Epistemological Disjunctivism and Closure-Based Radical Skepticism
In the last chapter we explored the merits of epistemological disjunctivism, both as an account of perceptual knowledge, and also as a putative undercutting anti-skeptical strategy. We saw that this proposal fared well when it came to underdeterminationRK-based radical skepticism. In this chapter we will further explore the epistemological disjunctivist anti-skeptical strategy by comparing it to some other anti-skeptical proposals in the literature. While our initial focus in this chapter will remain on underdeterminationRK-based radical skepticism, we will also examine how epistemological disjunctivism performs when it comes the closureRK-based formulation of the radical skeptical problem. As we will see, epistemological disjunctivism struggles with this variety of radical skepticism.
1. Anti-skeptical Contrasts (I): Rational Support Contextualism
In chapter 2 we encountered the attributer contextualist response to radical skepticism, and found that proposal inadequate, in particular with regard to underdeterminationRK-based radical skepticism. Recall that this proposal argued that while assertions of knowledge ascription sentences express truths relative to normal contexts of epistemic appraisal (where the epistemic standards operative for “knows” are undemanding), assertions of those very same sentences in skeptical contexts of epistemic appraisal (i.e., contexts in which radical skepticism is at issue, and hence where the epistemic standards operative for “knows” are very demanding) express falsehoods. There is thus a sense in which radical skepticism is correct, in that the skeptical assertion that we generally lack knowledge will express a truth in the context of epistemic appraisal in which it is uttered. Even so, there is also a sense in which radical skepticism is wrong, for our everyday assertions of knowledge ascription sentences will tend to express truths.
The core difficultly that this proposal faces is that the problem of radical skepticism doesn’t seem to in any way trade on a raising of the epistemic standards. That is, if the radical skeptic is right, then we fail to satisfy any epistemic standard for knowledge, even relatively low standards. We saw that we can bring this problem into sharp relief by considering how attributer contextualism would respond to underdeterminationRK-based radical skeptical paradox.
Recall that we can express the underdeterminationRK-based radical skeptical paradox in terms of the following collectively inconsistent triad:
THE UNDERDETERMINATIONRK-BASED SKEPTICAL PARADOX
(SU1) One cannot have rational support that favors one’s belief in an everyday proposition over an incompatible radical skeptical hypothesis.
(SU2) The underdeterminationRK principle.
(SU3) One has widespread rationally grounded everyday knowledge.
The underdeterminationRK principle at issue in (SU2) was formulated as follows:
THE UNDERDETERMINATIONRK PRINCIPLE
If S knows that p and q describe incompatible scenarios, and yet S lacks a rational basis that favors p over q, then S lacks rationally grounded knowledge that p.
Sticking to everyday contexts of epistemic appraisal in which assertions of knowledge ascription sentences express truths, we can see how attributer contextualists will retain (SU3), at least where that claim is suitably understood. But now we can ask the question: in virtue of what do such assertions express truths? That is, what epistemic desiderata are satisfied by the target beliefs to make these assertions true? In particular, do these beliefs enjoy rational support that favors them over skeptical alternatives, contra (SU1)? If attributer contextualists deny this, then unless they are willing to reject the underdeterminationRK principle (and thus (SU2)) they are subject to underdeterminationRK-based radical skepticism, with all that this entails. But on what basis would attributer contextualists deny this principle? If, on the other hand, they argue that our everyday beliefs do enjoy such favoring rational support over skeptical alternatives—and, again, there is nothing in the attributer contextualist thesis itself that motivates this claim—then we will need to see the argument for this contention.
The point is that attributer contextualists need to explain which of (SU1) or (SU2) they reject, and why. Moreover, notice that insofar as attributer contextualists can motivate the rejection of either of these claims, then attributer contextualism will become an idle cog as an anti-skeptical strategy. After all, if we can make sense of why one of these claims is false, then that will suffice all by itself to block the underdeterminationRK-based radical skeptical paradox. Any additional appeal to attributer contextualism would be superfluous (except to the extent that it serves some diagnostic role in explaining how the skeptical problem comes about).
Note that it is of course the case that in asking this question—in engaging in this kind epistemological enterprise altogether—we are, according to the attributer contextualist, in a skeptical context of epistemic appraisal in which high epistemic standards apply, and hence any assertions of knowledge ascription sentences that we make will tend to express falsehoods. But we shouldn’t let this fact distract us, since it is not materially relevant to the issues at hand. We can recognize this point about how attributer contextualism applies to assertions of knowledge ascription sentences in skeptical contexts of epistemic appraisal while nonetheless asking what epistemic desiderata are satisfied by an agent’s beliefs in everyday contexts of epistemic appraisal where assertions of knowledge ascription sentences tend to express truths. And the crux of the matter is that insofar as the underdeterminationRK principle is in play and these beliefs fail to enjoy rational support that favors the proposition believed over skeptical alternatives—in line with (SU1)—then the epistemic standing of these beliefs has been called into question.1
That, in any case, is roughly the line we took on the attributer contextualist anti-skeptical proposal in chapter 2. Now that we have epistemological disjunctivism on the table, however, we can consider a variant on the attributer contextualist anti-skeptical line that might fare better on this score. The key difficulty facing attributer contextualism as an anti-skeptical strategy is that while it generates the result that assertions of knowledge ascription sentences in everyday contexts of epistemic appraisal will tend to express truths, it does not back this up with a plausible epistemological story about how these assertions could express truths. That is, and especially bearing in mind the challenge to the epistemic standing of our beliefs that is posed by underdeterminationRK-based radical skepticism, in virtue of what do the beliefs in question count as knowledge?
