In current argumentation theory, the focus is not often on deliberative argumentation as such. Most modern theorists tend to see argumentation as a homogeneous phenomenon. But in recent years, there has been a tendency to differentiate more, especially in the works of Douglas Walton, who has defined different types of argumentative dialogue. However, we also need to differentiate in another way, namely on the basis of argumentative issues.
Aristotle did this when he defined the three main genres of rhetoric. And if we take a closer look at the nature of the issues in deliberative argumentation, several interesting implications will ensue. Deliberative argumentation will turn out to be at odds with assumptions widely accepted in current theories, such as pragma-dialectics and the model of “presumptive” reasoning advocated by Walton.
An essential fact about deliberative argumentation is that it is not about truth, but action. This fact has been cursorily acknowledged by some theorists, but not explored. Even Toulmin (1958), who made a strong case for distinguishing between argumentative fields, only considers arguments for claims like “Harry is a British citizen” and other constative propositions. Perelman and Olbrechts-Tyteca too fail to make a clear distinction. On the one hand, they emphasize that deliberative argumentation is “oriented toward the future” and “sets out to bring about some action or to prepare for it by acting, by discursive methods, on the minds of the hearers” (1969, 47); on the other hand, they consistently speak of “theses” presented for the audience’s assent. Characteristically, to find acknowledgement that the issues in deliberative argumentation are not propositions or theses, we must go to the textbook literature, including the work that Toulmin co-authored (1979). Educators remember what theorists like to forget: Deliberative argumentation is not about what is true, but about what to do.
A typical deliberative issue is (for the United States, at the time of writing), “starting a war on Iraq”, or “abolishing capital punishment”. It would be a categorial mistake to predicate truth, or falsehood, of these proposals. They are not propositions (assertions, constative statements); they do not predicate that anything is the case. Walton comes close to saying just that in his distinction between “practical” and “discursive” reasoning, when he states: “In the action type of critical discussion, the proposition is a practical ought-proposition that contains an imperative” (1996, 177). However, he blurs the distinction again by describing the deliberative issue as a proposition about what is prudential. The issue in deliberative argumentation is not a proposition; it is a proposal. It does not predicate a state of affairs, nor what ought to be the case; it proposes an action. It is like proposing a toast, or proposing marriage to someone. Proposals cannot be true or false.
All this is not to deny that deliberative argumentation usually involves a great deal of constative propositions, e.g., “Capital punishment reduces crime”. Such a claim may indeed be used as an argument in favour of capital punishment; but the ultimate issue at the top of the argumentative hierarchy is the decision on whether to have or not to have capital punishment. Similarly, the issues of recent referendums in Europe have not been propositions, but proposals to adopt the common currency, or to accept the treaty of Nice. Such issues cannot be formulated as constative statements, and they cannot have truth values. What we vote about is not the truth of a proposition, but the acceptance of a proposal.
It may seem formalistic to insist on this distinction. But it has important implications. One of them is that, strictly, there cannot be any logic of deliberative argumentation. This is because “logic” is about propositions, whereas deliberative argumentation is about proposals. And this accounts for another essential peculiarity of deliberative argumentation, namely what we may call its multi-dimensionality.
This term means that arguments for or against a proposal often belong to separate dimensions. If I propose marriage to someone, she might find me a prudential choice; but she might not love me. And even if she did love me, there would still be the fact that to marry her, I would have to break up my current family, which would be ethically questionable. So in deliberating upon my proposal of marriage, the chosen woman would have to do some mental juggling of arguments belonging to three dimensions: prudence, inclination, and ethics – and perhaps even more.
As we know from experience as well as from countless fictional narratives in literature, drama, or film, no logical rules can tell us how to put such heterogeneous arguments on a common denominator and calculate the net result. They lack commensurability. On capital punishment too there are many arguments on both sides, representing many dimensions. Some believe it reduces crime; others, that it does not. Both these arguments belong to the dimension of the socially advantageous, or, in Walton’s term, the “prudential”. But other arguments in the same debate belong to an ethical or religious dimension. Some argue that it is not fit for man to take another man’s life; others argue that God has ordained criminals to pay a life for a life. Again, the dimensions that the various arguments belong to lack commensurability.
