Microsoft Word - IL36.1 Blair-Copi JABrev.doc © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. Historical Note An Early Exchange on the Interpretation of Arguments in Texts Compiled by J. ANTHONY BLAIR Centre for Research in Reasoning, Argumentation and Rhetoric University of Windsor Windsor, ON Canada N9B 3P4 tblair@uwindsor.ca 1. Introduction Below is an exchange of letters that occurred in the late spring of 1981 between Professor Irving M. Copi, the author of the classic and widely used textbooks, Introduction to Logic and Symbolic Logic, and me (Tony Blair). Professor Copi died at the age of 85 in 2002. I came across the letters in 2015 while in the process of beginning, in my mid-70s, to discard some of the hundreds of files I have accumulated over the years. I suggest that the letters from Professor Copi might be documents of his- torical interest in the development of informal logic. I think readers will also share my admiration for Professor Copi’s cour- tesy, his generosity and openness towards a young assistant pro- fessor, his modesty, his intellectual curiosity, and the clarity of his informal writing. In late April of 1981 the Central Division of the American Philosophical Association held its annual meetings in Milwau- kee, Wisconsin. The meetings included a session on “the new logic course”, ‘new logic’ referring to the informal logical ap- proach to argument interpretation and evaluation that was then in its early years of development. My colleague Ralph Johnson and I attended, and Johnson gave a presentation in which he dis- cussed, among other things, the business of eliminating “clutter” when extracting its argument from a text. ‘Clutter’ was John- son’s term for the material in a text containing an argument that is extraneous to the argument—that might serve other commu- nicative functions than to express the argument, or that might consists of irrelevant ramblings, or both. Professor Copi was al- so present at that session and he had asked some questions of Johnson during the discussion of Johnson’s presentation. Blair, Copi © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 84 Later, on the plane back to Detroit, which has the major airport nearest to my home in Windsor, Ontario, I found myself seated beside Professor Copi. He had been a member of the Phi- losophy Department at the University of Michigan when I was a graduate student there in the early 1960s, but by 1981 he had moved to the University of Hawaii. He was flying to Detroit to visit one of his children, who then lived in Ann Arbor, Michi- gan, which is not far from the Detroit airport. We struck up a conversation about the new logic session at the APA, and he had some things to say about Johnson’s paper that he wanted me to convey to Johnson. (For some reason, Johnson was not on the same plane, or else was not seated with us.) He also asked me to send him some details about some textbooks that came up in our conversation. On April 28 I sent him the information about the books he’d asked about, and on May 12 he wrote back the first letter (Copi to Blair-1), hand-written, that appears below. I responded with (Blair to Copi), a long typed letter, on June 12, the second letter appearing below. And Professor Copi responded to my June 12 letter with (Copi to Blair-2), a typed letter dated June 29, the third letter below. 2. The exchange 2.1 Copi to Blair-1 University of Hawaii at Manoa Department of Philosophy 2530 Dole Street Honolulu, Hawaii 96822 � Cable Address UNIHAW May 12, 1981 Dear Professor Blair, Thank you for sending me the publishers’ addresses that I requested. I have ordered the three books from the publishers, since I am not happy with the service offered by our University bookstore. I too enjoyed our conversation on the plane. I hope you made clear to Ralph Johnson that I was asking questions about his paper rather than making criticisms of it. I have the same perplexity about informal logic that I have about several other things. This recurrent perplexity takes the old form: you can’t do A until you have done B, and it’s terribly hard to do B until after A has been done. This is pervasive, but applies especially to Johnson’s first desideratum: that of elimi- An Early Exchange on the Interpretation of Arguments © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 85 nating extraneous material in analyzing arguments (getting rid of clutter). For example, consider: This river is so beautiful. No sane person would want to build a dam here. There’s little beauty left in the world. The only good reason for building dams is to prevent floods. Some congressmen want to build a dam here. So some congressmen are insane. There’s an argument here—indeed a slightly concealed syllo- gism that in one of its (re)formulations is AII-1. You and I know syllogistic logic, so we can discard the 1st, 3rd, and 4th sentences as clutter. And we can convert the 2nd sentence & then obvert the result to give us an explicit major premiss to go with the 5th and 6th sentences to constitute the syllogism