But suppose we reconceived of attributer contextualism as being primarily concerned with the nature of the rational support available for our beliefs? In particular, imagine a form of attributer contextualism that maintained that the strength of rational support available to a subject could vary in response to changes in the context of epistemic appraisal. Relative to everyday contexts of epistemic appraisal, a subject could be truthfully ascribed very strong rational support for her beliefs, but relative to skeptical contexts of epistemic appraisal, in contrast, the very same subject could at best be truthfully ascribed only a very weak level of rational support for her beliefs. In this way one could maintain that the epistemic standards relevant to assessing the truth of assertions of knowledge ascription sentences don’t vary from one context of epistemic appraisal to another, but that nonetheless whether or not these assertions express truths is a variable matter. Call such a view rational support contextualism.
In order for this proposal to fare better than a standard version of attributer contextualism, we need to suppose that in everyday contexts of epistemic appraisal the rational support that can be truthfully ascribed to an agent’s belief can be factive in broadly the manner set out by epistemological disjunctivism. In this way rational support contextualism can lay claim to some of the anti-skeptical benefits that we saw accrued to epistemological disjunctivism in the last chapter, at least as regards underdeterminationRK-based radical skepticism. What happens on this view when radical skepticism is introduced to the context of epistemic appraisal is that the rational support that can be truthfully ascribed to a subject’s beliefs shrinks to being merely nonfactive support (e.g., that it seems to one as if p, rather than that one sees that p). This is why although the epistemic standards for assessing knowledge ascriptions do not change, it is nonetheless the case that assertions of knowledge ascriptions sentences will tend to express truths in everyday contexts of epistemic appraisal and express falsehoods in skeptical contexts of epistemic appraisal.
This is broadly the view defended by Ram Neta (2002; 2003), although he describes his view in terms of evidence rather than reasons, a difference that we will set to one side here. Here is Neta’s description of this proposal:
Suppose that we are talking on the phone and I claim to know that my basement floor is all wet. You ask me what evidence I have that it’s wet, and I reply that I can see that it’s wet. Here, my evidence reaches all the way out to the fact: I can’t see that my floor is all wet unless it is, in fact, all wet. There is, in this case, no epistemic gap between my evidence and my belief.
Nonetheless, when we face certain skeptical challenges, we move into a context of epistemic appraisal in which we can truthfully say that our evidence fails to support our beliefs about the external world. Once we are in such a context, we can’t find evidence for those beliefs, for none of our mental states count as evidence for such beliefs in that context. Thus, by issuing or confronting skeptical challenges, we create an epistemic gap between our beliefs about that world and our evidence for them. (Neta 2003, 27)
Here we get the essentials of the view. Exchanging talk of evidence with talk of reasons, Neta is claiming that relative to everyday contexts of epistemic appraisal the rational support one has for one’s beliefs can be factive, in essentially the manner that epistemological disjunctivism proposes, but that relative to skeptical contexts of epistemic appraisal this rational support diminishes to the point of being nonfactive. This is why, relative to skeptical contexts of epistemic appraisal, the radical skeptic’s claim that we lack knowledge expresses a truth (because such nonfactive rational support fails to favor our everyday beliefs over radical skeptical alternatives).
By allowing subjects to be truthfully ascribed factive rational support in everyday contexts of epistemic appraisal, rational support contextualism can plausibly evade the concern we raised above regarding the standard attributer contextualist anti-skeptical strategy. In particular, on this view we can explain why the underdeterminationRK-based skeptical challenge doesn’t go through, since in everyday contexts of epistemic appraisal subjects can be truthfully ascribed rational support that is strong enough to conflict with (SU1). Thus the underdeterminationRK principle can hold, and yet radical skepticism can nonetheless be resisted. Moreover, this view brings with it the usual advantage of attributer contextualist accounts, in that it can explain why radical skepticism can seem so compelling. After all, on this proposal radical skepticism is in a sense entirely correct, in that relative to contexts of epistemic appraisal in which radical skeptical error possibilities are at issue assertions of knowledge ascription sentences will tend to express falsehoods.
When it comes to radical skepticism, rational support contextualism thus has a key advantage over standard versions of attributer contextualism, in that it incorporates a plausible response to the specifically underdeterminationRK-based version of the radical skeptical paradox. Nonetheless, there are a number of problems with rational support contextualism, many of them shared with standard versions of attributer contextualism.2
First, like attributer contextualism, there is the worry that it is too simple a solution to the problem in hand. Recall that the attributer contextualist treatment of radical skepticism invited the response that if this problem simply trades on a raising of the epistemic standards applicable to the evaluations of knowledge ascription sentences, then surely this is something that we would have spotted long before now. Could we really have been so ignorant about how such a central term as “knows” functions? The same kind of worry surfaces with regard to rational support contextualism. If the extent of the rational support one’s beliefs enjoy is straightforwardly responsive to contextual factors in this way, such that the problem of radical skepticism turns on this kind of context sensitivity, then isn’t this something that we would have realized long ago?
A second problem that rational support contextualism shares with attributer contextualism concerns how it endorses an epistemic hierarchy of contexts. Recall that attributer contextualism concedes that radical skepticism, qua position, is correct relative to the high epistemic standards that are operative in skeptical contexts of epistemic appraisal. The same is also true of rational support contextualism, though here this claim is mediated by an appeal to the context sensitivity of our attributions of rational support. This concession severely undermines the anti-skeptical credentials of the view.
Recall our discussion of Wittgenstein’s account of the structure of reasons from chapter 3, where we contrasted this proposal with the ordinary language response to radical skepticism, as found in the work of thinkers such as J. L. Austin (1961). We noted there Barry Stroud’s (1984) point that radical skepticism, qua paradox, is entirely consistent with the claim that there are significant differences between our everyday practices of epistemic appraisal and the kinds of epistemic appraisal undertaken by the radical skeptic. After all, it is open to the radical skeptic to contend that the latter is a purified version of the former, such that if only we applied our epistemic principles in a thoroughgoing fashion—that is, if only we weren’t impeded by practical limitations of time, creativity, and so on—then we would be led to adopt radical skeptic standards of epistemic evaluation. We noted that this was a point that Wittgenstein saw but the ordinary language response to radical skepticism misses. For while both emphasize the differences between our quotidian and radical skeptical practices of epistemic appraisal, only the Wittgensteinian account of the structure of reasons demonstrates that the latter simply cannot be a purified version of the former (indeed, that the latter is deeply incoherent, in that it aspires to an impossible ideal).
