DEFENDING KLEIN ON CLOSURE AND SKEPTICISM

Transcription

1 E. J. COFFMAN DEFENDING KLEIN ON CLOSURE AND SKEPTICISM ABSTRACT. In this paper, I consider some issues involving a certain closure principle for Structural Justification, a relation between a cognitive subject and a proposition that s expressed by locutions like S has a source of justification for p and p is justifiable for S. I begin by summarizing recent work by Peter Klein that advances the thesis that the indicated closure principle is plausible but lacks Skeptical utility. I then assess objections to Klein s thesis based on work by Robert Audi and Anthony Brueckner. One finding is that the typical statement of the relevant closure principle can express a number of different closure principles, and that recognizing this helps to resolve certain disputes. 1. INTRODUCTION In this paper, I consider some issues involving a certain closure principle for Structural Justification. 1 Structural Justification is a relation that obtains between a cognitive subject and a proposition. It is expressed by locutions like S has a source of justification for p and p is justifiable for S. An important logical feature of this relation is that S s bearing it to p does not entail that S believes p: it s possible that S fail to believe a proposition that s justifiable for her (or, for which she has a source of justification). I begin by summarizing recent work by Peter Klein that advances the thesis that the relevant closure principle for Structural Justification is plausible but lacks Skeptical utility. 2 I then assess objections to Klein s thesis based on work by Robert Audi 3 and Anthony Brueckner. 4 One finding is that the typical statement of the relevant closure principle can express a number of different closure principles, and that recognizing this helps to resolve certain disputes. 2. KLEIN ON CLOSURE AND SKEPTICISM The line of thought I ll attempt to defend in what follows goes like this. 5 Consider the following closure principle for Structural Justification: Synthese (2006) 151: Springer 2006 DOI /s x

2 258 E. J. COFFMAN Closure: Where p and q are graspable propositions, if (i) S has a source of justification for p and (ii) it s obvious to S that p entails q, then S has a source of justification for q. 6 Two questions arise straightaway. The first is whether Closure is plausible. The answer is a qualified Yes. While Closure enjoys some initial intuitive appeal, there are also a number of putative counterexamples to it. Chief among these alleged counterexamples are Dretske-style Cases. 7 Here is an example (call it the Zoo Case ). On a trip to the zoo, Smith peers into a pen marked Zebras that contains some animals that look like zebras. Smith s current visual experience is itself a justification for the proposition that the animals in the pen are zebras. Plausibly, though, that visual experience isn t a justification for certain propositions obviously entailed by the lately mentioned one e.g., the proposition that the animals in the pen aren t cleverly disguised mules. Apparently, then, while Smith has a source of justification for the proposition that the animals in the pen are zebras, he lacks a source of justification for certain propositions obviously entailed by that proposition. In the light of such cases, Closure is at best dubious. Initially, Dretske-style Cases seem to cast doubt on Closure. Closure s proponents, however, can mount the following defense of their principle. First, they can argue that Closure is entailed by Claim: If (i) S has a source of justification for p and (ii) it s obvious to S that p entails q, then S has reflective access to good evidence for q. 8 Claim is tenable. Suppose that S satisfies Claim s antecedent. Clause (i) s satisfaction entails that S might, by reflecting on items already in her epistemic situation, come to hold a justified belief that p. That is, S s having a source of justification for p suffices for S s having reflective access to good evidence for p, evidence that would render justified a belief in p were S to so believe on its basis. Clause (ii) s satisfaction, on the other hand, presumably entails that S already holds a justified belief that p entails q. Given that clauses (i) and (ii) are satisfied, then, S at least has generative reflective access to both a justified belief that p and a justified belief that p entails q. 9 Finally, simply notice that such beliefs together constitute good evidence for q. Now, to see that Claim entails Closure, suppose that Claim is true and that S satisfies its antecedent (and so, satisfies Closure s antecedent, since those antecedents are the same). It follows, by

