Microsoft Word - 2. Prus Aberdein PROOF.JABedits.doc © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. Is Every Definition Persuasive? Douglas Walton on Persuasiveness of Definition JAKUB PRUŚ Jesuit University Ignatianum Krakow Poland jakub.prus@ignatianum.edu.pl ANDREW ABERDEIN Florida Institute of Technology Melbourne Florida, USA aberdein@fit.edu Abstract: “Is every definition persua- sive?” If essentialist views on defini- tion are rejected and a pragmatic account adopted, where defining is a speech act which fixes the meaning of a term, then a problem arises: if meanings are not fixed by the essence of being itself, is not every definition persuasive? To address the problem, we refer to Douglas Walton’s impres- sive intellectual heritage—specifically on the argumentative potential of definition. In finding some non- persuasive definitions, we show not every definition is persuasive. The persuasiveness lies not in syntactic or semantic properties, but the context. We present this pragmatic account and provide rules for analysing and evaluating persuasive definition—a promising direction for further re- search. Résumé: « Chaque définition est-elle convaincante ? » Si les perspectives essentialistes sur la définition sont rejetées et une explication pragma- tique est adoptée, où la définition est un acte de langage qui fixe le sens d'un terme, alors un problème se pose : si les sens ne sont pas fixés par l'essence de l'être même, toute défini- tion n'est-elle pas une définition persuasive? Pour résoudre le prob- lème, nous nous référons à l'héritage intellectuel impressionnant de Doug- las Walton, en particulier sur la possibilité argumentative de la défini- tion. Nous examinons des définitions non persuasives, nous montrons que les définitions ne sont pas toutes persuasives. La force de persuasion ne réside pas dans les propriétés syn- taxiques ou sémantiques, mais dans le contexte. Nous présentons cette explication pragmatique et fournis- sons des règles pour analyser et évaluer la définition persuasive - une direction prometteuse pour de plus amples recherches. Keywords: definition, essentialism, persuasive definition, pragmatism, seman- tic arguments, Charles Stevenson, Douglas Walton 26 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. 1. Introduction Twenty years ago, Douglas Walton posed a key question about persuasive definition: A problem for the new dialectic is how to define persuasive defi- nitions (the self-reference in the problem notwithstanding). Are all definitions persuasive? Or are lexical definitions different from persuasive definitions? According to the postmodernist view, all lexical definitions have an argumentative “spin” or bias to favor special interests or viewpoints. A central problem then is how to define the very expression “persuasive definition” itself so that some new dialectical classification system of the various kinds of definition can be given. So far, this problem is unsolved (Walton 2001, p. 127). This paper is an attempt to solve this problem. We will start with a brief introduction of the concept of persuasive definition (PD) and an overview of Walton’s views on PD. The works of Douglas Walton dedicated to definition and its role in argumentation are of great significance to the problem discussed here and will be re- ferred to throughout the paper. 2. A brief history of the concept of persuasive definition The idea of persuasive definition was introduced for the first time by Charles Stevenson in (1938) and developed in (1944). To be specific, Stevenson introduced two concepts: PD and persuasive quasi-definitions (PQD).1 PD changes the sense and/or reference of a term (“descriptive meaning” in Stevenson’s wording (1944, p. 54)) without changing its tone (“emotive meaning”). PQD changes tone only and leaves the sense and reference unchanged (1938, pp. 333–334; 1944, pp. 280–281). Stevenson’s theory of meaning has some difficulties, which won’t be analysed here,2 so it is better to 1 Also called “quasi-PD,” which may be misleading, for it is not a definition par excellence (Pruś 2021). 2 See more on Stevenson’s theory of meaning in (1944, p. 59–82). In simple words, it reduces the proposition “I like ice cream” to the expression of a Is Every Definition Persuasive? 27 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. analyse matters in Fregean terms (sense-reference-tone).3 There- fore an example of PD would be defining “racism” as “treating people unequally on the basis of their ethnicity” (very broad defi- nition) and consequently classifying someone as a “racist” because they vote for an electoral law privileging ethnic minorities—which is treating people unequally on the basis of their ethnicity. Such a definition would be persuasive for it widens the boundaries of correct application of a word with very strong negative connota- tions4 (‘racism’) so that it stigmatizes conduct that is ordinarily thought to be acceptable. An example of PQD would be defining “cowardice” as “a reasonable act of protecting oneself or one’s family or goods against any harm” in order to call a cowardly behaviour (negative connotation) a reasonable act (positive conno- tation). These two manoeuvres are often mixed and misleadingly named “persuasive definitions.” One of the most important treatments of PD since Stevenson is that of Tadeusz Pawłowski (a continuator of the Lvov–Warsaw School), who contributed to the concept of PD significantly. He defined PD as follows: “persuasive definitions are aimed to change the emotional value of the definiendum—this value is aimed to be changed with the emotional value of the definiens. This change may be made along with the change of the extension of the defini- tion” (Pawłowski 1978, p. 232). Pawłowski distinguished three types of persuasive definitions (1980, pp. 59–65): (a) PDs aimed at changing the extension of the definiendum (“true temperance is a bottle of claret with each meal and three double whiskies after each dinner”); (b) PDs intended to change the emotive charge of the definiendum, (“Whatever the artist spits is art”); (c) PDs in- tended to replace the definiendum used hitherto by another term, charged with emotive associations of the kind desired by the au- positive attitude towards ice creams. For more comments on Stevenson’s theory see: (Macagno and Walton 2008b, p. 207; 2008a, pp. 530–531). 