This point clearly has application to rational support contextualism (and, for that matter, attributer contextualism). For once we concede that radical skeptical epistemic assessments are relative to higher epistemic standards than normal epistemic assessments, then it is hard to resist the conclusion that radical skepticism is indeed offering us a genuine paradox. In particular, it seems to follow that relative to a purified version of our everyday standards of epistemic appraisal, a version that we have no way of insulating ourselves from, radical skepticism, qua position, is correct.
The contrast with the epistemological disjunctivist response to radical skepticism is telling in this regard. For what the epistemological disjunctivist is contending is that our everyday practices of paradigmatically treating the rational support agents have for their beliefs as factive is in fact entirely in order as it is. The story the radical skeptic tells us about why we should reject this quotidian picture turns out to trade on theoretical claims that the epistemological disjunctivist argues are highly dubious. There is thus no need to concede anything to the radical skeptic on this score, much less grant, as rational support/attributer contextualism does, that the radical skeptic effectively occupies the epistemic high ground.3
This point is clearest when it comes to the underdeterminationRK-based radical skeptical paradox. According to epistemological disjunctivism, it simply isn’t true that one of our fundamental epistemological commitments is that the rational support we have for our everyday beliefs fails to favor those beliefs over radical skeptical alternatives. One of the key claims that makes up the relevant radical skeptical inconsistent triad—namely, (SU1)—can thus be rejected. And notice that the rejection of this claim is not relativized to a particular context of epistemic appraisal. Rather, the contention is the underdeterminationRK-based radical skeptical paradox is illusory, in that it essentially turns on a dubious theoretical thesis that has been falsely presented as one of our fundamental epistemological commitments.
Notice that in making this point we needn’t deny that when one is in the grip of the problem of radical skepticism it can seem to one as if the rational support that is available to one has shrunk to the nonfactive (e.g., to the merely phenomenal). And of course, insofar as one believes that the rational support that is available to one is of this kind, then obviously one can no longer know everyday propositions in virtue of possessing the target factive rational support. The point, however, is that in such a case one is not now in the kind of paradigm epistemic conditions in which, according to epistemological disjunctivism, factive rational support is available.4 Moreover, notice that this is a completely self-imposed epistemic exile, in that there is in fact a strong philosophical basis for resisting this concession to the radical skeptic. That’s just the point of the anti-skeptical diagnosis that epistemological disjunctivism puts forward, that in conceding this much to the radical skeptic one is in effect allowing oneself to be hoodwinked. Thus although it is undeniably possible to get oneself into such an epistemically concessive state of mind, there is a sound philosophical basis to resist the charms of the radical skeptic in this regard, and hence we should do so.5
In responding to rational support contextualism and attributer contextualism in this way we are implicitly supposing that the view should be thought of as an undercutting anti-skeptical strategy, such that its appeal to everyday contexts of epistemic appraisal is meant to undermine the very idea that radical skepticism presents us with a genuine paradox (in that what the radical skeptic presents as one of our fundamental epistemic commitments is instead merely the product of faulty philosophical theory). And yet we could just as easily think of contextualist responses to radical skepticism of this kind as being overriding anti-skeptical strategies that offer us independent reasons to reject one of the core intuitive claims as issue in the radical skeptical paradox. (Indeed, we explored the option of conceiving of attributer contextualism along these lines in chapter 2.) Even then, I don’t think we will find the revisionist story put forward by these contextualist proposals as being compelling enough to effectively function as part of an anti-skeptical strategy. Moreover, as we noted in chapter 1, overriding anti-skeptical strategies are inevitably much less palatable than their undercutting counterparts. Hence if there is a plausible undercutting anti-skeptical strategy available—such as the kind of proposal offered by epistemological disjunctivism as regards the underdeterminationRK-based radical skeptical paradox—then there’s very little incentive to explore overriding anti-skeptical strategies.
A further problem facing rational support contextualism as an anti-skeptical proposal concerns the dialectical effectiveness of the appeal to contexts of epistemic appraisal. For notice that the anti-skeptical work is in fact done by the claim that relative to everyday contexts of epistemic appraisal agents can be properly said to be in possession of factive rational support for their beliefs. But if we can make sense of this idea at all, then why do we need to appeal to contexts of epistemic appraisal, at least as far as underdeterminationRK-based radical skepticism goes? After all, this idea all by itself will block this form of radical skepticism, as we noted above in our discussion of epistemological disjunctivism. So what is added by the appeal to contexts of epistemic appraisal?
The rational support contextualist will no doubt contend that the contextualist element of the view is required in order for the proposal to have diagnostic appeal. The thought would be that without this aspect of the thesis there is not a plausible story in play regarding why the radical skeptic’s assertions about our epistemic position can initially strike us as true. But we have already noted some deficiencies in the diagnostic element of the view, so this consideration is hardly decisive. Moreover, the relevant contrast is not merely with a view that allows factive rational support, à la epistemological disjunctivism, but that does not in addition offer any further anti-skeptical diagnostic story. Rather, the contrast is with such an alternative view when allied to its own anti-skeptical diagnostic story, which is exactly how we have developed the epistemological disjunctivist anti-skeptical proposal (at least as regards the underdeterminationRK-based radical skeptical paradox).
I contend that the epistemological disjunctivist diagnosis of where (underdeterminationRK-based) radical skepticism goes awry is much to be preferred, for at least two reasons. First, unlike rational support contextualism it offers us a straightforward and compelling undercutting response to the problem of (underdeterminationRK-based) radical skepticism. And, second, while on both views the anti-skeptical weight in this regard is carried by a claim about how our everyday beliefs can enjoy factive rational support, epistemological disjunctivism is not also burdened with a problematic and concessive appeal to contexts of epistemic appraisal (and in particular the idea of a skeptical context of epistemic appraisal).