By contrast, in a discussion of whether a certain proposition is true – that is, whether a certain predicate can be truthfully predicated of a certain subject – we only have to consider one dimension, namely the one represented by that predicate.
The insight that deliberative rhetoric is multidimensional is as old as rhetoric itself. We find it in Sophistic rhetoric, as in the “Rhetorica ad Alexandrum” (1937), the oldest extant book on rhetoric, once thought to be by Aristotle. This work, which predates Aristotle’s text with a few years, has the following checklist of dimensions in deliberative argument: “he who persuades must show that those things to which he exhorts are just, lawful, expedient, honourable, pleasant, and easy of accomplishment” (1421b). This type of rhetoric strikes some commentators as cynical or opportunistic in the way it suggests a battery of alternative ways to argue[i]; it has an air of “anything goes”. But this seeming opportunism represents the fundamental insight that we cannot decide hard issues by appealing to one general premise.
Aristotle, Plato’s student, who saw his task as that of turning textbook lore into a tekhne, tried to make deliberative debate neatly one-dimensional by declaring that “[t]he end of the deliberative speaker is the expedient or harmful [to sympheron kai blaberon] … all other considerations, such as justice and injustice, honour and disgrace, are included as accessory to this [symparalambanei]” (I, iii, 5; 1358b).
In Sophistic rhetoric, however, there are a diversity of equal, incommensurable dimensions, with no attempt to make one of them the master dimension or common denominator of all. I contend that the sophists are right: in actual deliberation, we find arguments belonging to all the dimensions, with no binding or intersubjective way to reduce them all to coefficients of the same denomination.
The insight that arguments may belong to several dimensions and hence not allow deductive inference to any conclusion or “truth” is central to sophistic thinking. It appears again in the great systematizer of Hellenistic rhetoric, Hermogenes (c. 150 A.D.), who states: “The practical issue is divided: legality; justice; advantage; feasibility; honour; consequence” (1995, 52). Conley rightly says of the Hermogenean system: “This is clearly a long way from the syllogism-based notion of rhetoric familiar from, say, Aristotle’s Rhetoric” (1990, 56). The diversity of the list, and the absence of “truth” from it, were no doubt some of the aspects of sophistic rhetoric that made Plato and others see rhetoric as opportunistic flattery and a method for turning black into white. We may compare this sophistic insight with the disillusioning discovery by the Pythagorean mathematicians of irrational numbers; for example, the relation between the diagonal and the sides of a square is irrational. This amounts to saying that there can be no common denominator, no commensurability, between them, i.e., the relation cannot be expressed by any fraction consisting of integers.
By contrast, economic cost and benefit are an example of commensurable entities, since both have the same denomination, namely money; they may therefore be reduced to one coefficient. Not so with the various arguments that are advanced about deliberative proposals such as adopting the Euro, abortion or capital punishment. There is no algorithm for tallying up the pros and cons.
This is why the distinction between and propositions and proposals is important. With propositions, we may, in principle, have deductive validity. A proposition is one-dimensional in that it asserts one predicate, and that is why the truth of that predication may follow from the truth of the premises. A proposal does not assert anything, but several propositions representing separate dimensions may be asserted as premises for or against the acceptance of the proposal.
This implies that in deliberative issues there is no deductive inference from premises to acceptance. This point is central to Perelman’s entire thinking about argumentation; indeed, he see the defining feature of “argumentation”, as opposed to “demonstration”, in the fact that argumentation is “noncompulsive”, i.e., deductive inference is not possible. By contrast, it may be possible to make a deductive inference to a constative proposition about a proposal, for example, that it will be economically beneficial; and this proposition may then be used in deliberative argumentation as an argument for adopting that policy. But there it will not have deductive force. There will always be other arguments in the matter, pertaining to other dimensions, and there is no deductive way to reduce the multiple, multidimensional arguments to one common denominator and deduce a net result.
