1. The Proselytizer and the Autodidact

On a particularly austere view of philosophy, philosophers are taken to be in the business of advancing arguments. Typically, the goal of such arguments is to rationally persuade or convince: to get one’s interlocutor, real or imagined, to consider and ultimately endorse (or give up) some philosophical commitment or other. Of course, not just any argument will do. Arguments require an audience; their intended recipients must feel their pull; they must find them plausible and compelling, which is to say, they must find each of their premises plausible and compelling. If an argument is only as strong as its weakest premise, then a strong argument is one such that its premises—all of them—are highly plausible (or at least that its intended recipient finds them so).

Descartes famously took this challenge to heart, setting out to construct arguments of epic proportions, arguments of such a kind that he doubted “there is any path open to human intelligence along which better ones can ever be found” (Descartes, 1641 [Reference Descartes2008], 5).Footnote ¹ In this way, we might see Descartes as in the business of providing knockdown arguments: deductively valid arguments whose conclusions compel assent on pain of irrationality.Footnote ² The game of providing such arguments—what Bergmann calls “airtight arguments”—arguments that would “persuade radical skeptics (by rational force!) to change their minds” (146)Footnote ³ is still very much alive in philosophy today. For some, it is what gives philosophy its distinctive rigor. Bergmann calls this way of settling philosophical disagreements the proselytizing approach:

That is the goal of the proselytizer, anyway. After all, it is not often that our opponents are satisfied.Footnote ⁴ Descartes, for all his ambitions, was certainly no exception here.

Perhaps, then, proselytizing ought not be the philosophers’ goal. Perhaps we should, with Friedrich Waismann, recognize that words like “argument” or “refutation” or “proof” are “dying words” in philosophy:

On Waismann’s view, philosophers mistakenly see themselves as akin to theorem provers, when really “what philosophers are concerned with is something different—neither discovering new propositions nor refuting false ones nor checking and re-checking them as scientists do” (1968, 2). According to Waismann, what philosophers are ultimately concerned with is articulating their “vision,” even if that vision is dressed in arguments.Footnote ⁵ Ultimately, the essence—and true power—of philosophy “lies in its freedom,” enabling us to “change our angle of vision” rather than to compel assent by sheer rational force (1968, 21). Arguments are just a hairsplitting distraction.

Clearly, for Waismann, the aim of philosophy is not to proselytize. Interestingly, neither is it for Bergmann, though for different reasons—a parallel I will return to in Section 4. He discloses that his book’s main aims are not to show how his approach—intuitionist particularist noninferential antiskepticismFootnote ⁶—satisfies or persuades the skeptic about the external world (146). Nor are his aims to persuade antiskeptics who do not share his specific views (of course, converts are more than welcome).Footnote ⁷ For Bergmann, any such effort amounts to a kind of proselytizing, which he regards as a largely “hopeless” (145) endeavor.

Rather than proselytize, Bergmann’s aims are instead to focus on “getting one’s own antiskeptical house in order,” which he understands as “teaching oneself how best to think about these things” (147). This is what he calls the autodidactic approach, which has two main aims:

At the heart of the autodidactic project, then, lies a careful examination of our philosophical intuitions concerning the various principles, premises, and views involved in both (i) and (ii). On Bergmann’s account, intuitions are thought of as intuitional seemings Footnote ⁸: a seeming that a proposition p is true or false, or that a philosophical principle is justified or unjustified, rational or irrational. Bergmann refers to our intuitions about skeptical and antiskeptical propositions and principles more specifically as epistemic intuitions—intuitions about epistemic value—just as he treats intuitions about moral value as moral intuitions.

Intuitions constitute evidence. While intuitions need not be consistent in the same way beliefs must—for example, there is nothing incoherent in having both a seeming that p and a seeming that not-p—some intuitions are rationally held, and some are not.Footnote ⁹ Hence, it is rational on Bergmann’s view to reject an argument’s premise if, upon very thorough reflection, it conflicts with one’s most strongly held philosophical intuitions. This is Bergmann’s antiskeptical strategy. Even if we do not know which premise of the skeptical argument to reject, we know that the conclusion is false because our epistemic intuitions—e.g., that we know a lot—are comparatively stronger than the intuitions we have in favor of their denial. Our most strongly held antiskeptical intuitions are evidence against the truth of such views.Footnote ¹⁰

Now, these moves may ultimately fail to persuade the skeptic (or the antiskeptic, for that matter), but it is important to recall that persuasion—the motive driving the proselytizing approach—is not Bergmann’s aim. As autodidacts, we curb the temptation to persuade or satisfy other philosophers; these are not our goals. Our goals are instead to consider, by our own lights, relying on our own intuitions, what is appealing or unappealing about various philosophical commitments.

