many texts now on the market, thinking about points and problems as you teach, creating innovative ways to teach reasoning skills and the love of reason. We, and your colleagues, would like to hear from you. Has our talk about moving in the direction of a journal scared you off? Perhaps our own conviction that there are lots of theoretical issues that deserve article- length treatment has led us to feature these items in a way that discourages brief and informal pieces, the airing of tentative proposals, the sharing of muddles and doubts and queries, the description of modest or untested innova- tions. If so, then we say to hell with journal talk. We'll drop it. If a journal is by definition stuffy and formal, inimical to openness and the easy sharing of ideas, the I LN will not become a journal while we are its editors. Let us simply pose this question: Do you have some half-finished paper or article or note on a subject of interest to our readers lying around? Then for everyone's sake, dust it off, revise it, complete it, and send it along to us. There are, it seems to us, just dozens of topics that require critical thought and inquiry: the whole problem of missing premises and how to supply them; the role of formal logic in the informal logic enterprise, and the relationship between the two; how best to achieve the aims of informal logic/critical thinking. We could go on to list others, but surely you have ideas of your own on these and other topics. this issue In this issue we feature a note from Doug Walton on the various models of argument now available, and a comment from Trudy Govier and Nick Griffin on Ralph Johnson's article about the principle of charity (lLN, iii.3). The bulk of the issue is the collection of examples for analysis originally intended for a special supplementary issue last Spring. We belatedly offer these examples now in the hope that they're better late than never. * 2 article What Is Logic About 1 Douglas Walton University of Winnipeg Mrs. Jones has her ear cupped to the wall adjoining the next apartment. She hears some tense, guttural pro- nouncements - a man's voice? Then she hears some higher pitched responses that a speech act theorist might describe as "aggrieved whining". The lower voice now breaks into loud staccato accusatory stabs of statements. Mrs. Jones can even make out some unprintable words. The hi!!her voice now responds with unmistakable screams, audible even to Mr. Jones, who is trying to read Maclean's magazine. The crescendo of voices is punctuated by a crash of crockery. "What's going on over there?" Jones queries his wife. Mrs. Jones replies, ''They're having an argument'" The paradigm probably most of us have of an argu- ment - at least those of us relatively uncontaminated by the study of logic - is that of a verbal interchange between or among a number of participants with (a) an adversarial or disputational flavour, and (b) heightened emotions, very often anger, being involved. Of course none of these items is absolutely essential. One can argue with oneself. One can have a friendly, or constructive argument. And one can argue unemotionally, in the style of Mr. Spock, the imperturbable Vulcan. Nonetheless, hot interpersonal dispute is among the commonest conceptions of argument. Let us call this model of argument the quarrel (more fully exposited in the work cited in note 8, Ch. 1). Accordin!! to the much more modest and sober, not to say austere, conception of argument favoured by twentieth-century logic, an argument is merely a set of propositions. This conception strips away the emotion, the interpersonal element, and even the adversarial notion of disputation. By this conception, an argument can even be some chalk marks on a blackboard or ink-marks on a page, according to some of the most determined exponents of austerity, at any rate. If we define logic to be the science of argument, which model of argument is better to start with? The first one is obviously rich in psycho-social information. Ann Landers would find lots there to be interested in. The second is very rich in mathematical results. Boole and sub- sequent generations of mathematicians have found lots there to be interested in. It is not too hard to see the fascination of each model for the critic of arguments. The second one admits of formal models that are decidable and complete. You can tell by objective tests which arguments are correct and which fall short of correctness. That is worth studying. The first one gives real-life case studies of actual arguments, refutations and fallacies. Critics have, however, pointed out limitations of each model. The first model is unstable, subjective, even unruly. Too often it seems impossible to tell who is mostly right or wrong, or even what the argu- ment is. The second model is provably correct as far as it goes, but it is questionable to what extent it applies to lively specimens of realistic argumentaion. Are we forced to choose between them? Sometimes exponents of one model will partially acknowledge the other. Gricean conversation theory argues that classical deductive logic is the right logic, but it needs to be trimmed with conversational niceties in order to approximate