The Unmaking of Wisdom

Andy Norman

How we compromise reason’s capacity to transform the human condition – Part 2: Recovering Reason

Part 1 of this two-part essay identified a common assumption about reason and then traced its origin and uptake. Plato suggested that reason consists in a judgment’s being supported with sufficient evidence, and this idea went on to decisively shape more than two thousand years of epistemological inquiry. The “Platonic model,” it was suggested, is actually the root cause of our collective failure to make significant progress toward achieving wisdom. Not only does it block more promising attempts to understand what it is to be reasonable—thereby depriving us of a clear view of a critical end—it actually breeds dismissive strains of skepticism, relativism, and dogmatism.

In Part 1 of this essay, I told the story of how a plausible but ultimately mistaken picture of reason seduced generations of philosophers. I’ve tried to show how this picture became an article of rationalist faith, how it created a set of persistent problems, and how it sidetracked the search for a genuinely enlightened and enlightening understanding of reason. The same misconception, I’ve argued, is the root cause of rationalism’s loss of traction in the wider marketplace of ideas. Uncritical acceptance of the Platonic model prevents rationalists from effectively promoting reason as a way of life, and from realizing reason’s vast promise to improve the human condition.

The diagnosis can be encapsulated like this: competent thinkers treat the need for reasons and evidence as situational, and they recognize that many claims stand in need of support but rightly treat many others as presumptively reasonable. Then the Platonic model comes along and characterizes reason as invariably conferred (or imparted). We accept the model for what look like good reasons and unwittingly generate an artificial and indiscriminate demand for backing. This effectively banishes presumptions from the space of reasons, rendering all judgments nonpresumptive, or “guilty until shown to be innocent.” Plato’s model, in other words, effectively puts the burden of proof on the claimant and does so no matter what the claim. But such a model renders any premise one might offer in support of a claim in similar need of support. It thus proves impossible—or at least unduly difficult—to make any dialectical headway. Reason regresses and skepticism looms because the Platonic conception rigs the reasoning game in the challenger’s favor.

This diagnosis allows us to understand why radical skepticism remains a persistent worry. It shows us why facile relativisms continue to flourish and how a faith-based dogmatism draws sustenance from the prevailing understanding. It explains the strange divorce between theory and practice and helps us grasp why scientists tend not to take epistemology seriously. It sheds light on rationalism’s periodic crises of integrity and illuminates the root cause of its failure to become a significant cultural force.

Treating the Root Cause

The dilemma of ultimate commitment articulates a line of thought that can make dogmatism seem the only alternative to skepticism. For either we leave everything open to question and the skeptic wins, or we treat some beliefs as immune to questioning—and thereby become dogmatists. Many take this one step further: because it is better to stand for something rather than nothing, unquestioning loyalty to certain beliefs is actually a virtue. The argument is fallacious, but why exactly?

To achieve clarity on this issue, we must give a name to the kind of question that drives regress skepticism. What sort of question is it that the skeptic employs in the imagined scenario? In Hume’s version of this scenario, the question is “Why?” The linguistic form of the gambit is immaterial, however: it could just as well be “How do you know?”; “What makes you think so?”; or “Prove it!” What’s essential is that the gambit suspend entitlement to the claim it targets—that is, render it inadmissible until such time as its rational standing has been redeemed. (It is assumed that such a gambit invariably places burden of proof on the claimant, but I will contest this assumption.)

At this point, it proves useful to introduce a little terminology. Let’s agree to use the term challenge to designate reasoning-game moves meant to take issue with claims or assumptions. More precisely, challenges are meant to suspend entitlement to the claims they target: they serve notice that the claim is viewed as problematic, that the claimant has reason-giving work to do before the challenger will admit it into evidence. Now, within the class of challenges, some present reasons against the claims they target, and some do not. We can call those that do “onus-bearing” challenges and those that don’t “bare.” Bare challenges, then, do not elaborate on the source of the challenger’s resistance to the claim: they simply demand that the claimant provide grounds (on pain, presumably, of losing the contested entitlement). Thus, where bare challenges call on a claimant to make good on a presumed burden of proof, onus-bearing challenges shoulder a presumed burden of disproof.