2. A Parity Problem for the Autodidact

Bergmann’s philosophical methodology raises some interesting questions. Here, I focus on his reasons for finding the proselytizer’s approach “hopeless”; I worry that these same reasons also undercut his case for favoring the autodidact approach.

Consider, then, why Bergmann takes the proselytizing approach to be hopeless in the first place: persistent, long-lasting dissensus in philosophy gives us reason to doubt that philosophical disputes can be rationally settled by argument (146). One natural elaboration of this thought is that the belief-forming method of the proselytizer—understood as the attempt to inspire rational belief change through philosophical argumentation—is epistemically unreliable.

Roughly speaking, we can think of a reliable method as one that, when employed in normal circumstances by normally functioning individuals, yields a higher ratio of true beliefs to false ones.Footnote ¹¹ One plausible test of reliability is whether a method produces consensus among the many individuals (philosophers, in this case) who competently use it over time. Of course, consensus does not guarantee truth; but the idea is that, under normal epistemic conditions, a method that reliably produces true beliefs should eventually lead to convergence on those truths among those who competently employ it. But—so the reasoning goes—philosophy, historically speaking, has produced very little consensus. Philosophical arguments rarely inspire rational belief change.Footnote ¹² The proselytizing approach might be more successful if history offered more cases in which competent disputants were rationally compelled by philosophical arguments (which is just another way of saying that knockdown arguments in philosophy were more common). But such instances are few and far between, if they exist at all. Thus, the proselytizing approach appears to be an unreliable guide to philosophical truth and, therefore, “hopeless” as a means of rationally settling philosophical disagreement.Footnote ¹³

But now a worry arises. Consider a central feature of the Bergmannian view: there is no intuition-independent position from which philosophical arguments can be motivated (cf. 259–260). Belief or disbelief in an argument’s premises ultimately rests on our philosophical intuitions. The success or failure of such arguments therefore depends on what intuitions we have and whether they are epistemically rational or irrational—occurring as they epistemically should or should not (see note 9). This, in fact, helps explain the persistent dissensus in philosophy: intuitions vary wildly, and a conclusion can always be resisted by rejecting the intuitions underlying one or more premises, or by discovering that one’s contrary intuitions are simply stronger. In this light, it is hardly surprising that robust convergence in philosophy is so rare.

And so, herein lies the problem: the proselytizer’s and the autodidact’s methods appear to rely on the same source—philosophical intuitions. But if the proselytizer’s method is deemed unreliable—“hopeless”— because it ultimately rests on intuitions which fail to produce consensus, then the autodidact’s method, resting on the very same source, must also be equally unreliable. The ground the autodidact stands on appears, then, to be no firmer than the proselytizer’s.Footnote ¹⁴

Now, one way Bergmann might want to resist this conclusion is to claim not that philosophical argumentation is an unreliable belief-forming method per se, but that the method is often incompetently employed. The idea is that while philosophical argumentation is not unreliable in itself, philosophers often incompetently employ it.

On its face, this may seem like an innocuous rejoinder: philosophers are not perfectly rational agents and can certainly misuse a method. But the claim becomes more contentious when we are reminded what, for the Bergmannian, this unreliability stems from—namely, the faulty intuitions of philosophers, intuitions that occur when they epistemically should not or that otherwise lead to beliefs that are not “fully externally rational” (252). Consider how, for example, Bergmann puts the difference between the intuitions of the skeptic and noninferential antiskeptic:

The “better evidence” here consists of the noninferential antiskeptic’s intuitions. The upshot is that the bad track record of philosophy is not due to an unreliable method but to facts about the philosophers employing it—more specifically, facts about their intuitions. Bergmann’s own intuitions, after all, have reliably led him to reject skepticism because they arise from properly functioning, truth-aimed cognitive faculties; they arise when they epistemically should.Footnote ¹⁵

This, however, only deepens concerns about the reliability of philosophical intuitions, for it seems to suggest that the Bergmannian has special reason to think their intuitions are more reliable or truth tracking than others. Yet, assuming normal epistemic conditions and properly functioning cognitive faculties, it is difficult to see how this could be; how, that is, one could accuse another of incompetently employing their intuitions without, in effect, presupposing that one’s own intuitions are by default truth aimed (a judgment which itself, according to Bergmann, turns on further intuitions).

