first proposal, and is developed further as the core of our second proposal.

Our second proposal is that respecifying argumentation theory on the basis of a process-relational metaphysics allows us to analyze the powerful presence, within argumentation, of that reality that is our “immediate experience” – despite the “illusory” overlay of solitary substances and mental representations, as formulated in verbal argumentation, from which argument analysis traditionally begins. Rather than understanding argumentation as disagreement between Cartesian subjects about diverse representational predicates, we can identify the diversity of ideas and values as intrinsic to the process by which arguers become who they are and how that coming-to-be continues in and through argumentive engagement. In other words: replacing a substantialist, individualist, and empiricist metaphysics with a process-relational metaphysics offers us a way of accounting for how particular ideas and values come to be a part of arguers’ process of coming-to-be, and how alternative ideas and values might be advocated more successfully than setting them out in opposition to those currently held.

In this essay, we introduce Whitehead’s process-relational metaphysics and briefly indicate the value of this alternative framework for clarifying, rather than “doing violence to,” immediate experience. Two conceptual shifts are needed at the start. First, although argumentation theory typically is considered to be an epistemological endeavor, we need to acknowledge that implicit – which is to say, unnoticed and unexamined – metaphysical presuppositions underlie all theory, including any epistemological theory. The conceptual shift that’s needed here is toward explicating these presuppositions and discerning their influence. Doing that takes us to the  second conceptual shift: We need to expand our theoretical resources for understanding the “clash about what is real and what is not” that, we believe, is operative within argumentation – usually, implicitly – and especially, when argumentation becomes obstructed by deep disagreement.[iii]

2. Whitehead’s Process-Relational Framework
We advocate this process-relational theory as alternative to traditional “modern” or “orthodox” philosophy, which (as he notes) relies upon a “subject and predicate, substance and quality, particular and universal” understanding of reality. This alternative enables us to reconsider modern philosophy’s characterization of humans as either passive recipients of sense-data or active imposers of form upon a sensory manifold. It also enables us to resist postmodern philosophy’s focus on the linguistic formulation of experience, which has reinforced argumentation theory’s proclivity for beginning analysis at the level of verbal, rather than experiential (affective and embodied), modes of being. In other words, theorizing argumentation within a process-relational metaphysical framework requires us to suspend acceptance of both empiricist and rationalistic presumptions. Once we understand this very different way of considering the environment, we can test its comparative efficacy by applying it to an example of argumentation about “immediate experience.” Thus, we begin by introducing terminology for some of the very basic claims of a process-relational framework.

Whitehead understands mind as an “actual occasion” rather than as a substance that requires only itself to exist – that is, which is independent of the material world, including mind’s physical embodiment. An actual occasion functions as a locus of response, and thus relation, to an environment that is not limited to present space-time. Rather, each occasion grasps, and draws from, past actuality as well as future possibility. Whitehead calls this grasping “prehension,” and cautions that unlike apprehension, which is a comparatively familiar mental activity, it is motivated by affective sensory attraction and repulsion, rather than cognition. Actual occasions continually form themselves as actual entities through retaining past prehensions while appropriating possibilities that are present to them as propositions.

Within this framework, propositions are not statements with a truth value. Rather, they are potential ways that occasions may come to be as actual entities. Selection of some propositions and deflection of others depends upon the interest and intensity of their anticipation, within the immediate experience of a prehending actual entity. Truth is still a useful category within this process-relational metaphysics, but it is not a matter of language that corresponds to reality (mentality to materiality) or coherence within an already accepted structure in the mind or in language. Rather, it pertains to correspondence between how an actual occasion (or group of occasions, which Whitehead calls a “nexus”) may be, and how it is. Although particular truth claims can be refuted, we cannot be certain that any particular claim is true. Those that resist efforts to refute them can be retained as, at least, closer to truth than refuted claims. Making an argument, then, is not a matter providing statements that correspond to how things are or should be in a pattern that results in having a valid, and even sound, argument. Rather, it is an activity of acknowledging the relative appeal of how things might come to be. Consideration and choice among possibilities is a response to the aesthetic and affective appeal with which they are present to prehension, rather than of calculative rationality. Selection or choice happens by relating to the more appealing alternative possibility, rather than by making a cognitive decision between opposing claims. It depends on an actual entity’s entertaining those possibilities as potentially providing a more fitting continuation with the past and future, rather than requiring a conceptual decision that’s constrained by already available ideas and values as they are asserted within an oppositional agenda.

3. A Case Study: The Tea Party
This conception of what happens in reasoning suggests a response to Jay Bernstein’s question in the editorial that provides our first epigraph. The context in which he finds that metaphysics and politics are difficult to separate is contemporary concern, within U.S. political argumentation, about the “Very Angry Tea Party.” The “seething anger” of the Tea Party, Bernstein argues, resists explanation through traditional logics of interest group pluralism. The Tea Party forwards no coherent policy proposals, nor does it protest in order to acquire political power. What matters about the Tea Party, and what no one has yet been able to explain, Bernstein argues, is the “exorbitant character” of its anger. Given the fury of its protests and how that fury “is already reshaping our political landscape,” he proposes that the important question is not what does the Tea Party want, but where does “such anger and such passionate attachment to wildly fantastic beliefs come from?”

Bernstein’s hypothesis is that the source of this anger is not merely political, but metaphysical. That is, the last several years of crisis and reform, disaster and response – particularly within the U.S. political economy – has shown that we are utterly dependent on government action, even as its limitations, corruptions, and incompetence have never been clearer. What has unraveled in these recent crisis-ridden years is the “belief that each individual is metaphysically self-sufficient, that one’s standing and being as a rational agent owes nothing to other individuals and institutions.” The autonomous individual has been revealed as an “artifact” manufactured by the “practices of modern life: the intimate family, the market economy, the liberal state.”

The poverty of the metaphysical commitments underwriting autonomous individualism has been exposed, and that creates an opening for Bernstein to propose an alternate metaphysical claim: Human subjectivity “only emerges through intersubjective relations.” Each of us is called into being by the other, wholly dependent on the other’s love for our freedom. Our independence is, therefore, “held in place and made possible by complementary structures of dependence.” Love, however, can go bad and when it does we realize that we are “absolutely dependent on someone for whom we ‘no longer count,’ we feel “vulnerable, needy, unanchored and without resource.” This vulnerability unleashes fury. We rage against our former love, proclaiming our independence, denying that we ever needed the other (whether personal or institutional) in the first place. This is the anger of the Tea Party. They are jilted lovers furious that they have been let down by their government, furious that they find themselves dependent and powerless. They feel all that comes with love’s betrayal: rage, disillusionment, sorrow, and confusion. Searching for the source of this betrayal becomes an obsession, expressed in terms of who has stolen their country and how they can get it back. Their anger leaves them epistemically vulnerable, ready to believe just about any conspiracy, any rumor, any fear-mongering appeal that can pinpoint the culprit.

Bernstein is careful not to imply that all political anger is metaphysically suspect. We ought to be angry at the “thoughtless greed of Wall Street bankers” and the “brutal carelessness of BP.” We have been betrayed. But there is a difference between moral indignation “raised by cruelty and injustice” and the “exorbitant and destructive” anger raised by resentment of the fact that we are inescapably interdependent. The former is an expression of concern that fosters moral community; the later seeks to destroy the institutions, such as town-hall meetings, which sustain community. Moral indignation leads to “creative, intelligent, non-violent” resistance; fear-induced rage towards the other leads to nihilistic terror. The Tea Party, thus far, has been a party of resentment. But if it traded its rage in regard to what has been taken from individuals (a sense of autonomy that does not correspond to the reality of the human condition) for indignation about how government has been corrupted so that it destroys real human needs, the seemingly intransigent opposition between left and right may be redirected from the diversity of ideas and values that attract and repel them, and toward common acknowledgment of the need for change in current political practices. Radicals of all stripes could be in “angry agreement” that democracy has indeed been hijacked by corporations, special interests, lobbyists, and self-serving, corrupt public servants. Their righteous indignation could be directed towards a common project of increasing public accountability and restoring self-government through increasing the opportunities for authentic public deliberation.

4.  The Nature of Argumentation: Two Insights
We find Bernstein’s diagnosis of “passionate attachment” persuasive and his call to “indignation” compelling (although we differ from some of the particulars of his argument). Putting his analysis into the process-relational metaphysical terminology we have introduced provides these insights into the nature of argumentation as a political practice:
(1) A process-relational account of argument is uniquely suited to understanding the dynamics of affective politics.
The effectiveness of the Tea Party’s fury in reshaping the US political landscape exemplifies a shift from a content-driven politics (ideas and values in opposition) to an affective politics (the lure of possibilities that attract or repel). The Tea Party’s significance, Bernstein makes clear, lies in the “exorbitant” character of its anger,” not in any concrete policy proposal or party platform it might forward. Affective politics are driven by image, tone, resonance, movement, and rhythm. Its governing terms are confidence, trust, support, and mood. It is a politics of the body, or more precisely of the becoming-body; a chosen coalescence of neurochemical reactions to environmental stimulations. Affective politics comes from a transfer of energy, of commitment among successive waves of actors, of how energy designs processes that serve as technologies of collaboration. Affective politics is a politics of relation in which the quality of life is increasingly defined in terms of modulating attachment, attunement, and attention.

Thus, Bernstein’s question – “where does” the Tea Party’s “anger and passionate attachment to wildly fantastic beliefs come from” – directs us toward an important contribution that a process-relational understanding of argumentation can make to the study of affective politics. Argumentation theory has had relatively little to say about the nature of affective attachment to particular claims and beliefs because it has operated from an overly cognitive account of the relationship between mind, body, and environment. This account treats them as distinct entities whose impact on the processes of reasoning and arguing is taken for granted and little understood. Moreover, argumentation has been treated as an exclusively cognitive and verbal activity that occurs in and through conscious reflection, despite growing evidence discrediting that view, as well as the increasing attention of many theoreticians to visual argumentation. A process-relational metaphysics, to the contrary, understands argumentation, in Erin Manning’s words, “as a complex passage from thought to feeling to concepts-in-prearticulation to events in the making” (2009, p. 5). A process-relational account understands thought not as a property of the mind, but as an activity of the minded body in dynamic response to, and thus in relation with, the diverse loci of allure and appeal that are continually emergent within its environment. In sum, one of the insights available to a process-relational analysis is that argumentation is very much more than is suggested by the final form it takes in language.

(2) Argumentation is inherently collaborative, not oppositional. Opposition is an artifact of substantialist metaphysics and the governmentality of liberalism that accompanies it.
A process-relational view provides us with a means to theorize our environment as a world that is made of events in dynamic relationship. Even seemingly solid and permanent objects are events, or better a series of events in the making, whose composition changes moment by moment. The continuity implied by the existence of enduring objects needs to be actively produced at every instant as a new event. The same is true of us and our perception of those objects. The persistent flow of perception and conception constitutes us anew as subjects. Each instant of every encounter is a new event and each of the selves to which it happens is also a fresh event. This does not entail that objects are created by our perception; it does entail that their shape and importance is formed in perceptual events of interaction with them. Objects-as-events are possibilities for choice, as they are present for perceptual or conceptual engaging of them by actual entities. This account reverses the Kantian assumption that “the world emerges from the subject.” A process-relational metaphysics reveals, instead, that “the subject emerges from the world.” We are born in the very course of our encounter with the world and are precipitated out this encounter, “like salt precipitated out of a solution” (Shaviro 2009, p. 21). For Whitehead there is no ontological difference between thoughts and things, between animate beings and so-called inanimate objects. The same goes for arguments and arguers. Each is grasped from the “buzzing world” of immediate experience, existing in a “democracy of fellow creatures.”

We have suggested that this grasping, which Whitehead calls prehension, is itself a description of argumentation. That is, argumentation is not simply a distinct activity that we can describe in process-relational terms. Rather, argumentation is at the heart of the process of becoming. Events are constituted through the creative interplay of past occasions of experience and the potentiality of anticipated experience. Their expression as propositions does not provide verbal assertion of goodness or truth, but does make choices available for the ongoing integration that constitutes actual entities. In other words, propositions are neither actual or fictive; they are ‘the tales that can be told about particular actualities’ from a given perspective, and that enter into the formation – the process that Whitehead calls concrescence – of that very perspective. As such, propositions are possible routes of actualization, vectors of nondeterministic change. (Shaviro 2009, p. 2, quoting Whitehead 1929/1978, p. 256).

Argumentation is the process of assembling and coalescing propositions. It is an essentially creative, collaborative activity, rather than a uniquely “human” activity, insofar as it is understood as an interaction with the environment (both past and anticipated) in which “what is real and what is not” depends not on autonomous individuals or their contexts and not on causation or cognition, but on affective, cognitive, cultural, and social response to the allure of what may be.

One conclusion that we draw from this account of argumentation is that the taken-for-granted understanding of argumentation as inherently oppositional is itself a proposition; a proposal – perhaps particularly appealing and attractive within our cultural and political environment – of how choice among possibilities happens. For instance, argumentation may be theorized as a critical discussion aimed at resolving a difference of opinion in which a protagonist defends a certain standpoint against the challenge of an antagonist who raises doubts about and objections to the acceptability of that standpoint (van Eemeren, Grootendorst, & Snoeck Henkemans 2002, p. 25). To characterize this theory as an interesting and perhaps appealing tale told of how reality happens does not mean that it is false. It does mean that it is a verbal abstraction, necessarily static (given the nature of both abstraction and verbality). As such – as an abstraction from immediate experience – we commit what Whitehead calls “the fallacy of misplaced concreteness” if we take it as real, or as a description or representation of the real. The fallacy to be avoided here is metaphysical, not logical: “misplaced concreteness” is to be avoided because it “does violence to. . . immediate experience” (Whitehead 1929/1978, p. 49).

Any theory portraying argumentation as oppositional in nature is, we contend (following Bernstein) a metaphysical “artifact manufactured by the “practices of modern life: the intimate family, the market economy, the liberal state.” It is no accident that the logic of opposition works so well to explain the machinations of these institutions. Opposition is the lifeblood of liberal governmentality, which requires that individuals be defined by their irreconcilable differences, standing ready to engage in total war if they are without the mediation of the state. Just as essential to the logic of opposition is the presumption that the state presents an ever present threat to the sovereignty of its citizens. In this respect the Tea Party’s rage is not an aberration of liberalism; it is a pathological expression of contemporary liberalism’s nature.

The test of a proposition is not whether it is true, coherent, or plausible. A proposition is “a lure for feeling”; a means to “pave the way along which the world advances into novelty” (Whitehead, 1929/1978, p. 187). Propositions should be assessed in terms of their aesthetic appeal, creativity, and potential for inventing novel platforms for collaboration. Argumentation theory, presently conceived within a substantialist metaphysics, can inform criticism of the Tea Party’s fury and demonstrate the irrationality of their “wildly fantastic beliefs,” but it cannot explain the nature of their “passionate attachments” nor propose a means to transcend the fierce logics of neoliberal governmentality that pervert them. We advocate adoption of Whitehead’s metaphysical theory as the framework for understanding argumentation as a relational process, rather than as a means for generating oppositional arguments, as the way of doing just that.

NOTES

[i] Unless identified otherwise, the quoted phrases from Whitehead in this essay are taken from this epigraph.
[ii] Arguably, this presumption continued to guide philosophical thinking about reasoning from classical to modern times, when it was expressed in René Descartes’ conception of humans as mental substances – solitary minds – whose thinking focuses on ideas (mental events) that describe or represent their material environment. It is also expressed in Immanuel Kant’s conception of humans as dictating the form of physical substance, which was taken to be independent of, and subservient to, mind. And it was expressed in David Hume’s and Thomas Hobbes’ conceptions of humans as passive recipients of sense data, and thus, of mentality as dependent upon materiality.
[iii] Concern with the challenges of “deep disagreement” to argumentation traces back to the germinal article by Robert Fogelin (1985). The editors of a special issue of Informal Logic in 2005 (which reprints that article) note that they hope to “spark renewed reflection on these sorts of fundamental questions” (Turner and Campolo, 2005, p. 2). See David Zarefsky’s (2010) paper for a current contribution to that reflection.  See also the discussion by Frans van Eemeren, Rob Grootendorst, Sally Jackson, & Scott Jacobs (1993, pp. 171-172) of the empirical challenges of deep disagreements to pragma-dialectics. We believe that these “fundamental questions” call for reflection on the metaphysical presuppositions that participants bring to argumentive engageme

REFERENCES
Bernstein, J.M. (2010, June 13). The very angry Tea Party. The New York Times. Retrieved from: http://opinionator.blogs.nytimes.com/2010/06/13/the-very-angry-tea-party/
Bernstein, J.M. (2010, June 22). The usefulness of anger: A response. The New York Times. Retrieved from: http://opinionator.blogs.nytimes.com/2010/06/22/the-usefulness-of-anger-a-response/
Eemeren, F.H. van, Grootendorst, R., Jackson, S.& Jacobs, S. (1993). Reconstructing Argumentative Discourse. Tuscaloosa: The University of Alabama Press.
Eemeren, F.H. van, Grootendorst, R. & Snoeck Henkemans, F. (2002). Argumentation: Analysis, Evaluation, Presentation. Mahwah, NJ: Lawrence Erlbaum Publishers.
Fogelin, R.J. (1985). The logic of deep disagreements. Informal Logic 7, 1-8.
Manning, E. (2009). Relationscapes: Art, Movement, Philosophy. Boston: MIT Press.
Shaviro, S. (2010). Without Criteria: Kant, Whitehead, Deleuze, and Aesthetics. Boston: MIT Press.
Turner, D. & Campolo, C. (Eds.). (2005). Introduction: Deep disagreement reexamined. [Special Issue]. Informal Logic 25(1), 1-2.
Wenzel, J.W. (1990). Three Perspectives on Argument: Rhetoric, Dialectic, Logic. In Perspectives on Argumentation: Essays in Honor of Wayne Brockriede. Prospect Heights, IL: Waveland Press, Inc.
Whitehead, A. N. (1929/1978). Process and Reality. New York: The Free Press.
Zarefsky, D. (2010, June). The appeal for transcendence: A possible response to deep disagreement. Paper presented at the Seventh International Conference on Argumentation, Amsterdam.