3 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM 259 Claim, that S has reflective access to good evidence for q. But S s having reflective access to good evidence for q suffices for S s having a source of justification for q. Claim thus entails Closure. Finally, proponents of Closure can reasonably maintain that Claim is more plausible than the thesis that a subject in a Dretskestyle Case lacks a source of justification for the relevant entailed proposition. This seems to be a successful defense of Closure against objections of this kind. At any rate, it seems to be the best possible defense of Closure against those objections. The second question is whether Closure can be used to construct a cogent Skeptical argument. Ultimately, the answer is No. This is because any Skeptical argument that s underwritten by Closure is, in the final analysis, question-begging. Let me explain. Consider the following representative Closure-based Skeptical argument (call it Argument ): Premise 1. S has a source of justification for an ordinary empirical proposition only if S has a source of justification for the proposition that he s not in a Skeptical scenario in which all ordinary empirical propositions are false. Premise 2. S does not have a source of justification for the proposition that he s not in a Skeptical scenario of the relevant kind. Conclusion. S does not have a source of justification for any ordinary empirical propositions. Neither of Argument s premises compels rational assent: each is such that one may reasonably withhold on it. The next move for Argument s proponent to make, then, is to present subarguments for those premises. Closure figures in the typical subargument for Premise 1. Given the maneuvers that Closure s proponents must execute when confronted with Dretske-style Cases, Premise 1 must ultimately be underwritten by Claim. This means that any subargument for Premise 2 must ultimately be presented in Claim s light. And that means that Premise 2 must ultimately be presented in light of the defenses of Claim and the thesis that Claim entails Closure laid out above. Those defenses, remember, commit one to regarding S s having reflective access to good evidence for a given proposition as both necessary and sufficient for S s having a source of justification for that proposition. The typical strategy for shoring up Premise 2 is to examine S s epistemic situation in an effort to show that none of the items in that situation is itself good evidence or a justification for the proposition that no relevant Skeptical hypothesis obtains. Now, prior to

4 260 E. J. COFFMAN officially defending Claim and the thesis that Claim entails Closure, a proponent of Argument ( Proponent ) is free to hold that having a source of justification for p requires having something that is itself a justification for p. Prior to officially defending those theses, then, Proponent can regard as cogent a subargument for Premise 2 that does no more than make plausible the claim that none of the items in S s epistemic situation is itself a justification for the proposition that no relevant Skeptical hypothesis obtains. In the course of officially defending Claim and the thesis that Claim entails Closure, however, Proponent affirms that a certain condition is both necessary and sufficient for having a source of justification for p viz., having reflective access to good evidence for p. Once this thesis is in play, Proponent s subargument for Premise 2 must do more than simply show that none of the items in S s epistemic situation is a justification for the proposition that no Skeptical hypothesis obtains; that subargument must also make plausible the claim that none of the items in S s epistemic situation affords S reflective access to good evidence for some ordinary empirical propositions. (The reason is this. In the present dialectical context, if the subargument for Premise 2 leaves it open whether S has reflective access to good evidence for some ordinary empirical propositions, that subargument leaves it open whether S has a source of justification for some ordinary empirical propositions. In light of Closure, this is to leave it open whether S has a source of justification for the proposition that no relevant Skeptical hypothesis obtains. Clearly, if the subargument for Premise 2 leaves it open whether S has a source of justification for the proposition that no relevant Skeptical hypothesis obtains, that subargument leaves it open whether Premise 2 is false, and so fails to shore up Premise 2.) In the present dialectical context, though, making plausible the claim that S lacks reflective access to good evidence for ordinary empirical propositions is tantamount to making plausible the very thesis Argument s main premises are supposed to establish. This is because that context includes the contention that lacking reflective access to good evidence for p suffices for lacking a source of justification for p. In sum, the most promising attempt to make the relevant kind of Skeptical argument cogent involves using its conclusion to shore up one of its premises. Any such Skeptical argument is question-begging, and thus uncogent. I ll now explain and assess some objections to this line of thought based on work by Robert Audi and Anthony Brueckner.