3 For a detailed analysis of Stevenson’s concept in Fregean terms see (Aberdein 2006, p. 157). 4 By “connotation” we do not mean Mill’s understanding of the term (System of Logic VII: 31), but rather the common usage, in which it is loosely synonymous with Frege’s ‘sense,’ Port Royal’s ‘intension’ or Stevenson’s ‘descriptive’ meaning. 28 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. thor of the definition (later called “synonymous definitions” (“‘Pa- riahs’ are people who completely depend on God—they are ‘peo- ple of God’ [Harijans]”).5 Stevenson’s concept and Pawłowski’s results were further developed by many researchers: Halldén (1960), Schiappa (1996; 2003), Burgess-Jackson (1995), Aberdein (1998; 2006), and Zarefsky (2006), to name but a few. Although Stevenson did not state clearly that PDs are “less re- spectable” than other types of definitions—even if they can be deceiving (1938, p. 331)—he rejected a “general disparagement on persuasive definitions” (1944, p. 215). This attitude, however, has not always been shared by more recent philosophers and logicians. For Richard Robinson, persuasive definition is “at best a mistake and at worst a lie, because it consists in getting someone to alter his valuations under the false impression that he is not altering his valuations but correcting his knowledge of the facts” (Robinson 1950, p. 170). And Patrick Hurley says that PD “masquerades as an honest assignment of meaning to a term” (2000, p. 96). Douglas Walton, however, often stood up for persuasive definition, and showed that it might be a rational and acceptable form of persua- sion. 3. Douglas Walton’s understanding of persuasive definition Walton sought to undermine the belief that PDs are fallacious or deceptive6 by—nomen omen—definition (2001, pp. 128–129). To support this, he analyses Stevenson’s example: if one broadens the extension of “cultured” in order to include an obviously uncultured 5 It is worth noting that we do not consider the second and the third types of PD in Pawłowski’s understanding as PD (in fact, the former is PQD)—for detailed comparison on PD and its various understanding see Table 1 in Section 5. 6 “Many might be inclined to think that persuasive definitions only represent clever tricks used by sophists, that they are of little practical interest, and that they would be easy to deal with if an arguer is confronted by them. Even a brief consideration of some actual examples quickly dispels these illusions” (Walton 2005, p. 164)—then he moves to case studies of “truly needy” (the term intro- duced by R. Reagan to cut social benefits) or the definition of “wetlands” (definition introduced by G. H. W. Bush to exclude some wetlands from the extension of the term “wetland” and allow investors to develop such land). Is Every Definition Persuasive? 29 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. man, and this intention is transparent, it is not deceptive at all. It is rather an argument, which might be evaluated by searching for possible reasons for or against such an understanding of “cultured” or by showing that such an “uncultured man” cannot be classified by the same term as that gentleman or that lady. Walton argues that there is always a “theory” behind the redefinition, which supports it (see, for example, a definition of abortion or rape, which is implied by a certain anthropological theory7), and there is nothing necessarily deceptive here. That’s why he says that “per- suasive definition should be treated as a particular kind of argu- ment” (2001, p. 118). Before we move on to discuss the evaluation of PDs, let us give an overview of how Walton understood such definitions. One of the formulations of the definition of PD given by Wal- ton is that “a definition will be shown to be persuasive when it leads the interlocutor to accept a proposition or action, analysed in terms of commitment, both action and propositional commitment” (Macagno and Walton 2008b, p. 204).8 This intuition seems very promising, since it involves the notion of “commitment,” which includes not only being emotionally attached to a certain defini- tion, but also being committed to a definition due to some benefits coming from it or its coherence with other beliefs which one holds. Thus, for example, the IAU definition of planet formulated in 2006, which excluded Pluto from the set of planets, was persua- sive.9 Therefore, in this formulation PD includes such definitional moves, which do not only influence—what Stevenson called— “emotive meaning,” but also “descriptive meaning.”10 On the other 7 (Schiappa 2003, pp. 61–68; 89–103, Aberdein 1998, Burgess-Jackson 1995). 8 For further analysis of action and propositional commitment see (Walton and Krabbe 1995, pp. 15–28). 9 See Walton’s analysis of the redefinition of “planet” and exclusion of Pluto (2008). The issue is still controversial today: Jim Bridenstine, until recently the Administrator of NASA, believes that Pluto is a planet. (https://www.youtube.com/watch?v=h9zM1vS1hVQ). 10 “Re-definitions can be applied to non-ethical terms with a persuasive effect. For instance, the re-definition of the legal–medical term ‘insanity’ can affect the evaluation of the classification, leading to an attitude of indulgence, for in- stance” (Macagno and Walton 2008a, p. 527). 