Argumentation Schemes for Presumptive Reasoning

Douglas Walton

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781136687051
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Communication Studies
• Index: Languages & Linguistics

Recent concerns with the evaluation of argumentation in informal logic and speech communication center around nondemonstrative arguments that lead to tentative or defeasible conclusions based on a balance of considerations. Such arguments do not appear to have structures of the kind traditionally identified with deductive and inductive reasoning, but are extremely common and are often called "plausible" or "presumptive, " meaning that they are only provisionally acceptable even when they are correct. How is one to judge, by some clearly defined standard, whether such arguments are correct or not in a given instance? The answer lies in what are called argumentation schemes -- forms of argument (structures of inference) that enable one to identify and evaluate common types of argumentation in everyday discourse. This book identifies 25 argumentation schemes for presumptive reasoning and matches a set of critical questions to each. These two elements -- the scheme and the questions -- are then used to evaluate a given argument in a particular case in relation to a context of dialogue in which the argument occurred. In recent writings on argumentation, there is a good deal of stress placed on how important argumentation schemes are in any attempt to evaluate common arguments in everyday reasoning as correct or fallacious, acceptable or questionable. However, the problem is that the literature thus far has not produced a precise and user-friendly enough analysis of the structures of the argumentation schemes themselves, nor have any of the documented accounts been as helpful, accessible, or systematic as they could be, especially in relation to presumptive reasoning. This book solves the problem by presenting the most common presumptive schemes in an orderly and clear way that makes them explicit and useful as precisely defined structures. As such, it will be an indispensable tool for researchers, students, and teachers in the areas of critical thinking, argumentation, speech communication, informal logic, and discourse analysis.

CHAPTER ONE

INTRODUCTION

Recent work in informal logic and argumentation has come to rely more and more on the idea that certain common forms of argumentation–like argument from precedent, argument from authority, argument from analogy, and so forth¹–are, in some instances, “valid” or correct modes of reasoning. If so, they must have structures or “forms.” But what are these forms of argument, or so-called argumentation schemes? That is the question addressed in this book.

Following the usual methods of logic, we would expect to find logical calculi, systems of propositional calculus, or the probability calculus, that would model these types of reasoning as valid or invalid. However, we will argue that this approach, at least by itself, is not the answer.² Instead, we hope to show, these argumentation schemes can best be revealed as normatively binding kinds of reasoning when seen as moves, or speech acts in the setting of dialogue. In this pragmatic framework, two participants are reasoning together in a goal-directed, interactive, conventionalized framework called a dialogue. An argument is evaluated as good (correct, reasonable) to the extent that it contributes to the goal of the dialogue. An argument is evaluated as bad (incorrect, fallacious) to the extent that it blocks the goals of the dialogue.

According to this type of analysis, each of the types of argumentation modelled will have a distinctive argumentation scheme (structure, form) that allows it to function as a way of making a point or shifting a burden of proof in a dialogue.

NEED FOR A SYSTEMATIC STUDY

The reliance on argumentation schemes as a central component in the normative structure of argument evaluation and analysis by theorists like van Eemeren and Grootendorst (1984; 1992) has acutely raised the need for a more systematic study of these schemes. In particular, what is vital is to see whether or how these schemes can function–like the forms of argument in deductive logic–to test or verify that instances of them in natural language argumentation are in some sense correct or reasonable. Although the term valid does not seem to be quite the right word to use with many of these argumentation schemes, still, when they are rightly or appropriately used, it appears that they are meeting some kind of standard of correctness of use. What is important to come to know is what this standard is, for the most common and widely recognized schemes especially, and how each of the schemes can be tested against this standard.

Blair (1991) put his finger on the central problem when he cited van Eemeren and Kruiger (1987) as identifying several different types of arguments like argument from analogy, causal arguments of various kinds, arguments from rules and principles, argument from consequences, and argument from precedent. But, Blair pointedly asked, what does it mean to say of such an argument that its premises are sufficient for its conclusion? According to the pragma-dialectical theory of van Eemeren and Grootendorst, Blair noted, “sufficiency is a function of appropriately meeting the critics’ challenges to premises and inferences” (p. 332). Blair also noted that this means that an argument can rightly be said to be sufficient for its conclusion in this sense when it meets its burden of proof³ relying on “what may be presumed without or accepted without further question” (p. 333). This answer seems exactly right, and points us in a promising direction. For many of the most common and basic types of argumentation schemes that require a new kind of analysis are inherently presumptive and dialectical in just the way Blair describes.

