The paradox of polemic and related interpretive phenomena

by Dave Maier

Recently I’ve been reading a couple of books attacking postmodernism and/or leftist politics, which the authors – not surprisingly – tie together as closely as they can, albeit from rather different perspectives, for maximum polemical effect. Maybe we’ll get into the gory details some other time (haven’t got very far into them yet), but for now let’s just examine a few things that struck me about the very idea of polemical interpretations. 

For example: I can never figure out whether I’m supposed to be the audience, invited to join the authors in combating these worthless and dangerous ideas, or instead whether I am the target, such that these books intend to smash my own position(s) to conceptual smithereens. If I picture myself beside the author as he fires bolts at his target, I find myself alarmed at the way he’s swinging that crossbow around: watch where you’re aiming that thing! But maybe I myself am really the target – in which case I watch in puzzlement as the bolts sail harmlessly off to one side (was that really aimed at me?). If it were consistently one rather than the other, I would simply toss the book aside as incompetent; but when the two flip back and forth like the duck and the rabbit, it makes me wonder about polemics in general. How exactly are they supposed to work?

The paradox of polemic

Polemic must be possible in principle. If we reject some idea or ideology as fundamentally misconceived, or actively destructive, it must surely be possible to say why this is. Indeed, a lot of what we read nowadays has this truculent tone, telling us why Those Other People are getting things completely wrong. However, effective polemic turns out to be very difficult to produce. The worst of the charges fail to apply to very much of the target, and what does clearly apply to all of it seems rather less bad. Often the author attempts to paper over these flaws by equivocation, poisoning the well, guilt by association, and other dodgy rhetorical techniques. Sometimes this is because the author’s point is too thin to support the full sweep of his categorical denunciation. But even when it seems that there is indeed a proper target for such treatment, the same thing often happens. Why is this?

The main reason for the difficulty of effective polemic, or so I believe, is inherent in the task of interpretation itself. The polemicist is trying to show that some set of ideas is utterly unacceptable to right-thinking people; but he also needs to show, on pain of irrelevance, that this phenomenon is worth paying enough attention even to reject as nuts. Who cares about what a mere handful of cranks believe? No one’s going to read the definitive takedown of the idea that since monarchy is the only legitimate form of government, the USA actually still belongs to the UK.

That is, once you’ve made clear that your target is a widespread phenomenon, you need to show that that crazy stuff is what these people actually believe. But if you describe it as the sort of piffle that only crazy people believe, then they won’t agree that they believe it in the first place, and quite naturally accuse you of erecting a straw man. The trick – the paradox of polemic – is to portray the relevant position as making just enough sense for your opponents and their defenders to accept it, while still retaining the obviously fatal flaw which justifies your dismissive and contemptuous tone.

This is naturally an aspect of the interpretive principle of charity. Charity is often misunderstood – maybe because the name makes it sound like “being nice” or something – as not allowing an interpreter to disagree strongly, or at all, with his interlocutor. But charity isn’t just being polite. After all, you don’t want to allow the target of your polemic to wriggle out of the supposedly devastating consequences of their beliefs, the punch line of your entire project. You need to get them on the hook, by getting them to agree to things that sound perfectly innocent and sensible (but actually, you intend to show, contain the seeds of their own destruction). This will also make your attack more relevant to your audience, as it will thereby show the real danger of these seductive premises. An attack on the idea of returning the USA to the UK can’t simply attack the inference from its monarchical premise to its unacceptable conclusion. That’s not a polemic; that’s just a reasoned disagreement. A more effective polemic would not only grease the skids from premise to conclusion, but also make that premise as superficially attractive as possible, the better to bait the hook, with the intent of springing the real counterargument later on, crushing the whole thing at once.

Charity is also the best method of avoiding an actual straw man. Again, charity is a general interpretive principle, not relevant simply to polemic. Consider Daniel Dennett’s formulation of his own rules for interpreting another thinker’s ideas, which begin with the charge that we first aim to summarize that thinker’s ideas so sensitively that they actually thank you for putting their views so insightfully, and teaching them things about their own views that they hadn’t even seen yet. Only then, he says, may you breathe even a word of protest. This is very hard to do, so it’s not too surprising that Dennett himself falls down on occasion. Nor is it surprising that in polemical interpretations of postmodernism, such charity – and, again, effective polemic – is virtually unheard of, in the race to tar the targeted cretins with comically self-undermining relativism or some such.