30 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. hand, he admitted that PDs often have some evaluative aspect, e.g. they “try to change our views about what is right or wrong” (2005, p. 168). Walton and Macagno analysed PD further, from the argumenta- tive perspective, and showed that PD consists of two argumenta- tion schemes: argument from values and argument from classifica- tion. This means that definitions used in discussion could be ana- lysed at the level of classification, that is, how one describes or classifies a given object. However, at the deeper level, it can also be analysed in terms of values, which urge the person to classify the object in a way that make it desirable (or not). The following quotation explains its argumentative mechanism: The reasoning process a definition follows to achieve this result will be described by means of the arguments from classification and from values. Persuasive and quasi-definitions involve an ap- praisal of the aspect of reality they refer to, and a modification of the classificatory criteria. This classification and appraisal is the ground for a decision to act. We represent the connection between an evaluation of a fragment of reality and the decision to act ac- cordingly as argument from values. Often persuasive definitions involve a conflict of values, in which the interlocutor founds his implicit argumentation upon a value that the interlocutor does not share. However, sometimes this conflict of values depends on the interlocutors’ arguing about two different realities, two different concepts named in the same fashion (Macagno and Walton 2008b, pp. 204–205). This gives us a clear understanding of when definition is persua- sive—when it leads the opponent to accept a given proposition or action in terms of commitment. Additionally, PD must, as quoted, assume certain values or a certain classification—which may be analysed with argumentation schemes. To summarise Walton’s understanding of PD we may distin- guish several features: (i) PD should be treated as a particular kind of argument; (ii) PD does not influence connotation alone; (iii) PDs often include ethical terms or emotionally laden words; (iv) PDs are often supported by argument from values or argument Is Every Definition Persuasive? 31 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. from classification. However, the dominant characteristic of PD is best expressed in the following passage: Persuasive definitions consist in a redefinition of a term, which is used to support a conclusion. It is, in other words, a redefinition of a term used argumentatively. While quasi-definitions concern an argumentation from values, persuasive definitions are aimed at trading on argumentation from values that is developed from the use of a redefined term. A concept, endoxically associated to an evaluation, is modified in order for a particular fragment of reality to be classified as falling within it (Macagno and Walton 2008a, pp. 545–546). A similar idea is expressed by Macagno and Walton in regard to words in general—words are persuasive due to their usage in argumentation: A word is persuasive … because its pragmatic role can only be explained by means of the argumentative process stemming from it and presupposed by the role the word plays in the communica- tion (Macagno and Walton 2008a, p. 531). This should be enough to characterise PD on Walton’s view. On the one hand he seems to accept Stevenson’s narrow understand- ing of PD (a definition manipulating the “emotive meaning” of the term), but on the other he presents a deep analysis of persuasive (re)definitions which do not include emotionally-laden words. We may also ask how other logicians have understood PD since Stevenson. Let us recall just a few: A. Schiappa makes the following comment on Stevenson, refer- ring to Perelman and Olbrechts–Tyteca: “Stevenson’s position can be reformulated as suggesting that advocates of new defi- nition of ‘X’ seek to alter which stimuli we denote as ‘X’ while preserving past patterns of action or behaviour taken toward stimuli called X. In this sense virtually all proposed new real definitions are persuasive definitions, since ‘all those who ar- gue in favour of some definition want it … to influence the use which would have been made of the concept had they not in- 32 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. tervened’ (Perelman and Olbrechts-Tyteca 1969, p. 213)” (Schiappa 1993, p. 414). B. Aberdein, in Fregean terms, indicates that PD are definitions that do not change the term itself or its tone, but change its sense, or its reference or both. If the definition changes the tone and leaves the sense, or the reference or both unchanged, then he calls it PQD. If it changes all three factors (tone, sense, reference) then it is “degenerate definition” (Aberdein 2006, p. 157). C. Definition is persuasive “if its acceptance offers a solution in the discussion and there is alternative definition of a given term” (Pruś 2019, 2021). These three understandings of persuasive definition are coherent— to a certain degree—with Walton’s characteristics. What is crucial here is that PDs are not necessarily deceptive or emotional, but they are aimed at acceptance of a given proposition (which might be the axis of dispute). Based on these analyses, we conclude that PDs cannot be re- duced to definitions which change the emotive meaning of a term and leave its descriptive meaning unchanged. They might be, as indicated by cases of definitions of “truly needy”, “rape”, “wet- land” and many others, persuasive not due to their changing emo- tive meaning (in Stevenson’s terms) or tone (in Frege’s terms), but rather due to their change to the extension of the definition (which could be labelled as “precising”—in Hurley’s classification—or “regulative”—in the Lvov–Warsaw School’s classification).11 Therefore, we may say we have clarified some of the doubts— there are such speech acts as definitions, and if and only if they try to change the denotation or sense of the term to support persua- sion, they are persuasive. Not all definitions do so,12 therefore, we 11 Both classifications will be discussed further below. 12 For example: the Einsteinian definition of energy, E=mc2 is not persuasive for it is not used to support persuasion (although one may imagine some particular circumstances, in which it could be used for this purpose—the notion of ‘ideal dictionary definition’ will be discussed further below). On the other hand, we may see how much scientists and philosophers discuss on the definition of Is Every Definition Persuasive? 