But as Blair noted, van Eemeren and Kruiger (1987) only give a few examples of these argumentation schemes, and clearly work is needed to more sharply define these types of argumentation before we can see better how they can be used as an aid to determine when an argument (of one of these types) is sufficient for its conclusion.

When we look to the literature on argumentation schemes, there is very little to be found. Quite a few argumentation schemes can be found in Perelman and Olbrechts-Tyteca (1969), but they are woven in with the general themes of the book. The single, outstanding best source is the doctoral dissertation of Hastings (1963), which presents nine “modes of reasoning” along with other “patterns of reasoning.” Each mode of reasoning represents what we call an argumentation scheme. Hastings presents an illustrative example of each from everyday argumentation, and supplies a set of critical questions to match each scheme. Thus, Hastings (1963) is a uniquely valuable source that in effect founded the systematic study of argumentation schemes as a subject for further development.

Aside from this unique resource, however, there is very little to go on. Kienpointner (1987) gave a useful historical and analytical overview of the subject. The ancient and medieval classifications of the different “topics” identify many of the same types of argumentation that we associate with argumentation schemes. However, their way of presenting these schemes, by mixing them with logical doctrines like essentialism, that are now outdated and unfamiliar to modern readers, detracts from the usefulness these accounts might have.

Other than the occasional paper on the subject, for example by van Eemeren and Kruiger (1987), and scattered remarks by van Eemeren and Grootendorst (1984; 1992), there is really nothing else to turn to in attempting to get a grasp of what argumentation schemes are, or how they can be applied.

Just recently however, a new book has appeared which gives a comprehensive typology of argumentation schemes, and identifies sixty of these schemes.

The typology of Kienpointner (1992) is made on the basis of the kind of warrant (Schlussregel) used in an argument to provide the transition or inference from the premise(s) to the conclusion. The typology (Kienpointner, 1992) has three main categories of argumentation schemes: (I) those where the warrant is one of use (Schlussregelbenützende Argumentation schemata), (II) inductive argumentation schemes, and (III) a third category comprising argument from example, analogy argument, and argument from authority. The first category is quite large, and comprises four subcategories: 1. classifying schemes–including definition, genus-species and part-whole arguments. 2. identity (Gleichheit)–including arguments based on similarity and difference. 3. against the proposition (Gegensatz) schemes–these are arguments used negatively for refutation (what Perelman and Olbrechts-Tyteca, 1969, called schemes of dissociation). 4. causal schemes.

Kienpointner’s typology turns out to be quite useful to help us gain a perspective on the variety of argumentation schemes studied in this book. We have also adopted the policy of identifying and classifying the argumentation schemes on the basis of the warrant or rule (we think of it as expressing a type of generalization) that links the premise(s) of an argument to the conclusion. However, the scope of our inquiry is more narrow than that of Kienpointner. We analyze only what we call presumptive argumentation schemes, (i.e., ones where the generalization is presumptive in nature, and meant to admit of exceptions). Therefore, we do not include, for example, inductive arguments, part-whole arguments, or genus-species arguments, presuming that (by and large, at any rate) these types of argumentation are not presumptive in nature.

A number of the argumentation schemes we study in the subsequent chapters do, however, fit nicely into Kienpointner’s typology. A number of them (especially the ad hominem variants in chapter 3) fit into the Gegensatz category. Several others are clearly causal in nature. Arguments based on example, authority, and analogy also fit into Kienpointner’s typology. However, we classify arguments from authority, or at least the subspecies we call the argument from expert opinion, as a species of argument from position to know (see chapter 3).

Thus there are some points of difference between Kienpointner’s way of classifying argumentation schemes and our own. But the main difference is that we are concentrating on what we call presumptive reasoning, and that makes the scope of our inquiry, and also, to some extent the basis on which it is made, different from Kienpointner’s. Certainly the problem remains of understanding how many of the most common of these argumentation schemes in everyday conversation are inherently different from the usual models of deductive and inductive reasoning–these are the presumptive argumentation schemes.

Perhaps part of the problem is that this kind of project never appealed to logicians very much because it seemed to have more to do with rhetoric and the discovery of persuasive arguments, for example, in making a speech. If so, it seemed questionable whether it was appropriate to try to see presumptive arguments as, in some sense, logically valid. More often, they have been portrayed as fallacious. This apparent ambiguity has dogged the subject from its beginnings.

