ISSA Proceedings 1998 – Pretending And Port-Royal Logic (Bad Reasoning And Pretend Reasoning)
The standard and approved ways of looking at fallacies start us off with a list of “tidy-looking dichotomies” (Austin 1964: 3). Reasoning is either good or bad, cogent or non-cogent, correct or incorrect, sound or unsound, valid or invalid. Sets like convincing or unconvincing can be found in some versions of the approved ways but not in all. And there are some versions that will include things like misleading, deceptive, and blighted, but their partner-words hardly ever show up.
It’s easy to see that one side of this division is positive and the other, negative. Cases of reasoning put on the positive side are cases of ‘good reasoning’. Cases put on the negative side are cases of ‘bad reasoning.’ Good reasoning is just good reasoning. Bad reasoning gets a special label. It is fallacious reasoning. It is easy to turn this all around and call a fallacy a case of bad reasoning. Sometimes it indisputably is. But sometimes it may not be. Or, anyhow, it may not be just a case of bad reasoning. Getting clear about the times when it is not and why it is not is what this paper is about. It’s a matter of being fair to fallacies.
First a word about some long standing complaints concerning the standard, approved ways. In the early 1980’s Woods and Walton complained that standard treatments of fallacies failed to provide a non-arbitrary way for sorting out cases of correct reasoning from fallacious ones. The standard treatments were mostly happy to take up the inherited list of names, usually in Latin but sometimes very colloquial – remember Flew’s “No-true-Scotsman” (1977: 47) – and give supposedly illustrative examples. Most of these examples were contrived or made up to suit the names and many were so obviously bad that they provided more fun than instruction. And some turned out to be not bad at all.
More importantly, however, clear guidelines and explanations for sorting, for putting this case on one side and another case on the opposite side, were said to be remarkably absent from the standard treatments. For Woods and Walton, this absence came, in large part, from the lack of an adequate model of correct reasoning or argument. They saw their job to be that of providing such a model. This would involve setting out precisely formulated rules, procedures, requirements, and the like for correct (good) reasoning. With such a model firmly in place, it should be a bit of snap to get non-arbitrary guidelines and explanations for sorting the cases of correct (good) reasoning from incorrect (bad, fallacious) reasoning (1982: v).
So we now have a sure-fire, fail-safe way for detecting cases of bad reasoning and for sorting them out into two piles. Moreover, we can give reasons for putting this case in one pile and that case in the other. Clearly, this is much better than what we are said to get in the standard treatments where it was mostly a matter of matching cases or samples with patterns.
The piles, however, have not changed. There are still two of them. And, by and large, we find in the pile of incorrect reasoning the same old list of patterns to which we have to match our samples. It may not be uninteresting to call attention to the similarity between this team’s complaints and those formulated more than three hundred years ago by another team, the Port-Royal team. Arnauld and Nicole, too, complained about the contrived character of the stock examples found in the standard treatments (1970: 49,53).
Like the contemporary team, they believed that the standard treatment was disorganized, heteroclite, and uninstructive. Sorting, as they inherited the business, was more a matter of tradition and habit than of reasoned detection. They, too, thought the business needed fixing.
To fix it, Arnauld and Nicole, like the other team, gave priority to the model of correct reasoning. Correct reasoning or good reasoning was the standard or the norm. Bad reasoning or fallacious reasoning was just the opposite, it was non-standard, abnormal reasoning.
However, the older team gave a bit of a twist to their proposed renovation and they started out at a different point. Good reasoning, normal reasoning was mostly unproblematic, indeed, it was natural. There were no bogs here. The model was all right. Trouble came up when they noticed just how well bad, fallacious reasoning fared – even the wisest of the wise could find themselves on the wrong side of the divide. So this team got started not by proposing to fix the model, but by asking why or how anyone reasoned badly and why or how anyone would be tempted to follow bad reasoning.
Given the fact that good reasoning was natural and that nothing could possibly go wrong in its operations or procedures – the reasoning machine, like a computer, pretty much guarantees impeccable results – Arnauld and Nicole looked to the users. They drew the conclusion that if reasoning goes bad, it must be because users put bad stuff in the machine. When they asked why bad stuff was put in the machine, they answered by pointing to defects in the users. They were, we are, afflicted with weak or unruly wills. And so they (we) let passion, bad faith, interests, and the like take over the nest and let the bad stuff in. This will be how and why reasoning becomes fallacious. Shared passions, bad faith, interests and the like, along with a penchant for the bad stuff anyway, make it easy for everyone to follow fallacious reasoning.