This is what seems to transpire in Bergmann’s face-off with the skeptic. For him, the skeptic is a paradigm example of a philosopher whose beliefs, while internally rational—in the sense that their beliefs are coherent, explanatory, and not viciously circular or contradictory—fail to be externally rational, because their beliefs do not correspond to the way things really are (233). The skeptic, he might put it, therefore incompetently employs the autodidactic method. But the skeptic will accuse Bergmann of much of the same. Dissensus in philosophy, then, is a lot worse, and a lot more entrenched, than we thought. The worry, then, is not merely that disregarding—or significantly discounting—consensus as a data point risks epistemic isolation for the autodidact, but rather that, since everything ultimately turns on intuitions from the Bergmannian standpoint, the evidential weight of consensus or dissensus cannot be determined any other way.

I worry that we can already see this at work in some of Bergmann’s discussions related to disagreement. When Bergmann, for instance, claims that the skeptic’s intuitions are “illusory epistemic intuitions” (217; cf. 125, 222), he does so on the basis of further intuitions alone. The issue becomes especially sharp when we turn from antiskeptical allies, like the disjunctivist, to perceptual skeptics, who—as Bergmann readily acknowledges—would claim that their evidence is superior and that it is Bergmann (and other antiskeptics) who are objectively irrational. With intuitions as our anchor, the only available response seems to be, as Bergmann puts it, that the skeptic has simply misjudged their ability to get at the truth. His basis for this judgment is, unsurprisingly, “The epistemic intuitions I have as a particularist noninferential antiskeptic” (248).Footnote ¹⁶

Of course, for Bergmann, this line of questioning misses the point. Consider how, on his view, one’s intuitions might change: you may come to see that the counterintuitive implications of your view are worse off than those of your opponent’s. Naturally, what counts as “worse off” here will inevitably depend on one’s epistemic intuitions about rationality and other related notions. But aside from outright contradiction (unless like Graham Priest we have strong dialetheist intuitions that support the existence of true contradictions), it is unclear what, rationally speaking, would compel, for example, a disjunctivist, to think that their rejection of the New Evil Demon intuitionFootnote ¹⁷ is somehow worse-off, less intuitive, or less rational, than Bergmann’s acceptance of it.

Bergmann is aware of this deadlock but remains unfazed. No philosopher has a way of settling philosophical disagreements in a manner that would be recognized by all parties as objectively satisfying. He does not see this as a pessimistic conclusion but as “realism about philosophical dialectic” (manuscript, 8). I agree that such a view need not entail pessimism, and I share Bergmann’s realist outlook. Perhaps, then, I have been unduly critical of his approach. Still, it is not unreasonable to think that some philosophical methodologies are better suited to sidestep or soften such deadlocks. The commonsense tradition in philosophy may be one unexpected contender. Let me elaborate.

3. Whither Common Sense?

The commonsense tradition is, of course, no monolith. But let me register what I take to be one of its central lessons, namely, that what makes commonsense beliefs (or commonsense knowledge) “common” in the first place is that such beliefs possess a certain publicity. Commonsense beliefs are public beliefs, shared touchstones that ground our inquiries and make communication possible; they are manifest in our everyday actions. Some, such as Susan Stebbing, have argued that even science cannot wholly dispense with common sense. “The validity of physics,” she writes, “rests upon the assumption that there are external objects, i.e., that propositions such as ‘I am now seeing the sun’ are true” (Stebbing, Reference Stebbing1929, 159).Footnote ¹⁸

The evidential power of common sense, limited though it may be, lies precisely in its publicness. The evidence I have that this table exists, that it is solid, is simply that here it is before me, resisting my weight as I type atop it. This evidence is accessible to you and others; it can be corroborated, confirmed, or verified in some manner. Indeed, as Moore once wrote, “That I am seeing [a table] needs no proof to me. If someone were to come & fail to see one, that would be a reason for me to doubt….” (Moore, Reference Moore, Moore and Lewy1962, 116).

While someone’s failure to see the table before me would give me reason to doubt my faculties—or theirs, depending on the case—someone’s failure to share my intuitions, whether skeptical or antiskeptical, would not serve as a reason for me to doubt my own. Nor would it be a reason to think my intuitions are truth-tracking: they are, after all, just my intuitions, and lack the public, confirmable status of common sense. We might think that the vindicating power of common sense lies precisely in this publicness. Intuitions, by contrast, understood as evidence, lack this intersubjective grounding. One can hear Moore exclaim that he is more certain that he has hands than he is of any strong perceptual intuitions regarding them.

