ISSA Proceedings 2006 – A Bottom-Up Approach To Argument Schemes: The Case Of Comparative Argument
The question of the internal structure of argumentation and the identification of the various argument schemes constitutes a central stake for argumentation studies. Analyzing argumentation requires that the analyst adopts a somewhat acrobatic but necessary median position in order to place himself at an intermediate level, between the “letter” of the argumentation (its very content, which is proper to a specific (text/discourse) and its “logical” structure (its possible translation into a general logical scheme, which misses most of the substance of the argumentation).
Distributing the various arguments we may be confronted to into general schemes, according to the nature of the relation which links the argument to the conclusion allows to distance oneself from the literal and specific content of an argumentative discourse in order to gain in abstraction. It then becomes possible to compare various argumentative speeches, dealing with various subject matters, but susceptible to mobilize argumentative strategies resorting to similar configurations of argument schemes.
Nevertheless, at the moment, researchers in the field of argumentation studies, and maybe in particular in the French-speaking sphere, have no systematic, coherent typology of argument schemes at their disposal.
The reference typologies are mostly directly inspired by that proposed by C. Perelman and L. Olbrechts-Tyteca. Their classification may be of great help because of the variety of the argument schemes it comprises (and because of the associated definitions), but it is weakened by a lack of coherence in the proposed classification criteria and therefore, by the heterogeneousness of the categories considered as argument schemes. Beyond these theoretical difficulties, the application of the model – trying to identify argument schemes used by arguers in everyday discussions – is far from being simple.
This paper will focus on a broad group of arguments defined by the fact that they are based on a comparison process. More precisely, on the basis of a similarity between two cases, such arguments focus on a characteristic of the case which constitutes the analogue, or the source, or phoros of the comparison, and extend it to the second case, which constitutes the primary subject, or thema, or target, of the argumentation.
In order to avoid the trap mentioned by Christian Plantin, who claims that “any proposition of synthesis of existing typologies finally results in an additional typology” (2005, p. 50), this paper will be limited to a non-exhaustive inventory of several parameters identified in various academic works, parameters which permit a sub-categorization of arguments based on a comparison.
Considering that an argument scheme is associated with a set of specific critical questions, we will test the relevance of such a sub-classification of comparative arguments for ordinary arguers. We will then investigate whether speakers, when engaged in an argumentative discussion, use “wide-spectrum” refutation strategies in order to counter an opponent’s comparison, or whether they use specific refutation strategies according to the sub-type they are confronted with.
1. Sub-classifications within comparative arguments.
1.1.A first distinctive feature which introduces a sub-classification in the group of arguments based on a comparison opposes:
– comparisons which parallel situations or cases issued from two heterogeneous domains of knowledge (see, for instance, a beautiful example from C.S. Lewis, quoted by Govier (2001, p.350-351):
“You can get a large audience together for a strip-tease act – that is, to watch a girl undress on stage. Now suppose you came to a country where you could fill a theatre simply by bringing a covered plate onto the stage and then slowly lifting the cover so as to let everyone see, just before the lights went out, that it contained a mutton chop or a bit of bacon, would you not think that in that country something had gone wrong with the appetite for food?)”:
– to comparisons which parallel situations from the same domain of knowledge (most of the comparisons of the a pari type seem to belong this category ; see the example quoted by Garssen (1994, p.106):
“You should give Miriam an expensive birthday present, because on Alice’s birthday you also gave her an expensive present”.
For some authors, observing the homogeneity or the heterogeneity of the domains of knowledge involved in the comparison allows the introduction of subtypes within the major type of “comparative arguments”; see for instance Snoeck-Henkemans 2003, who opposes figurative comparison and literal comparison on this criterion. But it may be given a more decisive importance, for instance by Perelman, who thus supports his distinction between comparison arguments (which he defines as a subtype of quasi-logic arguments) and arguments by analogy (a subtype of arguments establishing the structure of reality; see Perelman & Olbrechts-Tyteca 1988).
Although intuitively acceptable, in practice, the line between these two types of comparative arguments is often quite hard to draw. Maybe the distinction would be more adequately thought of as a gradual one, from arguments of comparison that bring together two cases from overlapping domains of reference, to arguments of comparison implying cases issued from maximally distant areas, with intermediate cases between these two extremes (for instance, when a comparison involves two situations within the same cultural area but temporally distant from one another – as in arguments from the precedent).