33 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. have answered the title question: not all definitions are persuasive. But this might not be as easy as it looks. The quotation from Schiappa indicates an intuition crucial to this problem, which we may formulate as the question: “does definition X say what X is or rather how we should understand X?” The first alternative leads us to “real” definitions, whereas the latter leads to “nominal” defini- tions. Therefore, PD should be characterized more broadly, as in (Macagno and Walton 2008b, p. 204), but with the additional condition presented in (Pruś 2021)—we will further call definition persuasive if “it is put forward to support one’s claim in the dis- cussion and there is alternative definition of a given term”. Now, we may ask when such definitions should be accepted and what could be the criteria for evaluating definitions. Rather than launch an investigation into the debate between two meta- physical traditions, essentialism and pragmatism, let us stick to the argumentative perspective in analysing definition.13 As Walton warns us: The rules for putting forward, challenging and accepting defini- tions clearly vary with the type of conversational exchange an ar- gument is supposed to be part of. So there are no easy or pat solu- tions to the problems of when definitions should be judged to be proper and when not (2001, p. 130). Additionally, Walton notes that “real” definitions are very power- ful tools in argumentation, due to the dissociation mechanism:14 The problem with essentialism is that by dissociation, it forces a dichotomy between nominal and real that is too sharp, like the di- chotomy between good and evil. The outcome is that the old form human, especially in the light of human evolution—this is a case, in which scientific definitions of Homo sapiens, based on reconstructions are in fact persuasive (see Lenartowicz and Koszteyn 2000). 13 For more on the debate on definition, see: (Walton 2005, pp. 173–184, 2008, pp. 134–135, Rosen 2015). 14 See more on the dissociation technique (Perelman and Olbrechts-Tyteca 1969, pp. 411–450) as a kind of semantic argument using dissociative PD (Pruś 2021) or as strategic manoeuvring (Feng, Zhao, and Feng 2021). 34 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. of essentialism supports a bad theory of philosophical argument that allows a philosophical arguer to preemptively and dogmati- cally dismiss all opposed views as ‘unreal’ (Walton 2005, p. 175). When there are two competitive definitions, e.g., definitions of death,15 and one is called “real,” for it grasps the very essence of death, then the other cannot be “real” or “true” at the same time— or so the arguer implies. Therefore, the existence of real defini- tions, and references to this concept in argument, makes all other definitions purely “nominal,” that is, “unreal.” There is of course much more to say about the pragmatism vs. essentialism debate—a detailed critique of essentialism from a pragmatic point of view can be found in (Schiappa 2003, pp. 167– 182). This can be summarized in two points: (i) we have no access (and probably never will) to the essence of things (things in them- selves), therefore objective discourse and evaluation of real defini- tion is beyond our reach; (ii) as Schiappa writes definitions are linguistic, and there is no way to escape the histori- cal contingency of any particular definitional proposition. The ‘re- alness’ of any proposed definition is theory bound; that is, the be- lief that a particular definition captures the ‘real’ nature of any given X is inextricably linked to a number of related beliefs that are held in a particular historical context and subject to possible revision (2003, p. 168). Of course, these arguments may also be subject to doubt (many metaphysicians would disagree with both statements). The analy- sis of this debate would be enough for a book, but we actually do not need it here—the following quotation shows that Walton, although he is not in favour of essentialism and he accepts Schiappa’s critique—allowed for a sort of “pragmatic essential- ism,” that is, he admits that there are central and contingent attrib- utes, and classical definitions are sometimes needed, useful, and 15 See Walton’s two books dedicated to the definition of death, in which he argues for a definition of death in terms of irreversible destruction of function— brain death—against a romantic idea of death, that is, when the heart stops beating (1979, 1980). Is Every Definition Persuasive? 35 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. possible to build. But Walton changes the emphasis (in James’s spirit)—it is not reality that makes these properties central, but we who decide that such and such a property is essential for this no- tion depending on the context (our identity, our needs and so on):16 I agree also that definitions are always inherently persuasive in na- ture, in ways that have not been fully recognized. On the other hand, I continue to think that there is some place for something like the notion of an essential attribute in a definition. When defin- ing a term by genus and difference (a good method in at least some instances, I would say), one needs to differentiate between the more important or central attributes and those that are less im- portant or less central to the meaning of the term. Some notion of essentiality is useful, I think. But maybe instead of talking about “essential” attributes, we should talk about attributes that are cen- tral or important for the purpose supposedly served by the defini- tion (Walton 2005, p. 174). To summarise, we may see that Walton stands for the pragmatic side in the essentialism–pragmatism debate. He accepts the cri- tique of essentialism but admits that it is useful and needed in certain contexts to hold that some properties are central for the given object, where others are contingent.17 It is, however, we who decide what is and what is not central for an object to be X, depending on our needs, possibilities and the consequences of a given definition. Having said this, we may conclude that Walton’s approach is pragmatic—he sees definition as a speech act, in which one puts forward a most central (to the definer) property or properties of the object X, which something must share to be X. 16 See William James’s example on the essential property of oil—it is different for a chemist, cook, firefighter and furniture manufacturer. Thus “central” property is always relativized to the context (James 1981, p. 962). 