When it comes to underdeterminationRK-based radical skepticism, epistemological disjunctivism is thus much more plausible, all things considered, than rational support contextualism. That said, as we will see below, once we consider not just underdeterminationRK-based radical skepticism but also closureRK-based radical skepticism, then the dialectical situation shifts somewhat toward views like rational support contextualism (and attributer contextualism too, for that matter). This is because epistemological disjunctivism struggles to handle this form of radical skepticism, but rational support contextualism fares better in this respect. As we will further see in chapter 7, however, the solution to this is not to abandon epistemological disjunctivism in favor of rational support contextualism, but rather to recognize the role that epistemological disjunctivism needs to play within a wider anti-skeptical strategy.6
2. Anti-skeptical Contrasts (II): Contrastivism
A second kind of proposal that shares some features with rational support contextualism, and that can also be usefully compared with epistemological disjunctivism, is the contrastivist response to radical skepticism, as most prominently defended by Jonathan Schaffer (2004; 2005; 2007; 2008).7 Like rational support contextualism, contrastivism is also often described as a form of attributer contextualism, although this description is in this case controversial.8
According to contrastivism, knowledge is to be understood not in the usual way as a binary relation between a subject and a proposition, but rather as a ternary relation between a subject, a proposition, and a contrast proposition. On this view, there is no such thing as knowledge that p simpliciter. Instead, knowledge is always to be understood as knowledge that p rather than q. Here is Schaffer:
It is widely assumed that knowledge is a binary relation of the form Ksp. The project of understanding knowledge then becomes the project of completing the schema “s knows that p iff.…” The traditional epistemologist supposes that there is one Ksp relation. [… The contrastivist says that there is none—the assumption that knowledge is a binary relation is an error due to the seductive pull of the surface grammar of a special form of utterance.
… Contrastivism is the view that knowledge is a ternary relation of the form Kspq, where q is a contrast proposition. (Schaffer 2004, 76–77)
We can get a better handle on what Schaffer has in mind in this regard by seeing the view in action with regard to the problem of radical skepticism. Here is what Schaffer writes in this regard, where for the “dogmatist” we can plug in most standard forms of (neo-Moorean) anti-skepticism:
Does G. E. Moore know that he has hands? Yes, says the dogmatist: Moore’s hands are right before his eyes. No, says the skeptic: for all Moore knows he could be a brain-in-a-vat. Yes and no, says the contrastivist: yes, Moore knows that he has hands rather than stumps; but no, Moore does not know that he has hands rather than vat-images of hands.
The dogmatist and the skeptic suppose that knowledge is a binary, categorical relation: s knows that p. The contrastivist says that knowledge is a ternary, contrastive relation: s knows that p rather than q. (Schaffer 2005, 235)
In particular, contrastivism claims that knowing that p rather than q is a matter of being able to discriminate between p and q.9 In this way, the contrastivist can account for the pull of radical skepticism without thereby conceding that we lack everyday knowledge. That is, our everyday knowledge that we have hands is perfectly fine, in that this is, presumably, knowledge that we have hands in contrast to mundane alternatives that are easy to discriminate between. In contrast, the radical skeptic is right that we don’t have knowledge that we have hands as opposed to radical skeptical alternatives, such as the BIV hypothesis, since in this case we lack the relevant discriminative capacities.
Note that the claim that we can conceive of our knowledge along contrastive lines in this way is not itself controversial. In particular, we need not dispute that it is true of some agents that they (contrastively) know that they have hands rather than some contrasts (e.g., stumps) but do not (contrastively) know that they have hands rather than some other contrasts (e.g., vat-hands). What makes contrastivism a controversial thesis that is in tension with standard ways of thinking about knowledge is the further claim that knowledge is only to be understood contrastively. Why should we opt for such a claim?10
To begin with, let’s set the problem to radical skepticism to one side and consider the merits of the view more generally. One advantage to contrastivism in this regard is that it seems to offer a very natural way of explaining perceptual knowledge. Think again of the “zebra” case that we encountered in chapter 5 (§2), where we have an agent faced with what looks like a zebra in a clearly marked zebra enclosure. Does our agent know that this creature is a zebra? Normally we would want to answer positively to this question, but matters become more complicated once we introduce the error possibility that the creature is a cleverly disguised mule. After all, by hypothesis, this agent cannot perceptually discriminate between zebras and cleverly disguised mules. Everyone can thus agree on at least this much: this agent can perceptually discriminate this zebra from lots of other nonzebra items (such as elephants), but she cannot perceptually discriminate zebras from cleverly disguised mules. There is thus an uncontentious sense in which our agent perceptually knows that what she is looking at is a zebra rather than, say, an elephant, but that she doesn’t perceptually know that what she is looking at is a zebra rather than a cleverly disguised mule.
So far so good for the contrastivist, though as yet there isn’t anything particularly controversial on the table. To make this claim controversial, we would need to add that there is nothing that our agent knows simpliciter in this case. What would prompt us to adopt the stronger claim?
Well, one clear motivation would be if we thought that there was a fundamental tension between our intuition that this agent would ordinarily know that what she is looking at is a zebra in these conditions, and the claim that she is unable to know that what she’s looking at is a zebra once the cleverly disguised mule error possibility is introduced. As we noted in chapter 5, once the closure principle (in some form) is in play, then it is very easy to set up a tension between these two claims, in that once we grant that our agent knows that what she is looking at is a zebra, then it’s just a short step to the further claim that she also knows that what she is looking at is not a cleverly disguised mule. But given that our agent cannot perceptually discriminate between zebras and cleverly disguised mules, this latter claim is questionable, and hence it calls into question whether we should attribute to her knowledge that what she is looking at is a zebra.