Perelman and Olbrechts-Tyteca anticipated this characteristic of deliberative argumentation. They point out that “the possibility of arguing in such a way as to reach opposite conclusions” will always exist “when the argumentation aims at bringing about an action which is the result of a deliberate choice among several possibilities” (46).
But much current theory has failed to follow this lead. In pragma-dialectics, some form of deductivism is central, i.e., a belief in a normative rule demanding that the conclusion should follow in a valid manner from the premises. One of the ten basic rules of pragma-dialectics states: “A party may use only arguments in its argumentation that are logically valid or capable of being validated by making explicit one or more unexpressed premises” (Rule 8 in van Eemeren et al., 1996, 284). But as we have just seen, because deliberative argumentation is about proposals and hence multidimensional, it does not allow for logical validity.
Pragma-dialecticians are aware of a difficulty and tend to point out that “valid” should be taken in a different sense. A footnote to the passage just quoted states that valid is used in “a broader sense”, so that there is no “dogmatic commitment” to deductivism. However, it never becomes quite clear in what broader sense “valid” is to be taken. There are sporadic comments, but they all deal with the kind of reservations about validity that are internal to the purely formal definition, e.g., concerns about how to deny valid status to a conclusion that tautologically repeats a premise. What we generally do not find in pragma-dialectics is a clear recognition that arguments in, e.g., ethical or political debate may be perfectly good and legitimate, and yet not be valid in any sense resembling deductive validity.
The qualification that arguments, if not logically valid, should be “capable of being validated by making explicit one or more unexpressed premises” does not fix this hole in the theory. The unexpressed premises thus imputed to people so that their arguments may be “validated” are, in many cases, premises that these people themselves would reject. For example, a British opponent of the Euro may believe in the argument that Sterling, as a symbol of national identity, should be preserved. But that person is not thereby committed to the premise that any symbol of national identity should be preserved. And only such a general premise would serve to “validate” his argument against the Euro. So the notion of “validating” arguments by reconstructing their unexpressed premises does not do justice to the way many people actually use arguments on deliberative issues.
Another example of a premise where this kind of validation would misrepresent the arguer’s own standpoint may be cited from a deliberative debate discussed in Jørgensen, Kock & Rørbech (1998). The issue was whether to ban surrogate motherhood arrangements. The opponent was Ms. Pia Kjærsgaard (later to become leader of the anti-immigrant Danish People’s Party, which recently has generated international attention). Her main argument was that a ban on surrogate motherhood arrangements would be a curtailment of personal freedom. Interestingly, this charismatic and powerful political leader lost the debate. But what concerns us more now is the fact she would never accept a general premise rejecting any curtailment of personal freedom. After all, any law curtails personal freedom. For example, her party has recently helped introducing new laws curtailing citizens’ rights to bring foreign spouses to the country.
Several theorists who sympathize with pragma-dialectics have sensed that its deductivist position is in need of qualification or defence. One such theorist is Leo Groarke (1999), who states, with praiseworthy explicitness, “natural language arguments should be understood as attempts to formulate deductive arguments” (1999, 2). He points out that validity in the relation between premises and conclusion only means that the conclusion preserves any certainty inherent in the premises, not that a certain conclusion can be drawn from uncertain premises. But even with this – perhaps rather obvious – qualification, deductivism is still at odds with the kind of arguments found in deliberative debate. And the way Groarke speaks of “inductivism” as the only alternative to deductivism indicates that in fact he only has argumentation about constative propositions in mind. The fact is that in deliberative debates we often hear arguments that are quite certain and legitimate, for example that if we adopt the Euro, we will not need to change our money when travelling to another member country; but in spite of such unassailable arguments, the conclusion, namely the adoption of the Euro, does not follow deductively (as was demonstrated when a majority of Danish voters rejected the Euro in September 2000).
