Microsoft Word - estheromero+Metaphorical+argumentation_PROOFS_2 (2)-2.doc © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Metaphorical Argumentation ESTHER ROMERO & BELÉN SORIA Departamento de Filosofía 1 Universidad de Granada Granada Spain eromero@ugr.es bsoria@ugr.es Abstract: It is a fact that novel metaphorical utterances appear in natural language argumentation. It seems, moreover, that these put forward metaphorical propositions that can have different roles (data, warrants or claims) in argument structure. There can even be good argumentation which is indispensably metaphorical. However, not all metaphor theories provide an explana- tion of metaphorical meaning compat- ible with these claims. In this article, we explain the three main views on metaphorical meaning and show, analysing some examples, their consequences for metaphorical argumentation. Our analysis shows that only the cognitive view can explain that there are arguments which can only be generated using novel metaphors. Résumé: C'est un fait que de nou- veaux énoncés métaphoriques appa- raissent dans l'argumentation en langage naturel. Il semble d'ailleurs que celles-ci mettent en avant des propositions métaphoriques qui peuvent avoir différents rôles (don- nées, licence inférentielle ou énoncés) dans la structure argumentative. Il peut même y avoir une bonne argu- mentation qui est indispensablement métaphorique. Cependant, pas toutes les théories de la métaphore ne fournissent une explication du sens métaphorique compatible avec ces affirmations. Dans cet article, nous expliquons les trois principaux points de vue sur le sens métaphorique et montrons, en analysant quelques exemples, leurs conséquences pour l'argumentation métaphorique. Notre analyse montre que seule la vision cognitive peut expliquer qu'il existe des arguments qui ne peuvent être générés qu'à l'aide de nouvelles métaphores. Keywords: argument structure, cognitive account of metaphorical argumenta- tion, deflationary account of metaphorical argumentation, metaphorical mean- ing, metaphorical proposition, novel metaphorical utterances, sceptical account of metaphorical argumentation 392 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. 1. Introduction The main objective of this paper is to defend that when novel metaphorical utterances appear in natural language argumentation this can have special features. However, argumentation theorists have not usually focused on the features of metaphorical argumen- tation.1 This low interest can be explained by the fact that meta- phorical utterances have traditionally been considered irreconcila- ble with argumentation. Metaphorical utterances should not be used in it. The most radical case of this attitude depends on the sceptical proposal on metaphorical meanings. Without metaphorical mean- ings, metaphorical utterances do not communicate any metaphori- cal proposition and, strictly speaking, there are no metaphorical arguments; a proposal that Sir Walter Scott already put in the mouth of Lady Hermione in his The Fortune of Nigel when she told Mistress Margaret, “Metaphors are no arguments, my pretty maiden” (sentence also quoted by Black at the beginning of his article “Metaphor” 1954-55, p. 273). Mistress Margaret claimed that she, a bird in the air, would rather be a lark than a weather- cock because the lark “sings while he is drifting down the summer breeze” while the weathercock “sticks fast yonder upon his iron perch, and just moves so much as to discharge his duty, and tell us which way the wind blows,” but this is not an argument because metaphors cannot communicate reasons nor conclusions. A less radical case of that attitude depends on the deflationary proposal on metaphorical meanings according to which metaphor- ical utterances communicate metaphorical propositions that could be communicated by means of literal utterances. Metaphorical argumentation is then replaceable by literal argumentation since metaphorical utterances are mere stylistic devices. Their justifica- tion comes from the pleasure the hearer gets when he is involved in solving puzzles. This justification, however, is often called into 1 We found some exceptions with works which deal with the relationship between argumentation and novel metaphor such as Perelman and Olbrechts- Tyteca (1958), and recently, Santibáñez (2010), Oswald and Rihs (2014), Xu and Wu (2014), Pollarolli and Rocci (2015), Wagemans (2016), Bilstrup Finsen, Steen and Wagemans (2019) and van Poppel (2021). Metaphorical Argumentation 393 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. question: certain pleasures are just distractions that can have un- wanted consequences for serious thought. In Locke’s words, (…) all the artificial and figurative application of words eloquence hath invented, are for nothing else but to insinuate wrong ideas, move the passions, and thereby mislead the judgment; and so indeed are perfect cheats: and therefore, (…) they are certainly, in all discourses that pretend to inform or instruct, wholly to be avoided; and where truth and knowledge are concerned, cannot but be thought a great fault, either of the language or person that makes use of them (1690, III, X, §34). By contrast, some scholars, following Black’s interaction theory, have argued for a cognitive proposal on metaphorical meaning according to which metaphorical utterances are not dangerous ornaments. Rather they communicate propositional contents that cannot be communicated by means of literal utterances and have a distinctive non-literal cognitive value. We are with those theorists and look into relevant examples of argumentation in everyday uses of language to show that there are cases of metaphorical argumen- tation with specific non-literal properties that make them irre- placeable by literal ones. The following example, taken from the film About a boy, illus- trates this type of cases: everyday communicative acts of arguing which include novel verbal metaphorical utterances. We call this type of cases ‘metaphorical argumentation.’ The scene takes place just at the beginning of the film and shows the protagonist, Will, being alone at home and watching the TV program Who wants to be a millionaire? When the host asks: Who wrote the sentence ‘No man is an island’? John Donne? John Milton? John F. Kennedy? Jon Bon Jovi?, Will answers: ‘Jon Bon Jovi; too easy’ and adds: (1) And, if I may say so, a complete load of bollocks. In my opinion, all men are islands. And what’s more, now’s the time to be one. This is an island age. A hundred years ago, you had to depend on other peo- ple. No one had TV or CDs or DVDs or videos... ...or home espresso makers. 