Abstract: It is often said that images are excellent persuasive means. However, if images are persuasive, can they also be argumentative? After discussing authors who have tried to fill the gap between rhetoric and argumentation (Perelman and Olbrechts-Tyteca, Reboul, Bonhomme), I will argue that the same figures or tropes can have both a persuasive and an argumentative function.
Keywords: metonymy, persuasion, visual argumentation, visual rhetoric.

1. Introduction
The relationship between visual rhetoric and visual argumentation is a topic to which several essays have been dedicated. Some scholars deal with it in a general way (Blair, 2004; Kjeldsen 2012). Others focus on figures or tropes in particular (for antithesis, van Belle, 2009). Indeed, it has becoming a sub-field in the domain of visual argumentation. That said, the way in which visual rhetoric and visual argumentation have been related is not completely satisfactory. I will try to show that most attempts to link rhetoric and argumentation are based on the assumption that figures of rhetoric are above all persuasive. This assumption has a dramatic consequence upon visual argumentation, specifically because one argument against visual argumentation is that images are merely persuasive. As a result, considering visual rhetoric as persuasive would not reinforce visual argumentation, but rather critiques against it. Furthermore, another critique must be taken into account: in the frequent case of mixed media, i.e. when an argument is displayed in both words and images (such as in ads or commercials), the text alone is supposed to be argumentative, while the image would be merely persuasive (Adam & Bonhomme, 2005, p. 194 & 217).

So, in the first part of this paper, I will examine some of the principal ways figures of rhetoric and argumentation have been related in order to determine the extent to which figures have been considered as arguments. Then, in the second part, I will argue that some figures of rhetoric can be persuasive and argumentative at the same time.
Simply stated, I am interested in the argumentativity of figures. In saying this, I am using a French concept (argumentativité) that was coined by Ducrot and is used in the French theory of argumentation in order to refer to figures (Bonhomme, 2009; Plantin, 2009). This concept essentially suggests that an utterance can have an argumentative value instead of being limited to providing merely informational value (Anscombre et Ducrot, 1986, p. 91). Such an argumentative value comes from the fact that we can find, in an enunciate, elements that allow for a given conclusion by way of a commonplace, which Ducrot calls a topos (Ducrot, 1992). However, this concept is used in a slightly different way when applied to figures: in this case, it refers to their argumentative value, which can be considered as persuasive or argumentative, in this case when figures provide reasons to support a claim. Note that in what follows, I use the adjective “argumentative” with this restrictive meaning, unlike those who use it in a broader way, i.e. including all mean of influencing the addressee.[i]

Yet, why is the issue of the argumentativity of figures so important? Simply put, if figures are considered to mainly have a persuasive role, it is hardly possible to see them as arguments, at least for those who believe argumentation and persuasion are mutually exclusive (Plantin 2012; Doury 2012; Micheli 2012).

It is generally accepted that persuasion is an important feature of images (Scott & Batra, 2003). It seems even that the syntagm “visual persuasion” is almost pleonastic since the supposed “essence” of image is closely related to persuasion (Hill, 2004). The problem, however, is that this understanding of images as persuasive does not have a positive connotation, as it is very often linked to propaganda. Propaganda and persuasion are indeed often seen as techniques for manipulating (Jowell & O’Donnel, 1992 ; Pratkanis & Aronson, 2001; Spangenburg & Moser, 2002), in particular regarding political posters (Seidman 2008) as well as advertising (Messaris 1997). This shows that we must be very careful when dealing with issues of visual persuasion. As we will see, this is all the more the case because figures of rhetoric are usually considered as persuasive, at least in French scholarship.

2. Figures of rhetoric and arguments
2.1 Perelman and Olbrechts-Tyteca
Perelman and Olbrechts-Tyteca are amongst the first to have drawn our attention to the relationship between figures of rhetoric and argumentation. These scholars were indeed interested in “showing why and how the use of certain figures of rhetoric can be explained by the need for argumentation” (Perelman & Olbrechts-Tyteca, 1970, p. 227). At its core, their theory aims to call into question the old understanding of figures of rhetoric as pure ornament, i.e. without any other function than “embellishment”. This would explain their need to distinguish between times when a figure is purely ornamental, and those when it may play a part in an argumentative process. For this reason, they consider “a figure to be argumentative if it brings about a change of perspective, and its use seems normal in relation to its new situation. If, on the other hand, the speech does not bring about the adherence of the hearer […], the figure will be considered an embellishment, a figure of style” (Perelman & Olbrechts-Tyteca, 1970, p. 229; authors’ emphasis).

To be sure, the idea of considering figures from an argumentative standpoint was an important step forward for the field. However, it is insufficient to say that a figure is argumentative simply if it is accepted. Insofar as Perelman and Olbrechts-Tyteca consider that “the same figure, recognizable from its structure, doesn’t necessarily produce the same argumentative effect” (Perelman & Olbrechts-Tyteca, 1970, p. 232), they proposed their own classification of figures, aimed at emphasizing how figures can help argumentation. They organized figures into three categories: choice, presence, and communion. Indeed, this classification has the purpose of showing that “the effect, or one of the effects, certain figures have in the presentation of data is to impose or suggest a choice, to increase the impression of presence, or to bring about communion with the audience” (Perelman & Olbrechts-Tyteca, 1970, p. 232-233).

From this point of view, figures are considered as argumentative if they increase the adherence of the audience, which is a consequence of the concept of argumentation developed by Perelman and Olbrechts-Tyteca, i.e. a concept aimed toward influencing a given audience (Plantin, 1990, p. 16).

2.1.1 Hypotyposis
Interestingly, the first example of a figure they give is hypotyposis. This figure has a lot to do with images. According to Fontanier, for instance, “Hypotyposis paints things in a such a lively and dynamic way that it puts them, so to say, in front of our eyes and turns a narrative or a description into an image, a painting, a tableau vivant” (Fontanier, 1968, p. 390). They comment on this figure by writing: “It is therefore a way of describing events that make them present to our conscience. Could we negate the eminent part it plays as a factor of persuasion?” (Perelman & Olbrechts-Tyteca, 1970, p. 226). And they added: “If we neglect this argumentative role played by figures, their study will quickly be a vain hobby”.

We can see in this quotation that hypotyposis is considered as a factor of persuasion. In turn, persuasion is assimilated to the argumentative role played by figures. The aim of the chapter on the relationship between figures of rhetoric and argumentation is indeed “to resituate argumentation figures in their proper place concerning the phenomenon of persuasion” (Perelman & Olbrechts-Tyteca, 1970, p. 231). Such a conception is not surprising, given that Perelman aims to reconcile rhetoric and argumentation. But it has, however, important consequences. From my point of view, playing a persuasive role is not enough to warrant seeing a figure as argumentative. If hypotyposis is eminently visual, we need to be sure that, beyond its effectiveness, it is also argumentative.
Yet within Perelman and Olbrechts-Tyteca’s own classification of figures, hypotyposis belongs to the category of figures of presence. Besides hypotyposis, other figures belong to the same category: ekphrasis and energeia, among others, since they have the same purpose: namely, to make the object of the discourse present (Perelman & Olbrechts-Tyteca, 1970, p. 235). However, once again, if such a figure is highly persuasive and contributes to the effectiveness of the discourse, is it also argumentative? I am not sure it is.

As we know, presence is very often visual. A well-known example of energeia – that Perelman and Olbrechts-Tyteca use when dealing with “presence” – is that of Caesar’s bloody tunic. This is a classic example that illustrates the use of concrete objects to move the audience (Perelman & Olbrechts-Tyteca, 1970, p. 157). Once more, I wonder whether such a very persuasive device can be considered as argumentative, since it is explicitly intended to move the audience through an appeal to pity. Aristotle described energeia as vividness, liveliness, “bringing-before-the-eyes”, (Rhetoric 1411b 24), but also limited its use and that of similar figures in so far as “it is not right to pervert the judge by moving him to anger or envy or pity” (Aristotle, Rhetoric 1354a 24-26). Unlike Cicero, Quintilian also wished to limit its use in courts (Quintilian, Institutio Oratoria, VI, 2, 1).

2.1.2 Phryne
A famous example of a similar rhetoric device is that of Phryne, a Greek courtesan known for her beauty. It has been said that Praxiteles used her as a model for his famous Aphrodite of Knide. She is also known for the legendary trial in which she was probably charged with impiety. According to some of the sources, such as Quintilian (Institutio Oratoria, II, 15, 9), the trail had a surprising turn of events. Just when it seemed that the verdict would be condemnation, her lawyer, Hypereides, (who was also, by the way, one of her lovers), removed Phryne’s robe and bore her breasts before the judges. Awe-struck by her beauty, and undoubtedly impressed with a sense of pity, they acquitted her.

The anecdote soon became a topos used to illustrate the persuasive power of rhetoric in Greek and Latin rhetoric treatises (Vouilloux, 1995, p. 102 & 109). It also illustrates quite well an appeal to pity based on sight (Lévy & Pernot, 1997, p. 6). For this reason, it is known to have inspired painters, like Baudouin and Gérôme (fig. 1).

Fig. 1. Gérôme, Phryné devant l‘Aéropage, 1861.

Not surprisingly, Gérôme’s painting has been used as an illustration in books on rhetoric and persuasion (fig. 2 & 3).

This shows again that we must be very careful when dealing with visual rhetoric and its relationship to argumentation. Hypotyposis and energeia belong, as we said, to figures of presence according to Perelman and Olbrechts-Tyteca’s classification. However, increasing the feeling of presence (Perelman & Olbrechts-Tyteca, 1970, p. 236) is not necessarily an argumentative tool. To round up the story about Phryne, it’s worth noting that after her acquittal, Athens published an official decree forbidding the use of the “appeal to pity” figure, in particular by exposing an accused individual to the judges (Lévy & Pernot, 1997, p. 6).

Fig. 2. Gérôme’s painting illustrating a book
Fig. 3. Gérôme’s painting illustrating a book

Once again, why is presence so effective? It must be said that the word “presence” is rather deceiving in this usage. For Perelman, it is important because it makes something more present and “enhance[s] the value of some of the elements of which one has actually been made conscious” (Perelman & Olbrechts-Tyteca, 1970, p. 235). This is true. However, why is visual presence so effective? One way of understanding this effect is that presence is evident, or even self-evident. It should be noted that the effect of presence can also be rendered by another rhetorical tool, enargeia, sometimes confused with energeia[ii]. Interestingly, when Cicero translated enargeia from Greek, he decided to invent a new word, instead of using adjectives available in Latin like clarus or perspicuus. As we know, the term created is “evidencia” (Lévy and Pernot, 1997, p. 10), based on videre, to see. Ironically, Perelman and Olbrechts-Tyteca – who have renewed the field of argumentation by explicitly rejecting the Cartesian concept of “évidence” (Perelman & Olbrechts-Tyteca, 1970, p. 4) – take for granted the argumentative value of presence as enargeia or evidencia!

The same holds true for another category of figures that, according to Perelman and Olbrechts-Tyteca, plays an argumentative role: that of communion. Its purpose is to create or confirm communion with the audience. Again, this is a very persuasive means. Also, from these examples, it should be clear that, for Perelman, the argumentativity of figures corresponds to their persuasiveness. Furthermore, Charles Hill, in his essay on the psychology of rhetorical images, shows that “vividness is almost a direct synonym of visualization”, and that “vividness enhances persuasiveness”, so that “vividness, emotional response and persuasion have all been shown to correlate to each other” (Hill, 2004, p. 32). So, even if presence is one of the four major rhetorical qualities of images – and is therefore crucial for visual argumentation (Kjeldsen, 2012, p. 240) – one can still wonder whether it is argumentative or persuasive. The problem, here, arises from Perelman’s understanding of argument as aiming to provoke or increase the adherence of the audience. Yet such an understanding doesn’t make it easy to distinguish between argumentative means (i.e. giving reasons to support a point of view) and non-argumentative means. Indeed, not all means used to influence an audience can be considered as argumentative. For this reason, it seems to me that it is not enough for visual argumentation to rely on The New Rhetoric to found the argumentativity of figures.

2.2 Reboul and Bonhomme
The same position has been adopted by some of Perelman and Olbrechts-Tyteca’s followers. As is often the case, followers have a tendency to exaggerate when they adopt a systematic idea, i.e. in this case, considering that all the figures can be understood as argumentative. For example, in his book Introduction à la rhétorique, Olivier Reboul dedicates a chapter to the argumentative role that the figures of rhetoric may play. When Reboul writes about the “argumentative strength” of a figure, he is above all referring to its persuasive force. For him, a figure is rhetorical only “to the extent that it contributes to persuading” (Reboul 1991, p. 121). Hence the fact that his chapter includes figures – like rhythm – that are based on the sound of the words. This is not surprising, given his objective. As he puts it, “the rhythm produces a feeling of obviousness able to satisfy the mind, but also to enroll it” (Reboul 1991, p. 124). Indeed, how would it be possible to claim that all figures can be argumentative? Only from a broadened understanding of argumentation associated to persuasion, but also to pleasure. According to Reboul, Perelman’s theory on the relationship between figures and argumentation “is too intellectualist, too oblivious of the figure pleasure, a pleasure deriving either from emotion or from comic, but always from pathos” (Reboul, 1991, p. 122).

Another interesting case in point is found in Marc Bonhomme. At the end of his book Les figures clés du discours, a few pages are dedicated to “argumentation through figures”, in which he posits that besides their aesthetic function, figures also have “a practical end oriented toward the productivity of utterances. In this case, figures are seen as argumentation tools, influencing the opinions of their addressees and stimulating their adherence to the discourse that has been produced. More precisely, they work like persuasive speech acts playing with reasoning (to persuade), but above all on the affects (to hit)” (Bonhomme 1998, p. 88). Such an understanding is again very close to that of Perelman.

This same author developed this issue in a paper focused on the argumentativity of figures. In the introduction, he explains that, for him, there are three ways of understanding the relationship between rhetoric and argumentation. The first one is convergence: an argumentative discourse is considered to be rhetorical if its aim is to persuade. The second is differentiation: from this point of view, a discourse can be seen as rhetorical without being argumentative. And the third one is inclusion: in this case, argumentativity is only one amongst the different dimensions of a rhetoric discourse. As a rhetorician, Bonhomme adopts this third option. This explains why he distinguishes five functions in a rhetoric discourse: aesthetic, phatic, pathemic, cognitive, and finally argumentative. According to the definition he gives, a rhetoric discourse plays “an argumentative function when, through different factors […] the figures contribute to persuasion, acting on the addressee’s capacity to change their behavior. When it succeeds, such persuasion reinforces their beliefs and their convictions” (Bonhomme 2009, § 20).

Fig. 4 The Island vineyard, advertisement, France Soir Magazine, 1984

According to this understanding, argumentation is a province of rhetoric, and rhetoric is (again) reduced to persuasion. This, in turn, has consequences on the way Bonhomme conceives of the argumentativity of visual figures. For example, for him, metonymy works as a transfer from agent to product, matter to product, product to place, and so on. He explains: «These isotopic transfers make it possible for advertising to manipulate the universe of the products so as to make them desirable for the public and trigger the act of buying» (Bonhomme 2009, § 46). An example given by Bonhomme in another paper (Bonhomme, 2008, p. 221) is an ad for a Corsican wine, The Island vineyard (fig. 4).

It relies upon the fact that the grapevine is shaped like the island of Corsica (fig. 5).

Hence Bonhomme’s analysis of the metonymy as a transfer from product to place. Here, it seems that this visual metonymy has a purely persuasive function, as it helps the consumer, at the moment of purchase

Fig. 4 The Island vineyard, advertisement, France Soir Magazine, 1984

choice, to associate wine and Corsica. Even though it is important to show that some figures play an important persuasive role in images, visual rhetoric cannot be confused, however, with visual argumentation if we consider the latter as providing reasons to support a claim. (Fig. 5 Map of Corsica)

2.2.1 Metonymy
In fact, Bonhomme’s conception of the argumentativity of figures depends on his theoretical presuppositions, namely the rhetorical approach he applies to figures. But, besides this understanding of the rhetorical function of metonymy as persuasive, others interpretations are possible. For instance, Christian Plantin suggests that the mechanism that explains how metonymies work is like the mechanism that makes it possible to derive a conclusion from an argument. “In the metonymy of effect, the designation of the effect is replaced by that of the cause associated to it. In argumentation through consequences, the value judgment given to a consequence is transferred to its cause. The laws governing this kind of substitution of signifiers in a trope are not different from those that conclude to the acceptability of a cause from that of its effect (argument by consequences). We could therefore speak of a metonymic argumentation” (Plantin 2009, § 22). For sure, there are many images corresponding to this kind of metonymy of effects and causes. Let me examine one (Fig. 6).

I previously focused on this ad precisely because it recycles a series of paintings by Magritte (La Belle Captive) (Roque, 1983, p. 111-113). Here, I will analyze its argumentativity. So I’ll first describe the contents and context of the ad. It is taken from an ad campaign used by a French savings that focuses on housing. The text in bold just below the house reads: “The Crédit Agricole savings housing plan is an investment to live at home”. And below the road sign that points to the bank, there reads an inscription: “common sense close to your home,” which served as a slogan as well as an identification code for the bank in the eighties, across multiple ad campaigns.

Fig. 6. Advertisement for Crédit Agricole, Havas Conseil, 1976

This ad represents a case of a mixed media argument. According to a classification I proposed, it is what I call a joint argument, i.e. an argument produced by using visual as well as verbal elements (Roque, 2012, p. 283). It is also important to note that in a joint argument, both parts (verbal and visual) contribute to the argument. In this case, the text alone doesn’t advance all the reasons to open a savings account: the body text, printed in small letters, is a description of the savings program. The text below the picture of a house is also informative, explaining the purpose of the savings plan (to live in one’s own house). And finally, the text “Common sense close to your home” could serve as a conclusion to the argument (in addition to its role of reinforcing the bank’s brand), but not as the argument itself.

The image is based on a famous painting by Magritte that shows a canvas painting of a house that blends into the landscape in its background. The image relies upon a visual pun that pivots on the word “plan”, namely a savings plan and a house plan on paper. Rhetorically, it corresponds to a visual syllepsis, since the same graphic element can be perceived as being simultaneously part of two distinct sets (Noguez, 1974, p. 120). Now the house plan blends into the land where it is to be built. We could see it a metonymy of effect or otherwise of product and place. I have also suggested elsewhere (Roque, 2005, p. 275-276) that it could be understood as a particular case of metonymy, i.e. a metalepsis, since there is an inversion of cause (a savings plan) and consequence (building a house): in the image, the house is presented as having already been built.

The Magrittean image is quite effective, since it shows that the plan to have one’s own house is not just a dream but can easily become real thanks to Crédit Agricole’s savings plan. It is very persuasive, too: the house has a strong presence and helps suggest that it is easy to turn a dream into a house. If we consider the image as persuasive, it would be interesting to ponder whether it is also argumentative. But first of all, what is the visual argument here? We could say that it is something along the lines of: a saving account is a good investment because soon you’ll be the owner of your own home. Therefore a savings account is a common sense investment. The reference to the “common sense” is important as a way of suggesting that opening a savings plan is a rational and good decision. Furthermore, if one accepts that the visual might be dialogic (Roque, 2008), I would like to suggest that this is the case here: the visual part of the argument also seems to be a proleptic[iii] response to a possible objection about time: how many years would I have to save money before having my own house? Yet the image collapses the distance between cause and effect, project and realization. Therefore it helps to think that a savings plan is a good investment.

So how are we to analyze the visual rhetoric used in such an ad? As persuasive or as argumentative? The response is: both. The syllepsis can be considered as persuasive, as it suggests that the house simultaneously belongs to representation (painted on a canvas) and reality (built in the estate). As for the metonymy, it can be seen as persuasive, like Bonhomme does, if we understand the metonymy as a transfer through contiguity, between the product (a house to be built thanks to a savings plan) and the place (the private housing estate where the house has to be/is built). Conversely, we can see it as argumentative, like Plantin does, in so far as the acceptability of the consequence (to be landlord) is transferred to the cause (to buy a savings plan). Finally, the prolepsis, when it is used to anticipate a possible objection, is argumentative, too. The conclusion we can draw from it is that the same figure, in this case a trope (metonymy), can be understood either as persuasive or as argumentative. Therefore, these points of view are not exclusive. The fact that some visual figures are persuasive doesn’t prevent them from also being argumentative, at least in some cases. This first conclusion already has an important consequence: visual images cannot be easily rejected from the field of visual argumentation for being persuasive if we succeed in showing that they also work argumentatively.

3. Peersuasion and argumentation
In a previously published paper, I made the following argument: since a figure can be persuasive and argumentative at the same time, a distinction should be made between a strong and a weak notion of visual argumentation. I proposed to call a visual argumentation “strong” when an image is fully argumentative, i.e. when it gives reasons in order to support (or criticize) a point of view. Conversely, it should be qualified as “weak” when it is merely persuasive and influences the addressee (Roque 2011, p. 98-99). Such a suggestion doesn’t seem satisfying any longer. Why? Because it supposes that it would be possible to clearly distinguish which images would be “purely” persuasive and which are “purely” argumentative. In practice, such a distinction is challenging to apply. It turns out that persuasive and argumentative elements are often closely combined. The reason for distinguishing between strong and weak visual argumentation was to fortify visual argumentation as a well-founded field because it excluded visual persuasion from it. However, such a view also presupposes that persuasion is not rational. But there are indeed cases of rational persuasion, sometimes even ones that use emotional means of arousal (O’Keefe, 2012).