17 He also resiles from the postmodernist view: “The new dialectical view recognizes the argumentative function of many definitions, unlike the essential- ist view, which sees meaning as fixed and objective. But the new dialectical view does not draw the postmodernist conclusion that all definitions, even highly loaded, persuasive, or coercive ones used to promote special interests, are equally justifiable” (2001, p. 127). 36 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. Therefore, a definition should be evaluated in the light of its ac- cordance with usage and its consequences compared to the needs of interlocutors. In other words, definition—which is always sup- ported by some theory—should be seen as an argument: From a logical point of view, the most important thing about properly handling cases like those above is that a persuasive defi- nition should be treated as an argument. It should be regarded as open to critical questioning and to the posing of counter- definitions. It should be regarded as having a burden of proof at- tached. It should be recognized that the audience or respondent of the redefinition should have the right to argue for retaining the ex- isting usage, if it seems to them to better represent their views on the matter (Walton 2001, p. 131). This leads us back to the problem with which we began. 4. The problem If we understand a definition in a pragmatic way, as Walton did, should we not claim that all definitions are persuasive, at least potentially? The quotation from Walton at the beginning of this paper raises the doubt, formulated in a “postmodern” spirit—if every definition comes from certain theories and supports other theories, then every definition is persuasive, for it is aimed at supporting its author’s beliefs or needs: According to the postmodernist view, all lexical definitions have an argumentative “spin” or bias to favor special interests or view- points. A central problem then is how to define the very expres- sion “persuasive definition” itself so that some new dialectical classification system of the various kinds of definition can be giv- en. So far, this problem is unsolved (Walton 2001, p. 127). In other words, if we accept that there is no “real” definition and that each definition should be evaluated by the discussants as to whether they accept it or not, in light of the consequences which follow from the definition, shouldn’t we say that every definition used in discussion is persuasive? For every definition implies some solutions or consequences and for many definitions there are Is Every Definition Persuasive? 37 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. alternative definitions, isn’t almost every definition persuasive? Therefore, we need to face the title question. Let us consider what the alternatives are. We may say: (i). Every definition is persuasive; or (ii). Not every definition is persuasive. To (i) we may formulate two comments. Firstly, does it mean that every definition is persuasive potentially or actually? Well, if we assume that it is the purpose of a definition that decides whether it is persuasive definition, reporting definition, precising definition, and so on, then every definition is potentially persuasive as well as potentially precising or reporting.18 Therefore, it is clear that we mean here being actually persuasive. The second objection to (i) is the following: why should we distinguish such a category of defi- nition as “persuasive”? If every definition is actually persuasive there is no sense in doing that, for they vary only by the degree of persuasiveness. This seems to be a strong argument against (i), if we do not want to abandon the concept of persuasive definition, which has been found useful in argumentation theory. Therefore, let us consider (ii). To support (ii) we need to indicate what makes a definition per- suasive (use of definition to support one’s claim in discussion) and to show whether it is consistent with the pragmatic understanding of meaning/definition. Therefore, we need to show that, although there can be persuasiveness in various definitions, even scientific or legal definitions, some distinctive properties of PD (some sig- nificant degree of persuasiveness) separate PDs from other defini- tions. We might find some support for (ii) in Walton’s work— although he rejects essentialism, he also does not share the “post- modern” approach: The postmodernist views are that definitions can change, that they are more persuasive in nature than tradition indicates, and that 18 We are grateful to Fabrizio Macagno for this insight. 38 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. they are bound to disciplines and other groups that have interests at stake. All these views seem to be justified, up to a point, by case studies of how the argumentation actually works when terms of public significance are redefined. But what seems to follow, that the term in question can freely be redefined by another inter- est group, simply to meet the rhetorical needs of the moment, seems wrong. The view that definitions can rightly be used for dif- ferent purposes in different contexts seems right. But the implica- tion that therefore a scientific definition is subject to redefinition in the same way as a political definition is, by pressure of compet- ing interest groups and majorities, does not seem to be right (2001, p. 124). Therefore we can conclude that Walton’s answer would be “no.” Not every definition is persuasive. How does he justify this claim? Walton distinguishes the so-called “postmodernist” understanding of definition, which for the purpose of this article we may formu- late in Foucauldian terms “every definition is a form of oppression and power, which is used to govern others, therefore every defini- tion is persuasive.” Although Walton rejects the essentialist under- standing of definition, he does not share the postmodernist idea of persuasiveness of every definition, that is, he does not agree to call all definition actually persuasive. What solution does he offer? He draws a line between these two alternatives, and shows that his approach, the new dialectics, is somewhere in between. This view admits that there is potential for deception in PDs, but, as already mentioned, this does not necessarily mean that they are fallacious. Walton says: According to the new dialectic, persuasion dialogue is a legitimate type of framework of argumentation, and there is nothing inher- ently fallacious or deceptive about persuasion in itself (2001, p. 128). So what does the new dialectic give us really? It requires us to recognise what types of dialogue we participate in. Is it political discourse during a campaign? Or rather is it a legal contract? Or is it a business argument on how to plan the budget for the next year? Preaching? Advertising? The new dialectic says that we should Is Every Definition Persuasive? 