Practical thinking is a tricky business. Its aim will never be fulfilled unless influence on practical attitudes is gained. These attitudes, though, are no neat propositional structures, as is sometimes suggested. Whether or not a living human being is willing to act in a certain way is determined by dispositions that are non-codified, non-transparent, habitual, embodied and emotional. To reflect upon such attitudes is roughly as complex as reflect­ing upon the agent’s moral identity.

This poses some problems for moral argumentation. In practical matters, justifying practical beliefs as “true” is not enough. The motiva­tional dimension cannot be ignored. This is the original field of classical rhetoric. Rhetorical methods are not designed to examine theoretical truths but for the purpose of practical decision making. This is why rhetoric and ethics have always been closely related. The aristotelian doctrine of lógos, páthos and êthos reminds us of the fact that speech is persuasive not due to its rationality only but also due to the “moral character” of the speaker and the emotional dispositions of the audience. The adoption of a practical attitude cannot be reached by deduction alone. It takes more to persuade and motivate a human being to act in a certain way.

This being the case, one should think that any conception of moral argu­mentation reduced to rational argu­mentation in a narrow sense will be incomplete. However, such conceptions of moral argumentation seem to be wide spread. One of the basic assumptions of cognitivist – and roughly, Kantian – ethics is that moral argumentation has to be built on reason alone, on “rational discourse”, as representatives of discourse ethics like Haber­mas would prefer to say. Any reference to emo­tions, then, has to be regarded as “merely rhetorical”.
In this contribution, I will ask for possibilities to reconcile the logical and the rhetorical dimension of moral argumentation. In particular, I will discuss how expressive speech can have a place in rational moral argu­mentation. Here, the important question will be how such speech can be part of moral argumentation and more than just emotional talk. I will first sketch what function logical reasoning is sup­posed to have in moral argumentation and why philosophy is often focused on this dimension (1). In a second step I will ask in how far we usually take rhetoric – in this case, expressive speech – to be relevant for morality (2). Third, I will try to outline a con­ception of moral argumentation that in­cludes logic and articulations of perspectives (3).

1. Cognitivism
The first question will be: Why is moral argumentation in philo­sophy so often regarded as a kind of rational argumentation more or less in the style of truth based reasoning?
The answer is, of course, not that philo­sophers assume that every-day moral communication is in actual fact “rational” in this narrow sense. The interpretation of moral argumentation as a kind of rational argu­mentation has normative sense. The idea – that we may call the cognitivist intuition – is this: By bringing out the logic of every-day moral communication one can set free its normative content. Logical reconstruction shows us what general moral principles are applied and what norms must be con­sidered as binding in the context of the normative systems sup­ported. So the ratio­nality of moral argu­mentation is not so much discovered but rather elaborated.

Justifying a normative claim, then, can only mean: showing that this claim satisfies the basic normative principles or showing how it fits into the presupposed system of norms. Ethical reflection turns into an attempt to ascribe some truth-value-like quality to normative claims. This, to be sure, does not mean that the peculiarities of moral debates are not accounted for. Of course claiming certain facts differs from claiming certain norms to be valid (and since Hume this difference is normally taken very strict). So normative logic is not epistemic but deontic. And the logical principles applied differ as well: e.g., principles of universability play a central role since the consistency of a normative system seems to depend on it. But despite these differences normative claims are treated as claims that transport a content that can be compared to the factual content of a descriptive statement. Sometimes this quality is called “rightness” or the “cognitive content” of moral claims: Just like descriptive claims are true if they correspond to certain facts, so normative claims can be true or “right”, if they express certain valid norms, i.e., a normative content that every “rational” person will accept (Habermas 1999).

One might say that this approach ignores the pluralism of the modern globalized world. But quite the contrary cognitivists argue that cognitivism is especially attractive in face of pluralism. Given a multitude of values and “ideas of the good life”, it seems to be the task of philosophical ethics to find a moral fundament independent of particular standpoints. Under the conditions of pluralism, many ethical perspectives have to be reconciled; and this cannot be established by falling back on particular ethical perspectives. What is needed is an “overlapping consensus” as John Rawls calls it (Rawls 1971). So the cognitivist approach that seems to be reductionist at first sight turns out to be the only option left in face of pluralism.

2. Ethical perspectives
Of course, all this does not mean that particular ethical perspectives just disappear. It is obvious that moral argumentation includes articulations of such perspectives, e. g., expressive speech. But the question has to be: How can such expressive speech be legitimately introduced into the kind of moral argumentation that philosophy tries to establish? Modern ethics seems to call for a cognitivist approach and this in turn seems to call for some sort of formal reasoning. Conse­quently, it might appear that moral argu­mentation has to be interpreted from a general stand­point. The idea might be that rational argu­mentation has come to an end as soon as, e. g., expressive – and “emotional” – speech comes into play. So the task is to show how such “perspectival” speech can have argumentative function.

To pave the way for an answer I will ask on what occasions we have no problems to accept the relevance of articulations of perspectives. Where do we usually locate such speech in moral contexts? I think the above-mentioned “ideas of the good” give us a hint. What I have in mind is this: Such speech has its natural place where human beings are initiated into a certain ethical practice. In order to communicate an “idea of the good” or a particular ethical perspective we have to use different means than logical arguments. – Let me explain.

It is a wide-spread neoaristotelian move in contemporary moral philosophy to focus on practice and character rather than on norms and rules. From this virtue ethical point of view the morality of a person is not constituted by the normative statements she rationally accepts or by the rules she is willing to obey but by the practice she is engaged in. A person’s moral identity is constituted by habits or dispositions instantiated in his or her action, and not just by “supporting claims”.

Given this perspective, a moral judgment can no longer be a matter of cognition alone; it must be a matter of practical wisdom and perception – the kind of competence that Aristotle has called phrónêsis. Acquiring a certain ethical practice goes along with acquiring a certain way of seeing. Here, moral judgment is highly con­textual. On particular occasions you do not have neutral perceptions of “what is the case” in the first step and moral reflections in the second step (which then can be based on “pure normative reasoning” of some sort). In fact, the two dimensions are entangled: moral judging, here, means per­ceiving a situation in a certain way. From this point of view, morality is a capacity to deal with multiple particular contexts in the right way; and a person satisfying this criterion has virtue.

But it is clear that the notion of “rightness” here is restricted to particular practices. The criteria of what counts as right are the criteria of particular communities and their “life forms” (as Witt­gen­stein calls it). This lack of universal validity is the central difference between the morally right and the ethically good: At first sight, the character-based approach apparently does not answer to the normative question of ethics at all. It rather tells us how the moral life of human beings really looks like.

Now, I do not want to start a discussion on virtue ethics here, but what is crucial to my argument is this: There is obviously no way of arguing in favor of an ethical practice or way of “moral seeing” by logical reasoning. Instead, the value of an ethical practice – as a practice having its purpose in itself – would have to be shown. John McDowell has elaborated this thought by referring to Witt­genstein’s reflections on rule-following. His example is, quite naturally, the case of moral education: When we are initiated into a way of moral perception we do not learn to act according to rules. McDowell writes: “In moral upbringing what one learns is […] to see situations in a special light, as constituting reasons for acting (McDowell 1978, p. 21). The decisive aspect of such a process of teaching such a way of moral seeing is expressed in the formula: “See it like this!” It is not a matter of saying what is “right” but of showing what is the point of it. There is no question if certain claims are justified. The aim is to make someone see what it means to consider something as valuable or “good” – what it is to take a certain ethical perspective. On such occasions one will apply “helpful juxtapositions of cases, descriptions with carefully chosen terms and carefully placed emphasis, and the like” (McDowell 1978, p. 21). In a process of this kind there is no guarantee that the aim is reached. “That, together with the importance of rhetorical skills to their successful deployment, sets them apart from the sorts of thing we typically regard as paradigms of argument.” (McDowell 1978, p. 22). Perhaps one might even say: To explain a particular ethical perspective it takes everything but argument.

3. Argumentation and articulation
Let us come back to the central question: How can articulations of ethical perspectives be part of “rational” moral argumentation as pre­ferred by philosophers? What we need is a conception of moral argumen­tation that accounts for both kinds of speech. On the one hand: logical persuasion, i.e., justifying claims as valid normative claims (which implies arguing from a moral standpoint that has overcome individual perspectives). On the other hand: articulations of ethical perspectives which implies expressing ways of “moral seeing” or particular “ideas of the good” (as familiar from moral education). How can these two elements at once be part of moral argumentation? I will try to at least outline an answer.

The reason why particular ethical perspectives cannot be ignored for moral argumentation is, of course, very simple: Moral conflicts are conflicts of ethical perspectives. The fact that normative ethics, especially the Kantian tradition, recommends a general moral standpoint to solve moral conflicts implies the thought that in case of moral conflict such a stand­point is lost. In other words, usually the moral conflict will rest on the very fact that there is no common ethical basis but opposing “ideas of the good”. Moral argumentation, then, starts with the collision of particular ethical perspectives and its aim is to reconcile these perspectives.

Still, from a strict cognitivist standpoint the “perspectival” aspect of moral claims has no cognitive content. According to the cognitivist a claim has cognitive content insofar as it overcomes perspectivity and expresses a possible general law. It is “rational” only if it fulfills the criteria of a formal procedure. Now one might say that this account is already given from a general moral point of view; the cognitivist takes, so to speak, the stand­point of the solution whereas the stand­points of the persons involved in the moral conflict are excluded. From this standpoint, however, the particular perspective and the universal claim are never separated. The point is: For the speaker, articulations of his or her very own ethical perspective will simply be utterances of “what is right”. The addressee, in turn, will not interpret the other one’s statements as showing him the way to a new idea of the good (i.e., he does not take the perspective of a disciple in moral education). In case of moral conflict, he will understand his opponent’s utterances as articulations of a particular standpoint that is unacceptable in some respect. In short, participants in a moral argument will take each other’s utterances either as articulations of universal rules (i.e., as right) or as articulations of a mere perspective (i.e., as wrong). More precisely, they will take their own utterances as articulations of universal rules (as right) and their opponents’ utterances as articulations of mere perspectives (as wrong).

If this is right, what can we answer to the question how expressive speech can have a place in moral argumentation (interpreted as reduced to logical reasoning for the sake of argument)? At least, we have a first clue: The normative claims to be justified in the course of argument and the articulations of ethical perspectives could be one and the same. From the standpoint of the participants, nor­mative claims do not necessarily have to be performed as normative claims in the first place. The speakers might just intend to express what their perspectives are like and then get involved in a normative argument. It is tempting to think that an actor who makes a normative claim must have known about the norm all from the beginning as if he had a stock of “rules” that he “follows” in his life. But the virtue ethics discussion reminds us that a practical attitude is not codified by nature. Expressing such an attitude – expressing an ethical perspective – is a creative act. The rules of action (the “maxims”, in Kant’s terminology) are not given as “ready made norms”. To put them to test of universability they have to be formulated.

Charles Taylor’s conception of articulation, inspired by Herder and Humboldt (Taylor 1980), can help to clarify this point. Taylor reminds us that expressing attitudes, is not something like describing “inner facts”. It does not mean to speak about things very hard to describe. Neither does it mean to “make explicit” rules. Just like ethical practice is not codified by nature, attitudes do not appear in sentential form. In this sense, attitudes are non-propositional; they have to be articulated: “articulations are attempts to formulate what is initially inchoate, or confused, or badly formulated” (Taylor 1977, p. 36). In this process, there is no constant object that is represented. When persons express their practical attitudes they rather fix what they want to accept as right. Taylor puts it this way: “To give a certain articulation is to shape our sense of what we desire or what we hold important in a certain way” (Taylor 1977, p. 36). To accept such an articulation, then, does not mean to accept a “rendering” as correct. It means accepting a certain interpretation as an adequate self ­interpretation (Taylor 1977, p. 37ff.).

What is the general picture that emerges? What in moral philosophy is sometimes called “cognitive content”, i.e., the content of the utterance as far as it can be generalized is only one side of the matter. Indeed, every utterance that appears as a normative claim in moral argumentation may at the same time be an expression of a particular perspective. We might call this the “ethical” or “expressive content” of the statement. But indeed, the term “content” might be misleading already since the articulation of ethical perspectives is not representational but productive speech, i.e., part of the formation of practical attitudes. Moral argu­mentation, in general, seems to include both dimensions: the production and the critical evaluation of norms. Attitudes are made public by expressive speech acts for the purpose of formation and further development. In other words, moral argumentation might be regarded as a process of intersubjective attitude formation by means of critical evaluation.

If this is right, then expressive speech and normative reasoning might in many cases be irreducibly entangled. In fact, the distinction between making normative claims and expressing one’s very own perspective might not even be a factual distinction. It depends on the perspective that is taken if an utterance is interpreted as a normative claim or as an expressive act. What is interpreted as a norm­ative claim from the perspective of a third person might be the articulation of a first person’s perspective who expresses his or her ethical standpoint. In this case, excluding expressive speech from moral discourse would mean to exclude this perspective. But this, in turn, would obviously amount to eliminate the ethical subject matter itself.

REFERENCES
Habermas, J. (1999). Richtigkeit versus Wahrheit. Zum Sinn der Soll­geltung moralischer Urteile und Normen. In J. Habermas, Wahrheit und Recht­fertigung. Philo­­so­phische Aufsätze (pp. 271-318), Frankfurt/Main: Suhr­kamp.
McDowell, J. (1978). Are Moral Requirements Hypothetical Imperatives? Proceedings of the Aristotelian Society, Supplementary Volume 52, 13-29.
Rawls, J. (1971). A Theory of Justice. Cambridge: Cambridge University Press.
Taylor, Ch. (1977).What is Human Agency? In Ch. Taylor Human Agency and Language. Philosophical Papers I (pp. 15-44), Cambridge: Cambridge University Press 1985.
Taylor, Ch. (1980). Theories of Meaning. Proceedings of the British Academy 66, 283-327.

1. Introduction
As we can see, authors working in the field of argumentation can develop quite different theories and models, especially in a pedagogical context. Let us assume that it would probably be useful to review and reflect on these theoretical achievements, not only for historical reasons but also to reflect on the limits and resources of previous models. This is what I would like to attempt here, with two models developed in two books that I wish to consider and compare. It has been a few years since I was stricken by the differences between these two Argumentation handbooks, books that of course I have used in classes, one from J. Michael Sproule, and the other by francophone authors from Québec, Nicole Toussaint and Gaston Ducasse, helped by pr. G. A. Legault. The first book is Argumentation. Language and its influence (1980), the other one is Apprendre à argumenter. Initiation à l’argumentation rationnelle écrite, théorie et exercices (1996).[i]

When I am mentioning « models » here, discussing specifically the S model and the TD model, I am describing and discussing the analytical tool that is furnished in these books by their respective authors in the aim of helping students to discern the main characteristics of a given argument. Armed with these analytical tools, students are supposed to be then able to analyze arguments. These books are both destined to an undergraduate public, but they can also be used at a professional graduate level. They both can be especially useful as first books in argumentation studies. If the theoretical level and the written explanations of Sproule’s book seem more theoretically advanced than those from Toussaint-Ducasse, the latter has more guidelines, schemas and details to help a beginner to grasp the argumentation domain; in that sense it can be said to be more “user friendly” then the other, more complex one.