So, instead of separating persuasive and argumentative aspects, it is more convenient to accept that they often work together. This is, nevertheless, a controversial issue. Some authors hold that persuasion and argumentation should be carefully separated. My opinion is that in some cases – and visual argumentation is certainly one of them – persuasion and argumentation intersect and are intertwined (Nettel & Roque, 2012). This understanding corresponds to that held by informal logicians, like Ralph Johnson and Tony Blair, who claim that argumentation is rational persuasion (Johnson 2000, p. 149-150; Blair, 2012). Blair’s analysis of different types of advertising is a good case in point (Blair, 2012, p. 75-77). In some advertisements, there is a mix of rational and non-rational – or irrational – reasons given for preferring one brand over others. Yet, if “the argument is the effective persuasive tool […] persuasion occurs through the use of arguments” (Blair, 2012, p. 76) and we have a case of rational persuasion.

Now, once we stop considering persuasion and argumentation as mutually exclusive, it becomes essential, when analyzing images, to determine whether or not persuasion is accompanied by a set of rational reasons provided to support a claim. Indeed, adversaries of visual argumentation could claim that in such cases, even though it is true that there is persuasion as well as argumentation, the persuasive role would be that of images.
For this reason it is important to better understand the relationship between figures of rhetoric and argumentation. Two different kinds of relationship have been envisaged: either figures help better present arguments, or figures are arguments themselves (Reboul, 1986, p. 184; Bonhomme, 1998, p. 88; Tindale, 2004, p. 59). In the first case, the relationship between figure and argumentation is extrinsic. In the second, it is intrinsic. When the relationship is extrinsic, the figure cannot be considered properly “argumentative”; it remains exterior to the argument and is merely persuasive most of the time. In the second case, it must be recalled that when a figure itself is an argument, this doesn’t necessarily mean that it cannot also be persuasive. Yet, what happens for the general relationship between persuasion and argumentation holds true, too, for the figures.
As we already saw, a trope like metonymy can be simultaneously seen as persuasive and argumentative. So it turns out that it is hardly possible to separate persuasive and argumentative aspects of a given figure. Furthermore, the distinction between extrinsic and intrinsic relationship is itself relative. Indeed, at the end of his 1986 paper, Reboul considers that the two different cases, extrinsic and intrinsic “are almost always indistinguishable” (Reboul, 1986, p. 186). For this reason, he relinquished the distinction when reprinting his paper as a chapter of his book (Reboul, 1991).

4. Conclusion
1. By examining the relationship between figures of rhetoric and argumentation, it turns out that, for most authors, when a figure is used in discourse, its function is primarily persuasive. Consequently, we must be careful when transposing their idea to the field of visual argumentation, since images are generally considered as more persuasive than argumentative.
2. This is particularly true for what Perelman and Olbrechts-Tyteca call figures of “presence” (hypotyposis, energeia). The fact that they are effective and impress the audience doesn’t necessarily transform them into argumentative tools.
3. Some figures (like metonymy) appear to be considered as persuasive and also argumentative. Furthermore, it is hardly possible to separate figures that would be persuasive from figures that would be argumentative.
4. If we admit that persuasion and argumentation are very often combined, the fact that many images are persuasive doesn’t prevent them from being simultaneously argumentative (at least in some cases). This point is quite important to counter the argument according to which images would be mainly persuasive. However, this raises the need to distinguish between these two complementary functions of images.
5. The concept of strategic maneuvering can be helpful here because it “refers to the continual efforts made in all moves that are carried out in argumentation discourse to keep the balance between reasonableness and effectiveness” (van Eemeren, 2010, p. 40). Similarly, I would like to suggest that something similar occurs in visual images. When there is a balance between reasonableness and effectiveness, visual images can be considered as successfully displaying a visual argument. But when effectiveness (i.e. persuasiveness – even though van Eemeren warns us that effectiveness and persuasiveness are not completely synonymous: van Eemeren, 2010, p. 39) gets the better of reasonableness, visual images are mainly persuasive.

Acknowledgements
I would like to thank Marc Bonhomme for giving me a reproduction of fig. 4.

NOTES
i. I will leave aside the complex issue of the relationship between verbal rhetoric and visual rhetoric, i.e. examining to what extent the verbal rhetoric terms can be transposed into visual rhetoric.
ii. Both deal with rhetoric and visuality, and their names are very similar. However, « energeia, » usually translated as « activity, » means « vividness, » while « enargeia » has the general meaning of visual clarity, but also pictorial vividness. As it has been noted, Aristotle uses the first one in his Rhetoric, not the second one (Zanker 1981, note 40 p. 307).
iii. On the prolepsis as persuasive and argumentative, see Nettel and Roque, 2012, p. 64-65.

References
Adam, J.-M., & Bonhomme, M. (2005). L’Argumentation publicitaire. Rhétorique de l’éloge et de la persuasion. Paris: Armand Colin.
Anscombre, J.-C., & Ducrot, O. (1986). Argumentativité et informativité. In M. Meyer (Ed.), De la métaphysique à la rhétorique (pp. 79-94). Brussels: Editions de l’Université de Bruxelles.
Aristotle. (2005). Poetics and Rhetoric, Ed. E. Garver. New York: Barnes & Noble.
Belle, H. van (2009). Playing with Oppositions. Verbal and visual antithesis in the media. In J. Ritola (Ed.), Argument Cultures: Proceedings of OSSA 09, CD-ROM (pp. 1-13), Windsor, ON: OSSA.
Blair, T. (2004). The Rhetoric of Visual Arguments. In C. A. Hill and M. Helmers (Eds.), Defining Visual Rhetorics (pp. 41-61). Mahwah, N.J.: Lawrence Erlbaum.
Blair, T. (2012). Argumentation as Rational Persuasion. Argumentation 26-1, 71-81.
Bonhomme, M. (2009). De l’argumentativité des figures de rhétorique. Argumentation et analyse du discours, 2, special issue, Rhétorique et argumentation (http://aad.revues.org/index495.html).
Bonhomme, M. (1998). Les figures clés du discours. Paris: Seuil.
Bonhomme, M. (2008). Peut-on parler de métonymie iconique? In S. Badir & Klinkenberg, J.-M. (Eds.), Figures de la figure. Sémiotique et rhétorique générale (pp. 215-228). Limoges: Presses Universitaires de Limoges.
Doury, M. (2012). Preaching to the converted. Why Argue when Everyone Agrees. Argumentation 26-1, 99-114.
Ducrot, O. (1992). Argumentation et persuasion. In W. De Mulder, F. Schuerewegen & L. Tasmowski (Eds.), Énonciation et parti-pris: Actes du Colloque d’Anvers, février 1990 (pp. 143-158). Amsterdam & Atlanta: Rodopi Bv Editions.
Eemeren, Frans H. van. (2010). Strategic Maneuvering in Argumentative Discourse. Amsterdam and Philadelphia: John Benjamins Publishing Company.
Fontanier, P. (1968). Les figures du discours. Paris: Flammarion.
Hill, Charles A. (2004). The Psychology of Rhetorical Images. In C. A. Hill and M. Helmers (Eds.), Defining Visual Rhetorics (pp. 25-40). Mahwah, N.J.: Lawrence Erlbaum.
Johnson, R. H. (2000). Manifest Rationality: A Pragmatic Theory of Argument. Mahwah, NJ/London: Lawrence Erlbaum.
Jowett, G. S. & O’Donnell, V. (1992). Propaganda and Persuasion. 2nd ed. Newbury Park, Cal. & London: Sage.
Kjeldsen, J.E. (2012). Pictorial Argumentation in Advertising: Visual Tropes and Figures as a Way of Creating Visual Argumentation. In F. H. van Eemeren & B. Garssen (Eds.), Topical Themes in Argumentation Theory. Twenty Exploratory Studies (pp. 239-255). Dordrecht, Heidelberg, London & New York: Springer.
Lévy, C. & Pernot, L. (1997). Phryné dévoilée. In C. Lévy & L. Pernod (Eds.), Dire l’évidence (Philosophie et rhétorique antiques) (pp. 5-12). Paris: L’Harmattan.
Messaris, P. (1997). Visual Persuasion. The Role of Images in Advertising. Thousands Oaks, Cal., London & New Delhi: Sage Publications.
Micheli, R. (2012). Arguing without Trying to Persuade? Elements for a Non-Persuasive Definition of Argumentation. Argumentation, 26-1, 115-126.
Nettel, A. L., & Roque, G. (2012). Persuasive Argumentation Versus Manipulation. Argumentation, 26-1, 55-69.
Noguez, D. (1974). Petite rhétorique de poche pour servir à la lecture des dessins dits “d’humour”. L’Art de masse n’existe pas. Special issue. Revue d’esthétique, 1974, 3/4, 107-137.
0’Keefe, D. (2012). Conviction, Persuasion and Argumentation: Untangling the Ends and Means of Influence. Argumentation, 26-1, 19-32.
Perelman, C. & Olbrechts-Tyteca L. (1970). Traité de l’argumentation: La Nouvelle Rhétorique. Brussels: Editions de l’Institut de Sociologie, Université Libre de Bruxelles.
Plantin, C. (1990). Essais sur l’argumentation. Introduction à l’étude linguistique de la parole argumentative. Paris: Kimé.
Plantin, Christian. 2009. Un lieu pour les figures dans la théorie de l’argumentation. In Argumentation et analyse du discours 2. http://aad.revues.org/215. Accessed 23 July 2011.
Plantin, C. (2012). Persuasion or Alignment? Argumentation, 26-1, 83-97.
Pratkanis, A., & Elliot Aronson E. (2001). Art of Propaganda: The Everyday Use and Abuse of Persuasion. New York: W.H. Freeman & Company.
Reboul, O. (1986). La figure et l’argument. In M. Meyer (Ed.), De la métaphysique à la rhétorique (pp. 175-187). Brussels: Editions de l’Université de Bruxelles.
Reboul, O. (1991). Introduction à la rhétorique: Théorie et pratique. Paris: Presses Universitaires de France.
Roque, G. (1983). Ceci n’est pas un Magritte. Essai sur Magritte et la publicité. Paris: Flammarion.
Roque, G. (2005). Sous le signe de Magritte. In J. Pier & J.-M. Schaeffer (Eds.), Métalepses. Entorses au pacte de la représentation (pp. 263-276). Paris: Editions de l’Ecole des Hautes Etudes en Sciences sociales.
Roque, G. (2008). Political Rhetoric in Visual Images. In E. Weigand (Ed.), Dialogue and Rhetoric (pp. 185-193). Amsterdam/Philadelphy: John Benjamins Publishing Company,
Roque, G. (2011). Rhétorique visuelle et argumentation visuelle. Semen 32, 91-106.
Roque, G. (2012). Visual Argumentation: A Further Reappraisal. In F. H. van Eemeren & B. Garssen (Eds.), Topical Themes in Argumentation Theory. Twenty Exploratory Studies (pp. 273-288). Dordrecht, Heidelberg, London & New York: Springer.
Scott L. M., & Batra R., (Eds.) (2003). Persuasive Imagery: A Consumer Response Perspective. Mahwah, N.J.: Lawrence Erlbaum.
Seidman, S. A. (2008). Posters, Propanganda and Persuasion in Election Campaigns around the World and through History. New York: Peter Lang.
Spangeburg, R., & Moser K. (2002). Propaganda. Understanding the Power of Persuasion. Berkeley Heights, NJ: Enslow Publishers, Inc.
Vouilloux, B. (1995). Des deux statuts rhétoriques de l’image et peut-être d’un troisième. In L.H. Hoek & K. Meerhoff (Eds.), Rhétorique et image. Textes en hommage à Á. Kibédi Varga (pp. 101-114). Amsterdam & Atlanta: Rodopi.
Zanker, G. (1981). Enargeia in the Ancient Criticism of Poetry. Rheinisches Museum fur Philologie 124, 297-311.

Abstract: Movie trailers are hybrid (combining narrative and advertising) audiovisual discourse genres that exploit a carefully selected re-montage of moving and still images, sound, music, voice-over, intertitles, etc. to persuade potential spectators/consumers that a forthcoming movie is worth watching. I hypothesize that movie trailers reach their goal by advancing monomodal (e.g. only pictorial or only verbal) and multimodal arguments and by employing monomodal and multimodal rhetorical schemas and tropes (e.g. metonymy and synecdoche).
Keywords: dispositio, elocutio, inventio, loci, movie trailers, metonymy, multimodal argumentation, multimodality, synecdoche.

1. Introduction
This is an exploratory study which looks at movie trailers as discourse genres from a rhetorical and argumentative point of view.
With this study, I wish to contribute to the research on visual/multimodal argumentation and the research on the relationship and isomorphism between rhetorical figures/tropes and argumentative topoi (or loci). On the one hand, the study on visual/multimodal argumentation has flourished since a special issue of Argumentation and Advocacy came out in 1996. This year marks a shift in the studies on argumentation: since then, scholars have become more and more aware of the fact that real argumentative discourses in real contexts do not convey arguments only verbally but exploit all the semiotic resources available to make their point and to persuade people. On the other hand, the study of the link between patterns of elocutio from ornatus (i.e. rhetorical figures and tropes) and patterns of inventio (i.e. argumentative loci) is not completely new. The author of the website Silva Rhetoricae puts into question the sharp division between tropes and topoi:
The difference between a figure and a topic of invention, then, may sometimes simply be a matter of degree, or it may be a matter of whether one views the strategy as one of expression of an idea (an issue of style) or the composition or discovery of an idea or argument (an issue of invention). The point is, we should recognize the close proximity of the figures and the topics of invention.

In order to understand the role of rhetorical figures/tropes, Fanhestock (1999, p. 23) suggests “shift[ing] the emphasis from what the figures are to what it is they do particularly well”, that is “epitomize lines of reasoning.” Also, Tindale (2004) says that figures are arguments if they engage the audience in a premise-conclusion process. More recently, Kjeldsen (2012) has investigated how tropes contribute to the inferential reconstruction of enthymemes in advertisements. He argues that pictorial rhetorical figures delimit the interpretation of the message of an advertisement and evoke the intended argument. I have tried to contribute to this line of research in Pollaroli and Rocci (forthcoming).

Movie trailers are an interesting discourse genre to be explored because of their multimodal and hybrid nature. Unfortunately, they have hardly ever raised scientific interest, as Carmen Maier (2011) complains about. Movie trailers are multimodal discourse genres because they combine meaning manifested through different semiotic modes such as moving and still images, sound, music, written and spoken language. As Dornaleteche Ruiz (2007) says, movie trailers are shows of other shows, they are audiovisual discourses anticipating and promoting other audiovisual discourses. Indeed, movie trailers are communicative practices that employ the same semiotic modes (and often the same media, especially when they are broadcast in cinemas) of the communicative practices they promote.

Movie trailers are hybrid because they combine the narrative nature of the movie they are constructed upon and the promotional nature of advertising; as Maier (2011, p. 141) says “trailers are designed to sell and tell a story.” The goal of movie trailers is to persuade potential consumers/spectators that a forthcoming movie is worth watching (Dusi, 2002; Kernan, 2004; Dornaleteche Ruiz, 2007, 2009; Maier, 2009, 2011). For this, they can be considered as a type of advertising, especially as a type of TV commercials (Dornaleteche Ruiz, 2007). The product is a movie, specifically it is a movie experience; in fact, one cannot properly ‘buy a movie’ as if it was a pair of shoes, but can go to the movies and watch it. In order to reach their advertising goal, movie trailers have to both give some information on the forthcoming movie to arouse the prospective consumer/spectator’s interest and leave out some other information to encourage the audience to go and watch the movie in the case they are interested in the story (or other features of the movie) and wish to know more about it. As Dornaleteche Ruiz (2007, p. 102) says, the marketing strategy of movie trailers is similar to those types of marketing (known as merchandising) that tempt the audience by offering an anticipation of the product (e.g. pieces of a new brand of cheese at the supermarket, free trails on websites that teach languages, demo of videogames sold with magazines) in order to ‘whet the appetite’ of the consumer. Movie trailers are appetizers of coming attractions (Kernan, 2004). In this study I wish to explore the hypotheses that:
1. Movie trailers are argumentative activity types;
2. Movie trailers employ multimodal arguments to fulfil their promotional goal;
3. Movie trailers employ multimodal rhetorical patterns from ornatus (e.g. synecdoche, metonymy, hyperbole, ellipsis);
4. The rhetorical patterns employed are argumentatively relevant, that is, they make the audience infer the arguments advanced in support of the standpoint put forward in the movie trailer.

This study does not present final results but only some preliminary results of a path of research that should be further developed.

2. Movie trailers are argumentative discourses
So far movie trailers have not been studied as argumentative discourses; yet, the persuasive purpose of film trailers is acknowledged among those few scholars that have written about them (Dusi, 2002; Kernan, 2004; Dornaleteche Ruiz, 2007, 2009; Maier, 2009, 2011).

Following Rigotti and Rocci’s (2006) model for communication contexts, movie trailers can be described as communicative activities which result from the application of the advertising interaction scheme – namely a culturally shared scheme of interaction which helps in achieving a goal – to the interaction field – namely the institutional reality defined by shared goals and commitments – of the market of movies. Broadly speaking, the goal of the people working in the market of movies is the positioning of a movie in the film market (Dornaleteche Ruiz, 2007, p. 100) in order for it to perform well at the box office in theatres. The goal of movie production companies is achieved only when spectators go and watch the movie in theatres; their goal will not be satisfied if spectators limit themselves in receiving the information provided in the trailer. Movie trailers are argumentative as advertisements are. Arguing that movie trailers are argumentative discourses because they are a specific type of advertising may not be easily accepted, especially among scholars who do not believe that advertisements can argue (see Blair, 1996, 2004). However, other scholars provide good reasons for claiming that advertisements argue (Pateman, 1980; Slade, 2002, 2003; Atkin & Richardson, 2005; Ripley, 2008; Rocci, 2008, 2009; van den Hoven, 2012; Kjeldsen, 2012; Mazzali-Lurati & Pollaroli 2014; Rocci, Mazzali-Lurati & Pollaroli, 2013; Wierda & Visser, 2013; Pollaroli & Rocci, forthcoming). The following quotation from Atkin and Richardson (2005, p. 167) clearly summarizes the position of these scholars:
Advertising discourse [is] per se argumentative given that advertising offers evidence – often implicit, indirect or semiotic support in addition to (largely non-requisite) premises – in defence of a contested or contestable position.

Ripley (2008) shows that advertising can be seen as argumentative from the perspective of different argumentation theories. Advertising for products, for instance, is, from a pragma-dialectical point of view, a single non-mixed difference of opinion (see van Eemeren, Grootendorst & Snoeck Henkemans, 2002; Wierda & Visser, 2013). Following this perspective, movie trailers can be seen as single non-mixed differences of opinion between a movie production company (the protagonist) and potential consumers/spectators (the antagonist). The standpoint often remains implicit, but it can be easily reconstructed from the context and verbalized as Movie X is worth watching in the theatre or You should watch movie X in the theatre. Moreover, it is often the case that the arguments are enthymematic and implicit, but the context and the recognizable overall purpose of the discourse make it possible to make them explicit and reconstruct the whole discourse as argumentative.

In order to fulfil their promotional goal, movie trailers advance arguments employing either the verbal, visual, or aural semiotic systems or a combination of them, that is they advance arguments multimodally. Although the scepticism about multimodal argumentation persists (Johnson, 2003; Blair, 1996, 2004; Jacobs, 2000), more and more scholars in argumentation theory claim that pictures, odours, sounds, moving images, etc. provide arguments in support of claims (Alcolea Banegas, 2009; Groarke, 2009; Kjeldsen, 2012; Dove, 2012; van den Hoven, 2012; Pollaroli & Rocci, forthcoming). For these scholars the argumentative role of discourse elements is independent from their manifestation in the verbal mode. The audience of multimodal argumentative discourses is able to recognize arguments manifested in other semiotic systems rather than the verbal one and to understand and correctly interpret the communicated message without translating it into words. Yet, analysts interested in the reconstruction of the claim(s) and argument(s) of multimodal argumentative discourses need to translate visual/aural/multimodal arguments into words; this may result in the loss of part of the original meaning. Seeing visual/aural/multimodal arguments as enthymemes may be a good starting point. Some scholars (Birdsell & Groarke, 1996, p. 6; Smith, 2007; Kjeldsen, 2007, 2012) claim that images can be enthymemes, that is rhetorical syllogisms that need the active participation of the audience to be completed with contextual-bound premises. The effectiveness of enthymemes relies on these contextual premises. Kjeldsen (2012, p. 241) sees images as “offer[ing] a rhetorical enthymematic process in which something is condensed or omitted, and, as a consequence, it is up to the spectator to provide the unspoken premises”.

3. Inventio and disposition in movie trailers
Movie trailers are composed of a carefully selected re-montage of dialogues, moving images, sounds, and music from the movie they promote and arrange them together with non-diegetic voice-over, shots and scenes created for the trailer only or original shots that were not included in the final editing of the movie, shots with information about the actors, the director, the production company, day of release, prizes that the movie has been awarded, empty black or white shots etc.