Another attempt to preserve some version of the normative validity requirement is based on the idea of arguments being presumptive or defeasible. Douglas Walton is the foremost exponent of this approach. However, the notion of presumptiveness is quite slippery. It is clear that presumptive reasoning is non-monotonic, in the sense that new arguments may come up so that debaters are no longer committed to the presumed conclusion. But what is the nature of this commitment to the presumed conclusion – as long as it lasts? It seems that there are two versions of this commitment, one weaker and one stronger. In the weak version, when an arguer offers an argument in support of a conclusion, then a burden of proof is shifted onto the respondent, who then has to question or attack the argument. By doing that, he can shift the burden back onto the other side. In the strong version of what presumption means, the respondent is committed to accepting the conclusion, in a presumptive way, unless he can find fault with the argument.
This latter meaning of presumption seems to be understood in the following statement by Walton, summarizing the views of van Eemeren and Grootendorst (1992): “If the hearer accepts the premises of the speaker’s argument, and the argument is an instance of a genuine and appropriate argumentation scheme (for the type of dialogue they are engaged in), then the hearer must or should (in some binding way) accept the conclusion” (1996, 10). Walton goes on to say that this “does not appear to be “validity” in the same sense in which the word is familiarly used in deductive (or even inductive) logic”. But still we find here the same general tendency as in the deductive model of argumentation: if an argument is “valid”, then it means that the hearer is in some way “bound” to accept the conclusion. Validity, even if it does not means deductive or monotonic validity, somehow means “bindingness” – although the precise nature of the binding commitment or burden is often hard to pin down.
I suggest that argumentation theory, at least as far as deliberative argumentation is concerned, needs to abandon the notion that the validity of an argument has to do with the conclusion being in some way binding. Plain deductivism, reconstrutionism, and presumptionism are all versions of the same deductivist way of thinking about argumentation. But for deliberative argumentation at least, this way of thinking is false. A look at any deliberative debate will show that the arguments used there may be perfectly good and legitimate, indeed that they may fully deserve the term “valid” – and yet the conclusion they support does not follow in any binding way. In most cases, not even the debater who uses a given argument in deliberation believes that the hearer should be bound by the conclusion. Moreover, respondents in deliberative argumentation often do not feel obliged to raising critical questions about their opponents’ arguments, either. This is not because they abandon their standpoint or shirk their duties as debaters. Just as often, it is because they recognize that the opponent has a legitimate argument; but, on the other hand, they believe they have arguments for their own standpoint that have greater weight.
The reason that deliberative debaters may think so is precisely that deliberative argumentation is multidimensional. This explains why arguments may be perfectly good and yet not binding.
In a recent paper by van Eemeren himself, with Peter Houtlosser (2000), we find an excellent example of deliberative argumentation that captures many of its central features. They quote a heated British debate on fox-hunting, which can be seen as illustration of how each side, precisely because of the multidimensionality of such debates, has legitimate arguments which carry some weight, but which cannot in themselves entail a conclusion.
The anti-hunters argue that fox-hunting is cruel, and they draw an analogy to cock-fighting and bear-baiting – both of which were banned long ago. The pro-hunters argue that a ban would unsettle popular rural traditions and have a divisive effect, “setting town against country”. Both these arguments are legitimate and carry some weight, yet neither of them is in itself sufficient to entail a conclusion. Even many of those who would use one of these arguments in a debate over this issue are probably not ready to accept a “reconstructed” general premise that would make their argument deductively valid; even die-hard fox-hunters hardly believe that any socially divisive policy should necessarily to be rejected. The abolitionist campaign in the United States 150 years ago was socially divisive and did set town against country; and even for an abolitionist like Lincoln himself, this argument no doubt was legitimate and had a certain weight. However, in the particular situation it was outweighed, for him and for many other Americans, by other considerations. Similarly, the cruelty argument is legitimate and yet not deductively valid. There are many cruel practices in our society, some of them traditional and some modern, but recognizing that they are cruel does not entail a commitment to having them all banned. Neither does the analogy to other cruel practices that have been banned entail such a commitment. One debater in van Eemeren and Houtlosser’s article offers further analogies such as horse races and “the far larger cruelty of factory farming”. However, many people who feel that there is indeed an amount of cruelty in horse racing and factory farming probably do not believe that they should eo ipso be banned. Thus, when theorists impute such an unexpressed belief to them in order to “validate” their argument, the theorists are at odds with how people actually think.