Adding the further contrastivist claim that there is no such thing as knowledge simpliciter—for example, knowledge simpliciter that the creature before one is a zebra—offers us a way of getting off this particular philosophical seesaw, albeit a highly revisionist solution. As we noted in chapter 5 (§§2 & 4) in our discussion of favoring versus discriminating epistemic support, however, there is a more straightforward way of dealing with the problem posed by the zebra case.
The crux of the matter is that we can treat the agent’s knowledge in this case as requiring different kinds of epistemic support depending on the error possibilities at issue. In normal conditions, our agent can know simpliciter that what she sees is a zebra in virtue of being able to perceptually discriminate between zebras and other items that could plausibly be in the neighborhood (e.g., elephants, etc.). Unless cleverly disguised mules are the kinds of thing that could plausibly be in the neighborhood, then our agent doesn’t need to be able to perceptually discriminate between zebras and cleverly disguised mules in order to know that what she see is a zebra.
Once the cleverly disguised mule error possibility is introduced, however, and our agent recognizes its incompatibility with what she currently believes, then it is incumbent upon her to rationally exclude this scenario. But that doesn’t require the ability to perceptually discriminate zebras from cleverly disguised mules. Instead, it requires only that our agent can draw on relevant background evidence to the extent that she has a better rational basis (i.e., favoring grounds) for believing that what she sees is a zebra rather than a cleverly disguised mule. (Alternatively, as we also saw in chapter 5, for the epistemological disjunctivist it might simply involve the citing of a relevant factive reason, at least insofar as the error possibility has not been rationally motivated.) In this way, our agent will—ordinarily anyway—be able to know both that what she is looking at is a zebra and that it is not a cleverly disguised mule. (And insofar as she is unable to know the latter proposition, then she will lose her knowledge of the former proposition.) Moreover, notice that we can capture this claim without having to adopt any form of epistemological revisionism, since there is no appeal here to contexts of epistemic appraisal (whether along attributer contextualist or rational support contextualist lines) or a retreat to a contrastivist construal of the knowledge relation. Instead all we are appealing to is the familiar idea that being aware of incompatible error possibilities can make rational demands on one—albeit, as we’ve just seen, rational demands that are easy to meet. Finally, recall that this point does not trade on epistemological disjunctivism or any other specific epistemological proposal, but is rather an entirely general thesis about the nature of the epistemic support available to agents in cases like this.
Once we recognize that there is in fact nothing particularly problematic or mysterious about the agent’s epistemic situation in the zebra case, what motivation is left for the revisionism advanced by contrastivism? Of course, we have bracketed the problem of radical skepticism here, and so that could be brought back into the frame as a motivating consideration. But notice that if the case for contrastivism as an epistemologically revisionist proposal rests squarely on the kind of anti-skepticism it offers, then that severely diminishes its dialectical effectiveness as an overriding anti-skeptical strategy. After all, the litmus test for anti-skeptical strategies of this kind is that they can offer compelling independent motivations for the epistemological revisionism that is being proposed in response to the problem of radical skepticism. If the motivation for this revisionism boils down to the need to deal with the problem of radical skepticism, then this isn’t going to be very persuasive.
Even setting this dialectical issue to one side, there is also the further point that we now have a proposal on the table—epistemological disjunctivism—according to which we can make sense of how agents can know everyday propositions over radical skeptical alternatives. This proposal is cast in terms of the possession of (factive) favoring rational support rather than discriminative capacities, but nonetheless this proposal offers us a way of accounting for our knowledge in this regard without resort to the kind of epistemological revisionism advanced by contrastivism. Moreover, recall that this is meant to be an undercutting response to the problem of radical skepticism (in its underdeterminationRK-based form at any rate), in that on this view the radical skeptical paradox is not bona fide, but instead trades on dubious theoretical claims that are masquerading as common sense. Epistemological disjunctivism thus has significant dialectical advantages over contrastivism as an anti-skeptical proposal.
Of course, we haven’t yet spelled out how epistemological disjunctivism deals with the closureRK-based radical skeptical problem, so there is potential for the scales to turn back more in contrastivism’s favor in this regard. We will return to compare and contrast epistemological disjunctivism and contrastivism with regard to closureRK-based radical skepticism below.11
3. Anti-skeptical Contrasts (III): Dogmatism
We now come to a third kind of anti-skeptical proposal that has been influential in the literature. This is known as dogmatism, and its foremost exponent is Jim Pryor (2000).12 Here is Pryor’s statement of this view:
The dogmatist about perceptual justification says that when it perceptually seems to you as if p is the case, you have a kind of justification for believing p that does not presuppose or rest on your justification for anything else, which could be cited in an argument … for p. To have this justification for believing p, you need only have an experience that represents p as being the case. No further awareness or reflection or background beliefs are required. (Pryor 2000, 519)
Thus far, the view may seem relatively unobjectionable, in that it does seem fairly plausible to suppose that its seeming to one as if p can give one reason to perceptually believe that p, and thus support a justified perceptual belief that p.
Where the view becomes controversial is with regard to how this basic thought is developed, particularly with regard to radical skepticism. The dogmatist maintains that the immediate justification for perceptually believing that p that arises out of it seeming to one as if p can suffice for perceptual knowledge. In particular, it provides defeasible support for perceptual knowledge that p, such that so long as there are not suitable defeaters present, then one perceptually knows that p. Furthermore, it is maintained that from this perceptual knowledge a subject can undertake closure-style inferences and thereby come to know the denials of radical skeptical hypotheses, such as the BIV hypothesis. Indeed, although dogmatists don’t express the point in these terms, they would presumably allow that the corresponding inferences involving the closureRK principle are fine too. Hence, subjects can in this way come to have rationally grounded knowledge of the denials of radical skeptical hypotheses by undertaking competent deductions from their knowledge of everyday propositions, where this latter knowledge is rationally grounded in the relevant perceptual seemings claim.