All this makes movie trailers something completely different from summaries of movies. The chronological structure of a movie in transformed into the mainly non-chronological structure of a trailer. Dornaleteche Ruiz (2007, p. 105) says that trailers may be constituted of ‘bracket syntagmas’ (Metz, 1989; Bateman, 2007) of the story that is told in the coming movie. Bracket syntagmas are shots put together because they represent examples of a reality, a topic, without chronological order and temporal link.

Maier (2009, p. 162) points out that consumers/spectators “evaluate” the characters, the relationships, the events, the film company, the actors, the director presented in the movie trailer, and consequently the movie advertised, visually. In fact, Maier defines “evaluative devices as being those verbal, visual and aural resources that inherently or contextually signal a process of appraisal” (2009, p. 165); thus, her concept of ‘evaluation’ is similar to ‘argumentation’. In my view, these are all diegetic and extra-diegetic visual (or multimodal) arguments. Examples of promotional evaluative devices in movie trailers are, for Maier (2009), the film company’s logo which “not only reminds the viewer of the company’s prestige, it may also be an indication of the quality or type of films created by the company” (p. 171) and the name of an actor, which has a similar effect to that of the film company’s logo. Maier (2009, p. 172) also points out that “no single semiotic mode is supposed to carry the whole or only evaluative information of a shot or scene. Visual, verbal and aural evaluative devices are co-deployed to maintain or subvert each others’ evaluative load both on the diegetic and non-diegetic levels.” These evaluative devices may be seen as the recurrent patterns of inventio that are employed in movie trailers.

How do these elements hold together in movie trailers as discourses? As Carmen Maier (2009, p. 161) points out “the whole structure of these film trailers is motivated by their promotional purpose.” This insightful remark can be better explained adopting the pragma-rhetorical perspective on discourses that Congruity Theory has developed (Rigotti, 2005; Rocci, 2005; see the literature cited in Mazzali-Lurati & Pollaroli, 2014). Following Congruity Theory, we see monomodal/multimodal discourses as complex acts governed by a superordinate act that corresponds to what the addresser does to the addressee with the discourse; all discourse elements are subordinate acts that contribute to fulfil the goal of the text as a whole. The promotional goal of movie trailers determines the complex multimodal act of the text – which is similar to that of advertising for product – and the functions fulfilled by the multimodal sequences of the movie trailer are subordinate to the advertising one. Multimodal sequences in audiovisual discourses are clusters of shots combined together with sound, music and other elements that form a unit; in order to determine the boundaries of each sequence we must look at changes in music, sound, images, etc. The voiceover may help in marking the multimodal sequences. I agree with Carmen Maier pointing out that all stages – or multimodal sequences – fulfil a ‘promotional’ function “through different informative means” (p. 144). From the perspective of Congruity Theory, the promotional function corresponds to the complex superordinate act whereas the informative means correspond to the subordinate acts.

In other words, movie trailers are multimodal argumentative discourses that perform the complex act that, for the purpose of this paper, we can name ‘the movie trailer act’. All multimodal subordinate units concur in performing the high-level act. Maier (2009) identifies different stages that fulfil specific functions in movie trailers. We will see some of them through the analysis of an example in Section 5.

Movie trailer act
(Addresser, Addressee, T)

Presupposition
Addresser is a motion picture company that produced movie X;
Addressee is a potential consumer/spectator;
T is a movie trailer having a propositional content Y which shows the movie story and other information about the movie.
Movie X will be available at time t. Addresser reasonably believes that movie X will satisfy a desire of Addressee.

Pragmatic effect
By stating T, Addresser commits himself in offering movie X and expresses the desire that
Addressee benefits from movie X.

The complex act determines the inferential process that the audience is invited to perform in order to correctly understand and interpret each multimodal sequence of a movie trailer. The meaning in movie trailers is condensed (Wildfeuer, 2014; see also Kjeldsen, 2012 and the enthymematic nature of visual/multimodal argumentation mentioned in Section 1) and the way multimodal sequences are arranged may seem incoherent and chaotic because, for instance, information about the production company is followed with brief shots from the movie and this is interrupted by information about the actors, etc. Indeed, Wildfeuer (2014) notes that the inferential work required by viewers in order to interpret a trailer is different from the inferential work they operate to interpret a movie. This is consistent, from a Congruity Theory perspective, with the very different superordinate complex acts that movie trailers and movies perform, respectively a promotional goal and an entertainment goal.[i] However, a link between the inferential work performed when watching a movie trailer and the process of interpretation of the promoted movie remains. Indeed, a movie trailers invites the audience to operate anticipatory hypotheses (Moeschler & Reboul, 2009) on the cinematic discourse that we are invited to watch in theatres.

4. Elocutio in movie trailers
Movie trailers employ patterns from elocutio, such as synecdoche, metonymies, hyperbole[ii], ellipsis (here I will focus only on metonymy and synecdoche for reasons of space).

In the last few decades, cognitive linguists have shown that traditional rhetorical figures and tropes are deep and pervasive structures of our thoughts through which people conceptualize and understand the world (Lakoff and Johnson, 2003[1980]; Barcelona, 2003; Ortony, 1993; Panther & Radden, 1999). Lakoff and Johnson (2003 [1980], p. 5), for instance, claim that “the essence of metaphor is understanding and experiencing one kind of thing in terms of another”. Stemming from this approach to metaphor, Forceville (1996) shows that the manifestation of a metaphor is not necessarily verbal but it can also be pictorial and multimodal: metaphors can be manifested by images and by a combination of different semiotic modes such as words and images, sound, moving images, etc. (see contributions in Forceville & Urios-Aparisi, 2009).

Metonymy is a substitution of one concept with another which plays a contiguous semantic role within the same frame (Bohnomme, 2005). The focus shifts from the proper concept and role to the substituted one. Metonymic concepts “usually involve[s] direct physical or causal associations” which are systematic and “grounded in our experience”. Indeed, it is possible to identify “certain general metonymic concepts in term of which we organize our thoughts and actions”; for example, the relations “producer for product”, “object used for user”, “controller for controlled”, “institution for people responsible”, “place for the institution” and “place for the event” (Lakoff & Johnson, 2003 [1980], p. 39). Consider, for example, the sentence She’s wearing an Armani in which the producer substitutes the product, or a TV commercial of a brand of water where the mountains from which the water springs are shown (metonymy of the origin-for-product type). Works on pictorial and multimodal metonymy (Forceville, 2009; Bonhomme & Lugrin, 2008; Urios-Aparisi, 2009; Yu, 2009; Villacañas & White, 2013; see also Forceville 1996) identify instances of metonymic relationships represented by visual elements in static or dynamic images in advertising texts.

Since Antiquity synecdoche has been recognized as a rhetorical figure independent from metonymy. Yet, already Quintilian noticed the little difference that exists between the two rhetorical tropes and that “it is but a short step between synecdoche and metonymy” (Institutio Oratoria VIII.VI.23). Burkhardt (2010, p. 247) laments that “a clear principle for the distinction between metonymy and synecdoche, which is more than 2,000 years old, is still missing”. Nerlich (2010) agrees and points out that it is a hard task to give a definite and agreed upon definition of synecdoche as well as to find its position in the realm of rhetorical figures. The distinction has been made even harder as synecdoche has been sometimes considered as a subtype of metonymy (Lakoff and Johnson, 2003 [1980]). For space reason, I cannot report all the characterizations and classifications that have been proposed on synecdoche, but I can plausibly claim that synecdoche is a structure of thought that substitutes the part for the whole (There where only ten heads today in the classroom) or the genus for the species (He has a temperature), the singular for the plural (The Roman won the battle), and vice versa.

Some research has been conducted on the manifestation of rhetorical patterns in audiovisuals, especially in movies and in TV commercials (Whittock, 1990; Forceville, 2007, 2009; Urios-Aparisi, 2009; Yu, 2009). Whittock (1990) lists nine ‘cinematic metaphors’ that include metonymy, synecdoche, explicit comparison and distortion. Forceville (2007) claims that metaphor can be manifested multimodally in TV commercials and metonymy (Forceville, 2009) is employed in movies when, for example, the spectator hears a sound that is connected with something that is not displayed on the screen (e.g. the creaking floorboards that stand for an unwelcome visitor) or the spectator watches a close-up of a part of the body (e.g. moving mouth) that stands for an action (e.g. talking). It follows that movie trailers as well may manifest rhetorical patterns such as synecdoche – parts of the movie stand for the whole movie – and metonymy – the director and the film production industry stand for the movie.

I hypothesize that these rhetorical patterns epitomize lines of reasoning, saying it with Fanhestock (1999), and make the viewer infer the intended argument, saying it with Kjeldsen (2012). For example a metonymy condenses an argument based on a locus from final cause or efficient cause and a synecdoche condenses an argument based on a locus from parts to whole.

5. A case study
In this section I will analyze a movie trailer that won the 15th Golden Trailer Awards for the ‘best in show’ trailer. It promotes the movie Gravity (2013) by Alfonso Cuarón.

This movie trailer is a one-minute 51 seconds audiovisual discourse composed of 7 multimodal sequences. A preliminary step for the analysis of audiovisuals is the transcription of the discourse into the written modes. The transcription is useful because it gives a synthetic representation of the linearity and strata of the audiovisual text (Casetti & Di Chio, 2009). The transcription table proposed here (table 1) is a simplified version of the transcription table presented in Rocci, Mazzali-Lurati & Pollaroli (2013) constructed on the basis of Baldry & Thibault (2006), Bateman (2007), and Casetti & Di Chio (2009).

The movie trailer for Gravity is composed of multimodal sequences that fulfil specific functions in the trailer (Maier, 2011) and concur to perform the overall promotional act of the discourse. Combining Maier’s functions and Congruity Theory, we can identify the act performed by each multimodal sequence.

The multimodal sequence 1, which lasts 4 seconds, shows the Warner Bros Pictures logo and accomplish what Maier (2011) calls the Promotional Identification function because gives non-diegetic information about the film company. The multimodal sequence 2 is diegetic and is composed of only one shot, that is one uninterrupted image, without editing cuts but with many frames. It lasts 1 minute 31 seconds and it shows an entire scene from the movie advertised. This multimodal sequence functions both as Orientation and as a Complication (always following Maier’s stages) because it introduces the characters and the situation and also what seems to be the disruptive event. The audience watches three astronauts working outside of the space shuttle Explorer. The mission control in Houston warns the team about debris in the space which do not last much in arriving. One of the astronauts is hit and seems dead, the astronaut Stone cannot unbuckle the belt that keeps her tied to the shuttle arm; while the astronaut Kowalski is trying to help Dr. Stone, the shuttle arm is broken by some debris and she starts tumbling through space. The spectator watching this sequence operates many inferences and anticipatory hypotheses (Moeschler & Reboul, 2009) about the plot and the chronological order of the events (is this the beginning of the movie or the end? What is the reason for the accident and the debris being around the Earth?) and the characters (Are those the only characters? How is the relationship between them? What happens to Dr. Stone after she is thrown away from the space shuttle?). The following multimodal sequences give extra-diegetic information.

Multimodal sequence 3 identifies the title of the movie thus specifying one of the elements presupposed in the ‘movie trailer act’ we have seen in Section 2. Multimodal sequence 4 identifies the famous actors playing the two characters the audience has just seen in multimodal sequence 2. The multimodal sequence 5 identifies the director. The multimodal sequence 6 gives information of the date of release in theatres and specifies a detail of the ‘movie trailer act’. Multimodal sequences 1 to 6 are composed of one shot each. Two shots compose MS7 in which some information is repeated (director, film company, actors) and some information is added about the music and the production. The overall act performed in this movie trailer is:

Movie trailer actGravity
(Addresser, Addressee, T)

Presupposition:
Warner Bros Pictures is a motion picture company that produced Gravity; Addressee is a potential consumer/spectator; T is a movie trailer having a propositional content Y which shows the movie story and other information about the movie. Gravity will be available on 10.04.2013. Warner Bros Pictures reasonably believes that Gravity will satisfy a desire of Addressee.

Pragmatic effect:
By stating T, Warner Bros Pictures commits himself in offering Gravity and expresses the desire that Addressee benefits from Gravity.

A reconstruction of the standpoint and the arguments following the pragma-dialectical analytical overview shows that the movie trailer benefits from a complex argumentative structure in which subordinate argumentation combines with multiple argumentation (van Eemeren, Grootendorst, 2004; van Eemeren, Grootendorst & Snoeck Henkemans, 2002).

1. You should watch Gravity (which will be released in theatres on 10.04.2013)
1.1 The movie Gravity is entertaining
1.1.1 The multimodal sequences (parts) that you are watching in the movie trailer are entertaining
1.2 Gravity is good (is a movie of high quality)
1.2.1 Sandra Bullock and George Clooney are starring
1.2.2 Gravity is directed by Alfonso Cuarón
1.2.3 Gravity is produced by Warner Bros. Pictures

The analytical overview shows that single aspects, or ‘parts’, of the movie are presented as details of quality; the quality of the parts of the movie is transferred to the movie as a whole and are presented as reasons for making Gravity worth watching.

The Argumentum Model of Topics (Rigotti & Greco Morasso, 2010; see the literature cited there) is helpful in making explicit the inferential path that links the arguments and the standpoint by making explicit the locus that licenses the premises-conclusion relation. According to Rigotti and Greco Morasso, arguments are composed of two equally important dimensions: the endoxical (also known as material or contextual) dimension and the logical (or procedural) dimension. In our case study, we see that the argument ‘The movie Gravity is entertaining’ (1.1) and ‘The multimodal sequences (parts) that you are watching in the movie trailer are entertaining’ (1.1.1) are linked by a synecdoche of the part-whole type that condenses a locus from parts to whole (figure 1). In the contextual dimension the endoxical premise ‘The multimodal sequences that you are watching in the movie trailer are parts of the movie Gravity’ combines with the factual premise (datum) ‘The multimodal sequences that you are watching in the movie trailer are entertaining’. The positive feature of being entertaining is transferred to the movie according to the maxim ‘If all parts share a property, the whole will inherit this property’.

The arguments ‘Sandra Bullock and George Clooney are starring’ (1.2.1), ‘Gravity is directed by Alfonso Cuarón’ (1.2.2), and ‘Gravity is produced by Warner Bros. Pictures’ (1.2.3) that support the evaluative standpoint ‘Gravity is good (is a movie of high quality)’ (1.2) are linked to the movie by a metonymical relation. Warner Bros. Pictures is the film production company that produces the movie, it is linked through a metonymy of the producer-for-product type and makes the viewer infer an argument licensed by a locus from efficient cause (figure 2). The director Alfonso Cuarón is also linked to the movie with a metonymy of the producer-for-product type and it is based on a locus from efficient cause as well (figure 3). Sandra Bullock and George Clooney are the actors that play the main characters of the movie; their link to the movie operates upon a metonymy and the line of reasoning is a locus from efficient cause (figure 4). In the three arguments the quality of the production company, the actors and the director which is accepted as an endoxical premise is transferred to the movie in accordance with the maxim ‘If a quality characterizes the efficient cause such quality characterizes the effect too’.

6. Conclusion
For now I am able to draw only some very preliminary conclusions that I will develop in future research.

Movie trailers can be reconstructed as argumentative discourses where the standpoint You should watch movie X in the theatre is supported by multimodal arguments. The multimodal sequences contribute in performing the overall act of movie trailers as discourses. The rhetorical patterns employed in movie trailers are argumentatively relevant, that is they make the viewers infer the intended argument licenses by a specific argument scheme or locus, e.g. synecdoche makes the view infer an argument licenses by a locus from parts to whole and metonymical relations make the viewer infer an argument licensed by a locus from efficient cause.

From the discussion and the presentation of the case study, I can draw the methodological consideration that a combination of approaches and disciplines is the only way to analyze complex audiovisual argumentative discourses.

NOTES
i. I am aware of the fact that the complex act performed by movies should not be easily dismissed and classified as ‘entertainment’. Indeed, Alcolea-Banegas (2009) and Chatman (1990) claim that movies can argue. However, I will not deal with this issue here because it exceeds the topic of this paper.
ii. Movie trailers exaggerate the film’s ‘plot’ “to maximise the viewer’s expectations and curiosity concerning various aspects of the film and not just the film’s story” (Maier, 2011, p. 145) and to raise doubt which are left unsolved “to trigger the viewers’ keener expectations and persuade them to see the whole film later on” (p. 146). For Dornaleteche Ruiz (2007, p. 105) the selection and montage of shots from the movie to realize a trailer is done with the objective of magnifying the movie and making its excellence stand out.

References
Alcolea-Banegas, J. (2009). Visual arguments in film. Argumentation, 23, 259-275.
Atkin, A. & Richardson, J. (2005). Constructing the (imagined) antagonist in advertising argumentation. In F. H. van Eemeren & P. Houtlosser (Eds.), Argumentation in practice (pp. 163-180). Amsterdam: John Benjamins.
Baldry, A. & Thibault, P. J. (2006). Multimodal transcription and text analysis. London: Equinox.
Barcelona, A. (Ed.) (2003). Metaphor and Metonymy at the crossroad. A cognitive perspective. Berlin/New York: Mouton de Gruyter.
Bateman, J.A. (2007). Towards a grande paradigmatique of film: Christian Metz reloaded. Semiotica, 167, 1/4, 13-64.
Birdsell, D. S. & Groarke, L. (1996). Toward a theory of visual argument. Argumentation & Advocacy, 33, 1-10.
Blair, J. A. (1996). The possibility and actuality of visual argument. Argumentation and Advocacy, 43, 103-113.
Blair, J. A. (2004). The rhetoric of visual arguments. In C. A. Hill & M. Helmers (Eds.), Defining visual rhetorics (pp. 41-61). Mahwah: Lawrence Erlbaum.
Bonhomme, M. (2005). Pragmatique des figures du discours. Paris: Honoré Champion Éditeur
Bonhomme, M. & Lugrin, G. (2008). L’interpretation et effets des figures visuelles dans la communication pubblicitaire: le cas de la metonymie et de la metaphore. Studies in Communication Sciences, 8(1), 238‐258.
Burkhardt, A. (2010). Between poetry and economy. Metonymy as a semantic principle. In A. Burkhardt & B. Nerlich (Eds.), T(r)opical truths. The epistemology of metaphor and other tropes (pp. 245-270). Berlin/New York: Mouton de Gruyter.
Casetti, F. & Di Chio, F. (2009). Analisi del film. Milano: Bompiani.
Chatman, S. (1990). Coming to terms. The rhetoric of narrative in fiction and film. New York: Cornell University Press.
Dornaleteche Ruiz, J. (2007). Definición y naturaleza del trailer cinematográfico. Pensar la Publicidad, 1/2, 99-116.
Dornaleteche Ruiz, J. (2009). El trailer cinematográfico: historia de un género publicitario en EE.UU. Pensar la Publicidad, 3/1, 163-180.
Dove, I. J. (2012). On images as evidence and arguments. In F.H. van Eemeren & B. Garssen (Eds.), Topical themes in argumentation theory (pp. 223-238). Amsterdam: Springer.
Dusi, N. (2002) Le forme del trailer come manipolazione intrasemiotica. In I. Pezzini (Ed.) Trailer, spot, clip, siti, banner. Le forme brevi della comunicazione audiovisiva (pp. 31-66). Roma: Meltemi.
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., & Snoeck Henkemans, A. F. (2002). Argumentation: Analysis, Evaluation, Presentation. Mahwah, NJ: Lawrence Erlbaum.
Fahnestock, J. (1999). Rhetorical figures in science. New York: Oxford University Press.
Forceville, C. (1996). Pictorial Metaphor in Advertising. London/New York: Routledge.
Forceville, C. (2007). Multimodal metaphor in ten Dutch TV commercials. The Public Journal of Semiotics, I(1), 15-34.
Forceville, C. (2009). Metonymy in visual and audiovisual discourse. In E. Ventola & A. J. Moya Guijarro (Eds), The World Told and the World Shown: Issues in Multisemiotics, (pp. 56-74). Basingstoke/New York: Palgrave MacMillan.
Forceville, C. & Urios-Aparisi, E. (Eds). (2009). Multimodal Metaphor. Berlin/New York: Mouton de Gruyter.
Groarke, L. (2009). Five theses on Toulmin and visual argument. In F. H. van Eemeren & B. Garssen (Eds.), Pondering on problems of argumentation (pp. 229-239). Berlin: Springer.
Hoven, P. J. van den (2012). The narrator and the interpreter in visual and verbal argumentation. In F. H. van Eemeren, & B. Garssen (Eds.), Topical themes in argumentation theory (pp. 257-271). Amsterdam: Springer.
Jacobs, S. (2000). Rhetoric and dialectic from the standpoint of normative pragmatics. Argumentation, 14, 261-286.
Johnson, R. H. (2003). Why ‘visual arguments’ aren’t arguments. Informal logic, 25, 1-13.
Kernan, L. (2004). Coming attractions. American Movie Trailers. Austin: University of Texas Press.
Kjeldsen, J. E. (2007). Visual argumentation in Scandinavian political advertising: A cognitive, contextual, and reception oriented approach. Argumentation and Advocacy, 43, 124-132.
Kjeldsen, J.E. (2012). Pictorial argumentation in advertising: visual tropes and figures as a way of creating visual argumentation. In F.H. van Eemeren & B. Garssen (Eds.), Topical themes in argumentation theory. Twenty exploratory studies (pp. 293-255). New York: Springer.
Lakoff, G. and M. Johnson. 2003 [1980]. Metaphor we live by. Chicago: Chicago University Press.
Maier, C. D. (2009). Visual evaluation in film trailers. Visual Communication, 8, 159.
Maier, C. (2011). Structure and function in the generic staging of film trailers. A multimodal analysis. In R. Piazza, M. Bednarek & F. Rossi (Eds.) Telecinematic discourse. Approaches to the language of films and television series (pp. 141-158). Amsterdam: John Benjamins.
Mazzali-Lurati, S. & Pollaroli, C. (2014). A pragma-semiotic analysis of advertisements as multimodal texts: a case study. In A. Maiorani & C. Christie (Eds.), Multimodal Epistemologies: Towards an Integrated Framework (pp. 66-82). London: Routledge.
Metz, C. (1989). Semiologia del cinema: saggi sulla significazione nel cinema. Milano: Garzanti.
Moeschler, J. & Reboul, A. (2009). Pragmatique du discours: dix ans après. Studia Universitatis Babeş-Bolyai, Philologia, Liv. 4, 5-28.
Nerlich. B. (2010). Synecdoche: A trope, a whole trope, and nothing but a trope?. In A. Burkhardt & B. Nerlich (Eds.), T(r)opical truths. The epistemology of metaphor and other tropes (pp. 297-319). Berlin/New York: Mouton de Gruyter.
Panther, K. U. & Radden, G. (Eds.) (1999). Metonymy in language and thought. Amsterdam/Philadelphia: John Benjamins.
Ortony, A. (Ed.) (1993). Metaphor and thought. Cambridge/new York: Cambridge University Press
Pateman, T. (1980). How to do things with images. An essay on the pragmatics of advertising. Theory and Society, 9(4), 603-622.
Pollaroli, C. & Rocci, A. (forthcoming). The argumentative relevance of pictorial and multimodal metaphor in advertising. To appear in Journal of Argumentation in Context. Quintilian, Institutio Oratoria.
Rigotti, E. (2005). Congruity theory and argumentation. Studies in communication sciences, 5, 75-96.
Rigotti, E. & Greco Morasso, S. (2010). Comparing the Argumentum Model of Topics to other contemporary approaches to argument schemes: the procedural and material components. Argumentation, 24(4), 489-512.
Rigotti, E. & Rocci, A. (2006). Towards a definition of communication context. Foundations of an interdisciplinary approach to communication, 6(2), 155-180.
Ripley, M. L. (2008). Argumentation theorists argue that an ad is an argument. Argumentation, 22, 507-519.
Rocci, A. (2005). Connective predicates in monologic and dialogic argumentation. Studies in communication sciences, 5, 97-118.
Rocci, A. (2008). Analysing and evaluating persuasive media discourse in context. In M. Burger (Ed), L’analyse linguistique des discours des medias: théories, méthods et enjeux (pp. 247-284). Québec : Nota Bene.
Rocci, A. (2009). Manoeuvring with voices: the poplyphonic framing of arguments in an institutional advertisement. In Frans H. van Eemeren (Ed.), Examining argumentation in context. Fifteen studies on strategic menoeuvring (pp. 257-284). Amsterdam: John Benjamins.
Rocci, A., Mazzali-Lurati, S. & Pollaroli, C. (2013). Is this the Italy we like? Multimodal argumentation in a Fiat Panda TV commercial. In B. Pennock-Speck & M.M. del Saz Rubio (Eds.), The multimodal analysis of television commercials (pp. 157-187). Valencia: Publicacions de la Universitat de València (PUV).
Slade, C. (2002). Reasons to buy: the logic of advertisements. Argumentation, 16, 157-178
Slade, C. (2003). Seeing reasons: visual argumentation in advertisements. Argumentation, 17, 145-160
Smith, V. (2007). Aristotle’s classical enthymeme and the visual argumentation of the twenty-first century. Argumentation and Advocacy, 43, 114-123
Tindale, C. (2004). Rhetorical argumentation. Principles of theory and practice. Thousand Oaks, CA: Sage Publications.
Urios-Aparisi, E. (2009). Interaction of multimodal metaphor and metonymy in TV commercials: Four case studies. In C. Forceville & E. Urios-Aparisi (Eds.), Multimodal metaphor (pp. 95-117). Berlin/New York: Mouton de Gruyter.
Villacañas, B. & White, M. (2013). Pictorial metonymy as creativity source in “Purificación García” advertising campaigns. Metaphor and the Social World, 3(2), 220-239.
Whittock, T. (1990). Metaphor and film. Cambridge: Cambridge University Press.
Wildfeuer, J. (2014). Neugier. Assoziation. Erinnerung. Zur imaginativen Leistung der Trailer in Das weiße Band. In H. P. Preußer (Ed.), Anschauen und Vorstellen. Gelenkte Imagination im Kino (pp. 317-337). Marburg: Schüren.
Wierda, R. & Visser, J. (2012). Direct-to-consumer advertisements for prescription drugs as an argumentative activity type. Journal of Argumentation in Context, 1(1), 81-96.
Yu, N. (2009). Nonverbal and multimodal manifestations of metaphors and metonymies: A case study. In C. Forceville & E. Urios-Aparisi (Eds.), Multimodal metaphor (pp. 119-143). Berlin/New York: Mouton de Gruyter.