The first case of arguments by comparison (when the reference areas overlap) is very close to argumentation by example, who may even be considered by some authors as belonging to the comparative argument schemes. For instance, for Plantin (1996, p.50), “Inductive argumentation analogically generalizes to all cases an observation drawn from a few cases”. Similarly, Amossy (2006), following Aristotle, equates argumentation by example and analogy (for instance, one chapter from her book L’argumentation dans le discours is entitled: “L’exemple, ou la preuve par l’analogie”). The same could be said of Schellens (1985) classification of argument schemes[i].
On the contrary, authors like Perelman (who holds argumentation by example to be another subtype within the arguments establishing the structure of reality; see Perelman & Olbrechts-Tyteca 1988), or like Govier (2001), or Kienpointner (who considers inductive arguments as belonging to the warrant-establishing argument schemes; see Kienpointner 1992), clearly distinguish between comparative arguments and arguments by example (the same could be said of the pragma-dialectical classification of argument schemes, according to which arguments from example belong to the symptomatic type).
1.2. Qualitative/quantitative comparison: another way of differentiating comparative arguments consists in opposing comparisons based on quantitative considerations (A is as p as B /A is more p or less p than B) and comparisons based on qualitative considerations (A is like B). For Perelman, qualitative comparisons are typical of what he calls “arguments by comparison”, as opposed to arguments by analogy” – that is, he tends to merge the first and the second criteria. He justifies that he considers comparison as a quasi-logic process as follows:
“By saying “his cheeks are red like apples” or “Paris has three times as many inhabitants as Brussels”, or “He is more beautiful than Adonis”, we compare realities in a way which seems more likely to be proved than a mere resemblance or analogy claim. This feeling is due to the underlying idea of measurement behind these utterances, even if the slightest criterion for proceeding to this measurement is lacking. In this sense, comparison arguments are quasi-logic arguments.” (Perelman & Olbrechts-Tyteca 1988, p.326; the translation is ours)
We assume that criterions (1) and (2), though often associated, should be distinguished, for a figurative comparison may be grounded on quantitative considerations – and conversely.
1.3. A third criterion within comparative arguments relates to the epistemic status of the compared cases: the analogy may involve two real facts, or a real fact and a hypothetical, invented one. As Govier (2001) points out, the fictitious nature of the analogue is acceptable because its first quality is not its veracity, but rather the fact of being consensually evaluated – be it positively or negatively – by the audience. Thus she opposes inductive analogies and a priori analogies, that is, analogies that are not empirically based. Once again, criterion 2 and criterion 3 are linked: quantitative comparisons usually make sense if both elements of the comparisons are hold to be true. Note that the epistemic status of the compared cases may be relevant in some qualitative comparisons: this is the reason why we consider the last two criterions separately.
1.4. A fourth opposition within comparison arguments distinguishes between those assuming an essentially positive function (i.e., supporting one of the arguer’s standpoints) and comparison arguments assuming an essentially negative function (i.e. rebutting the adversary’s argument). The negative use of analogy has been labelled “rebuttal analogy”, or “refutation by logical analogy”, which is, as per Govier (2001, p.357), an analogy which is designed to refute the opponent’s argument by showing that it is parallel to a second argument in which a comparable premise leads to a clearly unacceptable conclusion. The refutation by logical analogy may be seen as a subtype of the ad absurdum argument (see Eggs 1992, for example).
Some additional oppositions permit characterization of other subtypes within comparative arguments. For instance, as Declercq (1992, p.108-109) points out, when the comparative argument has a narrative form, it may consist in a parable or a fable. When the comparison is aimed emphasizing the differences rather than the similarities between two cases, it may result in an a contrario argument; if it is based on a double hierarchy, according to Perelman, it becomes a argument a fortiori. This paper will focus on the first three criterions only.
2. The evaluation of comparative arguments by ordinary speakers
Let us now turn to the question of the evaluation of comparative arguments by ordinary speakers. When engaged in an argumentative discussion, the rebuttal strategies used by arguers in order to challenge the opponent’s arguments are precious indicators of the argumentative norms they adhere to, and of their awareness of the various argument schemes involved in the argumentative process.
