UUnniivveerrssiittyy ooff WWiinnddssoorr SScchhoollaarrsshhiipp aatt UUWWiinnddssoorr OSSA Conference Archive OSSA 2 May 15th, 9:00 AM - May 17th, 5:00 PM WWhhoo iiss AAffrraaiidd ooff FFiigguurree ooff SSppeeeecchh?? Erik C W Krabbe University of Groningen Follow this and additional works at: https://scholar.uwindsor.ca/ossaarchive Part of the Philosophy Commons Krabbe, Erik C W, "Who is Afraid of Figure of Speech?" (1997). OSSA Conference Archive. 70. https://scholar.uwindsor.ca/ossaarchive/OSSA2/papersandcommentaries/70 This Paper is brought to you for free and open access by the Conferences and Conference Proceedings at Scholarship at UWindsor. It has been accepted for inclusion in OSSA Conference Archive by an authorized conference organizer of Scholarship at UWindsor. For more information, please contact [email protected]. WHO IS AFRAID OF FIGURE OF SPEECH? Erik C. W. Krabbe Faculty of Philosophy University of Groningen ©1998, Erik C. W. Krabbe Abstract: Aristotle's examples of the fallacy of Figure of Speech (or Form of Expression) are not very convincing to the modern reader. Most fallacy theorists have been happy to omit this fallacy from their accounts. But a study of Figure of Speech will lead one to find connections with twentieth- century analytical philosophy, where the idea that the apparent form of a sentence need not be its real logical form has been prominent. Other interesting issues concern the boundary between ambiguity and invalidity and the use of profiles of dialogue to describe the dialectics of this fallacy. *** 1. Introduction The term "Figure of Speech", when it denotes a fallacy, refers to the last one of the six fallacies dependent upon language on Aristotle's list in De Sophisticis Elenchis (On Sophistical Refutations). The name must have originated as a direct translation of the Latin term "figura dictionis", which again translates the original Greek term schêma tês lexeôs. The suggestion of a link with the figurative use of language is infelicitous, since such a connection, if it exists, can only be tenuous. The equivalent English name, "Form of Expression", is less misleading and more directly translates the Greek. For the modern reader, Aristotle's examples of this fallacy taste of the abstruse and the unconvincing. Partly this is caused by their extreme language dependency, in fact dependency upon peculiarities of the Greek language, which makes them stand in need of explanation of linguistic details in a way that is as pernicious with fallacy cases as it is with jokes. Partly also, this may be caused by the readers' unfamiliarity with the logical theory in the background: Aristotle's theory of the categories. These factors made the subject somewhat unappetizing for textbook authors. According to Charles Hamblin (1970: 25-26) "Very few modern writers even bother to mention it. Their difficulty, if they do so, is to find serious examples of it." But recently, in his book on fallacies of ambiguity, Douglas Walton devoted a chapter to the Fallacy of Figure of Speech (Walton, 1996, Hst. 5). According to Walton, the study of this fallacy shares some ground with "classical analytical philosophy in the twentieth century" where one often meets with the problem of misleading forms or misleading expressions (155). Could it be that the importance of figure of speech has been underestimated? Did Aristotle in these abstruse passages get hold of something that might still be needed for a better understanding of the practice of argument and its aberrations? In Sections 2 and 3 below this will be investigated. Another question is whether to subsume figure of speech in the class of fallacies of ambiguity. Wouldn't one rather categorize it as a failure of validity, a non sequitur? In Section 4 this question will lead to a more detailed classification of Aristotle's thirteen types of fallacy. Finally, at the end of Section 4, it will be briefly discussed what forms of dialectic interaction are adequate for those who meet with an (alleged) case of the fallacy of figure of speech within the context of a critical dialogue. 2. Aristotle's Treatment of the Fallacy of Figure of Speech Aristotle characterizes the fallacy of figure of speech as follows: Others [i.e,, other sophistical refutations] come about owing to the form of expression used, when what is really different is expressed in the same form, e.g. a masculine thing by a feminine termination, or a feminine thing by a masculine, or a neuter by either a masculine or a feminine; or again, when a quality is expressed by a termination proper to quantity or vice versa, or what is active by a passive word, or a state by an active word, and so forth with the other divisions previously laid down. For it is possible to use an expression to denote what does not belong to the class of actions as though it did so belong. Thus (e.g.) 