Reference, categories, and predication

Another interpretive principle relevant to our subject is that whenever you say anything at all, you are in some way manifesting both your beliefs and your meanings simultaneously. Take the statement that “Liberals believe that government is the best way to solve problems.” Here you are both indicating who “liberals” are (they’re the guys who believe this) and at the same time saying something about them (they [are the guys who] believe this); or, in linguistic terms, defining a predicate implicitly, on the one hand (here, “liberals”), and predicating something of a certain group (here, that they believe a particular thing), on the other. Our emphasis in the context may be on one or the other — and of course we occasionally provide explicit definitions of this or that term — but in the main, these things are inextricably intertwined, rendering the interpretive process unavoidably holistic: we attribute both belief and meaning in a single interpretation.

Here also the effective polemicist must not let his polemical conclusions interfere with a proper taxonomy. It is here that the danger of equivocation is most easily seen. For “liberals” and “postmodernists” believe other things as well as the things which define them, e.g, the ones we are rejecting as dangerous. If you go on to tell me that “liberals are anti-military” or “postmodernists don’t believe in truth,” this is both another attribution and another implicit definition. If our multiple definitions don’t match up – even if the term seems appropriate in each context – this is a problem. Are the “liberals” who reject free-market solutions the same “liberals” who are anti-military? As with charity, it is often safer to let the target group define themselves, even if that lets them leave out things we’d like to hang them with, if we can get them to do what amounts to hanging themselves instead. It will also help the reader, as I noted at the outset, keep track of whether he’s the audience or the target.

The no-true-Scotsman “fallacy”

What has become known as the no-true-Scotsman fallacy goes like this:

James: No Scotsman puts sugar on his porridge.

Ian: But my uncle Angus is a Scotsman, and he puts sugar on his porridge.

James: No true Scotsman puts sugar on his porridge.

As presented, this looks like a semantic dodge. Universal generalizations, like the ones we’ve been talking about, are vulnerable to counterexamples, and when Ian presents a clear counterexample to James’s general claim, it seems that James should withdraw or qualify it. But instead, he doubles down. The alleged counterexample is simply excluded – arbitrarily, it seems – from the relevant domain, thereby preserving the generalization. If we allow this, it looks like no counterexample is even conceptually possible; this makes James’s claim look unacceptably dogmatic.

But as Wikipedia notes, this is at best (that is, at worst) an “informal fallacy.” There’s no reason why a challenge to a general claim can’t be met by a restatement in slightly different terms. In James’s case, this seems motivated by sheer dogmatism; but this need not always be true. In fact, we often don’t need or even want so strong a generalization as to have no counterexamples at all (what liberal, for example, believes everything that “liberals” are supposed to believe?).

In fact, there’s a way of taking James’s statement in a way which is not even dodgy, let alone fallacious. Why can’t he be using it normatively rather than descriptively? That is, rather than a dogmatic claim, in the face of contradictory evidence, it’s a (perhaps rueful) expression of a normative attitude: this is how things should be even if, in certain regrettable cases, they’re not.

“The exception that proves the rule”

As just noted in the previous section, we need not always aim for exceptionless generalizations. Sometimes the exceptions can even help make clearer what we are actually claiming. There are a number of senses in which the name of this principle can be taken, so we have to be careful, but whatever we call it, it does seem that it picks out something interesting. (Pedants love pointing out that “prove” used to, or can still, mean “test” rather than “demonstrate,” and so the other meanings are wrong; but this is not interesting.) The clearest case (the “original” or “true” meaning, per Wikipedia), is e.g. a sign saying “Parking prohibited on Sundays.” The cited exception (“on Sundays”) only makes sense if there is a non-stated rule that parking is generally allowed. In this sense, that there is an exception “proves” that there is a rule to which it is an exception.

Another sense, more relevant to our own case, is what Wikipedia calls the “loose rhetorical sense” of the term. Here the thought is that an actual counterexample to a proposed rule does indeed make sense on its own terms (unlike “Parking prohibited on Sundays”), but, perhaps counterintuitively, lends positive support, albeit by implication, to the rule. (Wikipedia calls this a “rule of thumb,” rather than an actual rule, but of course “rules” can have exceptions and still be perfectly good rules – as Wittgenstein for one points out.) Here the rule that a local village is “always” quiet is lent support by the exception that there is a rock festival there once a year. That fact does not of course show that the village is indeed quiet on other days; but if you ask whether the village is quiet, and all you ever hear is that there’s a rock festival every year, that tends to imply that that’s the only reason it might not be (as opposed, for example, to the incessant baying of sheep). This means it probably is quiet when the rockers aren’t in town.

