RAIL

“The Way You Hear It Is the Way You Sing It”, by Jan Steen, c.1665

Seven people are sitting around the bar at the local college watering hole, when the bartender looks up from the taps and asks, “Say, how normative is logic, anyway?” From around the bar the patrons answer:

Primus: It isn’t normative at all, if by ‘logic’ all one means is the formal deductive logic taught by philosophers! In order to be normative, logic would have to be able to guide the reasoning (and the actions that flow from the reasoning) of real persons, but formal logic abstracts away from all the properties that real people have. It presents itself in symbols that do not even attempt to represent the complex particularities, contexts, histories, or lived social realities of actual persons. Instead, formal symbolic logic addresses itself exclusively to that restricted class of privileged white male elites who have the leisure to learn its complicated language. Learning this “insider’s game” reinforces their sense of entitlement because in mastering the operations of formal logic they come to believe–and in fact are taught–that their reasoning is superior to that of others who have not mastered it. Thus logic is twice over a sin. It is, first, unusable by marginalized peoples because, in it’s abstraction, it fails utterly to apply to their lives. Those who are seduced into learning it therefore waste the commodities of their labor and their time. Second, it becomes a tool of oppression at the hands of elitists who use it to subordinate others on the grounds of a so-called rational incapacity to think “logically”, that is in reality only a failure to think like themselves, to conform to their “truth”. We would be better off without such “logic” altogether.

Secundus: That’s a bit dismal and, I must say, a bit unkind too. Logic can be normative without being oppressive. Look, human reasoning is variable in its procedures and in its consequences. No one needs an education in order to see that much human reasoning is poor–riddled with fallacies, non-sequiturs, failures to appreciate relevant evidence, failures to see important implications, and so on. The evidence of that is all around, I’m afraid. Similarly, it takes no special insight to observe that failures of this sort often issue in decisions that have disastrous consequences, not just for the person who makes them, but for others as well–both close associates and fellow members of society alike. Ask anyone who has had to bail out a family member who got sucked into a pyramid marketing scheme!

If reasoning is a technique, or a set of techniques, then surely it can be done better or worse. Given the potential consequences, it seems clear that we have a responsibility to try to reason better. That’s what logic is, a study the ultimate aim of which is to make us better reasoners–not so that we can lord it over others but so that we can live better lives and help others to do the same. It is true that classical deductive logic doesn’t always seem to do that, but logic is much, much bigger in scope than classical deductive logic and has been for some time. That’s another point I wanted to challenge you on, Primus.

The study of Bayesian schemes, logics of belief revision, relevance logics, and many, many others too have become the heart of the discipline. No one understands the limits of classical formal logic better than contemporary logicians. Thus, to dismiss all logic on the grounds that the logic of Russell & Whitehead‘s Principia doesn’t capture a particular politically motivated definition of human reasoning is to fail substantially to understand what logic is. When we take into account its full scope, then it becomes clear that logic helps us see more features of our manifold types of reasoning and thus helps us see our own strengths and weaknesses as reasoners in a clearer light.

And by the way, tossing logic out won’t prevent oppression! If people have no tools for thinking, no methods for analyzing their own patterns of inference as well as the arguments of others, why then they will be exposed to the predations of con-artists, liars, fraudulent politicians and snake-oil salesmen of every stripe. Logic teaches people to think for themselves. Surely that’s not a bad thing!

Tertius: Hold on now. I think you may be overestimating what the study of logic can do for people. There’s no doubt that logic is powerful but that doesn’t make it normative with regard to human reasoning, or a force for good in society. You are right that Primus’s idea of logic is a bit outdated. That’s to be expected. Very few people who are critical of logic in this way have actually spent much time studying it. Logic is not what academicians outside of mathematics, computer science, and philosophical logic often seem to think it is. It is, as you say, much, much broader and a good deal more diverse. That said, modern logic never promised to enlighten us about human reasoning and it is a mistake to think that it did.