Bergmann sees his project as continuous with the commonsense tradition. Reid and Moore—two quite different philosophers of common sense, no doubt—are the philosophical heroes of his book, and Bergmann appeals to different aspects of their thought throughout. Intuitions are supposed to help strengthen this commonsense line.Footnote ¹⁹ However, as I see it, by treating intuitions as evidence, or by attributing them substantial evidential weight, Bergmann departs from this tradition by transforming evidence into a more inward-looking notion, rather than reinforcing its public and intersubjective character as emphasized by the commonsense tradition. This shift from the propositions of common sense to intuitions about them risks further entrenching philosophers within their own perspectives. The intuitionist particularist noninferential antiskeptic might indeed be on the side of common sense, but by making intuitions the center of attention, the approach may ultimately undermine the very accessibility and communicability that give common sense its philosophical edge.

4. The Limits of Argument

If I have understood Bergmann correctly, then I think he is probably right that, at the end of the day, perhaps all we are really doing in philosophy is affirming or reaffirming whatever intuitions we happen to hold.Footnote ²⁰ The modesty of this view, however, conceals something more epistemically immodest that I cannot help but recoil from: that some philosophical intuitions are just objectively more rational (externally rational) than others, that they may serve as evidence—as opposed to mere data—for the truth of our philosophical beliefs.

My unease comes into sharper relief when considering Bergmann’s own account of the possibility of rational change in intuitions. Although he is not particularly optimistic about the likelihood or frequency of such change, he does acknowledge that intuitions may shift “in clarity of focus, and in overall import”:

In this way, even if a rational change in intuitions is difficult to achieve, it can still be fruitful—particularly for the skeptic—to engage in discourse that prompts both sides to continually reassess their views. Such exchanges, Bergmann suggests, “might go some way toward helping skeptics to see that their skeptical arguments aren’t as strong as they may have supposed” (149–150).

While I agree that engagement of this sort is valuable, Bergmann’s remarks here strike me as overly optimistic. If, as he acknowledges, the skeptic is just as ready to dismiss the antiskeptic’s intuitions—even while conceding their appeal in the course of such exchanges—it is hard to see how merely setting out those intuitions could prompt them to recognize weaknesses in their own position. While I do not think such a shift is impossible, I have a hard time understanding how it could come from purely rational considerations alone. More plausibly, it would stem, to my mind, from a gradual shift in perspective or an openness to alternative frameworks—a “change [in] our angle of vision,” as Waismann put it.

On this point, it is worth returning to Waismann. Earlier, I introduced his picture of philosophy as a practice whose essence lies in “changing our angle of vision” rather than in proof or refutation. That vision aligns, in a striking way, with Bergmann’s autodidact method—though Bergman would almost certainly reject the comparison—since both thinkers reject the proselytizing approach and, as I see it, emphasize the profoundly personal nature of philosophical inquiry.

Recall that, for Waismann, philosophy’s freedom lies in changing our angle of vision—not in following a chain of logical proofs, but in achieving “a new and broader way of looking at things” (1968, 32). Bergmann likewise denies that the transformation involved with “teaching oneself how best to think about these things” (147) comes through sheer argumentative force. Instead, he emphasizes sustained self-teaching: careful reflection on which of one’s intuitions—my intuitions, your intuitions—are strongest, weakest, most rational, or less rational. In this narrow respect, their pictures converge. For both, the ultimate aim of philosophy is to determine, for oneself, how best to think about things—a project bound up, one might think, with articulating one’s philosophical vision.

This, I submit, is one of the most valuable aspects of Bergmann’s introduction of the proselytizing/autodidact distinction: it acknowledges something that I think philosophers should confront more directly—the limits of rational argument in our discipline. In that respect, I see a certain kinship, however uneasy, with Waismann’s conception of philosophy as the cultivation of vision rather than conquest of opponents by proof. Both Bergmann’s and Waismann’s approaches, in their different ways, grapple with what philosophical inquiry can meaningfully accomplish once we accept that rational persuasion often falls short.

Where they diverge—among other deep differences—is in what counts as philosophical progress. For Bergmann, how we get our antiskeptical house in order may be a deeply personal affair, but it is only successful, epistemically speaking, if one’s intuitions eventually crystalize into antiskeptical, noninferentialist, particularist form. The Bergmannian autodidact insists that some intuitions are objectively more rational than others and that these should guide us toward true philosophical beliefs—such as the falsity of skepticism. By contrast, the Waismannian autodidact, as I am construing them here,Footnote ²¹ refrains from such judgments and is content to let differing intuitions stand on equal epistemic footing.

For Bergmann, of course, resisting the skeptic’s premises on the grounds that they conflict with our (externally rational) intuitions is no more mysterious than resisting belief in a proposition when one’s evidence for its negation is stronger. While I cannot share his confidence in this idea—I lean more Waismannian than Bergmannian here—I am nonetheless certain that his intuitions have yielded philosophical insights well worth our careful consideration. And that, I assure you, is more than just a matter of it seeming to me so.