39 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. relativize criteria to the type of dialogue and other circumstances making up the context of the discussion. Walton says: It is not possible to make up any one set of rules to govern all such cases, because, according to the new dialectic, the rules will vary depending on the type of dialogue in which the parties in a case are supposedly engaging (2001, pp. 131–132). As already mentioned, there are types of dialogue in which defini- tions are not persuasive (but in certain contexts could be), and there are others where we do not see them as fallacious or decep- tive (for example in politics or advertisements). Walton also sug- gests that PDs should be treated with caution and gives some intuitions for evaluating them: But the potential for deception is there, as soon as the redefinition of the old term departs from the existing usage of the audience, or respondent of the argument. So from a logical point of view, we should be wary of persuasive definitions, and treat them with cau- tion (2001, p. 131). It’s worth distinguishing deceptiveness from fallaciousness here— a definitional move can be fallacious, in the sense of embodying some sort of logical mistake, without being the product of an intention to deceive and without having as a result the deception of anybody.19 Now we may ask what would be the criteria for evalu- ating PDs? Starting from case studies of PD of “culture,” “truly needy,” “wetland” and “rape,” Walton gives the general intuition, which was already quoted at the end of the previous section—a definition used in discussion should always be treated as an argu- ment (2001, p. 131). 5. Evaluation of Persuasive Definition Let us formulate a few rules when approaching definition in dis- cussion to help assess the persuasiveness of definition. 19 We are grateful to an anonymous reviewer for drawing our attention to this. 40 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. RULE 1. Definition used in discussion should be treated as an argument. Walton also states that the burden of proof is always attached to the definer, his respondent can always formulate critical questions to it, and the definer is obliged to face them. The respondent can always refer to the existing usage/accepted meaning and argue for it. The definer is obliged to justify why he introduced the new definition. RULE 2. The burden of proof is always on the definer. We may also derive from this another criterion—there must be sufficient reason for changing the definition, and the new defini- tion must be better than the previous one. For example, if the previous definition was blurry, the new definition cannot be blurry as well—unless it provides some other positive consequences. RULE 3. There must be a sufficient reason for replacing the existing meaning. We may also formulate a general rule from the requirement of the new dialectic to relativize assessment to context. Thus, if defini- tion occurs in a dialogue which does not permit persuasiveness of definition, then we must reject it, but if it permits PD, then we may not reject it simply because of its persuasiveness. RULE 4. Does the context in which the definition occurs al- low for the use of persuasive definition? Let us now summarize these few thoughts. PD can be distin- guished from the other types of definition due to its role in the discussion in which it was used: it is used to support a claim, which is controversial, rather than merely report the meaning of a given term, or precise it and so on. Once the goal of formulating a definition in a discussion has been identified, it is its role in that discussion, that makes the definition persuasive; that is, if the Is Every Definition Persuasive? 41 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. definition is put forward to support a controversial claim in the discussion (and an alternative definition exists), then the definition is persuasive. It is also important, if there are any alternative defi- nitions, whether one is presented as “the true one,” Since, as was already shown, the definition may serve as a premise in the argu- ment, which may sometimes tip the balance in favour of the argu- ment, every definition used in discussion should be treated as an argument itself; that is, examined in the light of the reason sup- porting it (e.g., the authority of the encyclopedia, etc.).20 The fact that it is the context that decides the persuasiveness of definition implies that we cannot find objective criteria of persuasiveness of definition for many prima facie non-persuasive definitions that may, nonetheless, be used persuasively in some extreme circum- stances or some possible worlds. This, however, does not mean that every definition is actually persuasive, for its persuasiveness is relativized to its use in the sentence/discussion. Therefore, the pragmatic understanding of definition (meaning) can be held without falling down a “postmodern” slippery slope and seeing all definitions as actually persuasive. Let us now compare the concept of PD formulated above with two widely accepted and well-grounded classifications of defini- tion. The first, rooted in the Anglo-American logic textbook tradi- tion, was formulated by Patrick Hurley, and Walton himself refers to it several times (2005, pp. 171–172; 2008, p. 133); the second was developed by Polish logicians, members of the Lvov-Warsaw School, at the beginning of the twentieth century. Hurley (2000, pp. 93–95) distinguishes five types of definition: stipulative definition (“assigning a meaning to a word for the first time”), lexical definition (“reporting the meaning that a word already has in a language e.g. by referring to a dictionary”), precis- ing definition (“reducing the vagueness of a word”), theoretical definition (“assigning a meaning to a word by suggesting a theory 20 The additional trickiness of PD is that because it is definition, some may treat it as needing no further justification. As Zarefsky puts it “to choose a definition is, in effect, to plead a cause, as if one were advancing a claim and offering support for it. But no explicit claim is offered and no support is provided” (2004, p. 618). 