394 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Actually, they didn’t have anything cool. Whereas now, you see... ...you can make yourself a little island paradise. With the right supplies and the right attitude... ...you can be sun-drenched, tropical, a magnet... ...for young Swedish tourists. (…) And I like to think that perhaps I am that kind of is- land. I like to think I’m pretty cool. I like to think I’m Ibiza. Will disagrees with John Donne’s famous (metaphorical) proposi- tion expressed by the utterance of ‘No man is an island’ and with those characters in the film who support it (in subsequent scenes). His argumentation is directed at justifying his claim that now is the time for a person to be an island using his happy way of living alone as evidence for it; a metaphorical claim that shows his disa- greement with the proposition that no man is an island. If we consider the cognitive proposals about metaphorical meaning, Will’s utterance of one of the declarative sentences included in the previous text, (2), (2) All men are islands expresses a metaphorical proposition that involves a metaphorical meaning. The utterance of (2) is metaphorical because it includes a focus, the expression in italics, which is used metaphorically and gets a metaphorical meaning. The focus refers to the source do- main, ISLAND, and ‘men’ is the frame of the metaphorical utterance that refers to the target domain, MAN. Using the source to talk about the target, the speaker constructs the metaphorical concept, MAN AS ISLAND. Will intends the addressee to see the target do- main from the perspective of some source domain if he wants to understand them.2 2 ‘Metaphorical meaning,’ ‘metaphorical utterance,’ ‘target domain,’ ‘metaphor- ical concept,’ ‘focus,’ ‘frame,’ ‘source domain,’ are terms normally used in the Metaphorical Argumentation 395 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Not all utterances in which a metaphorical concept is involved are like Will’s utterance of (2), an utterance that triggers a new metaphorical concept for its interpretation. The utterances of (3)- (4), examples taken from Lakoff and Johnson (1980/2003, p. 53)— (3) His theory has thousands of little rooms and long, winding corridors (4) These facts are the bricks and mortar of my theory —trigger for their interpretation a conventional metaphorical concept, THEORY AS BUILDING, that must be extended. However, although this concept is also triggered by the utterance of (5), (Lakoff and Johnson 1980/2003, p. 46): (5) We will show that theory to be without foundation This utterance is conventional and demands a literal interpretation. When utterances such as that of (5) are included in argumentation, this argumentation might be described as ‘metaphorical’ since it includes a lexicalized expression representing a ready-made part of the conventional metaphorical concept, but they are interpreted literally since their metaphorical meanings are conventionalized. We do not focus our study on this type of argumentation and do not use ‘metaphorical argumentation’ in this sense. In this paper we explore, instead, the role of novel metaphor in communicative acts of arguing to show that novel metaphorical utterances exemplified by (1)-(4), utterances that are interpreted metaphorically, can appear in communicative acts of arguing. Furthermore, adopting Toulmin’s (1958/2003) notation of the components of an argument, we show how they can operate in argumentative practice as utterances to express data, claims or warrants. To do that, we consider, in section 2, some theoretical aspects of argumentation in the service of the construction of an argument structure. In section 3, we expose the sceptical concep- theories on metaphor. For more information about them, see Romero and Soria (2021). 396 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. tion of metaphorical argumentation, from which it follows that it is not possible to argue by means of metaphorical utterances because they do not communicate metaphorical propositions. In section 4, we will see the deflationary conception of metaphorical argumen- tation in which the speaker communicates metaphorical proposi- tions that can be communicated literally. In section 5, we explain how, in order to interpret a novel metaphorical utterance, a new metaphorical concept is constructed and how it provides the meta- phorical proposition meant with a genuine cognitive value. This proposal allows us to argue, against the sceptical and deflationary proposals on metaphorical argumentation, that when novel meta- phorical utterances intervene in natural language argumentation, their specific insight has effects on the argument. For that reason, metaphorical argumentation deserves to be the object of a detailed study. 2. The structure of arguments Argumentation has been considered a rational discursive practice and arguments have been considered representations of the inferences that arguers make. The structure of the arguments is obtained by displaying a model of argument but not all models permit to construct the arguments that operate in argumentation. In the formal logic model, it is usual to consider an argument as a structure that indicates that a proposition would have to be the case if something else is the case; the first proposition or conclusion would be entailed by the second or premise. But not all inferences that arguers make can be represented by entailments and there are entailments that represent inferences that arguers never make. An informal logic model of argument such as Toulmin’s (1958/2003), instead, permits to produce arguments that represent material inferences, the kind of inferences made in argumentation. According to this model, the structure of arguments includes more elements than premises and conclusion and the entailment relation is not the only one that characterizes their relation. The elements of arguments, however, are propositions as in formal logic because propositions are the kind of items that possess logical properties such as the possibility of entering in relations of consequence and Metaphorical Argumentation 397 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. incompatibility with other items. In Toulmin’s model, the structure of the arguments includes several types of propositions. The claim or conclusion, C, is the proposition whose value the arguer is trying to establish. Whenever we make an assertion, we put forward a claim which can be questioned by our interlocutors. The first step in defending our claim is presenting our data, D. Data are the evidence, facts or information used to support or to justify the claim (Toulmin 1958/2003, p. 90). However, if the data provided are not sufficient to establish its value, we may be asked to articulate the relationship between the data and the claim, which is usually implicit. This relationship may be explicitly represented by the warrant, W. Warrants are hypothetical and general propositions that can act as bridges between data and claims, legitimizing the type of step required by the claim from the data. They can be expressed as follows: “Data such as D entitle one to draw conclusions, or make claims, such as C” (Toulmin 1958/2003, p. 91). Warrants, however, justify the step of certain data to a certain claim to different degrees: unequivocally, if the appropriate data are available and allow us to qualify our conclusion with the adverb ‘necessarily’; tentatively, if the step from data to conclusion is subject to conditions, exceptions or qualifications, which leads us to qualify the conclusion with modal terms such as ‘probably’ or ‘presumably.’ In this way the claim can be weakened or reinforced by a qualifier, Q, which indicates the strength of the relationship between the data and the claim. In addition, the arguments may involve rebuttal conditions, R, that mark the circumstances in which the general authority of the warrant is voided, preventing the derivation of the conclusion from the data. Although warrants are generally accepted as sufficient authority, sometimes it is not done immediately and requires a backing, B, from which the warrant can get the authority. A backing thus consists of a series of reasons for which to adopt the warrant. As backings are field-dependent (Toulmin 1958/2003, p. 96), they amount to categorical statements based, for example, on statistical reports, taxonomic systems, or legal rules. The visual representation of the structure of the arguments according to Toulmin (p. 97) is the following: 398 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Data So, Qualifier, Claim Since Unless rebuttal Warrant On account of a Backing In example (1) we can identify the following elements accord- ing to Toulmin’s structure of an argument: Figure 1. Argumentation of (1) from Toulmin’s model. In this example novel metaphorical utterances communicate meta- phorical propositions that act as data and conclusions and are part of the warrant. In argumentation theory, however, arguments are not always conceived to have all these elements. This happens when argu- mentation is conceived as an illocutionary act complex at a textual D: A hundred years ago, people had to depend on other people. No one had TV or CDs or DVDs or videos... or home espresso makers. Actually, they didn’t have anything cool. In contrast, now with the right supplies (TV, CDs, DVDs….) and the right attitude people can make themselves a little island paradise, a sun-drenched, tropi- cal island visited by young Swedish tourists. At least Will is that kind of island, in particular, he is Ibiza. W: Given that now a person with the right supplies doesn’t have to depend on other people, we can take it that this is an island age. C: Now is the time for a person to be an island. This is an island age. So, Q: Presu- mably Metaphorical Argumentation 399 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. level. In Pragma-Dialectics, for example, arguing as a speech act complex is giving at least one reason (datum) and having a warrant (van Eemeren and Grootendorst 1982; van Eemeren, Grootendorst and Kruiger 1984). If arguing were just that, the structure of the argument would only have the data and the warrant in Toulmin’s model. However, according to Pragma-Dialectics, there is a more general sense of argumentation which includes the above speech act complex (datum and warrant) as a phase or a move in a critical discussion. It is this type of argumentation that results in argu- ments with more ingredients than datum and warrant. The claim or conclusion, for example, is one more move excluded in arguing as an illocutionary kind. Another proposal of arguing as a speech act is that of the linguistic normative model of argumentation, LNMA (Bermejo-Luque 2011): to argue is to adduce reasons to justify some conclusion. In this case, the “warrant” is outside the illocu- tionary act of arguing while the conclusion is included. Regardless of their differences, these proposals of arguing as an illocutionary act complex are of great interest because both make it much clear- er that being a datum or being a claim depends on a second-order speech act which is possible thanks to the relationship of justifica- tion between the propositions expressed by individual speech acts. A mere sequence of utterances does not become a case of argu- mentation. If my verbal behaviour is questioned, the next move will be to give reasons in favour of my claim, and this is possible when a warrant is also communicated (often implicitly). With the analysis of the utterances of some expressions in (1) and other examples, we aim to show that, by everyday metaphori- cal uses of language, speakers convey metaphorical propositions which can take the roles of data, claim or warrants in natural lan- guage argumentation. This explanation, however, is incompatible with sceptical accounts of metaphorical meaning. 3. Sceptical accounts of metaphorical argumentation Many philosophers have rejected and still reject any non-literal propositional effect in metaphorical interpretation. The paradigm of this position is found in Davidson (1978). In his opinion, the metaphorical use of language is related only to the ordinary mean- 400 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. ings of words, to the literal meaning of the sentence used. In the case of metaphor, the speaker believes that the literal meaning of the sentence is obviously false or trivially true. (2) (2) All men are islands. expresses an obviously false literal proposition, while (6) (6) No man is an island. expresses a trivially true one.3 These literal propositions prompt us to see one thing as another. They invite us to attend to some like- ness (Davidson 1978, p. 40) but this invitation does not constitute a meaning. Being so, metaphor produces some distinctive non- propositional effects ‘caused,’ rather than ‘meant’ (1978, p. 46). There is no metaphorical meaning related to the source domain terms, ‘island’ in (2) and (6), because using and understanding a metaphor is, for Davidson, a creative effort that is not guided by rules and “the act of interpretation is itself a work of the imagina- tion” (1978, p. 31). Metaphor is not suitable for communication. Nowadays, Lepore and Stone (2015) also have a sceptical view of metaphorical meaning. Metaphorical effects are non- propositional and do not become part of the set of propositions in the common ground between speakers. For them, metaphorical propositions are incompatible with a pragmatic notion of speaker meaning which is determined by an intention aimed to produce certain effects by means of the recognition of speaker’s intention (Grice 1957/89, p. 219). The recognition of speaker’s intention to produce certain effects is a necessary condition for the audience, and a reason, to reach them. However, according to Lepore and Stone (2010, p. 170), when the speaker uses language metaphori- cally his intention is that the audience appreciate certain similari- ties as part of imagination, but the hearer’s appreciation of the similarities is achieved by how the world is rather than by means of the hearer’s recognition of that intention. Thus, metaphorical 3 (6) might well be used literally, for example in a conversation concerning the analytical properties of ‘man’ in which its truth is not obvious. Metaphorical Argumentation 401 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. effects cannot be metaphorical meanings in the speaker meaning and no metaphorical proposition is communicated. But then, how do they explain the fact that speakers can agree and disagree when they use metaphorical utterances? How do they explain the fact that Will disagrees with (6)? For sceptics, accept- ing or refusing a metaphorical utterance is accepting or refusing a cognitive image understood as the appreciation of certain similari- ty, rather than agreeing or disagreeing with a metaphorical propo- sition, but how is it possible to reject a similarity that you have already appreciated? Furthermore, how can Will disagree with (6) if the only proposition available is obviously true? If metaphorical meaning is not part of the proposition, no reasons can be adduced against it and no warrant can be communicated. From the rejection of metaphorical meaning follows the scepti- cal position on metaphorical argumentation. If there is no meta- phorical