'flourishing' is a word which in the form of its expression is like 'cutting' or 'building': yet the one denotes a certain quality-i.e. a certain condition-while the other denotes a certain action. In the same manner also in the other instances. (De Soph. El. 4: 166b 10-19)1 In Greek, as in German and Latin, nouns are divided into masculin, feminine, and neuter groups, each with characteristic paradigms for declension through cases. Generally, male subjects are referred to by words that follow a male paradigm, etc., but there are some exceptions, and it appears that a sophist may exploit these. Aristotle, however, does not provide us with an example of this particular abuse of form. The terms "quality", "quantity", "active"/"action", "passive", and "condition" refer to Aristotelean categories. One may wonder what particular forms of expression are supposed to be associated with each of these. For the categories of action and passion these must be the active and passive verb forms, as the following example clearly shows: Case 1 - Is it possible to be doing and to have done the same thing at the same time? - No. - But, you see, it is surely possible to be seeing and to have seen the same thing at the same time, and in the same aspect. (De Soph. El. 22: 178a 9-11) In such examples it is always the questioner who plays the role of the sophist and who tries to trip up the answerer. Here his tactic is to first let the answerer concede a general principle (that it is impossible to be doing what has been done), only to produce a jack-in-the-box "counterexample" against that very principle. The counterexample, however, owes any force it might have to the presupposition that "seeing something" is a special case of "doing something". But, an Aristotelean would be likely to say that "seeing" is not an action, but an affection ("passion"). In an ideal language, the verb that expresses "seeing" should have passive verb forms, not active ones as it has in Greek (excepting the future tense). For a contemporary readership this may all sound a bit far-fetched, but that this is actually what Aristotle had in mind as a solution is borne out by a second example: Case 2 - Is any mode of passivity a mode of activity? - No. - Then "he is cut", "he is burned", "he is struck by a sensible object"2 are alike in expression and all denote some form of passivity, while again "to say", "to run", "to see" are like one another in expression: but, you see, "to see" is surely a form of being struck by a sensible object; therefore it is at the same time a form of passivity and of activity. (De Soph. El. 22: 178a 11-16) In sum, the active conjugation of the verb "to see" produces misleading forms of expression, forms that make one mistake the denotation of "to see" for a kind of action, like the denotations of "to say" and "to walk". In truth, "seeing" is a kind of being struck by a sensible object and hence not an action, but an affection. Among the thirteen or so examples of the fallacy of figure of speech one finds in De Soph. El. 22 some cases, like those above, are strongly tied to peculiarities of the Greek language and to a mostly forgotten theory of categories, for instance those of action and affection. This makes them hard to follow, and often unconvincing. But some other cases are much easier to understand; to such an extent that in them the fallacy seems almost trivial. As is often the case in fallacy theory, we had better not take these examples as serious real life cases, but as caricatures that may help us to get a better understanding of the essential features of a type of fallacy. I shall quote two of these "trivial" examples and also reconstruct them as a fragment of dialogue. Case 3 Like the above are also the following arguments. It is asked if a man has lost what he once had and afterwards has not: for a man will no longer have ten dice even though he has only lost one die. (De Soph. El. 22: 178a 29-31) Reconstruction: - What one once had and afterwards has not, do we say that he has lost it? - Yes, thus we may define what it means to lose something. - Suppose, John has ten dice and loses only one of them. In that case John once had ten dice, but afterwards he has not. - Exactly. - So, according to our definition, John has lost ten dice. - Certainly. - But we supposed that he lost only one of them! - Good grief! Aristotle adds the following comment: No: rather it is that he has lost what he had before and has not now; but there is no necessity for him to have lost as much or as many things as he has not now [but had before]. So then, he [the questioner] asks the questions as to what he has, and draws the conclusion as to the whole number that he has [a quantity]: for ten is a number [a quantity]. If then he [the questioner] had asked to begin with, whether a man no longer having the number of things he once had has lost the whole number, no one would have granted it, but would have said 'Either the whole number or one [rather: some] of them'. (De Soph. El. 22: 178a 31-36) The form of the expression "ten dice" appears to be misleading: it is easily thought to denote one or other particular thing, whence it would be mistaken for a particular instance to which the definition of what it means to lose something might apply. In fact "ten dice" does not denote any particular thing but a particular quantity. The second trivial example is only briefly sketched by Aristotle. The succinct presentation of many examples, by the way, bears witness to their belonging to the common stock-in-trade. Case 4 