42 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. that gives a certain characterization to the entities that the term denotes”), and persuasive definition (“engendering a favourable or unfavourable attitude toward what is denoted by the definien- dum”). On Hurley’s classification, PD is a separate type of defini- tion, referring to purely emotional or manipulative definition (for example with the no true Scotsman structure, which adds an eval- uative adjective to the definiendum and then introduces a dissocia- tion). However, each of the remaining four types (stipulative, lexical, theoretical and precising) can also be persuasive depend- ing on the context. It’s worth noting that PDs are more likely to pretend to be precising definitions in Hurley’s classification, for the change of meaning is often justified as “precising” the unclear or deceptive meaning. Polish logicians divided definitions into two groups: reporting and projecting (Ajdukiewicz 1974, Kotarbiński 1986). The former is aimed at reporting (or describing) the meaning of a term, and the latter’s goal is to fix the meaning of a term. Projecting definitions can be further divided into constructive (fixing meaning for the first time, as in stipulative definitions) and regulative definitions (fixing a meaning of a term with respect to its previous or current meaning, so they are analogous to precising definitions). In the Lvov–Warsaw School’s classification, both reporting and project- ing definitions (regulative as well as constructing) might be per- suasive, but similarly to Hurley’s classification, it is more likely to see PD as regulative (when it pretends to precise misleading or ambiguous meaning) or constructive (when it introduces a new meaning for the first time). This shows that PDs cannot be ade- quately defined by their structure—but only by their context. It would be interesting to analyse the persuasiveness of definition not only on the propositional level (how something is defined) and on the pragmatic level (how definition is used), for not only definition may be persuasive, but also the act of defining itself—however, such analysis exceeds the scope of this paper.21 21 For the distinction between the pragmatic level and the propositional level of definition see: (Macagno and Walton 2014) and (Macagno 2013). It might be beneficial to see which acts of defining—as per Ennis’s distinction between stipulating, advocating, and reporting acts of defining (2020, pp. 260–268)—are or tend to be more persuasive. Is Every Definition Persuasive? 43 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. To support this, we may also refer to the Fregean terms in Ab- erdein’s interpretation. By employing Frege’s threefold distinc- tion, tone/sense/reference, instead of Stevenson’s emo- tive/descriptive meaning, Aberdein distinguishes sixteen possibili- ties of changing the definition (the term also may be changed itself). Aberdein uses (*) to mark the change of one of these com- ponents and (—) to mark keeping it fixed (2006, p. 157): Term Tone Sense Reference 1 — — — — Ideal Dictionary Definition 2 — — — * Persuasive Definition 3 — — * — 4 — — * * 5 — * — — Persuasive Quasi- Definition 6 — * — * 7 — * * — 8 — * * * Degenerate Definition 9 * — — — Dissociation, Euphemism, &c. 10 * — — * 11 * — * — 12 * — * * 13 * * — — 14 * * — * 15 * * * — 16 * * * * Table 1 Aberdein's Options for Change Now, the negative answer to the question “are all definitions per- suasive?” is supported by the first row of Table 1 (degenerate definition could be persuasive in some sense, and rows 9–16 are not strictly definitions22). It seems that there are, so called, “ideal dictionary definitions,” which are not persuasive, e.g., the Ein- 22 However, in Pawłowski’s typology some of them could be PD, e.g., rows 13– 14 would be PD of the third type—synonymous definition (1978). 44 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. steinian definition of energy or a definition of a full English break- fast: “fry up is a breakfast often served in the UK and Ireland, which consists of bacon, fried egg, sausage, mushrooms, baked beans, toast, grilled tomatoes.” It is hard to imagine the circum- stances in which they could be used persuasively—they fix the meaning of the term without changing the term itself nor its sense, reference, and tone (so the term, tone, sense, and reference all stay the same). Unless we are absurdly uncharitable, this seems to be something many dictionaries achieve for most of their definitions. A further question is whether definitions can be legitimate even if they do not meet this standard. Walton’s intuition according to which definitions should be treated as arguments is very useful, for it allows us to get out of a “definitional impasse” in the discussion. Now the question reduces to whether such an argument (called a “semantic argument”23) is a good one—especially when the PD is disguised, as such an argument may be deceptive and hard to evaluate (see for example Shackel’s Motte & Bailey arguments (Shackel, 2005) or arguments based on the No True Scotsman structure (Flew 1975)). This question may be very promising for further investigation—how should we evaluate such PDs or argu- ments based on such PDs? 6. Conclusion We have introduced the concept of definition from the perspective of its persuasiveness—that is its use in discussion to support one’s claim. We have analysed specifically the concept of persuasive definition, introduced by Stevenson and developed by many logi- cians and philosophers. Douglas Walton’s contribution to this topic was of great significance—especially his works on argu- ments based on persuasive definition. Walton has shown, based on his own research, but also referring to that of others, such as Hall- dén, Schiappa, or Zarefsky, that PDs are always supported by some arguments and support other arguments. Therefore, Walton’s intuition in this topic was to treat PDs (and definitions in general) 23 See the typology of semantic arguments that includes both arguments based on PD and classification arguments (Pruś 2021). Is Every Definition Persuasive? 