already mentioned. But somebody who didn’t know syllogistic logic could not sail so smoothly along and might throw out the 2nd sentence as clutter. In a longer passage the situation would be aggravated. My point is simply that unless one already knows some (more or less) formal logic, one will have a very hard time elim- inating clutter. And when one is able to eliminate clutter, one will be able to analyze arguments along the lines urged by Scriven, Thomas, et al. And if one is unable to eliminate clutter from “ordinary language arguments” one can work only with the stilted and skeletal arguments that inhabit old fashioned logic textbooks. Of course logical intuition can be used rather than prior logical knowledge to evict clutter. But if one has that stronger “{logical intuition}” one need not study logic! I may well have said all of this to you on the plane. But it has been much on my mind lately. I did enjoy attending the New Logic Course session in Milwaukee and even learned something there. I cannot say as much for other sessions I have attended over the years! With every good wish Irving Copi Blair, Copi © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 86 2.2 Blair to Copi 12 June 1981 Professor Irving M. Copi [Address] Dear Professor Copi, Thank you for your letter of May 12th. I have been thinking about what you wrote, and I would be interested in your reac- tions to some comments about your remarks. (1) About the need for some formal logic training (to de- velop deductive logical intuitions) in order to be able to elimi- nate extraneous material: a few comments. I was struck by your example and the way you proposed handling it. 1. This river is so beautiful. 2. No sane person would want to build a dam here. 3. There’s little beauty left in the world. 4. The only good reason for building dams is to prevent floods. 5. Some congressmen want to build a dam here. So 6. some congressmen are insane. You assert that our knowledge of deductive logic enables us to see right away that “2 and 5, therefore 6” is {a} valid syllogism, so we can discard 1, 3 and 4 as extraneous. I accept your first statement here, but I don’t think the inference you draw is war- ranted by it. I can perhaps best explain why by asking you to imagine the entire text of the example actually uttered by some- one in order to make a point. Presumably he is trying to con- vince us that the congressmen who want to build a dam on the river are mistaken (given everything he says, it’s doubtful he wants to convince us that they are literally insane). [The use of “so”, together with the presence of the deductively valid syllo- gism (2, 5: 6), are pretty decisive evidence of the intention to argue to some such conclusion (assuming the absence of contex- tual factors that would override this presumption).] Our arguer makes five other statements in a way that leads us to treat them as constituting some sort of unit—we might call it, “What He Has to Say About Some Congressmen’s Desire To Build a Dam on the River”. Our task is to extract from that unit (I’ll call it his “text”) the assertions belonging to his argument (or, since we have already identified his conclusion, we might say, “belonging to the support for his conclusion”), and set aside the assertions that are extraneous to the argument, or clutter. Actually, before we do that we have to decide whether we are interested in discerning our friend’s argument (the one he is An Early Exchange on the Interpretation of Arguments © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 87 offering, i.e., the conclusion he intends to defend and the state- ments he intends to be taken as support for it)—or whether we are interested in discerning some other argument in the text, not necessarily the same as his argument. (Examples of other possi- ble arguments: (a) the logically best argument using what he says and adding missing premises as required to tighten the case, given his stated premises, (b) the argument most likely to per- suade some specified audience, selecting from the statements he has made. There are other possibilities besides these two.] What our objective is will make a difference to what we count as clut- ter. (A) Suppose we are after our friend’s argument. We see (with our formal logic training) the valid syllogism (2,5:6), and since we’ve already identified 6 as his conclusion (only partly on the basis of the presence of the syllogism), we decide that 2 and 5 belong to his argument. 