We thus get a particularly robust form of neo-Mooreanism. Unlike the externalist neo-Mooreanism that we have previously encountered (e.g., in chapter 1), this proposal maintains not just that we can have knowledge of the denials of radical skeptical hypotheses, but also that we can have rationally grounded knowledge of the denials of radical skeptical hypotheses. Moreover, in contrast to the entitlement reading of hinge propositions that we examined in chapter 3, the claim is precisely that the rational support available for our beliefs in the denials of radical skeptical hypotheses can be dependent upon the defeasible rational support that we have for our everyday beliefs along with the relevant competent deduction. Unlike the entitlement reading, then, there is thus no need to deny the closureRK principle.
Consider first the suggestion that the immediate (defeasible) justification one gains for one’s perceptual belief that p from its seeming to one as if p can generate rationally grounded knowledge that one is not the victim of a radical skeptical hypothesis. At first blush, this can appear like little more than epistemic magic. We can bring this worry into sharper relief by asking how, specifically, the dogmatist would respond to the underdeterminationRK-based radical skeptical paradox. In particular, here is the crucial question: would the dogmatist concede that we lack rational support for our everyday perceptual beliefs that favors those beliefs over radical skeptical alternatives, in line with (SU1)?
It seems that the dogmatist must answer this question in the negative. Consider the pairing of an everyday perceptual belief and a radical skeptical hypothesis that we used in chapter 1: that one is presently seated at one’s desk as opposed to being the victim of the BIV radical skeptical hypothesis. According to the dogmatist, both one’s belief that one is presently seated at one’s desk and one’s belief that one is not the victim of the BIV radical skeptical hypothesis could amount to rationally grounded knowledge. But if that’s right, then it surely follows that one has a better rational basis for believing that one is presently seated at one’s desk than for believing that one is a BIV. The dogmatist is thus committed to rejecting (SU1).
Here is the crux. Whereas the epistemological disjunctivist has a sophisticated story to tell both how (SU1) could be false, and also why we might have thought that it was true, the dogmatist position, true to its name, has nothing so substantial to offer. Let us state the matter baldly: why should its merely seeming to one as if p give one better rational support for believing that p than for believing that one is the victim of a known to be incompatible radical skeptical hypothesis? Indeed, without the epistemological disjunctivist’s account of why the new evil demon intuition, and the insularity of reasons thesis that underlies it, is false, it is hard to see how the dogmatist can make any plausible case for why agents can possess this favoring rational support.
Note that the complaint I am raising here is not that the introduction of radical skeptical hypotheses defeats the defeasible rational support for believing that p, which the dogmatist claims accrues in virtue of it seeming to one as if p. In fact, I think we should grant to the dogmatist—in line with what we argued in chapter 5 (§4) regarding rationally unmotivated error possibilities—that the mere raising of an error possibility does not suffice to introduce a defeater. So the point is not that the dogmatist’s would-be rational support for rationally grounded knowledge that p is defeated by the introduction of the radical skeptical hypothesis.13 The claim is rather that the dogmatist is committed to treating the subject’s rational basis for believing that p as favoring this belief over radical skeptical alternatives, and yet the subject’s rational basis for believing that p according to the dogmatist simply doesn’t seem remotely robust enough to fulfill this requirement.
Epistemological disjunctivism, in contrast, doesn’t face this problem. This is because this view has a story to tell about how the rational support for our everyday beliefs can be factive in nature, such that it does favor those everyday beliefs over radical skeptical alternatives. This, coupled with the diagnostic story the epistemological disjunctivist offers regarding why we should reject (SU1), puts epistemological disjunctivism into a much stronger dialectical position with regard to radical skepticism than dogmatism.
We will return below to the comparison of epistemological disjunctivism and dogmatism qua anti-skeptical strategies, this time with regard to the closureRK-based radical skeptical paradox. As we will see, although epistemological disjunctivism has its problems in this regard, dogmatism is at least no better off on this front, and arguably offers an anti-skeptical strategy that is even less palatable.14
4. A Weakness in Epistemological Disjunctivism
I do not doubt that many will find the epistemological disjunctivist anti-skeptical proposal that we set out in chapter 5 unsatisfying. In particular, the idea that it is legitimate to regard factive favoring reasons as adequate responses to the presentation of radical skeptical hypotheses will surely strike many as dubious. As we will see, I have some sympathy for this way of reacting to the epistemological disjunctivist treatment of radical skepticism, since I agree that the account is problematic in a fundamental respect. But we need to carefully identify what is the source of the disquiet in this case, since some of the reasons why we might find this response to radical skepticism dubious are themselves suspect on closer inspection.
For example, we might find the appeal to factive favoring reasons in the context of radical skepticism problematic because we imagine that citing factive favoring grounds in this way is tantamount to (falsely) representing oneself as being able to discriminate between the scenario in which one finds oneself and the target skeptical scenario. We have already noted, however—see chapter 5, §4—that even when presented with rationally unmotivated error possibilities, one should still be wary about citing factive favoring rational support in an unqualified fashion, since it is apt to generate misleading conversational implicatures. In this case, for example, it would usually be more appropriate to qualify one’s assertion to make clear that in putting forward this factive favoring rational support one is not thereby representing oneself as being able to discriminate between current conditions and radical skeptical alternatives.
But it is not clear to me that we should in any case be thinking of the debate with the radical skeptic in terms of assertion and counterassertion in the first place. It is natural to characterize what is taking place in the “zebra” case in terms of some sort of conversational exchange between our agent and the (local) skeptic, but that is because we are implicitly supposing that the error possibility raised is rationally motivated. Once we imagine that it is completely lacking in rational motivation, then the conversational exchange becomes at best somewhat artificial.