Put another way, if reasoning cannot go bad in terms of, or on the level of, its rules, procedures, or requirements, something about the ways in which we follow the rules and procedures and satisfy the requirements must make what we did abnormal. Qualifiers like insincerely or self-interested, pick out the ways in which what we did become abnormal. So, as it is easy to see, sorting out cases of reasoning in the old way is clearly different from sorting them out in the new way. Still, there are only two piles, bad reasoning is another name for fallacious reasoning (sophisms or paralogisms are thrown in too). Patterns of fallacious or bad reasoning remain the same, and the positive terms really do all the work (Austin would say they wear the pants, but that expression may no longer be available).
In this paper, I will not take these teams as ‘stalking-horses.’ I mention them so as to show that both the dichotomies and the assimilation of fallacies to bad reasoning have been around for a long time and are still kicking (plus ça change …). But I mention them, too, because they make it easy to see that, should one team reproach us for having produced a case of bad reasoning, the things we can say, or have to say, if we are to get ourselves off the hook will be very different from the things we can, or have to say, should the other team lay charges. Our ‘outs’ on one side are really kind and gentle. On the other side, either we may not be able to get out or there may be no point in trying to get out.
Looking at the allowable pleas should take us some distance in making clear that there may be more to fallacies than bad reasoning. Having made that clear, it may turn out that we are not any better off. Still we should take a look.
Looking may be easier if we clean up the place and rearrange the furniture a bit. The list of “tidy-looking dichotomies” can be seen not as a list of opposed features of reasoning but as a list of criteria for applying the labels, good reasoning and bad or fallacious reasoning. If we take the approved list, which is good enough for our purposes, good reasoning will be reasoning that is cogent, correct, sound, valid, and the rest. Bad, fallacious reasoning, of course, will be the opposite or the negation of good reasoning. It will not be cogent. It will be incorrect, unsound, invalid, and the rest. So we have two piles and the criteria for putting cases of reasoning in one or the other.
We may want to take some of these items off the list or add others, but it is easy to see that the items are just the rules, procedures, or requirements of good reasoning. These will be the criteria. Set out in negative terms, as opposites, they will be the criteria for applying the negative, bad label. So reasoning will be said to be bad when the rules or procedures are not followed or the requirements not met, whatever those rules, procedures or requirements might be.
What are some of the ways in which we can not follow the rules and procedures or fail to meet the requirements? I suppose we could say that these things can be done in much the same way as spilling ink can be done: intentionally, deliberately, or on purpose. We may, of course run a red light or move a pawn three squares in any of these ways. But it is hard to think of anyone intentionally or deliberately, reasoning in a non-cogent or incorrect or invalid manner, or of producing an unsound case on purpose. So, maybe when a case of bad reasoning turns up, we will be better off to think of things like mistakes, errors, oversights, slips, blunders, misinterpretations, and the like.
Now, should anyone be charged with bad reasoning, it will be open to them to plead mistake, error, blunder, slip, and the rest. Such pleas may make them look stupid or silly or negligent, but the nice thing about these pleas is that they delimit a range of defects or shortfalls that can be, in principle, both detected and corrected. And when they are corrected, not only will anyone who slips be out of the frying pan, they may be out of the fire too. Things will be as they should. They will be back to normal.
So, taking fallacies to be cases of bad reasoning has an upside. If we have a full enough set of criteria, and if the criteria are clearly and precisely formulated, in principle if not always in fact, fallacies will be easy to detect. Moreover, if fallacies are cases of bad reasoning, fallacies will be both detectable and correctable. They can be fixed, made good, and that will be the end of it. It’s rather like moving the cursor back and deleting this or that letter, word or sentence, or adding something, or changing the order, and so on. Once it’s fixed, there’s nothing bad left. This is a pretty happy, kind and gentle story. Still some versions of the approved list include, on the negative side, criteria like misleading and deceptive. They, however, look quite different from all the other couples.
In the case made famous by Austin and Hart, Finney made a mistake in the taps with the result that Watkins was scalded (Austin 1979: 195-197). In the same way, we may make a mistake, for example, in the grounds we give for some claim we put forth. One possible result of our mistake may be to mislead or deceive our audience. However, if we see “scalding Watkins” not as something
Finney did, but rather as something that happened, as an accident, then, “misleading or deceiving the audience,” if it is the result of a mistake, will not be something we do. It, too, will be an accident. Looked at in this light, it is pretty clear that misleading and deceiving and all their neighbours can not be, just like that, criteria for grading or classifying cases of reasoning as bad any more than scalding Watkins by accident can be, without qualification, a criterion for putting Finney away.[i]
However, Finney would never had been brought to trial had it not occurred to the prosecutor that Finney might not have made a mistake in the taps at all. He might have turned the hot water tap just so as to scald Watkins. So, too, we may not overlook some piece of information, but withhold it (it’s hard to do that by mistake or inadvertence) just so as to mislead our audience. We may cook the books (this, too, is hard to do by mistake or inadvertence) just so as to deceive our audience or the accountants.