5 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM AUDI S CASE Robert Audi has presented a case that seems to impugn Closure. Klein s response is that Claim protects Closure against Audi s alleged counterexample. Currently, it s not clear who s right about what here. In this section, I ll try to carry this debate forward. Recall Closure: If (i) S has a source of justification for p and (ii) it s obvious to S that p entails q, then S has a source of justification for q. On one natural understanding, Audi s case is one in which S has a source of justification for p yet also holds a belief that would defeat a belief in q were S to form such a belief. 10 In such a case, it seems that S does not have a source of justification for q, that q is not justifiable for S. Here s an Audi-style Case that s suggested by Klein. 11 Jones, an amateur meteorologist, has a source of justification for believing that it ll rain tomorrow. Moreover, Jones (justifiedly) believes that all professional meteorologists are perfectly reliable weather predictors. Now, the proposition that it ll rain tomorrow obviously entails that any professional meteorologist who predicts otherwise is wrong. By Closure, then, Jones has a source of justification for the proposition that any professional meteorologist who predicts that it won t rain tomorrow is wrong. But that proposition isn t justifiable for Jones, since he holds a belief that would defeat a belief in that proposition were he to so believe. Klein s response is that reflection on Claim shows Audi s case to be incoherent. 12 Pace Audi, Jones indeed has a source of justification for the proposition that any professional meteorologist who predicts that it won t rain tomorrow is wrong. For Claim entails that Jones has reflective access to good evidence for the indicated proposition to wit, a belief that it ll rain tomorrow and a belief that the proposition that it ll rain tomorrow entails that any professional meteorologist who predicts otherwise is wrong. I think that the distinction between Prima Facie and Ultima Facie Doxastic (or, Belief) Justification sheds light on these issues. 13 We re led to make this distinction by the observation that there are two ways a belief can fail to be unqualifiedly (or, on balance ) justified. On the one hand, a belief can fail to be so justified by failing to be based on good evidence (i.e., evidence that renders the belief unqualifiedly justified other things equal ). (An example. Suppose I believe that you don t drink caffeinated beverages solely on the

6 262 E. J. COFFMAN basis of my belief that you just ordered water when many different beverages both caffeinated and non-caffeinated were available to you. The indicated evidence utterly fails to support my belief; my belief thus fails to be justified on balance.) On the other hand, a belief that s based on good evidence can fail to be unqualifiedly justified by being defeated. (An example. I (justifiedly) believe that Annie is in California. While walking on campus, I have a visual experience that constitutes good evidence for believing that Annie is on campus. I subsequently believe, on the basis of this experience, that Annie is on campus. In this case, my belief that Annie is in California defeats my belief that Annie is on campus (and, presumably, vice versa). My belief that Annie is on campus is based on good evidence, though it s not justified on balance.) Here are some terms that help us to distinguish among the lately mentioned epistemic properties. A belief that s not based on good evidence is not prima facie justified. A belief that s based on good evidence is prima facie justified. A belief that s based on good evidence but also defeated is merely prima facie justified. Finally, a belief that s both based on good evidence and undefeated is ultima facie justified. Now, on the plausible assumption that Structural Justification is to be understood partly in terms of Doxastic Justification, 14 the distinction between Prima and Ultima Facie Doxastic Justification yields a distinction between Prima and Ultima Facie Structural Justification. Roughly, to have a source of prima facie justification for p is to have reflective access to evidence that would render a belief in p prima facie justified were such a belief held on its basis. For ease of exposition, let s call such evidence a prima facie justifier for a belief in p. And to have a source of ultima facie justification for p is to have reflective access to evidence that would render a belief in p ultima facie justified were such a belief held on its basis. Again, for ease of exposition, let s call such evidence an ultima facie justifier for a belief in p. Interestingly, in the light of the distinction between Prima and Ultima Facie Structural Justification, we can see that the sentence labeled Closure above can express a number of different closure principles: each of its occurrences of justification may be read as either prima facie justification or ultima facie justification. This point raises the possibility that the thesis impugned by Audi s case differs from the thesis Klein thinks tenable. Now, if that were true, then we would have an attractive explanation of the fact that each

7 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM 263 theorist seems to be onto something in their exchange. But is this true? I think so. Let me try to argue this out. Klein is concerned to defend the closure principle that underwrites Premise 1 of Argument. To figure out exactly which principle that is, we need to figure out Premise 1 s content. Arguably, Argument s proponents should be understood to endorse the prima facie justification reading of Premise 1. This is because Argument s proponents should be understood to endorse the prima facie justification readings of Argument s second premise and conclusion. And that s because endorsing only the ultima facie justification readings of Premise 2 and Conclusion leads one to occupy a position that s less tenable than the position one occupies if one endorses the prima facie justification readings of those statements as well. Here s why. Suppose one endorses only the ultima facie justification readings of Premise 2 and Conclusion. One then leaves open a pair of baffling theses, to wit that one s having a source of prima facie justification for the proposition that one s not in a relevant Skeptical scenario nontrivially entails that one has a defeater for belief in that proposition, and that one s having a source of prima facie justification for some ordinary empirical proposition nontrivially entails that one has a defeater for belief in that proposition. A position that affirms only the ultima facie justification readings of Premise 2 and Conclusion leaves these claims open because such a position simultaneously leaves open the possibility that one have a source of prima facie justification for the indicated propositions and asserts the impossibility of one s having a source of ultima facie justification for those propositions. In short, endorsing only the ultima facie justification readings of Premise 2 and Conclusion leaves open the thesis that some contingent empirical propositions are such that there s a substantive logical connection between one s having a source of prima facie justification for them and one s having a defeater for belief in them. The latter thesis is exceedingly implausible (if not clearly incoherent): how could my now having a source of prima facie justification for (say) some ordinary empirical