Also there is the argument that 'a man may give what he has not got': for he has not got only one die. (De Soph. El. 22: 178a 36-38) Reconstruction: - Can anyone give what he has not got? - Of course not. - Suppose John has got two dice and Peter asks John to give him those dice. That could be the case, couldn't it? - Sure it could. - But John may decide to turn over only one die. - Certainly. - On the other hand, John has not got only one die. - No, he hasn't. - So John could give only one die, whereas he has not got only one die. - That seems to be the case. - But then John can give something what he has not got, namely "only one die": you are refuted. - Darn it! Aristotle's comment: No: rather it is that he has given, not what he had not got, but in a manner in which he had not got it, viz. [as] just the one. For the word 'only' does not signify a particular substance or quality or number [quantity] but a manner of relation, e.g. that it is not coupled with any other. It is therefore just as if he [the questioner] had asked 'Could a man give what he had not got?' and, on being given the answer 'No', were to ask if a man could give a thing quickly when he had not got it quickly, and, on this being granted, were to conclude that 'a man could give what he has not got'. It is quite evident that he has not proved his point: for to 'give quickly' is not to give a thing, but to give in a certain manner; and a man could certainly give a thing in a manner in which he has not got it, e. g. he might have got it with pleasure and give it with pain. (De Soph. El. 22: 178a 38-b 7) The expression "only one die", then, is misleading. For instance, in a sentence such as "Mary has only one die" it does not denote any particular substance but refers to a relation that the one die supposedly possessed by Mary bears to all other dice, namely that of not going together as far as possession by Mary is concerned. 3. How Serious is the Fallacy of Figure of Speech? The evidence presented thus far suffices to show that it is not excluded that someone somehow commits a fallacy of figure of speech. But the examples of this fallacy were either far-fetched or more or less trivial. They appeared far-fetched when the explanation of the delusion was based on an unfamiliar semantic theory of forms of expression, such as the theory that active and passive verb forms are, normally, associated with items from the categories of action and affection, respectively. They appeared trivial when the outlines of the underlying semantic theory were so obvious to us that we could not imagine how one would ever be fooled by such a fallacy; for instance, the semantic theory that "only one die" is not a singular term (i.e., that it is not a term suitable for reference to a definite individual). Are there other, and more serious, examples? A pretty serious example is found in a well-known text by John Stuart Mill. Hamblin, among others, quoted the case, and commented upon it: Case 5 It was given to J. S. Mill to make the greatest of modern contributions to this Fallacy by perpetrating a serious example of it himself. This was what the textbook writers were waiting for, and he is widely quoted. He said (Utilitarianism, ch. 4, p. 32) The only proof capable of being given that an object is visible, is that people actually see it. The only proof that a sound is audible, is that people hear it; and so of the other sources of our experience. In like manner, I apprehend, the sole evidence it is possible to produce that anything is desirable, is that people do actually desire it. But to say that something is visible or audible, is to say that people can see or hear it, whereas to say that something is desirable is to say that it is worthy of desire or, plainly, a good thing. Mill is misled by the termination '-able'. (1970: 26) This is certainly a serious and clear instance of fallacious reasoning based on figure of speech. According to Walton (1996: 161) there is, except for this one excellent case, no supply of serious cases that would fool an attentive and intelligent reasoner. Surely, one may easily make up some examples; for instance in a much used corner of ornithology: Case 6 Birds will become extinct. Tweety is a bird. So, presumably, Tweety will become extinct. In Case 6, it may be argued, the form of the expression "will become extinct" is misleadingly close to that of "will die". But anyhow it remains questionable, to say the least, to what extent such examples point out serious problems for those involved in actual reasoning. Among speakers of Dutch there used to be, in the seventies, some serious confusion about the meaning of the word "gijzelaar". The word definitely means "hostage", but many people were sure that it had to refer to a kidnapper or a hijacker. Their argument, based on the ending "aar", was very similar to Mill's. This argument would arguably constitute another serious real life example; however, in order not to burden the story with, besides some niceties of Greek genders, the details of Dutch word formation, I shall not repeat it here. The next real life example brings us close to philosophy. Who has never, in a philosophical mood, wondered why, in fact, the