These models have obviously been developed in a teaching context, but they are different in their orientation. Briefly stated, we can give the following precisions on the models. The S model distinguishes, among arguments, between descriptive, interpretive and evaluative arguments, meaning by interpretives, statements raising issues of definition, whereas of course descriptives are concerned with facts and states of affairs, and evaluatives are considering situations with the prism of some values used as more or less precise criteria. The TD model distinguishes, in terms of kinds or arguments available, between assertives, evaluatives and directives, meaning by this last element prescriptions, whereas the other categories overlap with those of Sproule. In each case we have three important categories that come out in the forefront of their respective model, but as we can see they disagree in one third of their respective categories; the Interpretive category is not to be found in TD and the Directive category is not to be found as such in the S model.

Each model represents a certain interest in its specificity. Probably because they differ, there still is a kind of compatibility of that plurality of tools in their capacity to analyse different arguments. Once we start using these kinds of tools, it is difficult to discard one of them as irrelevant, because they obviously have something complementary, as is showed by using them to analyze arguments.

There are basically two ways to look at this situation. The first strategy would be to try and combine them in a synthetic model. The second one would be to refer them to their interactive context of use, their pragmatic setting and respective teaching context. Developing the first briefly will lead us to the second strategy as being the more interesting one.
1 –  We could surmount this divergence by simply combining the different elements present, and forge a four-term model that keeps what they have in common and what is specific to each. We would then have descriptives, interpretives, evaluatives and directives (but no commissives – which would not be surprising since these authors do not interrogate the pragmatic dimension of argumentation (Van Eemeren & Grootendorst 2004, p. 62 f.).
2 –  But if we start to enlarge the model, we might as well add other dimensions too. It is probably more interesting to interrogate the specificity of each model and their raison d’être. To fuse the two models in one would be to sacrifice a certain level of simplicity that was probably a goal. These models obviously have been constructed to give a simplified and usable tool to students.

Focussing then on the first option, we will explore more systematically these models by looking more closely at some representative examples of their specific content, without pretending to furnish a complete and quantitative analysis of their respective work. Used and sometimes new copies of these books can easily be found at the time of this printing.

2. Briefly situating the authors and their respective interest
The present author has been teaching argumentation for a number of years (since 1997) inside an applied ethics perspective, in a graduate setting destined to professionals. This permits me to introduce an element that is important in both Sproule and Toussaint-Ducasse, and that is probably the reason of my previous interest in them: their strong integration and taking into account of what I generally call the ethico-moral dimension of human life and communication. These aspects manifest themselves differently in each handbook. In Sproule, ethical criteria are very important to judge the arguments, alongside the effects standard, the truth and rhetorical validity standard (Sproule 1980, p. 75-92). In Toussaint-Ducasse, the ethical interest manifests itself both by their choices of topics of discussion and by a stance given on the evaluative-directive pair among a total of three main categories, the third being the descriptive (Toussaint-Ducasse 1996, p. 32-89).

On another level of consideration, the careers of the different authors are not of the same amplitude. For those who would not know, Sproule was a speech communication professor for many years, and was named Emeritus professor of Speech communication in San Jose State University. He also published a number of articles and books (among which Sproule, 1992 and 1996), touching especially argumentation, rhetoric and propaganda issues in the public sphere; he was a dean of the College of Arts of Saint Louis University (starting in 2004) and is a recent past president (2007) of the very important NCA association in the United States, a country into which he certainly attained national and international status. With an excellent level of complexity and precision, Spoule’s book was obviously meant for students at the undergraduate level, maybe freshmen or the equivalent. Nicole Toussaint and Gaston Ducasse have been for many years college teachers preparing for the undergraduate level, but they have the merit to have been among the first to give some handbook of argumentative skills to francophone Québec students, and as such they had a good diffusion into a quite small population over all. Noteworthy is also the fact that their book was prepared with the help of an important ethics professor in French speaking Canada, Georges A.-Legault, well known for is applied ethics perspective oriented towards philosophical pragmatism and decision-making issues. This is probably the first time TD’s work is discussed at an international level. This having been said, that does not preclude the interest of looking at both these models, I hope to show why in the following.

S seems to be a tool constructed mostly for analysing documents, whereas TD is a tool servicing preferably a purpose of developing rational thought and writing skills, by providing structures of possible developments. But as things are standing, they both can also be used in the other way, since analysing and producing arguments often come together.

3. The Sproule model
To introduce the model, here we have to start with the general notions used. For Sproule, there is the basic and the extended argument (referring to Brandt, 1970). The basic argument is “the relationship of two terms via a name-relation pattern” (Sproule, 1980, p. 4). It is the simple declarative sentence by which two concepts or names are connected. For instance, the sentence “Smoking is harmful to your health” or “Dr Shintani is a good teacher” are basic arguments. This certainly can be reported back to basic attribution, as in Aristotle’s Peri Hermeneias (Aristotle 2004). Sproule proceeds then to define assumptions, elements seen as unstated and supporting visible arguments. The extended definition of the argument will then be “two or more basic arguments connected in such a way that one of them is a claim to be proved and the other (s) is (are) date offered in support of the claim” (Sproule 1980, p. 8). Then an argument can be said to have three composing elements: “the data, the reasoning process, and the conclusion”. Syllogism, enthymeme and the Toulmin model are briefly presented. For him, four different issues emerge in argumentation: issues of fact, of definition, of value…and of policy. For instance, if there is a conflict in faculty-administration relation, supposing that we have evaluated the situation to be bad, “the general policy issue becomes one of what should be done to dampen conflict and encourage cooperativeness…”(Sproule 1980, p. 19). We should already note that he will develop specific categories in his model only for the first three kinds of issues.

For his definition of the nature of meaning, he seems close to Charles S. Peirce: it is a triadic relationship between a referent, an interpreter and a symbol, but there is a second interpreter, the other person (Sproule 1980, p. 33). One useful distinction he gives is the one between positive terms and dialectical terms, taking back R. M. Weaver’s famous distinction. The first raise issues of fact; the others have what he calls nebulous referents, like justice or independence (Ibid. p. 34), and they can receive their meaning only in a dialectical way, by the interplay of questions and answers. Dialectical terms might be a necessary level of knowledge, but they carry important emotional overtones, and arguers tend to not define them satisfactorily (Sproule 1980, p. 36). Also noteworthy is the many functions of language: to report, to persuade, and there is an attitude-revealing function, a Self-revelation function, a relationship function with reference to Palo Alto (Watzlawick, Beavin and Jackson  1967).

We mentioned before that the S model distinguishes between descriptions, evaluations and interpretations, but we need to go into further detail. What is called a description draws first-order issues of fact, an evaluation draws first-order questions of value; interpretations draw first-order issues of definition (Sproule 1980, p. 69). In the first type of statements, we have appeal to facts, data and statistics, the problem we have is to ascertain if the facts alleged are or were the case. In the second, facts are regrouped and given meaning, they are united in an appropriate category (Sproule 1980, p. 142). Today, we would probably talk of framing issues here (Schön & Rein 1994). Statements that are evaluative for Sproule, using Rokeach’s well known perspective, are based on values defined as “a person’s notion of what is to be preferred” (Sproule 1980, p. 184).

Some of his material will help to better understand his perspective. In one example, we will see how the distinction between descriptions and definitions functions according to him. Sproule quotes an article from the New York Times, May 5, 1977, about the impeachment of Nixon. Without repeating the newspaper’s quote, I reproduce Sproule’s commentary to render visible his treatment.

(1)  “The initial questions raised by the data offered in this news article are ones of fact: Was Mr. Nixon cited as an unindicted co-conspirator? […] Did the House Judiciary Committee actually made the charge that Mr. Nixon participated in conspiracy to obstruct justice? Only when these factual issues are resolved can the reader proceed to the definitional question stated in the claim: Did Mr. Nixon commit an illegal act? The key observation to be made here is that while not everyone will accept the interpretive claim that “Mr. Nixon committed an illegal act”, they can be brought to agree that the House committee did allege his guilt.” (Sproule 1980, p. 71).

In this example, issues of fact as deployed in the legal sphere obtain meaning by being reconstructed as steps towards establishing the possible validity of an interpretive. While treating examples like this one, Sproule does not work most of the times by constructing and sequencing different propositions. As we can see here, starting with a substantial quote of a newspaper, he just reformulates the questions that can be raised. It is also interesting to note that this example, as many others in the book, is thoroughly legal and political in its nature, and gives voice to one very important type of recourse in any court of law, we could identify it as staying close to the facts while letting value elements play their part. Sproule situates his work inside what is called forensic debate (Sproule, 1980, p. 364). Other examples around what were immediately contemporary events in 1980, the Nixon impeachment, the Vietnam War and similar topic, abound in the book that refers copiously to Newsweek, The New York Times, Times etc.

For Sproule, a particular argument might raise first-order issues of fact AND subsidiary issues of value (noted 2). The same argument can also raise subsidiary issues of definition (noted 3). Every argument has potentially these three dimensions, present with differing importance. Sproule asserts that a specific prevalence would be present, meaning we will be able to distinguish what is of first order on this and that case. He admits that a combination of these issues is almost always present in complex argument. In practical use though, in some instances it is not easy to decide which aspect comes first, i.e. if this or that argument raises first-issue order of definition or of evaluation, for instance. Difficulties of the same kind might arise between arguments raising first-order issues of fact versus of evaluation. The tool can function nonetheless in general if we try and weigh carefully what is the most important use of the argument in the context.

Sproule does not discuss “framing” issues as such in this book, he does not make a technical use of this word that many authors report to E. Goffman (but see also Dewey 1925). Nonetheless, his use of the interpretive category understood as raising first-order issues of definition, is certainly a way to put some element in perspective and take into account something similar to the theme of framing inside his argumentation theory. This can be seen when he notes the fact that interpretations and evaluation sometimes overlap. For example, in a given election, descriptions give us the percentage of votes obtained by this or that candidate, whereas interpretations tell us “which candidate “won” the primary” or “which candidate did better/worse than expected”; the totals of the vote are then put into perspective by relating them to other opinions or facts (Sproule 1980, p. 144). Interpretives as raising definition issues certainly can be seen as framing devices, since they permit the grouping of facts under a category.

This role of the interpretive can be seen in an example about the Vietnam war (1965). Here Sproule refers to an expert on Vietnam history, Bernard Fall, according to which there were two different ways to see the forced moving of a million Vietnamese rural dwellers. In one narration of the event, by leaving the North of the country the refugees fled Viet Cong terrorism, but according to another interpretive, they were driven out by American bombing, which according to the author gives at the same time an argument against the war (Sproule 1980, p 145). Here either the communists are responsible for the fleeing refugees, or it is the Americans that are to blame. Without having to take a side, the author simply shows how each interpretive has different implications.  It is by the repetition of the examples that some position of the author (let us identify this as “liberal”, whatever that signifies) can be inferred, not because he would be dishonest in the treatment of the specific arguments.

Another example is not political: an anonymous writer (signing “Shy one buck”) writes to a newspaper column, “Dear Abby”. This person was in a grocery store, saw a woman arriving at the counter, having to pay, and then frantically looking into her purse, to declare out loud she was a dollar short. The writer to Abby’s column felt sorry and offered a dollar to help her. The woman expressed many thanks and insisted in writing the name and address of the giver, she also promised to send the dollar back to him by mail. Three weeks passed, there was nothing in the mail, so our guy writes to Abby and says: “…and I just didn’t peg her as the kind who would beat me out of a dollar” (Sproule 1980, p. 145). To better understand this nowadays, we would have to talk about a ten dollar bill. In any case, the “Dear Abby” person had then no difficulty in offering as an answer to the plaintiff three different interpretations of the same fact, one being the following: “She may have lost the paper with your name and address on it”. Using the same facts, Abby supplies different interpretations, placing the woman in alternative categories.

Interpretations also occur with comparisons, and with the use of metaphors; he gives the example of a strike being on one side compared to a hijacking or kidnapping, and on the other side to “a revolution for freedom of the small people against “the captains of industry” (Sproule 1980, p. 147). Comparisons act to construct reality, they are highly argumentative and are seen as an important strategy in defining issues. Sproule will also develop on analogy, argument from precedent, minimization-maximization as comparative tools that are also used under the interpretive mode. We might be interested to find one use of that last argumentative structure in an example about offshore drilling (no later than February 24, 1975, in Newsweek, p. 68), maximizing the economic benefits and minimizing the ecological aspects (Sproule 1980, p. 151).[ii] He notes three subtypes of this argumentative figure: playing on frequency, size and degree. When he comes to discuss causality (in the same chapter on interpretation), he makes a long detour by Mill’s canons of induction to go back again to political issues: different causes can be put forward to explain a given phenomena, and the way we assign cause plays an obvious role in defining the situation. He also treats arguments of sign, like the “tip of the iceberg”, arguments that predict future consequences, like the “domino theory” or the fear appeal, with reduction ad absurdum, humor, sarcasm, the argument of conspiracy, even the dilemma (with one horn of a situation being presented as less lethal than the other, for instance having to choose between freedom and starvation) and the antithesis. These argumentative figures (that could certainly by reconstructed as schemes) are all grouped under the interpretive category.

One of Sproule’s forces is the evaluation criteria for argumentation he puts forward. There is the effects standard, the truth standard, the Ethics standard, there is also the validity standards, and he constructs a mechanic for deciding rhetoric validity, putting literally into the balance asserted level of certainty and established level on certainty, we can then have an overstating or an understating of a claim. The argument will be declared valid if it is accurately qualified or understated (Sproule 1980, p. 88-89).
He recognizes that values are multiple and play a part, they can be attributed to persons. But he does not treat differently attribution of value to an end, to a mean, a state of being or a type of action. And if we regroup together all these inside what we could call figures of attribution, he does not take into account evaluation as such of X or Y in terms of specific criteria or in terms of specific values. The many different ways into which values can enter into a proposition are all lumped together.

4. T-D model
The book from Toussaint and Ducasse is a school handbook for students, most of the times of age 16-19, what is called in the province of Quebec (other provinces in Canada use a different teaching structure) the collegial or CEGEP level (Collège d’enseignement général et professionnel). This comes after the secondary school, for some students it constitutes a terminal degree, and for those who wish to obtain a University degree, the “college” level diploma is a mandatory requirement. A few other books are also available in Québec, they are especially used in one of the three required philosophy courses at the college level, the one that treats rational thinking and argumentation. The TD book comes with exercises, many examples, schematic representations etc. This book is also full of precise recommendations specifying how to proceed in the construction of an argumentative text. They tend to work by starting with propositional sentences, in the context of an argumentative development that is to be made afterwards.

They look especially at written argumentation, starting also with basic elements about attribution in ordinary language. Their general approach is dialectical in a sense that it involves taking explicitly into account the statement of a position (we would say a claim) on the basis of a problem-setting; the first step is the constructing of the position with its main arguments, including links between position and arguments. This leads to the formulation of opposing arguments, and to the answer or refutation of the arguments that go with that counter position or opposing claim. A good argumentation has to take into account the opposing side in a debate. They propose also to furnish a finale in reasserting the position taken and announced in the beginning. They aim at facilitating the construction of argumentative claims by students, while helping to see how an argument actually functions in different cases. The notion of “une problématique”, meaning the problematic, or the way a question is posed and pre-structures the discussion, is the necessary starting point in their perspective. As is also the idea of a controversial domain, an element that is required since we will not argue about the obvious or the uncontroversial, as we have learned since the beginnings of rational thinking.

Their vision of what is an argument is also quite specific. To three types of problem-settings, three types of statements and positions will correspond. There are assertive, evaluative and directive claims, which they call positions; most of the times the authors will aim at giving a precise and short formulation of the position/claim in a single proposition, including the argument used, for instance “The existence of unions was beneficial to workers because since their existence, the number of hours of work for a week has been reduced…” (Toussaint-Ducass, 1996, p. 119). For them, assertives are statements that answer to questions about determination of reality, they are deployed in a problematic about the existence or not of something or aiming to sustain or deny some attribution of a characteristic to some thing. To quote them: “The statement of an assertive position is a judgment that answers to a problematic question that is about the existence of a reality, its nature or the relationships between realities”[iii] (Toussaint-Ducasse 1996, p. 58). These statements can be categorical and certain, or hypothetical. The hypothetical is seen as something that could or could not exist, and its existence is seen as depending on a condition, that the argument will have to show. It is in that restricted sense that they take into account modality, but they do not discuss it as such. A certain assertive position will have to be confirmed by facts, meaning data acceptable by all. A hypothetical assertive will be justified by a realisable condition (like when people say: It would be possible to recycle more if the cities would furnish accessibility tools like recycling bins). As for evaluatives and prescriptives, they respectively come from a problematic of value and evaluation, and a problematic of what to do or not do, but they are not presented in modal terms.

Evaluatives do not come either as categorical or as hypothetical, and by their examples we can see in fact that argumentation in those guises tends to support a categorical affirmation of the positions taken. One example is the following:
(2) “The new reproductive techniques are more harmful than good for the human species (is declared harmful (néfaste) what provokes destruction of human life to satisfy a whimsical desire)” (Toussaint-Ducasse 1996, p. 201).