Logic is neutral. The subject of Russell & Whitehead‘s Principia was first and foremost mathematics, you know–not human reasoning. It is not the fault of logicians that philosophers (and others) adopted its tools for their work and cast them in a normative light that was never intended for them. At a fundamental level, logic is about the properties of various kinds of logical systems. Nothing more. It’s primary applications are within mathematics and computing, not in advising humans about how to become better reasoners, or about how to improve the social world. Everyone will agree that those are good and important goals, I think, but logic contributes to them only indirectly. Better, more pertinent advice for reaching those goals will have to come from somewhere else. Cognitive psychology may be a better place to look for advice on becoming a better reasoner. Economics and perhaps moral philosophy might help with the social goals, so long as they are done with one eye firmly fixed on the empirical. Wherever one looks for one’s answers to normative questions is fine, so long as one doesn’t expect to find them studying logic. That’s just not what it’s for.

Quartus: Well, listening to Secundus and Tertius I’d have to say the answer to the question of whether logic is normative or not is a decided “yes and no”. On the one hand, I think Tertius’s claim that logic is neutral, or plays no direct role in people’s lives is an exaggeration. Aristotle, whom nearly everyone sees as the one who got the ball rolling with respect to what we now call logic, did think that logic had practical import for human beings. (He also thought that rhetoric and dialectic were at least as important too, but let’s not get too distracted about that.) On the other hand, contemporary logic certainly has evolved in ways that make its application to ordinary life very difficult to grasp in any sort of immediate sense–at least for most of us. The centrality of formal methods in logic does come at the cost of transparent applicability to more concrete problems. We simply don’t reason in symbols, according to formal canons, deductive or otherwise. It is in that sense that Tertius is right. There’s not much in contemporary logic that is readily accessible to people in a way that would help them directly to live better lives. I disagree, though, that this means that human reasoning ought to be consigned to cognitive psychology, economics, or a handful of empirically minded philosophers. Scientific study is important and relevant, but it isn’t exactly normative. It tells us what seems to be the case right now with respect to human reasoning, not what ought to be the case, or to be honest even what really is the case in a settled way. (For the latter we would need empirical methods that were final in a way that no empirical methods currently are.) What we need are ways to improve our thinking.

If we take Aristotle’s thinking about the place of reasoning in a human life seriously then there does seems to be room for normativity of a logical sort. Certainly the history of the fallacies shows that there are logical considerations of a non-formal kind that apply to ordinary, everyday reasoning as done by regular people. Formal logical methods, deductive or otherwise, cannot tell me that my emotions are being appealed to in an illicit way, but an awareness of the general character of emotional arguments and the ways of their use can. Surely there’s room to talk about arguments of this and other sorts in a normative way? All we have to do is abandon the idea that formal methods exhaust the domain of logic. If it helps people–and if there is theoretical justification for it, as there seems to be–then why not do that? Let ‘argument’ be the central concept of this study, rather than formal systems, and let us talk about better and worse arguments of various types with as much clarity as we can. If we can do this, then there is room for a normative sort of logic that can guide people in their dealings with real, everyday arguments. I should think that this sort of logic stands a much better chance of securing the benefits that Secundus wants than pounding away at, say, set theory.

Quintus: You should have listened more to Primus, I think. As soon as you move the notion of argument to the center, you’re done with logic of either Secundus or Tertius’s sorts. What you propose isn’t logic at all. That’s okay though because arguments are what we should look at if we want to help people be more effective reasoners. The only place you go wrong is in thinking that that arguments can be better or worse in some objective sense. They can’t. They either work or they don’t work–which is to say that they either persuade or convince the persons one gives them to or they fail in that task. Success or failure in this regard isn’t about any sort of logical goodness, however imagined. It’s not even about being right, whatever that might mean. What it is about is getting others to adopt your position. In order to do that you have to appreciate all the things that Primus wants you to appreciate–all the historical, personal, social complexities that have to be addressed if you want to get a person to even want to listen in the first place. Argument is about interacting with others in this way, not about pronouncing objective judgments on the quality of their reasoning. The question of whether logic is normative or not is therefore an idle one. It isn’t logic that we care about. It’s arguments. The only questions that matter are practical and moral: 1) In giving an argument does the speaker deal respectfully and reasonably with her audience? and 2) Does her argument succeed or fail in winning their adherence? That’s it. Logical considerations matter only inasmuch as they might matter to the audience. And as I think we all agree, that’s not likely to be very much.

