Democracy, Deliberation, And The Owl Of Minerva Problem

January 15, 2017 Picture: Rick Wilking/Reuters.

This article is part of The Critique’s January/February 2017 Issue “Stick It To The Man: A Year Of Anglo-American Populist Revolt Against A Changing Culture And An Obtuse Political Establishment.”

The owl of Minerva flies only at dusk. This is a poetic way of saying that wisdom emerges only in hindsight, that understanding is always backward-looking. Our explanatory vocabularies are developed to account for and evaluate what has already transpired, what has surprised or offended us. And once we have alleviated one source of surprise or offense, new sources arise, ones that could be foreseen and understood only by means of conceptual tools yet to be developed. For those of us who are committed to a conception of democracy as fundamentally deliberative, this problem of wisdom’s retrospective orientation is of particular urgency.

Deliberative democrats hold that most of democracy’s value (its authority, legitimacy, and justification) depends on how citizens enact their public role. In a deliberative democracy, citizens shape political policy by means of processes of exchanging ideas, reasons, and arguments. In particular, deliberative democrats think that proper processes of public reason exchange among citizens tend to engender morally and epistemically better political outcomes; and this tendency is typically held to be the key to understanding democratic authority. That is, deliberative democrats explain the bindingness of democratic outcomes by appeal to the moral and epistemic dimensions of open public deliberation.

That democracy’s authority should be based in moral and epistemic considerations would have struck Plato, and many political thinkers since him, as absurd. On many traditional views, democracy is not a means for realizing morally and epistemically good results. Rather, democracy is seen as a way of muddling through, a mechanism for achieving political stability among a largely ignorant and internally factionalized population. Though it is the most fully represented view in the tradition of political philosophy, this understanding of democracy is no longer prevalent; indeed, it is rejected not only in theory, but in practice as well. It is not an overstatement to say that contemporary democracy is deliberative democracy, at least in aspiration.

“Contemporary democracy is deliberative democracy.”

To see this, consider that so many features of democratic political life depend for their intelligibility on the aspirations of good argumentative culture. News is no longer simply read, but is presented in a format of pro-and-con panel debate; journalists ask questions about reasons politicians have for policies, and test politicians’ views for internal consistency; public debates are organized prior to elections. These are all in the service of realizing a deliberative democracy, a mode of democratic politics where public argument is a (perhaps the) central civic activity. Note further that the deliberativist aspiration drives our criticism of what are generally taken to be democratically degenerative forms of political communication. Bias, spin, derp, lying, flip-flopping, glad-handing, and all the rest could hardly be regarded as deserving of political criticism except against the backdrop of the ideals of deliberative democracy. These terms would not be accusatory were it not presumed that democratic citizens are committed to a politics of epistemically sound public argumentation.

“News is no longer simply read, but is presented in a format of pro-and-con panel debate.”

Though the deliberativist aspiration is widespread, and arguably constitutive of contemporary democracy, it remains an aspiration. As we all know, public argumentation among democratic citizens is at best a mixed bag. Attempts to deliberativize democracy are fraught with hazards. For one thing, democratic citizens tend to discuss politics mainly with like-minded others. This occasions the difficulties associated with the phenomenon known as group polarization: as like-minded people speak to each other about their shared views, their positions shift towards more extreme versions. Put otherwise, under conditions of doxastic homogeneity, deliberation produces extremism. And as one’s views progress towards extremity, one grows increasingly unable to countenance the possibility of reasoned, informed, and sincere disagreement. That is, group polarization feeds what Julian Sanchez has described as epistemic closure, the incapacity to see views that run counter to one’s own as even intelligible. And as one’s views become epistemically closed, one will increasingly find one’s critics to be incompetent, insincere, dishonest, and ignorant. Their points will appear as mere noise or petulant emoting. Eventually, one will see fit to simply stop listening to those who espouse opposing views. As a result, there will be progressive all-around marginalization of unorthodox, unpopular, and unfamiliar political views. Those already least likely to get a hearing will be increasingly regarded as incapable of intelligible speech. Accordingly, the deliberativists’ central democratic mechanism seems to undermine democracy.

Of course, these difficulties become only more pronounced once it is noticed that public political deliberation is of necessity largely mediated by various modern communicative technologies. No deliberative democrat explicitly calls strictly for face-to-face discussions among citizens; public deliberation must be facilitated by intermediary institutions, such as social media, television news channels, websites, and online forums of many other kinds. Despite the fact that these communication platforms all offer the potential for robust argumentative exchange among citizens who may be geographically and politically distant from one another, in practice, they tend to exacerbate the vices outlined above. A casual survey of the comments thread of nearly any news site will show that polarization, closure, and marginalization are the norm. Well-run argument is beyond scarce on the internet.