Abstract: This paper aims at (1) introducing a teaching method of peer evaluation for argument especially for students who learn debating for the first time, and (2) examining their learning process. The curriculum consisted of fifteen classes (90 minutes) for a half-year period, and was used for college freshmen in the engineering department. After the classes, most students understood the importance of peer evaluation, and the average score of self-recognition toward peer evaluation became higher.
Keywords: Peer evaluation, College freshman, Debate, Argument, Learning process

1. Introduction
Recently, in Japan, argument education has drawn increasing attention from elementary to higher levels, as a means of cultivating argumentative skills as well as developing human resources in a globalized world. Argument skill is recognized as the framework which reflects thinking skills or thinking processes (Tomida & Maruno, 2004). Teaching how to argue with peers is the one of the important goals in higher education. In those classes, peer evaluation is sometimes introduced to improve learner’s individual ability as well as to develop community of practice. Nakano (2007) found that to cultivate argument skills learners need to learn the stratified argument skills step by step and apply those skills to specific appropriate situations. Through peer evaluation, learners can accumulate the knowledge and skill of argument by exchanging comments with each other. It helps learners to foster self-understanding about what they have learned and have not learned. Learners acquire the viewpoint of evaluator and find their own task, which leads deep understanding on complicated phenomena of argument (Nakano, 2013).

Previous research reported that peer evaluation is effective as a way to educational evaluation based on the new ability evaluation (Cousins & Whitmore，1988). Along with the popularization of E-Learning, a lot of programs and systems include evaluation in the learning process of WBT (Web Based Training). In ordinary classes, peer-evaluation and self-evaluation are used in bulletin board system (Nakahara et al., 2002), video-on-demand and web-database. These effects were tested in the research by learners’ satisfaction and motivation toward classes. However, empirical studies about how to teach peer evaluation in argument are scarce and its effect has not been sufficiently tested yet. The problem here is that teachers who have tried debate education experienced difficulties, as stating opinions to others is sometimes too hard for Japanese students mentally and technically (Inoue & Nakano, 2006; Nakano & Maruno, 2012).

The authors have done research on the new system of argument education using peer evaluation in these years. Nakano (2012) described the importance and the way of peer evaluation. In the author’s laboratory, the research on the effects of peer evaluation were conducted in 2011 (Hirata, 2012) and in 2012 (Shibata, 2013) Based on these studies, this paper aims at (1) introducing a teaching method of peer evaluation for argument especially for college students who learn debating for the first time, and (2) examining their learning process for two years.

2. The tool for peer evaluation for argument
2.1 Goal and criteria
The goal of peer evaluation is to foster students’ evaluation skills as well as their self-evaluation skills. As criteria, the two main categories “manner” (the content of argument) and “matter” (how to convey ones idea) in argument were selected for a tool for peer evaluation. Each category has five subordinate skills. The evaluation system with ten items of two categories was developed. It is simple and easy so that novice students take only 5 minutes to complete the evaluation. The system can be used as peer-evaluation as well as self-evaluation.

Table 1 The criteria of peer-evaluation for argument
*1 poor, 2 fair, 5 excellent

Table 1 shows the tool for peer evaluation for argument. The five items of “manner” are “voice production,” “speed,” “tone of voice,” “pause,” and “eye contact”. The ones of “matter” are “clearly stated claim,” “reasonable reason,” “example and data,” “organization,” and “interest”. Those items were extracted by the result of the author’s fifteen-year observation research for novice students. They are the items the novice students had common problem when they spoke in front of others. For quantitative evaluation, Five-point scale is used for evaluation; 1 is poor, 2 is fair, 5 is excellent. Along with this evaluation, students write about “good point” and “needs improvement” in free description as qualitative evaluation.

2.2 Procedure
2.2.1 Four steps of peer evaluation using a worksheet
There are four steps in peer evaluation. A worksheet is prepared according this procedure (see Appendix 1). The worksheet contains the following seven questions. Using this format of worksheet, the themes the students discuss were changed every class.

Q1: Please write your own opinion about “High School uniform should be abolished in Japan”
Q2: Please make a presentation using Q1 and evaluate members’opinion.
Q3: Please write the evaluators’’ comment about your opinion.
Q4: Please set your goal for the next presentation considering Q3.
Q5: Please analyze the best presentation in your group.
Q6: Did you change your opinion after sharing others’ opinion?
Q7: Why did you change, or didn’t you change in Q6?

In peer evaluation, first, a teacher makes a small group and decides a resolution. Students write their opinions in a worksheet in five to ten minutes (Appendix 1, Q1; step 1). After preparation, students decide the order of presentations in the group and they each make a presentation in about ten minutes. Students who are not presenters take memos and evaluate the presentation by filling in a worksheet (Q2; step 2). After presentations, students evaluate themselves, write about good points and improvement needed, and share the evaluation in the group in ten minutes. (Q3; step 3). Lastly, students discuss the gap between evaluations, set a goal about manner and matter, and analyze each other’s opinion (Q4-7; step 4). After all the groups finish, a teacher and students discuss consistency and fairness of evaluation in the class. To improve students’ skills, the teacher tells students to focus on the result of highest scores as strong points, and lower scores get close to the average.

2.2.2 Small step learning of manner and matter
As introduction, to learn peer evaluation effectively, two categories of manner and matter were used separately for the first time. After students used each category of five items, the complete version of ten items for manner and matter was used. When using a separate version for introduction, first manner and second matter is most effective, as students can evaluate manner base on their objective judgment. On the other hand, matter needs experience to judge the content. In the peer evaluation, the procedure is the same in manner and matter, so students can concentrate more on what they evaluate and get used to it.

2.3 Function and value
2.3.1 Understanding the gap between various evaluations
After exchanging ideas in a small group with around four students, students evaluate others’ presentations and their own as self-evaluation using the format shown in Table 1. When the group consists of four, one student will have three evaluations from others. The students can learn the variety of evaluations from others, and the gap between others’ evaluation and own self-evaluation at the same time. These multiphase feedbacks help students make an adjustment for improvement and understand what argument is.

2.3.2 The community of practice
Peer evaluation is effective to develop the community of practice in the class. Before introducing systemized peer evaluation, most of the students had trouble in making presentations and evaluations to unfamiliar students. A teacher explained that the importance of peer evaluation is not for just criticizing others, but respect other’s good points and improve by learning from others. Exchanging evaluations is the important part of communication, even though it is hard to say. In the class, the teacher always make consideration toward the students’ mood and tells them when they say something wrong.

3. Method
To clarify the change of students in the long term, the two research studies in two years were conducted. Research 1 is based on Hirata (2012), and research 2 is Shibata (2013).

Research 1
The questionnaire research was conducted in the subject of “Communication I” which aimed at cultivating debating skills and logical thinking for freshman in Fukuoka Institute of Technology. The number of students were 36 (M=36, F=0). After experiencing peer evaluation in the prior four classes, they answered the questionnaire in ten minutes after the class on June 16th, 2011.

This paper reports the result of one question for comparison with research 2. Question 1 is about the attitude toward peer evaluation. 1-1 “I’m good at peer evaluation”, 1-2 “I like peer evaluation”, 1-3 “everyone can learn peer evaluation”, 1-4 “I’d like to improve based on PE”, 1-5 “peer evaluation is important”, 1-6 “peer evaluation is useful in the future”.

Research 2
The second research study was conducted in the “Presentation” which aimed at cultivating presentation skills for sophomore students. The number of students were 40 (M=40, F=0). Most students are the same as the research 1. In the class, peer evaluations were used. To test the changes more closely, two questionnaire research studies were conducted after the first presentation at the middle stage (on May 17th and 24th, 2012) and second presentation at the final stage (July 12th and 19th, 2012) each taking ten minutes.

Question 1 is the same as research 1. In addition, this paper reports two more questions for further analysis. Question 2 is about the object of peer evaluation, and Question 3 is about the image of peer evaluation.

4. Result and discussion
4.1 Quantitate analysis of Question 1
Fig.1 shows the results of Question 1 conducted in research 1 and 2. The average scores of research 1 were as follows: 1-1，2.5(SD=.97) ; 1-2, 2.8(SD=.96) ; 1-3, 3.6(SD=.87); 1-4, 4.2(SD=.72) ; 1-5, 4.3(SD=.73) ; 1-6, 4.3(SD=.77). These results clarify that most of the students feel “they are not good at peer evaluation” and “they don’t like peer evaluation”, although they recognize the importance and it is needed for the future, and have motivation. At the time of research study 1, students only experienced peer evaluation four times in the classes, so they might have been unfamiliar with the new communication style of peer evaluation. This result implicates that the tool and system of peer evaluation proposed in this paper contributed to their learning in the classes.

The results of Research study 2 in the middle were as follows: 1-1, 2.5(SD=.86) ; 1-2, 2.6(SD=.87); 1-3, 3.9(SD=.87); 1-4, 4.1(SD=.88); 1-5, 4.1(SD=.89); 1-6, 4.2(SD=.90). The results of Research study 2 in the final were: 1-1, 3.0 (SD=.76); 1-2, 3.1(SD=.76); 1-3, 4.2(SD=.75); 1-4, 4.3(SD=.76); 1-5, 4.3(SD=.77); 1-6, 4.2(SD=.78). These results showed all the scores of Research 2-middle and final increased except for 1-6. The score of 1-1 and 1-2 which were lowest in average in the result 2-middle, increased most plus 0.5 point in each. These results show that the attitude changed positively through the presentation classes.

Fig. 1 The attitude toward peer evaluation in debate
*PE=peer evaluation

Comparing all the data of Research 1, 2-middle and 2-final in Table 1, we can see the gradual increase overall. There are three patterns in the result. One is the characteristic of 1-1 and 1-2, which are lowest of all and changed drastically through two years. Another is the items of 1-4, 1-5, and 1-6 which are highest and change little. The other is 1-3, which increased as the students became more experienced in the classes. Overall the scores of sophomore in Research 2-middle and final are higher than freshmen in the Research 1, which implicates that the learning using the tool and system of peer evaluation succeeds in helping students become motivated in the classes.

4.2 Qualitative analysis of Question 2 to 6
The answers of free description on Question 2 to 6 can be summarized as follows:

Question 2 What is the object of peer evaluation?
* To develop one’s merit and improve one’s demerit by cooperating with others
* To get interested and listen actively to others’ opinions
* To notice what I haven’t noticed by myself

Question 3 What is the image of peer evaluation?
* The good chance to reflect on myself
* To improve my skill
* To know my bad points
* I don’t have good image toward peer evaluation as I’m not good at evaluating others.

Question 4 Do you think you changed the image of peer evaluation from freshmen?
* I don’t know.
*I don’t remember.
*I had trouble in evaluating others when I was a freshman, but now I’ve gotten used to it and think deeply in peer evaluation

Question 5 When do you think you do peer evaluation in daily life?
* When I study with my friends
* In conversation
* Discussion watching TV news
* In driving

Question 6 What is the merit and demerit of peer evaluation?

About Question 2 and 3, these results show that most of the students understand the reason why they learn peer evaluation in the class, effectiveness in improving skills, and understanding others. Peer evaluation helps students concentrate on listening to others as they need to evaluate. This is one of the important factors in argument education in Japan. About Question 4, as stated in 4.1, the recognition toward peer evaluation became better and one of the students answered that he overcame the trouble in evaluating others and could concentrate much more on evaluation. In regards to Question 5, there are various answers and some students do peer evaluation in daily life, but others don’t. These differences in daily-life communication might affect the individual differences in the classes. As for Question 6, there are a lot of merits and demerits dividing evaluator and presenter. This result shows that the students understand the meaning of peer evaluation, but they consider it might a break relationship between classmates. Japanese students are hesitant to say their opinion directly. This problem is because they are not confident in what they feel or think enough to tell others.

Fig. 2 The three level of the objectives of peer evaluation (Nakano, 2012)

Nakano (2012) found that there are three levels of the objectives of peer evaluation shown in Fig.2: 【A】exchanging information about good and bad points of other’s skill and opinion (Sharing Information Level),【B】deeply understanding others’ opinion and value (Mutual Understanding Level), 【C】developing oneself by sharing and assimilate information and knowledge (Develop Level). The result of this paper follows this model. By developing the tool and system of peer-evaluation for Japanese novice students, most of the students feel positive toward peer evaluation. In the process, they have changed from just sharing information to gradually understanding others, and finally they develop themselves using the experience of peer evaluation. The result shows some students still feel trouble in evaluating others. This is caused by inexperience in their lives. These individual differences need to be researched.

5. Conclusion
This paper aimed at (1) introducing a teaching method of peer evaluation for argument especially for college students who learn debating for the first time, and (2) examining their learning process for two years. As for the attitude of students toward peer evaluation, they were getting used to evaluating each other. Through peer evaluation, they seemed more concentrated on arguments by listening to others’ opinion. At the same time, they judged their own opinion standing on the viewpoint of evaluator by evaluating others. These changes in the process are the essential points of peer evaluation. According to the results, the system for peer evaluation proposed in this paper fit the needs and levels of the students and worked properly as a tool for learning argument. On the other hand, some students still have a hard time in peer evaluation and lose confidence. In the future, a more systematic approach for the students who are not positive toward peer evaluation is needed.

References
Cousis, J.B. & Whitemore, E. (1998) Framing participatory evaluation. New Directions for Evaluation, 80: 5-23.
Hirata, K. (2012). The research on the peer evaluation in the “Communication I class” for college students. Fukuoka Institute of Technology. B.A. thesis. (in Japanese)
Inoue, N., & Nakano, M. (2006). The costs and benefits of participating in competitive debate activities: Differences between Japanese and American college students. In F. H. van Eemeren, M. D. Hazen, P. Houtlosser & D. C. Williams (Eds.), Contemporary Perspectives on Argumentation: Views from the Venice Argumentation Conference (pp.167-184). Amsterdam: Sic Sat.
Nakahara, J., Urashima, N., Nishimori, T., Suzuki, M., Imai, Y., Yamagiwa, K., & Nagata, T. (2002) The Development and Evaluation of Bulletin Board System Which Enables Learner’s Mutual Evaluation to Their Messages. Japan Society for Educational Technology, 26 (Suppl), 33-38. (in Japanese)
Nakano, M. (2007). How college students acquire argumentative skills in a community of practice. Cognitive Studies, 14(3), 398-408. (in Japanese)
Nakano, M. (2010). Introductory communication for college freshmen. Kyoto: Nakanishiya.
Nakano, M. (2012). Students Perception on Peer Evaluation in the Class for Argument Skills. Japan Society for Educational Technology,  37 (Suppl), 185-188. (in Japanese)
Nakano, M. (2013). Discussion: fostering students learning from each other in academia. In E. Tomida, & A. Tajima (Eds.), University Education: Psychological theories and practices for developing cross-boundary explanations (pp.111-126). Kyoto: Nakanishiya.
Nakano, M. & Maruno, S. (2012).The effect of debate training on argumentative skills: The developmental process of Japanese college students. Studies for the Learning Society. Issue 4.
Shibata, K. (2013). The research on the self and peer evaluation in the presentation class for college students. Fukuoka Institute of Technology. B.A. thesis. (in Japanese)
Tomida, E., & Maruno, S. (2004) Empirical Findings and Theoretical Background of Argument as Thinking. Japanese Psychological Review, 47 (2):187-209 (in Japanese)
Tomida, E., Mizukami, E., Morimoto, E., & Otsuka, H. (2010-2011). Learning Environment for Undergraduates to Enhance Spontaneous Growth of Communication Skills: Effects of group size. Educational technology research, 33(1・2), 121-130.

APPENDIX 1

Abstract: Climate scientist James Hansen’s use of we call ‘hinge terms’ – such as ‘dangerous’ and ‘tipping point’- operate to reconfigure argumentation on global warming by pre-scripting headlines of media coverage on scientific findings. Study of this case stands to elucidate an understudied aspect of the global warming controversy, as well as contribute to understanding of how ‘argument operators’ function to relocate arguments into different contexts, with potential implications for argumentation theory.
Keywords: global warming, argument activity type, rhetorical figures, James Hansen, rhetoric of science

1. Introduction
The intellectual roots of American argumentation scholarship intertwine with the tradition of public address criticism, a fact that helps account for the centrality of context in the work of prominent American scholars of argument (e.g., Newman 1961; Zarefsky 1990). The recent launch of the Dutch journal Argumentation in Context, along with a new book series by the same name, provides an occasion to explore how the American approach to criticism of public argument in situated contexts relates to new features of pragma-dialectics that emphasize contextual features of argumentation, such as the concept of “argumentative activity types” (van Eemeren & Houtlosser 2009).