This is even more so when it comes to the problem of radical skepticism. For as we have emphasized at a number of junctures, radical skepticism is not best thought of as a position—indeed, so construed it isn’t a particularly challenging philosophical problem at all—but rather as a putative paradox, where this is a problem that arises out of our own fundamental epistemic commitments. Accordingly, there is something very dubious about characterizing the debate with the skeptic in terms of our explicitly offering counterassertions in response to the skeptic’s assertions. The dispute with the “skeptic” is in fact a quarrel that is completely internal to our own conceptual realm, and hence it is an argument with ourselves. While we can represent this debate in terms of a clash with an imagined adversary, we should bear in mind that strictly speaking there isn’t anyone that we are arguing with. The issue of the conversational propriety, or otherwise, of our anti-skeptical assertions is thus to a large extent completely by the by.
But if the awkwardness of the epistemological disjunctivist response to radical skepticism does not lie in this area, then where does it lie? I think we can see the source of the problem once we consider how epistemological disjunctivism squares up to the closureRK-based formulation of radical skepticism. Recall that we formulated this version of the skeptical paradox in chapter 1 in terms of the following collectively inconsistent triad of claims:
THE CLOSURERK-BASED RADICAL SKEPTICAL PARADOX
(SC1) One is unable to have rationally grounded knowledge of the denials of radical skeptical hypotheses.
(SC2) The closureRK principle.
(SC3) One has widespread rationally grounded everyday knowledge.15
The closureRK principle at issue in (SC2) was in turn formulated as follows:
THE CLOSURERK PRINCIPLE
If S has rationally grounded knowledge that p, and S competently deduces from p that q, thereby forming a belief that q on this basis while retaining her rationally grounded knowledge that p, then S has rationally grounded knowledge that q.
We have seen how epistemological disjunctivism responds to underdeterminationRK-based radical skepticism, but how does it deal with the closureRK-based formulation of the problem?
Epistemological disjunctivism will obviously endorse (SC3), so the question is which of (SC1) and (SC2) do they reject? Let’s start with (SC1). Could epistemological disjunctivism reject this claim? Since epistemological disjunctivism holds that one can have perceptual knowledge that enjoys factive rational support—that is, rational support that entails the falsity of a radical skeptical hypothesis, such as the BIV hypothesis—it follows (contra the first premise of the underdeterminationRK-based skeptical paradox, (SU1)) that one can have better rational support for, say, one’s belief that one is currently seated at one’s desk than for the hypothesis that one is a BIV. As we noted in chapter 2, however, while this claim at least suggests the rejection of (SC1), in that it implies that an agent can know the denials of radical skeptical hypotheses, it does not entail it. This is because one can in principle have better (and even factive) reasons for believing a proposition over an alternative scenario, and know what one believes, without thereby knowing that the alternative scenario does not obtain. The epistemological disjunctivists’ rejection of (SU1) therefore does not commit them, by itself anyway, to rejecting (SC1). There is thus at least the logical space available to them on this score to maintain that agents are unable to know the denials of skeptical hypotheses as (SC1) alleges, even though they do possess rational support for their everyday beliefs that favors those beliefs over skeptical alternatives.
Although accepting (SC1) is at least a possibility for epistemological disjunctivism, it is clearly not a comfortable option to pursue. For this would on the face of it mean rejecting (SC2), and thus the closureRK principle. We have seen, however, that the case for this principle appears watertight. How could a competent deduction from one’s rationally grounded knowledge, which is by its nature an exemplary rational process, undermine the rational standing of one’s beliefs thereby deduced, such that the inferred beliefs do not amount to rationally grounded knowledge? The case for (SC2) thus looks irresistible. This brings us back to what the epistemological disjunctivist wants to say about (SC1). For while it might not follow from the epistemological disjunctivists’ rejection of (SU1) that (SC1) must be denied, unless they wish to reject (SC2), and thus the closureRK principle, then it seems that they are committed to opposing (SC1) after all.
So, given the choice between rejecting the closureRK principle, (SC2), and denying (SC1), the course of least resistance as far as the epistemological disjunctivist is concerned is to opt for the latter. Indeed, faced with these two options, the epistemological disjunctivist would be wise to embrace this option and argue that the ability to reject (SC1) is a considerable advantage of the view, in virtue of its anti-skeptical ramifications.16 This is, however, much easier said than done.
The reason for this is that the epistemological disjunctivist’s rejection of (SC1) seems far too strong. As we noted in chapter 1, we have a deep-seated intuition that (SC1) is true. In particular, we intuitively have no rational basis at all for regarding radical skeptical hypotheses as false. But consider now the rational basis on which the epistemological disjunctivist rejects (SC1). This includes a factive rational basis for one’s everyday beliefs that entails the falsity of radical skeptical hypotheses, combined with the exemplary rational process of competent deduction at issue in the closureRK principle. The upshot is surely rationally grounded knowledge that one is not the victim of a skeptical hypothesis—that one is not a BIV, say—which is supported by factive rational grounds. Far from lacking any rational basis for believing that we are not victims of radical skeptical hypotheses, according to epistemological disjunctivism, so conceived, we in fact have a decisive rational basis for these anti-skeptical beliefs. The worry, of course, is that this anti-skeptical conclusion seems far too robust.
There is nothing to stop the epistemological disjunctivist simply embracing this point as a mere consequence of their view, rather than a count against it. But it does at least suggest that we should step back and see if there are not other options available to us here. In particular, are there not ways of retaining the core idea behind epistemological disjunctivism—namely, that there can be perceptual knowledge supported by factive rational support, rational support that favors the proposition believed over even skeptical alternatives—which does not require the epistemological disjunctivist to in addition claim that we can know the denials of radical skeptical hypotheses? As we will see in the next chapter, the logical space we are seeking here is indeed available. In order to occupy it, however, we will need to combine epistemological disjunctivism with the Wittgensteinian response to radical skepticism that we explored in part 2.