8 264 E. J. COFFMAN proposition, p, nontrivially entail that I now also hold beliefs that would defeat a belief in p were I to so believe? Claims like this one should not be left open. Notice that endorsement of the prima facie justification readings of Premise 2 and Conclusion does not leave it open whether the lately mentioned baffling theses are true. Presumably, the statements that make up Argument should be understood to be necessarily true if true at all. Premise 1 is a claim about the logic of Structural Justification. And though I grant that things aren t as obvious in the case of Premise 2, that thesis looks to me to be a conceptual claim as well. Moreover, prominent theorists seem tempted to construe such claims as putative necessary truths. (For instance, in discussing the analogue of Premise 2 in an argument for the conclusion that no one knows any ordinary empirical propositions, Keith De- Rose asks How could I know such a thing? 15 This suggests that DeRose would construe Premise 2 of Argument as a putative necessary truth.) On the prima facie justification reading of Argument, then, both Premise 2 and Conclusion imply that it s impossible that S have a source of prima facie justification for propositions of a certain kind. Proponents of the prima facie justification reading of Argument are thus in a position to deny the existence of a non-trivial logical connection between one s having a source of prima facie justification for such propositions and one s having a defeater for beliefs in those propositions. Moreover, though one can evade formal commitment to the unpalatable thesis that no one has a source of prima facie justification for any ordinary empirical propositions by endorsing only the ultima facie justification readings of Premise 2 and Conclusion, those readings commit one to the only slightly more palatable thesis that most subjects don t have a source of prima facie justification for any ordinary empirical propositions. This is because the vast majority of subjects simply do not have experiences or beliefs that could function as defeaters for beliefs in ordinary empirical propositions. (Consider: most human cognizers have neither seen The Matrix nor thought about the possibility of an Evil Genius.) In sum, endorsing the prima facie justification readings of Premise 2 and Conclusion costs less than endorsing only the ultima facie justification readings of those statements. Charity thus counsels us to understand Argument s proponents as endorsing the prima facie justification readings of Premise 2 and Closure (and so, the prima facie justification reading of Premise 1).

9 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM 265 The upshot of all this is that Klein needs to defend only Closure PF : If (i) S has a source of prima facie justification for p and (ii) it s obvious to S that p entails q, then S has a source of prima facie justification for q. Now, while Audi-style Cases belie similar closure principles whose consequents involve the property of Ultima Facie Structural Justification, e.g., Closure UF : If (i) S has a source of prima facie justification for p and (ii) it s obvious to S that p entails q, then S has a source of ultima facie justification for q, they don t impugn Closure PF. In the case described above, Jones fails to have a source of justification for the proposition that any dissenting professional meteorologist is wrong because he holds a belief that would defeat a belief in that proposition were he to so believe. What defeaters defeat, though, is ultima facie justification: defeaters are irrelevant to prima facie justification. 16 Audi s case thus fails to connect with Closure PF. Let s take stock. While Audi-style Cases impugn some closure principles for Structural Justification, they seem not to impugn the principle Klein thinks tenable. On the other hand, Klein s treatment of Audi s case could be more concessive. While it s plausible that Jones has a source of prima facie justification for the proposition that any professional meteorologist who predicts rain for tomorrow is wrong, Jones doesn t have a source of ultima facie justification for that proposition. In an important (and perhaps the most natural) sense, then, Jones doesn t have a source of justification for the proposition that any dissenting professional meteorologist is wrong. Before moving on, I want to amplify one of this section s main points. I ve said that Klein needs to defend only Closure PF. It follows that Klein needs to endorse only one of various possible readings of the sentence labeled Claim above viz., one that involves only the property of Prima Facie Structural Justification: Claim PF : If (i) S has a source of prima facie justification for p and (ii) it s obvious to S that p entails q, then S has reflective access to a prima facie justifier for belief in q. 4. BRUECKNER S OBJECTIONS Brueckner lodges five objections against Klein s position. The first two objections we ll consider attack Claim. The remaining three

