Johnstone's View of Rhetorical and Dialectical Argument DOUGLAS WALTON Univers ity of Winnipeg Abstract: In the writings of Henry W. Johnstone, Jr. there can be found an evolv- ing and gradually more sophisticated discus- sion of the relationship between rhetorical and dialectical argument. Johnstone's view on these matters was highly original, and at odds with the prevailing logical empiricism of the time, much like Toulmin's views on argumentation in The Uses of Argument (1958). In view of the rising importance of the issue of the relationship between rheto- ric and informal logic, Johnstone's analysis of the argumentum ad hominem, and its re- lationship to Hamblin's notion of commit- ment, is especially worth careful considera- tion. Resume: On peut trouver dans les ecrits d'Henry W. Johnstone, Jr. une discussion sur la relation entre les arguments rhetorique et dialectique qui se developpe et se complexifie graduellement. Son approche etait tres originale et contrariait l'empiricisme logique de son temps, comme I' approche de Toulmin sur I' argumentation dans son oeuvre The Uses of Argument (1958). Etant donnee I 'importance croissante accordee au debat sur la relation entre la rhetorique et la logique non formeIle, son analyse de I' argumentum ad hominem et Ie rapport de celle-ci Ii. la notion d'engagement de Hamblin meritent un examen approfondi. Keywords: metaphilosophy, commitment, fallacies, philosophical argument, analytical philosophy, valid argument, persuasion, dialogue. Henry W. Johnstone Jr., Stephen Toulmin, and Michael Scriven are three philoso- phers who stood apart from the logical empiricism characteristic of the core of the analytical philosophy movement. Each made contributions to logic and philosophy of science, but none of them adopted the view of reasoning as logical deduction (and/or induction) from empirical facts. Scriven, objecting to the deductive-nomo- logical view of scientific explanation, used both everyday and scientific examples to argue for the importance of pragmatic factors. Toulmin emphasized the impor- tance of warrants that are not (absolute) universal generalizations in his analysis of reasoning. Johnstone (1978, p. 55) argued that philosophical reasoning cannot be based purely on the facts', but is really a form of ad hominem argumentation of the same kind used in rhetoric. 2 None of these views were, or even are, acceptable to the leading advocates oflogical empiricism, the dominant philosophy of the twen- tieth century. It is only now we can see what is really special about the contribu- tions of these three philosophers. In certain ways these three philosophers accepted the viewpoint and methods of analytical philosophy. But what m*es them especially interesting is how they ©lnformalLogic Vol. 21, No.1 (2001): pp. 51-60. 52 Doug/as Walton also stood apart from the more doctrinaire adherents of that viewpoint, expressing reservations and subtleties that makes their own views even more interesting. Johnstone's views were and are especially objectionable to the leading analytical philosophers who still tend to dominate the inner citadels of theoretical philosophi- cal research. Johnstone argued that philosophical argumentation is deeply rhetori- cal in a way that is bound to be offensive to many philosophers, who see them- selves as being in the pursuit of truth and knowledge. These philosophers draw a sharp line between persuasion on the one hand (with its dangers of sophistical argumentation) as the domain of rhetoric, and truth and knowledge as the domain of philosophy.> Johnstone's view, or what I take his view to imply, is that you can't really have the one without the other. This view implies, consistently with the allied views expressed by Aristotle, Richard Whately, and Chaim Perelman and Lucie Olbrechts-Tyteca, that rhetoric is much more closely tied to logic (espe- cially the informal variety) than has been traditionally recognized. Johnstone's view turned out to be an important forerunner of the current developments in argumentation theory and informal logic. Central to Johnstone's philosophical de- velopment were his gradually more subtle discussions, over the span of his career, of the close but complex relationship between logic and rhetoric. Some attempt is made here to give the reader a grasp of the continuing usefulness and importance of this discussion. 1. Development of Johnstone's Views Johnstone tells us (1978, p. 3) that his earliest characterization of rhetoric, based on his views about philosophical argumentation, was "impetuous and unfair." His later characterizations, he continues, have been a "series of modifications and softenings" of that characterization and that his present position is "just about the opposite" of the earliest one. These remarks are revealing. They show how Johnstone's views developed in an open-minded way over the years. They also show how the evolution of his views was woven through the strained relationship, and often outright conflict, that has characterized philosophy and rhetoric as fields. Having attended many papers and meetings in both fields, I have heard negative remarks made about the other field from time to time. Philosophers have long been, and still are, highly suspicious about rhetoric and are prone to saying that rhetoric has no concern for the truth of a matter, as contrasted of course with philosophy, portrayed as a subject where truth, evidence, and rational argument are central. Those in rhetoric are quick to say that philosophy is a highly abstract subject that has given up the use of empirical data, or contact with realistic prac- tical concerns. In philosophy, these