Considerable attention has been devoted in pragma-dialectics to understanding how context may “discipline” norms for judging the soundness of arguments that unfold within a particular argumentative activity type (van Eemeren & Houtlosser 2009, p. 15). Left understudied, however, is the question of what happens when an argument shifts from one activity type to another, and further, what moves by interlocutors might spur, or block, such shifts.

We use the term “argument operators” to refer to detectable moves that change argument modalities. Our focus here is on operators that relocate arguments within different normative contexts. While context is featured in various ways within the literature of argumentation (e.g. fields, argumentation activity types), it is normally taken to be a form of pre-figured ground that constrains or regulates what is possible within the given context. Our focus differs in that it calls attention to argumentative strategies that relocate an existing argument within a different context, thereby changing the norms and constraints that pertain to the argument.[i]

The specific argument operator that is our concern here is what we call the “hinge term,” and the case of climate scientist James Hansen’s argumentation on global warming provides an apt point of departure for our inquiry. The effect of the hinge term, as one type of argument operator, we contend, is to significantly affect the tenor and trajectory of climate change arguments. In particular, Hansen’s controversial use of hinge terms such as ‘dangerous’ and ‘tipping point’ in his peer reviewed journal articles operate to pivot his argumentation on global warming from the context of professional scientific discourse into the context of general public argument. In what follows, background on the Hansen case (in part two) paves the way for critical analysis of his strategic deployment of hinge terms (in part three). Part four draws lessons from the case study to sketch a speculative taxonomy of argument operators and open discussion about the possible utility of the concept. A concluding section reflects on how our intervention relates to ongoing work on argument context in pragma-dialectics.

2. From reticence to witnessing
Widely considered to be one of the world’s leading climate scientists, James Hansen began his research career by exploring how particulate matter in the Earth’s atmosphere refracts light from lunar eclipses (Matsushima, Zink & Hansen 1966). Shortly after completing his Ph.D. thesis at the University of Iowa, which dealt with properties of Venus’s clouds, Hansen realized that many of the same dynamics driving changes in Venus’s atmosphere might also be occurring on Earth. A decade of work from 1978 to 1988 that involved building a complex computer model of the Earth’s atmosphere led to Hansen’s first major public appearances as a scientist. As an official witness before the U.S. Congress during 1988 and 1989, Hansen declared with “99 percent confidence” that human carbon dioxide emissions were causing long-term warming in the Earth’s atmosphere.
Hansen’s bombshell congressional testimony provoked intense controversy and earned him the moniker “grandfather of climate change” – a role the scientist was not quick to embrace (McKie 2009). Following his first big splash as a public figure, Hansen (2009) “was firmly resolved to go back to pure science” and leave media appearances to “people who were more articulate and seemed to enjoy the process” (p. xvi). This retreat to the laboratory was consistent with Hansen’s (2007a) perspective on “scientific reticence,” a default rhetorical posture for scientists that involves a tendency to understate claims and emphasize the uncertain, open-ended nature of scientific knowledge (see also Ziman 2000).

For nearly a decade following his blockbuster congressional testimony in 1988/1989, Hansen practiced scientific reticence, publishing findings from his work on Global Circulation Models in peer reviewed journals and eschewing opportunities to appear in the media spotlight. Yet that posture changed in 1998, when Hansen agreed to participate in public debates on global warming with climate “contrarians” Patrick Michaels in New York City, and Richard Lindzen in Cambridge, Massachusetts (Mitchell & O’Donnell 2000). As Hansen (2009) explained his motivation for stepping out of the laboratory and into the public square for the first time since his famous congressional testimony, “I wanted to present and publish a table of the key differences between my position regarding global warming and the position of the contrarians” (p. xvi).

Hansen’s participation in the New York debate marked a turning point in his career, as afterward he increasingly embraced the role of a “public witness” to the dangers of global warming, especially following the birth of his first grandchild in 1999 (Hansen 2009, p. xii). This path would eventually lead to Hansen’s appearances at rock concerts and protest demonstrations with climate change activists (Eilperin & Mufson 2013). Also during this period, a subtle shift in the rhetorical arc of his scientific papers could be detected. For example, in an article published in Atmospheric Chemistry & Physics, Hansen and colleagues (Hansen et al. 2007a) repeatedly use the terms ‘tipping point’ and ‘dangerous’ to describe global warming. Major news outlets parroted those terms in headlines during the news cycle in which the paper was published:
* “Research finds that Earth’s climate is approaching ‘dangerous’ point” (PR Newswire 2007);
* “NASA Research Suggests Earth climate approaching dangerous point” (Space Daily 2007);
* “Earth nears tipping point on climate change.” (Spotts 2007)

Messages of danger are part of the stock and trade of newspapers, so Hansen was in effect pre-scripting headlines for general circulation and pivoting toward a different context and rhetorical stance. As a scientific argument became a public argument, the assessment of “facts” would move into a normative environment where questions of “value” and policy response would predictably arise. This netted Hansen a broader audience, but it complicated his voice as a scientist. Was he now acting as an advocate?

3. ‘Dangerous’ and ‘tipping point’ as hinge terms
By 2007, Hansen had become engaged fully in the rhetorical project of trying to invent ways of communicating the gravity of what he called the ‘climate catastrophe’ (2007b) to broader publics. In one open communication on his personal website, he mused:
A related alternative metaphor, perhaps less objectionable while still making the most basic point, comes to mind in connection with an image of crashing of massive ice sheets fronts into the sea – an image of relevance to both climate tipping points and consequences (sea level rise). Can these crashing glaciers serve as a Krystal Nacht, and wake us up to the inhumane consequences of averting our eyes? Alas, that metaphor probably would be greeted with the same reaction from the people who objected to the first. That reaction may have been spurred by the clever mischaracterization of the CEO, aiming to achieve just such a reaction. So far that seems to have been the story: the special interests have been cleverer than us, preventing the public from seeing the crisis that should be in view. It is hard for me to think of a different equally poignant example of the foreseeable consequence faced by fellow creatures on the planet. Suggestions are welcome. (Hansen 2007c)

This candid reflection laid bare for Hansen a fundamental dilemma facing scientists working on politically charged topics. The tradition of scientific reticence counsels restraint, yet the ethical calling to bear witness may demand more strident rhetoric. Ultimately, Hansen and colleagues settled on the terms ‘dangerous’ and ‘tipping point’ as red flags to heighten salience of the issue. While Hansen personally deployed such terms increasingly during public appearances, he also worked with his co-authors to pepper their scientific papers with these terms. For example, the previously mentioned Atmospheric Chemistry & Physics paper (Hansen et al. 2007) features 36 mentions of ‘dangerous’ in various contexts (see Table 1).

Table 1. Mentions of “dangerous” in Hansen et al. (2007). References with quotation marks are in blue, while references without quotation marks are in red.

Notably, the first four mentions of the term “dangerous” on the paper’s first page are accompanied by quotation marks, indicating perhaps some hesitance regarding use of the term. However, in the final five pages of the paper, these quotation marks drop out and dangerous appears as an unqualified adjective in 9 of 16 instances. In the penultimate discussion section, all three mentions of dangerous appear without quotations. This progression may reflect a common tendency of authors to move from a tentative to a more authoritative voice as their papers develop (Fahnestock 1998; Holmes 1997; Peacock 2002; Ruiying & Allison 2003), yet such maneuvers did not escape the notice of the peer review referees. In an interactive comment published in Atmospheric Chemistry & Physics Discussions, Hansen and colleagues (2007, p. S7351) note that referee #1 “expressed mild concern about terms such as ‘dangerous anthropogenic interference,’ ‘disruptive climate effects,’ and ‘tipping points.’”

Some of this pushback may have stemmed from the sheer number of ‘dangerous’ references in the paper. As Jeanne Fahnestock (1999, pp. 160-172) observes, strategic repetition of key terms (characterized by the classical rhetorical figure of ploche) can heighten the impact of scientific argumentation on audiences. Fahnestock points to Charles Darwin’s deployment of “subtler repetitions that declare identity in reference or the interconnections among phenomena” to illustrate how ploche can operate to heighten, in the terminology of Perelman and Olbrechts-Tyteca (1969, p. 144), a scientific argument’s ‘presence.’ Just as Darwin eschewed mere repetition, instead weaving different meanings through recurrent references, Hansen and colleagues deploy ‘dangerous’ in an array of different usages, in effect producing a pedagogy of dangerousness from which journalists could learn.

Previous scholarship has explored some of the rhetorical entailments associated with Hansen’s use of terms such as ‘tipping point.’ For example, Russill (2008, 2010) notes that in the global warming controversy, ‘tipping point’ tends to invoke the interests of future generations, as irreversible, runaway climate change would be most harmful to those not yet born. Yet as Figure 1 illustrates, concepts from the rhetorical tradition furnish a set of transformations that point to ways that Hansen’s hinge term strategy may carry even broader implications.

Figure 1. Rhetorical concepts illustrate ways that the hinge term ‘dangerous’ enables discourse to swing from one activity, genre, stasis or stance to another.

As the far left column suggests, successful deployment of the hinge term ‘dangerous’ enables discourse to swing from the argument activity type of scientific peer review to a different one – general public argument. In a related transformation, alterations in types of questions asked and goals pursued by interlocutors are marked by a shift in rhetorical genre. Whereas scientific discourse tends to follow patterns of reasoning associated with the forensic genre (rooted in the rhetorical tradition of adjudication in the law courts), public argument tends to feature epideictic (ceremonial) and deliberative (political decision-making) forms of reasoning (Fahnestock 1998). As the discourse migrates in this fashion, a further element of transformation occurs at the level of rhetorical stases, with argumentation “pulled” (Walsh 2010) from stases of fact and definition, into a different stasis point in which interlocutors debate how contingent value judgments relate to possible future courses of action. Aligned with all of these transformations is a concomitant shift evident in “rhetorical stance” (Booth 1963), as Hansen himself moves from self-identifying as a “reticent scientist” to a “public witness.”

4. Argument operators: nudge, pivot, or jump
The Hansen case calls to the fore three possible approaches to context-switching, and the response to Hansen sorts out to some degree according to which of these the audience senses his speech acts are aspiring to do. To identify these shifts we use intuitive language – common verbs, not adjectives – rather than terminology that aims for technical precision. These operators, we suggest, can do the following:
* Nudge an argument into a wider or narrower context, thereby expanding the range of rational strictures on relevance (see Walton 2003), but without introducing competing or conflicting accounts;[ii]
* Pivot strategically between competing or complementary contexts of rational assessment;
* Jump to an alternative context.

The first two of these may serve as bridges from one context to another, whereas the third makes a leap. The response to Hansen seems to depend in large part on which of these his readers are sensing. Being both a scientist and a citizen, Hansen might see his repeated invocation of the term ‘dangerous’ as a way of nudging his audience into a wider context that encompasses science but also the field of citizen action. While we do not reject such a characterization, our analysis picks up on what can be seen as a strategic pivot from one generic context, with its usual strictures and enablements to another. His critics seem inclined to see Hansen’s performance as a kind of abandonment – jumping ship, so to speak – by violating constraints of a professional context in order to play out the argument in a different context. They would no doubt see that characterization underlined by Hansen’s subsequent activism. In response, Hansen might point to the fact that the term ‘dangerous’ had been utilized previously in major scientific reports on climate change, and that ‘tipping point’ language was justified because it “conveys aspects of climate change that have been an impediment to public appreciation of the urgency of addressing human-caused global warning” (Hansen et al. 2007b, p. S7351). All of these considerations come to bear as we interrogate the kind of speech acts Hansen was deploying.

With one foot in the lab and one foot in the public media, it is quite possible that Hansen could be celebrated as exemplary of the “third culture” figure, who manages to speak persuasively across the boundaries between fields of expertise and contexts of public argument, contributing to a culture that consists of both experts and non-experts, and constituted in such a way that effective participation requires accepting the legitimacy of both empirical and interpretive methods (Lyne 2010). In that case we would have to see him in a rather different light than some of his critics have. He would be seen as crafting a distinctive voice that bridges, or “nudges” toward a more encompassing audience.

Because we regard arguments as something more than meaningful texts, we take their meaning to function in relation to human action. On this approach, the wider investigative terrain for argument operators is suggested by the speech act vocabulary, following John Searle (1969) and other speech act theorists (Austin 1975; van Eemeren & Grootendorst 1984), of locution, illocution, and perlocution which we translate as ways of posing the following general questions:
* Locution: What are you saying?
* Illocution: What are you doing?
* Perlocution: What effect are you having?

As we move down the list, each function presupposes what is listed above. That is, someone says in order to do in order to have some kind of effect. The last category, perlocution, can be variously understood as an actual consequence of an illocution, as an intended consequence of an illocution, or as a rationally foreseeable consequence of an illocution. We do not wish to exclude any of these from our consideration of ‘argumentative effects,’ that is, of the way speech acts influence ongoing or subsequent arguments or argumentative moves. Thus, the purview of this analysis would be possible interactions that can be taken as specifically relevant to an argument, but it would not include any other kind of effect (e.g. hurt feelings, anger, delight).

In reference to the “hinging” we are looking at here, the hinge effect is performed at all three levels of the speech act. In saying that conditions are dangerous, Hansen is making a shift in the argumentative context, with the effect that a number of entanglements – ranging from genre relevance to contestation of appropriateness – begin to work at once. But the nature of the shift is such that it can be interpreted in several different and contestable ways (see Figure 2).

Figure 2. Hinge term dynamics. Hinge terms enable arguments to swing between genres and stases, with associated shifts in the arguer’s rhetorical stance, artifact produced and activity type.

Returning to a generic reconstruction of the diagram presented in the previous section, we see five categories that that appear to move in unison. This, we believe, is why the instance of deploying a hinge term particularly invites attention. Understood as a speech act, Hansen’s repeated references to danger would reasonably be taken as a warning. And whether by intention or not, the illocutionary act of warning within one context has the perlocutionary consequence of pivoting the argument into another discourse frame. As the warning of danger breaks out of the confines of the presumptive scientific stasis, it produces a secondary perlocutionary effect of moving from “fact” to “value.” Moreover, the shift of stases has a gravitational pull that brings changes within each of the other categories (Walsh 2010). The text is now recontextualized as public argument, where it stirs controversy, and signals differences of genre, stasis, stance, and artifact. To the consternation of many, the line between science and public controversy begins to dissolve.

5. Scoping out the landscape
We have argued that the hinge term, in its functional sense, inflects toward, or toggles between, different registers of argument. What it “means,” in the most robust sense, is therefore what it does when affirmed or invoked, that is, what it does when introduced by a speech act. What it does to the argument is something more consequential than a mere figure of speech. In this case, for instance, the terms “danger” and “tipping point” cannot be sequestered as mere metaphors. Rather than non-literal flourishes, the introduction of such hinge terms into an argument is a speech act (or a set of speech acts) with the capacity to move arguments in a different direction, specifically toward different cognitive, affective, cultural, semiotic, or praxial registers. This can be done either as a deliberate strategy or an unintended consequence of the introduction of the term, and its consequences can be both foreseen and unforeseen. One of the consequences in this case was an inflection toward arguments about the objective limits of science and accusations that these terms had taken Hansen into a “subjective” frame, where their purchase as scientific claims were questioned.

In describing argument operators specific to this case, we have introduced strategies of context-shifting. Beyond the hinge term and related context shifters, this case leads us to anticipate other argument operators that have different modal functions in argument. Hansen was criticized for acting as an advocate. If he was acting as an advocate, at what point did that voice emerge? Was it when he dropped the quotation marks when using the word “dangerous”? Or were there gradations of his shift, perhaps subtly indicated, and when he engaged in a debate before academics not in his field? Depending on how that question is answered, his arguments are likely to be judged by one set of norms or by another. In argumentation literature, we observe that arguments are generally aligned with the intention of the arguer, and it is assumed that the arguer has a unitary voice, such that that person could be held responsible for inconsistencies or implications of the argument they are making. Moreover, we assume that the author’s intentions are framed with a particular normative context in mind. This would be the standard case of having a “voice” in an argument.

It is the arguer with the unitary voice that is typically assumed in philosophical discussions of rationality. The leading advocate of philosophical “inferentialism,” Robert Brandom, speaks of personal accountability in terms of “scorekeeping,” whereby participants in an argument constantly track and update the commitments and authorizations made by either party in order to make explicit the rational purport of any utterance (Brandom 2000). This is a dynamic way of thinking about argument as process, because it depends on the relationship between present and past assertions rather than on constructs in isolation. And this is a useful way of thinking about the trail of assertions as they chain out. But in view of the shifts of context, voice, and other functions of argument operators that we have been referencing, one might well ask if it is pragmatic to think of arguments only in terms of verbalized propositions by philosophically focused interlocutors. To understand the complexity of context and its relationship to argument, it might be useful to consider whether there are a number of different scoreboards and ways of scoring that are the very things at stake in many arguments (Lyne 2013). Public address scholars, who are observant of the relationship between propositional and non-propositional features of public argument, as well as the various ways that that rational arguments may play out, have something to bring to the table in laying out argument operators.

We know arguers modulate the voice they are using to advance an argument, sometimes by “ventriloquizing” the positions of others, or laying out the position of what another would say were they in top form. This kind of voicing is perhaps most clearly apparent when a surrogate stands in for a political candidate in a debate, where the aspiration would be to offer up the arguments the candidate would or could make. Somewhat differently, a defense attorney makes the best arguments possible, not because he or she necessarily believes them, but because they are thought to support the best case that might be made in defense of the client. Other arguments, we well know, are made “for the sake of argument,” without binding the hands of the arguer. We might well ask what are the ways of shifting in an out of any given frame of time-binding accountability.

The formal framing of a staged debate or of a courtroom trial generally eliminates any ambiguity about whether the arguments presented should be seen as isomorphic with those that the arguer would be personally and ethically accountable for making. In other cases ambiguity or confusion can arise, as when arguers shift between or among voices. So it would be worthwhile to pay attention to indicators of voice shiftings. These might be found in tonal changes, changes of body language, or stylistic changes – factors that have been of interest to students of public address but have generally been backgrounded in pragmatic analysis of argument.

We have seen from pragma-dialetics that arguments play out differently in different types of argumentative activities. Here we are suggesting that even within a given argument activity a shifting of voice can change the function of an argument. So in addition to context-shifters, other argument operators may need to be fleshed out. This is among the reasons we believe that the juncture between public address studies and argument studies may enrich both.

NOTES
i. We note that the term “operators,” as defined by computer programming languages, may show some elemental similarities to the ones we are describing, in that they allow manipulations of “semantic” as well as “syntactic” properties. At present, however, the language of “genre,” “stance,” “audience,” and so on, seem reserved for natural languages used in non-computational contexts. This is not to say that these could not be represented in binary code.
ii. Here we highlight “expansion” rather than shifts, but these are not always distinct, as Burke (1945) points out in commentary on “scope and reduction” (pp. 59-117).