10 266 E. J. COFFMAN attack Klein s treatment of Argument. I shall attempt to show that none of these objections succeeds. One of Brueckner s two objections to Claim runs as follows. 17 Suppose, for reductio, that Claim is true. Suppose also that a certain table looks red to S. It follows, by Claim, that S has reflective access to good evidence for the proposition that it s false that the table is white but illuminated by red light. But this is implausible. In the light of Section 3, we can see that this objection misfires. The only credible understanding of Brueckner s charge of implausibility is that it s implausible that S s having the relevant visual experience entails that S has reflective access to an ultima facie justifier for belief in the proposition that it s false that the table is white but illuminated by red light. As we ve seen, though, Klein needs to defend only Claim PF ; and Claim PF entails only S has reflective access to a prima facie justifier for belief in the proposition that it s false that the table is white but illuminated by red light. But that s not so implausible. Here s another putative reductio of Claim. 18 Suppose, for reductio, that Claim is true. Suppose also that each of the following propositions is such that S has a source of justification for it: and Joe is a man; Prob(x owns a gun/x is a man) = 0.7; Prob(x reads F ield and Stream/x owns a gun) = 0.7. Now, Claim seems to commit Klein to Principle: If (i) S has a source of justification for p and (ii) it s obvious to S that p (significantly) probabilifies q, then S has reflective access to good evidence for q. We ve now run headlong into absurdity. Since Joe s being a man obviously probabilifies Joe s owning a gun, Principle and our supposition entail that S has reflective access to good evidence for the proposition that Joe owns a gun. Moreover, since Joe s owning a gun obviously probabilifies Joe s reading Field and Stream, Principle entails that S has reflective access to good evidence for the proposition that Joe reads Field and Stream. But S doesn t currently have reflective access to good evidence for the proposition that Joe reads Field and Stream.

11 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM 267 In response, I ll show that there s available to Klein a coherent solution to Brueckner s puzzle that leaves his position intact. Just notice that Claim is compatible with the denial of Principle, to wit Possible: Possibly, (i) S has a source of justification for p and (ii) it s obvious to S that p probabilifies q but (iii) S doesn t have reflective access to good evidence for q. So, Claim doesn t entail Principle: one s accepting Claim doesn t formally commit one to Principle. Moreover, there s no other reason to think that Klein must endorse Principle. But then Klein can coherently solve Brueckner s puzzle by retaining Claim and eschewing Principle. Incidentally, this solution strikes me as not only coherent but also plausible. This is because Possible seems plausible. Suppose both that S has a source of justification for p and knows that q s likelihood given p is (say) 0.7. Does this entail that S has reflective access to good evidence for q? One might reasonably think not. For perhaps S has the bare minimum of Structural Justification for p. In such a case, it may be that S does not have reflective access to good evidence for q. If this is right, then Possible is true, and the suggested solution is not only coherent but plausible. Two of the remaining three objections attack Klein s treatment of Argument by attacking his thesis that Closure succumbs to Dretskestyle Cases unless it s defended by appeal to Claim. If this thesis is false, then we ve no reason to think that Argument is question-begging. For if this thesis is false, then we ve no reason to think that Closure s proponents must deploy Claim to defend Closure; and if Closure s proponents needn t deploy Claim to defend Closure, then we ve no reason to think Argument s proponents must use Argument s conclusion to support its second premise. One of these objections runs as follows. 19 According to Klein, Smith s visual experience in the Zoo Case is itself a justification for the proposition that the animal in the pen is a zebra but not for the proposition that the animal in the pen isn t a cleverly disguised mule. This is an essential component of Klein s treatment of the Zoo Case: that treatment is distinctive and intriguing because it attempts to simultaneously (i) honor the intuition that Smith s visual experience is a justification for the proposition that the animal in the pen is a zebra but not for the proposition that the animal in the pen isn t a cleverly disguised mule and (ii) preserve Closure. Now, while Klein s assessment of the Zoo Case isn t obviously implausible,