communication, there is no rational discursive practice as metaphorical argumentation would have to be. For sceptics it is not possible to argue metaphorically because metaphorical utter- ances do not express metaphorical propositions. As metaphorical utterances can only express literal proposi- tions, if one tries to argue with them, the argumentation would be at best literal. Now, since the literal propositions expressed by metaphorical utterances are either obviously false or trivially true, it does not seem likely that they can have or need epistemic justifi- cation. Metaphorical utterances cannot be part of an illocutionary act complex of arguing. If a metaphorical utterance expresses an obviously false propo- sition, it will not constitute an assertion. An assertion conditions what the speaker can say as a consequence of performing this type of speech act. As Brandom (1994) argues, an assertion commits the speaker to defend herself if challenged by an addressee. And she defends her claim by giving reasons. However, there is no reason that an arguer (e.g., Will in About a boy) can give to defend that he is literally Ibiza. Therefore, the literal proposition is not asserted, it is not communicated and can act neither as data nor as claim. Let’s imagine Will again. In the course of a conversation on people’s responsibilities to others, his interlocutor might intone (6). In this scenario, Will utters (7). 402 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. (7) I´m not affected by or responsible for others; I’m Ibiza. Understood from the sceptical position, the utterance of (7) in- cludes an assertion. Will asserts that he is not affected by or re- sponsible for others. However, he does not give any reason for that. His utterance does not include any data. Taken literally, the sentence ‘I’m Ibiza’ expresses a proposition that is obviously false. A person is an animated object, while Ibiza is an island. Nobody would believe the literal proposition that Will is Ibiza, nobody could assert it and no reasons can be provided to support it. When metaphorical utterances are obviously false in their literal interpretation, their interpretation cannot act as a datum in an argument, it cannot justify another proposition. Furthermore, it cannot act as a claim because strictly speaking no one can literally claim that he is Ibiza. Thus, the utterance of (8) (8) I’m Ibiza; I’m pretty cool and attract young Swedish tourists. would not be a case of argumentation from this theoretical per- spective. The sceptical proposal on metaphorical meaning cannot explain the presence of trivially true propositions expressed by metaphori- cal utterances in arguments either. The utterances of (9) or (10) (9) You should relax lockdown; no man is an island. (10) No man is an island; all people depend on others. cannot be explained as cases of argumentation. The utterance of (9) communicates a recommendation and an assertion but they cannot be a claim or a datum. The allegedly asserted literal propo- sition that man is not a landform cannot be the datum to recom- mend the relaxation of lockdown in times of coronavirus. How do we justify the step from human beings not being islands, some- thing trivial, to the claim that we should relax lockdown in times of coronavirus? In addition, the allegedly asserted literal proposi- tion that man is not a landform cannot be a claim in the utterance of (10) since an obvious truth does not require justification. In Metaphorical Argumentation 403 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. cases such as (9) and (10) there is no way in which we can estab- lish the kind of data-claim relation needed for communicative acts to become part of argumentation. No warrant is available. When metaphorical meanings are rejected, an explanation of the behaviour of the utterances of (7)-(10) as acts of arguing is unavailable. In everyday uses of language of this type, however, it seems that the speaker is actually giving reasons for her claims and this is possible only if the metaphorical utterances in each of them convey metaphorical propositions. Only with the metaphorical claim in (8) and (10) and the metaphorical data in (7) and (9) can the step from data to claim be available. Only when metaphorical utterances in the utterances of (7)-(10) are interpreted metaphori- cally we can say that their utterers are arguing. 4. Deflationary account of metaphorical argumentation According to the traditional account of metaphor, metaphorical propositions can be expressed by metaphorical utterances, but these can be replaced by utterances that express those same meta- phorical propositions literally. The metaphorical interpretation appeals to a substitution mechanism which holds that the focus of a metaphorical utterance is used instead of another literal expres- sion equivalent to it. In the utterance of (11), (11) Richard is a lion. ‘a lion’ is used instead of ‘brave’ because to be brave is the prop- erty of lions that can be attributed literally to Richard. This is a deflationary view of metaphorical content since it has no distinc- tive cognitive value. The immediate question that arises is why anyone would choose to communicate the intended meaning by using (11) if she can use (12)? (12) Richard is brave. 404 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. The main reason provided by the defendants of this view is that the listener enjoys deciphering the puzzle that the speaker express- es with the metaphorical expression (Black 1954-55, p. 281). From this view of metaphor as a form of expression whose val- ue is just ornamental, it would be accepted that there are meta- phorical arguments that involve at least one metaphorical proposi- tion. Among the aspects that are generally considered essential in argumentation, only the rhetorical ones are of value in metaphori- cal argumentation. The logical or dialectical aspects do not de- serve to be studied apart because their characteristics would be reducible to that of literal argumentation. Metaphorical arguments do not differ essentially from literal arguments. The value of met- aphorical argumentation is at most ornamental. The utterance of (13) or the utterance of (14) (13) Richard is English; then he is a lion. (Example adapted from Grice 1975/89) (14) Human beings are not responsible for their actions; they are machines. (Example taken from Wagemans 2016) can be analysed as metaphorical arguments. The speaker of (13) offers the assertion that Richard is English as a reason to justify the metaphorical conclusion that he is a lion and the speaker of (14) offers the metaphorical factual proposition that human beings are machines as a reason for his conclusion that human beings are not responsible for their actions. The reconstruction of these ar- guments under Toulmin’s model is portrayed in Figure 3. However, the utterances of (13) and (14) communicate the same propositions as the utterances of (15) and (16). (15) Richard is English; then he is brave (16) Human beings are not responsible for their actions. They are willnessless beings Metaphorical Argumentation 405 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Figure 2. Argumentation of (13) from Toulmin’s model. Figure 3. Argumentation of (14) from Toulmin’s model. The reconstruction of the arguments communicated by the utter- ance of (15) and (16) would give us an argumentative structure identical to those represented in Figures 2 and 3. ‘A lion’ and ‘brave’ contribute to the proposition with the same meaning be- cause in its metaphorical use, ‘lion’ means brave. Similarly, in its metaphorical use, ‘machine’ means willnessless being. The only additional move in the analysis of (13) and (14) has to do with the process of inverting the substitution in order to get the same literal propositions intended by the speaker and conveyed indirectly by their utterances. The process of inverting the substitution affects the conclusion in (13), the datum in (14) and the warrant in both. This illustrates that any theory of metaphor that establishes a substitution mechanism for metaphorical interpretation will imply, as Garssen (2009) argues, a deflationary approach to metaphorical argumentation which assigns it just non-propositional effects such as pleasure. This, however, does not seem to be very convincing. As Black says: “When in doubt about some peculiarity of lan- guage, attribute its existence to the pleasure it gives a reader. A D: Human beings are machines. D: Richard is British. So, Q, C: So, he is a lion. So, Q, C: So, necessarily, human beings are not responsible for their actions. W: Given that a man is British, we may take it that he is a lion. W: Given that a person is a machine, we can take it that he is not responsible for his actions. 406 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. principle that has the merit of working well in default of any evi- dence” (1954-55, p. 281). Another problem with this approach, as Black has also pointed out, is that it cannot explain the behaviour of all metaphorical utterances. What is the term that replaces ‘Ibiza’ in (7) and (8) or ‘island’ in (9) or (10)? Or if we look at example (1), what is the expression for which ‘island which is sun-drenched, tropical and a magnet for young Swedish tourists’ is replaced and that means literally what this expression means metaphorically? The interaction theory initiated by Black (1954/55) and devel- oped in more recent versions (Indurkhya 1986; Kittay 1987; Romero and Soria 1997-98) provide explanations about how the source domain acts by hiding, highlighting, introducing and reor- ganising aspects of the target domain. The cognitive value of metaphor, which arises from the conceptual interaction described by this approach, the mapping approach, makes it possible to account for metaphorical utterances in argumentation. If Black instead of thinking about the philosophers’ metaphors had thought about arguers’ metaphors, he would have started his famous article “Metaphor” by saying to draw attention to [an arguer]’s metaphors is to belittle him—like praising a logician for his beautiful handwriting. […] Yet the nature of the offence is unclear (Black 1954- 55, p. 273). Let us see why drawing attention to an arguer’s metaphors is not to belittle him. 5. Cognitive account of metaphorical argumentation According to Black, metaphorical utterances can only be explained if it is admitted that the properties of a concept in one domain can be projected upon another domain. In his words: The metaphorical utterance works by “projecting upon” the pri- mary subject a set of “associated implications,” comprised in the implicative complex, that are predictable of the secondary sub- ject (Black 1977, p. 442). Metaphorical Argumentation 407 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. In “Poverty is a crime,” Black (1977, p. 445) tells us, “assertions about crime are correlated one-to-one with corresponding state- ments about poverty” and this proposal is the basis of the mapping approach. Metaphor links two separate cognitive domains to con- ceptualise one as another. Some characteristics associated with the meaning of the focus change to be able to describe what is being talked about, the target domain. Let’s focus on how the interpretation of a novel metaphorical utterance such as the utterance of (2) can be explained according to the mapping approach. Its interpretation requires a contrast between the concepts involved, identifying the concept MAN as the target domain (Dt) and the concept ISLAND as the source domain (Ds) from which to describe the target domain. This contrast acti- vates metaphorical interpretation, making us project properties from Ds to Dt and conceptualise MAN AS ISLAND, a new metaphori- cal concept that alters the conceptual system stored in long-term memory provisionally. The projection is specified with a mapping, M, from the source domain, ISLAND, to the target domain, MAN. A domain can be represented by a set of terms forming its vocabu- lary, V, and by a set of sentences, Ss from the source domain and St from the target domain, which specifies how these terms give access to the information associated with the concept. The do- mains for ISLAND and for MAN can be represented as portrayed in Table 1. The interpretation of (2) depends on elaborating a partial admissible function F from the terms that belong to the source domain, the arguments of the function, to the terms that belong or that will belong to the target domain. This, applied to the example, would entail a partial function, F, between terms formed by pairs such as (‘island’ → ‘man’), (‘body of land’ → ‘individual human being’), (‘isolated’ → ‘independent’), (‘mainland’ → ‘fellow beings’), (‘climate’ → ‘culture’), (‘orography’ → ‘life conditions’). The application also consists of a subset of sentences from the source domain, S, which can be transformed coherently using F to information associated only with the target domain. 408 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Table 1: Representation of source and target domains involved in the construction of the metaphorical concept MAN AS ISLAND Source domain (Ds): ISLAND Ds = Vs = {‘island,’ ‘surrounded,’ ‘water,’ ‘separated,’ ‘mainland,’ ‘ocean,’ ‘wildlife,’ ‘climate,’ ‘land,’ ‘sun- drenched,’ ‘beach,’ ‘tourist,’ etc} Ss = [1s] An island is a body of land isolated from mainland by water, [2s] There are numerous islands in the ocean, lakes, and rivers around the world, [3s] Islands vary greatly (in origin, climate, orography…), [4s] Because of isolation, many islands have been home to some unusual and fascinating wildlife, [5s] For centuries, islands have been stopping places for ships, [6s] Some islands became notorious as pirate bases, [7s] Some islands were once part of mainland, others were formed by eruptions of volcanoes on the ocean floor, [8s] Today, some sun-drenched islands are often visited by tourists, [9s] Some islands are part of island chains, etc. Target domain (Dt): MAN (OR PERSON) Dt = Vt = {‘man,’ ‘human,’ ‘intelligent,’ ‘culture,’ ‘society,’ ‘marriage,’ ‘fellow beings,’ ‘cooperation,’ etc} St = [1t] Man is a human being, a highly intelligent primate, [2t] Man is an adult male human being, as distinguished from a woman and a child, [3t] There are numerous men in different countries around the world, [4t] Men vary greatly (in origin, culture, life conditions…), [5t] Man has a variety of needs (food, sex, …), [6t] For centuries, man’s cooperation with his fellow beings has been essential to get his needs fulfilled, [7t] Man cannot live separated from his fellow beings, [8t] Man’s social relations are often legally regulated (e.g. marriage, parenthood, inheritance), [9t] In certain cultures, sexual relations are permissible only under marriage, [10t] Man’s attitudes, beliefs, morals and ideals vary greatly, etc. In the example, S could be formed by sentences like [1s], and [3s] of Table 1 (in bold letters), sentences that include properties that are not literally applied to man. These sentences are transformable Metaphorical Argumentation 409 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. by F because each of its terms belongs to the arguments of this function or belongs directly to the vocabulary of the target domain. Source domain terms get the meaning of the target domain terms to which they are applied in F, meaning that is established in the metaphorically restructured target domain.4 When transforming the sentences of S, we find others only in terms of the target domain, [11t’] and [4t’], in Table 2 below. If the union of these sentences with the sentences of the target domain is coherent, that is, if this union is true in at least one model, then the sentences of S have been coherently transformed by means of F into sentences of the target domain. Coherence is an inferential requirement for mappings: we can only map the transformed information from the source domain that does not make our conception of the target domain incoherent. The mapping M for (2) generates the metaphorically restructured conception of MAN, MAN AS ISLAND (Dt’ or MANM for short), characterised by the structural constraints of Table 2. Table 2: Representation of the metaphorically restructured target domain, Dt’, in the interpretation of (2) Restructured target domain (Dt’ or MANM): MAN AS ISLAND [11t’] Man is an individual human being independent from his fellow beings. [4t’] Men vary greatly (in origin, culture, life conditions…), [6t’] For centuries, man’s cooperation with his fellow beings has been essential to get his needs fulfilled, [7t’] Man cannot live separated from his fellow beings. 4 In this way, the interaction approach dissolves the emergent properties problem of how properties not associated with the source concept can be activated; a problem that arises in categorization and other substitution views (Glucksberg 2003; Wilson and Carston 2006; Sperber and Wilson 2008). According to the mapping approach, the properties of the source domain do not have to be applied literally to the topic. Some characteristics of ISLAND change in order to describe MAN rather than ISLAND (Black 1954-55, p. 289). 410 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. With [4t’] and [6t’] nothing new is added to the target domain from the source domain, some information is reinforced and high- lighted by the relational similarities that are revealed by their alignment with the characteristics activated in the source domain. Since the information in [2t], [3t], [5t], [8t], [9t], and [10t] (see the right column of Table 1) is not selected, it is attenuated. In addi- tion, when the description of the target domain from the source domain adds information which is not present in the first one, but is consistent with it and relevant to understand the metaphorical utterance, novel properties emerge in the target domain as in the case of [11t’]. Metaphorical reconceptualization creates similarity, something which has been repeatedly defended by authors in the Blackian tradition (Indurkhya 1986; Romero and Soria 1997-98 and 2016; Gentner and Wolf 2000; Keating and Soria 2019). However, with the introduction of [11t’] the restructured target domain is not coherent unless we drop [7t]. In fact, that is the intention of the speaker in uttering (2). By contrast the intention of the speaker in uttering (6) is to deny [11t’] and keep [7t]. To inter- pret (2), assumption [7t’] is highlighted but only to be rejected. To interpret the utterance of (6) (included in the utterances of (9) and (10)), [11t’] is introduced and rejected at the same time. To appre- ciate the different ways to make the mapping in each case, let us compare Table 2 above with Table 2’ below. Table 2’: Representation of the metaphorically restructured target domain, Dt’, in the interpretation of (6), (9), (10) Restructured target domain (Dt’ or MANM): MAN AS ISLAND [11t’] Man is an individual human being independent from his fellow beings. [4t’] Men vary greatly (in origin, culture, life conditions…), [6t’] For centuries, man’s cooperation with his fellow beings has been essential to get his needs fulfilled, [7t’] Man cannot live separated from his fellow beings. Metaphorical Argumentation 411 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. These examples show that there are different possible elabora- tions of the mapping and the specific one involved in the proposi- tion intentionally communicated by the speaker will be derived by the hearer’s search of the speaker’s intended meaning and the evidence provided. The interpretation of the utterance of (1), involves the new conceptualization of man which takes into ac- count the more independent lifestyle of people in western society. It is a metaphorical conceptualization which involves the selection of [7t] to reject it rather than to endorse it and it is from this new perspective of man that Will can claim that this is an island age. Since assumption [7t] is part of the structural constraints of the concept MAN stored in long term memory, its rejection requires giving reasons to drop it from our conventional conceptualization and this is what Will is trying to do. He presents his lifestyle as an example of how a man can be an island if he is a certain kind of island and he elaborates a hyponym of the concept MAN AS ISLAND: the ad hoc metaphorical concept WILL-TYPE OF MAN AS IBIZA-TYPE OF ISLAND (see Table 3 below). 412 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Table 3: Representation of source and target domains and restructured target domain, Dt’, in the interpretation of (1) Source domain (Ds): IBIZA-TYPE OF ISLAND Ds = Vs = {‘Ibiza,’ ‘island,’ ‘beach,’ ‘sun- drenched,’ ‘beach,’ ‘tourist,’ ‘separated,’ ‘mainland,’ ‘Spain,’ ‘Balearic islands,’ ‘fun,’ ‘sexual freedom,’ etc} Ss = [1s] Ibiza is a Spanish island, a Spanish body of land isolated from Spain’s mainland by water, [2s] Ibiza is a sun-drenched island with fantastic conditions to enjoy life, [3s] Ibiza has beautiful beaches, [4s] Ibiza often receives young tourists in search of fun, [5s] In the sixties, foreigners arrived in Spanish islands such as Ibiza and changed the local lifestyle. [6s] In the sixties, Ibiza was a very attractive place for many young female tourists (Swedish, Danish, German…) with a look and behaviour very different from local women (they were generally called “las suecas” and became the symbol of sexual freedom in Spain). [7s] Ibiza is a body of land isolated from Spain’s mainland but part of an island chain, the Balearic Islands, etc. Target domain (Dt): Will-TYPE OF MAN Dt = Vt = {‘man,’ ‘young,’ ‘male,’ ‘TV,’ ‘home espresso maker,’ ‘independent’ ‘unmarried,’ ‘fellow beings,’ ‘sex without marriage,’ etc} St = [1t] Will-type-of-man is an individual human being independent from his fellow beings, [2t] Will lives alone in a fantastic house, [3t] Will has very good supplies at home (TV, DVDs, CDs, home espresso maker), [4t] Will is free to make decisions on his own, [5t] Will’s attitude makes him enjoy a very