5. Epistemological Disjunctivism and Its Competitors
At this juncture it will be worthwhile to compare epistemological disjunctivism as an anti-skeptical proposal with the three anti-skeptical strategies we looked at earlier: rational support contextualism, contrastivism, and dogmatism. We noted there that epistemological disjunctivism generally fares much better than competing anti-skeptical approaches when it comes to the underdeterminationRK-based radical skeptical paradox. Matters are not so straightforward, however, once we extend our anti-skeptical remit to cover the closureRK-based radical skeptical paradox too. The reason for this is that most other kinds of anti-skepticism aim to avoid precisely the kind of epistemic immodesty that epistemological disjunctivism seems committed to.
This is certainly true of rational support contextualism and contrastivism, both of which avoid at least an all-out commitment to our having rationally grounded knowledge of the denials of radical skeptical hypotheses (much less knowledge of these propositions that is rationally grounded in factive reasons). According to rational support contextualism, for example, in undertaking a competent deduction from one’s rationally grounded everyday knowledge to belief in the denial of a radical skeptical hypothesis, the subject must inevitably shift to a skeptical context of epistemic evaluation. Relative to this context of epistemic evaluation, however, it is no longer the case that one has rationally grounded knowledge of the everyday proposition (because one’s rational support shrinks to the merely nonfactive). In this way rational support contextualism, like its sister view attributer contextualism, can maintain that there is no context of epistemic evaluation relative to which an agent counts as having rationally grounded knowledge of the denials of radical skeptical hypotheses. Moreover, this claim is consistent not only with the closureRK principle but also with agents generally possessing rationally grounded knowledge of everyday propositions relative to nonskeptical contexts of epistemic appraisal. Insofar as we grant that it is a weakness in epistemological disjunctivism that it is committed to the epistemic immodesty of contending that we can have rationally grounded knowledge of the denials of radical skeptical hypotheses, then in this respect at least rational support contextualism has a dialectical advantage over epistemological disjunctivism as an anti-skeptical strategy.
Contrastivism also avoids the claim that agents can have rationally grounded knowledge of the denials of radical skeptical hypotheses. Indeed, since such knowledge is by its nature binary in form, contrastivism generates this result simply in virtue of its rejection of knowledge as a binary, as opposed to a ternary, relation. More specifically, contrastivism contends that while agents can know everyday propositions in contrast to nonskeptical contrast propositions, they cannot know these propositions in contrast to skeptical contrast propositions (i.e., propositions that concern radical skeptical hypotheses).
The challenge facing contrastivism is to explain how it can account for the intuitions driving the closureRK principle. Since this principle concerns a noncontrastivist conception of knowledge, the contrastivist must reject it. Even so, it will be desirable for the view to put in its place a suitably “contrastivized” version of this principle and show that this can satisfy the intuitions that motivated the closureRK principle. This is exactly what contrastivists have attempted to do. While I do not find what the contrastivists have proposed in this regard compelling, it would take me too far afield to argue for this claim here.17
Instead, let me simply make a concession to the contrastivist similar to the one I made to the rational support contextualist, which is that at least as regards closureRK-based radical skepticism contrastivism has dialectical advantages over epistemological disjunctivism. More precisely: to the extent that contrastivism can retain the motivations for the closureRK principle while avoiding the epistemic immodesty of epistemological disjunctivism, then it has significant dialectical advantages over epistemological disjunctivism when it comes to dealing with the closureRK-based radical skeptical problem.18
That leaves dogmatism. Fortunately for epistemological disjunctivism, this is one anti-skeptical proposal that doesn’t fare better when it comes to closureRK-based radical skepticism. For one thing, dogmatists are committed to essentially the same kind of epistemic immodesty that epistemological disjunctivists are committed, in that they allow that agents can have rationally grounded knowledge of the denials of radical skeptical hypotheses. This commitment on the dogmatism’s part is arguably both more and less problematic than the parallel commitment made by the epistemological disjunctivist, and for related reasons.
It is arguably less problematic since at least the dogmatist isn’t claiming, with the epistemological disjunctivist, that one has a factive rational basis for excluding radical skeptical hypotheses. But it is also arguably more problematic since without such a factive rational basis it is even harder to stomach the idea that one can have rationally grounded knowledge of such propositions. In particular, how can the kind of nonfactive defeasible rational basis that the dogmatist has in mind suffice for rationally grounded knowledge of the denials of radical skeptical hypotheses? I think we can thus reasonably conclude that dogmatism is at least no better off than epistemological disjunctivism when it comes to closureRK-based radical skepticism, and it may well be significantly worse off.
We will return to consider these three competitor anti-skeptical proposals again briefly in chapter 7. While we have just noted that at least two of these competitor proposals have advantages over epistemological disjunctivism, we will see in the next chapter that these advantages disappear once we recast epistemological disjunctivism as part of a wider anti-skeptical strategy that incorporates the Wittgensteinian insights that we gleaned in part 2.
In this chapter we have seen that the Achilles’ heel of epistemological disjunctivism as an anti-skeptical strategy lies in the response that it offers to the closureRK-based radical skeptical paradox. In particular, whereas the epistemological disjunctivist’s claim that one can possess a rational grounding for one’s everyday beliefs that favors those beliefs over radical skeptical alternatives is well motivated, such that epistemological disjunctivism can evade the problem posed by underdeterminationRK-based radical skepticism, a stronger epistemological thesis is required to deal with closureRK-based radical skepticism, and this thesis isn’t well motivated. In particular, what is required is the idea that one can have rationally grounded knowledge of the denials of radical skeptical hypotheses, and such a claim appears epistemically immodest in the extreme.
Is there a way for the epistemological disjunctivist to demur from advancing this stronger thesis while still being in a position to offer a response to closureRK-based radical skepticism? As we will see, there is a positive answer available for this question. In order for this positive answer to be available, however, we need to understand how epistemological disjunctivism can be combined with the Wittgensteinian account of the structure of reasons that we encountered in part 2.