world exists? Why should there not rather be nothing? In more serious cases of this mood, one may even wonder whether anything at all exists. To cope with such moments I used to avail myself of the following clear-cut argument: Case 7 Either something exists or nothing exists. But if nothing exists, there still is something that exists, namely Nothing. In any case something exists. This argument (which I hope you will pardon as a youthful lapse) is an instance of the fallacy of figure of speech analogous to Case 3 ("ten dice") and Case 4 ("only one die"). Just like the expressions "ten dice" and "only one die" the word "nothing" can delude one into believing that one is dealing with a singular term denoting some particular entity. This delusion originates from the similarity in form between sentences containing these expressions and sentences containing real singular terms instead.3 Perhaps philosophy constitutes the most promising source of serious cases. We may return to Aristotle for a nice philosophical argument displaying the use of a misleading form of expression. This argument is known as the "Third Man" and was aimed at Plato's idealism. Case 8 Again, there is the proof that there is a 'third man' distinct from Man and from individual men. But that is a fallacy [added by the translator], for 'Man', and indeed every general predicate, denotes not an individual substance, but a particular quality, or the being related to something in a particular manner, or something of that sort. (De Soph. El. 22: 178b 36-40) The use of the Third Man in dialogue can be pictured as follows: - Do you accept Plato's idealism, namely that all individuals of a particular kind partake in the idea of that kind and for that very reason belong to that kind? - Surely. - Hence all men are men because they partake in one idea: Man itself? - Exactly. - Man itself must be a man. - Evidently. - So Man itself and all individual men are all men and therefore of one kind. Then the reason for this must be that they all partake in some superior idea: MAN. - That must be the case. - But by the same reasoning we then need to assume an idea MAN! that is again superior and in which MAN, Man, and all individual men partake, and then an idea MAN!!, and so on. - That seems unavoidable. But I'm worried by this crowd. In this dialogue use is made of a misleading form of expression ("Man itself" is treated as a singular term), but the setting differs from the other fragments, The questioner in the present dialogue is not a sophist trying to trip up an innocent answerer, but a philosopher (perhaps Aristotle himself) who presents an adherent of idealism with a predicament based on his own principles. It is an argument ex concessis, a nonfallacious type of the argumentum ad hominem. If a fallacy was committed in this case, it must have been committed by the answerer through granting the basic tenets of idealism. Perhaps it would be better to reconstruct the case as one in which the fallacy of figure of speech is not committed within the dialogue itself, but is somehow wrapped up in the initial concession (granting idealism). The technique used by the questioner, however, is based on a misinterpretation of a misleading form of expression ("Man itself"), and in this respect there is no difference with the earlier fallacious cases. The only difference is that, because of the initial concession, this misinterpretation was now sanctioned to start with. The Third Man argument shows us how the fallacy of figure of speech is intimately connected with a fundamental philosophical discussion. The connection with philosophical problems and points of view is even more obvious when twentieth century analytical philosophy is brought into the picture. Bertrand Russell's Theory of Descriptions provides us with a prime example. According to Russell (1905) a definite description such as "the present queen of The Netherlands" does not qualify as a genuine singular term (or proper name). The form of any sentence in which such a description occurs superficially, but misleadingly, resembles the form of sentences in which the description has been replaced by a genuine singular term. Suppose, for the sake of simplicity, that "Beatrix" qualifies as a genuine singular term, then, according to Russell, their would be only a superficial and misleading similarity in form between the sentences "The present queen of The Netherlands is an eloquent speaker" and "Beatrix is an eloquent speaker". The last sentence expresses a singular proposition (a proposition to the effect that a certain predicate applies to a certain individual), whereas the first has a hidden complexity and is actually equivalent to a conjunction of three sentences: there is a present queen of The Netherlands and there are no two different present queens of The Netherlands and every present queen of The Netherlands is an eloquent speaker. Being misled through a superficial similarity of form between sentences that are logically completely different, one may be tempted to assume that for each definite description there exists (or subsists, or whatever) some entity to which the description refers. Thus one may be inclined to envisage such