the do's and don't's of natural discourse. [1] On the other hand, some ~ho stress the study of real- life actual argumentation may concede that formal logic has its place. It's just that arbitrary designation of a set of propositions as argument does not go far enough. It is a legitimate - but informal - task to determine what the argument is, even before it gets processed further. But the question remains whether we have to choose between these two models of argument. Are there other alternatives? Aristotle, the founder of the subject of logic, distin- quished two models of argument, neither of which is precisely identical with either of the pair above. Aristotle defined a demonstrative argument as one in which the premisses are better known than the conclusion, so that the conclusion may be established on the basis of the premisses. This is an asymmetrical model of argument. If p is a correct argument for q, then q cannot be a correct argument for p. It is also irreflexive. The classical inference pattern "p, therefore p" cannot be correct according to the demonstra- tive model of argument. Aristotle defined a dialectical argument as one in which the premisses are presumed to be true, or thought to be true by the wise or some other source that falls short of guaranteeing that the premisses are known to be true. These facts about Aristotle are well known, but they are worth reviewing because they posit two models of argument distinct from the quarrel or the purely deductive model. In modern treatments, the first model is akin to the model of epistemic logic developed notably by Hintikka. [2] The second has been formalized in recent times by the dialectical games of Hamblin. [3] According to the dialectical model, an argument is a two or many- person game with a set of rules that defines permissible moves in orderly sequence, and a win-strategy. Each move is a proposition, indexed to a participant. These dialectical and demonstrative models of argu- ment are a nice compromise because they capture the personal element, the give-and-take of disputation, and the directionality of reasoning. But at the same time the rules are clear, and the model is amenable to decision procedures to determine correctness or failure of correctness. Kripke 3 has even given an interpretation of the intuitionistic calculus that would seem to make it a very good model of one kind of demonstrative argument. [4] A major problem is that there are many formal models of dialectical and demonstrative reasoning. So the application problem is very much with us. Which of these is most applicable to realistic argumentation where fallacies and other good or bad steps of reasoning take place? The realistic models of the quarrel, or even the discussion, or Socratic disputation, or debate cannot be left behind. Even the model of argument as a set of propositions is incorpo- rated into the dialectical and demonstrative models. If all four models of argument so far identified have a legitimate role to play in the theory of argument, do we not seem to be enmeshed in a hopeless pluralism? Not to mention the inductive-deductive pluralism discussed in recent issues of this newsletter! [5] Is there some common root to these various models? In essence, we are asking: What is logic about? I will not try to settle this question. Suffice it to say that it is my own opinion that we will only be able to work towards an answer to it by means of a more attentive study of the so~alled informal fallacies - traditional, significant sophisms of argument that provide benchmarks for the analysis of argument. Logic, argument, and fallacy - the three concepts are closely connected. But how closely? Charles Kielkopf has warned us that there may be fallacy (at least of the traditional sort, like ad baculum) without argument. [6] But perhaps more narrowly and properly construed, a fallacy should be a fallacious argument. Certainly logic is about arguments, and thereby about fallacies. Without pursuing these interconnections further, let me pose one problem about them. Mrs. Jones, ear cupped to the wall again, hears what appears to be the higher voice saying, "George, you're so inconsistent. You tell me not to back-seat drive, and then the other day you criticized my failure to signal a turn. You're always lecturing me on the foolishness of smoking, and you can't give up the habit yourself, ... " Mr. Jones looks up, "What's going on?" Mrs. Jones replies, "She just accused him by means of the circumstantial ad hominem." [7] Here we have an argument, and a very interesting one at that. George stands accused, not of logical inconsistensy, but of an action-theoretic circumstantial conflict that may, or may not, be reducible to some logical inconsistency. In a nutshell, he is accused of failing to practise what he preaches. This lapse, if not defensible, may indeed be a serious ethical failure or at least evidence of one. But despite the traditional ad hominem label, is it really a lapse of logic? Are George's arguments incorrect because of his actions? A hard question, but if the answer is to be "yes", it is equally hard to see how the argument can be eluci- dated by any of the four preceding models