The owl of Minerva flies only at dusk. Only after we have identified these pathologies and suffered their consequences can we think about how to mitigate them. There are many fixes on offer among theorists of democracy. Some argue in favor of new and demanding civic duties that require citizens to read widely across the political spectrum. Others propose institutional interventions, ranging from the enactment of “equal time” laws for news outlets and websites, to the creation of a new national holiday devoted to professionally-orchestrated public deliberation events. Empirically-minded theorists of deliberative democracy are working vigilantly on these matters. It is safe to say that there is no easy way to inoculate deliberative democracy against these pathologies. But even if there were, we worry that another, even more foreboding difficulty lurks. There is reason to think that public argument itself, even when institutional distortions do not pervert it, yields its own pathologies.

Consider the following. Some arguments fail because they run on false premises. Other arguments fail because they draw an obviously unwarranted conclusion from their premises, much in the manner in which a magician pulls a rabbit out of his hat. In such cases, it isn’t difficult to see that something has gone awry. But some cases of the latter kind of failure aren’t so obviously failures. These are cases of fallacious argument. Fallacies are arguments that we tend to regard as good, but in fact are not. We have to work to see them as failures, and developing the ability to see them as failures requires us to craft concepts with which to diagnose the ways in which they go wrong. This calls us to theorize arguments. The task is notoriously difficult, as the proliferation of textbooks and college courses on Critical Thinking suggests. One trouble is that the variety of seemingly good (but in fact bad) arguments is considerably wider than the variety of diagnostic names we have for them. Moreover, this variety is itself continually moving and growing. Our apparently endlessly creative linguistic capacities occasions a similarly capacious field for the creation of new forms of fallacious argumentation. As a consequence, it is often only in retrospect — after the debates are over, votes are cast, and decisions are made — that the illusions can be revealed for what they are. And with the proliferation of communication outlets and argumentative forums, argumentation theory can hardly keep up.

“It is often only in retrospect — after the debates are over, votes are cast, and decisions are made — that the illusions can be revealed for what they are.”

One reason why our theories have a hard time keeping up is that our best models of argument take them centrally to be dialogues, between two people or parties, who exchange reasons, each with the purpose of changing the other’s mind. But this dyadic (two-sided) model is no longer fitting. It leaves out of its purview the fact that argumentative dialogues, especially when occurring by means of modern technological mediation, are performed largely for the sake of onlooking audiences. The two discussants may reply to each other, but their objective is actually to move the audience. Once we see this triadic (three-sided) structure to political argumentation, many otherwise strange phenomena start to make sense.

Consider the textbook straw man fallacy. In the case of the straw man, one takes one’s opponent’s view and restates it in a form that is more easily criticizable. One then goes after the new (and worse) version of the view with justly critical lines about it. And then one closes the discussion. For sure, this would not convince the opponent, as they would only say that this criticized version of the view is not their own. But a straw man argument can move an onlooking audience, those who may not be familiar with the issue under debate, who may not be particularly sympathetic with one side, or who may just be looking for a moment of easy clarity with the issue. The straw man strategy gives them what they are looking for.

Consider, further, that much of the textbook vocabulary concerning fallacies has made it into the vernacular. One of particular note is that of the ad hominem, the fallacy of inferring that someone is wrong from the fact that they exhibit some irrelevant personal vice. It is common to find in popular political discourse charges of the ad hominem. To be sure, the prohibitions on name-calling and insulting an interlocutor’s person in the midst of a debate is very old; but now that we have a name for the prohibited strategy, we have a critical tool to invoke in the midst of argument. That the vocabulary of “ad hominem” and “straw man” has entered the political vernacular means that public arguers have additional means with which to sort the good reasons from the bad.

But notice that when the argumentative strategy of invoking a fallacy name is used, it occurs as yet one more move in the developing argumentative exchange. One invokes the straw man or the ad hominem as a way of criticizing one’s interlocutor. So, when Donald Trump was criticized in the Republican Primary debates for his use of the ad hominem, he interpreted the criticism simply as more naysaying to contend with in the argument; he did not take the criticism to be targeting his style of engagement. Three specific examples are of note. The first is when Senator Rand Paul (R-KY) chastised Trump for insulting people for their looks. Trump responded, “I have never attacked Rand Paul on his looks . . . . And believe me, there’s plenty of subject matter there”. The second is when Senator Ted Cruz (R-TX) impersonated Trump to open the Iowa debates (Trump was absent). Cruz began: “Let me say that I’m a maniac, and everyone on this stage is stupid, fat, and ugly. And Ben (Carson), you’re a terrible surgeon . . . . And now that we’ve got the Donald Trump portion out of the way . . . . ”. Third and finally, at the very late Texas primary debate, Trump acknowledged the effects his attacks on others, and noted, “So far, I cannot believe how civil it’s been up here”.