The example questions not only the deductivist account of argumentation, but also the presumptionist theory. That theory would hold that if a debater points out the cruelty in fox hunting and argues that it should therefore be banned, then that presumption stands, and the opponent should then carry the burden of proof and refute the argument. But none of the pro-fox-hunters in the debate seem to have tried to refute the cruelty argument, in fact they may tacitly have recognized its legitimacy; instead, they meet it with an argument belonging to another dimension, i.e., the social good of hallowed traditions and the avoidance of divisive laws. Thus an ethical argument is countered, not cancelled, with social arguments. One may see all these arguments as acceptable and having at least some weight – and many people probably do. This is tantamount to saying that none of them is logically valid or “binding”, not even in the “presumptive” way.
A final, paradigmatic example may be in order. In an article titled “The Right to Live vs. the Right to Die: No Single Yardstick”, columnist Ellen Goodman (1986) describes two cases of people who have wished to be allowed to die by starvation. One is an 85-year-old man in Syracuse, N.Y., who has recently had a stroke, and who has deliberately stopped eating. The administrators of the nursing home where he lives want to force-feed him, and they take the case to court, but Justice Miller of the State Supreme Court rules against them, writing in his ruling, “I will not, against his wishes, order this man to be operated upon and/or to be force-fed”. Goodman comments that she approves of this ruling. The fact that the man wishes to die of starvation is a legitimate reason in favour of letting him die – but not a reason that deductively entails the decision taken.
Here Groarke’s point about deductive validity being only certainly-preserving, not certainty-establishing, is irrelevant: the man certainly wishes to die, and this is certainly a legitimate argument, yet the decision does not follow deductively. It would obviously be false to “reconstruct” a general unexpressed premise underlying the Justice’s decision (and Goodman’s approval of it) to the effect that “all persons who wish to die of starvation should be allowed to do so”. The premise we may reconstruct is rather that a person’s wish to die of starvation is a reason in favour of letting that person do so. No more, no less.
That this is so is brought out explicitly by Goodman’s second example: a 26-year old woman in California, severely handicapped by cerebral palsy, wants to be allowed to starve herself to death. Yet here the judge denies her request. And Goodman agrees with this decision too. But there is no inconsistency. It is much more reasonable to say that in both cases, she (and the judges handling the cases) hold the premise, stated before, that a person’s wish to die of starvation is a reason in favour of letting that person do so. No more, no less. It is a premise with some weight in both cases, but in neither case does this premise, which is undoubtedly true and certain, deductively entail the conclusion. In one case, this reason is on the “winning” side of the argument; in the other case, on the losing side. But in both cases, it is legitimate and has a certain weight.
In defence of the deductivist account, one might rightly point out that the patient’s own wish is not the only premise in either of the two cases. The deductivist might then say that when this premise is added to the other pertinent premises in each of the cases, then the conclusion in each of the two cases follows deductively. In other words, for the old man one would say something like this: his own wish, his advanced age and the nature of his illness together entail the conclusion that he should be allowed to die. For the young woman, her youth would be one of the premises that, in spite of her own wish, deductively entail the opposite conclusion.
It is easy to see how artificial such an account would be. How does one add up the premises favouring a certain conclusion, and how does one subtract the ones favouring the opposite conclusion? How old does one have to be to be allowed to starve oneself to death? How ill? We would need an algorithm assigning a specific weight to each premise, using the same common denominator for all the premises, and we would need a rule defining just how much weight on one side would be needed to constitute a deductively valid inference. The two cases in question were both brought to court and decided there, but obviously no such formula exists in the laws of the two states. Even if it did, it is obvious that a rule stating just how much weight is needed to make a conclusion deductively “valid” would be quite arbitrary; a different threshold value might just as well have been chosen.