They take into account and discuss one counter-argument:
(3) “Yes, but they also permit to prevent and cure genetic diseases”,
to which they answer by giving another counter-argument:
(4) “But the danger of genetic selection is greater then the benefices of preventing and curing grave sicknesses” (Ibid.)

We should note that (2) is an affirmative assertion, even though some validation is seen as required and is offered inside the handbook. We should also note that (3) is also backed by some elements in the text, but (4) is more again a general evaluative assertion that would require more clarification, which they develop only a little. One massive statement seem to be refuted by another massive statement, we are passing from Charybdis to Scylla. Hypothetical statements might be required in those kinds of issues.

Arguments of this kind are said to rely on value judgments, and they can be backed by consequential arguments (called pragmatic by explicit reference to Perelman) or “facts corresponding to a non pragmatic evaluation”, i.e. referring to norms, values or principles (Toussaint-Ducasse 1996, p. 77). Directives raise issues of how to act, they prescribe or forbid some way of acting or behaving. According to them, a Directive argument can be justified as a moral obligation by recourse to a general norm, or it can be justified as a necessary means to a justified end (Toussaint-Ducasse 1996, 200-201). This gives us a total of six argumentative structures in to which rational argumentation is supposed to occur or can occur in written developments.

An example will show how they would have difficulty to stay neutral on some specific important issues. Translated in English, it would go like this: “Feminine and Masculine characters are more acquired then innate, as we ca see by Margaret Mead’s study on three ethnical groups of New Guinea that shows different ways to be a woman or a man” (Toussaint-Ducasse 1996, p. 109).  The authors proceed systematically, while explaining this example, first to the clarification of the statement, then to the clarification of the binding relationship between the argument and the position. This binding leads to clarify the content of Mead’s study, giving details about the ethnic groups to which she refers, which leads to an intermediate conclusion showing that qualities and roles are not universal, which permits the main conclusion as to the acquired aspect of gendered behaviour. This position is seen by them as a certain Assertive position backed by confirmation in the facts of a well known research. Of course, thus formulated the position can seem to be backed by the Mead study.

In their model, the normative or deontological argument surfaces for the two types of statements that have to do with the ethico-moral dimension, i.e. the evaluative and the directive, as a counterpoint to the pragmatic or teleological argument that is also an option in both types of statements. This seems to confirm the closeness, almost redundant character of the evaluative and of the prescriptive to one another as categories in their model. Six elements then sum up every argument according to TD: possible or actual facts, ends and means, norms and values.

Let us note also that this closeness between evaluation and prescription is discussed in ethical theory, in the same movement as their difference is also debated. The question of the relationship between norms and values is also a difficult one, since if the values can inspire norms, many different norms or rules of behaviour can claim to be manifestations or realizations of a single value. Sproule notes that this often goes hand in hand with different definitions, or the different play of interpretives (Sproule 1980, p. 199-201, section “Conflicts based on the same value”). We also have a similar structure for the assertives, since we can put facts and data as backing, but we can also put forward a realizable condition, referring here again to action, as the model does for the evaluative and the directive. Facts can then be seen as analogues to a norm in the domain of reality issues.

Throughout their book, Toussaint and Ducasse emphasize the importance of the validation link between argument and position. Its importance goes with its fragility in many situations, where it is in need of reinforcing. They do not have the interpretive category, and they do not raise the question of the constellation of terms used to discuss an issue including figures of speech, metaphors and names, what is called framing in communication and media studies and in some trends of rhetorical studies or in social sciences more generally (see for instance Tversky and Kahnemann 1981). But they use definition constantly in their work. Their use of the problematic and of problematization understood as problem-setting takes into account this dimension that is sometimes called framing, but from a philosophical point of view. For instance, they will define “the development of an assertive position” as the “manifestation of the meaning that we give it. It includes the clarification of the meaning of the key words with descriptive definitions and illustrations, and the clarification of the general meaning of the statement of the position” (Toussaint-Ducasse 1996, p. 61). In other words, the reflective use of definition will help clarify and develop an argumentative position.

Distinguishing as they do between evaluation and prescription is interesting, but taken in itself it would not be sufficient. For instance, there are nuances to take into account on each side. After all, it is one thing to attribute value, it is another to judge according to a certain value positively or negatively, which is evaluation properly speaking (Dewey 1939). And it is another thing again to intimate a certain course of action. For that matter, their classification does not take into account the differences between prescribing, giving an order, pleading for a practical solution, suggesting a course of action, etc.

5. Concluding remarks
If the Sproule book can be used to indirectly document its readers about the United States of circa 1975, giving us information about the Vietnam War and the Nixon era, the TD book can be used to document the general questions and ideas discussed and abundant in the young Québec population since the late seventies through the 1990s.

The professors of philosophy that are Toussaint and Ducasse rely on definition to develop argumentative strengths; the speech communication professor that is Sproule shows clearly the political use of interpretation in the understanding of political events.

We should note that the two models agree on the importance and the specificity of statements about issues of fact. In Toussaint-Ducasse, something is missing compared to Sproule, the interpretive, and some element is added, the prescription that they call the directive.

(1) The models have specific features that say something of their usefulness
We can say that TD emphasize the ethico-moral by giving it two thirds of their attention already in terms of the categories they put forward.  In TD the examples forcibly have a tendency to be taken inside the vast domain of moral issues. This happens while discussing possible positions about ethical issues. We have to see a correlation between their emphasis on ethico-moral issues and the fact that two of three of their main categories are relevant to those kinds of issues (the Evaluatives and the Directives). Their model might then be specially useful for working on a corpus of moral or ethical judgments.

In S the examples and problems treated are set in terms of more broadly construed political dilemmas. Here again, a correlation has to be seen between the importance of the Interpretive category in politics generally speaking and the fact that the preferred examples are taken into that domain. This says also something about its possible usefulness.

The categories that are specific to each model (Interpretive for Sproule and Directive for Toussaint-Ducasse) have a structuring importance in their respective theories, they serve as grouping and organizing structures; in that sense, each of the elements of the two triads work as classifiers of arguments, and also as selecting tools for picking up and constructing examples. But since Definition (or Interpretives) and Prescription (or Directives) are what distinguishes them respectively the one from the other, they also give us the specificity of their respective approaches. In that regard, they work as classifiers of their theories of argumentation taken globally.

In Sproule we have four types of issues, one being the policy issue and concerning action generally speaking, He does not develop this domain by looking at a specific type of propositions, like T-D is doing by focussing on prescriptions. This is probably because policy issues are seen by Sproule as too complex to reside only in the explicit directives or prescriptions.  He has specific chapters towards the end of the book to discuss policy analysis that are in fact the culmination of the volume. These chapters treat “what should be done” in terms of “the use of argument to establish or refute a policy position”, which is really more than just prescribe a specific course of action. He especially shows how on policy issues, the three levels are necessarily present and intertwined in a complex manner.

One of its strength compared to the TD model is its taking into account of the interpretive. We could say that without considering it as such, he touches the framing questions but limits it to the grouping factor of a series of facts and by saying it is the language used that raise issues of definition. What he lets on the side is what is called framing more broadly speaking today: namely the use of this and that term, name, adjective or category in the way to discuss an issue. Framing also encompasses the problem setting of a specific issue seen as the way that the problem or question is formulated, this is close to TD’s intentions.

The TD model gives a general structure that provide us with a very basic outlook of argumentative reasoning that is easy enough to help develop some argumentation skills for beginners. It puts emphasis mostly on the problem of the relationship between the claim being made and the arguments that sustain it. Its way to deal with definition issues is to render conscious and reflexive the meaning of the concepts used, by inserting into a writing strategy the question of the meaning of the terms discussed.

(2) Limits of these models
We do not find in these books a theorizing of the speech act dimension of argumentation, as we find in the books from Van Eemeren and colleagues. We do not achieve the clarity and precision of Walton on the analytical-logical aspect, and neither the rhetorical clout of Perelman or the explicit wish of keeping together the logical, the dialectical and the rhetorical, as we find in Tindale (Walton, Reed and Macagno 2008; Perelman 1977; Tindale 1999). In terms of handbooks, another book seems more rightly designed for decision makers (Rieke, Sillars and Peterson 2008), even though nothing of the kind is available in French.

Neither of these models really takes into account modality and rebuttals, as Toulmin did in his celebrated 1958 book. We noted that TD distinguished, in matters of fact, the certain and the hypothetical, but there is no reason why the evaluatives and directives should not also be theorized in terms of modality, a thing they avoid. TD is closer to an inferential logic by their insistence on validation links and inference. Sproule takes into account the ethos – pathos – logos triad, whereas TD neglects it. We cannot say that they are very close to informal logic in the sense the expression took in the last decades. The TD model lacks some developed discussion of induction, deduction and abduction, basic reasoning skills that are forcibly required in an informal logic perspective, as we can see for instance in Walton’s books or elsewhere.

Of course, Sproule recognizes the distinction between evaluations and prescriptions, but he does not give it a specific treatment. In fact, as we have seen, Sproule underscores the notable difficulties in some contexts to bind together an evaluation with a specific practical position. Even if it is given a great importance in both works, we can not say that the value issues are clearly situated in them, they are supposed to be already understood.

(3) Their respective context of use
In the case of Sproule, almost all of his examples are taken from the political domain. This goes with his well known interest for the political sphere, as we can see by his list of books and articles, especially his work on propaganda issues. If we keep in mind that political actors are supposedly experts in defining the terms of a public discussion, and if we remind ourselves of the necessity in which politicians are situated to frame problems and solutions according to their party’s way of defining the issues at hand, we will not be surprised by this emphasis that is visible in the sheer structure of the analytical tool that Sproule provides (Reese, Gandy and Grant 2003). Said in other terms, he has a model that fits well with the purpose of looking at the political sphere and to policy issues in particular.  As for T-D, their immediate context is clearly that of ethical discussions properly speaking inside philosophy classes, even though the book presents itself mostly as an introduction to argumentation for undergraduates. The examples are simple to understand, and do not require a high level of knowledge, for instance of the recent history, but they require and contribute to an ethical consciousness of debated questions. Some examples touch at the political, but considered from a moral point of view. Their analytical tool then reflects this privileged domain of discussion which concerns ethical discussions and issues, mostly to be held in classes.

The field-dependency and field-relatedness of argumentation is something well established since Toulmin’s 1958 groundbreaking work. Does our work here show a field dependency not only of argumentative practices, but also of theoretical work about argumentation? We certainly have showed a correlation between preferred domain of interrogation and the categories put in the forefront, even though we did not select a quantitative approach and have not endeavoured to treat exhaustively their respective material. Until further verification then, it would seem that the preferred field of application and research has “selected” the required dominant categories, in one case the Interpretive, in the other the dual system of Evaluatives and Directives, respectively useful especially to understand in some way political phenomena, or to orient action and evaluate practices. To consider things in the opposite direction (the categories constructed permitting to select domains and preferred examples) would only be to consider the other face of the same coin.

NOTES
[i] Sproule (1980) and Toussaint-Ducasse (1996), respectively the S model and the TD model for the ends of this disussion. See references for the bibliographical details.
[ii] Many different terms could be used to describe identifiable argumentative procedures, like the ad baculum, etc. Instead of using the “scheme” word here, that is used with great efficacy by Walton and colleagues theses days, or to talk of topoï, that could also be valid but would refer us to Aristotle, Cicero and the other classics, we prefer to use here, for our describing purposes, “argumentative structures” that seems general enough to take into account the work of Sproule and Toussaint-Ducasse.
[iii] Personal translation, as for the following.

REFERENCES
Aristotle, On Interpretation (2004). E. M. Edgehill tr. http://classics.mit.edu//Aristotle/interpretation.html
Brandt,  William J. (1970). The Rhetoric of Argumentation. Indianapolis, Bobbs Merrill.
Dewey, J. (1925). Experience and nature. John Dewey, The later works, 1. Carbondale, Southern Illinois University Press, 1991.
Dewey, J. (1939). A Theory of Valuation. Chicago, University of Chicago Press.
Perelman, C. (1977). L’emprire rhétorique. Rhétorique et argumentation. Paris, Vrin.
Reese, S. D.., Gandy, O.H. & Grant, A.E. (2007). Framing Public Life. Perspectives on Media and our understanding of the Social World. New York, Routledge.
Rieke, R.D., Sillars, M.O. & Peterson, T.R. (⁷2008). Argument and Critical Decision Making. Boston, Allen & Bacon.
Schön, D. A. & Rein, M. (1994). Frame Reflection. Toward the Resolution of Intractable Policy Controversies. New York, Basic Books.
Sproule, J. M. (1980). Argument. Language and its influence. New York, McGraw-Hill.
Sproule, J. M. (1992). Propaganda and Democracy.. The American Experience of Media and Mass persuasion. Cambridge and New York, Cambridge University Press.
Sproule, J. M. (1996). Speechmaking. Rhetorical competence in a postmodern world. Madison (Wis), Brown & Benchmarking Publishers.
Tindale, Christopher W (1999). Acts of Arguing. A Rhetorical Model of Argument. Albany (NY), State University of New York.
Toulmin, S. E. (1958). The uses of argument. Cambridge, Cambridge University Press.
Toussaint, N & Ducasse, G. (With the collaboration of Georges A. Legault) (1996). Apprendre à argumenter. Initiation à l’argumentation rationnelle écrite. Sainte-Foy, Le Griffon d’Argile.
Tversky, A. and Kahneman, D. (1981). “The Framing of Decisions and the Psychology of Choice”. Science. Vol 211(4481) pp.453-458
Van Eemeren, F. H. & Grootendorst, R. (2004). A Systematic Theory of Argumentation. The pragmai-dialectical approach. Cambridge & New York, Cambridge University Press.
Walton, D., Reed, C. & Macagno, F. (2008). Argumentation Schemes. Cambridge and New York, Cambridge University Press.
Watzlawick, P., Beavin-Bavelas, J. and Jackson, D.D. (1967). Pragmatics of Human communication. A Study of Interaction Patterns, Pathologies, and Paradoxes. New York, Norton & Co.

1. Introduction
Argumentative practices in various forums for computer-mediated, or online, communication have been an object of increasing interest among argumentation researchers (see, inter alia, Aakhus 2002a, 2002b, Amossy this volume, Chaput & Campos 2007, Doury 2005, Jackson 1998, Lewiński 2010, Weger & Aakhus 2003). In accordance with the descriptive and normative functions of argumentation theory, such studies combine, in a more or less balanced manner, analysis of some modes or patterns of argumentation characteristic of online formats for discussion with attempts at evaluating the patterns under study, or the format at large, against a certain idealised context for argumentative discussion (such as the pragma-dialectical model of a critical discussion). In this paper, I focus on one pattern of argumentation – the collective antagonist – that can be distinguished in discussions held in political Web-forums accessible through Google Groups. In the pattern of the collective antagonist groups of individual arguers jointly criticise argumentation advanced by other arguers. The goal of the paper is to give a pragma-dialectical account of this pattern in both descriptive and normative terms. Hence the main questions to be addressed are: How can pragma-dialectics contribute to a more subtle understanding of a pattern of collective criticism? Is collective criticism conducive or obstructive to realising reasonable forms of argumentation embodied in the ideal model of a critical discussion? Finally, what are the possible challenges that the analysis and evaluation of collective online criticism opens for argumentation theory?

In order to address these questions, I will proceed in four basic steps. First (section 2), I will describe these characteristics of online discussion forums that are directly relevant to the task of investigating and assessing collective criticism. Second, (section 3), I will analyse the pattern of the collective antagonist on the basis of a fragment of an actual online discussion. Third (section 4), I will examine the potential of collective online criticism for supporting reasonable argumentative discussions. Finally (section 5), I will mention some methodological and theoretical challenges that the analysis and evaluation of online discussions can pose to argumentation theory, and pragma-dialectics in particular.

2. Online discussions as informal multi-party deliberations
Asynchronous online discussions, in which users “post” (i.e., send), read and reply to publicly available messages in a form similar to e-mail (i.e., without rigorous time and space constraints), belong to the oldest yet still very popular technologies of computer-mediated communication. Today, systems such as Google Groups (http://groups.google. com/) provide a unified Web-based design for accessing two important sub-types of online asynchronous discussions: Web-forums, which are hosted on Google servers, and the independent Usenet newsgroups, to which Google provides only a popular gateway. The range of topics discussed in such forums is virtually unlimited, and politics has a prominent place among them.[i] Online political discussions held via Google forums are informal, grassroots initiatives hosted and administrated by politically engaged Internet users which are in no explicit and direct way connected to any institutional decision-making processes. Because of that, such political discussions are a specimen of informal public deliberations, in which opinions are publicly expressed, challenged, defended and criticised, without the aim of arriving at some explicitly declared final outcomes.[ii]

Two interrelated characteristics of such argumentative forums for informal online deliberation are of special importance to analysing patterns of collective argumentation: first, online forums allow for participation of large groups of discussants and, second, this participation is predominantly unregulated.