independent lifestyle, etc. Metaphorical Argumentation 413 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Restructured target domain (Dt’ or Will-TYPE OF MANM): WILL-TYPE OF MAN AS IBIZA-TYPE OF ISLAND [1t’] Will-type-of-man is an individual human being independent from his fellow beings, [2t’] Will lives alone in a fantastic house, [3t’] Will has very good supplies at home (TV, DVDs, CDs, home espresso maker), [4t’] Will-type-of-man is free to make decisions on his own, [5t’] Will’s attitude makes him enjoy a very independent lifestyle, [6t’] Will-type-of-man often receives occasional visitors in search of fun, (new coming from [4s]) [7t’] Will-type-of-man is a very attractive man for occasional female beautiful visitors with the right attitude (sexual freedom typical of “las suecas” in Spain) to enjoy his good conditions with him but with no intention to get permanent local rights. (new coming from [6s]) If through the narration of Will’s story, we can represent as true or probably true the metaphorical proposition that today there are men of a certain type that are islands of a certain type, it is not true that no man is an island. If some object within the extension of the concept MAN (a Will-type of-man) belongs to the extension of the concept WILL-TYPE OF MAN AS IBIZA-TYPE OF ISLAND, there is a reason to claim that, unlike other non-island ages, this is an island age, an age in which, with the right supplies and the right attitude, some people can be an Ibiza-type-of-island. For the speaker to communicate his intended proposition by the utterance of “this is an island age,” the term ‘island’ gets a shifted provisional meaning which depends on the metaphorical context created for the occa- sion of the utterance. This is possible since the metaphorically restructured target concept causes a shift in the context of interpre- tation of this metaphorical utterance. The terms from the source domain will be interpreted, among other things, from the concept MAN AS ISLAND and more particularly from the concept WILL-TYPE OF MAN AS IBIZA-TYPE OF ISLAND. From this new context, a new meaning is produced, at least for the terms from the source domain used in (1). Thus, ‘island’ does not mean in that context, BODY OF LAND ISOLATED FROM MAINLAND, rather, it acquires the meaning that man has in MANM and ‘Ibiza’ and ‘that kind of island’ do not mean in that context, TROPICAL SUN-DRENCHED ISLAND, rather, 414 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. they acquire the meaning that Will has in WILL-TYPE-OF-MANM. This provisional metaphorical meaning is conceivable only from the metaphorically restructured concepts MANM and WILL-TYPE- OF-MANM. This provisional metaphorical meaning is part of the meaning that the speaker intends to communicate and that the listener obtains following pragmatic principles. The recognition of the speaker’s intention is essential in the reconceptualization of the target domain from which the metaphorical utterance is interpret- ed. It is not the similarities of the world that are recovered but rather the similarities that the speaker wants us to recognise to be able to claim that this is an island age. These metaphorical mean- ings have peculiar characteristics that allow us to defend the cog- nitive value of the metaphorical use of language. Only from this metaphorical context of interpretation can we interpret (1) and other utterances in the film, such as the one which appears at the very end when Will utters (17) (17) Every man is an island. And I stand by that. But clearly, some men are part of island chains. Below the surface of the ocean they’re actually connected. With this utterance, the novel metaphorical concept is extended. [8t’] is introduced and we get an extended restructured target domain (see Table 4 below). Table 4: Representation of the metaphorically restructured target domain, Dt’, in the interpretation of (17) Extended restructured target domain (Dt’ or WILL-TYPE OF MANM): Will- TYPE OF MAN AS IBIZA-TYPE OF ISLAND [1t’] Will-type-of-man is an individual human being independent from his fellow beings, [2t’]…-[7t’] [8t’] Today, Will-type-of-man can be legally independent from his fellow beings and yet can keep connected below the surface to some of them. (new, coming from [7s]) Metaphorical Argumentation 415 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. To interpret (17), [8t’] is introduced in the target domain to highlight the kind of “independent connection” that man can enjoy today. This assumption is now part of the restructured target do- main which highlights an aspect of Ibiza that becomes relevant in what he intends to communicate. WILL-TYPE-OF-MAN AS IBIZA, AN ISLAND IN THE BALEARIC ISLANDS becomes part of the context of interpretation to get the proposition intentionally communicated by the speaker with the utterance of ‘some men are part of island chains. Below the surface of the ocean they’re actually connected.’ At the end of the film, Will still claims that he is Ibiza but he realizes that his being part of an island chain metaphorically im- plicates that his friends are the other Balearic Islands which are Ibiza-type-of-islands. More objects within the extension of the concept MAN (some women and a child included) belong to the extension of the concept WILL-TYPE OF MAN AS IBIZA-TYPE OF ISLAND and thus, there is more evidence to claim that this is an island age. The topic of the film About a boy is neither Will’s life nor Ibiza. The narration of Will’s story is used to determine the target domain, WILL-TYPE-OF-MAN, and Ibiza is used to communicate the source domain, THAT KIND OF ISLAND or IBIZA-TYPE-OF-ISLAND. WILL-TYPE-OF-MAN AS IBIZA-TYPE-OF-ISLAND is a new and ex- tended metaphorical concept, it is cumulatively constructed as the narration of Will’s story evolves. And it is the context from which to get also the metaphorical conclusion: THIS IS THE TIME TO BE A MAN AS ISLAND. In this way, metaphorical arguments have, like some of the metaphorical utterances they are elaborated with, a distinctive cognitive value. This value is reflected in the fact that the provi- sional metaphorical meaning of the focus is a meaning that de- pends on the context that is generated by the interpretation of the metaphorical utterance, a context characterised by a new or ex- tended metaphorical concept. Metaphorical utterances in argumen- tation communicate data, conclusions, or even warrants. Their assessment depends on the characteristics of the metaphorical propositions conveyed, propositions with which speakers commit themselves in a new context in which metaphorical provisional meanings arise. This context is important not only because it is of 416 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. use for the interpretation and analysis of argumentation, but also because it is the context from which the justification of the relation between propositions is assessed. If we evaluate from the new metaphorical context the meta- phorical argumentation of the example (1), we can say that dialec- tically it is good because something that many accept, that no man is an island, is incompatible with what the speaker claims, that this is an island age, that now is the time to be an island. Whether or not it is a good attempt to show that the claim is correct depends on whether the reasons justify it, as Bermejo-Luque (2011) would say. The warrant itself could be called into question since being able to be something does not mean that, for that reason, this is the time to be what we can be. Another issue is that not all metaphori- cal data seem to be acceptable. For example, nowadays not every- one can be a man-as-island since not everyone can have the neces- sary conditions to be one. In any case, what could be argued against (6) is that there are men-as-islands. However, once we have obtained the metaphorical propositions from the new meta- phorical context, the assessment of the metaphorical argumenta- tion of the example (1) from this context is due to causes unrelated to the metaphoricity of the utterances. 