It is evident, I think, that the regress skeptic is imagined to employ bare challenges—iterated demands that the claimant provide supporting reasons. (“If I ask why you believe any matter of fact that you relate, you must provide some reason.”) Need we fear the prospect of indefinitely iterated onus-bearing challenges? No, for there is no reason to think that the grounds for doubt needed to power such challenges are indefinitely available—any more than there is reason to think that grounds for belief go on ad infinitum.

We can now pose the key question: can bare challenges be iterated indefinitely and successfully saddle the claimant with a burden of proof each time? The answer should by now be clear: not if our picture of the space of reasons makes room for presumptions. For as long as some claims are presumptive (that is, rational by default in the absence of telling reasons against them), a claimant can in principle shift the burden of proof by invoking presumptive premises. If the challenger lacks the grounds for doubt needed to shift the onus back, he or she will often (and properly) concede the point. Thus does real reasoning come to an end: not with first principles, articles of faith, indubitable beliefs, perceptual judgments, or synthetic a priori truths but with ordinary presumptive premises. Of course, challengers sometimes muster grounds for doubt, and the game then continues. Eventually, though, presumptive premises will put one player or the other in a position to win. Should a challenger again demand a warrant, even after the burden of (dis)proof has shifted to him, the claimant is well within his rights to counter: “It seems reasonable enough. Why do you doubt it?” Under such conditions, the reply suffices to disarm the challenge.

Presumptions, then, turn out to have an interesting property: they are immune to bare challenge. The argument is straightforward. Presumptions are, so to speak, rational by default. Hence it is not up to the claimant to provide reasons for them; it is up to the challenger to provide reasons against them. The burden of disproof, in other words, rests on the challenger, and bare demands for supporting evidence are inappropriate. They don’t arise. The fact that they may sometimes be turned aside with a simple “Why do you doubt it?” shows this. Thus, when they target presumptions, bare challengers mislocate the burden of proof. Presumptions are bare challenge immune practically by definition.

This can seem like a radical suggestion, but it is really just common sense. After all, were the right to issue bare challenges (and suspend entitlement thereby) truly unlimited, the challenger would have a formulaic strategy for undermining any entitlement whatsoever: just iterate bare challenges until the claimant gives up. Some claims must be bare-challenge immune, or the reasoning game would be degenerate.

Isn’t it imperative, though, that we not place claims beyond the reach of questions? Might we not, by acknowledging bare-challenge immunity, inadvertently sanction dogmatism? Actually, no: acknowledging this commonsense kind of challenge immunity lends no comfort whatever to closed-mindedness. Recognizing that some claims are rationally immune to bare challenge does not place any claim beyond the reach of all challenges. It’s only the latter, not the former, that we need fear. Open-mindedness is important, and it requires that we leave everything open to possible undermining by onus-bearing challenge, but it does not require claimants to bear the burden of proof in all circumstances. It can be fully rational, in some circumstances, to reject demands for evidence.

Philosophers and rationalists have understandably shied away from the idea of challenge immunity. The idea that it might be rational to reject a demand for evidence has been similarly taboo. We have seen, however, that we must confront and overturn these taboos, for a certain kind of challenge immunity is actually constitutive of reason: if our reasoning practices are to maintain their integrity, certain demands for evidence must be rejected. It’s the only way to bring what William James called “skeptical balance” to our understanding of reason. “Question everything” might be a decent recipe for cleverness, but wisdom demands that we wield our questions more judiciously.