According to Eemeren, Grootendorst and Snoeck Henkemans (2002), the main critical question to ask about argumentation based on analogy is:
“Are there any significant differences between Z and X? Such differences can be pointed out in two ways: by claiming that Z has a certain characteristic that X does not have, or vice versa. Both forms of criticism are serious charges because basing argumentation on a relation of analogy assumes that X and Z share all characteristics relevant to the argument.” (p.99-100)
This critical question guides most of the general refutations of comparative arguments, directly – by pointing out the differences presented as essential between the compared cases, or indirectly – by proposing a counter-analogy presented as more conclusive than the opponent’s one because of its better adequacy to the target.
Nevertheless, these general strategies of refutation of comparative arguments (beyond wide-spectrum refutation devices such as the French very common injunction “il faut comparer ce qui est comparable”) may take different forms according to the subtype of comparative argument involved in the discussion.
As far as the epistemic status of the comparison elements is concerned, one may expect that it entails specific modes of evaluation in an argumentative discussion. As Govier (2001) puts it, “Some aspects of inductive analogies [that is, analogies that are empirically grounded] make their evaluation different. The most obvious of these is that in the inductive analogy, the analogue must describe something real, and the quoted facts must be genuine. Imaginary examples are fine for a priori analogies, but not for inductive ones. The similarities on which inductive analogies are based are between empirical aspects of the primary subject and the analogue. We cannot determine the extent of the similarity merely by reflecting on structural features, as we can for a priori analogies.’ (…) Another significant fact about inductive analogies is that the cumulative effect of similarities is an important factor. In an a priori analogy, what is important is that the similarities relevant to the conclusion hold. If they do, it does not matter whether there are many further similarities or none at all. But in the inductive analogy, the sheer number//of similarities does matter. The closer the two cases, in detail, the more likely it is that the inferred conclusion will be true. This means that the evaluation of inductive analogies depends more on factual background knowledge than does the evaluation of a priori analogie” (p.370-371).
The importance of factual background knowledge is illustrated in the following example, where the author challenges the parallel drawn by a participant in a newsgroup between the US embargoes on Iraq and on Cuba.
Certainly not, the embargo that strikes Iraq, associated with an intensive bombing, is in no way comparable to the embargo endured by Cuba, neither in its historical conditions, nor in its field of application.
Comparing the embargo on Iraq with the embargo on Cuba is a nonsense, they are essentially different.
Embargo on Cuba is an economic embargo imposed by the United States since 1960 and Cuba, member of the COMECON until 1989 has always traded more or less freely with the USSR and other Comecon members plus China. As a symbol of the communist resistance to the American imperialism Cuba was even very generously supported by USSR. The harder period for Cuba was the end of the soviet era in 1989 but since Castro decided to open the economy at the beginning of the 90s, the European Union started investing in and trading with Cuba. In 2001 Cuba welcomed over 2 millions tourists and Cuba was the first South America country to use Euro as a trading money with the EU countries. So please, try to compare what can be!
In this message, the comparison between the embargo on Cuba and the embargo on Iraq is challenged based on a detailed analysis of the historical context of the embargo on Cuba. The difference with the embargo on Iraq is not explicitly stated but the addressee is expected to infer it owing to his knowledge of its historical context.
The same observation can be made for quantitative comparisons, the refutation of which may focus on the balance between similarities and differences between the terms of the comparison, or on the accuracy of the quantitative data.
Let’s now turn to the question of the homogeneity versus heterogeneity of the areas involved in the comparison.
Confronted with comparative arguments which parallel cases issued from highly heterogeneous domains of knowledge, the refutation cannot relevantly focus on factual differences between the compared cases, nor on the degree of proximity between them. On the other hand, the refutation may rather focus on what Perelman calls the “interaction” between the phoros and the thema provoked by the comparison. Besides the characteristic of the phoros that is meant to be transferred onto the thema, a phenomenon of contagion may be observed between other features of the phoros which extend to the thema of the argumentation (Perelman & Olbrechts-Tyteca 1988, p.508).