Large-scale participation is afforded by the technological design of open online forums (or Usenet newsgroups): since any (registered) Internet user can join and leave discussion at any point, the pool of discussants may be quite considerable. Moreover, various (groups of) participants can be simultaneously developing several lines of discussion; in this way, the main topical thread of a discussion can fork out into many sub-threads. Taking such considerations into account, Marcoccia (2004) proposes that online discussions should be analysed as “on-line polylogues” with a complex “participation framework.” As he notes, polylogues in general are characterised, on the one hand, by the “lack of collective focusing,” since there is often no one centre or main thrust of discussion, and, on the other hand, by “the existence of varied focuses,” since discussants can focus on specific parts of interaction, for example by participating exclusively in selected sub-discussions (Marcoccia 2004, p. 118; see Kerbrat-Orecchioni 2004).

What is unique to informal online polylogues is that compared to many institutionalised forms of multi-party deliberation they contain hardly any explicit procedural regulations. No clear “rules of order” – known in many highly formalised institutional polylogues, such as parliamentary debates – which discipline the exchange of arguments and criticisms are stated for online political discussions. Therefore, elements such as the order of speakers, the length and the shape of their contributions (type of allowed, or even required, arguments and criticisms), the possibilities to address criticisms and develop arguments, the overall length of discussion, etc., are not prescribed, but rather are left to be decided by the discussants themselves.[iii] Online discussions are thus open, emergent activities in which exchanges of arguments and critical reactions develop freely in accordance with the direction a discussion takes depending on the online arguers’ ongoing participation (or lack thereof).[iv]

The combination of factors such as freedom of access and participation, opportunity to involve in many-to-many interactions and lack of strict regulation and moderation, make it possible for various lines of online discussions to overlap and affect one another in a somewhat disorganised manner. Therefore – especially when compared to tightly regulated one-on-one dialogic exchanges – computer-mediated polylogues have been considered as rather chaotic forums characterised by disrupted global topical relevance and local turn-to-turn adjacency (Herring 1999). Notably, the patterns of responding in multi-party asynchronous online discussions are quite peculiar:
…there is not a one-to-one correspondence between an initiation and its response. Multiple responses are often directed at a single initiating message, and single messages may respond to more than one initiating message, especially in asynchronous CMC [Computer-Mediated Communication – ML], where longer messages tend to contain multiple conversational moves […]. Moreover, many initiations receive no response. (Herring 1999, online)

Shortly, argumentative discussions in various Web-forums (or Usenet newsgroups) are online polylogues with fluid participation and convoluted patterns of conversation (Herring 1999, Marcoccia 2004).

Still, there are other noticeable qualities of such online discussion forums that to a certain extent counterbalance the apparent chaos of unregulated polylogues. Notably, these forums support asynchronous rather than real-time communication, so there are no time (and space) constraints to reflect on and advance arguments and criticisms. Moreover, individual contributions to discussions (“posts”), are usually recorded, numbered, and organised in topical threads (or discussion trees). This is important since, as has been observed, “the record of exchanges often available to participants in online debate […] allows careful consideration of the development of ongoing arguments” (Dahlberg 2001, online).

Altogether, despite noticeable deviations from a neat dialogical structure consisting of dovetailed adjacency pairs (such as argument–critical reaction), online multi-party discussions can still be seen as organised and patterned around the vital characteristics described above. In the pragma-dialectical view, such characteristics are methodically grasped as restrictions and opportunities of an argumentative activity type of online discussion forums (Lewiński 2010).

3. The pattern of the collective antagonist
The goal of this section is to describe the pattern of the collective antagonist that can be identified in online political discussions on the basis of their close argumentative analysis. The analysis presented below follows methods of qualitative study of argumentative discourse developed within the pragma-dialectical theory of argumentation (van Eemeren, Grootendorst, Jackson, & Jacobs 1993).

Discussion 1 took place in the last weeks of the 2008 presidential campaign in the United States on one of the discussion forums available through Google Groups: PoliticalForum. It was sparked by a campaign event in which Barack Obama, during a meeting with residents of a neighbourhood in Ohio on 12 October 2008 (only three days before the final presidential debate), was asked by “Joe the Plumber” about his tax plans as a future president. The “plumber” suggested that the new tax proposals would negatively affect his plans to expand the small plumbing business he was working in. In response, among other things, Obama explained that tax would only be levied on businesses bringing more than $250.000 a year in revenue and added: “I think when you spread the wealth around, it’s good for everybody.”[v] The event quickly became a hot campaign topic, and was mentioned a number of times by Obama’s Republican opponent John McCain during the last presidential debate.

(Discussion 1)[vi]
nobama thinks he is robin hood
http://groups.google.com/group/PoliticalForum/browse_frm/thread/e33251a56f53930f/d7781d4f78961e69?tvc=1#d7781d4f78961e69

1. mark Oct 15 2008, 11:45 am
when asked by a plumber if his was going to raise his taxes, barry said he had no problem taking his money to spread the wealth.
socialism but we all know barry is indeed a socialist.

2. Travis Oct 15 2008, 12:48 pm
Heis.  You just didn’t spell it right.  Robbing Hoodwinking

3. jenius Oct 15 2008, 1:47 pm
any one who knows anything know that Obama is only going to raise taxes on those who make more that$250,000 a year. to me thats a good
plan.  I am disabled and living on a fixed income.  I bet thats agreeable to most people too. that is why same old Mccain is not going
to win this election. vote for Obama,a vote to justice and equality for the poor and the middle class. Jenius

4. Lone Wolf Oct 15 2008, 2:29 pm
The multimillionaire, that supported the bailout of corporate crooks with the funds of those they ripped off, and who receives more
donations from Wall St than McCain. That Mr Equality. Wake up my friend, the Dems and the GOP are two sides of the same coin. Obama is
an unmittigated lying low life reprobate.

5. mark Oct 16 2008, 0:20 am
2/3 of those being taxed by barry are small businesses who will either be forced to reduce staff, or close their doors.  since the small
business is the backbone of our economy, please tell me how this is a good thing.
oh yeah and let us not forget that he will repeal the Bush taxcuts, so he is raising everyones taxes.

6. Gaar Oct 16 2008, 0:22 am
On Oct 15, 4:20 pm, mark <marsupialm…@sbcglobal.net> wrote:
> oh yeah and let us not forget that he will repeal the Bush taxcuts, so
> he is raising everyones taxes.
Actually, he now claims he won’t do that.

11. Jenius Oct 16 2008, 11:56 am
thats a complete falsehood, read the plan. anyway if your business is making that much you should be paying more taxes, and may not even
qualify as a small business anymore. Jenius

24. Hollywood Oct 16 2008, 3:42 pm
mark,
Are you a complete idiot? What percentage of “small businesses” have a profit of $250,000.oo after all deductible expenses? WTF are you
called “small business”?

29. Lone Wolf Oct 17 2008, 10:03 am
The backbone of the US was heavy industry, steel smelting and car manufacturing to earn export dollars, not small business that operates
within the domestic economy and does nothing to improve US trade deficit.

Why do you bother listening to what Obama says, he is making it up as he goes along? He is craven populist, what do you expect him to say?
BTW. The US is screwed

Discussion 1 is initiated by mark’s comment regarding Obama’s meeting with “Joe the Plumber.” In this very context (the last days of the election campaign), a statement that ‘barry [Barack Obama – ML] is a socialist” or, more precisely, that Obama endorses a “socialistic” tax plan to “spread the wealth,” can be directly reconstructed as an argument for a standpoint “one should not vote for Obama.” After Travis’ affirmative remark in turn 2, the main difference of opinion in this discussion is made explicit in Jenius’ turn 3. Jenius advances a standpoint opposite to mark’s: one should “vote for Obama,” because his policies promote “justice and equality for the poor and the middle class” and, in particular, his tax proposal is “a good plan.” Following Lone Wolf’s short and outspoken call for a third way in American politics (turn 4: one should vote for neither Obama nor McCain, because “the Dems and the GOP are two sides of the same coin”), mark responds to Jenius’ challenge in message 5 by advancing a complex of arguments that can be schematically pictured in the following way (see Figure 1).

Figure 1 Structure of mark’s argumentation in turn 5
(Above the line is the main standpoint and the main argument reconstructed from mark’s initiating post 1; “-/ 3. Jenius” means that the discussant Jenius in turn 3 reacts critically to this element of mark’s argumentation.)

As an analytic overview of mark’s arguments in figure 1 shows, his short message contains a rather complex argumentation structure. The bone of contention here is the sub-standpoint (1) that Obama’s tax plan is not good, expressed by means of a rhetorical question of sorts (“please tell me how this is a good thing”). This sub-standpoint is supported by a multiple structure consisting of two independent arguments: (1.1) Obama’s plan will lead to a collapse of the American economy and, apart from that, (1.2) it leads to a universal tax rise (an unexpressed premise for both of these arguments is that none of these is a good thing ). The former argument is further supported by a long subordinative structure, in which many premises are left unexpressed (but are reconstructible on the basis of the entire discussion or general background knowledge). The latter argument is supported by a fairly simple coordinative structure: Obama is planning to raise taxes for both small businesses (1.2.1a) and rich people (1.2.1b), so “he is raising everyones taxes” (1.2).

Mark’s post receives four direct responses, all of them critical: by Gaar (6), Jenius (11), Hollywood (24), and Lone Wolf (29). In this way, a collection of individual participants to an online polylogue criticises distinct parts of complex argumentation advanced by another arguer, thereby creating “the collective antagonist.” Moreover, each of these reactions opens a new sub-discussion: this is how discussion 1 splits into four simultaneously held sub-disputes regarding four different elements of mark’s argumentation put forward in turn 5.
Individual arguers’ joining forces leading to a collective construction of argumentation is a well-known phenomenon in group discussion usually studied under the label of “tag-team argument” (Brashers & Meyers 1989, Canary, Brossmann, & Seibold 1987). However, whereas the study of tag-team argument was focused on a joint construction of complex argumentation structures in the context of face-to-face, small group decision-making, what is evident in discussion 1 is joint criticism of an argumentation structure in a pseudonymous and mediated context of large group discussion which is not (immediately) aimed at generating a decision to act in any particular way. Moreover, while tag-teams have been analysed as neatly delineated groups with consistent, opposing standpoints to defend, the collective criticism here is collective only in the sense of the object of criticism. Gaar (turn 6), Jenius (turn 11), Hollywood (turn 24), and Lone Wolf (turn 29) team up to criticise mark’s argument advanced in turn 5, but otherwise they do not seem to be jointly defending any one consistent position. Gaar, in fact, similarly to Travis (turn 2) seems to be sympathetic with mark’s anti-Obama opinions; his criticism against the content of facts adduced by mark is thus more of a correction of the position he otherwise agrees with. By contrast, both Jenius and Hollywood attack mark from a pro-Obama point of view; in this sense, they create a regular tag-team which jointly produces complex argumentation (next to complex criticism). Yet differently, Lone Wolf argues both against pro- and anti-Obama position, and thus stands alone, aligning with one of the main positions in the discussions only when criticisms are to be voiced against the other position.

Despite such differences with clearly defined tag-teams, there is some kind of regularity in this rather complicated web of critical reactions: different critical respondents precisely target different elements of the same piece of complex argumentation. One can say that in this case arguers enact a horizontal criticism: even though the criticisms of Gaar, Jenius, Hollywood, and Lone Wolf are clearly voiced one after another, rather than simultaneously, they do not create a sequence of critical reactions in which one of the critics picks up where another left. In this way, every critical reaction seems independent from another, at least in terms of their argumentative import. As a result, online discussants create one line of comprehensive attack against another discussant’s arguments expressed in one single message.[viii] Characteristically to online discussions, such multiplied criticism does not lead to a final resolution of the expressed differences of opinions: the separate sub-discussions that the criticisms of Gaar, Jenius, Hollywood, and Lone Wolf instigate are not concluded, but instead fade away when discussants stop contributing to them.

4. Evaluation of collective criticism
It has been stressed by pragma-dialecticians that smooth implementation of the ideal model of a critical discussion usually faces serious obstacles in actual circumstances (van Eemeren et al. 1993, pp. 30-34). One of such obstacles may result from a tension between the competing demands for open participation and reasonableness in public discourse.[ix] Jackson (1998), who analysed conditions for argumentation in Usenet discussions, grasped this tension by referring to two first rules of a critical discussion. On the one hand, in accordance with rule 1 (“Discussants may not prevent each other from advancing standpoints or from calling standpoints into question”; van Eemeren & Grootendorst 2004, p. 190), arguers should be able to freely exercise their unconditional right to voice objections against others’ position. On the other hand, following rule 2 (“Discussants who advance a standpoint may not refuse to defend this standpoint when requested to do so”; van Eemeren & Grootendorst 2004, p. 191), arguers should meet their conditional obligation to defend one’s own position when challenged. According to Jackson (1998, p. 189), meeting these two conditions simultaneously in open forums for online deliberation, such as Usenet groups, may be difficult due to the characteristics of their design. It is exactly because such forums are open for everyone to enter discussions by advancing and criticising opinions without restrictions rule 1 for a critical discussion can easily be followed. It is equally easy, however, to abandon or shift discussions and thus evade the burden of proof associated with one’s challenged opinions, violating rule 2.

These general observations seem to apply well to the pattern of the collective antagonist. On the one hand, the pattern of the collective antagonist is conducive to realizing reasonable forms of argumentation, because multiple criticism enhances critical testing of public opinions. Standpoints and arguments expressed on Web-forums can be unlimitedly called into question, to the satisfaction of rule 1. This is the case even if some kind of disorderliness in online arguers’ critical reactions can be noticed. As argued above, a collective of critics is not necessarily a tag-team acting consistently towards one common purpose, but rather a certain strategic alliance that comes into being in a particular dialectical situation. However, even if this alliance is purely opportunistic and temporary (or even coincidental), it plays an important dialectical role. From the perspective of a critical discussion, such joint production of criticisms allows for the collectively “optimal use of the right to attack” (van Eemeren & Grootendorst 2004, pp. 151-152). Since online forums give abundant opportunities to react critically to argumentation in as many ways as possible by as many people as possible, factors such as lack of individual ingenuity in launching comprehensive criticism are of lesser importance. In effect, the potential for open public scrutiny of the opinions and arguments advanced increases.

Yet on the other hand, the pattern of collective criticism can be deemed obstructive to realizing reasonable forms of argumentation on a few weighty accounts. Most obviously, in order to be reasonable, individual objections adding up to one collective line of argumentative criticism should be good, relevant objections. This can be seen as a precondition for the potential for critical public scrutiny to be actually realized. This precondition is certainly not universally met. Analysts of online discussions noticed that the minimally designed, open and loosely regulated forums for multi-party discussion are susceptible to unqualified and irrelevant objections (Jackson 1998, pp. 190-193), and the resulting “micro-level digression” and “macro-level drift” of discussions away for the issues that are supposed to be discussed (Aakhus 2002a, p. 127). Critical reactions can also involve a straw man, that is, an illegitimate reformulation of the criticised opinions and arguments (Lewiński 2010, ch. 9). Moreover, as often pointed out, an opportunity for uninhibited critical uses of online technology is also an opportunity for getting away with rampant abuses of it, among which the use of derogatory, abusive language (so called flaming) seems to be the most notorious (see Amossy, this volume). Furthermore, multiple criticisms can be repetitive, which is the case when various individual antagonists propose no more than stylistically different variants of basically the same objection. Shortly, individual critical reactions making up one collective antagonist can simply be fallacious.

The study of fallacious criticisms in online discussions is not, however, where the evaluation of multiplied criticism should end. That is because even if individual criticisms voiced by different arguers are reasonable in the sense of being relevant, relatively civil, and original (as is largely the case in discussion 1), the entire collective criticism can still be problematic in terms of its impact on the quality of public discussions that goes beyond fallaciousness of particular argumentative moves. The problem lies also in the design of open online forums for informal deliberation. In such forums, multiple criticisms can easily overwhelm defences that are in fact strong, or perhaps even conclusive. One way of grasping this problem is to analyse it as a difficulty that online discussions create for arguers willing to observe rule 2 of a critical discussion.

The point is, that for an arguer confronted with a collective antagonist on a Web-forum it may be very difficult, or indeed impossible, to satisfactorily discharge the burden of proof by consistently addressing all criticisms. This is partly due to the polylogical character of online discussions in which lines of attack and defence may become terribly convoluted. It is certainly much easier for an argumentation analyst, than it is for an actual arguer involved in an ongoing multi-party discussion, to reconstruct a consistent, ordered pattern in critical reactions. Moreover, the lack of any moderator who links all developing sub-threads back to the main standpoint discussed adds to the difficulty of tracing and addressing all criticisms as one coherent whole.