12 268 E. J. COFFMAN it commits him to implausible views regarding other cases. Consider the following case, for instance. Smith now has a visual experience that s a justification for the proposition that there s a physical object. Presumably, Klein s assessment of the Zoo Case commits him to thinking that Smith s current visual experience isn t a justification for the proposition that it s false that all objects are nonphysical. Notice, however, that the proposition that there s a physical object is obviously logically equivalent to the proposition that it s false that all objects are non-physical. So, Klein s assessment of the Zoo Case commits him to the absurdity that the property of being a justification for isn t closed under obvious logical equivalence. In short, either Klein retracts his assessment of the Zoo Case or he doesn t. Either way, he can t reasonably maintain that Closure succumbs to Dretske-style Cases unless it s defended by appeal to Claim. But then we ve no reason to think that Argument can t be made cogent. By my lights, this is Brueckner s best objection. I think that it can be resisted, though. This objection depends on the claims that (i) it s absurd to reject the thesis that being a justification for is closed under obvious logical equivalence, and that (ii) Klein s treatment of Dretske-style Cases requires him to reject that thesis. For each of these assumptions, there s at least one response available to a Kleinian. First, one might regard as a live option rejecting the thesis that being a justification for is closed under obvious equivalence. Second (and more plausibly), one can point out that Klein s treatment of Dretske-style Cases is compatible with the thesis that being a justification for is closed under obvious equivalence. For that treatment is incompatible only with the stronger thesis that being a justification for is closed under obvious entailment. Strictly speaking, then, Klein s treatment of Dretske-style Cases doesn t commit him to denying that being a justification for is closed under obvious equivalence. Brueckner has failed to provide a reason for thinking that Klein s treatment of the Zoo Case commits him to embracing an absurdity. The other of the two lately mentioned objections runs as follows. 20 Klein contends that Closure is impugned by Dretske-style Cases unless it s defended by appeal to Claim. By Brueckner s lights, a proponent of Argument (again, Proponent ) may reasonably deny that Dretske-style Cases impugn Closure without thereby incurring a commitment to Claim. With respect to the Zoo Case, for example, Proponent may reasonably deny that Smith has a source of

13 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM 269 justification for the proposition that the animals in the pen are zebras, and so may reasonably maintain that the putative counterexample fails to connect with Closure. This obviates an appeal to Claim, and may well lead to Argument s being made cogent. This response is unsuccessful. Suppose Proponent attempts to defend Argument s cogency by denying that Smith s visual experience is a source of justification for the proposition that the animals in the pen are zebras. This denial itself renders Argument uncogent. If it s possible to have a source of justification for an ordinary empirical proposition, then Smith s visual experience in the Zoo Case is such a source. Denying at Argument s outset that Smith s experience is a source of justification for the relevant proposition is thus tantamount to denying at Argument s outset that a subject can have a source of justification for an ordinary empirical proposition. So, the envisaged defense of Premise 1 seems to invoke Argument s conclusion at Argument s outset. But then, far from serving to defend Argument s cogency, the envisaged response suffices to render Argument uncogent. That response is thus unsuccessful. The last of Brueckner s objections attempts to show that Klein s treatment of Argument entails the absurdity that the Modus Tollens inference pattern is fallacious. 21 This objection runs as follows. According to Klein, any Skeptical argument underwritten by Closure has the following three characteristics: It exemplifies the Modus Tollens inference pattern. It contains a premise to the effect that an arbitrarily selected subject doesn t have a source of justification for the proposition that Skeptical hypotheses of a certain kind are false. The indicated premise can be supported only by invoking the argument s conclusion. Whatever reason Klein gives for thinking that every such Skeptical argument has the third characteristic will commit him to the claim that every Modus Tollens inference has that characteristic, and so is fallacious. But that s absurd. Pace Brueckner, Klein s treatment of Argument holds no general implications for the Modus Tollens inference pattern. To see this, think carefully about the nature of Klein s criticism of Argument. It s that Argument s main premises fail to compel rational assent, and that the most promising attempt to shore them up involves deploying the very thesis those premises aim to establish. This