For Goodman too, a whole set of considerations explains why she thinks differently of the two cases. But that is precisely the nature of making decisions, whether in court, in politics, in ethics or in everyday life. In a situation where several considerations or premises simultaneously play legitimate parts, the demand that conclusions must be as certain as their premises is doomed to failure, or forced to resort to artificial ad hoc constructions. The only natural way to account for argumentation in such situations is to say that a number of arguments or premises are all legitimate and relevant, but that there is no deductively valid link from the relevant premises to any conclusion.
Indeed, we might argue that the use of the term “valid” in logic is a misnomer, and that the term might be much better employed for precisely those arguments that are legitimate without being deductively valid or cogent. Instead, “valid” arguments would be those that speak with some weight for the conclusion.
If one follows Walton’s account, one might object that these cases still allow of a semi-deductivist or “presumptionist” interpretation. A patient’s own wish to die of starvation, we might say, creates a presumption that the patient be allowed to do so – unless there are other factors that negate this presumption. Thus we have a valid inference of the “presumptive” or “defeasible” kind.
The answer to this account is that there are always other factors. They do not arrive out of the blue; they are always there already. But in neither of the two cases do these other factors that may plead for the opposite decision negate the legitimacy of the patient’s own wish. That wish remains a legitimate argument of some weight, even if we decide that there are other arguments of greater weight that plead for the opposite decision. The idea that we either have to negate and demolish an argument, or else accept the conclusion for which it pleads, is false.
The two cases described by Goodman are telling examples of how the making of decisions in politics, ethics or law is better described by the term “casuistry”, as defined by Jonsen and Toulmin (1988), than by a model based on the deductive application of general principles.
And since we now have broadened our scope from deliberation to ethics and law, we might point to the theory of legal reasoning proposed by Robert Alexy (1978) – a theory based on the insight that judicial decisions, at least in the “interesting” cases, cannot be deductive.
It seems that we need an alternative metaphor for thinking not only about deliberative argumentation, but more broadly about how we discuss decisions – instead of the old metaphors that have to do with “chains” of reasoning or lines of argument that “bind” the opponent.
The ancient forensic image of the scales in which opposing arguments are weighed is a good starting point, emphasizing as it does that deliberation is related to libra, the Latin word for scales. However, while this image is illustrative of some features of deliberative debate, it is misleading about others.
Some interesting aspects well illustrated by the “scales” image are:
1. In deliberative debate there is no deductive or “valid” demonstration of the claim, in the sense of “valid” defined by logic. Indeed, deliberative debaters often do not proceed from “premises” to “conclusion”, as logicians do, but the other way around, i.e., they begin with a standpoint for which they then try to find arguments. To apply the scales image, people generally have a preconceived wish to tip the scales one way or the other, and they look for weights to throw into one of the two pans.
2. Arguments used by deliberative debaters defy evaluation by binary standards such as valid/invalid, or sound/unsound; in deliberative debate there will generally be some arguments on both sides that have some weight.
3. Deliberative debaters generally do not intend or pretend that their arguments make their proposal logically valid or “binding”. This calls into question the way argumentation theorists “reconstruct” deliberative arguments by introducing “validating” premises.
4. Deliberative debate is usually not linear, i.e., it is usually not limited to the establishment or refutation of one “clinching” argument. This is because there are no clinching arguments in deliberative debate, which again is why there are often several arguments on each side. Staking all on one line of argument in the belief that if the opponent accepts that, he must also accept the conclusion, is illusory.