Misleading and deception are things that we can do. Sometimes they happen as the result of something else we did. Sometimes we do them inadvertently. Then they are things we did not exactly do. Sometimes, however, we do them by design. We can work hard at doing them. Now we have a rather different story.
Suppose misleading and deception, for example, are not put on the list of criteria just so as to warn us off bad reasoning, to point to the bad things that might result from bad reasoning. They are put there because they really are criteria for applying the label bad, fallacious reasoning.
Of course, the label can not be applied if misleading and deception are thought to be the result of a mistake of some kind. So, should someone say: “Your argument is bad, fallacious because it is misleading and deceptive,” they must believe that misleading and deception were part of my plan and not at all the result or consequence of some mistake or slip or whatever. What can I do? Deny it? (But what will I deny?) Apologize? Hang my head in shame? Run away? Make amends? Whatever I do, I may never be able to fix the damage I am said to have done. If my plan was to mislead or deceive by arguing as I argued, I can not correct anything. There is nothing really correctable, nothing that I would want to correct. This, then, is pretty clearly a very different story. Maybe it is a darker, more distressing story, one that may make us look pretty bad. But it is a story that must be told.
Arnauld and Nicole championed this story. They, too, put fallacies in the bad reasoning pile, but their criteria for doing so turn out to have little to do with the rules, procedures, and requirements of reasoning or argument. Although we should take time to fully savour the procedures followed by the Port-Royal logicians, how they say one thing and then take it back, it may be enough to recall that they praised Aristotle for having picked out the fallacy of ignoratio elenchi and then went on to chide him for having put his foot in it. Arnauld and Nicole were really not so much concerned about Aristotle’s putting his foot in it as they were intent on pointing out how his foot got there.
Aristotle, Arnauld and Nicole said, refuted Parmenides and Melissus by attributing to them doctrines they did not hold. Then they went on to say that Aristotle insincerely reported these doctrines. In saying this, they clearly rule out such things as distraction (Plato rang him on his cell phone just as he was transcribing these doctrines), poor light (power failure), fragmentary documentation (library rats made a meal of the manuscripts), mistakes (Aristotle took the wrong books). But they clearly ruled in the possibility that Aristotle put contaminated materials in the machine. His premises were false, but they say he made them that way. So Aristotle could not plead accident, mishap, or anything resembling these pleas. Insincere reporting is not something that can be done in any of those ways anymore than tying a string across the top of the stairs can be done unintentionally, accidently, or inadvertently (Austin 1979: 275).
Anyhow, Arnauld and Nicole, pressed or not by the time-frame of their little wager, were anxious to say something else. Namely that insincerity and all the other bad ways of doing things have darker, maybe deeper, causes in the form of passion, bad faith, and the desire to be right (for a fuller list see: Arnauld and Nicole 1970: 304ss). And all this comes from a weak and unruly will, not from stupidity or ignorance of the rules, procedures or requirements of good reasoning. If we still want to call a fallacy a case of bad reasoning, we will have to add more rules to our list. But such rules will have little to do with reasoning and much to do with the ways we expect everyone to do the things everyone does. Violating these rules gets us into another kind of trouble. Trouble that we can hardly get out of. In a way this is the line Arnauld and Nicole took. They say that we expect l’homme de bien to be sincere (1970: 304). But they give the line a rather special pitch or twist and it becomes unclear whether we are willing to let our passions, interests, and desire to be right run the show or whether we are so inclined. Willingness calls for reformation, more stringent rules, or a larger, vigilant police force. Inclination calls perhaps for genetic engineering or outside help.
Outside help is what Arnauld and Nicole called for. They made this appeal because they saw part of our nature as disposing us, even pushing us to give first place to our passions, interests, tribe, and the rest. So, with this story, either nothing we say can get us out of what we did – what can we say to the charge of insincere reporting, other than denying it? – or what we did comes naturally and there is not much point in trying to put a stop on doing what comes naturally (although we might want to cover it up or hide it). I have no inclination to follow Arnauld and Nicole along either of these paths, even though they may have correctly taken the lay of the land and set the terms of a wager. But I do believe Arnauld and Nicole were on to something that their inherited commitment to the dichotomy – good reasoning-bad reasoning – may have made it difficult to fully articulate. They alleged that Aristotle was really interested in making himself look good by making his predecessors look bad, not in getting at the truth of the matter. They said, in other words, that Aristotle was putting down his predecessors, not arguing with them. They went on to say that Aristotle consequently messed up his argument. That he used false premises, premises he made up. The result of which was a clear-cut case of bad, fallacious reasoning.