1, 3 and 4 don’t fit into the deduc- tive argument but he did assert them as part of the text in which 2,5:6 occurs: 1 immediately preceded 2, and 3 and 4 were as- serted between 2 and 5. This is some evidence that he intended 1, 3 and 4 to be part of his argument too. Furthermore, we can see how a reasonable person would see preserving the beauty of a river as a reason for not wanting to build a dam on it (especial- ly if there is little beauty left in the world and granting the as- sumption that the dam will ruin the beauty of the river). Also, we can see that flood-prevention is one good reason for building dams and we can understand how someone who has never heard of dams for other purposes (e.g., for irrigation, hydroelectricity, water supply) but had been exposed to many dams built for flood prevention might think that the only good reason for build- ing dams is to prevent floods; and if it happens to be true of the proposed dam that our friend opposes that it will serve no signif- icant flood-prevention function, then we can see how he might understandably take 4 to be a reason for opposing this dam. An alternative way to see how our friend wanted 4 included in his argument is this: he knows some dams are built to prevent floods, and he knows that a dam on this river would serve no flood-prevention functions, and he is eager to persuade us that the congressmen are wrong to want to put a dam on it, so he in- cludes 4 to try to sway us. The fact is that he did assert 4 in the context of defending 6, so there is a presumption in favour of taking 4 to be intended as part of the argument: given these two plausible interpretations of how 4 could be taken to be intended as part of his argument, it seems to me legitimate to count it in. If we include 4, then we can also see a way of reading 1 and 3 into his argument. If the dam would serve no useful purpose, and if it would destroy a beautiful river, in a world with little Blair, Copi © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 88 beauty left in it, then there is some reason to think that anyone who wanted to build this dam is mistaken or wrong-headed—if not literally insane. Taking these considerations into account, I can see inter- preting our friend as intending 1, 2 and 4 as support for 2, which then, together with 5 entails 6. (To get this interpretation it is necessary to add in the unexpressed premise, 4′, that building this dam will not prevent floods, but that move is not objection- able in principle: we often have to add unexpressed premises to get valid syllogisms.) I’ve perhaps gone into unnecessary detail, but my point is this: using deductive validity as the measure of what belongs to this person’s argument would lead to excluding from it premises that he probably intended to be taken as part of it. (At least, he can be fairly interpreted as having had this intention.) So the constraint of deductive validity is too strong, in such cases. Knowing AII-1 certainly helps us here, for we built our interpre- tation around 2,5:6. But it does not serve the purpose you pro- posed this example to illustrate—namely to tell us what in the text is clutter and what in the text belongs to the argument. (B) Suppose we are interested not in what our friend may have intended to argue, but whether his conclusion is true—and so whether what he says gives us good reason to accept it. Are the congressmen who want to build a dam on the river wrong? Well, they’re wrong if a dam shouldn’t be built there. It strikes me as relevant that building a dam there would destroy natural beauty; and the force of this aesthetic factor is strengthened if natural beauty, which is a valuable thing, is rare. So I can see 1 and 3 as relevant to the conclusion. Also, if building this dam will do nothing to prevent floods, then 4 is a relevant considera- tion—and indeed, if true, a powerful one. Again, all of this would fit as support for 2. The immedi- ate (unexpressed) premise would be the assertion that there is no good reason at all to want to build a dam on this river. Once again, then, using valid deductive argument forms as our criterion for what belongs to the argument would result in our throwing out as clutter assertions which actually belong to the argument. In this case of both (A), what our friend intended to argue, and (B), the best argument for his conclusion, it helps to have some knowledge of deductive logic, but we would err if we restricted ourselves to what could be putt into a deductively valid argument. I think (A) and (B) are the two purposes we generally have in mind when we want to interpret and assess arguments from a logical point of view. (2) Your example contained as part of the overall argu- ment a deductive syllogism. What do we do in the case of induc- An Early Exchange on the Interpretation of Arguments © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 89 tive arguments (or other kinds of arguments, if there are any)? Here our knowledge of deductive logic would not apply—and it would be important to realize that there are arguments to which it does not apply, for otherwise we would be trying to force non- deductive arguments into deductive patterns and as a result we would distort those arguments So this is another limit to the ap- plicability of deductive logic to the task of extracting an argu- ment from the extraneous material which accompanies it. (3) I understand the dilemma you speak of: you can’t do A until you’ve done B, and vice versa. I am inclined to identify the bootstraps as a sense for relevance rather than a sense for deduc- tive validity. It’s the former that is