attitudes are very old, and were expressed very forcefully by ancient philosophers such as Plato, who condemned the soph- ists. The criticism was that the sophists used rhetoric deceptively, to make the weaker case seem the stronger. Their concern, according to this criticism, was for profit and not for truth. Johnstone's View of Rhetorical and Dialectical Argument 53 The advent of informal logic and argumentation theory in recent years has called for a re-assessment of the relationship between philosophy and rhetoric. Now philosophy is starting to advance beyond its obsession with deductive logic, truth-values, and the purely semantic approach to argument. There is recognition that arguments need to be evaluated in a context of use. This pragmatic approach calls for paying attention to how an argument was used for some purpose in a given context of discourse. Fallacies like straw man and ad hominem now come to be centrally important in judging philosophical arguments. The striking aspect of Johnstone's writings on metaphilosophy and rhetoric shows how deeply he early on recognized the importance of such pragmatic concerns. Indeed, he was virtu- ally alone in this and out of step with the dominant conventional wisdom of the time. His writings on meta-philosophy were seriously radical at the time even though, in other respects, he was a member of the logical establishment. He had made contributions to formal logic and co-authored a leading textbook in math- ematical logic (Anderson and Johnstone, 1962). Perhaps only someone so well respected for his work on formal logic could get some grudging acceptance for writing such unfashionable views about philosophical argumentation. Johnstone was well out of the mainstream in attaching such importance to the argumentum ad hominem, even seeing it as a reasonable type of argument that is characteristic of philosophical argumentation. Johnstone (1978, p. 134) wrote that he regarded argumentum ad hominem as "the only valid argument in philosophy." He defined this form of argument as " the criticism of a position in terms of its own presup- positions." This definition of ad hominem, along with Johnstone's statement of its central importance as a philosophical argument, also indicates why the straw man argument has to be important on his view. Philosophical argumentation, on Johnstone's view, is a kind of dialogue exchange in which the one party uses as premises what she takes to be the views expressed or implied by the other party. Hence the notion of an arguer's position is central to Johnstone's theory of philo- sophical argument. This notion of arguer's position was later shown to be funda- mental to the dialectical concept of argument developed by Hamblin (1970). Hamblin showed, through the analysis of the forms of argumentation associated with the traditional informal fallacies, that the central concept needed for the evaluation of fallacies in formal dialectical models is that of an arguer's commitment (or "posi- tion," as it could also be called). The following schematic diagram shows how Hamblin's notion of commitment can be deployed in explicating Johnstone's theory of the logic of the argumentum ad hominem. Let's say that Philosopher X attacks Philosopher Y using an ad hominem argu- ment. In a typical case of this sort, Philosopher Y is not making a direct (abusive) attack on the personal character of Philosopher Y (although that is one type of ad hominem argument). Nor is Philosopher Y attacking Philosopher X's argument by citing empirical evidence, or arguing that X's argument lacks support by empirical evidence. Instead, in typical philosophical disputation, Philosopher Y is attacking the philosophical argumentation of Philosopher X. Philosopher Y is extracting the 54 Douglas Walton Philosopher X Philosopher Y X's argument evidence from the previously expressed commitments of Philosopher X, using this evidence to argue that Philosopher X's argument is deficient, and open to criticism. Johnstone's insight was to recognize that the pattern of argumentation modeled in the schematic diagram above was not only typical of philosophical argumentation. But also, the radical aspect of his theory was his recognition that this pattern of argumentation can actually be classified as fitting the form of the argumentum ad hominem. 2. Commitment and Ad Hominem Johnstone understood the importance of the notion of commitment in argumenta- tion, and expressly analyzed the notion. Johnstone (1959, p. 124) used the illustra- tion of parenthood to explain commitment. According to Johnstone, parental obli- gations to a child arise from what can be called the parent's commitment. But according to Johnstone (p. 124), this same notion of commitment transfers to the kind of case in which a valid ad hominem argument is addressed to a philosophical ;;trguer. What Johnstone apparently meant by drawing this comparison is that the ad hominem argument of the kind used so commonly in philosophical argumenta- tion is essentially a commitment-based type of argument. It is based on what are taken to be (rightly or wrongly) the expressed or implied commitments of an arguer. How are these commitments determined in a given case? The answer is that you have to look at what the arguer said. If there is a record of the text of discourse, as there often is in cases of philosophical argumentation, you have to examine the actual wording of the text carefully. To judge what the arguer's com- mitments are, you not only have to use valid logical reasoning by drawing infer- ences, you also have to look at what was said, and judge what that discourse can be taken to