Sextus: Ah, but even if you’re right, Quintus, there’s still a role for logic–at least of the kind Quartus favors. You said the speaker has to deal with her audience in a reasonable way. You also want to focus on what “works” but what works sometimes is a matter of whether the argument is good or bad in some sort of logical sense. I don’t share your pessimism about people’s level of care for logically good arguments. You say an argument is good in the only sense that matters if the person to whom it is directed accepts it, but that can’t be right. After all, do we not sometimes think people blameworthy for accepting bad arguments? Do we not say things like “I can’t believe you bought that line!” or “Didn’t you consider this or that objection?” These kinds of considerations suggest that the quality of the reasoning, quite apart from its mode of presentation, really does matter. It can’t be purely subjective. If so, then there may be room for something approaching the kind of normativity that Quartus wants. I think that maybe what you’re really worried about, Quintus, is that Quartus’ sense of ‘argument’ is not really free of the social baggage that vexes Primus. That’s easily remedied. We just have to change our thinking in the direction of expanding the concept of argument beyond the boundaries of the old definition inherited from classical logic. Once we do, it will quickly become clear that the normativity in question is of a different sort.

There’s good reason for this expansion too. An honest look at the way we reason seems to require it. We change our minds for lots of causes and draw conclusions from lots of things. Not all of them are verbal. Sometimes, for example, it is a picture that convinces us that, say, war is wrong. Why can that picture not be treated as an argument? For that matter, why isn’t my mother’s arched eyebrow–which always convinced me that whatever I was doing put me in imminent danger of a trip to the corner–an argument too? If these kinds of things are arguments, then they can be used in better or worse ways. Thus the norms of our “logic” have to be expanded to cover them too.

If I’m right in this then I don’t think I’m that far off from what you want, Quintus, since the normativity in question is going to be of a largely ethical sort. It will involve prescriptions of a moral kind. For example, to use a materially doctored photo to elicit a false conclusion in the minds of the audience is prohibited, not because it violates a logical rule, but because it violates a moral principle of truth-telling.

The only thing we really need to be clear about is that the norms at work in such prohibitions are generated by the community of arguers. They are the product of human discourse, not a fixed feature of the universe. To put it another way, they are neither subjective nor objective, but intersubjective. If we are vigilant about thinking of normativity in this way then we have less of a worry about the old socially problematic norms of “capital-R” Reasoning creeping back in, right?

Severus: I don’t know…I’ve been sitting here listening to you all the whole time I can’t say I’m wholly convinced by any of you. At this point I’ve heard two or three conceptions of logic and as many conceptions of normativity. Just now ‘argument’ exploded too. At this point I have to wonder if this dispute isn’t merely verbal. All of you are quibbling over descriptions over what are essentially the same phenomena. None of you wants to license illicit argumentative coercion, and all of you are concerned to equip others with the ability to resist it. You’re just differing over the precise definitions of terms like ‘coercion’, ‘reasonable’, ‘logic’, ‘normative’ and ‘argument’. Maybe you’d all be better off if you stopped hobby-horsing with the definitions and just got to the business of talking about what follows from what and why?

All: WHAT??? Raah! Rabble rabble rabble!

“I can see I hit a nerve with this one.”, the bartender said, “Tell you what: Next round’s on the house.”