45 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. as an argument, due to its argumentative/persuasive potential. This however raises a question—is every definition persuasive? In this paper we tried to answer this question negatively—we have shown that persuasive definitions may be distinguished from definitions in general; that is to say, there are definitions that are not persuasive. We have shown that this notion is pragmatic in nature, for there are no syntactic or semantic properties that make definition persuasive—it always depends on the context of use of the definition. On the example of two well-known classifications of definition, we showed that every type of definition, regardless of its structure or aim, can be—in certain contexts—persuasive. We have also formulated some basic rules, which may be helpful in future to formulate some criteria for evaluating persuasive definitions. References Aberdein, Andrew. 1998. Persuasive definition. In Argumentation and rhetoric, eds. Christopher W. Tindale, Hans V. Hansen and Athena V. Colman. Newport News, VA: Vale. Aberdein, Andrew. 2006. Raising the tone: Definition, bullshit, and the definition of bullshit. In Bullshit and philosophy: Guaranteed to get perfect results every time, eds. Gary L. Hardcastle and George A. Reisch. Chicago and La Salle, IL: Open Court. Ajdukiewicz, Kazimierz. 1974. Pragmatic logic. Dordrecht: Reidel. Burgess-Jackson, Keith. 1995. Rape and persuasive definition. Canadian Journal of Philosophy 25(3): 415–454. Ennis, Robert H. 2020. Definition. In Studies in critical thinking, ed. J. Anthony Blair, 249–270. Windsor, ON: Windsor Studies in Argu- mentation. Feng, Jieyun, Fan Zhao, and Aiqing Feng. 2021. Strategic manoeuvring by dissociation in corporate crisis communication: The case of the 2017 United Airlines’ passenger dragging-off incident. Argumenta- tion 35(2): 321–338. Flew, Antony. 1975. Thinking about thinking. London: Fontana. Halldén, Sören. 1960. True love, true humour and true religion: A semantic study. Lund: Gleerup. 46 Pruś & Aberdein © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. Hurley, Patrick J. 2000. A concise introduction to logic. 7th ed. Belmont, CA: Wadsworth. James, William. 1981. The principles of psychology. Cambridge, MA: Harvard University Press. Original edition, 1890. Kotarbiński, Tadeusz. 1986. Elementy teorii poznania, logiki formalnej i metodologii nauk. Warsaw: Państwowe Wydawnictwo Naukowe. Lenartowicz, Piotr, and Jolanta Koszteyn. 2000. Fossil Hominids - an Empirical Premise of the Descriptive Definition of homo sapiens. Fo- rum Philosophicum 5: 141–176. Macagno, Fabrizio. 2013. The acts and strategies of defining. In What do we know about the world? Rhetorical and argumentative perspec- tives, eds. G. Kišiček and Ž. Žagar, 63–78. Windsor, ON: Windsor Studies in Argumentation. Macagno, Fabrizio, and Douglas Walton. 2008a. The argumentative structure of persuasive definitions. Ethical Theory and Moral Prac- tice 11(5): 525–549 Macagno, Fabrizio, and Douglas Walton. 2008b. Persuasive definitions: Values, meanings and implicit disagreements. Informal Logic 28(3): 203–228. Macagno, Fabrizio, and Douglas Walton. 2014. Emotive language in argumentation. New York: Cambridge University Press. Mill, John Stuart. 1874. A system of logic, ratiocinative and inductive, being a connected view of the principles of evidence and the methods of scientific investigation. 8th ed. New York, NY: Harper & Brothers. Pawłowski, Tadeusz. 1978. Tworzenie pojęć i definiowanie w naukach humanistycznych. Warsaw: Państwowe Wydawnictwo Naukowe. Pawłowski, Tadeusz. 1980. Concept formation in the humanities and the social sciences. Dordecht: Reidel. Perelman, Chaïm, and Lucie Olbrechts-Tyteca. 1969. The new rhetoric: A treatise on argumentation. Translated by John Wilkinson and Pur- cell Weaver. Notre Dame, IN: University of Notre Dame Press. Pruś, Jakub. 2019. Argumentacja semantyczna – podstawowe pojęcia i problemy definicyjne. Res Rhetorica 6(4): 56–76. Pruś, Jakub. 2021. How Can Modifications of Meaning Influence Argu- mentation? The Concept and Typology of Semantic Arguments. Ar- gumentation 35 (3): 483–508. Robinson, Richard. 1950. Definition. Oxford: Clarendon Press. Rosen, Gideon. 2015. Real definition. Analytic Philosophy 56(3): 189– 209. Schiappa, Edward. 1993. Arguing about definitions. Argumentation 7(4): 403–417. Is Every Definition Persuasive? 47 © Jakub Pruś & Andrew Aberdein. Informal Logic, Vol. 42, No. 1 (2022), pp. 25–47. Schiappa, Edward. 1996. Towards a pragmatic approach to definition: ‘Wetlands’ and the politics of meaning. In Environmental pragma- tism, eds. Andrew Light and Eric Katz, 209–230. London: Routledge. Schiappa, Edward. 2003. Defining reality: Definitions and the politics of meaning. Carbondale, IL: Southern Illinois University Press. Shackel, Nicholas. 2005. The vacuity of postmodernist methodology. Metaphilosophy, 36(3): 295–320. Stevenson, Charles L. 1938. Persuasive definitions. Mind 47(3): 331– 350. Stevenson, Charles L. 1944. Ethics and language. New Haven, CT: Yale University Press. Walton, Douglas. 1979. On defining death: An analytic study of the concept of death in philosophy and medical ethics. Montréal, QC: McGill-Queens University Press. Walton, Douglas. 1980. Brain death: Ethical considerations. West Lafayette, IN: Purdue University Press. Walton, Douglas. 2001. Persuasive definitions and public policy argu- ments. Argumentation and Advocacy 37(3): 117–132. Walton, Douglas. 2005. Deceptive arguments containing persuasive language and persuasive definitions. Argumentation 19(2): 159–186. Walton, Douglas. 2008. The case of redefining ‘planet’ to exclude Pluto. Informal Logic 28(2): 129–154. Walton, Douglas, and Erik C.W. Krabbe. 1995. Commitment in dia- logue: Basic concepts of interpersonal reasoning. Albany, NY: State University of New York Press. Zarefsky, David. 2006. Strategic maneuvering through persuasive defini- tions: Implications for dialectic and rhetoric. Argumentation 20(4): 399–416.