entities as the present king of France, the golden mountain, and even the round square. Alexius Meinong (1904) cheerfully accepted this point of view and elaborated its consequences in a Theory of Objects. Russell's theory, however, provides an alternative view that permits one to jettison the "Meinongian underworld". Upon comparing Russell's well-known theory wih Aristotle's reflections on figure of speech, we see a straightforward parallel between Russell's treatment of definite descriptions and Aristotle's analysis of Cases 3, 4 and 8 where the expressions "ten dice", "only one die", and "Man itself" were mistaken for genuine singular terms. To these cases we may add Case 7, were "nothing" is misinterpreted in a similar way. In sum, adherents of Russell's theory may decry the assumption of a Meinongian underworld of objects as based on nothing more than a fallacy: the fallacy of figure of speech! The parallel reaches even further. The notion of being misled by a form of expression presupposes that, besides the logical misconceptions to which one may fall victim, there is also a correct conception of the logic of an expression. Charging one with the fallacy of figure of speech therefore presupposes some logical theory to provide a criterion by which correct and incorrect interpretations of a sentence may be distinguished. In Aristotle's analysis some version of the theory of categories fills the bill, whereas for Russell this theory is the Theory of Descriptions. Without, at least some rudimentary, logical theory a charge of having committed a fallacy of figure of speech makes no sense. It is a major theme of analytical philosophy in this century that the form of expression of a sentence need not coincide with its true logical form. Ludwig Wittgenstein (1922: 4.002) summarizes the point in the slogan "Language disguises the thought." Shortly after he refers to Russell: "It is the merit of Russell's to have shown that the apparent logical form of the proposition need not be its real form." The same theme features in Gilbert Ryle's seminal paper "Systematically Misleading Expressions" (1931). Ryle discusses several species of misleading expressions: (i) Quasi-ontological Statements, such as "God exists" (15) 4 and "Mr. Pickwick is a fiction" (17), where neither "exist" nor "being a fiction" figures as a genuine predicate; (ii) Quasi-Platonic Statements, such as "Unpunctuality is reprehensible" and "Virtue is its own reward" (20), where "unpunctuality" and "virtue" constitute examples of spurious singular terms; (iii) Quasi-referential 'The'-phrases, such as "Jones hates the thought of going to the hospital" (29), "The victory of the Labour Party would have surprised me" (31, 32), and "The whale is not a fish but a mammal" (32), where we do not need to assume entities for the descriptions to denote. All these cases can be thought of as occasions for a fallacy of figure of speech. Ryle does not mention the fallacy of figure of speech, but is aware of a connection between his theme and fallacies in general: "Paralogisms and antinomies are the evidence that an expression is systematically misleading." (35) Anthony Flew, in his introduction to the volume of Logic and Language that contains a reprint of Ryle's paper (1968), stresses the importance of this theme for (modern British) philosophy: ...it is from this and upon this central and fundamental discovery that all the other characteristic doctrines and assumptions of modern British philosophy have been developed and founded. It has been realized that expressions may be grammatically similar and yet logically different. (7) As a corollary one may add that further study of the fallacy of figure of speech is likely to help us understand an important and as far as I see unfinished chapter of the analytical tradition. 4. A Fallacy of Ambiguity or a Non Sequitur? Aristotle classifies figure of speech as a fallacy dependent upon language, (para tên lexin, see De Soph. El. 4: 165b 23-24). Later he divides the six types of fallacy dependent upon language into two groups: For of the fallacies that consist in language, some depend upon a double meaning, e. g. ambiguity of words and of phrases, and the fallacy of like verbal forms (for we habitually speak of everything as though it were a particular substance)-while fallacies of combination and division and accent arise because the phrase in question or the term as altered is not the same as was intended. (De Soph. El. 6:168a 24-28) The first group consists of those para to ditton, i.e., bound up with something double). Aristotle explicitly mentions equivocation (ambiguity of words, lexical ambiguity), and refers in all likelihood to amphiboly (ambiguity of phrases, structural ambiguity); thirdly, he clearly assigns figure of speech to this category, referring to it by the rare but rather perspicuous term (homoioschêmosynê, i.e., similarity of form. This term was translated by Pickard-Cambridge, in the passage just quoted, as "fallacy of like verbal forms" and by Forster (Aristotle 1965) as "similarity of formation". The second group comprises composition and division and the fallacy of accent. These fallacies are not bound up with double meanings