As mentioned in the previous section, one cannot assume that the collective antagonist is always concurrent with the existence of clearly delineated tag-teams that consistently support or oppose one explicitly formulated position. Instead, teams of arguers and critics can “gang up” for one specific round of collective criticisms, and then dissipate in the ensuing polylogue. Such lack of clearly defined, continuous argumentative roles throughout an entire online discussion is important from the perspective of weighting pros and cons in multi-party deliberation. That is because critical objections, even if they are not parts of one consistent position (as is the case in discussion 1) or even when they amount to a collection of fragmented “hit-and-run” strategies (see Aakhus 2002b, Weger & Aakhus 2003), can be still argumentatively forceful, since they multiply the defendant’s burden of proof. By contrast, for positive positions to prevail over the course of deliberation, they need to remain consistent (see Meyers, Brashers, & Hanner 2000).

To conclude – the pattern of the collective antagonist points to a certain imbalance in the opportunities for an advantageous management of the burden of proof. Arguers aiming at a strategic advantage in online deliberations can easily position themselves as parts of the collective antagonist, in which case they do not acquire heavy burden of proof. By contrast, arguers faced with such collective antagonist may find it exceedingly difficult to discharge their multiplied burden of proof: regardless of their individual willingness and ability to do so, in the context of open online forums for deliberation they may find it hard to fully comply with rule 2 for a critical discussion, i.e., to address all criticisms. Apart from the reasons just mentioned, that is the case because such forums provide no tools and regulations that would prevent the imbalance in managing the burden of proof from happening. One such regulation may be a requirement that additional criticisms are only allowed after the protagonist of a standpoint had been given proper chance to address the previously voiced objections. Another might be a requirement that every criticism has to be a “constructive criticism”: one can attack a given position only if one is able to present and defend a relatively stronger position of one’s own.

5. Analysis of online polylogues as a challenge to argumentation theory
In pragma-dialectics all argumentation is reconstructed from the perspective of a critical discussion: an ideal dialogue between the protagonist and the antagonist, who orderly take turns and thus move from a confrontation through opening and argumentation stages to a conclusion. That means that actual argumentation taking place in various communicative activity types (van Eemeren & Houtlosser 2005, van Eemeren, Houtlosser, Ihnen, & Lewiński 2010) is always approached as a more or less imperfect instance of a critical discussion: whether actually occurring between interacting discussants or merely presupposed in one arguer’s monological argumentation. An open problem to be discussed here is that when employed in the analysis and evaluation of fragmented online polylogues, an ideal critical discussion is a useful, but possibly simplified heuristics. It is useful in the process of analysis of discourse, for it provides a comprehensive overview of analytically relevant moves in online disputes. In particular, the model specifies various types of critical reactions that can be performed in argumentative discussions (see Lewiński 2010, ch. 7). Moreover, in the evaluative sense, it allows to spot the departures from ideal forms of argumentation and thus to trace the limitations and imperfections of various actual contexts for (online) argumentation (Aakhus 2002a, 2002b, Jackson 1998, Weger & Aakhus 2003).

Despite a well-documented usefulness of a critical discussion in reconstructing and assessing any form of actual argumentation, including online discussions, its application (or, indeed, the application of any other dialogical model of argumentation, such as Walton’s (1998) “dialogue types”) in examining online polylogues may face serious challenges. That is primarily because for dialectical approaches argumentation is basically seen as an instance of a dyadic exchange.[x] In fact, however, actual argumentative dialogue may take many forms: from simple one-on-one interactions, to activities where a third party interferes to regulate discussions, to complex multi-party exchanges. Indeed, activity types in which third parties play a significant role (for instance mediation sessions and legal trials; see, e.g. van Eemeren et al. 1993, ch. 6) have been consistently and overall successfully studied from a pragma-dialectical perspective. In general, various kinds of multi-party discussions have been approached in pragma-dialectics as variations or collections of fundamentally two-party exchanges (van Rees, 2003). By contrast, the conversation structure of an online polylogue, as described above, may significantly exceed the limits of a dyadic structure.[xi] That implies, inter alia, that arguers can face different difficulties and make use of different affordances than in a dyadic exchange. For example, arguers can attempt to respond to a number of argumentative objections, possibly raised from a few distinct or even incommensurable positions, in one online post. In such a situation, what seems to be a rather sloppy defence when analysed and evaluated from the perspective of one singular discussion (say, A against B), can be the strongest possible argumentative move when taken in the entirety of the polylogue (e.g., A against B and C and D, where B, C, and D make up one collective antagonist of A’s standpoint, but at the same time hold mutually conflicting positions).

More in particular, when it comes to the reconstruction of an online polylogue in pragma-dialectical terms, two options seem to be at an analyst’s disposal, none of them fully satisfactory: The first is the reduction of a polylogue to two clearly delineated camps (one critical discussion between the collective protagonist and the collective antagonist). In this case, however, an analyst simplifies the disagreement space and reduces it to a dialectical pro and contra, while certain “third way” may in fact be advocated by some arguers (see, e.g., contributions of Lone Wolf in discussion 1). The second is the reduction of a polylogue to many simultaneously held one-on-one critical discussions (see van Rees, 2003). In this case, an analyst abstracts from the net of often overlapping discussions that may affect each other in subtle yet important ways.

What follows from such possible complications in the pragma-dialectical reconstruction of online polylogues is that the very notion of strategic manoeuvring (in its strict sense defined by van Eemeren & Houtlosser 1999, pp. 485-486) is not as adequately applicable in the analysis of the polylogical practices as it may be in dialogues. If one gave up the idea that an online polylogue can be always justifiably reconstructed as a discussion between discrete and consistent collective parties (pro and con in case of two parties), then it would be difficult to speak of strategies in the sense of methodical and coordinated attempts at influencing the outcome of a discussion by one of the parties to a discussion. Global strategies (or simply strategies in the proper sense of the word) are not really possible in a chaotic, unpredictable environment in which clear notions of pro and con do not fully apply and argumentative roles constantly fluctuate. Rather – assuming that online arguers still act strategically despite such difficulties – one should speak of local strategies (or tactics) aimed at a rhetorical advantage, implemented in fragmented pieces of inconclusive argumentative exchanges. Further, if participation in a polylogue is reconstructed as participation in many simultaneously held dialogues, then strategic manoeuvring can be happening not only within these reconstructed dialogues, but also across the dialogues, since doing something in one discussion may be primarily directed toward gaining advantage in another discussion. This may happen, for example, when by arguing in one sub-discussion of a polylogue an arguer aims (primarily) at establishing starting points useful in another sub-discussion (with different participants). The idea of strategic maneuvering across discussions, however, stretches the meaning of the term beyond its grounding in one dialectical encounter.

I treat such complications in analysing and evaluating online polylogues as open questions for future consideration, questions to which here I can hardly give even a tentative answer. Still, hoping that analogies do make strong arguments now and then, I would point out that playing one game of chess for three is different than playing three simultaneous regular games of chess between two players. Quite manifestly, the strategies utilised in such chess for three can be decidedly different from regular chess. One such prominent strategy – unavailable in one-on-one contests – is making alliances, i.e., teaming up against another player. However, also the very rules of the game require some modifications and additions. Therefore, if indeed accurate, this analogy points to a need for considering an ideal model of argumentation not limited to a dyadic view of argumentative interactions.

6. Conclusion
The goal of this exploratory paper was to give a pragma-dialectical account of the phenomenon of the collective antagonist observable in online political discussions. To this end, collective criticism has been analysed as a pattern of argumentation afforded by some crucial qualities of open online forums for informal large-scale deliberation, such as the possibility to involve in many-to-many interactions and lack of effective regulation. When assessed from the perspective of a critical discussion, multiplied, collective criticism seems to be good and bad at the same time. It is critical in the sense of the opportunities for comprehensive public scrutiny of political opinions that antagonists of these opinions have, but it is not quite critical in the sense of the opportunities for protagonists to positively discharge their burden of proof and thus conclude discussions with a critically applauded result. Moreover, multi-party online discussions pose some challenges to dialectical approaches to argumentation, according to which a paradigm for analysing and evaluating argumentation is a dyadic discussion between a pro and contra party. Such intricacies of argumentative analysis and evaluation, as well as challenges that may be difficult to overcome, make online political discussions a fascinating object of research for argumentation theorists.

NOTES
[i] Many political Usenet groups rank high among the ‘Top 100 text newsgroups by postings’ (see http://www.newsadmin.com/top100tmsgs.asp). Newsgroups explicitly labelled as political in top 20 include it.politica (Italian, #6), fr.soc.politique (French, #15), pl.soc.polityka (Polish, #17), and alt.politics (English, #18), (consulted 15-07-2010).
[ii] It is an established practice among political theorists to distinguish between two basic goals and, in effect, two general kinds of deliberation: decision-making and opinion-formation. Among others, Fraser contends that deliberation aimed (solely) at opinion-formation amounts to political “discourse [that] does not eventuate in binding, sovereign decisions authorizing the use of state power; [but] on the contrary, […] eventuates in ‘public opinion,’ critical commentary on authorized decision-making that transpires elsewhere” (Fraser 1990, pp. 74-75).
[iii] Netiquette (see, e.g., http://www.dtcc.edu/cs/rfc1855.html), as well as charters of particular forums, do provide some basic guidelines meant to regulate online discussions, but, firstly, they are often not strictly enforced and, secondly, they exhibit a certain “bias towards particular, agonistic forms of discourse” (Dahlberg 2001, online).
[iv] Since the forums for informal online deliberation discussed in this paper belong to grassroots activities underlain by the ideas of free, and free-wheeling, Internet communication, they are, in principle, not moderated.
[v] See, e.g., http://en.wikipedia.org/wiki/Joe_the_plumber.
[vi] Note that due to the topical rather than purely chronological structuring of the conversations even posts far removed in the numbered sequence can be direct responses to some previous posts. All the messages are quoted verbatim, without any editorial corrections.
[vii] Van Eemeren & Grootendorst (2004, p. 4) and Snoeck Henkemans (1992), distinguish between three basic types of complex argumentation structures: multiple (convergent), coordinative (linked), and subordinative (chained).
[viii] Apart from the horizontal variant of collective criticism, one can also distinguish a vertical variant, in which a group of arguers acts in sequence by deepening the previously voiced criticisms against one element of their opponent’s argumentation.
[ix] Jacobs (2003) refers to these two possibly conflicting demands as “two values of openness in argumentation theory”: “freedom of participation” and “freedom of inquiry.”
[x] Discourse analysts studying polylogues point out and criticise a general and “deep-rooted tendency to associate interaction with interaction between two people, considered as the prototype of all forms of interaction” (Kerbrat-Orecchioni 2004: 2). Bonevac (2003) addresses specifically the problem of analysing multi-party discourses in “essentially dualistic” pragma-dialectical approach.
[xi] Conversely, some informal logicians such as Blair (1998), have seen “the limits of the dialogue model of argument” in “solo arguments” performed in contexts of monologues or “non-engaged dialogues.”

REFERENCES
Aakhus, M. (2002a). Modeling reconstruction in groupware technology. In F. H. van Eemeren (Ed.), Advances in pragma-dialectis (pp. 121-136). Amsterdam: Sic Sat.
Aakhus, M. (2002b). The design of forums for online public deliberation and the consequences for argumentation. The Kentucky Journal of Communication, 21(2), 139-149.
Amossy, R. (this volume). Polemical discourse on the Net: “Flames” in argumentation.
Blair, J. A. (1998). The limits of the dialogue model of argument. Argumentation, 12(2), 325-339.
Bonevac, D. (2003). Pragma-dialectics and beyond. Argumentation, 17(4), 451-459.
Brashers, D., & Meyers, R. A. (1989). Tag-team argument and group decision-making: A preliminary investigation. In B. E. Gronbeck (Ed.), Spheres of argument: Proceedings of the Sixth SCA/AFA Conference on Argumentation (pp. 542-550). Annandale, VA: Speech Communication Association.
Canary, D. J., Brossmann, B. G., & Seibold, D. R. (1987). Argument structures in decision-making groups. Southern Speech Communication Journal, 53, 18-37.
Chaput, M., & Campos, M. (2007). A pragma-dialectical analysis of online political argumentation. In F. H. van Eemeren, J. A. Blair, C. A. Willard & B. Garssen (Eds.), Proceedings of the sixth conference of the International Society for the Study of Argumentation (pp. 235-240). Amsterdam: Sic Sat.
Dahlberg, L. (2001). Computer-mediated communication and the public sphere: A critical analysis. Journal of Computer-Mediated Communication, 7(1), available online: http://jcmc.indiana.edu/vol7/issue1/dahlberg.html
Doury, M. (2005). The accusation of amalgame as a meta-argumentative refutation. In F. H. van Eemeren & P. Houtlosser (Eds.), Argumentation in practice (pp. 145-161). Amsterdam: John Benjamins.
Eemeren, F. H. van, & Grootendorst, R. (2004). A systematic theory of argumentation: The pragma-dialectical approach. Cambridge: Cambridge University Press.
Eemeren, F. H. van, Grootendorst, R., Jackson, S., & Jacobs, S. (1993). Reconstructing argumentative discourse. Tuscaloosa: University of Alabama Press.
Eemeren, F. H. van, & Houtlosser, P. (1999). Strategic manoeuvring in argumentative discourse. Discourse Studies, 1(4), 479-497.
Eemeren, F. H. van, & Houtlosser, P. (2005). Theoretical construction and argumentative reality: An analytic model of critical discussion and conventionalised types of argumentative activity. In D. Hitchcock (Ed.), The uses of argument: Proceedings of a conference at McMaster University, 18-21 May 2005 (pp. 75-84). Hamilton.
Eemeren, F.H. van, Houtlosser, P., Ihnen, C., & Lewiński, M. (2010). Contextual considerations in the evaluation of argumentation. In C. Reed & C. Tindale (Eds.), Dialectics, dialogue and argumentation: An examination of Douglas Walton’s theories of reasoning and argument (pp. 115-132). London: College Publications.
Fraser, N. (1990). Rethinking the public sphere: A contribution to the critique of actually existing democracy. Social Text, 25/26, 56-80.
Herring, S.C. (1999). Interactional coherence in CMC. Journal of Computer-Mediated Communication, 4(4), available online: http://jcmc.indiana.edu/vol4/issue4/herring.html.
Jackson, S. (1998). Disputation by design. Argumentation, 12(2), 183-198.
Jacobs, S. (2003). Two conceptions of openness in argumentation theory. In F.H. van Eemeren, J.A. Blair, Ch.A. Willard & A.F. Snoeck Henkemans (Eds.), Proceedings of the Fifth Conference of the International Society for the Study of Argumentation (pp. 553-556). Amsterdam: Sic Sat.
Kerbrat-Orecchioni, C. (2004). Introducing polylogue. Journal of Pragmatics, 36(1), 1-24.
Lewiński, M. (2010). Internet political discussion forums as an argumentative activity type: A pragma-dialectical analysis of online forms of strategic manoeuvring with critical reactions. Amsterdam: Sic Sat.   
Marcoccia, M. (2004). On-line polylogues: conversation structure and participation framework in internet newsgroups. Journal of Pragmatics, 36(1), 115–145.
Meyers, R. A., Brashers, D. E., & Hanner, J. (2000). Majority/minority influence: Identifying argumentative patterns and predicting argument-outcome links. Journal of Communication, 50(4), 3-30.
Rees, M.A. van. (2003). Within pragma-dialectics: Comments on Bonevac. Argumentation, 17(4), 461-464.
Snoeck Henkemans, A. F. (1992). Analysing complex argumentation: The reconstruction of multiple and coordinatively compound argumentation in a critical discussion. Amsterdam: Sic Sat.
Walton, D. N. (1998). The new dialectic: Conversational contexts of argument. Toronto: University of Toronto Press.
Weger, Jr., H., & Aakhus, M. (2003). Arguing in Internet chat rooms: Argumentative adaptations to chat room design and some consequences for public deliberation at a distance. Argumentation and Advocacy, 40(1), 23-38.

1. Introduction
About two years ago, one of the authors of this paper [i] once wrote another paper discussing the dialectical approach within Pragma-Dialectics and Blair and Johnson’s informal logic theory. In a section of that paper, he made the following two points about Johnson’s notion of dialectical tier: “The dialectical tier within an argument marks that the thesis is critically established, and a dialectical history of an argument reveals that the argument is critically developed.” And “the requirement of manifest rationality can be regarded as requiring a process of critical testing for seeking the strongest or the most appealing reasons and better arguments” (Xie, 2008). Both points, unfortunately, brought back Johnson’s negative comments in their later correspondence. Johnson’s remark on the first point is “This is not clear to me”, and on the second, “Not sure of this”.

Besides the author’s disappointment, still there are interesting topics emerged for further investigation. Why does Johnson disagree with this interpretation of his dialectical tier? And what is the relation between the dialectical tier and the critical scrutiny function in argument? In this paper we would like to dig deeper on these issues. We will begin with explaining the critical view of argument, and then re-examine the above two points based on a careful reading of Johnson’s own views on the dialectical tier and manifest rationality. On that basis, we will then try to further explore the critical dimension within dialectical tier by bridging together the critical view of argument and Johnson’s theory of argument. After that, we conclude with some remarks on exploring the critical dimension within the study of argument.