Arnauld and Nicole might have said that Aristotle was not exactly arguing, but doing something else. Had they really said something like this, they would have been able to see that the opposition is not between good and bad reasoning but between reasoning and something else, the something else being that which was really done. Then they might have drawn the distinction between a case that is a faux raisonnement and a case that is a raisonnement faux. It would, then, have been clear that the inherited dichotomies, patterns and stock samples put different fish in the same kettle. Arnauld and Nicole, of course, did not follow this line. But it is a line Austin followed and was quite good at following. That is where the next pull, not to say light, will come from. “That chap over there, he’s all right I suppose, he’s cleaning the windows, eh?” “Ah, him, he’s pretending to be cleaning the windows right enough, cleaning ‘em a treat too: but I see him taking note of the valuables through ‘em all the time” (Austin 1979: 259).
Pretending to be a hyena at a party and pretending to be cleaning the windows are different cases. Success at the party depends on satisfying one set of criteria. Success at the windows depends on another, different set of criteria.
Pretending to be a hyena and taking a bite out of someone’s calf, “taking a fair sized piece right out of it” (Austin 1979: 256) will be carrying the pretence too far, but really cleaning the windows may be a necessary part of pretending to be cleaning the windows. The case is different, too, when on the stage we pretend to saw someone in half. Here delivering the genuine article will get us on the news or guarantee our 15 minutes of grim glory. There are, I believe, some important lessons to be drawn out. The lessons do not require imitating Austin’s “hounding down the minutiae.” Broad outline should be good enough.
Firstly, successfully pretending to be a hyena supposes a certain transparency in the pretending so that there will be little possibility of confusing the pretence with the genuine article. Something like this is certainly true of the magician also. What we admire is the skill in bringing the trick off, where bringing the trick off is being like the genuine article without the grim consequences. Cleaning the windows is different still. Here the pretence can not be transparent in just these ways. Indeed the window cleaner, if the pretence is to be any good, had better deliver the genuine article.
If the genuine article has to be delivered, why talk about pretence? Because the pretence covers over something else, it dissembles what is really being done (taking note of the valuables). If this is the case, there is no room for the window cleaner to slack-off or make mistakes. Austin’s window cleaner clearly muffed his pretence. Something about the performance gave it away. Window-cleaners don’t ordinarily stop cleaning the windows to take notes. They know the difference between a squeegee and a mop, Windex and motor oil. The successful pretend window cleaner does such a good job that no one knows what they were up to – until the valuables disappear.
Secondly, it is clear that pretending to clean the windows and taking note of the household treasures are not opposed to one another in any intelligible way. One is used to mask, dissimulate the other. We do one thing so as to hide the fact that we are really doing something else. This is where Arnauld and Nicole get back in. Aristotle, they might now say, just pretended to argue while what he was up to all the time was a ‘put-down’ of his predecessors. He cast the ‘put down’ in the form of Bocardo (or maybe it was Barbara), it looked good and genuine and got by for a long time. Parmenides and Melissus did look bad until someone checked the original documents. So we might pretend to argue while all we are really doing is putting someone down, venting our passion, protecting or promoting our interests, or any of the other things we do but do not want to come right out and do. Clearly, if we are to do these things, if they are to get by, our pretence had better be convincing. It had better be enough like the genuine article to get by.
The next case is different. Here there is an opposition between the genuine and the sham, the spurious, or the fake. What is hidden or dissimulated is the sham, spurious or fake article. An example might be, if we are materially inclined, an imitation Rollex or a counterfeit bank note. These things had better be enough like the genuine article to get us to be tempted by a good deal, or to give the man two tens for the twenty without asking any questions. The more intellectually inclined will want another example. Here we need only think of a bad argument disguised as a good one or that is passed off as a good one. In this case, the resemblance hides the fake, spurious article. In the first case, the genuine article hides or dissimulates something else. It may be hard to keep these apart, but they are clearly different cases.
When we encounter a case of bad reasoning what we ordinarily do is pick out the mistakes, errors, slips, and the like and then correct them or try to get them corrected. Sometimes we get thanked for doing this, perhaps not right away, but thanks usually do come eventually. We may, of course, also chide the bad reasoner on not knowing things we think they ought to know, or for not interpreting information in the ways, given their training and experience, they are expected to interpret information. We may chide them on many other things, all of which will be related in some way to getting the argument fixed and reasoning straight or back on track.