References
Austin, J. L. (1975). How to do things with words. Oxford: Oxford University Press.
Booth, W. C. (1963). The rhetorical stance. College Composition and Communication, 14, 139-145.
Brandom, R. (2000). Making it explicit: Reasoning, representation, and discursive commitment. Cambridge, MA: Harvard University Press.
Burke, K. (1945). A grammar of motives. Berkeley: University of California Press.
Eemeren, F. H. van & Grootendorst, R. (1984). Speech acts in argumentative discussions. Dordrecht: Foris.
Eemeren, F. H van & Houtlosser, P. (2009). Strategic maneuvering: Examining argumentation in context. In F. H. van Eemeren & P. Houtlosser (Eds.), Examining Argumentation in Context (pp. 1-24). Amsterdam: John Benjamins.
Eilperin, J. & Mufson, S. (2013). Activists arrested at White House protesting Keystone pipeline. Washington Post, February 13.
Fahnestock, J. (1999). Rhetorical figures in science. Oxford: Oxford University Press.
Fahnestock, J. (1998). Accommodating science: The rhetorical life of scientific facts. Written Communication, 15, 330-350.
Hansen, J. E. (2009). Storms of my grandchildren. New York: Bloomsbury Books.
Hansen, J. E. (2007a). Scientific reticence and sea level rise. Environmental Research Letters, 2, 1-6.
Hansen, J. E. (2007b). Climate catastrophe. New Scientist, July 28, 30-34.
Hansen, J. E. (2007c). Averting our eyes. Web communication, November 28. Retrieved from http://www.columbia.edu/~jeh1/mailings/2007/20071128_AvertingOurEyes.pdf
Hansen, J. E., et al. (2007a). Dangerous human made interference with climate: A GISS modelE study. Atmospheric Chemistry & Physics, 7, 2287-2312.
Hansen, J. E., et al. (2007b). Interactive comment on ‘Dangerous human-made interference with climate: A GISS modelE study,’ Atmospheric Chemistry and Physics Discussions, 6, S7350-S7354.
Holmes, R. (1997). Genre analysis and the social sciences: An investigation of the structure of research article discussion sections in three disciplines. English for Specific Purposes, 16, 321-337.
Lyne, J. (2013). Having a whole battery of concepts: Thinking rhetorically about the norms of reason. Social Epistemology Review and Reply Collective, 2(1), 143-148.
Lyne, J. (2010). Rhetoric and the third culture: scientists as arguers and critics. In M. J. Porrovecchio (Ed.), Reengaging the Prospect of Rhetoric (pp. 113-132). New York: Routledge.
Matsushima, S., Zink, J.R., & Hansen, J.E. (1966). Atmospheric extinction by dust particles as determined from three-color photometry of the lunar eclipse of 19 December 1964. The Astronomical Journal, 71, 103-110.
Mitchell, G.R. & O’Donnell, T. (2000). Full transcript of inaugural AARST science policy forum, New York Hilton, Friday 20 November 1998, 7-9 pm. Social Epistemology 14, 131-180.
McKie, R. (2009). ‘We have only four years left to act on climate change – America has to lead. Observer (London), January 17.
Newman, R. P. (1961). Recognition of Communist China: A study in argument. New York: Macmillan.
Peacock, M. (2002). Communicative moves in the discussion section of research articles. System, 30, 479-497.
Perelman, C. & Olbrechts-Tyteca, L. (1969). The new rhetoric: A treatise on argumentation. J. Wilkinson and P. Weaver (Trans.). Notre Dame: University of Notre Dame Press.
PR Newswire (2007). Research finds that Earth’s climate is approaching ‘dangerous’ point. May 31.
Ruiying, Y. & Allison, D. (2003). Research articles in applied linguistics: Moving from results to conclusions. English for Specific Purposes, 22, 365-385.
Russill, C. (2010). Truth and opinion in climate change discourse: The Gore-Hansen disagreement. Public Understanding of Science, 20, 796-809.
Russill, C. (2008). Tipping point forewarnings in climate change communication: Some implications of an emerging trend. Environmental Communication, 2, 133-153.
Searle, J. (1969). Speech acts: An essay in the philosophy of language. Cambridge: Cambridge University Press.
Space Daily (2007). NASA research suggests Earth climate approaching dangerous point. May 31.
Spotts, P. N. (2007). Earth nears tipping point on climate change. Christian Science Monitor, May 30, 2.
Walsh, L. (2010). Visual strategies for establishing ethos across the ‘is/ought’ divide in the IPCC’s report Climate Change 2007: Summary for Policy Makers. Poroi, 6, http://dx.doi.org/10.13008/2151-2957.1066
Walton, D. (2003). Relevance in argumentation. London: Routledge.
Zarefsky, D. (1990). Lincoln, Douglas, and slavery: In the crucible of public debate. Chicago: University of Chicago Press.
Ziman, J. (2000). Are debatable scientific questions debatable? Social Epistemology, 14, 187-199.

Abstract: This paper investigates the role of verbs of appearance as argumentative indicators analysing the uses of the Italian verb sembrare (‘seem’) in a sample of 40 texts chosen from a corpus of reviews, editorials and comment posts. An analysis conducted within the framework of the Argumentum Model of Topics, shows that the verb, in its evidential-inferential uses, indicates specific argument schemes of the symptomatic as well as the causal type.
Keywords: argumentative indicators, Argumentum Model of Topics, causal argumentation, inferential evidentiality, Pragma-Dialectics, symptomatic argumentation, syntagmatic argument schemes, verbs of appearance

1. Introduction
This paper addresses the relations between verbs of appearance and argument schemes, taking as an example the Italian verb sembrare (‘to seem’) in its function as an argumentative indicator[i]. In the framework of Pragma-Dialectics, the notion of argumentative indicators has been defined as including “all words and expressions that refer to any of the moves that are significant to the argumentation process” (van Eemeren, Houtlosser & Snoeck Henkemans, 2007, p. 2). Such argumentative clues can belong to different classes of linguistic items, ranging from verbs to conjunctions and to various kinds of discourse markers[ii]. Within Pragma-Dialectics, argumentative indicators have been considered, above all, from the point of view of the analyst facing the task of argumentative reconstruction. In this perspective, it has been underlined that indicators may work at different levels, signaling, for example, the engagement of the interactants in a particular stage of a critical discussion[iii], argumentative moves or the presence of a particular argumentation scheme. From a linguistic point of view, it is crucial to acknowledge that the usefulness of indicators for the analyst depends on their usefulness for the participants engaged in an argumentative interaction. Like other aspects of textual or conversational structure, the construction of argumentative relations at the different levels mentioned above is, in the first place, the participants’ task; functional categories are emic, not etic (Pike 1954). What justifies the attribution of an indicator function to a linguistic expression is, then, the potential of the expression to guide interlocutors and readers in this task. In any particular context, this potential will depend both on the expression’s functions coded in a relatively stable manner in the linguistic system (e.g. in the lexicon or in the domain of recurrent syntactic constructions and discourse routines) and on the specific pragmatic configuration (Bazzanella & Miecznikowski 2009) the expression is used in. As we will argue in our paper, corpus-based linguistic analysis, focused on single expressions and their contexts of occurrence, can fruitfully contribute to a better understanding of argumentative indicators in this sense.

Like other verbs of appearance interlinguistically (e.g. English to seem, Spanish parecer), the verb sembrare has been attributed an evidential function in the linguistic literature when occurring in certain syntactic and pragmatic contexts[iv]. Evidentials specify “the kind of justification for a factual claim which is available to the person making that claim […]” (Anderson, 1986, p. 274). The typological analysis of evidential systems has shown that frequently grammaticalized types of justifications for assertions, otherwise called information sources, means/ways of acquiring knowledge or modes of knowing, are direct experience (eventually distinguished according to perceptual modality), inference, and report/hearsay (cf. Willett 1988). Research on lexical evidentials (e.g. Squartini 2007) suggests that these cognitive categories are relevant also in linguistic systems that do not grammaticalize evidentiality, and it is in this line of thinking that the notion of evidentiality is currently used to analyze the semantics of appearance verbs.

Evidentiality and argumentation are related because the justification of claims is, of course, the defining feature of one of the central moves in argumentative discourse. However, an important difference between evidentially marked utterances and full-fledged argumentative moves is that, in the former case, the speaker signals the presence of evidence in favor of his or her assertion and categorizes that evidence in a generic fashion, whereas in the latter case, the speaker establishes a discourse relation between the assertion and one or more specific arguments given in the text. By consequence, speakers can use evidentials both to support argumentation, contributing to establish argument-conclusion relations present in a critical discussion, and as an alternative to argumentation, merely suggesting the relevance of evidence without actually formulating any arguments. Recent studies at the semantic-argumentative interface (Miecznikowski, 2011; Rocci, 2008, 2012, 2013) have concentrated on the argumentation supporting function of modal and evidential expressions, arguing that, in argumentative contexts, these expressions function as indicators strengthening and categorizing argument-conclusion relations. One of the basic ideas is that the evidential categorization of modes of knowing in an utterance restricts the range of argument schemes with which the utterance is compatible. In the present analysis, we will develop this idea, showing that sembrare constructions preferentially occur with certain argument schemes and insisting in the role of the verb’s lexical meaning at this regard. Argument schemes will be analyzed and reconstructed using the Argumentum Model of Topics (Rigotti & Greco Morasso 2010).

In section 3, after having presented our data, we will provide an overview of the syntactic constructions of sembrare associated with evidential meanings and explain why these constructions are good candidates to function as argumentative indicators. We will then focus on sembrare as an indicator of argument schemes. We will discuss existing research on copulative constructions with appearance verbs as indicators of argument schemes (section 4), before presenting the results of our corpus study (section 5).

2. Data
The data considered in this paper consist of 40 texts taken from a mixed corpus of reviews, editorials and posts published in the comment spaces associated with reviews and editorials.[v]. The texts in our corpus have been collected from the Italian daily newspapers La Stampa and La Repubblica and from four thematic websites about art exhibitions (www.mostreinmostra.it), music (www.fullsong.it), haute cuisine (www.passionegourmet.it) and consumer electronics (www.digital.it).

The choice of these text genres is motivated by the important role argumentation plays in them and by the variety of activity fields they cover. In editorials, journalists express an opinion, mostly on a political matter, backing it up by arguments. In reviews, experts or consumers evaluate an object on the basis of firsthand experience as well as field-specific knowledge and values (Miecznikowski, in press). Comment spaces allow for a lot of variation in terms of text genres. Argumentation is common in most types of posts, however. On one hand, users react to the standpoints and arguments put forward in the text they comment on; on the other hand, on the metacommunicative level, users formulate opinions about the text as such, usually backing up their judgment by at least one argument[vi].

3. Sembrare constructions
The verb sembrare semantically presupposes two participants, namely an experiencer and an experienced. The experience in question can be entirely mental or involve perception.

The mental/perceptual process undergone by the experiencer is expressed by various syntactic constructions in which the experiencer role is either expressed by an indirect object NP or left implicit. The main form-function patterns attested with sembrare are the following:
I. Copula constructions asserting similarity between two elements (a, b), the first having a set of properties identical to a set of properties of another individual:
1. [Marco]a sembra [suo padre]b .
‘Marco looks like his father’.

II. Copula constructions and infinitive constructions asserting the existence of clues to attribute a property B to an individual a and warranting the implicature, under certain circumstances, that the speaker indeed attributes B to a:
2. [Marco]a (mi) sembra [affamato/aver fame]B .
‘Marco seems hungry/to be hungry (to me)’.

In (2), the speaker states that Marco has a set of (unspecified) properties that normally warrant the attribution of the property ‘to be hungry’. Without contextual clues to the contrary, the hearer may infer that the experiencer (here: the speaker) holds the weak belief that Marco is hungry.

III. Constructions with a complement clause in subject function. These directly and explicitly attribute a belief to the experiencer, presupposing that this belief is based on available evidence:
3. (Mi) sembra [che Marco sia stanco]p.
‘It seems (to me) that Marco is tired’.

In type I contexts, the experiencer usually coincides with the speaker and is left implicit. The experience encoded by sembrare is that of grasping the results of a process of comparison and the verb does not have an evidential function in this construction[vii].

In contexts of the types II and III sembrare can fulfill evidential functions under two conditions. The first condition is that the experiencer hold the (albeit weak) belief p. This depends on context in II, whereas the experiencer’s holding a belief is encoded grammatically in III, where the complement clause strongly suggests the presence of a proposition, i.e. of a third order entity that can be attributed a truth value and thus become a term of a belief relation[viii] When this condition is fulfilled, sembrare denotes a complex situation in which someone holds a belief on the basis of some available evidence. The second condition is that the experiencer coincide with the speaker and that the experience take place in the moment of speech. In that case, exemplified by (2) and (3) above, the verb has a performative character (Faller 2002), i.e. knowledge acquisition is not reported, but presented as achieved in the moment of speech, and the relation between p and the available evidence is mapped onto the ongoing speech event.

When sembrare is used evidentially, it always signals an indirect mode of knowing, i.e. either inference or hearsay/report. In this paper, we will be concerned especially with the verb’s inferential uses. Example (2) above is a typical case: if the speaker holds the belief that Marco is hungry, this belief is based on a reasoning process that takes into account a set of Marco’s properties in combination with further, more general, premises. In what follows, we will take a closer look at the type of reasoning sembrare is compatible with.

4. Symptomatic argumentation
In the pragma-dialectic approach, three main types of argument schemes are distinguished, namely those based on a symptomatic relation, those based on a relation of analogy and those based on a causal relation (van Eemeren & Grootendorst, 1992, pp. 98-99). In symptomatic argumentation, the argument (minor premise) and the standpoint have a common referent (X) but different predicates, as visualized in the scheme:

Y is true of X
Because Z is true of X
ANDZ is typical (characteristic/symptomatic) of Y
(van Eemeren and Grootendorst, 1992, p. 98)

The property attributed to ‘X’ in the minor premise is a symptom of the property ascribed to it in the standpoint. The major premise states the association between entities or situations which justifies the relation between the argument and the standpoint. The critical questions underlying symptomatic argumentation are the following:

– Is Z indeed typical of Y?
– Is Z not also typical of somethingelse (Y’)?
(van Eemeren and Grootendorst, 1992, p. 99)

According to Garssen (1997, p. 77-101) the category of symptomatic argumentation encompasses different subtypes of arguments such as those based on a classification, on genus-species relations, on definition and on evaluation critieria.

Van Eemeren, Houtlosser & Snoeck Henkemans (2007, p. 160) identify copulative constructions in which the predicative is an adjective or noun containing the copula to be, or its modal variants to seem/appear, as particularly suitable to form the standpoint or the minor premise in a symptomatic argumentation. According to these scholars, the abovementioned copulative constructions are good candidates to signal symptomatic argumentation because the copula normally refers to states rather than to events or processes, mirroring the nature of symptomatic argumentation, which is about qualities and features rather than about events or processes.

In analogy with van Eemeren’s, Houtlosser’s & Snoeck Henkemans’ (2007) proposal, also Italian sembrare can be hypothesized to be associated with symptomatic argument schemes. Lexical semantic arguments lend further support to this hypothesis. One of the core elements of the meaning of sembrare is the idea of similarity. This idea is present not only in the type I contexts discussed in the previous section, but also in inferential uses. In the type II contexts, in particular, the identification of clues to the presence of a property B often relies on a process of categorization by which a specific individual or situation is matched to a category (proto)type:
(4) Sembra una beffa la conclusione del processo Mills-Berlusconi. Dopo anni di preparazione, mesi di udienze, non abbiamo neanche un verdetto sulla colpevolezza o meno dell’ex premier Berlusconi.
‘The conclusion of the Berlusconi Mills’ trial seems a farce. After years of preparation, months of hearings, we do not even have a verdict on the guiltiness or innocence of the former Prime Minister Berlusconi’.
(La Repubblica, editorial, February 2012)

In example (4), the speaker categorizes a trial as a farce. One plausible reconstruction of this process of categorization is that the author compares what he has observed to his idea of typical farces:
The conclusion of the Mills’Berlusconi trial seems a farce
Because after months of preparation the trial has not produced a verdict (i.e. no goal has been reached and, by consequent, the participants’ acts appears to be meaningless) (and it is typical of farces that one cannot recognize any sense in people’s acting).

The schema of similarity activated by sembrare fosters the establishment of a link between the minor premise, in which a property is attributed to the first term of comparison, and the major premise, in which the same property is recognized as being typical of the classes of farces.

5. Sembrare and argument schemes in editorials, reviews and comments
5.1 Analytical approach
Sembrare occurs 52 times in our corpus. 39 occurrences are performative; among these, 2 are of type I construction, 17 of type II and 20 of type III. In order to find out which are the argument schemes compatible with sembrare, we have analyzed the local co- and context of all tokens in order to determine plausible implicit premises and have reconstructed the inferential relations applying the Argumentum Model of Topics (Rigotti, 2006, Rigotti, 2009a, Rigotti & Greco-Morasso, 2010).

Compared to the pragma-dialectical approach to argument schemes illustrated in the preceding section, AMT allows for a more detailed analysis of implicit premises. According to AMT, the inferential structure of any argumentation presupposes the presence of both procedural and material premises. Procedural premises have the form of maxims that define the inferential connections at issue. They are based on loci, pieces of an ontology shared by the speech community which “bind the truth value of the standpoint to the acceptance by the considered public of propositions referring to specified aspects of the ontology of the standpoint” (Rigotti, 2006, p. 527). Material premises are of two types: the endoxon, a major premise that refers to shared general knowledge and is often left implicit, and the datum, a factual (minor) premise that is often (but not necessarily) made explicit. In order to generate relevant arguments, as represented in the schema in fig. 1, procedural and material components must be combined in a double syllogistic structure (Fig.1):

Fig.1: The Argumentum Model of Topics.

5.2 Sembrare as an indicator of symptomatic argumentation
Our data confirm the role of sembrare as an indicator of symptomatic relations. The verb is indeed compatible with symptomatic argumentation in each of its constructions. More specifically, the attested subtypes of argument schemes exploit ontological relations from definition, from the parts to the whole, from implications and from concomitances.

To illustrate this group of argument schemes, we will reconstruct an example taken from an editorial of the Italian daily newspaper La Stampa about a speech in support of democracy as a prerequisite for peace, which Pope Wojtyła delivered in occasion of the disorders in Iraq during 2003:
(5) Dunque siamo grati dal profondo del cuore a Giovanni Paolo II per la costanza e la determinazione con cui ha levato la voce (una voce anche fisicamente piu’ alta e chiara, sembra che stia assai meglio ed è questo un altro motivo di consolazione).
‘Therefore we are deeply grateful to John Paul II for the persistence and the determinacy with which he has raised his voice (a voice also physically louder and clearer, it seems that he is in much better health and this comforts us even more).’
(La Stampa, editorial, April 2003)

In (5), the verb sembrare indicates that that the speaker is committed to the proposition ‘John Paul II is in much better health’ on the basis of the fact that the Pope’s voice is louder and clearer than before. This piece of evidence is a datum made explicit in the text. As to the ontological relationship between a loud voice and a state of good health, it can be conceptualized in different manners. The example might be analyzed as an instance of reasoning from the effect to the cause, if we view a loud voice as a result of the proper functioning of a healthy organism. Alternatively, it could be hypothesized that good health and a loud and clear voice are properties that are frequently associated in the experience of the speaker and the hearer, giving rise to argumentation by concomitance.

Yet another solution could be proposed, in virtue of the fact that the journalist, in this text, has chosen to institute John Paul’s voice as a discourse referent and to attribute a property to it. The journalist seems to underline the object-like status of the Pope’s voice, rather than the event of the Pope using his voice. For this reason, a part-whole relationship might be relevant in this example. If we assume that the voice is a relevant part of a person and that loudness and clearness are synonyms of healthiness when applied to a voice, the property of healthiness can be transferred from the voice to the entire person, through a maxim like the one proposed in the following reconstruction (Fig. 2):

Fig 2. Argumentative reconstruction exploiting a locus from the parts to the whole

The validity of the transfer is, of course, questionable. As underlined by van Eemeren & Garssen (2009), only absolute structure-dependent properties, such as those expressing colours or materials, are always transferrable. The choice of sembrare, which signals weak commitment, is congruent with such a context.

5.4. Sembrare as an indicator of causal argumentation
As we have seen discussing the preceding example, symptomatic argumentation does not exclude causal schemes (from the effect to the cause). In a number of contexts, however, causality – be it from the effect to the cause or from the final cause (Rigotti 2009b) – is even the most prominent ontological relation warranting the inferential transition from argument to conclusion. We have found cases of this type mostly in contexts in which speakers refer to the field of human action. In this use of sembrare, the preferred syntactic construction in the corpus is the complement clause construction.

The example we propose is taken from a post published on the website of the Italian daily newspaper La Repubblica, which comments on an editorial about Silvio Berlusconi’s defeat in the 2011 elections:
(6) La saga SB [Silvio Berlusconi] è stata una tragedia italiana che ha fatto rivivere atteggiamenti machisti ed incolti che ci hanno riportato indietro di decenni quando il nostro Paese nuotava ancora nell’analfabetismo e le nonne si stupivano della nuova invenzione della televisione. Fortunatamente sembra che il Paese sia uscito dallo stato ipnotico in cui i vari programmi televisivi lo avevano affogato.
‘The saga of SB [Silvio Berlusconi] has been a tragedy characterized by a revival of machism and uncultivated attitudes that have taken us decades back, when our country was still swimming in illiteracy and grandmothers were amazed in front of the new invention of television. Luckily, it seems that the country has woken up from the hypnotic state in which the various television programs had drowned it.’
(La Repubblica, post commenting on an editorial, June 2011)

The author claims that the country has got out of ‘the hypnotic state in which the various television programs had drowned it’. The arguments supporting this claim are largely left implicit, which is related to the highly interactive and inter-textual situation typical of forum discussions. In order to reconstruct the writer’s argumentation, we have supplied the missing premises on the basis of linguistic and contextual clues and we have interpreted the metaphorical expression “getting out of an hypnotic state”, hypothesizing that the author intends to stress the citizens’ regaining consciousness and agency (Fig. 3):

Fig. 3: Argumentative reconstruction exploiting a locus from causes

The fact that citizens have not reelected Berlusconi is highly salient in this comment space and can therefore function as a datum although it is not mentioned. The presence of the adverb ‘luckily’ in the standpoint as well as the claim that the country was in a state of backwardness due to Berlusconi’s government show that the author considers Berlusconi’s defeat as an advantage for the Italian people, an opinion that emerges also in other parts of the text. Considering La Repubblica’s political orientation, the author can assume that many readers share this opinion as an endoxon. The maxim at work is causal and is part of an ontology of human action (agents normally act in such a way as to obtain results that are advantageous for them), making it possible to reconstruct the pragmatic reasoning of agents. As a result, a certain state of mind of agents is infered from these agents’ deeds. Like in (5), the reasoning is defeasible, due to the defeasibility of the maxim (agents may act without being fully aware of their acts’ consequences).

5.5. Discussion
The data we have examined shows that sembrare can indicate symptomatic argumentation in any of its constructions, while it tends to be associated to causal relations only in the most pragmaticalized one (the one in which it functions most clearly as a propositional operator, rather than as a predicate attributed to a specific subject). The semantic relationship between causal reasoning and the lexical meaning feature /similarity/ is also rather weak. Both observations lead to the hypothesis that the possibility to express causal reasoning might be mediated by the dominant evidential function of the complement clause construction, which shifts language users’ attention from the lexeme’s core meaning to the pragmatic operation of indicating an indirect mode of knowing.