The renaissance thinker Lorenzo Valla, a harsh critic of the medieval mode of thinking that aimed at logical proof in human or theological matters, made this point eloquently when he wrote, in a commentary on the medieval philosopher Boethius: “What is more inept than arguing the way the philosophers do, where, if one word is wrong, the whole case falls? The orator, on the other hand, uses many reasons of various kinds, he brings in opposites, he cites examples, he compares similar phenomena and forces even the hidden truth to appear. How miserable and inept is the general who lets the entire outcome of the war depend on the life of one single soldier! The fight should be conducted across the whole front, and if one soldier falls, or if one squadron is destroyed, others and still others are at hand. This is what Boethius should have done, but like so many others he was too deep in love with dialectics” (Valla, 1970, 113)[ii].
All this is well illustrated by the “scales” image. However, even more interesting are some features of deliberative argumentation that this image misrepresents:
5. The total weight of the arguments on any side cannot be calculated by adding and subtracting the weights of all arguments on both sides. Arguments in deliberative debate lack commensurability, i.e., they cannot be put on a common denominator in any binding way. Attempts to tally up the relative merits of alternative proposals in an objective fashion, e.g., in terms of economic cost and benefit, are thus illusory.
6. This is because the weight of each argument is a subjective or “phenomenological” property relative to each member of the audience. As we saw, one may acknowledge that fox-hunting has some social value, and at the same time feel that the suffering inflicted on the foxes argues against fox-hunting with greater weight. But for the person sitting next to you it might be the other way around.
7. However, even if there is no binding or formal way to define a “common denominator” for the pros and cons on a given deliberative issue, people nevertheless may have to decide between the two sides. And somehow they manage. Sometimes they even change sides after listening to argumentation. Apparently they do find a way to put the arguments on the same scales and assess which side has more weight. But this cannot be done formally; which way the scale tips is, for each person, a “phenomenological” property, resulting from the total impact of all the rhetorical stimuli which that person has received. In deliberative argumentation gravity, too, is relative. Weight in deliberative argumentation is a matter of degrees: deliberative arguments are not either valid or invalid, but have more or less weight. But that weight is relative to the person who judges it, and that person’s judgment is influenced by the rhetoric that is used to either enhance or reduce that weight. Enhancing the weight of an argument is what Aristotle called “auxesis” and Latin rhetoricians “amplificatio”; reduction is “meiosis”. The insight that the weight of an argument may be enhanced or reduced by degrees, and for each member of the audience individually, is one of the defining features of rhetoric; the insight that arguments belong to many dimensions is the other.
The very fact of multidimensionality in deliberation, which makes deduction in any form impossible, also makes rhetoric necessary. Deliberative argumentation is full of arguments on both sides that all have a certain weight – except that their weight is anything but certain or definite, but changeable and relative.
An important implication of all this for the normative evaluation of actual political debate is the following.
The idea that the other side may have legitimate arguments of some weight is abhorrent to many politicians. They tend to ignore, misrepresent or offhandedly dismiss any argument that can be made against their own policies or in favour of their opponents’. However, many voters probably tend to think that on most contested issues, there is in fact something to be said on both sides. Such voters will want to know what it is and to get a chance to evaluate the relative merits of all reasonable arguments. Politicians who flatly deny that the opposition may have a point, and who maintain that their own policies are unassailable, are not credible in such people’s eyes.
Argumentation theory should teach deliberative arguers to acknowledge legitimate arguments on the opposite side. It should also point out that the acceptance of some of the opponent’s arguments does not entail a commitment to the opponent’s proposals or policies. It should keep a vigilant eye on debaters who tend to suppress or misrepresent arguments made by the opponent; this is something that pragma-dialectics has always emphasized, and rightly so. And they should show that the necessary function of rhetoric is to find the available means of persuasion on both sides and to help audiences form their subjective assessments of their relative weight. This would in turn help democracies sustain the credibility of political processes currently threatened by polarization, non-participation and cynicism.
NOTES
[i] Manfred Fuhrmann, who edited the only modern text of this work, has also written an introduction to classical rhetoric in which he indignantly dismisses it: “Seine Lehre ist radikal relativistisch – nicht aus der Fülle des eigenen Schöpfertums, wie die des Gorgias, sondern aus dem baren Opportunismus. Seine aalglatte Routine kennt nichts als eine Vielfalt von Situationen, die ein Redner zu meistern hat, und sie sucht für eine jede von ihnen möglichst viele und hilfreiche Argumente an die Hand zu geben, gute und schlechte, wie es sich trifft; eine Bewertung und Auslese nach irgendwelchen Prinzipien findet nicht statt” (Fuhrmann, 1984, 29).