When we encounter a case where someone is taken to be pretending to argue, we may indeed look for mistakes, errors or slips that will give the little pretence away (nothing so gross as a bearded queen on a bank note), but, also, ordinarily, we will try to bring out into the open whatever it was they were hiding (the fake bank note, the bad argument, the break-in, the put down, et cetera). We will chide them not on their mistakes or slips (we may be happy that they forgot that the queen does not have a beard), but on the abuse they have made of a perfectly honourable trade, window cleaning, engraving, arguing, or whatever. We will denounce them or turn them in. When we do, we do not expect any thanks from them (pace Hegel). Indeed, when we do point out the pretence, reveal the agenda or party line, all we may get is denial – “That’s not what I was doing” or a shifting of the load, “You don’t understand.”[ii]
Of course, there may be cases where we do pick out the mistakes, errors, slips, and the rest without getting any thanks at all and, as it turns out, we lose the promotion or the job to boot. Then we may wonder what the bad reasoner was really up to in the first place.
This takes us back to the beginning. In the De sophisticis elenchis, Aristotle distinguished not between good reasoning and bad reasoning but between genuine reasoning and reasoning that only seemed to be so. Reasoning that only seemed genuine is sham reasoning. Sham reasoning, Aristotle said, makes up a different class of arguments. Another name for sham reasoning was fallacious reasoning.
Right at the start, then, there were more than two piles into which arguments could be sorted. It will be easy to see why there are more piles if we remember Aristotle’s striking example of the difference between the genuine and the sham: there are beautiful people who are beautiful thanks to their beauty, while others seem to be so by dint of embellishing themselves (164b, 21).
Clearly looking beautiful is not the opposite of beautiful, so we will not call the embellished person an ugly person but a person who looks beautiful by dint of embellishment. So, too, the opposite of a fallacy will not be a good argument, it will be a genuine argument. If this is the way it is, when we sort out cases of arguments, taking into account the distinctions and differences that show up, we will need more and different piles than the approved dichotomies allow for.
Moreover, it will appear that a fallacy need not be a bad argument, but only an argument that looks genuine by dint of a certain likeness to the genuine. Indeed, it may be the case that a fallacious argument is a good argument. But, whatever the case, there will have been both dissimulation and abuse of sorts (misuse and misconduct). This, without doubt, is what got Socrates’ ire up.
Equating fallacies with bad reasoning will be unfair to fallacies. We will, moreover, be making a mistake, not as serious perhaps as the one that comes from assimilating winning a war to sneezing or horses to beds, but still serious enough (cf. Austin 1979: 179). It will be one that misleads us. For we may look for mistakes when we ought to be looking for misuse or abuse, we may call for corrections when apologies are needed and call for apologies when corrections will do. This is unfair to fallacies.
To be fair to fallacies, we may have to work harder at keeping grading separate from sorting, classifying or cap-fitting. Determining worth and merit using some scale – good to bad, strong to weak, 10 to 1, or whatever – is clearly different from putting things, good or bad, strong or weak, 10’s or 1’s, into different piles or putting caps on the different cases that appear. It may happen that some stock caps, off-the-shelf, made up in the shop before the customers come in, don’t really fit. Can we pretend that the caps we have on hand will fit every head that appears? To be fair and honest we may have to tailor-make some caps.
i. Moreover, it is equally clear that the thing to do in Finney’s case is to fix the taps, to make them such that mistaking one for the other will be less likely. In the case of mistaken grounds, whether anyone gets taken in by them or not, the thing to do will be to correct the mistake or get it corrected. But even setting things straight will not be a guarantee against misleading and deception. After all, an impeccable calculation may mislead. Remember Austin’s 3.75 men needed for building a cistern (1979: 194) and many of us know that telling someone your place of birth may mislead them about your first language or your genetic stock.
ii. Check the letters in the New York Review of Books where reasoned exchange is the major product. We find reproaches like blasé disregard, academic arrogance, hint, innuendo, caricature, larded with political and religious motives, bluff and posturing, and protective wall of unbreakable a priori conclusions. Such reproaches rarely give rise to corrections or straightforward rebuttals, they are mostly rebuffed, denied.
Arnauld, A. and Nicole, P. (1970/1683). La Logique ou L’Art de penser. Paris: Flammarion.
Austin, J.L. (1964). Sense and Sensibilia. Oxford: Oxford University Press.
Austin, J.L. (1979). Philosophical Writings. Oxford: Oxford University Press.
Flew, A. (1977). Thinking Straight. Buffalo, New York: Prometheus Books.
Woods, J. and Walton, D. (1982). Argument: The Logic of the Fallacies. Toronto: McGraw-Hill Ryerson.