2. Critical View of Argument
As preliminaries for the discussions in the remainder of this paper, we will start by making it clear what we mean by “critical”. By this term we want to refer to a view of argument, which claims that arguing for a thesis involves taking into account not simply the reasons in favor of it but also (some) reasons against it. To further articulate this critical view of argument, we will unpack it into three specific but related levels in our understanding of argument.

First, it is nearly superfluous to say that arguments need to take into account reasons in favor of the conclusion; and this has already long been well recognized in our understanding of argument. However, there has also been another strand which values arguments as taking into account of reasons against the conclusion. As Keith has observed, “only the participation of the other in resisting, contesting and challenging the claims” can make argument distinguished from persuasion (Keith 1995, p. 172). And Meiland put forward a similar idea in this way, “the fundamental idea behind all argumentation is this: a possible reason that survives serious objections is a good reason for accepting the belief in question” (Meiland 1981, p. 26). These ideas, as we understand, could be phrased more briefly as this: arguments are intrinsically or conceptually critical.

Second, besides the reasons in favor of the conclusion, why should we bother to take into account reasons against it? The most natural answers are, to improve the strength of argument, by testing and detecting possible flaws in our ways of arguing, or/and to make a better case for the thesis defended in the argument, by rejecting opposing points of view and by weighing and balancing positive and negative considerations. To cite Scott’s words, arguments “must be extended in testing, not only for consistency, but also toward completeness” (Scott 1987, p. 68). That is to say, more specifically, to function persuasively or to better achieve its pragmatic and practical goals, the act of arguing should involve a process of critical scrutiny to seek for the strongest or most appealing reasons and better argument. Hence, not only the criticisms and other forms of reasons against the conclusion “relate to the creation of argument and the being of argument” (Scott 1987, p. 70), but also the arguments themselves are generically and functionally required to be critical.

Third, it has also been long acknowledged that a key indicator of argument’s cogency is how well or adequately it can, or actually does, take into account of reasons against its conclusion. Toulmin has endorsed this idea when he claimed that “a sound argument, a well-grounded or firmly-backed claim, is one which will stand up to criticism” (Toulmin 1958, p. 8). So does Perelman when he makes it clear that “the strength of an argument depends…upon the objections; and upon the manner in which they can be refuted” (Perelman 1982, p. 140). The idea underlying these views is that arguments are normatively appreciated to be critical.

All these three points, that arguments are intrinsically or conceptually critical, generically and functionally required to be critical, and normatively appreciated to be critical, are the embodiment of the critical view of argument we are going to discuss in this paper. They are closely interrelated, but can be endorsed separately and differently by scholars in their diverse theories of argument. But, are we here just confusing, as many might be wondering, the famous distinction of argument as process and argument as product? We believe this is a fair but misleading question, but still some further clarifications are indispensable. Firstly, the critical view of argument we explained above is not some new conceptualization of argument, but, to some extent, a general and overall view or perspective, from which we could understand our practice of arguing by specifying or emphasizing some of its particular aspects or characteristics. In particular, the critical view of argument gives prominence to the critical scrutiny function of argument (i.e. through taking into account both reasons in favor of and reasons against its conclusion), and stresses some specific features related to this function (e.g. normatively appreciated to be critical). Given this clarification, we might say that a critical view of argument could be comparable to a rhetorical view of argument or a dialectical view of argument, which also focus on some particular function of argument and its related characteristics. Secondly, the distinction of argument as process/product is another, nevertheless quite different conceptual framework to understand our practice of arguing. It has a special focus on the different stages or phases of the production of our argument. Therefore, it is now easy to see that these are two distinct theoretical ways of analyzing argument. They are overlapped or interlaced framework since they are all about understanding our practice of arguing, but they could not be confused as the same. More specifically, the critical view of argument could be embodied in both the product and the process level of argument, as we have just spelt it out in this section.[ii]

3. How Critical is the Dialectical Tier?
After a general clarification of what we mean by “critical”, now we turn to Johnson’s original notion of dialectical tier. According to his pragmatic theory of argument, a complete or paradigmatic argument has an “illative core-dialectical tier” structure. Based on this new concept, an argument needs not only an illative core, in which the arguer puts forward the reasons that support the thesis in argument, but also a dialectical tier, in which the arguer anticipates and defends against existent or possible objections and deals with the alternative positions that are incompatible with or threatening to the establishment of the thesis (Johnson 2000, pp. 164-169). Moreover, within this dialectical tier, the arguer discharges his/her dialectical obligations and fulfills the requirement of manifest rationality, and thereby exhibits himself/herself as a competent practitioner of argumentation.

Considering that objections normally present challenges, difficulties or possible impediments to the argument’s achieving its purpose, and given that alternative positions usually bring the arguer some counter-considerations about his/her argument or conclusion, we can easily tend to understand both of them as materials negatively relevant to the argument, i.e. both of them function as reasons/considerations against the tenability of the conclusion in argument. Given this understanding, it will be so natural to link the dialectical tier with the critical view of argument. We can easily think that including the dialectical tier within an argument indeed shows that the conclusion is critically established, since it indicates so obviously our taking into account not only the reasons in favor of the thesis, but also (some) reasons against it.

This interpretation also appears to have some plausibility within Johnson’s own articulations in his theory. Firstly, Johnson claims explicitly that arguer must take account of objections and opposing points of view when constructing arguments (Johnson 1996b, p.107), holding that “they are not supererogatory efforts”, but some kind of “dialectical obligations” (Johnson 2000, p. 157). It is in this way that the need to discharge these obligations renders necessity to the presence of dialectical tier within the concept of argument, and consequently, arguments without dialectical tier are suggested to be regarded as “unfinished, incomplete” (Johnson 2000, p. 166). So it seems reasonable to say that Johnson has endorsed the view that arguments are intrinsically or conceptually critical. Secondly, Johnson also holds that “criticism and revision are both internal to the process of arguing. They are not externalities that may or may not happen…they are integral parts of the process of arguing” … and “[in the practice of argumentation] … the strength of the better reasoning, and that alone, has determined the outcome” (Johnson 2000, pp. 157-160). So it seems likewise to be the case that Johnson approves the idea that arguments themselves are generically and functionally required to be critical. Thirdly, Johnson also believes that “a controversial thesis can not be adequately supported if its supports failed to surpass its objections and alternative positions” (Johnson 1996b, p. 107), and sees the ability of an argument to withstand objections and criticisms as a crucial test of its real value, “the test of the argument is a strong objection, the stronger the objection, the better the test” (Johnson 2007b). From this we could find as well that Johnson is apt to accept the idea that arguments are normatively appreciated to be critical. Based on these observations, can we then conclude, as we expected, that the notion of dialectical tier indeed embodies or manifests the critical view of argument? Here Johnson’s own answer is a negative one, as already hinted in our introduction, “not sure of this.” We might wonder, however, for what reasons does he think it is not sure? And where and why do the dialectical tier and critical view of argument go apart?

To answer these questions, we need to further reveal another part of the story in Johnson’s theory. That is, Johnson indeed intends to require that “the arguer responds to all materials, if possible” (Johnson 2001). By “all materials”, he requires the arguer to deal with positive and neutral materials which are simply questions or which only aim at clarification or understanding (Johnson 2001). Moreover, he still claims that “the arguer must respond even to criticisms which he believes (or knows) are misguided” (Johnson 1996b, p.108), or he/she must respond to all those objections “the audience is known to harbor, whether reasonable or not” (Johnson & Blair 2006, p. xv). Besides, he also believes that the arguer is obliged to respond an objection “even though he might well be justified in not responding to it.” (Johnson 2001) Or, “we would expect to hear how an arguer handles a well-known objection, even if it is not likely to cast serious doubt on the cogency of the argument” (Johnson 2000, p. 333).

By revealing these possibilities that in dialectical tier the objections which are neutral, misguided, unreasonable or unlikely to affect the cogency of our argument will all be dealt with, we have to admit, contrary to our expectation, that Johnson’s notion of dialectical tier does not embody our critical view of arguments, neither does the presence of dialectical tier really indicate that the conclusion is critically established. Within the critical view, although we value the process of critical scrutiny intrinsic in argument and truly appreciate the constructive merits of reasons against the tenability of the thesis in question, only those reasons which are relevant to the establishment of our thesis or to the improvement of the strength of argument require and deserve our concern. In other words, we need to take account of reasons against our thesis, but we only do that subject to the purpose of seeking the strongest or most appealing reasons to make a better argument for our thesis in question. More specifically, the process of critical scrutiny only consists of weighing and balancing positive and negative reasons from which we can directly or indirectly gain improvements or revisions for our argument. However, materials like misguided, unreasonable objections and those which are unlikely to affect the cogency of our argument are essentially irrelevant, thus they have no constructive values with respect to the improvement of our argument. Therefore, within a truly critical view of argument, those materials do not require or deserve our concern to deal with in the process of critical scrutiny. Given this clarification, now it could be confirmed that, in spite of their prima facie similarities, Johnson’s notion of dialectical tier does not accord with the critical view of argument.

But, we may still wonder, why does Johnson intend to include responses to those materials as internal to the process of arguing, even though responding to them would bring no revisions and betterments to our argument? Moreover, the efforts of dealing with those materials would even possibly and easily turn out to be a risk of wasting arguer’s energy and cognitive resources, but why does Johnson still want to regard those efforts as obligatory but not supererogatory? To resolve these doubts, we need to further investigate Johnson’s understanding and justification of dialectical tier. And by probing into these issues, we can also better reveal the deep discrepancies between dialectical tier and critical view of argument.

4. Why the Dialectical Tier is not Critical?
The notion of dialectical tier, needless to say, is one of the most controversial topics in recent argumentation studies, “no other concept in the recent literature on argumentation has attracted so much notice” (Leff 2003). Among a lot of disputes surrounding it, Johnson took pains to clarify, revise and justify his own ideas. We will in this section investigate his justification of arguer’s dialectical obligations, which emerges as pivotal with respect to our current discussion. If there are no obligations incurred to the arguer to respond dialectical materials, there will obviously be no inclusion of dialectical tier within argument. And what kinds of materials are in need of response in such a tier will surely depend on why and how these obligations are incurred.

In the development of his theory, Johnson used different strategies to justify dialectical obligations. At the very beginning, dialectical obligations come from the requirement of sufficiency. It is initiated from the consideration of, or the need of “defending your (own) argument” when constructing arguments (Johnson & Blair 1983, p. 195). Later it is required explicitly as “obligations” when he (and Blair) started to treat argumentation as dialectical (Johnson 1996b, p. 100). As he puts it, it is an “aspect of sufficiency” which makes arguer obliged to include defenses against actual and possible objections. Otherwise an argument will not only fail to be a good one because of being “in violation of the sufficiency requirement”, but will also be regarded as “incomplete” (Johnson 1996b, p. 100).

However, a few years later, Johnson proposed “dictates of rationality” as a related but slightly different justification for dialectical obligations. “If the arguer really wishes to persuade the other rationally, the arguer is obliged to take account of these objections, these opposing points of view, these criticisms”, and “if she does not deal with the objections and criticisms, then to that degree her argument is not going to satisfy the dictates of rationality” (Johnson 1996a, p. 354). But what kind of rationality is coming to dictate? Johnson believes that a “bare-bones specification of rationality” will be adequate and could allow him to develop his own theory of argument. It is “the disposition to, and the action of using, giving and-or acting on the basis of reasons” (Johnson 2000, p. 161). Based on this understanding, rational arguers are those who have “the ability to engage in the practice of giving and receiving reason” (Johnson 2000, p. 14). Accordingly, following the dictate of this rationality, arguers are required, obviously and naturally, to give (good) reasons, and only use (good) reasons, to justify or defend their thesis in the practice of argumentation. However, but why do we still need to, and even be obliged to, consider negative reasons, or to deal with objections and alternatives? Considering and dealing with them are obviously not only efforts of giving and using reasons, but indeed efforts of providing more reasons, and efforts of giving and weighing different kinds of reasons?

The answer was finally given, a few years later, when Johnson realized that “the idea of rationality alone cannot illuminate the practice of argumentation” (Johnson 1996b, p. 114). He then started to construct his new idea of argument as manifest rationality, with which he tried to characterize argument as “patently and openly rational”. More specifically, “it would not only be rational, it must also be seen to be/appear rational”. By this characterization, he claimed that “participants in the practice of argumentation not only exercise their rationality but they need to be seen to be so doing” (Johnson 2000, p. 164). Furthermore, arguers are required to care about both “the inner reality and the outward appearance” of argument, and to “exhibit what it is to be rational” in a way of “to give reasons; to weigh objections; to revise over them or to reject them”, because “all of this describes a vintage performance of rationality” (Johnson 2000, p. 163). This articulation of manifest rationality provides Johnson a better way to justify dialectical obligations: “if the arguer were obligated only by the dictates of rationality (rather than those of manifest rationality), then one might ignore criticism” (Johnson 1995, p. 260). Finally, “manifest rationality is why the arguer is obligated to respond to objections and criticisms from others, and not ignore them or sweep them under the carpet,” because otherwise “it would not only not be rational; it would not look rational” (Johnson 2000, p. 164, italics original), and it would become “in most contexts, a failure not just of rationality but to make that rationality manifest” (Johnson 2007a).

At the end of this brief detour on Johnson’s justifications of dialectical obligation, we come to the finding that his ultimate explanation and justification of dialectical obligation rely on his characterization of manifest rationality. It is by this requirement that we can incur dialectical obligations upon arguer and explain the necessity of dialectical tier. Moreover, based on these observations, our primary issue of probing into the discrepancies between dialectical tier and critical view of argument could also be better illuminated now. As Johnson has made it clearly, “The constraint I call manifest rationality requires that the arguer respond to all material, if possible. If there is an objection and the arguer doesn’t respond to it, then even though he might well be justified in not responding to it, the argument will not have the appearance of rationality” (Johnson 2001). And it is “from the perspective of the requirement of manifest rationality, the arguer is obliged to respond even to criticisms that are regarded as misguided, because to ignore such criticisms compromises the appearance of rationality” (Johnson 2000, p. 270).

As indicated by Kauffeld, Johnson has assigned, by his characterization of manifest rationality, “the priority to rationality as the primary internal good realized through the activity of argumentation”, thus “supposes a priori that argumentation is governed by an overriding commitment to rationality which identifies its practitioners and dictates their probative obligations” (Kauffeld 2007). We agree with this analysis, and will further demonstrate that this is where Johnson’s dialectical tier and the critical view of argument start to diverge from each other. In our practice of argumentation, while the critical view of argument assigns the uppermost importance to the seeking for the strongest or most appealing reasons and better argument, Johnson gives priority to making manifest our rationality over the improvement of argument quality. In his theory, arguing means not only to persuade the other, but to “rationally persuade the other”. And “rationally persuade the other” requires not only the arguer to use good reasons or better argument, but at same time, to cherish rationality and to increase the amount of rationality in the whole world. In other words, in the process of arguing there is a more important underlying presumption that “the arguer and the critic have each exercised reasoning powers” (Johnson 2000, p. 162, italics added). Accordingly, it is by this reason we can better understand Johnson’s inclusion within dialectical tier of the responses to materials which are not directed or relevant to the betterment of argument. Because, “if the critic’s objections have been found wanting, then the arguer will have to exercise his reasoning powers to show this…” (Johnson 2000, p. 162). As a result, a judgment that an objection is misguided may have been well established, from which not only the arguer makes his rationality/reasoning power exercised and manifest, but also that his critic will learn something thereby improve his own rationality/reasoning power. As Johnson has envisaged, “if it turns out that criticism is easily responded to, then the critic will have learned that the criticism was not so good”, or “the respondent realized that the point of her criticism is not able to devastate the opponent, nor yet the argument” (Johnson 2001). And in this way, more importantly, “the participants are more rational and the amount of rationality has increased,” and in the end “the world becomes a slightly more rational place” (Johnson 2000, p. 162).

5. Why not Make Dialectical Tier Critical?
Although manifest rationality counts as the most essential groundwork for Johnson’s articulation and justification of dialectical tier, many argumentation theorists, strangely and interestingly, are apt to discuss Johnson’s notion of dialectical tier while brushing aside his idea of manifest rationality. Given that this idea actually explains where and why his theory and the critical view part their company, in this section we intend to scrutinize it further with a critical eye.

Johnson’s justification of dialectical obligation by conceiving argument as manifest rationality is unique and theoretically coherent. If argument is really an exercise of manifest rationality which requires its participants to make their own rationality manifest and improved, and to make the amount of our rationality increased, then a dialectical tier is undoubtedly needed for our conceptualization of argument. And so do the arguers have obligations to respond all materials where there is any possibility to get our rationality manifest, exercised and increased. However, in order to make this line of justification more persuasive and adequate, we think still more developments or even revisions are needed. To achieve this goal, we will try to bridge and integrate the dialectical tier with the critical view of argument, in the following two respects.