imply. This view of commitment-based argumentation is now com- mon in argumentation and informal logic. But when Johnstone wrote about it in 1959, it was far from common. It was quite an unusual and radical view that is Johnstone's View of Rhetorical and Dialectical Argument 55 clearly an important predecessor of the notion of commitment later developed by Hamblin. Johnstone's expressed thesis (\959, p. 76) was that all valid philosophical arguments are ad hominem. This thesis, of course, is provocative because the conventional view of the time was that the argumentum ad hominem is a fallacy. What seems to be the implied conclusion is that all valid philosophical arguments are fallacious. However, when you analyze what Johnstone meant by argumentum ad hominem, it actually comes out as meaning what has now been called argument from commitment. This form of argument can be illustrated by the following kind of case. Bob tries to persuade Mary what a good position communism is, using expressions like "Power to the people!" Bob then argues that the Post Office should be run by private enterprise. Mary then comments on what she sees as an inconsistency in Bob's argument. She remarks, "Well, Bob, I take it from your arguments that you are committed to communism. Though, wouldn't arguing that the Post Office should be run by private enterprise conflict with the normal and expected communist commitment to agencies like the Post Office being run by the state?" In Johnstone's analysis, Mary's argument would be a valid ad hominem argument against Bob. Depending on the details of the case, it could be quite a reasonable form of criticism in which Mary raises questions about the internal consistency of Bob's position. Indeed it would seem, as long as Mary doesn't push the argument too hard and as long as Bob really did say all those positive things about communism, the argument could be quite reasonable. Of course, the case is quite oversimplified. Johnstone used many more detailed cases of serious philosophical arguments to illustrate the point. But the idea is fairly straightfor- ward. Johnstone saw argumentum ad hominem as argument from commitment. As such, based on observations about real philosophical disputes, he moved to the quite reasonable thesis that all valid philosophical arguments are ad hominem. Once you understand it, the thesis makes perfectly good sense, and it can be supported by good evidence. Yet it seemed counterintuitive and even shocking at the time. The implications of Johnstone's shocking thesis about the ad hominem are sweeping. For one thing, it suggests quite a different view of philosophical argu- mentation as commitment-based. It also suggests that arguments formerly thought to be fallacious, like ad hominem and straw man, are fundamental to all philosophi- cal argumentation as reasonable arguments. Of course, it is not being ruled out that such arguments can sometimes be fallacious. What was new are the views that they are not always fallacious, and that inmany common cases of philosophi- cal argumentation, they are quite reasonable and appropriate. 3. Rational Persuasion Another thesis that Johnstone argued for, one that seemed radical at the time and still does to many, is that philosophical argument is centrally based on what can be classified as a kind of rational persuasion. Rational persuasion involves two par- 56 Douglas Walton ties, the persuader and the party to be persuaded. So the notion of philosophical argument as being based on rational persuasion is dialectical. Argument is seen as a relational process between two parties. This dialectical view of argument is opposed to the monolectical view of argument so typically expressed in typical accounts in logic textbooks. On this account, an argument is a set of propositions, and no essential use is made ofthe conclusion being a "claim" made by one party or doubted by a second party. On this persuasion view, philosophical argument is commitment-based. Rational persuasion takes place when the first party produces a valid argument for her conclusion based only on premises that are commitments of the second party. This view of argument as rational persuasion is now com- monplace in argumentation theory and informal logic. But when Johnstone advo- cated it, the view was very definitely quite unusual. It would have been quite unsettling to mainstream analytical philosophers of the time, and especially to logi- cal positivists, who saw themselves as engaged in using logic to objectively pursue the truth. The idea that they were doing some kind of "persuading" would have appeared quite horrible to them. Persuading was equated with rhetoric, sophists, and generally with a "subjective" mode of argument that was not in tune with the logical point of view. Once it is seen how central the notion of rational persuasion was for Johnstone's view of philosophical argument, a number of things come out much more clearly about his central philosophical position. For one thing, it becomes evident how important rhetoric was for him, and how closely rhetoric and philosophy were tied together. On the view of philosophical argument as dialectical, and as centrally involving rational persuasion attempts, the notion of a philosopher using reasoned and reasonable rhetorical argumentation to advocate her position comes to the fore. The following thesis comes to seem plausible. There is nothing wrong with a philosopher using rhetoric, in the form of argumentation meant to be persuasive, in order to support a claim that other philosophers may disagree with. However, even though such argumentation can be described as rhetorical, it