ARISTOTLE’S TOPICS

The history of the systematic study of argumentation schemes begins with Aristotle’s (1939) Topics. The purpose of this treatise was to apprehend what Aristotle called dialectical reasoning, in which two parties reason together on the basis of premises that are endoxa, usually translated as “generally accepted opinions” (Topica 100M8). Dialectical reasoning contrasts with demonstration, a kind of reasoning which “proceeds from premises which are true and primary,” for example, the first principles of science (100a28).

Aristotle also recognized a third kind of reasoning he called contentious (or eristic), which is “based on opinions which appear to be generally accepted but are not really so” (Topica 100b26).

In the Topics, Aristotle was taking a point of view on argumentation that subsequently came to be pushed to one side and forgotten. He saw reasoning as being used in different ways in different types of dialogue. And he set out to study and evaluate the uses of different types of argumentation as used within the norms or conventions of these different types of dialogue.

Such an approach has long seemed unfamiliar and even alien, and especially so in the 20th century, where logic, and reasoning generally, is seen exclusively from a semantic and formalistic point of view. Modern symbolic logic, a development of Aristotle’s syllogistic (deductive) logic, came to abstract reasoning from its uses in a context of dialogue entirely. According to the formalistic viewpoint that became very dominant in the positivistic era of the first half of the 20th century, an “argument” came to be seen as a set of propositions with “truth-values” attached to them. The idea of the use of an argument in different types of conventionalized dialogues, or structures of interactive reasoning, dropped out of the picture. The idea was reverted to in Wittgenstein’s “language-games” in his later philosophy, but has not begun to be investigated in any systematic way until the recent advent of argumentation as a field.

For these reasons, it is difficult for a 20th-century reader, especially one familiar with formal logic in some form, to have much of a grasp of what Aristotle was up to when he wrote about dialectical argumentation. The ideas in the Prior Analytics are familiar, because Aristotle was presenting a theory of objective validity, concerned with universal and absolute propositions, for example, of the form “All F are G,” where the individuals that have property F all have property G (without considering exceptions). However in the Topics, Aristotle was presenting a theory of dialectical argumentation, in which an argument is evaluated relative to another person (Topics 155b10).

However, although dialectical argumentation, by its nature, involves two people reasoning together, Aristotle was not attempting to consider any particular pair of individuals, or the varying reactions of all individuals. Instead, according to Evans (1977), he is “contrasting the art which pays attention to the views of each individual, however eccentric these views be, with that which organizes and selects certain views as typical and specially relevant to the subject under consideration” (p. 76). Thus Aristotle’s position is not an extreme relativism, but a qualified relativism that judges the worth of an argument in relation to an endoxon, which could perhaps be called a plausible or typical point of view on an issue where opinions differ. An argument is judged good or not, in relation to a point of view.

As Evans (1977) pointed out, the translation of the word endoxon presents difficulties. Evans did not think that it represents probability of any sort, and can instead be taken, in a basic sense, as meaning “representative of someone’s view” (p. 78). Aristotle also makes a distinction between a qualified and an unqualified endoxon (Topica 159a39 - 159bl). Something endoxon “may be so either without qualification or in a way which is qualified by reference to some person (Evans, 1977, p. 80). Evans came to the view that the unqualified endoxon could be translated as “absolute plausibility, unqualified by reference to this person” (p. 83), whereas the qualified endoxon refers to something that is plausible to a particular person, or participant in dialectical argumentation. The plausible, in other words, can take an absolute or relative form.

Those schooled in traditional (formal) logic might have difficulty in seeing how this notion of plausibility represents or relates to some logical, as opposed to rhetorical method of evaluating reasoning. But evidently, Aristotle thought that it does. According to De Pater (1968, p. 188), the Aristotelian topic or topos has two functions. In its selective function, a topic is a device to find arguments within a possible set of types of arguments. Using this function, an orator can run over the list of topics, representing different kinds of argumentation, and pick out the best or most appropriate one to make his case on a given issue. This is more of a rhetorical or inventive function of a topic. Van Eemeren, Grootendorst and Kruiger (1987) called this function a “tactical aid” or “search formula,” in setting up an argument (p. 65).

In its guarantee function, the topic is a kind of inference link that grants “the plausibility of the step from arguments to controversial claim” (Kienpointner, 1987, p. 280). The guarantee function uses the argumentation scheme as a bridging structure of inference or Toulmin warrant that connects a set of premises to a conclusion. Used as a guarantee function, an argumentation scheme works in much the same way a deductive form of inference like modus ponens does, in sanctioning a form of argumentation as valid or i...