The first aspect on which we want to cast our doubt is concerning the rationality at play in Johnson’s theory. Johnson understood rationality as “the disposition to, or ability of using, giving, and-or acting on the basis of reasons”, and accordingly “to be rational means to be able to engage in the giving and receiving of reasons” (Johnson 2000, p. 14). On that basis, argumentation is seen as an exercise of manifest rationality which is valued by its virtu of embracing, cherishing, increasing and exhibiting rationality (Johnson 2000, pp. 162-3). However, when we take into considerations of dialectical materials, no matter they are relevant or irrelevant, and no matter whether dealing with them leads ultimately to the revision of argument or to the exercises or improvements of someone’s reasoning power, but do these efforts really make manifest the above kind of rationality and result in our being more rational in Johnson’s sense? We suspect that by dealing with dialectical materials we do much more than that. First, as Ohler has observed, in considering criticisms and objections we are actually “putting more reasons at play” (Ohler 2003). Second, we would like to add, in responding them and revising our argument accordingly, we are not only making manifest our ability to give and receive reasons, but are also exhibiting our ability to weigh, compare and balance among different reasons. In other words, if there is some kind of rationality that has been embraced, cherished and exhibited in argumentation, it is definitely not just the disposition or ability of using, giving and acting on the basis of reasons. Therefore, for better capturing the reality with Johnson’s notion of manifest rationality, we propose a richer sense of rationality as “the disposition or ability to be responsive to reasons”. By this term we want to refer to a sense of rationality which is much more complex than Johnson’s bare-bones specification of giving and receiving reasons. It will further include those abilities of quantifying reasons, of measuring the quality/force of reasons, and of regulating the interaction among reasons. These are, in our view, what we really manifested in our dealing with dialectical materials. In particular, when arguers are required to make manifest their ability to be responsive to reasons, it is obvious that they will firstly be responsive to the varieties of reasons. This means they will not only provide reasons of their own, but also take into account reasons from the others (i.e. taking account of dialectical materials), and specifically, they will have to consider both reasons in favor of and against the thesis (i.e. dealing with objections/criticisms/ alternatives). Furthermore, they will in this process also be responsive to the quality/force of reasons, and to the interaction between different reasons. This means they will be able to weigh the force of different reasons, to value them differently with respect to strength, and to accept, improve or reject them accordingly, and at last, to balance among these reasons thereby to find the strongest or most appealing reasons and better arguments (i.e. revising and improving his arguments).

The second aspect we think in need of further development is related to the normative requirements generated by manifest rationality. Johnson has made great efforts, in his recent works, to specify what the arguer is actually obliged to do in order to make rationality manifest, i.e. the Specification Problem, and to resolve in which way we can judge that the arguer has adequately fulfilled these obligations, i.e. the Dialectical Excellence problem. Undoubtedly, Johnson’s exploration on these issues is profound and elaborate, and his achievements are valuable. But a meticulous critic can still find something wanting in his solutions.

Firstly, Johnson ignores, to our understanding, the exploration on a more general normative aspect with respect to the ways of fulfilling the requirement of manifest rationality. That is, in argumentation it is required that rationality “must be seen to be done”, but can we do that in an unreasonable/irrational way? Or, what is the right/acceptable way of making rationality manifest? This is a reasonable question. It was well hinted by Ohler’s accusation that responding to criticisms which are believed or known to be misguided is “in one important sense of the word quite irrational” (Ohler 2003). And it was also perfectly embodied in van Eemeren’s suspicion that we can even try to fulfill the requirement of manifest rationality “by arguing in what Perelman calls a ‘quasi-logical’, and sometimes fallacious way…[or] there may be techniques of purporting to deal with all criticisms while responding in fact only to those that are most easy to answer. You can pretend to deal with all objections without actually treating them satisfactorily” (van Eemeren 2001). However, Johnson does not propose any general norms governing our ways of making manifest our rationality. We believe that this is an important issue in need of development, and more importantly, it is where many others started to misunderstand manifest rationality and thereby interpreted or criticized it as a rhetorical requirement (Hansen 2002, van Eemeren 2001). We will not here accuse Johnson of not providing any ideas on this issue, for you can find some relevant basic ideas underlying his recent discussions, that is, “firstly, choose the right dialectical materials, and then deal with them in an adequate way”. Nevertheless, what could count as right and adequate still leaves room for different interpretations and misunderstandings. And we suspect that, based on his own articulation of manifest rationality, he is not able to exclude those misunderstandings. If rationality only means an ability to engage in giving and receiving reasons, then an elaborate selection and arrangement of responses to some insignificant, unimportant or easy criticisms can still exhibit our being able to give and receive reasons. Moreover, when we use some sort of techniques to successfully pretend that we’ve already satisfactorily dealt with all objections, it likewise has the same effect that our arguing appears to be rational, or that our rationality is seen to be done. The point we want to indicate here is that there is a lack of normative constraints intrinsic to the ways of fulfilling the requirement of manifest rationality. And we believe a solution can be found if we bring closer Johnson’s dialectical tier and manifest rationality with the critical view of argument. If we could understand manifest rationality as essentially a similar requirement of seeking most appealing reasons or better argument, we will have to recognize that the process of arguing is not simply an accumulation of different or more reasons, nor is it a subtle construction by dealing with materials which are faked, deceitful or not genuine with respect to our thesis or argument. In other words, through such integration we can understand dialectical tier as an embodiment of critical scrutiny function, which will intrinsically set out some inner constraints for its process, and in this way it will help us to build up a better and clearly fixed fence that could keep many misunderstandings and interpretations away.

Secondly, Johnson’s specification of arguer’s dialectical obligations seems to be problematic. In his recent works, Johnson wants to develop a specific method which does not “all depend on context” (Johnson 2007a) to determine the arguer’s dialectical obligations. To realize this goal, he seems to have set out two principles. On the one hand, it appears that he endorses the principle that arguer’s obligations are incurred with respect to the dialectical material’s capacity/strength to undermine the argument. This principle is incarnated in his ways of prioritizing dialectical materials (Johnson 2001) and in his ways of unpacking of “The Standard Objections” by the criteria of “proximity/strength/salience” (Johnson 2007a). The underlying motivation for this principle, obviously, is to detect the strength or viability of argument and thus to revise it for a better one. On the other hand, he also seems to endorse another principle that arguer’s obligations are incurred with respect to the requirement of making rationality manifest. This principle is perfectly embodied in his requirement of dealing with neutral, positive, misguided and unreasonable materials, which might only be request for clarifications or with no effect on weakening or threatening the argument. The motivation for this principle, as we have already indicated, is to make our rationality exercised, exhibited and improved. But can these two principles be well integrated together in his theory? By requiring the responses to neutral, positive and misguided materials, it appears that Johnson explicitly makes the latter principle outweigh the former, for those materials are obviously not qualified as having any effect on detecting the strength and viability of the arguer’s argument. Nevertheless, by claiming that “the intervention of the other is seen to lead to the improvement of the product…a better argument, a more rational product” (Johnson 2000, p. 161, italics added), and by explaining the reason for dealing with neutral and positive materials as that “still there are times when this material will result in the arguer’s having to modify or clarify the argument, which will result in its being a better argument” (Johnson 2001, italics added), it seems that he has also reversely put his first principle over the second.

Despite this vague combination of two principles, many scholars also regard his second principle as misleading or harmful. As Adler has complained, it is “imposing excessively burdensome costs on arguers” (Adler 2004, p. 281). Similar to them, we are not well persuaded on this principle either. Firstly, we also have suspicions about the meaning or uses of dealing with misguided and unreasonable materials in the process of arguing. Secondly, even if dealing with neutral and positive materials can possibly result in our argument’s being a better one, it is definitely not what usually and frequently happens in reality. And here again we want to urge an integration of Johnson’s theory with critical view of argument. In doing so, we will suggest a hierarchy for those two principles, and thereby to eliminate the vagueness and to revise his specification of dialectical obligations. Within a critical view, an argument is an embodiment of the process of critical scrutiny for the truth/acceptability of a thesis, thus we will take the arguer’s foremost concern to be the seeking for most appealing reasons and better arguments. With this in mind, we should elevate the first principle over the second. That is, we should incur dialectical obligations only with respect to their relevance and capacity to our potential revisions or improvements of argument. Accordingly, we would like to suggest that we narrow down the scope and contents of dialectical obligations, by discarding those irrelevant, unreasonable, misguided or neutral materials, with which if we deal we can only make some clarifications or corrections of the others, or can only make our rationality exhibited, exercised. In other words, by the act of arguing we rationally and critically justify our thesis, and in the same process at the same time, we also make our ability manifest and the whole world more rational. But this manifest aspect comes as spontaneous and secondary. We do not need to intentionally exercise or perform that ability by taking every chances or possibilities, especially when some of them will bring no benefits with regard to our argument under consideration, and some of them will sometimes even result in detours and hindrance. To be brief, the search for more appealing reasons or better argument should outweigh the exhibition of our ability to reason and argue, and the justification of one thesis in question should surpass the desire of manifestness of our rationality. Actually, realizing the former will simultaneously realize the latter, while asymmetrically, realizing the latter will normally and easily go beyond the scope of the former.

6. Conclusion: Exploring the Critical Dimension within Study of Argument
In this paper we started with a curiosity to probe into the relationship between dialectical tier and the critical scrutiny function in argument. By a careful reading of Johnson’s theory, we disconfirmed our conjecture that the inclusion of a dialectical tier in argument means the thesis is critically established. However, we also urged to bridge Johnson’s theory and the critical view of argument, and thereby to make dialectical tier critical in nature. It is, to some extent, a promising proposal for the improvements of his theory, as well as the resolutions of some theoretical problems.

Based on this case study of Johnson’s theory, we will conclude this paper by claiming that the critical view of argument is important and promising, and more serious and thorough study should be done on the critical dimension within our study of argument. Here by critical dimension we refer to those theoretical aspects that are developed from the endorsement of critical view of argument. What are those aspects? And what are the issues which will emerge as pivotal in those aspects? Bringing forward a comprehensive framework will go beyond the limit of this paper, here we can only sort out some important theoretical questions which deserve our better reflections.

What are the underlying assumptions or justifications of the critical view of argument? The arguers as fallible? Or/and with a fallibilist attitude? Or/and every thesis is fallible? Or/and every argument is vulnerable? What are their implications in our theories of argument?
How is the function of critical scrutiny performed? What is the mechanism underlying the interaction of different reasons, especially, between reasons for and against (such as in conductive argument)? By what principles or methods can we judge some of them outweigh the other?
How is the critical dimension embedded differently in different theories of argument? For what reasons? How can we use critical dimension as a better perspective to further indicate their theoretical divergences, and also to better bridge them?

NOTES
[i] The work in this paper is supported by the Chinese MOE Project of Key Research Institute of Humanities and Social Sciences at Universities (2009JJD720022), and by the Chinese MOE Project of Humanities and Social Sciences (10YJC72040003).
[ii] Thanks to two anonymous reviewers for their helpful comments and suggestions for our revisions here. And we would also like to respond to their criticisms with regard to our ‘mischaracterizing’ of Johnson’s theory. That is, it seems that we are using our process-oriented understanding of argument to unjustly misinterpret and criticize Johnson’s ideas, which are, as he himself has clearly claimed, product-oriented. Our responses will consist of the following three points. First, as we have already clarified in the second section, the critical view of argument which we proposed and articulated in this paper is not process-oriented or a process-conception of argument, but a general view or perspective of argument which could be embodied in both the process and product of argument. Second, although Johnson’s view of argument, generally speaking, could be regarded as product-focused, we believe that his theory has also clearly and inevitably involved “an appreciation of argument as a process” (Johnson 2000, p. xi). Even though Johnson himself does endorse explicitly the argument as process/product distinction, and claimed that his theory is product-oriented, his discussions of many issues within his theory are still falling back on the process level of argument. For example, his articulation of manifest rationality is unpacked into the process of arguing, and his resolution to the fundamental Specification Problem (of dialectical obligations) is based on a division of the process of arguing into “phases of constructing argument and revising argument” (Johnson 2001). Third, in this paper the topics we mainly discussed are the idea of rationality, the justification of dialectical obligation, and the normative requirements of incurring dialectical obligations, none of them are restricted to the product or process level of argument. Neither do our comparison and integration between Johnson’s ideas on dialectical tier and the critical view of argument. Therefore, given the above three points, it is now easy to see that it might be inappropriate or misleading to consider the merits and arguments in this paper using the framework of argument as product/process distinction, since it is a different or, to some extent, irrelevant framework. And it is not really the case that we are just reading Johnson’s argument-as-product ideas from an argument-as-process view

REFERENCES
Adler, J. E. (2004). Shedding Dialectical Tiers: A Social-Epistemic View. Argumentation 18, 279-293.Eemeren, F. H.van, (2001). Commentary on Johnson’s More on Arguers and Their Dialectical Obligations. in H. V. Hansen, C. W. Tindale, J. A. Blair and R. H. Johnson (Eds.), Argumentation and Its Applications, Proceedings of the 4th OSSA Conference, Ontario: CD-ROM.
Hansen, H. (2002). An Exploration of Johnson’s Sense of “Argument”, Argumentation 16, 263-276.
Johnson, R. H. (1995). The principle of vulnerability. Informal Logic 17, 259-269.
Johnson, R. H. (1996a). The Need for a Dialectical Tier in Arguments, in Lecture Notes in Computer Science, vol.1085. Proceedings of the International Conference on Formal and Applied Practical Reasoning (pp. 349-360), London: Springer-Verlag.
Johnson, R. H. (1996b). The Rise of Informal Logic. Newport News: Vale Press.
Johnson, R. H. (2000). Manifest Rationality: A Pragmatic Theory of Argument. Mahwah: Lawrence Erlbaum Associates.
Johnson, R. H. (2001). More on Arguers and Their Dialectical Obligations. In H. V. Hansen, C. W. Tindale, J. A. Blair and R. H. Johnson (Eds.), Argumentation and Its Applications, Proceedings of the 4th OSSA Conference, Ontario: CD-ROM.
Johnson, R. H. (2007a). Anticipating Objections as a Way of Coping With Dissensus, in H.V. Hansen, C.W. Tindale, J.A. Blair, R.H. Johnson, D.M. Godden (Eds.). Dissensus and the Search for Common Ground. Proceedings of the 7th OSSA Conference, University of Windsor. CD-ROM.
Johnson, R. H. (2007b). Responding to Objections. In F. H. van Eemeren, J. A. Blair, C. A. Willard & B. Garssen (eds.), Proceedings of the Sixth conference of the International Society for the Study of Argumentation (pp.717-722), Amsterdam: Sic Sat.
Johnson, R. H. & Blair, J. A. (1983). Logical Self-defense, Toronto: McGraw-Hill Ryerson.
Johnson, R. H. & Blair, J. A. (2006). Logical Self-defense. New York: Idea Press.
Kauffeld, F. (2007). Two views of the Necessity to Manifest Rationality in Argumentation. In H.V. Hansen, C.W. Tindale, J.A. Blair, R.H. Johnson, D.M. Godden (Eds.). Dissensus and the Search for Common Ground. Proceedings of the 7th OSSA Conference, University of Windsor. CD-ROM.
Keith, W. (1995). Argument Practices. Argumentation 9, 163-179.
Leff,  M. (2003). Commentary on A. Ohler’s ‘Rationality in Argumentation’. In J. A. Blair, D. Farr, H. V. Hansen, R. H. Johnson & C. W. Tindale (eds.), Informal Logic @ 25: Proceedings of the Windsor Conference. Windsor, Ontario: CD-ROM.
Meiland, J. (1981). College Thinking: How to Get the Best out of College. New York: The New American Library.
Ohler, A. (2003). Rationality in Argumentation. In J. A. Blair, D. Farr, H. V. Hansen, R. H. Johnson & C. W. Tindale (eds.), Informal Logic @ 25: Proceedings of the Windsor Conference. Windsor, Ontario: CD-ROM.
O’Keefe, D. (1977). Two Concepts of Argument. Journal of the American Forensic Association 13, 121-128.
Perelman, C. (1982). The Realm of Rhetoric. Notre Dame: University of Notre Dame Press.
Scott, R. (1987). Argument as a Critical Art. Argumentation 1, 57-71.
Toulmin, S. E. (1958). The Uses of Argument. Cambridge: Cambridge University Press.
Xie, Y. (2008). Dialectic within Pragma-Dialectics and Informal Logic, in Takeshi Suzuki, Takayuki Kato & Aya Kubota (Eds.), Argumentation, the Law and Justice: Proceedings of the 3rd Tokyo Conference on Argumentation (pp.280-286), Tokyo: Japan Debate Association.