Such an interaction can be subject to criticism, as illustrated by the following message, in which the author criticizes a comparison made by French politicians between students engaged in a protest movement and toothpaste (that is, between unquestionably heterogeneous elements). The comparison is the following:
“Students are like toothpaste: once out from the tube, there is no getting them back”.
The comparison is criticized in a post on a political Internet newsgroup as follows:
Newsgroup: fr.education.divers, fr.soc.politique
No need to be a qualified teacher in literature to realize that the image is not trivial, but clearly insulting. The compared term (students, a human population hence a priori worthy of respect) and the comparing term (toothpaste) are merged owing to a shared sordid characteristic: their alleged capacity of escaping in an uncontrolled flood which cannot be contained. Given the -established – level zero of consciousness of toothpaste, the students protest movement would be, as toothpaste is, submitted to obscure physical laws, whose only ‘raison d’être’ (like the law of the slice of bread and butter that always falls on the bad side) is to cause maximum trouble to the Minister.
The comparison between students and toothpaste is denounced as activating more than the alleged shared characteristic (the fact that both are ‘impossible to contain’): these additional features are ‘the lack of consciousness’ and the obedience to physical laws oriented towards causing maximal trouble to the environment. Thus the criticized comparison is seen as revealing a highly negative perception of the students’ protest movement.
Here, the type of criticism of the comparison used in example 1 would make no sense, since from a factual point of view, there is no doubt that students differ very much from tooth paste.
As a conclusion, it seems quite hazardous to try and systematically match a specific evaluation device with each subtype of comparative argument. We would rather assume that the criticism of a comparative argument is likely to take specific forms if one considers very distant subtypes of comparative arguments, as in examples 1 and 2; the extreme poles of the opposition being a priori qualitative figurative comparative arguments, as opposed to quantitative, literal, inductive comparative arguments.
We will conclude with the observation that comparative arguments of the latter type, in polemical contexts, often elicit criticisms which require an ever-increasing degree of factual similarity between the compared elements. Such a criticism, in the end, may result in an outright rejection of comparison as an argument, on the ground that, following the French expression, “comparaison n’est pas raison”, for comparative arguments always involves some kind of shift, which makes them specifically vulnerable to refutation.
[i] According to Kienpointner, ‘Les arguments peuvent-ils faire l’objet d’une classification exhaustive? Sur la complétude des typologies d’argument’ (Conference on ‘Structures argumentatives et types d’arguments’, Paris, May, 26th, 2006).
Amossy, R. (2006). L’argumentation dans le discours. Paris: Armand Colin.
Declercq, G. (1992). L’art d’argumenter. Structures rhétoriques et littéraires. Paris: Editions Universitaires.
Eemeren, F. H., Grootendorst, R. & Snoeck Henkemans, F. (2002). Argumentation Analysis, Evaluation, Presentation. Mahwah, New Jersey / London,Lawrence Erlbaum Associates.
Eggs, E. (2002). Grammaire du discours argumentatif. Paris: Kimé.
Garssen, B. (1994). Recognizing argumentation schemes. In: F.H. van Eemeren & R. Grootendorst (Eds), Studies in Pragma-dialectics. Amsterdam: SICSAT, 105-111.
Govier, T. (2001). A Practical Study of Argument. Belmont: Wadsworth (5th ed.).
Kienpointner, M. (1992). How to Classify Arguments. In: F. H. van Emeren, R. Grootendorst, J. A. Blair, C. A. Willard (Eds), Argumentation illuminated. Amsterdam: SICSAT, 178-188.
Perelman, C. & Olbrechts-Tyteca, L. (1988). Traité de l’argumentation. La nouvelle rhétorique. Bruxelles: éditions de l’université de Bruxelles.
Plantin, C. (1996). L’argumentation. Paris: Seuil (Mémo).
Plantin, C. (2005). L’argumentation. Paris: Presses Universitaires de France (Que Sais-je ?).
Schellens, J. (1985). Redelijke Argumenten. Utrecht.
Snoeck-Henkemans (2003). Indicators of Analogy Argumentation. In: F. H. van Emeren, J. A. Blair, C. Willard & A. F. Snoeck-Henkemans (Eds), Proceedings of the Fifth Conference of the International Society for the Study of Argumentation. Amsterdam: SICSAT, 969-973.