of argument. George may even admit that he can't give up smoking and that he is thereby circumstantially inconsistent. He may still maintain his condemnation of smoking is, in itself, sound. Is his argument good, bad, or partially both? We might say that the argument is O.K., but that George's own personal advocacy of it is questionable. In other words, according to one model of argument -an impersonal one - the argument is good. According to another model - a person-relative one - the argument can be criticized negatively. In short, we are back to a relativity of plura- listic models. [8] Just as worrisome, we are on the border- line between the logic of argument and the ethics of argu- ment. It is not entirely clear that the lapse, if there is one, is a failure of logic as opposed to a moral incorrectness of Goerge's actions. Should the circumstantial ad hominem be taken out of the logic textbooks and put into the ethics textbooks? I do not think so. Not yet anyway. If only by dint of the inertia of a tradition in which: there is some wisdom, it should not be turfed out too hastily. The concept of argu- ment is fluid and unsettled in such a way as to accommo- date questionable characters like the circumstantial ad hominem. Still, one cannot but suspect that George is being criticized more for his morals than for his logic. Notes [1] H. P. Grice, "Logic and Conversation," in The Logic of Grammar, ed. Donald Davidson and Gilbert Harman, Encino, California, Dickenson, 1975, 64-74. [2] Jaakko Hintikka, Knowledge and Belief, Ithaca, N.Y., Cornell University Press, 1962. [3] C. L. Hamblin, Fallacies, London, Methuen, 1970. [4] Saul Kripke, "Semantical Analysis of Intuition- istic Logic I," in Formal Systems and Recursive Functions ed. J. N. Crossley and M. A. E. Dummett, Amsterdam: North-Holland, 1965,92-130. [5] David Hitchcock, "Deduction, Induction and Conduction,"ILN, iii.2, 1981,7-15. [6] Charles Kielkopf, "Relevant Appeals to Force, Pity, and Popular Pieties,"ILN, ii.2, 1980, 1-5. [7] John Woods and Douglas Walton, "Ad Hominem," The Philosophical Forum, 8, 1977, 1-20. [8] John Woods and Douglas Walton, Argument: The Logic of the Fallacies, Toronto and New York McGraw-Hill-Ryerson, 1980. .sb ' 'i,e' editor's note After his review of the four models of argument - the quarrel, the set of propositions, the demonstration 4 and the dialectical interchange - Professor Walton won- ders if there is a common root, and announces that "We will only be able to work towards an answer" by studying the informal fallacies more attentively. The only support he gives for this appears to be the comment that logic, argument and fallacy are closely connected, and that, pace Charles Kielkopf (lLN, ii.21. "properly construed, a fallacy should be a fallacious argument". For one who would deny any necessary connection between fallacy and argument, this won't do; so assuredly Walton owes us fur- ther support here. But beyond that, his intriguing sug- gestion that the notion of fallacy will unlock the mystery of the concept of argument, and thence explain what logic is, merits amplification. And it is perhaps doubly deserving of development in light of a couple of recent challenges to the adequacy of our analyses of informal fallacy. One I'm thinking of is Maurice Finocchiaro's grim indictment of the handling of fallacies in textbooks: In summary, textbook accounts of fallacies are basically misconceived, partly because their concept of fallacy is internally incoherent, partly because the various alleged fallacious practices have not been shown to be fallacies, partly because their classification of fallacies is unsatis- factory, and partly because their examples are artificial. (American Philosophical Quarterly, Vol. 18, No.1, January 1981, p. 18.) The other is the chapter on informal fallacies in Karel Lambert and William Ulrich's recent text. The Nature of Argument (Macmillan, 1980), Lambert and Ulrich conclude, . .. we are suggesting that until a general charac· terization of informal fallacies can be given which enables one to tell with respect to any argument whether or not it exhibits one of the informal fallacies, knowing how to label certain paradigm cases of this or that mistake in rea· soning is not really useful for determining whether a given argument is acceptable. (p. 28.) In the face of these dissatisfactions with the development of the theory of informal fallacies, it looks as though the burden of proof shifts to Walton's side. Finally, Walton's puzzle about how to handle the circumstantial ad hominem is indeed perplexing, but how. does it bear on the issue of whether the study of the . informal fallacies is the correct route to the heart of the concept of argument, and thence the explanation of what logic is? Walton's answer to this question might be that the informal fallacies exist; they are committed here, there and everywhere, and so they are the raw material from which we must start. We must be empirical, and start our analysis from what we know to be errors in arguments. Let's look and see. The trouble with this answer is, as I've indicated, some hold that when we claim there are fallacies, we're making things up. So who is right? ~ J.A. B.