What is troubling here is that in these moments the vocabulary for criticizing the mode of argumentation has become merely one more tactic in the argument, one more way to discredit another’s views. The whole point of developing the diagnostic language of fallacies is to create a vocabulary with which we can argue about the argument itself, rather than the first-order claims at issue within it. In particular, the Cruz moment is important, because, in accusing Trump of relying on the ad hominem, Cruz is himself attacking Trump, the person. That is, the impersonation of Trump’s ad hominem is itself a form of ad hominem. And so, instead of being a tool for evaluating the mode of the debate, the fallacy-charge falls back into the debate itself; it becomes yet one more tactic internal to the dispute. Trump obviously sees this point very clearly, as he plays with insulting (by not insulting) Rand Paul, and goes out of his way to acknowledge and congratulate himself that he (and the others) haven’t been name-calling in the Texas debate.

We argue in our natural languages, and so often when we argue, we argue over economies, animals, environments, poverty, and so on. But arguments are structured collections of statements that are alleged to manifest certain kinds of logical relations; consequently, they, too, can be the subject of scrutiny and disagreement. And often in order to evaluate a claim about, say, poverty, we need to attend specifically to the argument alleged to support it. In order to discuss arguments, as arguments, we must develop a language about the argumentative use of language. That is, we must develop a metalanguage. The objective in developing a metalanguage about argument is to enable us to talk about a given argument’s quality without taking a side in the debate over the truth of its conclusion. Accordingly, with the metalanguage in place, we can assess the quality of a given argument without reference to our own view of the matter under dispute. Among other things, the metalanguage enables us to criticize the arguments offered by people with whom we agree, and it similarly allows us to recognize that sometimes a powerful argument can be produced for a conclusion that we know is false. But perhaps most importantly, the metalanguage enables certain crucial self-critical assessments; it is by means of the metalanguage that we can assess our arguments as lacking without thereby adopting a skeptical stance with regard to our own first-order positions. Put more simply, it is by means of the metalanguage about argument that we can stand above the fray of our first-order disputes, as it were, and ascend to a relatively impartial plane from which to assess not the matter under dispute, but the dispute itself. One might say that rationality itself depends upon our ability to competently wield a metalanguage about reasoning, debate, and argument.

The problem is that when the concepts of the metalanguage are used as first-order tools in an ongoing argument, the impartiality of the metalanguage is dissolved. And so with Trump, the language of fallacies became for all involved in the debates yet one more competing view about which to wrangle. The metalanguage for assessing the mode of dispute was dragged into the dispute itself, and, predictably, the rationality of the exchanges dissolved precipitously.

The owl of Minerva flies only at dusk. Only after the day is done, after the argument is over, are the tools of wisdom available. The tools of argument assessment, when applied in the midst of the argument, are mistaken for, or are appropriated as, yet more first-order claims. They are entered into the breach, and so can no longer assess it. Subsequent debate flies free of evaluative and rational constraints. And what passes for argument then is mere power.

Call it the owl of Minerva problem. Our vocabulary for assessing public argument must run behind the practices of public argument. By the time our evaluative tools are fashioned, the tactics they were designed to diagnose have already had their effect. What’s worse, the metalanguage for assessing argumentation needs to be somewhat technical, precise, and somewhat unwieldy. This renders the concepts in the metalanguage especially vulnerable for first-order deployment. All of this poses an acute difficulty for deliberative democracy. Citizens in a deliberative democracy must be able to argue well. And in order to argue well citizens not only need to be able to share, exchange, and evaluate first-order reasons; they need also to be able to scrutinize each other’s argumentative performances. That is, in a deliberative democracy, citizens need to develop a command of an adequate metalanguage about argumentation. However, in the rough-and-tumble of instant online political debate and constant inducements to “sound off,” “call in,” and “be heard,” there is little occasion for stepping back and reflecting on the prevailing modes of political argument; the metalanguage enters the political vernacular almost always as a first-order deployment, and so is most often already distorted. The deeper and dimmer worry is that proper public argumentation is too demanding for creatures like us, since it requires that we be double-minded about ourselves in the midst of argument about things that matter most. Perhaps there’s a slight consolation in the thought that the bleakest implications of the considerations offered above would challenge not only deliberative democracy, but political association as such.