or with any kind of doubleness. In all these cases some word or expression is taken for some other word or expression that may resemble it, yet remains clearly distinguishable from it, at least in spoken language.5 On the evidence of the passage quoted, then, it seems that Aristotle subsumes the fallacy of figure of speech among those depending on ambiguity. Yet doubleness with figure of speech must be distinguished from doubleness with equivocation and amphiboly. In the latter two cases the fallacy depends on one and the same expression's admitting of two different, but both prima facie acceptable, interpretations. For an analysis of a case of such a fallacy both interpretations are needed. For instance, a certain premise may only be acceptable if the expression is understood in one sense, whereas the conclusion could follow only if it were taken in the other sense. But cases of figure of speech do not, in general, admit of two bona fide interpretations. If one wants to describe these fallacies in terms of two interpretations - something Aristotle seems to imply by calling them "bound up with doubleness"-one of them must be the correct interpretation, whereas the other is incorrect and due to some misconception of the logical form of a statement, which misconception again is induced by a superficial resemblance with staments of a different logical form. Thus we reach a quite different conclusion: even though figure of speech may, together with equivocation and amphiboly, be classified as a fallacy of doubleness, it should not, like the other two, be analyzed as a fallacy of ambiguity, but rather as a fallacy of faulty logic. In terms of pragma-dialectical rules (Van Eemeren and Grootendorst 1992), figure of speech violates not Rule 10 (prescribing an adequate use of language, p. 196) but Rule 8 (prescribing validity, p. 169). But then we are confronted with another problem. If figure of speech is not a fallacy of ambiguity, but of invalidity, so that matters of validity are seen to be bound up with language, why did Aristotle classify other types of fallacy that also turn on a failure of validity as fallacies independent of language? Why did he split up the non sequitur fallacies into groups assigned to different main divisions? When figure of speech is language dependent one would expect the same to hold for accident and for consequent (which according to De Soph. El. 6: 168b 27-28 constitutes a special subtype of accident) and for secundum quid.6 Are not all matters of validity sensitive to language? A possible answer can be found starting from Hamblin's understanding of the motive behind the classification of fallacies as either language dependent or language independent: What does distinguish the refutations dependent on language is that they all arise from the fact that language is an imperfect instrument for the expression of our thoughts: the others could, in theory, arise even in a perfect language. (1970: 81) The use of an Aristotelian perfect language would exclude the occurrence of a fallacy of figure of speech. Such a language would be free of misleading declensions and conjugations; nor would there be any doubt about the category to which each phrase is to be referred, i.e., whether it denotes an individual, or a quality, or a quantity, or whatever. In other words, the forms of expression in an Aristotelian perfect language would exactly match the theory of categories. But even under these circumstances one could not exclude the fallacy of consequent and other non sequitur fallacies, these could occur on a cognitive, instead of a linguistic, level. Compare this with a formal language for predicate logic, use of which would outrule a confusion between general and singular terms, but which by itself does not shield the user from errors in deduction. From this point of view, the border line between language dependent and language independent fallacies will itself depend upon the linguistic-logical theory of which the fallacy theorist disposes. For Aristotle this was mainly the theory of categories and for that reason he could not envisage an ideal language that would outrule, for instance, fallacies of accident. A stronger linguistic-logical theory, however, might bring these fallacies within the class of fallacies dependent upon language, indeed as other cases of figure of speech, since given that stronger theory a logically perfect language would exclude them. The classification of figure of speech as a non sequitur will largely determine the profile of dialogue for cases in which one disputant has charged the other with this fallacy. Elsewhere I sketched a profile of dialogue for non sequitur (Krabbe, 1995) and one would expect this profile to give us a survey of the most notable options for disputants involved in a verbal exchange on account of an (alleged) case of figure of speech. One of the options of attack consists of the presentation of a counterexample against the fallacious reasoning, i. e., a possible world in which the premises would be true whereas the conclusion would be false. Presumably, this option would not work very well in cases of figure of speech, since the confusion about the correct interpretation