Nevertheless, that functional generalization is not complete. Even in complement clause constructions, sembrare is not compatible with any argument scheme, and symptomatic and causal arguments share some relevant features. One of these is that the various argument schemes of this group are based on loci that we can define “syntagmatic”, following Rigotti (2006):
we speak of syntagmatic loci to indicate all the classes of arguments that refer to aspects that are ontologically linked to the standpoint, either directly or indirectly, such as[..] the relationship between the whole and its constituent parts; included in this group of loci are also the classes of arguments which assume as their hooking point those pieces of world, traditionally called causes, effects, circumstances and concomitances, that condition the state of affairs the standpoint refers to.
(Rigotti, 2006, p. 528)

The term syntagmatic loci has been adopted in the AMT framework (e.g. Rigotti, 2007) to oppose these to the paradigmatic ones, in which the argument and the standpoint refer to ontologically independent states of affairs and are rather linked by relations in absentia such as opposition or analogy. The AMT model distinguishes, moreover, the intermediate class of complex loci encompassing those cases which present features of both syntagmatic and paradigmatic argument schemes. A typical example of a complex locus is the locus from authority, which establishes a causal relation between the qualities of an author and the truth of his or her discourse, while there is no direct ontological relation between the state of affairs referred to in the standpoint and the communicative situation in which the authoritative discourse is uttered.[ix]

Sembrare appears to be compatible with syntagmatic loci and, in the hearsay reading of the complement clause construction, with the complex locus of authority as well (e.g. A quanto dicono, sembra che la sinistra vincerà le elezioni, ‘According to what they say, the right wing will win the elections’).

Another restriction, which regards causality, is that sembrare is not equally compatible with any causal argument scheme. We have found several instances of argumentation from the effect to the cause, but none from the cause to the effect, neither in inferences concerning the past or present nor in predictions. The following set of constructed examples illustrates this tendency. Whereas the conclusion introduced by sembra in (7a) can easily be derived from the premise expressed in the preceding statement, this is not the case in (7b), where sembra (in contrast to other solutions such as deve ‘must’) is acceptable only if additional perceptual or hearsay evidence is assumed to be available in the context:

(7a) Marco ha una faccia stanchissima. Sembra che abbia fatto tardi ieri sera .
‘Marco has a very tired face. It seems he went to bed late, yesterday night.’

(7b) ?Marco ha fatto tardi ieri sera. Sembra che sia stanchissimo. [perceptual or hearsay evidence required].
‘?Marco went to bed late yesterday night. It seems that he is really tired’.

In predictions, inferential sembrare seems to be less acceptable with the future tense than when it is combined with a periphrasis such as stare per, which indicates a phase immediately prior to an event, or with alethic dovere ‘must’ with future reference, which indicates a situation that will cause an event:

(8a) (Mi) sembra che stia per/debba cadere. ‘
(To me), it looks as if he/she/it is about to fall.’

(8b)?(Mi) sembra che cadrà. ‘
(To me), it looks as if he/she/it will fall.’

A possible explanation of these patterns is a temporal one: by choosing inferential sembrare speakers typically signal that the available datum allows to infer a simultaneous state of affairs. This is compatible with the basic scheme of symptomatic argumentation (cf. section 4) and is evident in the cases illustrated by the examples (1) to (5) discussed in previous sections; but this analysis applies also to (a). The extension to causal inferences about the past illustrated by (6) and (7) could be mediated by the passato prossimo, since one of the functions of this tense is to denote a resultant state. The resultant state is, by the way, communicatively highly relevant in our example (6). We are aware of apparent exceptions to this generalization such as the use of sembrare in weather forecasts or with the passato remoto:

(9) (observing the sky): Sembra che pioverà.
‘It seems it will rain.’

(10) Mi sembra che il centro commerciale fu costruito negli anni ’70.
‘As far as I know, the shopping mall was built in the Seventies’.

However, these examples may be considered instances of mixed loci that share less properties with inferential uses of sembrare than with the verb’s hearsay uses, which, according to our data, are not subject to any temporal restriction. In (10), a context type that is not attested in our corpus, the knowledge source is recall from memory, whereas (9), for cultural reasons, may be framed as a semiotic practice of sign reading rather than being an instance of genuine causal reasoning[x]. Further research on appearance verbs expressing inferences about the past and the future is needed to corroborate this hypothesis.

6. Conclusion
The empirical study presented in this paper has shown that evidential uses of Italian sembrare can be used to introduce a standpoint and that they constrain the set of relevant argument schemes. The lexical meaning of sembrare makes this verb compatible with symptomatic as well as certain causal argument schemes which may be subsumed under the wider category of syntagmatic or mixed argument schemes. According to a hypothesis that has to be checked against a larger and more varied set of data, inferential uses (a) show a preference to express a temporal relation of simultaneity between the datum and the conclusion, which (b) can be extended to reasonings about non simultaneous causes and effects, especially when the verb is combined with temporal and modal markers that encode a posteriority or anteriority relation between an event and a state[xi].

Lexical semantic analysis, syntactic analysis and the argumentative reconstruction of texts are all necessary to understand which inferential processes are encoded by evidential constructions and to define their function as argumentative indicators in discourse. Perception and appearance verbs combine epistemic stance marking and evidential meanings and often occur in contexts in which the justifications at the basis of the uttered proposition are left implicit. Their polysemy and dependance on syntactic constructions calls for a fine-grained, context-sensitive semantic analysis.

The investigation of evidential and modal verbs usefully completes the growing body of research on discourse markers as argumentative indicators. Discourse markers, for example conclusion introducing connectives or concessive markers are useful to the analyst to recognize stance and argumentative moves, while evidentials and modals appear to be particularly relevant to argumentative analysis with regard to stancetaking and argument schemes.

Acknowledgements
We would like to thank the participant to the 8th Conference of the International Society for the Study of Argumentation (ISSA) for their very helpful comments and suggestions.

NOTES
i. The study presented is part of a research on the relationship between inferential uses of perception verbs and argumentation conducted at the Università della Svizzera italiana (“From perception to inference. Evidential, argumentative and textual aspects of perception predicates in Italian”, SNF grant n.141350, direction: Johanna Miecznikowski and Andrea Rocci, cf. http://www.perc-inferenza.ch).
ii. Discourse markers are particles, connectives, sentence adverbs or more complex lexical expressions that do not contribute to the propositional content of their host utterance, are syntactically poorly integrated and whose primary function is to relate utterances to their co- and context at the textual, inferential or interactional level. See Bazzanella (2006) for a more detailed discussion of the category and Miecznikowski et al., 2009, for a corpus based analysis focussed on argumentative functions of the discourse connective allora in Italian.
iii. According to the Pragma-Dialectical framework (e.g. van Eemeren & Grootendorst 1992), argumentation takes place within the context of a critical discussion involving protagonists and antagonists that critically test standpoints in order to reduce a difference of opinion. According to that model, the subtasks, or stages, defining a critical discussion are the confrontation stage (a difference of opinion is made explicit), the opening stage (the interactants commit themselves to resolve the difference of opinion and agree upon some basic assumptions and rules), the argumentation stage (arguments are put forward to justify or refute standpoints), and the concluding stage.
iv. Appearance verbs and evidential uses of perception verbs have been studied in Romance and Germanic languages by Usoniene, 2001, Pietrandrea, 2005, Cornillie, 2007, 2009, Aijmer, 2009, Diewald & Smirnova, 2010, Strik Lievers, 2012, Musi, in press a, b. For a diachronic perspective cf. Gisborne & Holmes, 2007 and Whitt, 2011 on English and Musi, 2014 on Italian sembrare.
v. The corpus has been compiled within the project From perception to inference. We would like to thank Martina Cameroni, Giuliana Di Febo and Francesca Saltamacchia for their contribution to data collection.
vi. See Miecznikowski & Musi (submitted), who adopt a genre perspective to investigate the relationship between reviews published online and the posts published in the corresponding comment spaces.
vii. The process of comparison is presupposed by the propositional content of p (similarity), whereas evidential operators are independent of the content of the proposition in their scope. In fact, in (1), the speaker commits herself to asserting the results of the comparison process, leaving the mode of knowing proper unspecified: (1) is both compatible with a situation in which the speaker has seen how Marco and Marco’s father look and infers the similarity relation on that basis, and with a situation in which the speaker has come to know about the resemblance between father and son by hearsay.
viii. According to Lyons’ classification of ontological entities (1977, pp. 438-452), taken up also in Functional Discourse Grammar (Dik, 1997), propositions are third order entities which can be judged in terms of truth value, whereas (differently from second order entities, i.e. states of affairs) they cannot be located in space and time.
ix. Cicero proposes, in his Topica (see Riposati, 1947, pp. 34-35), a distinction between intrinsic loci (alii in eo ipso de quo agitur haerent, ‘some [loci] are linked to the subject of the discussion’), and extrinsic loci (alii assumuntur extrinsecus, ‘other [loci] are derived from outside’). This topical taxonomy has been further elaborated by Boethius in his De Topiciis Differentiis (see Stump, 2004), who also suggests a third category of loci medii situated between the intrinsic and the extrinsic loci.
x. It may be relevant, at this regard, that Italian modal verbs behave atypically as well in meteorological contexts, as shows the use of deve in Deve piovere ‘it will rain’, discussed by Squartini, 2004 and Rocci, 2013:143.
xi. As far as future reference is concerned, the role played by lexical and modal verbs implying posteriority relations has been examined by Miecznikowski, under review, on the basis of an Italian corpus of economic predictions.

References
Toulmin, S. E. (1958). The uses of argument. Cambridge: Cambridge University Press.
Benoit, W. L. (1989). Attorney argumentation and Supreme Court opinions. Argumentation and Advocacy, 26, 22-38.
Eemeren, F. H. van, & Grootendorst, R. (1993b). The history of the argumentum ad hominem since the seventeenth century. In E. C. W. Krabbe, R. J. Dalitz & P. A. Smit (Eds.), Empirical Logic and Public Debate. Essays in Honour of Else M. Barth (pp. 49-68). Amsterdam: Rodopi.
Aijmer, K. (2009). Seem and evidentiality. Functions of Language, 16 (1), 63-88.
Anderson, L. B. (1986). Evidentials, paths of change, and mental maps: Typologically regular asymmetries. In W. Chafe & J. Nichols (Eds.), Evidentiality: The linguistic coding of epistemology (pp. 273-312). Norwood (NJ): Ablex.
Bazzanella, C. (2006). Discourse markers in Italian: towards a ‘compositional’ meaning. In K. Fischer (Ed.), Approaches to discourse particles (pp. 449-464). Amsterdam: Elsevier.
Bazzanella, C., & Miecznikowski, J. (2009). Central/peripheral functions of allora and ‘overall pragmatic configuration’. In M. B. Mosegaard Hansen & J. Visconti (Eds.), Current Trends in Diachronic Semantics and Pragmatics (pp. 107-121). Oxford: Emerald.
Cornillie, B. (2007). Evidentiality and Epistemic Modality in Spanish (Semi-)Auxiliaries: A Cognitive-Functional Approach. Berlin/New York: Mouton de Gruyter.
Cornillie, B. (2009). Evidentiality and Epistemic Modality: On the Close Relationship between Two Different Categories. Functions of Language, 16 (1), 44-62.
Diewald, G. & Smirnova, E. (2010). Evidentiality in German: Linguistic Realization and Regularities in Grammaticalization. Berlin/New York: Mouton de Gruyter.
Dik, S. (1997). The Theory of Functional Grammar. Berlin/New York: Mouton de Gruyter.
Eemeren, F. H. van, & Grootendorst, R. (1992). Argumentation, Communication and fallacies. A pragma-dialectical Perspective. Hillsdale: Lawrence Erlbaum.
Eemeren, F. H. van, Houtlosser, P., & Snoeck Henkemans, F. (2007). Argumentative indicators in Discourse. A Pragma-Dialectical Study. Amsterdam: Springer.
Eemeren, F. H. van, & Garssen, B. (2009). The fallacies of composition and division revisited. Cogency, 1(1), 23-42.
Faller, M.T. (2002). Semantics and pragmatics of evidentials in Cuzco Quechua, PhD Thesis Stanford.
University, Stanford. Garssen, B. (1997). Argumentatieschema’s in pragma-dialectisch perspectif. Een theoretisch en empirisch onderzoek. Amsterdam: IFOTT.
Gisborne, N. & Holmes, J. 2007. A History of English Evidential Verbs of Appearance. English Language and Linguistics, 11 (1),1-29.
Lyons, J. (1977). Semantics. Cambridge: Cambridge University Press.
Miecznikowski J., Gili Fivela, B., Bazzanella C. (2009). Words in context. Agreeing and disagreeing with allora. In G. Gobber et al. (Eds.), Word Meaning in Argumentative Dialogue, vol. 1 (= L’analisi linguistica e letteraria XVI, 2008/1, special issue), 205-218.
Miecznikowski, J. (2011). Construction types and argumentative functions of possibility modals: evidence from Italian. In F. H. van Eemeren, B. Garssen, D. Godden & G. Mitchell (Eds.), Proceedings of the 7th Conference of the International Society for the Study of Argumentation (pp. 1284-1297). Amsterdam: Rozenberg/SicSat.
Miecznikowski, J. (in press). L’argomentazione nelle recensioni on-line. In B. G. Fivela, E. Pistolesi & R. Pugliese (Eds.), Parole, gesti, interpretazioni. Studi linguistici per Carla Bazzanella. Roma: Aracne.
Miecznikowski, J. (under review). Envisager le futur dans les prévisions économiques. In L. Baranzini, J. Sánchez Méndez & L. de Saussure (Eds.), Le futur dans les langues romanes, Peter Lang.
Miecznikowski, J. & Musi, E. (under review). Text genres adapted to new contexts: the case of online reviews. In M. Casoni, S. Christopher, A. Kamber, J. Miecznikowski, E. Pandolfi & A. Rocci (Eds.), Proceedings of the Vals-Asla meeting “Language norms in context”, 12th-14th February 2014, Special Issue of Bulletin suisse de linguistique appliquée.
Musi, E. (2014).Verbi d’apparenza tra semantica e sintassi: il caso di sembrare in italiano antico. AION Linguistica 3.
Musi, E. (in press a). Evidential modals at the semantic-argumentative interface: appearance verbs as indicators of defeasible argumentation. In Informal Logic, 34 (3).
Musi, E. (in press b). Sembrare tra semantica e sintassi. Proposta di un’annotazione multilivello. In Proceedings of the XLVIII International SLI (Società linguistica italiana) Conference “Livelli di analisi e fenomeni d’interfaccia”, 26th-28th September 2013.
Pietrandrea, P. (2005). Epistemic Modality: Functional Properties and the Italian System. Amsterdam-Philadelphia: John Benjamins.
Pike, K. L. (1954). Emic and Etic Standpoints for the Description of Behavior. In K. L. Pike (Ed.), Language in Relation to a Unified Theory of the Structure of Human Behavior (pp. 8-28). Glendale: Summer Institute of Linguistics.
Rigotti, E. (2006). Relevance of Context-Bound Loci to Topical Potential in the Argumentation Stage. Argumentation, 20, 519-540.
Rigotti, E. (2007). Can classical topics be revived within the contemporary theory of argumentation? In F. H. van Eemeren, A.J. Blair, F. Snoeck Henkemans & C. Willards (Eds.), Proceedings of the Sixth Conference of the International Society for the Study of Argumentation (pp. 1155-1164). Amsterdam: Rozenberg/SicSat.
Rigotti, E. (2009a). Whether and how Classical Topic can be Revived within Contemporary Argumentation Theory. In F. H. van Eemeren & B. Garssen (Eds.), Pondering on Problems of Argumentation (pp. 157-178). Amsterdam: Springer.
Rigotti, E. (2009b). Locus a causa finali. Special Issue of L’analisi linguistica e letteraria, 559-576.
Rigotti, E., & Greco Morasso, S. (2010). Comparing the Argumentum Model of Topics to Other Contemporary Approaches to Argument Schemes: The Procedural and Material Compo¬nents. Argumentation, 24(4), 489-512. Riposati, B. (1947). Studi sui Topica di Cicerone. Milano: Vita e Pensiero.
Rocci, A. (2008). Modality and its conversational backgrounds in the reconstruction of argumentation. Argumentation, 22(2), 165-189.
Rocci, A. (2012). Modality and argumentative discourse relations: a study of the Italian necessity modal dovere. Journal of Pragmatics, 44(15), 2129–2149.
Rocci, A. (2013). Modal conversational backgrounds and evidential bases in predictions: the view from the Italian modals. In K. Jaszczolt & L. de Saussure (Eds.), Time: Language, Cognition and Reality. Oxford: Oxford University Press.
Squartini, M. (2004). Disentangling evidentiality and epistemic modality in Romance. Lingua, 114, 873-895.
Squartini, M. (Ed.). (2007). Evidentiality between lexicon and grammar. Special Issue of Italian Journal of Linguistics 19 (1).
Strik Lievers, F. 2012. Sembra, ma non è. Studio semantico-lessicale sui verbi con complemento predicativo. Firenze: Accademia della Crusca.
Stump, E. (2004). Boethiu’s ‘De Topiciis Differentiis’. Ithaca, NY: Cornell University Press.
Usoniene, A. (2001). On direct/indirect Perception with Verbs of Seeing and Seeming in English and Lithuanian. Working Paper, 48, 163-182.
Whitt, R. J. (2011). (Inter)Subjectivity and Evidential Perception Verbs in English and German. Journal of Pragmatics, 43(1), 347-360.

Abstract: Our paper aims to examine several aspects of the epideictic genre according to the tradition of the Brussels School of Rhetoric. We study, at first, the confused notions as a specific material for the rhetorical art, and, in particular, for the epideictic genre as they contribute to create the social concord. Then, we establish a relationship between disagreement and epideictic genre after the Perelman’s New Rhetoric. Here, our idea is to show how disagreement feeds the argumentative nature of this third rhetorical genre. In a democratic society, the epideictic genre needs to work well to allow disagreement; and likewise, disagreement requires always a well-functioning epideictic. According to Perelman, if the epideictic genre constitutes the foundations of the rhetorical system, or even its “crowning”, it is also the center, the mobile part of this system, in other words: its limbs.
Keywords: Chaim Perelman, confused notions, concord, disagreement, epideictic genre, Eugène Dupréel, rationality, rhetoric.

1. Introduction
Our paper aims to examine several aspects of the epideictic genre according to the tradition of the “Brussels School of Rhetoric” started with Eugène Dupréel and Chaim Perelman. We study how, in the epideictic genre, the “confused notions” contribute to create social concord. The relationship between disagreement and epideictic genre in Perelman’s New Rhetoric will then be considered to show how disagreement feeds the argumentative nature of this third rhetorical genre.

To start with, taking as a frame the perspective of Emmanuelle Danblon, in which rhetoric is a technè and the orator is a craftsman, we would like to show how the “confused notions” (in the sense given by Eugène Dupréel) could be shaped in a specific way, according to the desired rhetorical purpose, to become efficient tools, which will be destined to a “good use” by the orator.

2. “Using value” of confused notions and its role in the epideictic genre
2.1. Origins of the confused notions
Already before the First World War, Eugène Dupréel had suggested a re-establishment of the “confused thought”, wishing to exceed the classical dichotomy clarity vs. darkness. Confusion and instability, like clarity and stability, are essential components of some notions, especially values as justice, happiness, merit or freedom. In Dupréel’s conception, notions are not a reflection of the world but a tool with an acting value:
Avant d’être classées comme connaissances claires ou confuses, les connaissances servent à quelque chose, à la vie des individus et des sociétés; les mensonges même ont leur utilité, on ne les produirait pas sans cela. La connaissance est donc une valeur d’action. […] Une notion, tout ce que désigne un mot ou une phrase, cela n’est pas élaboré par un souci de correspondance avec un objet réel, c’est un instrument dont on se sert et dont la valeur se mesure d’abord à son rendement. (Dupréel, 1949, p. 332).

Before being classified as clear or confused knowledge, knowledge is used to something, in the lives of persons and societies; lies even have their uses, they will not happen without it. Knowledge is therefore an acting value. […]. A notion, everything that refers to a word or phrase, is not developed by a desire to match with a real object; it is a tool that is used and its value is measured primarily to performance[i].

Notions contain an extensible semantical core that allows us to progress towards a practical knowledge. Actually, the function conferred to the confused notions is to allow an agreement in domains where formal demonstration is impossible (i.e. the Humanities), and in particular to allow adherence to a philosophical truth. Indeed, due to the great precariousness of this kind of truth, that adherence is its only support:
Ne travaillant pas, comme le savant, entre une intention précisée et un mode de vérification fixé d’avance, ne déterminant qu’en cours de route son intention, le philosophe verra toujours son œuvre moins formellement accomplie et non formellement vérifiée: en fait il ne peut compter que sur l’adhésion gagnée, sur l’accord avec lui-même et l’accord avec les autres esprits, ce qui n’est jamais un critère, mais un état de chose, difficile et précaire. […] Au contraire, la valeur d’une vérité philosophique aura bien plus besoin, pour s’imposer, de l’unanimité dans l’adhésion car, en dehors de la conviction de celui qui la découvre, cette approbation d’autrui est en fait son seul appui; or, c’est justement cette adhésion qui se montre plus précaire et moins probable. (Dupréel, 1939, pp. 289-290).