Unreason has long exploited an unresolved tension in rationalism to provide a convenient excuse for closed-mindedness. I have rendered this excuse in the form of an argument that I call “the dilemma of ultimate commitment.” The notions of presumption and bare challenge, however, allow us to see precisely why this argument fails. For it is now plain that the argument relies on an equivocation—a surreptitious slide from the notion of a claim’s being immune to bare challenge to that of a claim’s being immune to all challenges. With the distinction in hand, however, the problem simply dissolves: skepticism is avoided because presumptions halt the regress of bare challenges, and dogmatism is avoided because presumptions, like all claims, remain forever open to possible undermining by onus-bearing challenges.

Thus fortified, we understand better what it is to be openminded and see clearly that it does not require us to embrace relativism, compromise rational standards, or lack conviction. We can take reason seriously without fear that serious examination will unravel its requirements. Logical considerations do not preclude genuine, thoroughgoing open-mindedness; rationality need not allow, let alone require, dogmatic commitment. Thus is closedmindedness deprived of an influential excuse, and a major obstacle cleared from the path to collective wisdom.

The New Socratic Model

But if groundedness is not the hallmark of rationality, then what is? How should we conceive of reason? What standard should a follower of reason apply when measuring the worth of a claim to knowledge? In what follows, my aim is to provide not a final answer but a decent approximation of one. The idea is to point the way to a more adequate conception—to suggest where philosophers might focus their efforts to articulate a rigorous understanding of reason, and to suggest how the rest of us might best understand its requirements in the meantime.

The way forward, it turns out, begins with a return to Socrates. By elaborating the Socratic model of reason just a little, we arrive at a simple, viable, and genuinely wisdom-conducive alternative to the Platonic model. This alternative, I think, has all the qualities of a suitable successor.

Socrates operated on the premise that a claim must withstand questioning to merit rational assent. But which questions are these? What kind of questioning must a claim withstand? Presumably, it is the questions that threaten rational permissibility that must be considered. I have used the term challenge to designate such questions and argued that they arise, not invariably, but on a situational basis. This suggests one way we might articulate a Socratic understanding of reason. We might say, for example, that a claim is reasonable just in case the challenges to it that arise (in that context) can be met. The explanatory task then fissions into two: first, we must explain the notion of a challenge’s arising, and second, we must clarify what it means to say that a challenge can be met.

The first task presents no serious difficulty once we have divided the class of challenges into those that are onus-bearing and those that are bare. For the former arise only when the requisite grounds for doubt are available. Bare challenges only arise when—as we have seen—the claim challenged is not presumptive. Thus, the task reduces to that of defining the class of presumptions.

Here we are free to experiment with any number of options. We could say, for example, that the class of presumptions encompasses common knowledge, or the elements of the epistemic status quo, or perhaps what a given community of inquiry—perhaps the community of scientific inquirers—happens to take as rationally permissible. Alternatively, we might say that it makes sense to treat as presumptive anything that the relevant experts endorse or anything that seems sufficiently likely, given common background knowledge. (I am not endorsing any of these proposals at present, just mentioning them for the sake of illustration.) Perhaps none of these proposals gets it exactly right, but any one of them would put us considerably closer to an adequate understanding of reason than the Platonic model, which treats the class of presumptions as empty and thereby makes for a degenerate reason-giving game. At any rate, it is not necessary at present to define the class of presumptions precisely, for presumption is a weak and easily overridden status. Though we need some notion of presumption to get the reason-giving game underway, the game itself has the final say on what remains, or becomes, a rational entitlement. This is as it should be: epistemic conditions, not epistemological theorizing, should have the final word.

As to our second task: What does it mean to say that the challenges to a claim can be met? It means that each such challenge can be turned away, either directly—by answering the question contained in the challenge—or indirectly, by showing that the challenge does not really arise. In both cases, the way to determine whether the response to the challenge meets it is to look and see whether it (the response) “undoes” the entitlement-suspending force of the challenge. Of course, this proposal relies on local contextual discernment. Again, though, this is precisely as it should be. Theory can direct attention to relevant considerations, but ultimately, situated practical judgment must be accorded room to do what it does best.