6. Conclusion As we have seen, the sceptical or deflationary proposals on meta- phorical meaning lead to a sceptical or deflationary view on meta- phorical argumentation. If metaphorical meaning is rejected, it is impossible to explain cases of metaphorical argumentation that, in fact, continually occur in our daily uses of language. If instead the metaphorical meaning is admitted and characterized as the mean- ing of another expression, the logical aspects of metaphorical argumentation are identical to those of the literal. Examples of metaphorical uses of language such as (1) show that not all metaphorical utterances can be literally paraphrased and thus any explanation of their interpretation not only requires accepting the existence of metaphorical meanings but also, as maintained by interaction theorists, that these do not coincide with the literal meanings of other expressions. Many metaphorical Metaphorical Argumentation 417 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. utterances are irreducible to the literal and intervene in our daily arguments. By the same token, metaphorical argumentation cannot be reduced to literal argumentation. Without the specific cognitive insights of these metaphorical utterances, the arguments in ques- tion would not be possible. Acknowledgements Research for this article was supported by project PGC2018- 098236-B-I00. Our interest on the role of metaphor in argumenta- tion has evolved in the course of extended discussions about ar- gumentation with the members of the research group ActArg (https://speechactofarguing.wordpress.com/). We would like to thank them for their enthusiastic and rigorous discussions and especially Lilian Bermejo for her continuous encouragement and her interest in our work. We are also very grateful to two anony- mous reviewers for their insightful comments to a previous ver- sion. References Black, Max. 1954-55. Metaphor. Proceedings of the Aristotelian Society 55: 273-294. Black, Max. 1977. More about metaphor. Dialectica 31(3-4): 431-456. Bermejo-Luque, Lilian. 2011. Giving reasons: A linguistic pragmatic approach to argumentation theory. Dordrecht: Springer. Bilstrup Finsen, Andreas, Gerard J. Steen and Jean H. M. Wagemans. 2019. An argumentative reconstruction of the computer metaphor of the brain. Journal of Argumentation in Context 8(3): 317-335. Brandom, Robert. 1994. Making it explicit: Reasoning, representing, and discursive commitment. Cambridge, MA: Harvard University Press. Davidson, Donald. 1978. What metaphors mean. Critical Inquiry 5(1): 31-47. Eemeren, Frans H. van and Rob Grootendorst. 1982. The speech acts of arguing and convincing in externalized discussions. Journal of Pragmatics 6: 1-24. Eemeren, Frans H. van, Rob Grootendorst and Tajark Kruiger. 1984. The study of argumentation. New York: Irvington Publishers, Inc. Garssen, Bart J. 2009. Comparing the incomparable: Figurative analo- gies in a dialectical Testing procedure. In Pondering on problems of 418 Romero and Soria © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. argumentation. Twenty essays on theoretical issues, eds. Frans H. van Eemeren and Bart Garssen, 133–140. Dordrecht: Springer. Gentner, Dedre and Phillip Wolf. 2000. Metaphor and knowledge change. In Cognitive dynamics: Conceptual change in humans and machines, eds. Eric Dietrich and Art Markman, 295-342. Mahwah, NJ: Laurence Erlbaum Associates. Glucksberg, Sam. 2003. The psycholinguistics of metaphor. Trends in Cognitive Sciences 7: 92-96. Grice, H. Paul. 1957/89. Meaning. The Philosophical Review 64: 377– 88. In Studies in the way of words, 213-223. Cambridge, MA: Har- vard University Press. Grice, H. Paul. 1975/89. Logic and conversation. In Studies in the way of words, 22-40. Cambridge, MA: Harvard University Press. Indurkhya, Bipin. 1986. Constrained semantic transference: A formal theory of metaphors. Synthese 68(3): 515-51. Keating, John and Belén Soria. 2019. Populist metaphorical utterances. In Populist discourse: Critical approaches to contemporary politics, eds. Encarnación Hidalgo-Tenorio, Miguel Ángel Benítez-Castro and Francesca de Cesare, 209-226. Oxford: Routledge. Kittay Eva F. 1987. Metaphor. Its cognitive force and linguistic struc- ture. Oxford: Clarendon Press. Lakoff, George and Mark Johnson. 1980/2003. Metaphors we live by. Chicago: University of Chicago Press. Lepore, Ernie and Matthew Stone. 2010. Against metaphorical meaning. Topoi 29(2): 165-180. Lepore, Ernie and Matthew Stone. 2015. Imagination and convention: Distinguishing grammar and inference in language. Oxford: Oxford University Press. Locke, John. 1690/1999. Essay concerning human understanding. The Pennsylvania State University. Oswald, Steve and Alain Rihs. 2014. Metaphor as argument: Rhetorical and epistemic advantages of extended metaphors. Argumentation 28(2): 133-159. Perelman, Chaïm and Lucie Olbrechts-Tyteca. 1958. Traité de l’argumentation: La nouvelle rhétorique. Paris: Presses Universitaires de France. Pollarolli, Chiara and Andrea Rocci. 2015. The argumentative relevance of pictorial and multimodal metaphor in advertising. Journal of Ar- gumentation in Context 4(2): 158–199. Poppel, Lotte van. 2021. The study of metaphor in argumentation theory. Argumentation 35: 177-208. Romero, Esther and Belén Soria. 1997-98. Stylistic analysis and novel metaphor. Pragmalingüística 5-6: 373-389. Metaphorical Argumentation 419 © Esther Romero and Belén Soria. Informal Logic, Vol. 41, No. 3 (2021), pp. 391–419. Romero, Esther and Belén Soria. 2016. Against Lepore and Stone’s sceptic account of metaphorical meaning. Croatian Journal of Phi- losophy 16(2): 145–172. Romero, Esther and Belén Soria. 2021. Philosophy of language and metaphor. In The Cambridge handbook of the philosophy of lan- guage, ed. Piotr Stalmaszczyk, 639–658. New York: Cambridge Uni- versity Press. Santibáñez, Cristián. 2010. Metaphors and argumentation: The case of Chilean parliamentarian media participation. Journal of Pragmatics 42: 973-989. Sperber, Dan and Deidre Wilson. 2008. A deflationary account of meta- phor. In The Cambridge handbook of metaphor and thought, ed. Raymond W. Gibbs, 84-105. New York: Cambridge University Press. Toulmin, Stephen. 1958/2003. The uses of argument. Cambridge: Cam- bridge University Press. Wagemans, Jean. 2016. Analyzing metaphor in argumentative discourse. Rivista Italiana di Filosofia del Linguaggio 2: 79-94. Wilson, Deidre and Robyn Carston. 2006. Metaphor, relevance and the ‘emergent property’ issue. Mind and Language 21(3): 404-433. Xu, Cihua and Yicheng Wu. 2014. Metaphors in the perspective of argumentation. Journal of Pragmatics 62: 68-76.