should also be expected to meet criteria of reasonableness. For example, it should be open-minded. It should be open to questioning, and it should be expected to respond appropri- ately to relevant questions and legitimate criticisms. Johnstone even proposed such a criterion in a paper on the ethics of rhetoric. What Johnstone (1981, p. 310) called the basic imperative of ethical rhetoric is "So act in each instance as to encourage, rather than suppress, the capacity to persuade and to be persuaded, whether the capacity in question is yours or another's." This principle foreshad- owed later normative constraints on persuasion dialogue like the rules for critical discussion set out by van Eemeren and Grootendorst (1987). Johnstone saw the basic imperative as implying other obligations on a rational arguer. For example, he cited (1981, p. 310) a duty of openness in argumentation. He also saw the basic imperative as being incompatible with the use of appeals to force or threats in place of rational persuasion (p. 311). In formulating these principles of rational persuasion, Johnstone was an important precursor not only of rules for critical Johnstone's View of Rhetorical and Dialectical Argument 57 discussion. He had also articulated how such rules of rational persuasion dialogue have normative function in the analysis and evaluation of arguments associated with the traditional informal fallacies. 4. Rhetoric and Dialectic Johnstone's work provokes one to think more deeply about a subject that has recently come into prominence through the advent of recent work on argumenta- tion. This subject is the relationship between rhetoric and dialectic. Dialectic is an ancient subject. It was well known tothe Greek philosophers. Plato and Aristotle often discuss the relationship between rhetoric and dialectic, for example. But dialectic fell out of favor (or even notice) after the fall ofthe ancient world. Nowa- days many (at least those who are not insiders to argumentation theory) associate it with Hegel or Marx. However, in argumentation theory, the new dialectic is the new pragmatic framework for analyzing and evaluating an argument in a commu- nicative context of use. The new dialectic has arisen out of Grice's conversational principles (Grice, 1975), and out of Hamblin's (1970) systems of formal dialectic, used to analyze logical fallacies. Dialectic is a normative subject because its use is to provide an abstract standard or model (often a formal model) that exhibits how an argument (or other speech act) should properly be used according to the guide- lines and definitions in the model. Dialectic is really a branch oflogic. It represents applied or practical logic of a kind that can be used to evaluate how an argument should be used for some communicative purpose in a context of discourse. Many might feel that the notion of dialectic is antiquated, and is a mere philosopher's artifact. But that perception is changing, as the new field of computational dialec- tics become established within computer science. 4 The advent of this new field in computing has legitimized and justified dialectic as a serious scientific field with many useful applications. Once dialectic is recognized as an important field of knowledge and technol- ogy in its own right, and as a new science of applied logic and pragmatics, the question is posed, How does dialectic relate to rhetoric? Both subjects have to do with persuasion. Both subjects take argument as their central concern. Both sub- jects analyze and evaluate argumentation in a context of discourse and view an argument as an interpersonal relationship between two parties. Typically in both subjects, the basic framework of argument use is the same. You have one party with a thesis or viewpoint, and you have another party (or audience) who does not, at least initially accept that thesis or viewpoint.The concern is with the at- tempts of the first party to get the other party to come to accept that thesis or viewpoint by means of using arguments. These elements account for the similari- ties and common ground. What are the perceived differences? The main difference appears to be that dialectic is a branch of logic. Dialectic rules on whether an argument in a given case should be judged to be logically correct or not. A central concern of dialectic 58 Douglas Walton is with fallacies. A fallacy is an argument (or move made in argumentation) that appears to be correct, or has some element of deception, but is actually incorrect. One of the main jobs of dialectic is to identify, analyze and evaluate fallacies. Dialectic, in a word, is normative. It judges whether arguments are right or wrong according to standards of how an argument should be put forward or responded to. Rhetoric, in contrast, appears to be a practical and empirical subject. The aim of rhetoric is centrally to give a speaker advice on how to put forward an argu- ment most convincingly or persuasively, to get your audience to do what you want, or to get them to accept your viewpoint. I write this hesitantly, because my field is logic, not rhetoric, and I feel it is really up to those in the field of rhetoric to define their field, and to agree on what its central aims should be. For centuries leading theorists in both philosophy and rhetoric have argued about the relationship between logic and rhetoric. As mentioned above, the issue was much discussed in ancient Greece and Rome. It would be unwise to be dogmatic about it, and it is best to write about "perceived differences" rather than to claim to have the final word. Johnstone's discussion of the issue is so useful precisely because