As Wikipedia keeps telling us, the “rule” so “proven” is not a strict rule, and not “proven” either. But that seems less important than they make it out to be. If to mention it at all “draw[s] attention to the rarity of the exception, and to establish” the usual (if not literally exceptionless) state of affairs, then that’s what we’re talking about. The relevance of this phenomenon to our subject should be clear, so I won’t belabor it.

Bait-and-switch (motte and bailey) diagnoses and their symmetrical nature

It’s pretty clear from the name what “bait and switch” means: it’s a form of equivocation, a clear fallacy. I agree to something unobjectionable, and then you use it as if it meant something bad, without further argument. Flag on the play! As popularized a few years back by philosopher Nicholas Shackel, a “motte and bailey” doctrine is very similar: one advances a controversial take on some idea (the “bailey” — see the link for the medieval-fortress explanation and a diagram), but then when under attack, one retreats to the much more defensible “motte,” a virtually truistic version of what one pretends to be the same thing. Not simply a fallacy, in Shackel’s telling, this technique allows one to lend undeserved support to the controversial idea simply by association with the truism through repetition and purposeful vagueness. (See his post for further elaboration, which I am eliding here for brevity’s sake.)

Such diagnoses are very useful in polemical contexts, if you can get them to stick. No one’s going to admit to equivocation, especially as a conscious argumentative strategy, but if you can rightly (charitably) attribute both meanings to your opponent, and show how he dances from one to the other, that should do the trick. But what I have noticed in particular cases, and which I suspect may be true in general, is that these cases are perfectly symmetrical: an attempt to make an accusation of equivocation in this way is itself vulnerable to analogous countercharges.

Let’s see how this works. Richard Rorty and other critics of metaphysical realism are often accused (e.g by Shackel himself) of equivocation of this kind. Rorty is taken to insist, for example, that “objective truth” is of no interest: a shocking claim! But on other occasions, when pressed, he agrees to obvious truisms about truth and objectivity, on pain of admitting to the most facile idealism and/or relativism. (My advisor Akeel Bilgrami told me that they used to call Rorty “Eeyore” after the grumpy donkey in Winnie-the-Pooh; I can just imagine Rorty saying “oh, all right, I guess so” in an Eeyore voice, when agreeing to truisms of this kind.) Back and forth: a motte and bailey.

But from the pragmatist perspective, it is the realist who is equivocating. (The case is complicated by Rorty’s actual errors by my lights, errors entirely misunderstood and exploited by his critics in a deplorably uncharitable manner. What follows is my own take on the pragmatist attitude.) Naturally if you say something unobjectionable or obvious about “objective truth,” then I’ll shrug and agree with you (truisms like “saying doesn’t make it so” or “belief can be false”); but if you say something that sounds contentious or otherwise dodgy (e.g. metaphysical doctrines like “metaphysical realism”), I’ll object, especially if you seem to be regarding it as entailed by or identical to the banal truisms.

This makes the situation almost entirely symmetrical. Pragmatists can seem to realists to jump back and forth between accepting realism and rejecting it in favor of controversial philosophical doctrines; but realists seem to pragmatists like me to jump back and forth between accepting mere truisms (or pressing me to do so, as if I didn’t already), on the one hand, and going beyond them to controversial philosophical doctrines like metaphysical realism, which are not at all the same thing.

It’s true that our inherited way of talking about these things makes the relevant differences (between truism and philosophical doctrine) difficult to parse. This is what results in the rhetorical differences between “postmodernism” and its realist critics. The postmodern “bailey” (or, more specifically, the relevant aspect of a “Troll’s Truism”, as in Shackel’s post) is explicitly described as an “exciting falsehood” (as opposed to the boring truism which lends it the only credibility it can manage). But when realists seem to go beyond the truisms to what pragmatists see as controversial, they never present it that way. Metaphysical realism isn’t an exciting discovery; for realists, it’s simply the non-controversial but rigorous philosophical manifestation of that same truistic thought, and thus nothing to shout from the rooftops – until, that is, crazy people reject it (and even then, it sounds funny to shout something so obviously “commonsensical”).

This doesn’t mean that neither side is right and it’s a draw. (It may be an impasse; but impasses aren’t draws.) It just means that the mere diagnosis of baiting-and-switching cannot do all the work (which we really should have known already). In my view, the metaphilosophical attitudes which tend to go together with the philosophical doctrine of metaphysical realism keep one from seeing the real problem, making the issue more one of vision than of argument – yet another reason to distrust the false finality of polemical interpretation and getting back to the real task at hand.