Not working, as the scientist, between a specified purpose and a verification mode fixed beforehand, determining only on the way his intention, the philosopher will always see his work less formally completed and not formally checked: actually he can only rely on membership earned, on agreement with himself and the agreement with the other spirits, which is never a criterion, but a state of things, difficult and precarious. […] On the contrary, the value of a philosophical truth will much more need to impose unanimity in membership because, apart from the conviction of the person who discovers it, the approval of others is in fact his only support; however, it is this membership that is more precarious and less likely.

To be able to adjust the scope of the notion to a context of use, one needs to require to the reasonable, which Dupréel called “excellence confuse” (Dupréel, 1949, p. 294). Human being is able to make choices without dogmatism, because a way exists to review these choices (Dupréel, 1949, p. 295). For instance, a part of Dupréel’s Traité de morale touches on the values of justice and honor as confused notions. According to him, confusion is a fact that allows to act in a living and human world.

Dupréel speaks about a tool, and not about a material. Moreover, he devotes very little attention to which technè has to be optionally used to transform these confused notions into a tool. His students, Chaim Perelman and Lucie Olbrechts-Tyteca, make this point in their article “Les notions et l’argumentation”. Returning on what are exactly confused notions, they explain that argumentation involves playing on its plasticity, through two technaï: either opposing two notions through flexibility on the one hand and curing on the other hand, or extending the semantical core of a single notion.

For the first case, the orator presents to the audience an opposition between two notions: he offers his own view as modern, flexible and rich in potential, while the conception of his opponent is downgraded as old, frozen and outdated. In the second case, and for the notions which the value is clearly established and prior the argumentation, another technè is used: the extension of the notion (with amplification or restriction of its semantical core):
Cette technique qui consiste à figer le concept de l’adversaire tout en donnant plus de souplesse à celui qu’on défend, est généralement adoptée lorsque l’appréciation sur le concept doit résulter, en partie au moins, de l’argumentation. Par contre, dans le cas où la valeur de la notion est nettement établie et préalable à l’argumentation, c’est une autre technique portant plutôt sur l’extension de la notion, qui est généralement employée. (Perelman & Olbrechts-Tyteca, 1989 [1955], p. 136).

This technique consisting in freezing the concept of the adversary while providing more flexibility to one we defend, is generally used when appreciating that the concept must result, partially at least, from the argumentation. Contrariwise, if the value of the notion is clearly established prior to the argument, this is another technique involving the extension of the concept, which is generally used.

The common values, celebrated in the epideictic genre, are included in this last kind of notions. Public discourses celebrate those values to preserve social concord – homonoia for the Greeks. They are destined to introduce a proairesis, a disposition to act in a good way. In this case, notions are amplified to the maximum in order to appear, as blatant as the sensitive evidence (Danblon, 2002, 130-134). On the other hand, regarding the deliberative genre, decisions have to be taken for the good functioning of the city; regarding the forensic genre, decisions concern the establishment and qualification of past events. Both decisions are bouleutics and derive from public debates. The purpose of the technè is either to make a choice between two notions, or to narrow the semantical core of a notion, questioning respectively what is useful or what is just in a specific case.

Places where confused notions can be found might be compared to a kind of “marketplace”, in which the orator can somehow shop around; this metaphor was previously used by Wilhelmus De Pater, talking of Aristotle’s Topics (De Pater, 1965)[ii]. These stores could take the form of the law to be interpreted or great universal declarations like, e.g., Human Rights. Indeed, those expressions of topoi, as commonly accepted premises, form the starting point of the argumentative reasoning. The confusion of the notion allows, as Perelman said, to an agreement on the formula even if disagreements subsist on the interpretation. In that way, we might say it becomes more a tool for concord than a tool for agreement.

In Perelman and Olbrechts-Tyteca’s paper, notions are thus presented as tools for persuasion, but after they have been shaped by the technè in accordance with the rhetorical purposes. One may suggest that the confused notions exist beforehand, in the “marketplace”, as raw materials to be shaped, and finally become a tool. We would like to go further on that process that allows to precise the conception of rhetoric as a craft.

2.2. Rhetoric as a craft, “using value” of the confused notions as a material fort the rhetorical art
Emmanuelle Danblon, in L’homme rhétorique, recalls Vernant’s work about craft in Myth and Thought among the Greeks, and then applies it to the rhetorical art:
D’un point de vue naturaliste, la rhétorique se révèle être l’art de tous les artisanats. Elle n’est pas d’une discipline, elle est de toutes les disciplines. Elle exerce l’homme à utiliser son environnement naturel: celui des sociétés humaines. (Danblon, 2013, p. 84).

From a naturalistic point of view, the rhetoric appears to be the art of all crafts. It belongs not to a single discipline, it belongs to all disciplines. It exerts the man to use his natural environment: the human societies.

Following Vernant, in an antique conception of the work (in a craftsmen’s society), the point is the using value of the artefact, not its market value. This artefact matches with a special need for a specific user. The question of this need, the purpose of the craft is dominant in the process, and much more important than the technè implemented:
The artisan and his skill exist for the sake of the product, the product for the sake of the need. It could not be otherwise, as long as the product of work was considered only from the point of view of its use value, not its exchange value. As for its use value, the product is defined by its service to the person who uses it. (Vernant, 2006 [1965], pp. 295-296).

For Danblon, in that framework, the rationality of the craft is directly linked to its efficiency. And so it goes in the rhetorical art, whose worth emerges only if its efficiency is sufficient to impact on man’s action and on the running of the City. As far as the rhetorical activity is concerned, the purpose is to take decisions, and, in Aristotle’s conception, decisions that lead to Happiness in the City.

In the classical Greek society of the 5th century, where the first theories of that discipline emerged, the place given to the craftsman has moved. It became associated to menial tasks, whereas the craftsman, before, had occupied a much more prestigious and prevalent position. At the same time, Sophists were leading the first technical reflections about rhetorical technè. That technè was quite different of the craftsman’s technè: while the craftsman implements a poïésis (he creates an artefact out of himself), the orator commits a praxis (he acts on the world) (Vernant, 2006 [1965], p. 291). However, as Danblon has noticed, the category of “using value” is very relevant to us. Furthermore, it could directly be linked to Dupréel’s acting value.

Vernant adds that this model of craft, transferred to intellectual matters, leads to a model of “demiurgic creation” mentioned by Plato and Aristotle. The spirit of the final product exists outside of the craftsman, because it’s defined by its uses: the house (built) preexists at the future house to be built, such as vases, and other artefacts in general. What is important is not the market value but the benefit for the user: to be safe, to carry water… So there’s something like a “matrix”, available for the craftsman, allowing varied shapes of materials. Craftsman’s activity is, according to Vernant, guided by an eidos, prior, fixed and immutable:
The technè aims, in effect, to produce an eidos, such as health or a house, in a certain matter. Such a production presupposes the exercise of a dunamis for which the technè, in a sense, provides the method of use. (Vernant, 2006 [1965], p. 289).

To maintain the parallel with rhetoric, confused notions as materials could be shaped according to the context and the purpose, since technè, as we said, depends on the type of decision to be generated. The orator draws on his store, the topical heritage which we mentioned previously, where he could find raw materials. If the orator is a craftsman, that store contains the eidè with which he needs to practice his art.

But that conception of eidos might directly lead to a Platonic vision, and seems hardly compatible with the efficiency sought by the Sophists or with Dupréel’s acting value. However, if the eidos is linked to a using value, and that shaping confused notions allows creating new eidè, this hurdle is avoided. Indeed, the orator’s marketplace is only composed of shaped material that could be shaped again, according to the uses encountered or to be encountered, whose meaning will never be defined once and for all. Actually, in the rhetorical art, there is not any raw material: topical heritage is linked to a specific period and is constituted by uses; always moving, and liable to be modified by critics. The dynamic aspect of the notions prevents them from being treated as Platonic ideas.

This point of the “using value” leads to another question: the good use of confused notions, in particular in the epideictic genre. Values, confused notions by excellence, keep a privileged relationship with this genre. Perelman has noted that confused notions without critique leads directly to propaganda; so it is necessary to implement them in a whole rhetorical system.

3. Epideictic as a condition of disagreement in Perelman’s New Rhetoric
From their early works, and contrary to popular belief even in our scientific field, Perelman and Olbrechts-Tyteca give a prominent and leading position to the epideictic genre. There is something very intuitive in their minds. For them, the epideictic is the first of the three genres: even before the deliberative and the judicial. However, Perelman and Olbrechts-Tyteca do not ignore the specific gaps of the epideictic genre in comparison to the two other genres. These gaps give the epideictic a special and marginal nature. In the epideictic genre, there is no opponent, no controversial issue, no debate, and no decision-making.

As a genre of circumstance, the epideictic seems secondary, even unimportant in the rhetorical perspective. In a certain sense: a soft and “feminine” genre (against the two others, which are considered more “virile”). We think usually that the epideictic orator speaks in order to say nothing because the subject of the discourse is not controversial; everything in the speech has already been deliberated on. Perelman and Olbrechts-Tyteca oppose this opinion. There is a real ambiguity because they appear to make a marginal (and not “serious”) genre a primary one. What’s more, they denounce the misunderstanding of epideictic. They propose to rediscover its rhetorical and argumentative nature: its place in the field of argumentation.

For them, the consequences of this misunderstanding were dramatic for rhetoric as a discipline. They make a link between the dismemberment of rhetoric in particular since the nineteenth century, and the negative perception of the epideictic genre in public opinion and scientific field. We can read what Perelman and Olbrechts-Tyteca wrote about it:
C’est cette incompréhension du rôle et de la nature du discours épidictique – qui, ne l’oublions pas, existait bel et bien, et s’imposait donc à l’attention – qui a encouragé le développement des considérations littéraires en rhétorique et a favorisé, entre autres causes, l’écartèlement de celle-ci entre deux tendances, l’une philosophique […], l’autre littéraire. (Perelman & Olbrechts-Tyteca, 1952 [1950], pp. 15-16).

It’s this misunderstanding of the role and the nature of epideictic discourse – which, let us not forget, existed and therefore was well known – which encouraged the development of literary considerations in rhetoric, and encouraged, with other implications, the breakup of rhetoric into two tendencies: one philosophical […] and the other, literary.

For Perelman and Olbrechts-Tyteca rhetoric has to be understood as something coherent and efficient. This requires above all, an understanding of the epideictic genre as a place of “communion” and as a mood of gathering. We could suppose that Perelman came to discover rhetoric (and therefore epideictic) through his reflections on legal agreement between two sides as well as the conditions necessary to find this agreement.

However, this would be an incorrect interpretation. Upon closer examination, we find that Perelman is not interested, first of all, by agreement, but by disagreement. He is especially interested in how disagreement can give rise to argumentative invention and rhetorical opportunities. For him, disagreement is not a drama, the sign of an error, or the evidence of our irrationality. He is radically opposed to Descartes and all the radical positivists. Perelman argues that there may be two (or x) contrary positions on the same subject without any of these having to be necessarily irrational. Argumentative rationality can also be found in the exploration of disagreement between the parties. For Chaim Perelman, it would be misleading to identify agreement with good choice and/or rationality.

A large part of Perelman’s work aims to analyze the possibilities of a reasonable disagreement; and how such a disagreement can be explored through argumentation. This is how Perelman presents his intellectual itinerary in a letter to the young Marcel Côté (a Canadian doctoral candidate) dated from January 1982:
L’inspiration fondamentale pour l’élaboration de la théorie de l’argumentation ne me vient pas du droit mais de la philosophie [la question étant] d’où vient le désaccord entre les philosophies. Ce n’est qu’à partir de 1953 que j’ai commencé à m’intéresser sérieusement au raisonnement juridique. (Perelman, 1982).

The fundamental inspiration for in the elaboration of a theory of argumentation does not come to me from law but from philosophy; [the question, for me, to find] where the disagreement between the two philosophies has its source. It is only from 1953 onwards that I became interested in legal reasoning.

To recapitulate, Chaim Perelman encountered rhetoric and epideictic through the lens of disagreement. However, one of his first texts on rhetoric, “Logique et rhétorique” (published in 1950, and co-authored with Lucie Olbrechts-Tyteca), provides a clear focus on epideictic to rehabilitate it.

Interested in the concept and practice of disagreement, Perelman focuses on the genre, which seems most radically distinct from disagreement and which is the least clearly argumentative of the three genres. There is something contradictory here. That is why we need to assume a political and rhetorical link between disagreement and epideictic. A link that Perelman did not explain, but which is implied in his work; a crucial link for understanding what rhetoric really is. That is to say, to see rhetoric as a truly “human work” that can lead the way for a “sense of responsibility and freedom” (Perelman & Olbrechts-Tyteca, 1952 [1950], pp. 42-43). The Perelmanian idea, because it is humanistic aims to express the connection between disagreement and the epideictic genre; to challenge the apparent dichotomy between the two. In a democratic society, epideictic needs to work well to allow disagreement; and likewise, disagreement requires always a well-functioning epideictic. This idea is represented in the table below (see fig. 1).

To be clear: we need to ask ourselves, what would disagreement without epideictic? It would be, no doubt, a permanent cacophony; civil conflict, and maybe even chaos. This is why, it is always necessary to regularly nourish the intensity of adherence to certain values to ensure the communion around these values. In the same way, what would epideictic genre be, without disagreement? It would certainly be a dictatorship of enforced agreement and all forms of propaganda and authoritarianism.

It is for this reason that rhetorical argumentation only has sense if one places value on adherence. At the same time, this adherence, by nature conditional (i.e. it is a fact, not a right), must exclude the use of violence or coercion. Rhetorical and political balance hangs on this relationship.

Fig. 1: The epideictic genre and the disagreement

Perelman does not give the epideictic genre a unique place: he even gives it two. He makes the epideictic genre the basis of his system of rhetoric: since without epideictic, no rhetoric is possible. Furthermore, he also makes it the center of his system: since without epideictic, no disagreement, nor justification, is possible. Summing up, according to Perelman, it is the role of this genre, which is seen as marginal, to ensure the functioning of the whole system of rhetoric around it. Not only does the epideictic genre make rhetoric possible; but also it makes rhetoric practical, and even practicable. It constitutes the roots and the living substance of rhetoric as in the diagram below (fig. 2). This stark and revealing distinction is laid out in the two paragraphs from the programmatic article quoted previously:
Ne voyant pas nettement de but au discours épidictique, les anciens étaient donc enclins à le considérer uniquement comme une sorte de spectacle, visant au plaisir des spectateurs et à la gloire de l’orateur, par la mise en valeur des subtilités de sa technique. Celle-ci devient donc un but en soi. Aristote lui-même [la critique est peu charitable, mais passons] ne semble saisir que l’aspect agrément, apparat, du discours épidictique. Il ne perçoit pas que les prémisses sur lesquelles s’appuient les discours délibératifs et judiciaires, dont l’objet lui paraît si important, sont des jugements de valeur. Or ces prémisses, il faut que le discours épidictique les soutienne, les confirme. (Perelman & Olbrechts-Tyteca, 1952 [1950], p 14).

Not seeing a clear objective for epideictic discourse, the ancients were thus inclined to consider it only as a sort of spectacle, which pleased spectators and gave glory to the orator, through the showcasing of the orator’s subtle techniques. In this way it thus became a goal in and of itself. Aristotle himself [in an unkind critique, but let’s leave this aside] appears to understand only its pleasing aspect, its pomp and circumstance. He does not understand that the premises on which deliberative and judiciary discourses base themselves, and whose function he values so much, are in fact value judgments. However, these premises must be sustained and confirmed by epideictic discourse.

Without epideictic discourse to support or confirm certain values, which are seen as important for a certain community, speakers would be unable of making value judgments. Speakers would be deprived of the capacity to argue. In fact, the formulation of judgments in the deliberative or judicial arena implies always the availability of values for judgment, principles to criticize, and commonplaces to denounce. Without epideictic discourse, without roots, without premises at our disposal, no one could ever formulate anything but senseless and valueless discourses.

Fig. 2: Basis and center of Rhetoric

However, if the epideictic genre constitutes the foundations of the rhetorical system, or even its “crowning”, it is also the center, the mobile part of this system, in other words: its limbs. This is why the third genre of rhetoric enables the articulation of the whole edifice of rhetoric. It helps rhetoric to be applied and tested. In other words, the epideictic is not only an enabling condition of the judicial and deliberative discourses, their roots, but it is also the very source of their permanent vitality. Indeed, the epideictic seeks to create a “communion” between free and responsible citizens:
Cette communion ne détermine pas un choix immédiat, détermine toutefois des choix virtuels. Le combat que livre l’orateur épidictique est un combat contre des objections futures; c’est un effort pour maintenir la place de certains jugements de valeur dans la hiérarchie ou éventuellement leur conférer un statut supérieur. […] Aussi le genre épidictique est-il central dans la rhétorique. (Perelman & Olbrechts-Tyteca, 1952 [1950], p 14).

This communion while it does not determine an immediate choice, it does however determine virtual choices. The struggle which the epideictic orator leads, is a struggle against future objections; it is an effort to maintain the place of certain value judgments in the hierarchy, or maybe to give them a superior status. […] In this way, the epideictic genre is central in rhetoric.

This genre ensures the stability and the circulation of values. It articulates the continuity and coherence between the past, present and future of the community. In this regard, Perelman goes further than Aristotle. On the one hand, he makes epideictic discourse a place of dialogue between this three temporalities; on the other hand, he makes of it a place, which, in this dialogue, opens the way for a struggle to come, based on these same values. This struggle cannot always take place here and now, because it is neither the time nor the place. This is implied in the rules of the genre. Hence, the deliberative and judicial genres exist to offer an arena for this struggle to take place in the future.

From now on, we can say that the epideictic genre cannot be placed outside the field of argumentation. Adherence now and elsewhere is not pre-established. It would be an illusion to believe that the conditions for a communion of conscience could be inscribed in the nature of things. At the same time, if the struggle is delayed for now, it is to allow epideictic discourses to protect the community against itself, against all the threats of discord, fear, and disenchantment. This is why the epideictic genre, is in no case a collection of empty commonplaces or trivialities beyond discussion.

4. Conclusion
In a bold way, and to conclude, we could say that Perelman underlines the precarious character of values and adherence to these, which is present in the epideictic genre. He invites us to recognize this fragility as an opportunity and not as a drama.
The act of speaking to reinforce the established order does not seek to deny the existence of problems. Neither is it a question of denying the fragility of the values that are being defended. On the contrary, the aim is to manifest the fact that there is a problem and that the values being defended are indeed fragile ones.
Concretely, if there would be no problem, and if values would not be fragile, or confused, there would simply be no need to speak up to set the problem in context.

NOTES
i. Unless otherwise specified, the translations are done by the authors of the paper.
ii. We would like to thank Emmanuelle Danblon and Victor Ferry for this reference.

References
Aristotle (1967 [1938]). Rhétorique (vol. 1, trans. by M. Dufour). Paris: Les Belles Lettres.
Danblon, E. (2002). Rhétorique et rationalité: Essai sur l’émergence de la critique et de la persuasion. Bruxelles: Éd. de l’Université de Bruxelles.
Danblon, E. (2013). L’homme rhétorique: Culture, raison, action. Paris: Éditions du Cerf.
De Pater, W.A. (1965). Les topiques d’Aristote et la dialectique platonicienne: Méthodologie de la définition. Fribourg: Ed. St. Paul.
Dominicy, M. (2006). Perelman und die Brüsseler Schule. In J. Kopperschmidt & E. Eggs (Eds.), Die neue rhetorik: Studien zu Chaim Perelman (pp. 73-134). Munich: W. Fink Verlag.
Dupréel, E. (1939). Esquisse d’une philosophie des valeurs. Paris: Librairie Félix Alcan.
Dupréel, E. (1949). Essais pluralistes. Paris: PUF.
Mayeur, I. (2015). Les notions confuses comme matière de l’activité rhétorique: postérité d’un héritage dupréelien. In L. Nicolas (Ed.), Le fragile et le flou: Apprivoiser la précarité, un art rhétorique, forthcoming.
Nicolas, L. (2009). La fonction héroïque: parole épidictique et enjeux de qualification. Rhetorica: A journal of the History of Rhetoric, XXVII, 2, 115-141.
Perelman, Ch. & Olbrechts-Tyteca, L. (1952). Rhétorique et philosophie: Pour une théorie de l’argumentation en philosophie. Bruxelles: Presses universitaires de Bruxelles.
Perelman, Ch. & Olbrechts-Tyteca, L. (1989 [1955]). Les notions et l’argumentation. In Rhétoriques (pp. 123-150). Bruxelles: Presses universitaires de Bruxelles.
Perelman, Ch. & Olbrechts-Tyteca, L. (1969 [1958]). The new rhetoric: A treatise on argumentation (trans. by J. Wilkinson and P. Weaver). Notre Dame: University of Notre Dame Press.
Perelman, Ch. (1978). L’usage et l’abus des notions confuses. Logique et Analyse, 81, 3-17.
Perelman, Ch. (1979). L’originalité de la pensée d’Eugène Dupréel (à l’occasion du centenaire de sa naissance). Bulletin de la classe des Lettres et des Sciences morales et politiques, LXV, 60-71.
Perelman, Ch. (1982). Letter to Marcel Côté (January 4, 1982). Archives and Libraries of the ULB – Fonds Chaim Perelman: BE.ULB – A&B – ARCH.089PP (box 21).
Vernant, J.-P. (2006 [1965]). Myth and thought among the Greeks (trans. by J. Lloyd & J. Fort). New York: Zone Books.
Sennett, R. (2008). The craftsman. New Haven: Yale University Press.