he doesn't oversimplify. It is a question he asked over and over again during the course of his carreer. He gradually peeled off new layers of subtlety surrounding the issue. Many of us in the field of logic begin with a particular view of rhetoric. This view is to sharply draw a hard and fast line between rhetoric and dialectic. Dialectic is normative. It deals with how an argument should (ideally or correctly) be put forward or responded to. Rhetoric is psychological, descriptive and practical. It tells you how to put forward the most psychologically persuasive argument, never mind whether the argument is logically correct. This simplistic view of the rela- tionship between rhetoric and dialectic is refuted by Johnstone (1981). This paper, unusual for someone with a background in logic, argued that there are normative principles of rhetorical discourse. These normative principles, according to Johnstone's theory, prescribe duties for both speaker and hearer. The central duty called the basic imperative (quoted above), requires that a participant in rhetorical discourse should take a stand and not just capitulate to the rhetorical arguments of others. Another important duty cited by Johnstone (see above) is that of openness. I ought to be open to being persuaded, and I ought not to use violence in place of responding to persuasion by persuasion. These principles have normative status, and they do relate to traditional logical fallacies. For example, the principle not to use violence relates to the argumentum ad baculum or fallacy of appeal to force. Johnstone's set of ethical principles for rhetorical discourse shows that rheto- ric should not be seen as a purely psychological, as opposed to a normative, subject. What is shown is that given cases of rhetorical discourse can be evaluated normatively according to ethical principles. In some cases, argumentation in rhe- torical discourse could be rightly classified as propaganda, for example, on the grounds that the argumentation fails the normative requirement of being open in Johnstone's framework. Normatively, the argumentation could fail to meet the Johnstone's View of Rhetorical and Dialectical Argument 59 duties set down by Johnstone's ethical requirements for rhetorical argumentation. By using such normative standards, argumentation in particular cases can be analyzed. It could be graded, so to speak. Some rhetorical argument could be judged to be better, or more successful than others, in meeting ethical standards of discourse. For example, a newspaper editorial could be criticized for being persua- sive in a way that is not open. Of course, these judgments are normative. How- ever, they represent a standard by which given cases of the use of rhetoric can be evaluated from a critical perspective. An argument that is heavily biased to one side, that refuses to fairly consider opposed arguments, or that uses threats or sanctions in place of arguments based on the views put forward by the other side, can be criticized as falling short. Some instances of rhetorical argumentation, according to Johnstone's view, can be said to be better than others. Johnstone's view makes the relationship between rhetoric and dialectic more complicated than the simplistic view we began with above. Normative evaluation, on his view, is no longer confined exclusively to dialectic. Rhetoric, following Johnstone's view, should have its own normative standards. This view is not the final word on an issue that, by all indications, will be fought out even more inten- sively in the future than it was in the past. But it is a mature view that yields deep insight to the complex, controversial and often troubled relationship between dia- lectic and rhetoric. Notes I According to the so-called "verification principle" of the logical empiricists, a statement is only meaningful ifit is verifiable. According to at least a narrow interpretation of this principle, the only kinds of statements that are verifiable are tautologies (logical truths) and empirical state- ments. 2 In Johnstone (1978, p. 45), the statement was made. "My position is that all philosophical arguments are ad hominem." ) Johnstone (1978, p. 62) noted that the view that the philosophy does not have the function of persuasion is "characteristic of at least the extreme form of positivism. 4 The first workshop on computational dialectics was held in Scattle in 1994 and the second one was held in 1996 in Bonn, Germany. The aim ofthe third meeting, hcld in Berlin in 2000, was to discuss inter-agent argumentation and negotiation protocols. as well a'l approaches of computa- tional dialectics that model group decision making processes. The website for the Third Work- shop on Computational Dialectics is . References Anderson, John M. and Henry W. Johnstone, Jr. Natural Deduction: The Logical Basis of Axiom Systems. Belmont, CA: Wadsworth, 1962. Eemeren, Frans H. van and Rob Grootendorst. "Fallacies in Pragma-Dialectical Perspec- tive." Argumentation I, 1987: 283-30 I. 60 Douglas Walton Grice, H. Paul. "Logic and Conversation," in Syntax and Semantics, Vol. 3. Peter Cole and Jerry L. Morgan, eds. New York: Academic Press, 1975. Hamblin, Charles. Fallacies. London: Methuen, 1970. Johnstone, Henry W., Jr. Philosophy and Argument. University Park, PA: The Pennsyl- vania State University Press, 1959. -. Validity and Rhetoric in Philosophical Argument. University Park, PA: The Dia- logue Press of Man and World, 1978. -. "Toward an Ethics of Rhetoric." Communication 6,1981: 305-314. Douglas Walton Department o/Philosophy, University o/Winnipeg Winnipeg, Manitoba, Canada R3B 2£9 d. walton@lIwinnipeg.ca