14 270 E. J. COFFMAN criticism plainly concerns the contents of Argument s premises, not Argument s formal structure. Klein s pessimism about the prospects for making Argument cogent is thus compatible with the thesis that lots of good arguments exemplify the Modus Tollens inference pattern. 5. CONCLUSION Audi-style cases help us to see that the sentence labeled Closure above can express a number of different closure principles, some of which are implausible. Fortunately for Klein, Audi-style cases do not impugn the closure principle he thinks tenable. As for Brueckner s objections, I believe I have shown that none clearly succeeds. So far as I can see, then, both Klein s favored closure principle and his treatment of Skeptical arguments of the relevant kind remain intact. ACKNOWLEDGEMENTS This paper was read at a meeting of the Notre Dame Graduate Student Philosophy Paper Club and a meeting of the Central Division of the American Philosophical Association (2004). Thanks to those who attended these sessions for helpful discussion of the issues involved in this paper. Special thanks to Fred Adams, Beau Branson, Todd Long, and Zach Manis. NOTES 1 For a very helpful discussion of this relation, see Audi (1991). 2 See Klein (1995). 3 See (inter alia) Chapter 6 of Audi (2003). 4 See Brueckner (2000). 5 Incidentally, there is a lot of interesting material in Klein (1995) that I won t engage here. This is mainly because I do not wish to become deeply embroiled in Kleinian exegesis in this paper. My aim is simply to describe and defend an interesting position that is reasonably attributed to Klein on the basis of that article. 6 For ease of exposition, I shall henceforth suppress the graspability restriction when stating and discussing Closure and related principles. 7 See (inter alia) Dretske (1970). For Klein s treatment, see Klein (1995, ).

15 DEFENDING KLEIN ON CLOSURE AND SKEPTICISM For Klein s discussion of (something in the neighborhood of ) this thesis, see Klein (1995, ). 9 So far as I know, the term generative reflective access was coined by Robert Audi. See Audi (1991). 10 Here, the relation expressed by the term defeat is that of Rationality (or, Justificational) Defeat, as opposed to Warrant (or, Knowledge) Defeat (the latter being the phenomenon with which Defeasibility theorists of Knowledge are primarily concerned). The terms of this relation are beliefs and experiences, on the one hand, and beliefs, on the other. That is, beliefs and experiences are (Rationality) Defeaters and beliefs are (Rationality) Defeatees. Roughly, a belief or experience B 1 defeats a belief B 2 iff one s having B 1 makes one s holding B 2 subjectively epistemically irrational. For detailed treatments of this relation, see (inter alia) Chapter 11 of Plantinga (2000), Chapter 3 of Bergmann (1997), and Pollock (1986, 37ff.). 11 Klein (1995, ). 12 See Klein (1995, ). 13 One of the best available discussions of this distinction is found in Senor (1996). 14 Plausibly, Structural Justification is to be understood partly in terms of Propositional Justification a relation between a cognitive subject and a proposition expressed by locutions like p is justified for S and S has a justification for p roughly as follows: S has a source of justification for p (or, p is justifiable for S) iff S s epistemic situation is such that she might by reflection alone come to have a justification for p, something that justifies p for her. Plausibly again, Propositional Justification is to be understood partly in terms of Doxastic Justification a property of beliefs expressed by locutions like S s belief B is justified and S s belief that p is justified roughly as follows: S has a justification for p (or, p is justified for S) iff S has evidence, E (i.e., experiences or beliefs), such that were S to believe p (solely) on E s basis that belief would be justified. (For a defense of this explication of Propositional Justification in terms of Doxastic Justification, see my The Logic of the Main Varieties of Epistemic Justification (not yet committed for publication).) 15 DeRose (1995, 1). 16 Cf. Senor (1996, 554). 17 Brueckner (2000, 143). 18 Brueckner (2000, ). 19 What follows is based on an objection developed in Brueckner (2000, 144). 20 Brueckner (2000, 149). 21 Brueckner 2000, ( ). REFERENCES Audi, Robert: 1991, Structural Justification, The Journal of Philosophical Research 16, Audi, Robert: 2003, Epistemology, Routledge, New York. Bergmann, Michael: 1997, lnternalism, Externalism, and Epistemic Defeat, University of Notre Dame Ph.D. dissertation.

16 272 E. J. COFFMAN Brueckner, Anthony: 2000, Klein on Closure and Skepticism, Philosophical Studies 98, DeRose, Keith: 1995, Solving the Skeptical Problem, The Philosophical Review 104, Dretske, Fred: 1970, Epistemic Operators, Journal of Philosophy 67, Klein, Peter: 1995, Skepticism and Closure: Why the Evil Genius Argument Fails, Philosophical Topics 23, Plantinga, Alvin: 2000, Warranted Christian Belief, Oxford University Press, New York. Pollock, John: 1986, Contemporary Theories of Knowledge, Rowman Littlefield, Totowa, NJ. Senor, Thomas: 1996, The Prima/Ultima Facie Justification Distinction in Epistemology, Philosophy and Phenomenological Research 56, Department of Philosophy University of Notre Dame Notre Dame, IN USA ecoffman@nd.edu