[ii] Quid enim ineptius philosophorum more ut si uno verbo sit erratum tota causa periclitemur? At orator multis et variis rationibus utitur, affert contraria, exempla repetit, similitudines comparat et cogit etiam latitantem prodire veritatem. Quam miser ac pauper imperator est qui omnem fortunam belli in anima in unius militis ponit! Universitate pugnandum est et si quis miles concidit aut si qua turma plofligata est, alia subinde atque alia sufficienda. Hoc modo agendum Boethio erat, qui ut plurimi alii nimio amore dialectice deceptus est.
REFERENCES
Alexy, Robert (1978). Theorie der juristischen Argumentation. Frankfurt a.M.: Suhrkamp.
Conley, T.M. (1990). Rhetoric in the European tradition. Chigaco: The University of Chigaco Press.
Aristotle (1926). Volume xxii. The art of rhetoric. Translated by J.H. Freese. (Loeb Classical Library, 193.) Cambridge, Mass.: Harvard University Press.
Aristotle (1937). Volume xvi. Problems, books 22-38. Rhetorica ad Alexandrum. Translated by W. S. Hett, H. Rackham. (Loeb Classical Library, 317.) Cambridge, Mass.: Harvard University Press.
Eemeren, F.H. van, & Grootendorst, R. (1992). Argumentation, communicatio, and fallacies. Mahwah, New Jersey: Lawrence Erlbaum Associates.
Eemeren, F.H. van, Grootendorst, R., & Snoeck Henkemans, F. (1996). Fundamentals of argumentation theory: a handbook of historical backgrounds and historical developments. Mahwah, New Jersey: Lawrence Erlbaum Associates.
Eemeren, F.H. van, & Houtlosser, Peter (2000). Argumentation, interpretation, rhetoric. http://www.argumentation.spb.ru/2000 1/papers/1 2000p1.htm.
Fuhrmann, M. (1984). Die antike Rhetorik: eine Einführung. München: Artemis-Verlag.
Goodman, E. (1986). The right to live vs. the right to die: no single yardstick. In W.F. Smith & R.D. Liedlich (Eds.), From thought to theme. A rhetoric and reader for college English. Eighth Edition (pp. 386-388). San Diego: Harcourt Brace Jovanovich.
Groarke, L. (1999). Deductivism within pragma-dialectics. Argumentation, 13, 1-16.
Heath, M. (1995). Hermogenes on issues: strategies of arguments in later Greek rhetoric. Oxford: Oxford university Press.
Jonsen, A., & Toulmin, S.E. (1988). The abuse of casuistry: a history of moral reasoning. Los Angeles: University of California Press.
Jørgensen, C., Kock, C., & Rørbech, L. (1998). Rhetoric that shifts vores: an exploratory study of persuasion in issue-oriented public debates. Political Communication, 15, 283-299.
Perelman, C., & Olbrechts-Tyteca, L. (1969). The new rhetoric: a treatise on argumentation. Translated by John Wilkinson and Purcell Weaver. Notre Dame: University of Notre Dame Press. (Original: Traité de l’argumentation: la nouvelle rhétorique (1958). Paris: Presses universitaires de France.)
Toulmin, S.E. (1958). The uses of argument. Cambridge: Cambridge University Press.
Toulmin, S.E., Rieke, R., & Janik, A. (1979). An introduction to reasoning. New York: Macmillan.
Valla, L. (1970): De vero falsoque bono. Critical edition by Maristella de Panizza Lorch. Bari: Adriatica Editrice.
Walton, D. N. (1996). Argumentation schemes for presumptive reasoning. Mahwah, N.J.: Lawrence Erlbaum Associates.