at the heart of the logical in- tuition we’re trying to instil when we teach a one-shot logic course to undergraduates, I think. Assuredly, one way for X and Y to be relevant to Z is for Z to be entailed by X-plus-Y. But do all assertions (statements, facts) that are relevant to others entail them? It is not clear to me that they do. The facts that a river is beautiful and that building a dam on the river would destroy that beauty do not entail that the dam should not be built. Yet they do appear to be relevant to the issue of whether the dam should be built; at least, they are argu- ably so. What perplexes me is how to analyze relevance; and also, how to teach a sense for relevance. Perhaps studying some for- mal logic helps to sharpen one’s sense for relevance. I think people interested in informal logic ought to be open to that pos- sibility. There is a danger in this approach—one I have encoun- tered when teaching—namely, students find the open-and-shut character of deductively valid argument forms so attractive that they want to assimilate all arguments to it. (I’m aware that this might just indicate flaws in my teaching.) But possibly this dan- ger can be compensated for. Still, having conveyed a sense for deductive validity, we still face the task of conveying a sense for relevance in other forms. How is that to be done? Well, this has turned into a long letter. I would be grateful if you’d let me know about clear mistakes in my thinking, and I’d be most interested to know if you find these arguments at all persuasive. Since I have referred to your letter in some detail, I am returning a copy of it so you can check my interpretations of what you said. With best regards, Yours sincerely, Tony Blair Blair, Copi © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 90 2.3 Copi to Blair-2 University of Hawaii at Manoa Department of Philosophy 2530 Dole Street Honolulu, Hawaii 96822 � Cable Address UNIHAW June 29, 1981 Dear Professor Blair, Thank you for your letter of June 12. I have reread it sev- eral times and must agree with practically every one of the points you make. I have been in a rut for a very long time—being concerned with whether a given argument really establishes its conclusion, rather than worrying about the obviously prior question as to what the given argument is. Despite having recently worked at mastering the kind of analysis urged by recent writers on infor- mal logic, I am not really sure that an old dog can be taught new tricks. I just dashed off the little argument I sent you to illustrate the point I had in mind. If we credit the arguer with the unstated premise A, to the effect that There is no need for flood prevention here. I suppose the analysis of the passage would go like this (using your numeration of the separate propositions that were explicitly stated[)]: 1 3 4 A î í 2 5 ê 6 In order for 4 to be a serious part if the argument rather than clutter (or just “chatter” in this case) it is clear that something like A must be credited to the arguer. Perhaps a useful stab at your notion of relevance might be the following: p is relevant to q iff there are plausible propositions p1, p2, ... pn such that p, p1, p2, ... pn ∴ q is a better argu- ment than p1, p2, ... pn ∴ q. An Early Exchange on the Interpretation of Arguments © Blair, Copi. Informal Logic, Vol. 36, No. 1 (2016), pp. 83-91. 91 I don’t like it much because I don’t see how to make the notion of “better argument” effective. And I also don’t like it because any p can be shown to be relevant in that sense to any q by tak- ing p1 to be “if p then q”. Score: informal logic 1, Copi 0. I should make one final remark. The “Principle of charity” can be stretched too far, I think. We can always credit the arguer with some hidden agenda according to which there is no clutter at all, with unstated premisses always available to “connect” his clutter to the conclusion he wants to establish. In conclusion, I don’t know what the next step is at the present time. I have already indicated that I find your arguments persuasive. If you do discover the essence of relevance, I hope you’ll let me know. With every good wish, Sincerely, Irving Copi 3. Conclusion There was no further correspondence on these matters. I wrote to Professor Copi again in May 1985 inviting him to submit an Abstract for the first conference of the International Society for the Study of Argumentation (ISSA) at the University of Am- sterdam to be held in June 3-5, 1986. He wrote back in June say- ing that while he’d like to go to Amsterdam, he really was too busy. Nevertheless, he apparently was persuaded to go by a later appeal from the Amsterdam organizers, Frans van Eemeren and Rob Grootendorst. However the need for a major operation in March 1986, from which he could not be sure by early June that he would be sufficiently recovered to attend, forced him to beg off. So unfortunately, although Professor Copi was quite famil- iar with informal logic, he did not develop a similar kind of con- tact with the larger argumentation community.