The “New Socratic” model of reason, then, can be simply put: claims to knowledge are reasonable just in case they can withstand questioning—that is, if they are so situated in the space of reasons that arising challenges can be met. This account has some nice features. For one, it squares with how reasoning, at its best, really goes. Also, the traditional, Platonic conception lives on as a limiting case of reason properly understood. For where the claim in question stands in need of backing—as controversial and interesting claims usually do—and the arising bare challenge can be met directly, the claim appears (and in a sense is) reasonable by virtue of a supporting reason. The New Socratic model thus subsumes the Platonic model in much the way that relativity
theory subsumed Newtonian physics. The New Socratic model is in essence a more encompassing truth.

Where the Platonic model struggled to account for basic beliefs, though, the New Socratic model accounts for them easily and elegantly. Claims are “contextually basic” when, in the context in question, sensible challenges to them do not arise: when onus-bearing challenges fail to arise because grounds for doubt are contingently unavailable, and bare challenges fail to arise because the claim is nevertheless presumptive. Basic belief, in other words, is simply the special case where “All arising challenges can be met” is trivially satisfied. Where no sensible challenges arise, it follows that all that do arise can be met. It is for this reason—and this reason only—that “Spring follows winter,” “Happiness is good,” “The sun will rise tomorrow,” “Emeralds are green,” and countless other claims are properly treated as basic. The New Socratic model handles basic belief without breaking a sweat.

These two features of the New Socratic model—its subsumption of the Platonic model, and its straightforward handling of basic belief—bode well. For as historians and philosophers of science have noted, successor paradigms are generally constrained by the need to cast a coherent narrative explaining their predecessor’s successes and failures. This is precisely what the New Socratic model provides: a compelling account of why the Platonic model could have seemed so natural and intuitive, and a clear view of where and why that model breaks down. The New Socratic model duplicates the successes of the Platonic paradigm, while also addressing the anomalies that long distressed that paradigm.

The rationalist project is not dead. Yes, a disastrous conceptual error hobbled it for two and a half millennia. But there is no reason why it cannot again become a powerful force for enlightenment and progress. Perhaps it will remake our collective understanding of reason and spark a movement that brings widespread wisdom to the challenges of tomorrow. But will this really happen? That depends. An adequate understanding of reason can only illuminate the path; it is up to us to travel it.

Sources and Further Reading

  • Bartley, W.W. The Retreat to Commitment. Chicago: Open Court, 1984.
  • MacIntyre, Alasdair. The Tasks of Philosophy. Cambridge: Cambridge University Press, 2006.
  • Norman, Andrew. “Regress and the Doctrine of Epistemic Original Sin,” The Philosophical Quarterly 47, No. 189 (1997).
  • Note: Specialists in epistemology will want to know whether contemporary coherence, reliability, and context theories of justification have moved beyond the Platonic conception of reason. In fact, they have not. See Bonjour, Lawrence, The Structure of Empirical Knowledge (Cambridge: Harvard University Press, 1985, p.18); Goldman, Alvin, “What Is Justified Belief?” in G. Pappas, ed., Justification and Knowledge (New York: Springer, 1979, p.2); and Toulmin, Stephen, The Uses of Argument (Cambridge: Cambridge University Press, 1958, pp. 11-12).


Andy Norman

Andy Norman is professor of philosophy at Carnegie Mellon University (CMU) and director of CMU’s Humanist Initiative. He is working on a book titled Unhinged: Faith, Ideology, and the Ethics of Belief, Boosting Mental Immune Response in the Ideological Animal.

How we compromise reason’s capacity to transform the human condition – Part 2: Recovering Reason Part 1 of this two-part essay identified a common assumption about reason and then traced its origin and uptake. Plato suggested that reason consists in a judgment’s being supported with sufficient evidence, and this idea went on to decisively shape …

This article is available to subscribers only.
Subscribe now or log in to read this article.